diff --git "a/last-checkpoint/trainer_state.json" "b/last-checkpoint/trainer_state.json"
new file mode 100644--- /dev/null
+++ "b/last-checkpoint/trainer_state.json"
@@ -0,0 +1,62050 @@
+{
+  "best_global_step": null,
+  "best_metric": null,
+  "best_model_checkpoint": null,
+  "epoch": 0.8198576636000694,
+  "eval_steps": 2951,
+  "global_step": 8856,
+  "is_hyper_param_search": false,
+  "is_local_process_zero": true,
+  "is_world_process_zero": true,
+  "log_history": [
+    {
+      "epoch": 0.00010168155877829607,
+      "grad_norm": 3.390625,
+      "learning_rate": 0.00011299435028248587,
+      "loss": 3.2031,
+      "step": 1
+    },
+    {
+      "epoch": 0.00020336311755659214,
+      "grad_norm": 3.359375,
+      "learning_rate": 0.00022598870056497175,
+      "loss": 3.0147,
+      "step": 2
+    },
+    {
+      "epoch": 0.0003050446763348882,
+      "grad_norm": 3.171875,
+      "learning_rate": 0.00033898305084745765,
+      "loss": 3.1702,
+      "step": 3
+    },
+    {
+      "epoch": 0.0004067262351131843,
+      "grad_norm": 3.421875,
+      "learning_rate": 0.0004519774011299435,
+      "loss": 3.1697,
+      "step": 4
+    },
+    {
+      "epoch": 0.0005084077938914803,
+      "grad_norm": 3.515625,
+      "learning_rate": 0.0005649717514124294,
+      "loss": 3.099,
+      "step": 5
+    },
+    {
+      "epoch": 0.0006100893526697764,
+      "grad_norm": 3.484375,
+      "learning_rate": 0.0006779661016949153,
+      "loss": 3.1276,
+      "step": 6
+    },
+    {
+      "epoch": 0.0007117709114480725,
+      "grad_norm": 3.21875,
+      "learning_rate": 0.0007909604519774012,
+      "loss": 3.1475,
+      "step": 7
+    },
+    {
+      "epoch": 0.0008134524702263686,
+      "grad_norm": 3.375,
+      "learning_rate": 0.000903954802259887,
+      "loss": 3.0928,
+      "step": 8
+    },
+    {
+      "epoch": 0.0009151340290046646,
+      "grad_norm": 3.09375,
+      "learning_rate": 0.001016949152542373,
+      "loss": 3.028,
+      "step": 9
+    },
+    {
+      "epoch": 0.0010168155877829607,
+      "grad_norm": 3.21875,
+      "learning_rate": 0.0011299435028248588,
+      "loss": 3.0019,
+      "step": 10
+    },
+    {
+      "epoch": 0.0011184971465612568,
+      "grad_norm": 3.359375,
+      "learning_rate": 0.0012429378531073447,
+      "loss": 2.999,
+      "step": 11
+    },
+    {
+      "epoch": 0.0012201787053395528,
+      "grad_norm": 3.171875,
+      "learning_rate": 0.0013559322033898306,
+      "loss": 3.0901,
+      "step": 12
+    },
+    {
+      "epoch": 0.001321860264117849,
+      "grad_norm": 3.046875,
+      "learning_rate": 0.0014689265536723165,
+      "loss": 2.9991,
+      "step": 13
+    },
+    {
+      "epoch": 0.001423541822896145,
+      "grad_norm": 3.1875,
+      "learning_rate": 0.0015819209039548024,
+      "loss": 2.9065,
+      "step": 14
+    },
+    {
+      "epoch": 0.001525223381674441,
+      "grad_norm": 2.546875,
+      "learning_rate": 0.001694915254237288,
+      "loss": 2.8828,
+      "step": 15
+    },
+    {
+      "epoch": 0.0016269049404527372,
+      "grad_norm": 3.03125,
+      "learning_rate": 0.001807909604519774,
+      "loss": 2.8672,
+      "step": 16
+    },
+    {
+      "epoch": 0.0017285864992310333,
+      "grad_norm": 3.3125,
+      "learning_rate": 0.0019209039548022599,
+      "loss": 2.8512,
+      "step": 17
+    },
+    {
+      "epoch": 0.0018302680580093292,
+      "grad_norm": 3.390625,
+      "learning_rate": 0.002033898305084746,
+      "loss": 2.8261,
+      "step": 18
+    },
+    {
+      "epoch": 0.0019319496167876254,
+      "grad_norm": 2.96875,
+      "learning_rate": 0.0021468926553672315,
+      "loss": 2.7482,
+      "step": 19
+    },
+    {
+      "epoch": 0.0020336311755659213,
+      "grad_norm": 2.328125,
+      "learning_rate": 0.0022598870056497176,
+      "loss": 2.5005,
+      "step": 20
+    },
+    {
+      "epoch": 0.0021353127343442177,
+      "grad_norm": 2.265625,
+      "learning_rate": 0.0023728813559322037,
+      "loss": 2.6218,
+      "step": 21
+    },
+    {
+      "epoch": 0.0022369942931225136,
+      "grad_norm": 1.6171875,
+      "learning_rate": 0.0024858757062146894,
+      "loss": 2.4375,
+      "step": 22
+    },
+    {
+      "epoch": 0.0023386758519008096,
+      "grad_norm": 1.109375,
+      "learning_rate": 0.002598870056497175,
+      "loss": 2.3563,
+      "step": 23
+    },
+    {
+      "epoch": 0.0024403574106791055,
+      "grad_norm": 1.09375,
+      "learning_rate": 0.002711864406779661,
+      "loss": 2.4473,
+      "step": 24
+    },
+    {
+      "epoch": 0.002542038969457402,
+      "grad_norm": 0.953125,
+      "learning_rate": 0.002824858757062147,
+      "loss": 2.4006,
+      "step": 25
+    },
+    {
+      "epoch": 0.002643720528235698,
+      "grad_norm": 0.76171875,
+      "learning_rate": 0.002937853107344633,
+      "loss": 2.4106,
+      "step": 26
+    },
+    {
+      "epoch": 0.0027454020870139938,
+      "grad_norm": 0.60546875,
+      "learning_rate": 0.003050847457627119,
+      "loss": 2.3258,
+      "step": 27
+    },
+    {
+      "epoch": 0.00284708364579229,
+      "grad_norm": 0.3671875,
+      "learning_rate": 0.003163841807909605,
+      "loss": 2.2928,
+      "step": 28
+    },
+    {
+      "epoch": 0.002948765204570586,
+      "grad_norm": 0.40625,
+      "learning_rate": 0.00327683615819209,
+      "loss": 2.3009,
+      "step": 29
+    },
+    {
+      "epoch": 0.003050446763348882,
+      "grad_norm": 0.267578125,
+      "learning_rate": 0.003389830508474576,
+      "loss": 2.2469,
+      "step": 30
+    },
+    {
+      "epoch": 0.0031521283221271784,
+      "grad_norm": 0.21484375,
+      "learning_rate": 0.0035028248587570623,
+      "loss": 2.2102,
+      "step": 31
+    },
+    {
+      "epoch": 0.0032538098809054743,
+      "grad_norm": 0.1923828125,
+      "learning_rate": 0.003615819209039548,
+      "loss": 2.2146,
+      "step": 32
+    },
+    {
+      "epoch": 0.0033554914396837702,
+      "grad_norm": 0.1787109375,
+      "learning_rate": 0.003728813559322034,
+      "loss": 2.2294,
+      "step": 33
+    },
+    {
+      "epoch": 0.0034571729984620666,
+      "grad_norm": 0.162109375,
+      "learning_rate": 0.0038418079096045198,
+      "loss": 2.2151,
+      "step": 34
+    },
+    {
+      "epoch": 0.0035588545572403625,
+      "grad_norm": 0.146484375,
+      "learning_rate": 0.003954802259887006,
+      "loss": 2.2166,
+      "step": 35
+    },
+    {
+      "epoch": 0.0036605361160186585,
+      "grad_norm": 0.1298828125,
+      "learning_rate": 0.004067796610169492,
+      "loss": 2.2415,
+      "step": 36
+    },
+    {
+      "epoch": 0.003762217674796955,
+      "grad_norm": 0.1337890625,
+      "learning_rate": 0.004180790960451978,
+      "loss": 2.1685,
+      "step": 37
+    },
+    {
+      "epoch": 0.003863899233575251,
+      "grad_norm": 0.150390625,
+      "learning_rate": 0.004293785310734463,
+      "loss": 2.1851,
+      "step": 38
+    },
+    {
+      "epoch": 0.003965580792353547,
+      "grad_norm": 0.193359375,
+      "learning_rate": 0.004406779661016949,
+      "loss": 2.165,
+      "step": 39
+    },
+    {
+      "epoch": 0.004067262351131843,
+      "grad_norm": 0.1923828125,
+      "learning_rate": 0.004519774011299435,
+      "loss": 2.177,
+      "step": 40
+    },
+    {
+      "epoch": 0.004168943909910139,
+      "grad_norm": 0.19921875,
+      "learning_rate": 0.004632768361581921,
+      "loss": 2.1863,
+      "step": 41
+    },
+    {
+      "epoch": 0.004270625468688435,
+      "grad_norm": 0.1650390625,
+      "learning_rate": 0.004745762711864407,
+      "loss": 2.1063,
+      "step": 42
+    },
+    {
+      "epoch": 0.004372307027466731,
+      "grad_norm": 0.1650390625,
+      "learning_rate": 0.004858757062146893,
+      "loss": 2.1679,
+      "step": 43
+    },
+    {
+      "epoch": 0.004473988586245027,
+      "grad_norm": 0.1904296875,
+      "learning_rate": 0.004971751412429379,
+      "loss": 2.151,
+      "step": 44
+    },
+    {
+      "epoch": 0.004575670145023323,
+      "grad_norm": 0.201171875,
+      "learning_rate": 0.005084745762711864,
+      "loss": 2.1011,
+      "step": 45
+    },
+    {
+      "epoch": 0.004677351703801619,
+      "grad_norm": 0.1826171875,
+      "learning_rate": 0.00519774011299435,
+      "loss": 2.1333,
+      "step": 46
+    },
+    {
+      "epoch": 0.0047790332625799155,
+      "grad_norm": 0.1806640625,
+      "learning_rate": 0.005310734463276836,
+      "loss": 2.0644,
+      "step": 47
+    },
+    {
+      "epoch": 0.004880714821358211,
+      "grad_norm": 0.16796875,
+      "learning_rate": 0.005423728813559322,
+      "loss": 2.1194,
+      "step": 48
+    },
+    {
+      "epoch": 0.004982396380136507,
+      "grad_norm": 0.1845703125,
+      "learning_rate": 0.005536723163841808,
+      "loss": 2.1437,
+      "step": 49
+    },
+    {
+      "epoch": 0.005084077938914804,
+      "grad_norm": 0.166015625,
+      "learning_rate": 0.005649717514124294,
+      "loss": 2.1265,
+      "step": 50
+    },
+    {
+      "epoch": 0.005185759497693099,
+      "grad_norm": 0.154296875,
+      "learning_rate": 0.00576271186440678,
+      "loss": 2.1426,
+      "step": 51
+    },
+    {
+      "epoch": 0.005287441056471396,
+      "grad_norm": 0.1650390625,
+      "learning_rate": 0.005875706214689266,
+      "loss": 2.1135,
+      "step": 52
+    },
+    {
+      "epoch": 0.005389122615249692,
+      "grad_norm": 0.19140625,
+      "learning_rate": 0.005988700564971752,
+      "loss": 2.0959,
+      "step": 53
+    },
+    {
+      "epoch": 0.0054908041740279875,
+      "grad_norm": 0.1904296875,
+      "learning_rate": 0.006101694915254238,
+      "loss": 2.1376,
+      "step": 54
+    },
+    {
+      "epoch": 0.005592485732806284,
+      "grad_norm": 0.1884765625,
+      "learning_rate": 0.0062146892655367235,
+      "loss": 2.0737,
+      "step": 55
+    },
+    {
+      "epoch": 0.00569416729158458,
+      "grad_norm": 0.1708984375,
+      "learning_rate": 0.00632768361581921,
+      "loss": 2.1113,
+      "step": 56
+    },
+    {
+      "epoch": 0.005795848850362876,
+      "grad_norm": 0.19140625,
+      "learning_rate": 0.006440677966101695,
+      "loss": 2.1405,
+      "step": 57
+    },
+    {
+      "epoch": 0.005897530409141172,
+      "grad_norm": 0.1923828125,
+      "learning_rate": 0.00655367231638418,
+      "loss": 2.1187,
+      "step": 58
+    },
+    {
+      "epoch": 0.0059992119679194685,
+      "grad_norm": 0.1953125,
+      "learning_rate": 0.006666666666666666,
+      "loss": 2.1115,
+      "step": 59
+    },
+    {
+      "epoch": 0.006100893526697764,
+      "grad_norm": 0.2001953125,
+      "learning_rate": 0.006779661016949152,
+      "loss": 2.138,
+      "step": 60
+    },
+    {
+      "epoch": 0.00620257508547606,
+      "grad_norm": 0.21484375,
+      "learning_rate": 0.0068926553672316385,
+      "loss": 2.1772,
+      "step": 61
+    },
+    {
+      "epoch": 0.006304256644254357,
+      "grad_norm": 0.22265625,
+      "learning_rate": 0.007005649717514125,
+      "loss": 2.1571,
+      "step": 62
+    },
+    {
+      "epoch": 0.006405938203032652,
+      "grad_norm": 0.2080078125,
+      "learning_rate": 0.00711864406779661,
+      "loss": 2.1095,
+      "step": 63
+    },
+    {
+      "epoch": 0.006507619761810949,
+      "grad_norm": 0.1904296875,
+      "learning_rate": 0.007231638418079096,
+      "loss": 2.0307,
+      "step": 64
+    },
+    {
+      "epoch": 0.006609301320589245,
+      "grad_norm": 0.205078125,
+      "learning_rate": 0.007344632768361582,
+      "loss": 2.0889,
+      "step": 65
+    },
+    {
+      "epoch": 0.0067109828793675405,
+      "grad_norm": 0.1953125,
+      "learning_rate": 0.007457627118644068,
+      "loss": 2.0393,
+      "step": 66
+    },
+    {
+      "epoch": 0.006812664438145837,
+      "grad_norm": 0.2138671875,
+      "learning_rate": 0.007570621468926554,
+      "loss": 2.067,
+      "step": 67
+    },
+    {
+      "epoch": 0.006914345996924133,
+      "grad_norm": 0.2060546875,
+      "learning_rate": 0.0076836158192090396,
+      "loss": 2.1042,
+      "step": 68
+    },
+    {
+      "epoch": 0.007016027555702429,
+      "grad_norm": 0.201171875,
+      "learning_rate": 0.007796610169491526,
+      "loss": 2.081,
+      "step": 69
+    },
+    {
+      "epoch": 0.007117709114480725,
+      "grad_norm": 0.205078125,
+      "learning_rate": 0.007909604519774013,
+      "loss": 2.0768,
+      "step": 70
+    },
+    {
+      "epoch": 0.0072193906732590215,
+      "grad_norm": 0.23828125,
+      "learning_rate": 0.008022598870056498,
+      "loss": 2.1325,
+      "step": 71
+    },
+    {
+      "epoch": 0.007321072232037317,
+      "grad_norm": 0.220703125,
+      "learning_rate": 0.008135593220338983,
+      "loss": 2.1196,
+      "step": 72
+    },
+    {
+      "epoch": 0.007422753790815613,
+      "grad_norm": 0.2314453125,
+      "learning_rate": 0.008248587570621468,
+      "loss": 2.1501,
+      "step": 73
+    },
+    {
+      "epoch": 0.00752443534959391,
+      "grad_norm": 0.1953125,
+      "learning_rate": 0.008361581920903955,
+      "loss": 2.0384,
+      "step": 74
+    },
+    {
+      "epoch": 0.007626116908372205,
+      "grad_norm": 0.2109375,
+      "learning_rate": 0.00847457627118644,
+      "loss": 2.0667,
+      "step": 75
+    },
+    {
+      "epoch": 0.007727798467150502,
+      "grad_norm": 0.203125,
+      "learning_rate": 0.008587570621468926,
+      "loss": 2.1045,
+      "step": 76
+    },
+    {
+      "epoch": 0.007829480025928797,
+      "grad_norm": 0.2080078125,
+      "learning_rate": 0.008700564971751413,
+      "loss": 2.077,
+      "step": 77
+    },
+    {
+      "epoch": 0.007931161584707094,
+      "grad_norm": 0.212890625,
+      "learning_rate": 0.008813559322033898,
+      "loss": 2.0602,
+      "step": 78
+    },
+    {
+      "epoch": 0.00803284314348539,
+      "grad_norm": 0.22265625,
+      "learning_rate": 0.008926553672316385,
+      "loss": 2.0987,
+      "step": 79
+    },
+    {
+      "epoch": 0.008134524702263685,
+      "grad_norm": 0.2294921875,
+      "learning_rate": 0.00903954802259887,
+      "loss": 2.1236,
+      "step": 80
+    },
+    {
+      "epoch": 0.008236206261041983,
+      "grad_norm": 0.21875,
+      "learning_rate": 0.009152542372881356,
+      "loss": 2.0861,
+      "step": 81
+    },
+    {
+      "epoch": 0.008337887819820278,
+      "grad_norm": 0.19921875,
+      "learning_rate": 0.009265536723163843,
+      "loss": 2.0189,
+      "step": 82
+    },
+    {
+      "epoch": 0.008439569378598574,
+      "grad_norm": 0.2421875,
+      "learning_rate": 0.009378531073446328,
+      "loss": 2.0509,
+      "step": 83
+    },
+    {
+      "epoch": 0.00854125093737687,
+      "grad_norm": 0.2177734375,
+      "learning_rate": 0.009491525423728815,
+      "loss": 2.0923,
+      "step": 84
+    },
+    {
+      "epoch": 0.008642932496155166,
+      "grad_norm": 0.2177734375,
+      "learning_rate": 0.0096045197740113,
+      "loss": 2.0488,
+      "step": 85
+    },
+    {
+      "epoch": 0.008744614054933462,
+      "grad_norm": 0.2236328125,
+      "learning_rate": 0.009717514124293785,
+      "loss": 2.0658,
+      "step": 86
+    },
+    {
+      "epoch": 0.008846295613711759,
+      "grad_norm": 0.26171875,
+      "learning_rate": 0.009830508474576272,
+      "loss": 2.032,
+      "step": 87
+    },
+    {
+      "epoch": 0.008947977172490055,
+      "grad_norm": 0.232421875,
+      "learning_rate": 0.009943502824858758,
+      "loss": 2.0085,
+      "step": 88
+    },
+    {
+      "epoch": 0.00904965873126835,
+      "grad_norm": 0.259765625,
+      "learning_rate": 0.010056497175141245,
+      "loss": 2.0155,
+      "step": 89
+    },
+    {
+      "epoch": 0.009151340290046646,
+      "grad_norm": 0.2412109375,
+      "learning_rate": 0.010169491525423728,
+      "loss": 2.047,
+      "step": 90
+    },
+    {
+      "epoch": 0.009253021848824943,
+      "grad_norm": 0.2431640625,
+      "learning_rate": 0.010282485875706215,
+      "loss": 2.0138,
+      "step": 91
+    },
+    {
+      "epoch": 0.009354703407603238,
+      "grad_norm": 0.224609375,
+      "learning_rate": 0.0103954802259887,
+      "loss": 2.0701,
+      "step": 92
+    },
+    {
+      "epoch": 0.009456384966381534,
+      "grad_norm": 0.23046875,
+      "learning_rate": 0.010508474576271187,
+      "loss": 1.9933,
+      "step": 93
+    },
+    {
+      "epoch": 0.009558066525159831,
+      "grad_norm": 0.23046875,
+      "learning_rate": 0.010621468926553673,
+      "loss": 2.0354,
+      "step": 94
+    },
+    {
+      "epoch": 0.009659748083938127,
+      "grad_norm": 0.2353515625,
+      "learning_rate": 0.01073446327683616,
+      "loss": 2.0852,
+      "step": 95
+    },
+    {
+      "epoch": 0.009761429642716422,
+      "grad_norm": 0.25,
+      "learning_rate": 0.010847457627118645,
+      "loss": 2.0171,
+      "step": 96
+    },
+    {
+      "epoch": 0.00986311120149472,
+      "grad_norm": 0.240234375,
+      "learning_rate": 0.010960451977401128,
+      "loss": 1.9798,
+      "step": 97
+    },
+    {
+      "epoch": 0.009964792760273015,
+      "grad_norm": 0.2158203125,
+      "learning_rate": 0.011073446327683615,
+      "loss": 2.0219,
+      "step": 98
+    },
+    {
+      "epoch": 0.01006647431905131,
+      "grad_norm": 0.2421875,
+      "learning_rate": 0.0111864406779661,
+      "loss": 2.0537,
+      "step": 99
+    },
+    {
+      "epoch": 0.010168155877829608,
+      "grad_norm": 0.2392578125,
+      "learning_rate": 0.011299435028248588,
+      "loss": 2.124,
+      "step": 100
+    },
+    {
+      "epoch": 0.010269837436607903,
+      "grad_norm": 0.259765625,
+      "learning_rate": 0.011412429378531073,
+      "loss": 2.0611,
+      "step": 101
+    },
+    {
+      "epoch": 0.010371518995386199,
+      "grad_norm": 0.2294921875,
+      "learning_rate": 0.01152542372881356,
+      "loss": 2.0089,
+      "step": 102
+    },
+    {
+      "epoch": 0.010473200554164496,
+      "grad_norm": 0.2275390625,
+      "learning_rate": 0.011638418079096045,
+      "loss": 2.0065,
+      "step": 103
+    },
+    {
+      "epoch": 0.010574882112942791,
+      "grad_norm": 0.2412109375,
+      "learning_rate": 0.011751412429378532,
+      "loss": 1.9934,
+      "step": 104
+    },
+    {
+      "epoch": 0.010676563671721087,
+      "grad_norm": 0.248046875,
+      "learning_rate": 0.011864406779661017,
+      "loss": 2.0497,
+      "step": 105
+    },
+    {
+      "epoch": 0.010778245230499384,
+      "grad_norm": 0.23828125,
+      "learning_rate": 0.011977401129943504,
+      "loss": 2.0126,
+      "step": 106
+    },
+    {
+      "epoch": 0.01087992678927768,
+      "grad_norm": 0.27734375,
+      "learning_rate": 0.012090395480225988,
+      "loss": 2.0156,
+      "step": 107
+    },
+    {
+      "epoch": 0.010981608348055975,
+      "grad_norm": 0.2294921875,
+      "learning_rate": 0.012203389830508476,
+      "loss": 2.0018,
+      "step": 108
+    },
+    {
+      "epoch": 0.011083289906834272,
+      "grad_norm": 0.294921875,
+      "learning_rate": 0.01231638418079096,
+      "loss": 1.99,
+      "step": 109
+    },
+    {
+      "epoch": 0.011184971465612568,
+      "grad_norm": 0.296875,
+      "learning_rate": 0.012429378531073447,
+      "loss": 2.0445,
+      "step": 110
+    },
+    {
+      "epoch": 0.011286653024390863,
+      "grad_norm": 0.32421875,
+      "learning_rate": 0.012542372881355932,
+      "loss": 2.0848,
+      "step": 111
+    },
+    {
+      "epoch": 0.01138833458316916,
+      "grad_norm": 0.318359375,
+      "learning_rate": 0.01265536723163842,
+      "loss": 2.0249,
+      "step": 112
+    },
+    {
+      "epoch": 0.011490016141947456,
+      "grad_norm": 0.318359375,
+      "learning_rate": 0.012768361581920904,
+      "loss": 2.0644,
+      "step": 113
+    },
+    {
+      "epoch": 0.011591697700725751,
+      "grad_norm": 0.306640625,
+      "learning_rate": 0.01288135593220339,
+      "loss": 2.0365,
+      "step": 114
+    },
+    {
+      "epoch": 0.011693379259504049,
+      "grad_norm": 0.3046875,
+      "learning_rate": 0.012994350282485877,
+      "loss": 2.0184,
+      "step": 115
+    },
+    {
+      "epoch": 0.011795060818282344,
+      "grad_norm": 0.296875,
+      "learning_rate": 0.01310734463276836,
+      "loss": 2.0587,
+      "step": 116
+    },
+    {
+      "epoch": 0.01189674237706064,
+      "grad_norm": 0.265625,
+      "learning_rate": 0.013220338983050847,
+      "loss": 2.0332,
+      "step": 117
+    },
+    {
+      "epoch": 0.011998423935838937,
+      "grad_norm": 0.267578125,
+      "learning_rate": 0.013333333333333332,
+      "loss": 1.9925,
+      "step": 118
+    },
+    {
+      "epoch": 0.012100105494617232,
+      "grad_norm": 0.29296875,
+      "learning_rate": 0.01344632768361582,
+      "loss": 2.008,
+      "step": 119
+    },
+    {
+      "epoch": 0.012201787053395528,
+      "grad_norm": 0.279296875,
+      "learning_rate": 0.013559322033898305,
+      "loss": 2.0001,
+      "step": 120
+    },
+    {
+      "epoch": 0.012303468612173825,
+      "grad_norm": 0.271484375,
+      "learning_rate": 0.013672316384180792,
+      "loss": 2.0134,
+      "step": 121
+    },
+    {
+      "epoch": 0.01240515017095212,
+      "grad_norm": 0.28125,
+      "learning_rate": 0.013785310734463277,
+      "loss": 2.0145,
+      "step": 122
+    },
+    {
+      "epoch": 0.012506831729730416,
+      "grad_norm": 0.306640625,
+      "learning_rate": 0.013898305084745764,
+      "loss": 2.0826,
+      "step": 123
+    },
+    {
+      "epoch": 0.012608513288508713,
+      "grad_norm": 0.30859375,
+      "learning_rate": 0.01401129943502825,
+      "loss": 2.0491,
+      "step": 124
+    },
+    {
+      "epoch": 0.012710194847287009,
+      "grad_norm": 0.306640625,
+      "learning_rate": 0.014124293785310736,
+      "loss": 2.0546,
+      "step": 125
+    },
+    {
+      "epoch": 0.012811876406065304,
+      "grad_norm": 0.3046875,
+      "learning_rate": 0.01423728813559322,
+      "loss": 1.9846,
+      "step": 126
+    },
+    {
+      "epoch": 0.012913557964843602,
+      "grad_norm": 0.2890625,
+      "learning_rate": 0.014350282485875707,
+      "loss": 2.0048,
+      "step": 127
+    },
+    {
+      "epoch": 0.013015239523621897,
+      "grad_norm": 0.2734375,
+      "learning_rate": 0.014463276836158192,
+      "loss": 2.0125,
+      "step": 128
+    },
+    {
+      "epoch": 0.013116921082400193,
+      "grad_norm": 0.294921875,
+      "learning_rate": 0.014576271186440677,
+      "loss": 2.0268,
+      "step": 129
+    },
+    {
+      "epoch": 0.01321860264117849,
+      "grad_norm": 0.279296875,
+      "learning_rate": 0.014689265536723164,
+      "loss": 2.0574,
+      "step": 130
+    },
+    {
+      "epoch": 0.013320284199956785,
+      "grad_norm": 0.3046875,
+      "learning_rate": 0.01480225988700565,
+      "loss": 2.0416,
+      "step": 131
+    },
+    {
+      "epoch": 0.013421965758735081,
+      "grad_norm": 0.294921875,
+      "learning_rate": 0.014915254237288136,
+      "loss": 2.0474,
+      "step": 132
+    },
+    {
+      "epoch": 0.013523647317513378,
+      "grad_norm": 0.3125,
+      "learning_rate": 0.015028248587570622,
+      "loss": 2.0576,
+      "step": 133
+    },
+    {
+      "epoch": 0.013625328876291674,
+      "grad_norm": 0.30078125,
+      "learning_rate": 0.015141242937853109,
+      "loss": 1.9987,
+      "step": 134
+    },
+    {
+      "epoch": 0.01372701043506997,
+      "grad_norm": 0.30859375,
+      "learning_rate": 0.015254237288135592,
+      "loss": 1.9953,
+      "step": 135
+    },
+    {
+      "epoch": 0.013828691993848266,
+      "grad_norm": 0.29296875,
+      "learning_rate": 0.015367231638418079,
+      "loss": 2.0199,
+      "step": 136
+    },
+    {
+      "epoch": 0.013930373552626562,
+      "grad_norm": 0.3203125,
+      "learning_rate": 0.015480225988700564,
+      "loss": 2.0736,
+      "step": 137
+    },
+    {
+      "epoch": 0.014032055111404857,
+      "grad_norm": 0.337890625,
+      "learning_rate": 0.015593220338983051,
+      "loss": 1.9911,
+      "step": 138
+    },
+    {
+      "epoch": 0.014133736670183155,
+      "grad_norm": 0.34375,
+      "learning_rate": 0.015706214689265537,
+      "loss": 2.0036,
+      "step": 139
+    },
+    {
+      "epoch": 0.01423541822896145,
+      "grad_norm": 0.31640625,
+      "learning_rate": 0.015819209039548025,
+      "loss": 2.0224,
+      "step": 140
+    },
+    {
+      "epoch": 0.014337099787739746,
+      "grad_norm": 0.30078125,
+      "learning_rate": 0.015932203389830507,
+      "loss": 1.9657,
+      "step": 141
+    },
+    {
+      "epoch": 0.014438781346518043,
+      "grad_norm": 0.314453125,
+      "learning_rate": 0.016045197740112996,
+      "loss": 2.039,
+      "step": 142
+    },
+    {
+      "epoch": 0.014540462905296338,
+      "grad_norm": 0.30859375,
+      "learning_rate": 0.01615819209039548,
+      "loss": 2.0156,
+      "step": 143
+    },
+    {
+      "epoch": 0.014642144464074634,
+      "grad_norm": 0.31640625,
+      "learning_rate": 0.016271186440677966,
+      "loss": 2.0155,
+      "step": 144
+    },
+    {
+      "epoch": 0.014743826022852931,
+      "grad_norm": 0.296875,
+      "learning_rate": 0.01638418079096045,
+      "loss": 1.9882,
+      "step": 145
+    },
+    {
+      "epoch": 0.014845507581631227,
+      "grad_norm": 0.310546875,
+      "learning_rate": 0.016497175141242937,
+      "loss": 1.96,
+      "step": 146
+    },
+    {
+      "epoch": 0.014947189140409522,
+      "grad_norm": 0.296875,
+      "learning_rate": 0.016610169491525426,
+      "loss": 2.0671,
+      "step": 147
+    },
+    {
+      "epoch": 0.01504887069918782,
+      "grad_norm": 0.294921875,
+      "learning_rate": 0.01672316384180791,
+      "loss": 2.0444,
+      "step": 148
+    },
+    {
+      "epoch": 0.015150552257966115,
+      "grad_norm": 0.267578125,
+      "learning_rate": 0.016836158192090396,
+      "loss": 2.0075,
+      "step": 149
+    },
+    {
+      "epoch": 0.01525223381674441,
+      "grad_norm": 0.279296875,
+      "learning_rate": 0.01694915254237288,
+      "loss": 1.958,
+      "step": 150
+    },
+    {
+      "epoch": 0.015353915375522706,
+      "grad_norm": 0.271484375,
+      "learning_rate": 0.017062146892655367,
+      "loss": 1.97,
+      "step": 151
+    },
+    {
+      "epoch": 0.015455596934301003,
+      "grad_norm": 0.3125,
+      "learning_rate": 0.017175141242937852,
+      "loss": 1.9909,
+      "step": 152
+    },
+    {
+      "epoch": 0.015557278493079299,
+      "grad_norm": 0.26953125,
+      "learning_rate": 0.01728813559322034,
+      "loss": 1.9762,
+      "step": 153
+    },
+    {
+      "epoch": 0.015658960051857594,
+      "grad_norm": 0.287109375,
+      "learning_rate": 0.017401129943502826,
+      "loss": 1.9928,
+      "step": 154
+    },
+    {
+      "epoch": 0.01576064161063589,
+      "grad_norm": 0.271484375,
+      "learning_rate": 0.01751412429378531,
+      "loss": 1.9856,
+      "step": 155
+    },
+    {
+      "epoch": 0.01586232316941419,
+      "grad_norm": 0.302734375,
+      "learning_rate": 0.017627118644067796,
+      "loss": 2.0208,
+      "step": 156
+    },
+    {
+      "epoch": 0.015964004728192482,
+      "grad_norm": 0.30078125,
+      "learning_rate": 0.017740112994350285,
+      "loss": 2.064,
+      "step": 157
+    },
+    {
+      "epoch": 0.01606568628697078,
+      "grad_norm": 0.3046875,
+      "learning_rate": 0.01785310734463277,
+      "loss": 1.9849,
+      "step": 158
+    },
+    {
+      "epoch": 0.016167367845749077,
+      "grad_norm": 0.2734375,
+      "learning_rate": 0.017966101694915255,
+      "loss": 1.9804,
+      "step": 159
+    },
+    {
+      "epoch": 0.01626904940452737,
+      "grad_norm": 0.27734375,
+      "learning_rate": 0.01807909604519774,
+      "loss": 1.9864,
+      "step": 160
+    },
+    {
+      "epoch": 0.016370730963305668,
+      "grad_norm": 0.296875,
+      "learning_rate": 0.018192090395480226,
+      "loss": 2.0279,
+      "step": 161
+    },
+    {
+      "epoch": 0.016472412522083965,
+      "grad_norm": 0.326171875,
+      "learning_rate": 0.01830508474576271,
+      "loss": 2.0376,
+      "step": 162
+    },
+    {
+      "epoch": 0.01657409408086226,
+      "grad_norm": 0.291015625,
+      "learning_rate": 0.018418079096045196,
+      "loss": 2.0915,
+      "step": 163
+    },
+    {
+      "epoch": 0.016675775639640556,
+      "grad_norm": 0.2890625,
+      "learning_rate": 0.018531073446327685,
+      "loss": 2.0137,
+      "step": 164
+    },
+    {
+      "epoch": 0.016777457198418853,
+      "grad_norm": 0.2890625,
+      "learning_rate": 0.01864406779661017,
+      "loss": 1.9784,
+      "step": 165
+    },
+    {
+      "epoch": 0.016879138757197147,
+      "grad_norm": 0.28515625,
+      "learning_rate": 0.018757062146892656,
+      "loss": 2.0182,
+      "step": 166
+    },
+    {
+      "epoch": 0.016980820315975444,
+      "grad_norm": 0.275390625,
+      "learning_rate": 0.01887005649717514,
+      "loss": 1.9851,
+      "step": 167
+    },
+    {
+      "epoch": 0.01708250187475374,
+      "grad_norm": 0.302734375,
+      "learning_rate": 0.01898305084745763,
+      "loss": 2.0214,
+      "step": 168
+    },
+    {
+      "epoch": 0.017184183433532035,
+      "grad_norm": 0.294921875,
+      "learning_rate": 0.01909604519774011,
+      "loss": 2.0199,
+      "step": 169
+    },
+    {
+      "epoch": 0.017285864992310333,
+      "grad_norm": 0.310546875,
+      "learning_rate": 0.0192090395480226,
+      "loss": 2.0289,
+      "step": 170
+    },
+    {
+      "epoch": 0.01738754655108863,
+      "grad_norm": 0.298828125,
+      "learning_rate": 0.019322033898305085,
+      "loss": 1.9159,
+      "step": 171
+    },
+    {
+      "epoch": 0.017489228109866924,
+      "grad_norm": 0.34765625,
+      "learning_rate": 0.01943502824858757,
+      "loss": 2.0473,
+      "step": 172
+    },
+    {
+      "epoch": 0.01759090966864522,
+      "grad_norm": 0.33984375,
+      "learning_rate": 0.019548022598870056,
+      "loss": 2.0882,
+      "step": 173
+    },
+    {
+      "epoch": 0.017692591227423518,
+      "grad_norm": 0.349609375,
+      "learning_rate": 0.019661016949152545,
+      "loss": 1.9956,
+      "step": 174
+    },
+    {
+      "epoch": 0.017794272786201812,
+      "grad_norm": 0.318359375,
+      "learning_rate": 0.01977401129943503,
+      "loss": 1.9571,
+      "step": 175
+    },
+    {
+      "epoch": 0.01789595434498011,
+      "grad_norm": 0.369140625,
+      "learning_rate": 0.019887005649717515,
+      "loss": 2.0089,
+      "step": 176
+    },
+    {
+      "epoch": 0.017997635903758406,
+      "grad_norm": 0.32421875,
+      "learning_rate": 0.02,
+      "loss": 2.0011,
+      "step": 177
+    },
+    {
+      "epoch": 0.0180993174625367,
+      "grad_norm": 0.330078125,
+      "learning_rate": 0.02,
+      "loss": 2.007,
+      "step": 178
+    },
+    {
+      "epoch": 0.018200999021314997,
+      "grad_norm": 0.306640625,
+      "learning_rate": 0.02,
+      "loss": 1.9815,
+      "step": 179
+    },
+    {
+      "epoch": 0.01830268058009329,
+      "grad_norm": 0.345703125,
+      "learning_rate": 0.02,
+      "loss": 2.062,
+      "step": 180
+    },
+    {
+      "epoch": 0.01840436213887159,
+      "grad_norm": 0.310546875,
+      "learning_rate": 0.02,
+      "loss": 1.9555,
+      "step": 181
+    },
+    {
+      "epoch": 0.018506043697649886,
+      "grad_norm": 0.32421875,
+      "learning_rate": 0.02,
+      "loss": 2.0017,
+      "step": 182
+    },
+    {
+      "epoch": 0.01860772525642818,
+      "grad_norm": 0.32421875,
+      "learning_rate": 0.02,
+      "loss": 2.0067,
+      "step": 183
+    },
+    {
+      "epoch": 0.018709406815206477,
+      "grad_norm": 0.3359375,
+      "learning_rate": 0.02,
+      "loss": 2.0034,
+      "step": 184
+    },
+    {
+      "epoch": 0.018811088373984774,
+      "grad_norm": 0.28125,
+      "learning_rate": 0.02,
+      "loss": 1.9166,
+      "step": 185
+    },
+    {
+      "epoch": 0.018912769932763068,
+      "grad_norm": 0.310546875,
+      "learning_rate": 0.02,
+      "loss": 2.0066,
+      "step": 186
+    },
+    {
+      "epoch": 0.019014451491541365,
+      "grad_norm": 0.30078125,
+      "learning_rate": 0.02,
+      "loss": 1.9807,
+      "step": 187
+    },
+    {
+      "epoch": 0.019116133050319662,
+      "grad_norm": 0.294921875,
+      "learning_rate": 0.02,
+      "loss": 1.9777,
+      "step": 188
+    },
+    {
+      "epoch": 0.019217814609097956,
+      "grad_norm": 0.302734375,
+      "learning_rate": 0.02,
+      "loss": 2.005,
+      "step": 189
+    },
+    {
+      "epoch": 0.019319496167876253,
+      "grad_norm": 0.3046875,
+      "learning_rate": 0.02,
+      "loss": 2.0068,
+      "step": 190
+    },
+    {
+      "epoch": 0.01942117772665455,
+      "grad_norm": 0.29296875,
+      "learning_rate": 0.02,
+      "loss": 2.0115,
+      "step": 191
+    },
+    {
+      "epoch": 0.019522859285432844,
+      "grad_norm": 0.28125,
+      "learning_rate": 0.02,
+      "loss": 2.0295,
+      "step": 192
+    },
+    {
+      "epoch": 0.01962454084421114,
+      "grad_norm": 0.28125,
+      "learning_rate": 0.02,
+      "loss": 1.9956,
+      "step": 193
+    },
+    {
+      "epoch": 0.01972622240298944,
+      "grad_norm": 0.26171875,
+      "learning_rate": 0.02,
+      "loss": 1.9922,
+      "step": 194
+    },
+    {
+      "epoch": 0.019827903961767732,
+      "grad_norm": 0.25,
+      "learning_rate": 0.02,
+      "loss": 2.0226,
+      "step": 195
+    },
+    {
+      "epoch": 0.01992958552054603,
+      "grad_norm": 0.267578125,
+      "learning_rate": 0.02,
+      "loss": 2.0312,
+      "step": 196
+    },
+    {
+      "epoch": 0.020031267079324327,
+      "grad_norm": 0.26171875,
+      "learning_rate": 0.02,
+      "loss": 1.971,
+      "step": 197
+    },
+    {
+      "epoch": 0.02013294863810262,
+      "grad_norm": 0.29296875,
+      "learning_rate": 0.02,
+      "loss": 1.9968,
+      "step": 198
+    },
+    {
+      "epoch": 0.020234630196880918,
+      "grad_norm": 0.3203125,
+      "learning_rate": 0.02,
+      "loss": 2.0127,
+      "step": 199
+    },
+    {
+      "epoch": 0.020336311755659215,
+      "grad_norm": 0.34765625,
+      "learning_rate": 0.02,
+      "loss": 1.97,
+      "step": 200
+    },
+    {
+      "epoch": 0.02043799331443751,
+      "grad_norm": 0.345703125,
+      "learning_rate": 0.02,
+      "loss": 2.0109,
+      "step": 201
+    },
+    {
+      "epoch": 0.020539674873215806,
+      "grad_norm": 0.365234375,
+      "learning_rate": 0.02,
+      "loss": 2.034,
+      "step": 202
+    },
+    {
+      "epoch": 0.020641356431994103,
+      "grad_norm": 0.38671875,
+      "learning_rate": 0.02,
+      "loss": 2.1007,
+      "step": 203
+    },
+    {
+      "epoch": 0.020743037990772397,
+      "grad_norm": 0.32421875,
+      "learning_rate": 0.02,
+      "loss": 1.9621,
+      "step": 204
+    },
+    {
+      "epoch": 0.020844719549550694,
+      "grad_norm": 0.3046875,
+      "learning_rate": 0.02,
+      "loss": 1.98,
+      "step": 205
+    },
+    {
+      "epoch": 0.02094640110832899,
+      "grad_norm": 0.283203125,
+      "learning_rate": 0.02,
+      "loss": 1.9535,
+      "step": 206
+    },
+    {
+      "epoch": 0.021048082667107285,
+      "grad_norm": 0.3046875,
+      "learning_rate": 0.02,
+      "loss": 1.982,
+      "step": 207
+    },
+    {
+      "epoch": 0.021149764225885583,
+      "grad_norm": 0.28515625,
+      "learning_rate": 0.02,
+      "loss": 1.9799,
+      "step": 208
+    },
+    {
+      "epoch": 0.02125144578466388,
+      "grad_norm": 0.3203125,
+      "learning_rate": 0.02,
+      "loss": 1.9682,
+      "step": 209
+    },
+    {
+      "epoch": 0.021353127343442174,
+      "grad_norm": 0.298828125,
+      "learning_rate": 0.02,
+      "loss": 2.0116,
+      "step": 210
+    },
+    {
+      "epoch": 0.02145480890222047,
+      "grad_norm": 0.29296875,
+      "learning_rate": 0.02,
+      "loss": 1.982,
+      "step": 211
+    },
+    {
+      "epoch": 0.021556490460998768,
+      "grad_norm": 0.271484375,
+      "learning_rate": 0.02,
+      "loss": 1.9805,
+      "step": 212
+    },
+    {
+      "epoch": 0.021658172019777062,
+      "grad_norm": 0.26953125,
+      "learning_rate": 0.02,
+      "loss": 1.9555,
+      "step": 213
+    },
+    {
+      "epoch": 0.02175985357855536,
+      "grad_norm": 0.244140625,
+      "learning_rate": 0.02,
+      "loss": 1.9305,
+      "step": 214
+    },
+    {
+      "epoch": 0.021861535137333656,
+      "grad_norm": 0.25390625,
+      "learning_rate": 0.02,
+      "loss": 1.9542,
+      "step": 215
+    },
+    {
+      "epoch": 0.02196321669611195,
+      "grad_norm": 0.263671875,
+      "learning_rate": 0.02,
+      "loss": 1.931,
+      "step": 216
+    },
+    {
+      "epoch": 0.022064898254890247,
+      "grad_norm": 0.265625,
+      "learning_rate": 0.02,
+      "loss": 2.0024,
+      "step": 217
+    },
+    {
+      "epoch": 0.022166579813668544,
+      "grad_norm": 0.275390625,
+      "learning_rate": 0.02,
+      "loss": 1.9948,
+      "step": 218
+    },
+    {
+      "epoch": 0.022268261372446838,
+      "grad_norm": 0.27734375,
+      "learning_rate": 0.02,
+      "loss": 2.0353,
+      "step": 219
+    },
+    {
+      "epoch": 0.022369942931225135,
+      "grad_norm": 0.283203125,
+      "learning_rate": 0.02,
+      "loss": 1.9739,
+      "step": 220
+    },
+    {
+      "epoch": 0.022471624490003433,
+      "grad_norm": 0.271484375,
+      "learning_rate": 0.02,
+      "loss": 2.0091,
+      "step": 221
+    },
+    {
+      "epoch": 0.022573306048781727,
+      "grad_norm": 0.275390625,
+      "learning_rate": 0.02,
+      "loss": 2.0047,
+      "step": 222
+    },
+    {
+      "epoch": 0.022674987607560024,
+      "grad_norm": 0.263671875,
+      "learning_rate": 0.02,
+      "loss": 1.9813,
+      "step": 223
+    },
+    {
+      "epoch": 0.02277666916633832,
+      "grad_norm": 0.30078125,
+      "learning_rate": 0.02,
+      "loss": 1.9593,
+      "step": 224
+    },
+    {
+      "epoch": 0.022878350725116615,
+      "grad_norm": 0.267578125,
+      "learning_rate": 0.02,
+      "loss": 1.9584,
+      "step": 225
+    },
+    {
+      "epoch": 0.022980032283894912,
+      "grad_norm": 0.29296875,
+      "learning_rate": 0.02,
+      "loss": 1.9614,
+      "step": 226
+    },
+    {
+      "epoch": 0.02308171384267321,
+      "grad_norm": 0.251953125,
+      "learning_rate": 0.02,
+      "loss": 1.9228,
+      "step": 227
+    },
+    {
+      "epoch": 0.023183395401451503,
+      "grad_norm": 0.291015625,
+      "learning_rate": 0.02,
+      "loss": 1.9252,
+      "step": 228
+    },
+    {
+      "epoch": 0.0232850769602298,
+      "grad_norm": 0.263671875,
+      "learning_rate": 0.02,
+      "loss": 1.955,
+      "step": 229
+    },
+    {
+      "epoch": 0.023386758519008097,
+      "grad_norm": 0.302734375,
+      "learning_rate": 0.02,
+      "loss": 1.9955,
+      "step": 230
+    },
+    {
+      "epoch": 0.02348844007778639,
+      "grad_norm": 0.279296875,
+      "learning_rate": 0.02,
+      "loss": 2.006,
+      "step": 231
+    },
+    {
+      "epoch": 0.02359012163656469,
+      "grad_norm": 0.306640625,
+      "learning_rate": 0.02,
+      "loss": 2.0029,
+      "step": 232
+    },
+    {
+      "epoch": 0.023691803195342986,
+      "grad_norm": 0.29296875,
+      "learning_rate": 0.02,
+      "loss": 2.0084,
+      "step": 233
+    },
+    {
+      "epoch": 0.02379348475412128,
+      "grad_norm": 0.30078125,
+      "learning_rate": 0.02,
+      "loss": 2.0177,
+      "step": 234
+    },
+    {
+      "epoch": 0.023895166312899577,
+      "grad_norm": 0.2734375,
+      "learning_rate": 0.02,
+      "loss": 2.0138,
+      "step": 235
+    },
+    {
+      "epoch": 0.023996847871677874,
+      "grad_norm": 0.267578125,
+      "learning_rate": 0.02,
+      "loss": 1.9433,
+      "step": 236
+    },
+    {
+      "epoch": 0.024098529430456168,
+      "grad_norm": 0.259765625,
+      "learning_rate": 0.02,
+      "loss": 1.9131,
+      "step": 237
+    },
+    {
+      "epoch": 0.024200210989234465,
+      "grad_norm": 0.26953125,
+      "learning_rate": 0.02,
+      "loss": 1.9632,
+      "step": 238
+    },
+    {
+      "epoch": 0.024301892548012762,
+      "grad_norm": 0.275390625,
+      "learning_rate": 0.02,
+      "loss": 1.9729,
+      "step": 239
+    },
+    {
+      "epoch": 0.024403574106791056,
+      "grad_norm": 0.2490234375,
+      "learning_rate": 0.02,
+      "loss": 1.9081,
+      "step": 240
+    },
+    {
+      "epoch": 0.024505255665569353,
+      "grad_norm": 0.2431640625,
+      "learning_rate": 0.02,
+      "loss": 1.938,
+      "step": 241
+    },
+    {
+      "epoch": 0.02460693722434765,
+      "grad_norm": 0.23828125,
+      "learning_rate": 0.02,
+      "loss": 1.9033,
+      "step": 242
+    },
+    {
+      "epoch": 0.024708618783125944,
+      "grad_norm": 0.26171875,
+      "learning_rate": 0.02,
+      "loss": 1.9676,
+      "step": 243
+    },
+    {
+      "epoch": 0.02481030034190424,
+      "grad_norm": 0.2392578125,
+      "learning_rate": 0.02,
+      "loss": 1.9472,
+      "step": 244
+    },
+    {
+      "epoch": 0.02491198190068254,
+      "grad_norm": 0.2470703125,
+      "learning_rate": 0.02,
+      "loss": 1.9551,
+      "step": 245
+    },
+    {
+      "epoch": 0.025013663459460832,
+      "grad_norm": 0.2431640625,
+      "learning_rate": 0.02,
+      "loss": 1.9475,
+      "step": 246
+    },
+    {
+      "epoch": 0.02511534501823913,
+      "grad_norm": 0.2578125,
+      "learning_rate": 0.02,
+      "loss": 1.9453,
+      "step": 247
+    },
+    {
+      "epoch": 0.025217026577017427,
+      "grad_norm": 0.2421875,
+      "learning_rate": 0.02,
+      "loss": 1.9622,
+      "step": 248
+    },
+    {
+      "epoch": 0.02531870813579572,
+      "grad_norm": 0.2412109375,
+      "learning_rate": 0.02,
+      "loss": 1.952,
+      "step": 249
+    },
+    {
+      "epoch": 0.025420389694574018,
+      "grad_norm": 0.234375,
+      "learning_rate": 0.02,
+      "loss": 1.9272,
+      "step": 250
+    },
+    {
+      "epoch": 0.025522071253352315,
+      "grad_norm": 0.26171875,
+      "learning_rate": 0.02,
+      "loss": 1.9224,
+      "step": 251
+    },
+    {
+      "epoch": 0.02562375281213061,
+      "grad_norm": 0.26171875,
+      "learning_rate": 0.02,
+      "loss": 1.9779,
+      "step": 252
+    },
+    {
+      "epoch": 0.025725434370908906,
+      "grad_norm": 0.27734375,
+      "learning_rate": 0.02,
+      "loss": 1.981,
+      "step": 253
+    },
+    {
+      "epoch": 0.025827115929687203,
+      "grad_norm": 0.2421875,
+      "learning_rate": 0.02,
+      "loss": 1.9122,
+      "step": 254
+    },
+    {
+      "epoch": 0.025928797488465497,
+      "grad_norm": 0.26953125,
+      "learning_rate": 0.02,
+      "loss": 1.9942,
+      "step": 255
+    },
+    {
+      "epoch": 0.026030479047243794,
+      "grad_norm": 0.2734375,
+      "learning_rate": 0.02,
+      "loss": 1.965,
+      "step": 256
+    },
+    {
+      "epoch": 0.02613216060602209,
+      "grad_norm": 0.26953125,
+      "learning_rate": 0.02,
+      "loss": 1.9614,
+      "step": 257
+    },
+    {
+      "epoch": 0.026233842164800385,
+      "grad_norm": 0.212890625,
+      "learning_rate": 0.02,
+      "loss": 1.9624,
+      "step": 258
+    },
+    {
+      "epoch": 0.026335523723578683,
+      "grad_norm": 0.23046875,
+      "learning_rate": 0.02,
+      "loss": 1.9564,
+      "step": 259
+    },
+    {
+      "epoch": 0.02643720528235698,
+      "grad_norm": 0.21875,
+      "learning_rate": 0.02,
+      "loss": 1.9381,
+      "step": 260
+    },
+    {
+      "epoch": 0.026538886841135274,
+      "grad_norm": 0.22265625,
+      "learning_rate": 0.02,
+      "loss": 1.9782,
+      "step": 261
+    },
+    {
+      "epoch": 0.02664056839991357,
+      "grad_norm": 0.2060546875,
+      "learning_rate": 0.02,
+      "loss": 2.0137,
+      "step": 262
+    },
+    {
+      "epoch": 0.026742249958691868,
+      "grad_norm": 0.2216796875,
+      "learning_rate": 0.02,
+      "loss": 1.9617,
+      "step": 263
+    },
+    {
+      "epoch": 0.026843931517470162,
+      "grad_norm": 0.20703125,
+      "learning_rate": 0.02,
+      "loss": 1.9304,
+      "step": 264
+    },
+    {
+      "epoch": 0.02694561307624846,
+      "grad_norm": 0.2158203125,
+      "learning_rate": 0.02,
+      "loss": 1.9319,
+      "step": 265
+    },
+    {
+      "epoch": 0.027047294635026756,
+      "grad_norm": 0.2177734375,
+      "learning_rate": 0.02,
+      "loss": 2.0122,
+      "step": 266
+    },
+    {
+      "epoch": 0.02714897619380505,
+      "grad_norm": 0.2392578125,
+      "learning_rate": 0.02,
+      "loss": 1.9253,
+      "step": 267
+    },
+    {
+      "epoch": 0.027250657752583347,
+      "grad_norm": 0.2578125,
+      "learning_rate": 0.02,
+      "loss": 2.0081,
+      "step": 268
+    },
+    {
+      "epoch": 0.027352339311361645,
+      "grad_norm": 0.2890625,
+      "learning_rate": 0.02,
+      "loss": 1.997,
+      "step": 269
+    },
+    {
+      "epoch": 0.02745402087013994,
+      "grad_norm": 0.279296875,
+      "learning_rate": 0.02,
+      "loss": 1.883,
+      "step": 270
+    },
+    {
+      "epoch": 0.027555702428918236,
+      "grad_norm": 0.291015625,
+      "learning_rate": 0.02,
+      "loss": 1.9693,
+      "step": 271
+    },
+    {
+      "epoch": 0.027657383987696533,
+      "grad_norm": 0.27734375,
+      "learning_rate": 0.02,
+      "loss": 1.9459,
+      "step": 272
+    },
+    {
+      "epoch": 0.027759065546474827,
+      "grad_norm": 0.2578125,
+      "learning_rate": 0.02,
+      "loss": 1.9048,
+      "step": 273
+    },
+    {
+      "epoch": 0.027860747105253124,
+      "grad_norm": 0.27734375,
+      "learning_rate": 0.02,
+      "loss": 1.9607,
+      "step": 274
+    },
+    {
+      "epoch": 0.02796242866403142,
+      "grad_norm": 0.26953125,
+      "learning_rate": 0.02,
+      "loss": 1.9306,
+      "step": 275
+    },
+    {
+      "epoch": 0.028064110222809715,
+      "grad_norm": 0.267578125,
+      "learning_rate": 0.02,
+      "loss": 1.9706,
+      "step": 276
+    },
+    {
+      "epoch": 0.028165791781588012,
+      "grad_norm": 0.2470703125,
+      "learning_rate": 0.02,
+      "loss": 1.9179,
+      "step": 277
+    },
+    {
+      "epoch": 0.02826747334036631,
+      "grad_norm": 0.24609375,
+      "learning_rate": 0.02,
+      "loss": 1.9612,
+      "step": 278
+    },
+    {
+      "epoch": 0.028369154899144603,
+      "grad_norm": 0.25390625,
+      "learning_rate": 0.02,
+      "loss": 1.9409,
+      "step": 279
+    },
+    {
+      "epoch": 0.0284708364579229,
+      "grad_norm": 0.2578125,
+      "learning_rate": 0.02,
+      "loss": 1.9908,
+      "step": 280
+    },
+    {
+      "epoch": 0.028572518016701198,
+      "grad_norm": 0.2353515625,
+      "learning_rate": 0.02,
+      "loss": 1.8841,
+      "step": 281
+    },
+    {
+      "epoch": 0.02867419957547949,
+      "grad_norm": 0.2373046875,
+      "learning_rate": 0.02,
+      "loss": 1.8824,
+      "step": 282
+    },
+    {
+      "epoch": 0.02877588113425779,
+      "grad_norm": 0.23046875,
+      "learning_rate": 0.02,
+      "loss": 1.9361,
+      "step": 283
+    },
+    {
+      "epoch": 0.028877562693036086,
+      "grad_norm": 0.25390625,
+      "learning_rate": 0.02,
+      "loss": 1.92,
+      "step": 284
+    },
+    {
+      "epoch": 0.02897924425181438,
+      "grad_norm": 0.240234375,
+      "learning_rate": 0.02,
+      "loss": 1.9297,
+      "step": 285
+    },
+    {
+      "epoch": 0.029080925810592677,
+      "grad_norm": 0.2451171875,
+      "learning_rate": 0.02,
+      "loss": 1.8725,
+      "step": 286
+    },
+    {
+      "epoch": 0.029182607369370974,
+      "grad_norm": 0.2294921875,
+      "learning_rate": 0.02,
+      "loss": 1.9288,
+      "step": 287
+    },
+    {
+      "epoch": 0.029284288928149268,
+      "grad_norm": 0.24609375,
+      "learning_rate": 0.02,
+      "loss": 1.9688,
+      "step": 288
+    },
+    {
+      "epoch": 0.029385970486927565,
+      "grad_norm": 0.232421875,
+      "learning_rate": 0.02,
+      "loss": 1.9505,
+      "step": 289
+    },
+    {
+      "epoch": 0.029487652045705862,
+      "grad_norm": 0.2333984375,
+      "learning_rate": 0.02,
+      "loss": 1.9474,
+      "step": 290
+    },
+    {
+      "epoch": 0.029589333604484156,
+      "grad_norm": 0.2265625,
+      "learning_rate": 0.02,
+      "loss": 1.8796,
+      "step": 291
+    },
+    {
+      "epoch": 0.029691015163262453,
+      "grad_norm": 0.263671875,
+      "learning_rate": 0.02,
+      "loss": 2.0346,
+      "step": 292
+    },
+    {
+      "epoch": 0.02979269672204075,
+      "grad_norm": 0.2578125,
+      "learning_rate": 0.02,
+      "loss": 1.9779,
+      "step": 293
+    },
+    {
+      "epoch": 0.029894378280819044,
+      "grad_norm": 0.25390625,
+      "learning_rate": 0.02,
+      "loss": 1.9338,
+      "step": 294
+    },
+    {
+      "epoch": 0.02999605983959734,
+      "grad_norm": 0.255859375,
+      "learning_rate": 0.02,
+      "loss": 1.9454,
+      "step": 295
+    },
+    {
+      "epoch": 0.03009774139837564,
+      "grad_norm": 0.234375,
+      "learning_rate": 0.02,
+      "loss": 1.9089,
+      "step": 296
+    },
+    {
+      "epoch": 0.030199422957153933,
+      "grad_norm": 0.2314453125,
+      "learning_rate": 0.02,
+      "loss": 1.9519,
+      "step": 297
+    },
+    {
+      "epoch": 0.03030110451593223,
+      "grad_norm": 0.240234375,
+      "learning_rate": 0.02,
+      "loss": 1.9262,
+      "step": 298
+    },
+    {
+      "epoch": 0.030402786074710524,
+      "grad_norm": 0.2275390625,
+      "learning_rate": 0.02,
+      "loss": 1.9226,
+      "step": 299
+    },
+    {
+      "epoch": 0.03050446763348882,
+      "grad_norm": 0.240234375,
+      "learning_rate": 0.02,
+      "loss": 1.8828,
+      "step": 300
+    },
+    {
+      "epoch": 0.030606149192267118,
+      "grad_norm": 0.2353515625,
+      "learning_rate": 0.02,
+      "loss": 1.9003,
+      "step": 301
+    },
+    {
+      "epoch": 0.030707830751045412,
+      "grad_norm": 0.232421875,
+      "learning_rate": 0.02,
+      "loss": 1.9077,
+      "step": 302
+    },
+    {
+      "epoch": 0.03080951230982371,
+      "grad_norm": 0.2353515625,
+      "learning_rate": 0.02,
+      "loss": 2.0387,
+      "step": 303
+    },
+    {
+      "epoch": 0.030911193868602006,
+      "grad_norm": 0.212890625,
+      "learning_rate": 0.02,
+      "loss": 1.8985,
+      "step": 304
+    },
+    {
+      "epoch": 0.0310128754273803,
+      "grad_norm": 0.220703125,
+      "learning_rate": 0.02,
+      "loss": 1.9086,
+      "step": 305
+    },
+    {
+      "epoch": 0.031114556986158597,
+      "grad_norm": 0.2216796875,
+      "learning_rate": 0.02,
+      "loss": 2.0058,
+      "step": 306
+    },
+    {
+      "epoch": 0.031216238544936895,
+      "grad_norm": 0.2353515625,
+      "learning_rate": 0.02,
+      "loss": 2.0162,
+      "step": 307
+    },
+    {
+      "epoch": 0.03131792010371519,
+      "grad_norm": 0.2255859375,
+      "learning_rate": 0.02,
+      "loss": 1.9123,
+      "step": 308
+    },
+    {
+      "epoch": 0.03141960166249349,
+      "grad_norm": 0.2265625,
+      "learning_rate": 0.02,
+      "loss": 1.9021,
+      "step": 309
+    },
+    {
+      "epoch": 0.03152128322127178,
+      "grad_norm": 0.22265625,
+      "learning_rate": 0.02,
+      "loss": 1.899,
+      "step": 310
+    },
+    {
+      "epoch": 0.03162296478005008,
+      "grad_norm": 0.2255859375,
+      "learning_rate": 0.02,
+      "loss": 1.8917,
+      "step": 311
+    },
+    {
+      "epoch": 0.03172464633882838,
+      "grad_norm": 0.232421875,
+      "learning_rate": 0.02,
+      "loss": 2.0207,
+      "step": 312
+    },
+    {
+      "epoch": 0.03182632789760667,
+      "grad_norm": 0.248046875,
+      "learning_rate": 0.02,
+      "loss": 1.9945,
+      "step": 313
+    },
+    {
+      "epoch": 0.031928009456384965,
+      "grad_norm": 0.2451171875,
+      "learning_rate": 0.02,
+      "loss": 1.9405,
+      "step": 314
+    },
+    {
+      "epoch": 0.032029691015163265,
+      "grad_norm": 0.2177734375,
+      "learning_rate": 0.02,
+      "loss": 1.879,
+      "step": 315
+    },
+    {
+      "epoch": 0.03213137257394156,
+      "grad_norm": 0.2216796875,
+      "learning_rate": 0.02,
+      "loss": 1.9376,
+      "step": 316
+    },
+    {
+      "epoch": 0.03223305413271985,
+      "grad_norm": 0.2080078125,
+      "learning_rate": 0.02,
+      "loss": 1.9242,
+      "step": 317
+    },
+    {
+      "epoch": 0.032334735691498154,
+      "grad_norm": 0.2080078125,
+      "learning_rate": 0.02,
+      "loss": 1.9547,
+      "step": 318
+    },
+    {
+      "epoch": 0.03243641725027645,
+      "grad_norm": 0.1982421875,
+      "learning_rate": 0.02,
+      "loss": 1.881,
+      "step": 319
+    },
+    {
+      "epoch": 0.03253809880905474,
+      "grad_norm": 0.21484375,
+      "learning_rate": 0.02,
+      "loss": 2.008,
+      "step": 320
+    },
+    {
+      "epoch": 0.03263978036783304,
+      "grad_norm": 0.205078125,
+      "learning_rate": 0.02,
+      "loss": 1.9909,
+      "step": 321
+    },
+    {
+      "epoch": 0.032741461926611336,
+      "grad_norm": 0.2080078125,
+      "learning_rate": 0.02,
+      "loss": 1.9824,
+      "step": 322
+    },
+    {
+      "epoch": 0.03284314348538963,
+      "grad_norm": 0.1904296875,
+      "learning_rate": 0.02,
+      "loss": 1.9,
+      "step": 323
+    },
+    {
+      "epoch": 0.03294482504416793,
+      "grad_norm": 0.205078125,
+      "learning_rate": 0.02,
+      "loss": 1.9193,
+      "step": 324
+    },
+    {
+      "epoch": 0.033046506602946224,
+      "grad_norm": 0.2138671875,
+      "learning_rate": 0.02,
+      "loss": 1.9114,
+      "step": 325
+    },
+    {
+      "epoch": 0.03314818816172452,
+      "grad_norm": 0.2060546875,
+      "learning_rate": 0.02,
+      "loss": 1.8746,
+      "step": 326
+    },
+    {
+      "epoch": 0.03324986972050282,
+      "grad_norm": 0.208984375,
+      "learning_rate": 0.02,
+      "loss": 1.9326,
+      "step": 327
+    },
+    {
+      "epoch": 0.03335155127928111,
+      "grad_norm": 0.19140625,
+      "learning_rate": 0.02,
+      "loss": 1.885,
+      "step": 328
+    },
+    {
+      "epoch": 0.033453232838059406,
+      "grad_norm": 0.2060546875,
+      "learning_rate": 0.02,
+      "loss": 1.863,
+      "step": 329
+    },
+    {
+      "epoch": 0.03355491439683771,
+      "grad_norm": 0.181640625,
+      "learning_rate": 0.02,
+      "loss": 1.9164,
+      "step": 330
+    },
+    {
+      "epoch": 0.033656595955616,
+      "grad_norm": 0.1982421875,
+      "learning_rate": 0.02,
+      "loss": 1.8594,
+      "step": 331
+    },
+    {
+      "epoch": 0.033758277514394294,
+      "grad_norm": 0.189453125,
+      "learning_rate": 0.02,
+      "loss": 1.8836,
+      "step": 332
+    },
+    {
+      "epoch": 0.033859959073172595,
+      "grad_norm": 0.2099609375,
+      "learning_rate": 0.02,
+      "loss": 1.9283,
+      "step": 333
+    },
+    {
+      "epoch": 0.03396164063195089,
+      "grad_norm": 0.19921875,
+      "learning_rate": 0.02,
+      "loss": 1.8882,
+      "step": 334
+    },
+    {
+      "epoch": 0.03406332219072918,
+      "grad_norm": 0.2138671875,
+      "learning_rate": 0.02,
+      "loss": 1.8417,
+      "step": 335
+    },
+    {
+      "epoch": 0.03416500374950748,
+      "grad_norm": 0.1943359375,
+      "learning_rate": 0.02,
+      "loss": 1.8984,
+      "step": 336
+    },
+    {
+      "epoch": 0.03426668530828578,
+      "grad_norm": 0.20703125,
+      "learning_rate": 0.02,
+      "loss": 1.867,
+      "step": 337
+    },
+    {
+      "epoch": 0.03436836686706407,
+      "grad_norm": 0.1962890625,
+      "learning_rate": 0.02,
+      "loss": 1.9648,
+      "step": 338
+    },
+    {
+      "epoch": 0.03447004842584237,
+      "grad_norm": 0.216796875,
+      "learning_rate": 0.02,
+      "loss": 1.8923,
+      "step": 339
+    },
+    {
+      "epoch": 0.034571729984620665,
+      "grad_norm": 0.220703125,
+      "learning_rate": 0.02,
+      "loss": 1.8713,
+      "step": 340
+    },
+    {
+      "epoch": 0.03467341154339896,
+      "grad_norm": 0.23046875,
+      "learning_rate": 0.02,
+      "loss": 1.9215,
+      "step": 341
+    },
+    {
+      "epoch": 0.03477509310217726,
+      "grad_norm": 0.2197265625,
+      "learning_rate": 0.02,
+      "loss": 1.9153,
+      "step": 342
+    },
+    {
+      "epoch": 0.03487677466095555,
+      "grad_norm": 0.2255859375,
+      "learning_rate": 0.02,
+      "loss": 1.8511,
+      "step": 343
+    },
+    {
+      "epoch": 0.03497845621973385,
+      "grad_norm": 0.224609375,
+      "learning_rate": 0.02,
+      "loss": 1.9254,
+      "step": 344
+    },
+    {
+      "epoch": 0.03508013777851215,
+      "grad_norm": 0.2412109375,
+      "learning_rate": 0.02,
+      "loss": 1.9455,
+      "step": 345
+    },
+    {
+      "epoch": 0.03518181933729044,
+      "grad_norm": 0.2294921875,
+      "learning_rate": 0.02,
+      "loss": 1.8929,
+      "step": 346
+    },
+    {
+      "epoch": 0.035283500896068735,
+      "grad_norm": 0.2353515625,
+      "learning_rate": 0.02,
+      "loss": 1.9297,
+      "step": 347
+    },
+    {
+      "epoch": 0.035385182454847036,
+      "grad_norm": 0.2216796875,
+      "learning_rate": 0.02,
+      "loss": 1.9091,
+      "step": 348
+    },
+    {
+      "epoch": 0.03548686401362533,
+      "grad_norm": 0.2255859375,
+      "learning_rate": 0.02,
+      "loss": 1.9613,
+      "step": 349
+    },
+    {
+      "epoch": 0.035588545572403624,
+      "grad_norm": 0.2373046875,
+      "learning_rate": 0.02,
+      "loss": 1.9152,
+      "step": 350
+    },
+    {
+      "epoch": 0.035690227131181924,
+      "grad_norm": 0.2236328125,
+      "learning_rate": 0.02,
+      "loss": 1.9256,
+      "step": 351
+    },
+    {
+      "epoch": 0.03579190868996022,
+      "grad_norm": 0.21875,
+      "learning_rate": 0.02,
+      "loss": 1.8975,
+      "step": 352
+    },
+    {
+      "epoch": 0.03589359024873851,
+      "grad_norm": 0.2373046875,
+      "learning_rate": 0.02,
+      "loss": 1.9908,
+      "step": 353
+    },
+    {
+      "epoch": 0.03599527180751681,
+      "grad_norm": 0.2060546875,
+      "learning_rate": 0.02,
+      "loss": 1.8537,
+      "step": 354
+    },
+    {
+      "epoch": 0.036096953366295106,
+      "grad_norm": 0.2099609375,
+      "learning_rate": 0.02,
+      "loss": 1.9156,
+      "step": 355
+    },
+    {
+      "epoch": 0.0361986349250734,
+      "grad_norm": 0.205078125,
+      "learning_rate": 0.02,
+      "loss": 1.91,
+      "step": 356
+    },
+    {
+      "epoch": 0.0363003164838517,
+      "grad_norm": 0.2041015625,
+      "learning_rate": 0.02,
+      "loss": 1.9222,
+      "step": 357
+    },
+    {
+      "epoch": 0.036401998042629995,
+      "grad_norm": 0.2021484375,
+      "learning_rate": 0.02,
+      "loss": 1.8976,
+      "step": 358
+    },
+    {
+      "epoch": 0.03650367960140829,
+      "grad_norm": 0.2060546875,
+      "learning_rate": 0.02,
+      "loss": 1.898,
+      "step": 359
+    },
+    {
+      "epoch": 0.03660536116018658,
+      "grad_norm": 0.2001953125,
+      "learning_rate": 0.02,
+      "loss": 1.9171,
+      "step": 360
+    },
+    {
+      "epoch": 0.03670704271896488,
+      "grad_norm": 0.2060546875,
+      "learning_rate": 0.02,
+      "loss": 1.9262,
+      "step": 361
+    },
+    {
+      "epoch": 0.03680872427774318,
+      "grad_norm": 0.236328125,
+      "learning_rate": 0.02,
+      "loss": 1.9066,
+      "step": 362
+    },
+    {
+      "epoch": 0.03691040583652147,
+      "grad_norm": 0.224609375,
+      "learning_rate": 0.02,
+      "loss": 1.8847,
+      "step": 363
+    },
+    {
+      "epoch": 0.03701208739529977,
+      "grad_norm": 0.2275390625,
+      "learning_rate": 0.02,
+      "loss": 1.9028,
+      "step": 364
+    },
+    {
+      "epoch": 0.037113768954078065,
+      "grad_norm": 0.2236328125,
+      "learning_rate": 0.02,
+      "loss": 1.8746,
+      "step": 365
+    },
+    {
+      "epoch": 0.03721545051285636,
+      "grad_norm": 0.2314453125,
+      "learning_rate": 0.02,
+      "loss": 1.8823,
+      "step": 366
+    },
+    {
+      "epoch": 0.03731713207163466,
+      "grad_norm": 0.2373046875,
+      "learning_rate": 0.02,
+      "loss": 1.9134,
+      "step": 367
+    },
+    {
+      "epoch": 0.03741881363041295,
+      "grad_norm": 0.2314453125,
+      "learning_rate": 0.02,
+      "loss": 1.9073,
+      "step": 368
+    },
+    {
+      "epoch": 0.03752049518919125,
+      "grad_norm": 0.23046875,
+      "learning_rate": 0.02,
+      "loss": 1.8681,
+      "step": 369
+    },
+    {
+      "epoch": 0.03762217674796955,
+      "grad_norm": 0.25,
+      "learning_rate": 0.02,
+      "loss": 1.9754,
+      "step": 370
+    },
+    {
+      "epoch": 0.03772385830674784,
+      "grad_norm": 0.2490234375,
+      "learning_rate": 0.02,
+      "loss": 1.9868,
+      "step": 371
+    },
+    {
+      "epoch": 0.037825539865526135,
+      "grad_norm": 0.2392578125,
+      "learning_rate": 0.02,
+      "loss": 1.9113,
+      "step": 372
+    },
+    {
+      "epoch": 0.037927221424304436,
+      "grad_norm": 0.22265625,
+      "learning_rate": 0.02,
+      "loss": 1.8791,
+      "step": 373
+    },
+    {
+      "epoch": 0.03802890298308273,
+      "grad_norm": 0.2255859375,
+      "learning_rate": 0.02,
+      "loss": 1.9005,
+      "step": 374
+    },
+    {
+      "epoch": 0.03813058454186102,
+      "grad_norm": 0.2236328125,
+      "learning_rate": 0.02,
+      "loss": 1.8517,
+      "step": 375
+    },
+    {
+      "epoch": 0.038232266100639324,
+      "grad_norm": 0.2255859375,
+      "learning_rate": 0.02,
+      "loss": 1.8787,
+      "step": 376
+    },
+    {
+      "epoch": 0.03833394765941762,
+      "grad_norm": 0.2353515625,
+      "learning_rate": 0.02,
+      "loss": 1.8634,
+      "step": 377
+    },
+    {
+      "epoch": 0.03843562921819591,
+      "grad_norm": 0.234375,
+      "learning_rate": 0.02,
+      "loss": 1.8804,
+      "step": 378
+    },
+    {
+      "epoch": 0.03853731077697421,
+      "grad_norm": 0.228515625,
+      "learning_rate": 0.02,
+      "loss": 1.8802,
+      "step": 379
+    },
+    {
+      "epoch": 0.038638992335752506,
+      "grad_norm": 0.224609375,
+      "learning_rate": 0.02,
+      "loss": 1.8983,
+      "step": 380
+    },
+    {
+      "epoch": 0.0387406738945308,
+      "grad_norm": 0.232421875,
+      "learning_rate": 0.02,
+      "loss": 1.8915,
+      "step": 381
+    },
+    {
+      "epoch": 0.0388423554533091,
+      "grad_norm": 0.2470703125,
+      "learning_rate": 0.02,
+      "loss": 1.9942,
+      "step": 382
+    },
+    {
+      "epoch": 0.038944037012087394,
+      "grad_norm": 0.220703125,
+      "learning_rate": 0.02,
+      "loss": 1.852,
+      "step": 383
+    },
+    {
+      "epoch": 0.03904571857086569,
+      "grad_norm": 0.2119140625,
+      "learning_rate": 0.02,
+      "loss": 1.8505,
+      "step": 384
+    },
+    {
+      "epoch": 0.03914740012964399,
+      "grad_norm": 0.216796875,
+      "learning_rate": 0.02,
+      "loss": 1.8551,
+      "step": 385
+    },
+    {
+      "epoch": 0.03924908168842228,
+      "grad_norm": 0.2255859375,
+      "learning_rate": 0.02,
+      "loss": 1.864,
+      "step": 386
+    },
+    {
+      "epoch": 0.039350763247200576,
+      "grad_norm": 0.2255859375,
+      "learning_rate": 0.02,
+      "loss": 1.8806,
+      "step": 387
+    },
+    {
+      "epoch": 0.03945244480597888,
+      "grad_norm": 0.2177734375,
+      "learning_rate": 0.02,
+      "loss": 1.8932,
+      "step": 388
+    },
+    {
+      "epoch": 0.03955412636475717,
+      "grad_norm": 0.220703125,
+      "learning_rate": 0.02,
+      "loss": 1.9155,
+      "step": 389
+    },
+    {
+      "epoch": 0.039655807923535465,
+      "grad_norm": 0.201171875,
+      "learning_rate": 0.02,
+      "loss": 1.9133,
+      "step": 390
+    },
+    {
+      "epoch": 0.039757489482313765,
+      "grad_norm": 0.203125,
+      "learning_rate": 0.02,
+      "loss": 1.9024,
+      "step": 391
+    },
+    {
+      "epoch": 0.03985917104109206,
+      "grad_norm": 0.2080078125,
+      "learning_rate": 0.02,
+      "loss": 1.9514,
+      "step": 392
+    },
+    {
+      "epoch": 0.03996085259987035,
+      "grad_norm": 0.1923828125,
+      "learning_rate": 0.02,
+      "loss": 1.8982,
+      "step": 393
+    },
+    {
+      "epoch": 0.040062534158648654,
+      "grad_norm": 0.19921875,
+      "learning_rate": 0.02,
+      "loss": 1.914,
+      "step": 394
+    },
+    {
+      "epoch": 0.04016421571742695,
+      "grad_norm": 0.21484375,
+      "learning_rate": 0.02,
+      "loss": 1.9033,
+      "step": 395
+    },
+    {
+      "epoch": 0.04026589727620524,
+      "grad_norm": 0.2119140625,
+      "learning_rate": 0.02,
+      "loss": 1.8651,
+      "step": 396
+    },
+    {
+      "epoch": 0.04036757883498354,
+      "grad_norm": 0.2314453125,
+      "learning_rate": 0.02,
+      "loss": 1.9016,
+      "step": 397
+    },
+    {
+      "epoch": 0.040469260393761836,
+      "grad_norm": 0.224609375,
+      "learning_rate": 0.02,
+      "loss": 1.9074,
+      "step": 398
+    },
+    {
+      "epoch": 0.04057094195254013,
+      "grad_norm": 0.228515625,
+      "learning_rate": 0.02,
+      "loss": 1.9165,
+      "step": 399
+    },
+    {
+      "epoch": 0.04067262351131843,
+      "grad_norm": 0.2109375,
+      "learning_rate": 0.02,
+      "loss": 1.8182,
+      "step": 400
+    },
+    {
+      "epoch": 0.040774305070096724,
+      "grad_norm": 0.2119140625,
+      "learning_rate": 0.02,
+      "loss": 1.8673,
+      "step": 401
+    },
+    {
+      "epoch": 0.04087598662887502,
+      "grad_norm": 0.2177734375,
+      "learning_rate": 0.02,
+      "loss": 1.8879,
+      "step": 402
+    },
+    {
+      "epoch": 0.04097766818765332,
+      "grad_norm": 0.2119140625,
+      "learning_rate": 0.02,
+      "loss": 1.8446,
+      "step": 403
+    },
+    {
+      "epoch": 0.04107934974643161,
+      "grad_norm": 0.2080078125,
+      "learning_rate": 0.02,
+      "loss": 1.8926,
+      "step": 404
+    },
+    {
+      "epoch": 0.041181031305209906,
+      "grad_norm": 0.1982421875,
+      "learning_rate": 0.02,
+      "loss": 1.8921,
+      "step": 405
+    },
+    {
+      "epoch": 0.04128271286398821,
+      "grad_norm": 0.205078125,
+      "learning_rate": 0.02,
+      "loss": 1.8637,
+      "step": 406
+    },
+    {
+      "epoch": 0.0413843944227665,
+      "grad_norm": 0.2109375,
+      "learning_rate": 0.02,
+      "loss": 1.937,
+      "step": 407
+    },
+    {
+      "epoch": 0.041486075981544794,
+      "grad_norm": 0.2060546875,
+      "learning_rate": 0.02,
+      "loss": 1.9151,
+      "step": 408
+    },
+    {
+      "epoch": 0.041587757540323095,
+      "grad_norm": 0.2060546875,
+      "learning_rate": 0.02,
+      "loss": 1.8858,
+      "step": 409
+    },
+    {
+      "epoch": 0.04168943909910139,
+      "grad_norm": 0.205078125,
+      "learning_rate": 0.02,
+      "loss": 1.8709,
+      "step": 410
+    },
+    {
+      "epoch": 0.04179112065787968,
+      "grad_norm": 0.20703125,
+      "learning_rate": 0.02,
+      "loss": 1.8554,
+      "step": 411
+    },
+    {
+      "epoch": 0.04189280221665798,
+      "grad_norm": 0.224609375,
+      "learning_rate": 0.02,
+      "loss": 2.001,
+      "step": 412
+    },
+    {
+      "epoch": 0.04199448377543628,
+      "grad_norm": 0.2080078125,
+      "learning_rate": 0.02,
+      "loss": 1.8804,
+      "step": 413
+    },
+    {
+      "epoch": 0.04209616533421457,
+      "grad_norm": 0.19921875,
+      "learning_rate": 0.02,
+      "loss": 1.7878,
+      "step": 414
+    },
+    {
+      "epoch": 0.04219784689299287,
+      "grad_norm": 0.2001953125,
+      "learning_rate": 0.02,
+      "loss": 1.9061,
+      "step": 415
+    },
+    {
+      "epoch": 0.042299528451771165,
+      "grad_norm": 0.224609375,
+      "learning_rate": 0.02,
+      "loss": 1.9463,
+      "step": 416
+    },
+    {
+      "epoch": 0.04240121001054946,
+      "grad_norm": 0.21484375,
+      "learning_rate": 0.02,
+      "loss": 1.8504,
+      "step": 417
+    },
+    {
+      "epoch": 0.04250289156932776,
+      "grad_norm": 0.208984375,
+      "learning_rate": 0.02,
+      "loss": 1.8812,
+      "step": 418
+    },
+    {
+      "epoch": 0.04260457312810605,
+      "grad_norm": 0.21484375,
+      "learning_rate": 0.02,
+      "loss": 1.8442,
+      "step": 419
+    },
+    {
+      "epoch": 0.04270625468688435,
+      "grad_norm": 0.2080078125,
+      "learning_rate": 0.02,
+      "loss": 1.9442,
+      "step": 420
+    },
+    {
+      "epoch": 0.04280793624566265,
+      "grad_norm": 0.2177734375,
+      "learning_rate": 0.02,
+      "loss": 1.9365,
+      "step": 421
+    },
+    {
+      "epoch": 0.04290961780444094,
+      "grad_norm": 0.2119140625,
+      "learning_rate": 0.02,
+      "loss": 1.8515,
+      "step": 422
+    },
+    {
+      "epoch": 0.043011299363219235,
+      "grad_norm": 0.220703125,
+      "learning_rate": 0.02,
+      "loss": 1.887,
+      "step": 423
+    },
+    {
+      "epoch": 0.043112980921997536,
+      "grad_norm": 0.2255859375,
+      "learning_rate": 0.02,
+      "loss": 1.934,
+      "step": 424
+    },
+    {
+      "epoch": 0.04321466248077583,
+      "grad_norm": 0.2353515625,
+      "learning_rate": 0.02,
+      "loss": 1.8694,
+      "step": 425
+    },
+    {
+      "epoch": 0.043316344039554124,
+      "grad_norm": 0.2353515625,
+      "learning_rate": 0.02,
+      "loss": 1.8941,
+      "step": 426
+    },
+    {
+      "epoch": 0.043418025598332424,
+      "grad_norm": 0.234375,
+      "learning_rate": 0.02,
+      "loss": 1.8909,
+      "step": 427
+    },
+    {
+      "epoch": 0.04351970715711072,
+      "grad_norm": 0.2216796875,
+      "learning_rate": 0.02,
+      "loss": 1.9057,
+      "step": 428
+    },
+    {
+      "epoch": 0.04362138871588901,
+      "grad_norm": 0.2255859375,
+      "learning_rate": 0.02,
+      "loss": 1.8465,
+      "step": 429
+    },
+    {
+      "epoch": 0.04372307027466731,
+      "grad_norm": 0.2109375,
+      "learning_rate": 0.02,
+      "loss": 1.8173,
+      "step": 430
+    },
+    {
+      "epoch": 0.043824751833445606,
+      "grad_norm": 0.22265625,
+      "learning_rate": 0.02,
+      "loss": 1.9444,
+      "step": 431
+    },
+    {
+      "epoch": 0.0439264333922239,
+      "grad_norm": 0.2041015625,
+      "learning_rate": 0.02,
+      "loss": 1.8878,
+      "step": 432
+    },
+    {
+      "epoch": 0.0440281149510022,
+      "grad_norm": 0.2021484375,
+      "learning_rate": 0.02,
+      "loss": 1.9324,
+      "step": 433
+    },
+    {
+      "epoch": 0.044129796509780495,
+      "grad_norm": 0.216796875,
+      "learning_rate": 0.02,
+      "loss": 1.8873,
+      "step": 434
+    },
+    {
+      "epoch": 0.04423147806855879,
+      "grad_norm": 0.208984375,
+      "learning_rate": 0.02,
+      "loss": 1.9502,
+      "step": 435
+    },
+    {
+      "epoch": 0.04433315962733709,
+      "grad_norm": 0.2060546875,
+      "learning_rate": 0.02,
+      "loss": 1.9115,
+      "step": 436
+    },
+    {
+      "epoch": 0.04443484118611538,
+      "grad_norm": 0.197265625,
+      "learning_rate": 0.02,
+      "loss": 1.845,
+      "step": 437
+    },
+    {
+      "epoch": 0.044536522744893677,
+      "grad_norm": 0.2001953125,
+      "learning_rate": 0.02,
+      "loss": 1.956,
+      "step": 438
+    },
+    {
+      "epoch": 0.04463820430367198,
+      "grad_norm": 0.1962890625,
+      "learning_rate": 0.02,
+      "loss": 1.8409,
+      "step": 439
+    },
+    {
+      "epoch": 0.04473988586245027,
+      "grad_norm": 0.205078125,
+      "learning_rate": 0.02,
+      "loss": 1.9024,
+      "step": 440
+    },
+    {
+      "epoch": 0.044841567421228565,
+      "grad_norm": 0.220703125,
+      "learning_rate": 0.02,
+      "loss": 1.8784,
+      "step": 441
+    },
+    {
+      "epoch": 0.044943248980006865,
+      "grad_norm": 0.23046875,
+      "learning_rate": 0.02,
+      "loss": 1.8811,
+      "step": 442
+    },
+    {
+      "epoch": 0.04504493053878516,
+      "grad_norm": 0.2431640625,
+      "learning_rate": 0.02,
+      "loss": 1.9148,
+      "step": 443
+    },
+    {
+      "epoch": 0.04514661209756345,
+      "grad_norm": 0.234375,
+      "learning_rate": 0.02,
+      "loss": 1.8644,
+      "step": 444
+    },
+    {
+      "epoch": 0.045248293656341754,
+      "grad_norm": 0.2255859375,
+      "learning_rate": 0.02,
+      "loss": 1.9102,
+      "step": 445
+    },
+    {
+      "epoch": 0.04534997521512005,
+      "grad_norm": 0.228515625,
+      "learning_rate": 0.02,
+      "loss": 1.8952,
+      "step": 446
+    },
+    {
+      "epoch": 0.04545165677389834,
+      "grad_norm": 0.25,
+      "learning_rate": 0.02,
+      "loss": 1.9325,
+      "step": 447
+    },
+    {
+      "epoch": 0.04555333833267664,
+      "grad_norm": 0.21484375,
+      "learning_rate": 0.02,
+      "loss": 1.868,
+      "step": 448
+    },
+    {
+      "epoch": 0.045655019891454936,
+      "grad_norm": 0.2138671875,
+      "learning_rate": 0.02,
+      "loss": 1.9058,
+      "step": 449
+    },
+    {
+      "epoch": 0.04575670145023323,
+      "grad_norm": 0.228515625,
+      "learning_rate": 0.02,
+      "loss": 1.9316,
+      "step": 450
+    },
+    {
+      "epoch": 0.04585838300901153,
+      "grad_norm": 0.234375,
+      "learning_rate": 0.02,
+      "loss": 1.8801,
+      "step": 451
+    },
+    {
+      "epoch": 0.045960064567789824,
+      "grad_norm": 0.2177734375,
+      "learning_rate": 0.02,
+      "loss": 1.8815,
+      "step": 452
+    },
+    {
+      "epoch": 0.04606174612656812,
+      "grad_norm": 0.1982421875,
+      "learning_rate": 0.02,
+      "loss": 1.8847,
+      "step": 453
+    },
+    {
+      "epoch": 0.04616342768534642,
+      "grad_norm": 0.205078125,
+      "learning_rate": 0.02,
+      "loss": 1.9624,
+      "step": 454
+    },
+    {
+      "epoch": 0.04626510924412471,
+      "grad_norm": 0.193359375,
+      "learning_rate": 0.02,
+      "loss": 1.9427,
+      "step": 455
+    },
+    {
+      "epoch": 0.046366790802903006,
+      "grad_norm": 0.19921875,
+      "learning_rate": 0.02,
+      "loss": 1.8908,
+      "step": 456
+    },
+    {
+      "epoch": 0.04646847236168131,
+      "grad_norm": 0.2138671875,
+      "learning_rate": 0.02,
+      "loss": 1.9354,
+      "step": 457
+    },
+    {
+      "epoch": 0.0465701539204596,
+      "grad_norm": 0.2041015625,
+      "learning_rate": 0.02,
+      "loss": 1.8774,
+      "step": 458
+    },
+    {
+      "epoch": 0.046671835479237894,
+      "grad_norm": 0.1826171875,
+      "learning_rate": 0.02,
+      "loss": 1.8669,
+      "step": 459
+    },
+    {
+      "epoch": 0.046773517038016195,
+      "grad_norm": 0.2060546875,
+      "learning_rate": 0.02,
+      "loss": 1.9547,
+      "step": 460
+    },
+    {
+      "epoch": 0.04687519859679449,
+      "grad_norm": 0.181640625,
+      "learning_rate": 0.02,
+      "loss": 1.8196,
+      "step": 461
+    },
+    {
+      "epoch": 0.04697688015557278,
+      "grad_norm": 0.201171875,
+      "learning_rate": 0.02,
+      "loss": 1.9181,
+      "step": 462
+    },
+    {
+      "epoch": 0.04707856171435108,
+      "grad_norm": 0.1982421875,
+      "learning_rate": 0.02,
+      "loss": 1.8848,
+      "step": 463
+    },
+    {
+      "epoch": 0.04718024327312938,
+      "grad_norm": 0.2119140625,
+      "learning_rate": 0.02,
+      "loss": 1.8992,
+      "step": 464
+    },
+    {
+      "epoch": 0.04728192483190767,
+      "grad_norm": 0.2138671875,
+      "learning_rate": 0.02,
+      "loss": 1.8566,
+      "step": 465
+    },
+    {
+      "epoch": 0.04738360639068597,
+      "grad_norm": 0.20703125,
+      "learning_rate": 0.02,
+      "loss": 1.8819,
+      "step": 466
+    },
+    {
+      "epoch": 0.047485287949464265,
+      "grad_norm": 0.1884765625,
+      "learning_rate": 0.02,
+      "loss": 1.8654,
+      "step": 467
+    },
+    {
+      "epoch": 0.04758696950824256,
+      "grad_norm": 0.1962890625,
+      "learning_rate": 0.02,
+      "loss": 1.9501,
+      "step": 468
+    },
+    {
+      "epoch": 0.04768865106702086,
+      "grad_norm": 0.1904296875,
+      "learning_rate": 0.02,
+      "loss": 1.885,
+      "step": 469
+    },
+    {
+      "epoch": 0.04779033262579915,
+      "grad_norm": 0.1943359375,
+      "learning_rate": 0.02,
+      "loss": 1.8242,
+      "step": 470
+    },
+    {
+      "epoch": 0.04789201418457745,
+      "grad_norm": 0.19921875,
+      "learning_rate": 0.02,
+      "loss": 1.9175,
+      "step": 471
+    },
+    {
+      "epoch": 0.04799369574335575,
+      "grad_norm": 0.205078125,
+      "learning_rate": 0.02,
+      "loss": 1.9177,
+      "step": 472
+    },
+    {
+      "epoch": 0.04809537730213404,
+      "grad_norm": 0.1806640625,
+      "learning_rate": 0.02,
+      "loss": 1.8493,
+      "step": 473
+    },
+    {
+      "epoch": 0.048197058860912335,
+      "grad_norm": 0.1748046875,
+      "learning_rate": 0.02,
+      "loss": 1.8796,
+      "step": 474
+    },
+    {
+      "epoch": 0.048298740419690636,
+      "grad_norm": 0.171875,
+      "learning_rate": 0.02,
+      "loss": 1.8145,
+      "step": 475
+    },
+    {
+      "epoch": 0.04840042197846893,
+      "grad_norm": 0.1748046875,
+      "learning_rate": 0.02,
+      "loss": 1.8888,
+      "step": 476
+    },
+    {
+      "epoch": 0.048502103537247224,
+      "grad_norm": 0.16796875,
+      "learning_rate": 0.02,
+      "loss": 1.9088,
+      "step": 477
+    },
+    {
+      "epoch": 0.048603785096025524,
+      "grad_norm": 0.185546875,
+      "learning_rate": 0.02,
+      "loss": 1.9166,
+      "step": 478
+    },
+    {
+      "epoch": 0.04870546665480382,
+      "grad_norm": 0.1884765625,
+      "learning_rate": 0.02,
+      "loss": 1.8738,
+      "step": 479
+    },
+    {
+      "epoch": 0.04880714821358211,
+      "grad_norm": 0.19921875,
+      "learning_rate": 0.02,
+      "loss": 1.8741,
+      "step": 480
+    },
+    {
+      "epoch": 0.04890882977236041,
+      "grad_norm": 0.21484375,
+      "learning_rate": 0.02,
+      "loss": 1.9206,
+      "step": 481
+    },
+    {
+      "epoch": 0.049010511331138706,
+      "grad_norm": 0.2177734375,
+      "learning_rate": 0.02,
+      "loss": 1.898,
+      "step": 482
+    },
+    {
+      "epoch": 0.049112192889917,
+      "grad_norm": 0.203125,
+      "learning_rate": 0.02,
+      "loss": 1.9377,
+      "step": 483
+    },
+    {
+      "epoch": 0.0492138744486953,
+      "grad_norm": 0.2236328125,
+      "learning_rate": 0.02,
+      "loss": 1.8967,
+      "step": 484
+    },
+    {
+      "epoch": 0.049315556007473595,
+      "grad_norm": 0.21484375,
+      "learning_rate": 0.02,
+      "loss": 1.8982,
+      "step": 485
+    },
+    {
+      "epoch": 0.04941723756625189,
+      "grad_norm": 0.22265625,
+      "learning_rate": 0.02,
+      "loss": 1.8574,
+      "step": 486
+    },
+    {
+      "epoch": 0.04951891912503019,
+      "grad_norm": 0.224609375,
+      "learning_rate": 0.02,
+      "loss": 1.85,
+      "step": 487
+    },
+    {
+      "epoch": 0.04962060068380848,
+      "grad_norm": 0.21875,
+      "learning_rate": 0.02,
+      "loss": 1.8345,
+      "step": 488
+    },
+    {
+      "epoch": 0.04972228224258678,
+      "grad_norm": 0.2060546875,
+      "learning_rate": 0.02,
+      "loss": 1.8557,
+      "step": 489
+    },
+    {
+      "epoch": 0.04982396380136508,
+      "grad_norm": 0.220703125,
+      "learning_rate": 0.02,
+      "loss": 1.8801,
+      "step": 490
+    },
+    {
+      "epoch": 0.04992564536014337,
+      "grad_norm": 0.220703125,
+      "learning_rate": 0.02,
+      "loss": 1.8832,
+      "step": 491
+    },
+    {
+      "epoch": 0.050027326918921665,
+      "grad_norm": 0.208984375,
+      "learning_rate": 0.02,
+      "loss": 1.9192,
+      "step": 492
+    },
+    {
+      "epoch": 0.050129008477699966,
+      "grad_norm": 0.216796875,
+      "learning_rate": 0.02,
+      "loss": 1.9003,
+      "step": 493
+    },
+    {
+      "epoch": 0.05023069003647826,
+      "grad_norm": 0.2158203125,
+      "learning_rate": 0.02,
+      "loss": 1.8599,
+      "step": 494
+    },
+    {
+      "epoch": 0.05033237159525655,
+      "grad_norm": 0.208984375,
+      "learning_rate": 0.02,
+      "loss": 1.8821,
+      "step": 495
+    },
+    {
+      "epoch": 0.050434053154034854,
+      "grad_norm": 0.201171875,
+      "learning_rate": 0.02,
+      "loss": 1.8741,
+      "step": 496
+    },
+    {
+      "epoch": 0.05053573471281315,
+      "grad_norm": 0.2216796875,
+      "learning_rate": 0.02,
+      "loss": 1.8974,
+      "step": 497
+    },
+    {
+      "epoch": 0.05063741627159144,
+      "grad_norm": 0.2001953125,
+      "learning_rate": 0.02,
+      "loss": 1.8582,
+      "step": 498
+    },
+    {
+      "epoch": 0.05073909783036974,
+      "grad_norm": 0.20703125,
+      "learning_rate": 0.02,
+      "loss": 1.8548,
+      "step": 499
+    },
+    {
+      "epoch": 0.050840779389148036,
+      "grad_norm": 0.2099609375,
+      "learning_rate": 0.02,
+      "loss": 1.8275,
+      "step": 500
+    },
+    {
+      "epoch": 0.05094246094792633,
+      "grad_norm": 0.20703125,
+      "learning_rate": 0.02,
+      "loss": 1.8852,
+      "step": 501
+    },
+    {
+      "epoch": 0.05104414250670463,
+      "grad_norm": 0.1953125,
+      "learning_rate": 0.02,
+      "loss": 1.8189,
+      "step": 502
+    },
+    {
+      "epoch": 0.051145824065482924,
+      "grad_norm": 0.2001953125,
+      "learning_rate": 0.02,
+      "loss": 1.8342,
+      "step": 503
+    },
+    {
+      "epoch": 0.05124750562426122,
+      "grad_norm": 0.2021484375,
+      "learning_rate": 0.02,
+      "loss": 1.8366,
+      "step": 504
+    },
+    {
+      "epoch": 0.05134918718303952,
+      "grad_norm": 0.2138671875,
+      "learning_rate": 0.02,
+      "loss": 1.8983,
+      "step": 505
+    },
+    {
+      "epoch": 0.05145086874181781,
+      "grad_norm": 0.19921875,
+      "learning_rate": 0.02,
+      "loss": 1.8734,
+      "step": 506
+    },
+    {
+      "epoch": 0.051552550300596106,
+      "grad_norm": 0.2001953125,
+      "learning_rate": 0.02,
+      "loss": 1.8972,
+      "step": 507
+    },
+    {
+      "epoch": 0.05165423185937441,
+      "grad_norm": 0.205078125,
+      "learning_rate": 0.02,
+      "loss": 1.9273,
+      "step": 508
+    },
+    {
+      "epoch": 0.0517559134181527,
+      "grad_norm": 0.201171875,
+      "learning_rate": 0.02,
+      "loss": 1.8707,
+      "step": 509
+    },
+    {
+      "epoch": 0.051857594976930994,
+      "grad_norm": 0.208984375,
+      "learning_rate": 0.02,
+      "loss": 1.96,
+      "step": 510
+    },
+    {
+      "epoch": 0.051959276535709295,
+      "grad_norm": 0.2080078125,
+      "learning_rate": 0.02,
+      "loss": 1.8575,
+      "step": 511
+    },
+    {
+      "epoch": 0.05206095809448759,
+      "grad_norm": 0.205078125,
+      "learning_rate": 0.02,
+      "loss": 1.8271,
+      "step": 512
+    },
+    {
+      "epoch": 0.05216263965326588,
+      "grad_norm": 0.20703125,
+      "learning_rate": 0.02,
+      "loss": 1.8723,
+      "step": 513
+    },
+    {
+      "epoch": 0.05226432121204418,
+      "grad_norm": 0.205078125,
+      "learning_rate": 0.02,
+      "loss": 1.8493,
+      "step": 514
+    },
+    {
+      "epoch": 0.05236600277082248,
+      "grad_norm": 0.2138671875,
+      "learning_rate": 0.02,
+      "loss": 1.8842,
+      "step": 515
+    },
+    {
+      "epoch": 0.05246768432960077,
+      "grad_norm": 0.21875,
+      "learning_rate": 0.02,
+      "loss": 1.8515,
+      "step": 516
+    },
+    {
+      "epoch": 0.05256936588837907,
+      "grad_norm": 0.2431640625,
+      "learning_rate": 0.02,
+      "loss": 1.8979,
+      "step": 517
+    },
+    {
+      "epoch": 0.052671047447157365,
+      "grad_norm": 0.21484375,
+      "learning_rate": 0.02,
+      "loss": 1.9194,
+      "step": 518
+    },
+    {
+      "epoch": 0.05277272900593566,
+      "grad_norm": 0.2197265625,
+      "learning_rate": 0.02,
+      "loss": 1.9144,
+      "step": 519
+    },
+    {
+      "epoch": 0.05287441056471396,
+      "grad_norm": 0.21484375,
+      "learning_rate": 0.02,
+      "loss": 1.8921,
+      "step": 520
+    },
+    {
+      "epoch": 0.052976092123492254,
+      "grad_norm": 0.19921875,
+      "learning_rate": 0.02,
+      "loss": 1.8319,
+      "step": 521
+    },
+    {
+      "epoch": 0.05307777368227055,
+      "grad_norm": 0.197265625,
+      "learning_rate": 0.02,
+      "loss": 1.8916,
+      "step": 522
+    },
+    {
+      "epoch": 0.05317945524104885,
+      "grad_norm": 0.201171875,
+      "learning_rate": 0.02,
+      "loss": 1.8827,
+      "step": 523
+    },
+    {
+      "epoch": 0.05328113679982714,
+      "grad_norm": 0.203125,
+      "learning_rate": 0.02,
+      "loss": 1.8601,
+      "step": 524
+    },
+    {
+      "epoch": 0.053382818358605436,
+      "grad_norm": 0.203125,
+      "learning_rate": 0.02,
+      "loss": 1.8542,
+      "step": 525
+    },
+    {
+      "epoch": 0.053484499917383736,
+      "grad_norm": 0.228515625,
+      "learning_rate": 0.02,
+      "loss": 1.8244,
+      "step": 526
+    },
+    {
+      "epoch": 0.05358618147616203,
+      "grad_norm": 0.205078125,
+      "learning_rate": 0.02,
+      "loss": 1.8316,
+      "step": 527
+    },
+    {
+      "epoch": 0.053687863034940324,
+      "grad_norm": 0.1865234375,
+      "learning_rate": 0.02,
+      "loss": 1.8592,
+      "step": 528
+    },
+    {
+      "epoch": 0.053789544593718625,
+      "grad_norm": 0.193359375,
+      "learning_rate": 0.02,
+      "loss": 1.8283,
+      "step": 529
+    },
+    {
+      "epoch": 0.05389122615249692,
+      "grad_norm": 0.1962890625,
+      "learning_rate": 0.02,
+      "loss": 1.926,
+      "step": 530
+    },
+    {
+      "epoch": 0.05399290771127521,
+      "grad_norm": 0.205078125,
+      "learning_rate": 0.02,
+      "loss": 1.8904,
+      "step": 531
+    },
+    {
+      "epoch": 0.05409458927005351,
+      "grad_norm": 0.201171875,
+      "learning_rate": 0.02,
+      "loss": 1.9216,
+      "step": 532
+    },
+    {
+      "epoch": 0.054196270828831807,
+      "grad_norm": 0.197265625,
+      "learning_rate": 0.02,
+      "loss": 1.8909,
+      "step": 533
+    },
+    {
+      "epoch": 0.0542979523876101,
+      "grad_norm": 0.1787109375,
+      "learning_rate": 0.02,
+      "loss": 1.8765,
+      "step": 534
+    },
+    {
+      "epoch": 0.0543996339463884,
+      "grad_norm": 0.1806640625,
+      "learning_rate": 0.02,
+      "loss": 1.8119,
+      "step": 535
+    },
+    {
+      "epoch": 0.054501315505166695,
+      "grad_norm": 0.181640625,
+      "learning_rate": 0.02,
+      "loss": 1.8851,
+      "step": 536
+    },
+    {
+      "epoch": 0.05460299706394499,
+      "grad_norm": 0.1865234375,
+      "learning_rate": 0.02,
+      "loss": 1.8522,
+      "step": 537
+    },
+    {
+      "epoch": 0.05470467862272329,
+      "grad_norm": 0.1728515625,
+      "learning_rate": 0.02,
+      "loss": 1.8735,
+      "step": 538
+    },
+    {
+      "epoch": 0.05480636018150158,
+      "grad_norm": 0.16796875,
+      "learning_rate": 0.02,
+      "loss": 1.8601,
+      "step": 539
+    },
+    {
+      "epoch": 0.05490804174027988,
+      "grad_norm": 0.158203125,
+      "learning_rate": 0.02,
+      "loss": 1.8188,
+      "step": 540
+    },
+    {
+      "epoch": 0.05500972329905818,
+      "grad_norm": 0.177734375,
+      "learning_rate": 0.02,
+      "loss": 1.8671,
+      "step": 541
+    },
+    {
+      "epoch": 0.05511140485783647,
+      "grad_norm": 0.169921875,
+      "learning_rate": 0.02,
+      "loss": 1.9311,
+      "step": 542
+    },
+    {
+      "epoch": 0.055213086416614765,
+      "grad_norm": 0.1748046875,
+      "learning_rate": 0.02,
+      "loss": 1.9234,
+      "step": 543
+    },
+    {
+      "epoch": 0.055314767975393066,
+      "grad_norm": 0.1767578125,
+      "learning_rate": 0.02,
+      "loss": 1.864,
+      "step": 544
+    },
+    {
+      "epoch": 0.05541644953417136,
+      "grad_norm": 0.1962890625,
+      "learning_rate": 0.02,
+      "loss": 1.9139,
+      "step": 545
+    },
+    {
+      "epoch": 0.05551813109294965,
+      "grad_norm": 0.1875,
+      "learning_rate": 0.02,
+      "loss": 1.859,
+      "step": 546
+    },
+    {
+      "epoch": 0.055619812651727954,
+      "grad_norm": 0.1923828125,
+      "learning_rate": 0.02,
+      "loss": 1.8353,
+      "step": 547
+    },
+    {
+      "epoch": 0.05572149421050625,
+      "grad_norm": 0.1865234375,
+      "learning_rate": 0.02,
+      "loss": 1.9553,
+      "step": 548
+    },
+    {
+      "epoch": 0.05582317576928454,
+      "grad_norm": 0.181640625,
+      "learning_rate": 0.02,
+      "loss": 1.8461,
+      "step": 549
+    },
+    {
+      "epoch": 0.05592485732806284,
+      "grad_norm": 0.18359375,
+      "learning_rate": 0.02,
+      "loss": 1.84,
+      "step": 550
+    },
+    {
+      "epoch": 0.056026538886841136,
+      "grad_norm": 0.2021484375,
+      "learning_rate": 0.02,
+      "loss": 1.8416,
+      "step": 551
+    },
+    {
+      "epoch": 0.05612822044561943,
+      "grad_norm": 0.19140625,
+      "learning_rate": 0.02,
+      "loss": 1.8286,
+      "step": 552
+    },
+    {
+      "epoch": 0.05622990200439773,
+      "grad_norm": 0.2021484375,
+      "learning_rate": 0.02,
+      "loss": 1.854,
+      "step": 553
+    },
+    {
+      "epoch": 0.056331583563176024,
+      "grad_norm": 0.19921875,
+      "learning_rate": 0.02,
+      "loss": 1.8725,
+      "step": 554
+    },
+    {
+      "epoch": 0.05643326512195432,
+      "grad_norm": 0.1953125,
+      "learning_rate": 0.02,
+      "loss": 1.8643,
+      "step": 555
+    },
+    {
+      "epoch": 0.05653494668073262,
+      "grad_norm": 0.1982421875,
+      "learning_rate": 0.02,
+      "loss": 1.8885,
+      "step": 556
+    },
+    {
+      "epoch": 0.05663662823951091,
+      "grad_norm": 0.185546875,
+      "learning_rate": 0.02,
+      "loss": 1.8842,
+      "step": 557
+    },
+    {
+      "epoch": 0.056738309798289206,
+      "grad_norm": 0.1953125,
+      "learning_rate": 0.02,
+      "loss": 1.8473,
+      "step": 558
+    },
+    {
+      "epoch": 0.05683999135706751,
+      "grad_norm": 0.1875,
+      "learning_rate": 0.02,
+      "loss": 1.7754,
+      "step": 559
+    },
+    {
+      "epoch": 0.0569416729158458,
+      "grad_norm": 0.1923828125,
+      "learning_rate": 0.02,
+      "loss": 1.8843,
+      "step": 560
+    },
+    {
+      "epoch": 0.057043354474624094,
+      "grad_norm": 0.185546875,
+      "learning_rate": 0.02,
+      "loss": 1.8368,
+      "step": 561
+    },
+    {
+      "epoch": 0.057145036033402395,
+      "grad_norm": 0.1982421875,
+      "learning_rate": 0.02,
+      "loss": 1.8479,
+      "step": 562
+    },
+    {
+      "epoch": 0.05724671759218069,
+      "grad_norm": 0.212890625,
+      "learning_rate": 0.02,
+      "loss": 1.9351,
+      "step": 563
+    },
+    {
+      "epoch": 0.05734839915095898,
+      "grad_norm": 0.2109375,
+      "learning_rate": 0.02,
+      "loss": 1.8939,
+      "step": 564
+    },
+    {
+      "epoch": 0.05745008070973728,
+      "grad_norm": 0.1904296875,
+      "learning_rate": 0.02,
+      "loss": 1.8028,
+      "step": 565
+    },
+    {
+      "epoch": 0.05755176226851558,
+      "grad_norm": 0.201171875,
+      "learning_rate": 0.02,
+      "loss": 1.8764,
+      "step": 566
+    },
+    {
+      "epoch": 0.05765344382729387,
+      "grad_norm": 0.1923828125,
+      "learning_rate": 0.02,
+      "loss": 1.8186,
+      "step": 567
+    },
+    {
+      "epoch": 0.05775512538607217,
+      "grad_norm": 0.1943359375,
+      "learning_rate": 0.02,
+      "loss": 1.882,
+      "step": 568
+    },
+    {
+      "epoch": 0.057856806944850465,
+      "grad_norm": 0.1943359375,
+      "learning_rate": 0.02,
+      "loss": 1.8457,
+      "step": 569
+    },
+    {
+      "epoch": 0.05795848850362876,
+      "grad_norm": 0.1845703125,
+      "learning_rate": 0.02,
+      "loss": 1.8292,
+      "step": 570
+    },
+    {
+      "epoch": 0.05806017006240706,
+      "grad_norm": 0.1884765625,
+      "learning_rate": 0.02,
+      "loss": 1.8868,
+      "step": 571
+    },
+    {
+      "epoch": 0.058161851621185354,
+      "grad_norm": 0.1787109375,
+      "learning_rate": 0.02,
+      "loss": 1.851,
+      "step": 572
+    },
+    {
+      "epoch": 0.05826353317996365,
+      "grad_norm": 0.1904296875,
+      "learning_rate": 0.02,
+      "loss": 1.8639,
+      "step": 573
+    },
+    {
+      "epoch": 0.05836521473874195,
+      "grad_norm": 0.1787109375,
+      "learning_rate": 0.02,
+      "loss": 1.7678,
+      "step": 574
+    },
+    {
+      "epoch": 0.05846689629752024,
+      "grad_norm": 0.1953125,
+      "learning_rate": 0.02,
+      "loss": 1.8709,
+      "step": 575
+    },
+    {
+      "epoch": 0.058568577856298536,
+      "grad_norm": 0.18359375,
+      "learning_rate": 0.02,
+      "loss": 1.8733,
+      "step": 576
+    },
+    {
+      "epoch": 0.058670259415076836,
+      "grad_norm": 0.1962890625,
+      "learning_rate": 0.02,
+      "loss": 1.8186,
+      "step": 577
+    },
+    {
+      "epoch": 0.05877194097385513,
+      "grad_norm": 0.2041015625,
+      "learning_rate": 0.02,
+      "loss": 1.9111,
+      "step": 578
+    },
+    {
+      "epoch": 0.058873622532633424,
+      "grad_norm": 0.2080078125,
+      "learning_rate": 0.02,
+      "loss": 1.8715,
+      "step": 579
+    },
+    {
+      "epoch": 0.058975304091411725,
+      "grad_norm": 0.2001953125,
+      "learning_rate": 0.02,
+      "loss": 1.8842,
+      "step": 580
+    },
+    {
+      "epoch": 0.05907698565019002,
+      "grad_norm": 0.2080078125,
+      "learning_rate": 0.02,
+      "loss": 1.8512,
+      "step": 581
+    },
+    {
+      "epoch": 0.05917866720896831,
+      "grad_norm": 0.22265625,
+      "learning_rate": 0.02,
+      "loss": 1.9205,
+      "step": 582
+    },
+    {
+      "epoch": 0.05928034876774661,
+      "grad_norm": 0.212890625,
+      "learning_rate": 0.02,
+      "loss": 1.8521,
+      "step": 583
+    },
+    {
+      "epoch": 0.05938203032652491,
+      "grad_norm": 0.2158203125,
+      "learning_rate": 0.02,
+      "loss": 1.819,
+      "step": 584
+    },
+    {
+      "epoch": 0.0594837118853032,
+      "grad_norm": 0.212890625,
+      "learning_rate": 0.02,
+      "loss": 1.8738,
+      "step": 585
+    },
+    {
+      "epoch": 0.0595853934440815,
+      "grad_norm": 0.224609375,
+      "learning_rate": 0.02,
+      "loss": 1.8838,
+      "step": 586
+    },
+    {
+      "epoch": 0.059687075002859795,
+      "grad_norm": 0.216796875,
+      "learning_rate": 0.02,
+      "loss": 1.8294,
+      "step": 587
+    },
+    {
+      "epoch": 0.05978875656163809,
+      "grad_norm": 0.212890625,
+      "learning_rate": 0.02,
+      "loss": 1.8372,
+      "step": 588
+    },
+    {
+      "epoch": 0.05989043812041639,
+      "grad_norm": 0.224609375,
+      "learning_rate": 0.02,
+      "loss": 1.8074,
+      "step": 589
+    },
+    {
+      "epoch": 0.05999211967919468,
+      "grad_norm": 0.2080078125,
+      "learning_rate": 0.02,
+      "loss": 1.7526,
+      "step": 590
+    },
+    {
+      "epoch": 0.06009380123797298,
+      "grad_norm": 0.20703125,
+      "learning_rate": 0.02,
+      "loss": 1.8414,
+      "step": 591
+    },
+    {
+      "epoch": 0.06019548279675128,
+      "grad_norm": 0.2109375,
+      "learning_rate": 0.02,
+      "loss": 1.8939,
+      "step": 592
+    },
+    {
+      "epoch": 0.06029716435552957,
+      "grad_norm": 0.1904296875,
+      "learning_rate": 0.02,
+      "loss": 1.8106,
+      "step": 593
+    },
+    {
+      "epoch": 0.060398845914307865,
+      "grad_norm": 0.1865234375,
+      "learning_rate": 0.02,
+      "loss": 1.8324,
+      "step": 594
+    },
+    {
+      "epoch": 0.060500527473086166,
+      "grad_norm": 0.1904296875,
+      "learning_rate": 0.02,
+      "loss": 1.854,
+      "step": 595
+    },
+    {
+      "epoch": 0.06060220903186446,
+      "grad_norm": 0.1845703125,
+      "learning_rate": 0.02,
+      "loss": 1.8131,
+      "step": 596
+    },
+    {
+      "epoch": 0.06070389059064275,
+      "grad_norm": 0.1865234375,
+      "learning_rate": 0.02,
+      "loss": 1.8018,
+      "step": 597
+    },
+    {
+      "epoch": 0.06080557214942105,
+      "grad_norm": 0.1787109375,
+      "learning_rate": 0.02,
+      "loss": 1.8308,
+      "step": 598
+    },
+    {
+      "epoch": 0.06090725370819935,
+      "grad_norm": 0.1865234375,
+      "learning_rate": 0.02,
+      "loss": 1.7933,
+      "step": 599
+    },
+    {
+      "epoch": 0.06100893526697764,
+      "grad_norm": 0.171875,
+      "learning_rate": 0.02,
+      "loss": 1.808,
+      "step": 600
+    },
+    {
+      "epoch": 0.061110616825755935,
+      "grad_norm": 0.1806640625,
+      "learning_rate": 0.02,
+      "loss": 1.8487,
+      "step": 601
+    },
+    {
+      "epoch": 0.061212298384534236,
+      "grad_norm": 0.1904296875,
+      "learning_rate": 0.02,
+      "loss": 1.9196,
+      "step": 602
+    },
+    {
+      "epoch": 0.06131397994331253,
+      "grad_norm": 0.1953125,
+      "learning_rate": 0.02,
+      "loss": 1.8699,
+      "step": 603
+    },
+    {
+      "epoch": 0.061415661502090824,
+      "grad_norm": 0.1884765625,
+      "learning_rate": 0.02,
+      "loss": 1.8149,
+      "step": 604
+    },
+    {
+      "epoch": 0.061517343060869124,
+      "grad_norm": 0.1904296875,
+      "learning_rate": 0.02,
+      "loss": 1.8576,
+      "step": 605
+    },
+    {
+      "epoch": 0.06161902461964742,
+      "grad_norm": 0.1904296875,
+      "learning_rate": 0.02,
+      "loss": 1.9418,
+      "step": 606
+    },
+    {
+      "epoch": 0.06172070617842571,
+      "grad_norm": 0.1865234375,
+      "learning_rate": 0.02,
+      "loss": 1.8505,
+      "step": 607
+    },
+    {
+      "epoch": 0.06182238773720401,
+      "grad_norm": 0.1865234375,
+      "learning_rate": 0.02,
+      "loss": 1.798,
+      "step": 608
+    },
+    {
+      "epoch": 0.061924069295982306,
+      "grad_norm": 0.185546875,
+      "learning_rate": 0.02,
+      "loss": 1.8397,
+      "step": 609
+    },
+    {
+      "epoch": 0.0620257508547606,
+      "grad_norm": 0.181640625,
+      "learning_rate": 0.02,
+      "loss": 1.8976,
+      "step": 610
+    },
+    {
+      "epoch": 0.0621274324135389,
+      "grad_norm": 0.1806640625,
+      "learning_rate": 0.02,
+      "loss": 1.8797,
+      "step": 611
+    },
+    {
+      "epoch": 0.062229113972317195,
+      "grad_norm": 0.1748046875,
+      "learning_rate": 0.02,
+      "loss": 1.8658,
+      "step": 612
+    },
+    {
+      "epoch": 0.06233079553109549,
+      "grad_norm": 0.1728515625,
+      "learning_rate": 0.02,
+      "loss": 1.8707,
+      "step": 613
+    },
+    {
+      "epoch": 0.06243247708987379,
+      "grad_norm": 0.1845703125,
+      "learning_rate": 0.02,
+      "loss": 1.8478,
+      "step": 614
+    },
+    {
+      "epoch": 0.06253415864865208,
+      "grad_norm": 0.18359375,
+      "learning_rate": 0.02,
+      "loss": 1.8331,
+      "step": 615
+    },
+    {
+      "epoch": 0.06263584020743038,
+      "grad_norm": 0.19921875,
+      "learning_rate": 0.02,
+      "loss": 1.8719,
+      "step": 616
+    },
+    {
+      "epoch": 0.06273752176620867,
+      "grad_norm": 0.2080078125,
+      "learning_rate": 0.02,
+      "loss": 1.9068,
+      "step": 617
+    },
+    {
+      "epoch": 0.06283920332498698,
+      "grad_norm": 0.19140625,
+      "learning_rate": 0.02,
+      "loss": 1.8235,
+      "step": 618
+    },
+    {
+      "epoch": 0.06294088488376527,
+      "grad_norm": 0.177734375,
+      "learning_rate": 0.02,
+      "loss": 1.8041,
+      "step": 619
+    },
+    {
+      "epoch": 0.06304256644254357,
+      "grad_norm": 0.169921875,
+      "learning_rate": 0.02,
+      "loss": 1.8535,
+      "step": 620
+    },
+    {
+      "epoch": 0.06314424800132186,
+      "grad_norm": 0.1689453125,
+      "learning_rate": 0.02,
+      "loss": 1.7843,
+      "step": 621
+    },
+    {
+      "epoch": 0.06324592956010015,
+      "grad_norm": 0.181640625,
+      "learning_rate": 0.02,
+      "loss": 1.8573,
+      "step": 622
+    },
+    {
+      "epoch": 0.06334761111887845,
+      "grad_norm": 0.193359375,
+      "learning_rate": 0.02,
+      "loss": 1.8991,
+      "step": 623
+    },
+    {
+      "epoch": 0.06344929267765675,
+      "grad_norm": 0.1865234375,
+      "learning_rate": 0.02,
+      "loss": 1.8153,
+      "step": 624
+    },
+    {
+      "epoch": 0.06355097423643505,
+      "grad_norm": 0.181640625,
+      "learning_rate": 0.02,
+      "loss": 1.7881,
+      "step": 625
+    },
+    {
+      "epoch": 0.06365265579521334,
+      "grad_norm": 0.189453125,
+      "learning_rate": 0.02,
+      "loss": 1.8484,
+      "step": 626
+    },
+    {
+      "epoch": 0.06375433735399164,
+      "grad_norm": 0.177734375,
+      "learning_rate": 0.02,
+      "loss": 1.8426,
+      "step": 627
+    },
+    {
+      "epoch": 0.06385601891276993,
+      "grad_norm": 0.1728515625,
+      "learning_rate": 0.02,
+      "loss": 1.8266,
+      "step": 628
+    },
+    {
+      "epoch": 0.06395770047154822,
+      "grad_norm": 0.19921875,
+      "learning_rate": 0.02,
+      "loss": 1.8731,
+      "step": 629
+    },
+    {
+      "epoch": 0.06405938203032653,
+      "grad_norm": 0.193359375,
+      "learning_rate": 0.02,
+      "loss": 1.8761,
+      "step": 630
+    },
+    {
+      "epoch": 0.06416106358910482,
+      "grad_norm": 0.18359375,
+      "learning_rate": 0.02,
+      "loss": 1.8348,
+      "step": 631
+    },
+    {
+      "epoch": 0.06426274514788312,
+      "grad_norm": 0.1884765625,
+      "learning_rate": 0.02,
+      "loss": 1.8366,
+      "step": 632
+    },
+    {
+      "epoch": 0.06436442670666141,
+      "grad_norm": 0.197265625,
+      "learning_rate": 0.02,
+      "loss": 1.8757,
+      "step": 633
+    },
+    {
+      "epoch": 0.0644661082654397,
+      "grad_norm": 0.1884765625,
+      "learning_rate": 0.02,
+      "loss": 1.8282,
+      "step": 634
+    },
+    {
+      "epoch": 0.064567789824218,
+      "grad_norm": 0.19140625,
+      "learning_rate": 0.02,
+      "loss": 1.8026,
+      "step": 635
+    },
+    {
+      "epoch": 0.06466947138299631,
+      "grad_norm": 0.2001953125,
+      "learning_rate": 0.02,
+      "loss": 1.8798,
+      "step": 636
+    },
+    {
+      "epoch": 0.0647711529417746,
+      "grad_norm": 0.1865234375,
+      "learning_rate": 0.02,
+      "loss": 1.763,
+      "step": 637
+    },
+    {
+      "epoch": 0.0648728345005529,
+      "grad_norm": 0.18359375,
+      "learning_rate": 0.02,
+      "loss": 1.7795,
+      "step": 638
+    },
+    {
+      "epoch": 0.06497451605933119,
+      "grad_norm": 0.171875,
+      "learning_rate": 0.02,
+      "loss": 1.8312,
+      "step": 639
+    },
+    {
+      "epoch": 0.06507619761810948,
+      "grad_norm": 0.1923828125,
+      "learning_rate": 0.02,
+      "loss": 1.8387,
+      "step": 640
+    },
+    {
+      "epoch": 0.06517787917688778,
+      "grad_norm": 0.197265625,
+      "learning_rate": 0.02,
+      "loss": 1.8017,
+      "step": 641
+    },
+    {
+      "epoch": 0.06527956073566608,
+      "grad_norm": 0.1826171875,
+      "learning_rate": 0.02,
+      "loss": 1.8566,
+      "step": 642
+    },
+    {
+      "epoch": 0.06538124229444438,
+      "grad_norm": 0.1982421875,
+      "learning_rate": 0.02,
+      "loss": 1.7853,
+      "step": 643
+    },
+    {
+      "epoch": 0.06548292385322267,
+      "grad_norm": 0.185546875,
+      "learning_rate": 0.02,
+      "loss": 1.8474,
+      "step": 644
+    },
+    {
+      "epoch": 0.06558460541200097,
+      "grad_norm": 0.177734375,
+      "learning_rate": 0.02,
+      "loss": 1.8743,
+      "step": 645
+    },
+    {
+      "epoch": 0.06568628697077926,
+      "grad_norm": 0.185546875,
+      "learning_rate": 0.02,
+      "loss": 1.8229,
+      "step": 646
+    },
+    {
+      "epoch": 0.06578796852955755,
+      "grad_norm": 0.189453125,
+      "learning_rate": 0.02,
+      "loss": 1.8764,
+      "step": 647
+    },
+    {
+      "epoch": 0.06588965008833586,
+      "grad_norm": 0.203125,
+      "learning_rate": 0.02,
+      "loss": 1.8945,
+      "step": 648
+    },
+    {
+      "epoch": 0.06599133164711415,
+      "grad_norm": 0.193359375,
+      "learning_rate": 0.02,
+      "loss": 1.824,
+      "step": 649
+    },
+    {
+      "epoch": 0.06609301320589245,
+      "grad_norm": 0.1953125,
+      "learning_rate": 0.02,
+      "loss": 1.7852,
+      "step": 650
+    },
+    {
+      "epoch": 0.06619469476467074,
+      "grad_norm": 0.1875,
+      "learning_rate": 0.02,
+      "loss": 1.8277,
+      "step": 651
+    },
+    {
+      "epoch": 0.06629637632344904,
+      "grad_norm": 0.2060546875,
+      "learning_rate": 0.02,
+      "loss": 1.8476,
+      "step": 652
+    },
+    {
+      "epoch": 0.06639805788222733,
+      "grad_norm": 0.201171875,
+      "learning_rate": 0.02,
+      "loss": 1.8818,
+      "step": 653
+    },
+    {
+      "epoch": 0.06649973944100564,
+      "grad_norm": 0.193359375,
+      "learning_rate": 0.02,
+      "loss": 1.8788,
+      "step": 654
+    },
+    {
+      "epoch": 0.06660142099978393,
+      "grad_norm": 0.1923828125,
+      "learning_rate": 0.02,
+      "loss": 1.8708,
+      "step": 655
+    },
+    {
+      "epoch": 0.06670310255856222,
+      "grad_norm": 0.208984375,
+      "learning_rate": 0.02,
+      "loss": 1.8532,
+      "step": 656
+    },
+    {
+      "epoch": 0.06680478411734052,
+      "grad_norm": 0.1904296875,
+      "learning_rate": 0.02,
+      "loss": 1.8254,
+      "step": 657
+    },
+    {
+      "epoch": 0.06690646567611881,
+      "grad_norm": 0.193359375,
+      "learning_rate": 0.02,
+      "loss": 1.8561,
+      "step": 658
+    },
+    {
+      "epoch": 0.0670081472348971,
+      "grad_norm": 0.1875,
+      "learning_rate": 0.02,
+      "loss": 1.8459,
+      "step": 659
+    },
+    {
+      "epoch": 0.06710982879367541,
+      "grad_norm": 0.1806640625,
+      "learning_rate": 0.02,
+      "loss": 1.7987,
+      "step": 660
+    },
+    {
+      "epoch": 0.06721151035245371,
+      "grad_norm": 0.185546875,
+      "learning_rate": 0.02,
+      "loss": 1.8576,
+      "step": 661
+    },
+    {
+      "epoch": 0.067313191911232,
+      "grad_norm": 0.1796875,
+      "learning_rate": 0.02,
+      "loss": 1.8265,
+      "step": 662
+    },
+    {
+      "epoch": 0.0674148734700103,
+      "grad_norm": 0.1787109375,
+      "learning_rate": 0.02,
+      "loss": 1.8151,
+      "step": 663
+    },
+    {
+      "epoch": 0.06751655502878859,
+      "grad_norm": 0.1904296875,
+      "learning_rate": 0.02,
+      "loss": 1.8283,
+      "step": 664
+    },
+    {
+      "epoch": 0.06761823658756688,
+      "grad_norm": 0.201171875,
+      "learning_rate": 0.02,
+      "loss": 1.8181,
+      "step": 665
+    },
+    {
+      "epoch": 0.06771991814634519,
+      "grad_norm": 0.189453125,
+      "learning_rate": 0.02,
+      "loss": 1.8193,
+      "step": 666
+    },
+    {
+      "epoch": 0.06782159970512348,
+      "grad_norm": 0.197265625,
+      "learning_rate": 0.02,
+      "loss": 1.8483,
+      "step": 667
+    },
+    {
+      "epoch": 0.06792328126390178,
+      "grad_norm": 0.1845703125,
+      "learning_rate": 0.02,
+      "loss": 1.8552,
+      "step": 668
+    },
+    {
+      "epoch": 0.06802496282268007,
+      "grad_norm": 0.1806640625,
+      "learning_rate": 0.02,
+      "loss": 1.8108,
+      "step": 669
+    },
+    {
+      "epoch": 0.06812664438145836,
+      "grad_norm": 0.1875,
+      "learning_rate": 0.02,
+      "loss": 1.8038,
+      "step": 670
+    },
+    {
+      "epoch": 0.06822832594023666,
+      "grad_norm": 0.1923828125,
+      "learning_rate": 0.02,
+      "loss": 1.8213,
+      "step": 671
+    },
+    {
+      "epoch": 0.06833000749901497,
+      "grad_norm": 0.19921875,
+      "learning_rate": 0.02,
+      "loss": 1.8612,
+      "step": 672
+    },
+    {
+      "epoch": 0.06843168905779326,
+      "grad_norm": 0.1923828125,
+      "learning_rate": 0.02,
+      "loss": 1.8215,
+      "step": 673
+    },
+    {
+      "epoch": 0.06853337061657155,
+      "grad_norm": 0.2001953125,
+      "learning_rate": 0.02,
+      "loss": 1.8086,
+      "step": 674
+    },
+    {
+      "epoch": 0.06863505217534985,
+      "grad_norm": 0.20703125,
+      "learning_rate": 0.02,
+      "loss": 1.8234,
+      "step": 675
+    },
+    {
+      "epoch": 0.06873673373412814,
+      "grad_norm": 0.2021484375,
+      "learning_rate": 0.02,
+      "loss": 1.7992,
+      "step": 676
+    },
+    {
+      "epoch": 0.06883841529290644,
+      "grad_norm": 0.197265625,
+      "learning_rate": 0.02,
+      "loss": 1.8377,
+      "step": 677
+    },
+    {
+      "epoch": 0.06894009685168474,
+      "grad_norm": 0.19921875,
+      "learning_rate": 0.02,
+      "loss": 1.8476,
+      "step": 678
+    },
+    {
+      "epoch": 0.06904177841046304,
+      "grad_norm": 0.1865234375,
+      "learning_rate": 0.02,
+      "loss": 1.7821,
+      "step": 679
+    },
+    {
+      "epoch": 0.06914345996924133,
+      "grad_norm": 0.1875,
+      "learning_rate": 0.02,
+      "loss": 1.8701,
+      "step": 680
+    },
+    {
+      "epoch": 0.06924514152801962,
+      "grad_norm": 0.17578125,
+      "learning_rate": 0.02,
+      "loss": 1.8449,
+      "step": 681
+    },
+    {
+      "epoch": 0.06934682308679792,
+      "grad_norm": 0.1875,
+      "learning_rate": 0.02,
+      "loss": 1.9272,
+      "step": 682
+    },
+    {
+      "epoch": 0.06944850464557621,
+      "grad_norm": 0.189453125,
+      "learning_rate": 0.02,
+      "loss": 1.7936,
+      "step": 683
+    },
+    {
+      "epoch": 0.06955018620435452,
+      "grad_norm": 0.185546875,
+      "learning_rate": 0.02,
+      "loss": 1.8335,
+      "step": 684
+    },
+    {
+      "epoch": 0.06965186776313281,
+      "grad_norm": 0.1728515625,
+      "learning_rate": 0.02,
+      "loss": 1.8328,
+      "step": 685
+    },
+    {
+      "epoch": 0.0697535493219111,
+      "grad_norm": 0.1826171875,
+      "learning_rate": 0.02,
+      "loss": 1.8895,
+      "step": 686
+    },
+    {
+      "epoch": 0.0698552308806894,
+      "grad_norm": 0.16796875,
+      "learning_rate": 0.02,
+      "loss": 1.7895,
+      "step": 687
+    },
+    {
+      "epoch": 0.0699569124394677,
+      "grad_norm": 0.181640625,
+      "learning_rate": 0.02,
+      "loss": 1.8795,
+      "step": 688
+    },
+    {
+      "epoch": 0.07005859399824599,
+      "grad_norm": 0.1787109375,
+      "learning_rate": 0.02,
+      "loss": 1.8543,
+      "step": 689
+    },
+    {
+      "epoch": 0.0701602755570243,
+      "grad_norm": 0.1650390625,
+      "learning_rate": 0.02,
+      "loss": 1.733,
+      "step": 690
+    },
+    {
+      "epoch": 0.07026195711580259,
+      "grad_norm": 0.166015625,
+      "learning_rate": 0.02,
+      "loss": 1.8614,
+      "step": 691
+    },
+    {
+      "epoch": 0.07036363867458088,
+      "grad_norm": 0.171875,
+      "learning_rate": 0.02,
+      "loss": 1.7814,
+      "step": 692
+    },
+    {
+      "epoch": 0.07046532023335918,
+      "grad_norm": 0.173828125,
+      "learning_rate": 0.02,
+      "loss": 1.8387,
+      "step": 693
+    },
+    {
+      "epoch": 0.07056700179213747,
+      "grad_norm": 0.17578125,
+      "learning_rate": 0.02,
+      "loss": 1.8363,
+      "step": 694
+    },
+    {
+      "epoch": 0.07066868335091576,
+      "grad_norm": 0.1845703125,
+      "learning_rate": 0.02,
+      "loss": 1.9104,
+      "step": 695
+    },
+    {
+      "epoch": 0.07077036490969407,
+      "grad_norm": 0.1787109375,
+      "learning_rate": 0.02,
+      "loss": 1.7387,
+      "step": 696
+    },
+    {
+      "epoch": 0.07087204646847237,
+      "grad_norm": 0.1982421875,
+      "learning_rate": 0.02,
+      "loss": 1.9323,
+      "step": 697
+    },
+    {
+      "epoch": 0.07097372802725066,
+      "grad_norm": 0.1806640625,
+      "learning_rate": 0.02,
+      "loss": 1.8411,
+      "step": 698
+    },
+    {
+      "epoch": 0.07107540958602895,
+      "grad_norm": 0.181640625,
+      "learning_rate": 0.02,
+      "loss": 1.8653,
+      "step": 699
+    },
+    {
+      "epoch": 0.07117709114480725,
+      "grad_norm": 0.189453125,
+      "learning_rate": 0.02,
+      "loss": 1.7791,
+      "step": 700
+    },
+    {
+      "epoch": 0.07127877270358554,
+      "grad_norm": 0.1904296875,
+      "learning_rate": 0.02,
+      "loss": 1.7871,
+      "step": 701
+    },
+    {
+      "epoch": 0.07138045426236385,
+      "grad_norm": 0.19140625,
+      "learning_rate": 0.02,
+      "loss": 1.8244,
+      "step": 702
+    },
+    {
+      "epoch": 0.07148213582114214,
+      "grad_norm": 0.1806640625,
+      "learning_rate": 0.02,
+      "loss": 1.7784,
+      "step": 703
+    },
+    {
+      "epoch": 0.07158381737992044,
+      "grad_norm": 0.1904296875,
+      "learning_rate": 0.02,
+      "loss": 1.8928,
+      "step": 704
+    },
+    {
+      "epoch": 0.07168549893869873,
+      "grad_norm": 0.177734375,
+      "learning_rate": 0.02,
+      "loss": 1.8151,
+      "step": 705
+    },
+    {
+      "epoch": 0.07178718049747702,
+      "grad_norm": 0.185546875,
+      "learning_rate": 0.02,
+      "loss": 1.8431,
+      "step": 706
+    },
+    {
+      "epoch": 0.07188886205625532,
+      "grad_norm": 0.181640625,
+      "learning_rate": 0.02,
+      "loss": 1.8198,
+      "step": 707
+    },
+    {
+      "epoch": 0.07199054361503363,
+      "grad_norm": 0.1884765625,
+      "learning_rate": 0.02,
+      "loss": 1.8371,
+      "step": 708
+    },
+    {
+      "epoch": 0.07209222517381192,
+      "grad_norm": 0.1826171875,
+      "learning_rate": 0.02,
+      "loss": 1.8308,
+      "step": 709
+    },
+    {
+      "epoch": 0.07219390673259021,
+      "grad_norm": 0.1875,
+      "learning_rate": 0.02,
+      "loss": 1.835,
+      "step": 710
+    },
+    {
+      "epoch": 0.0722955882913685,
+      "grad_norm": 0.1865234375,
+      "learning_rate": 0.02,
+      "loss": 1.8718,
+      "step": 711
+    },
+    {
+      "epoch": 0.0723972698501468,
+      "grad_norm": 0.1865234375,
+      "learning_rate": 0.02,
+      "loss": 1.7849,
+      "step": 712
+    },
+    {
+      "epoch": 0.0724989514089251,
+      "grad_norm": 0.1884765625,
+      "learning_rate": 0.02,
+      "loss": 1.7968,
+      "step": 713
+    },
+    {
+      "epoch": 0.0726006329677034,
+      "grad_norm": 0.1962890625,
+      "learning_rate": 0.02,
+      "loss": 1.7529,
+      "step": 714
+    },
+    {
+      "epoch": 0.0727023145264817,
+      "grad_norm": 0.1923828125,
+      "learning_rate": 0.02,
+      "loss": 1.8578,
+      "step": 715
+    },
+    {
+      "epoch": 0.07280399608525999,
+      "grad_norm": 0.203125,
+      "learning_rate": 0.02,
+      "loss": 1.8665,
+      "step": 716
+    },
+    {
+      "epoch": 0.07290567764403828,
+      "grad_norm": 0.2138671875,
+      "learning_rate": 0.02,
+      "loss": 1.8932,
+      "step": 717
+    },
+    {
+      "epoch": 0.07300735920281658,
+      "grad_norm": 0.2060546875,
+      "learning_rate": 0.02,
+      "loss": 1.7948,
+      "step": 718
+    },
+    {
+      "epoch": 0.07310904076159487,
+      "grad_norm": 0.1982421875,
+      "learning_rate": 0.02,
+      "loss": 1.8575,
+      "step": 719
+    },
+    {
+      "epoch": 0.07321072232037316,
+      "grad_norm": 0.2080078125,
+      "learning_rate": 0.02,
+      "loss": 1.808,
+      "step": 720
+    },
+    {
+      "epoch": 0.07331240387915147,
+      "grad_norm": 0.220703125,
+      "learning_rate": 0.02,
+      "loss": 1.8696,
+      "step": 721
+    },
+    {
+      "epoch": 0.07341408543792977,
+      "grad_norm": 0.20703125,
+      "learning_rate": 0.02,
+      "loss": 1.8024,
+      "step": 722
+    },
+    {
+      "epoch": 0.07351576699670806,
+      "grad_norm": 0.1962890625,
+      "learning_rate": 0.02,
+      "loss": 1.7953,
+      "step": 723
+    },
+    {
+      "epoch": 0.07361744855548635,
+      "grad_norm": 0.18359375,
+      "learning_rate": 0.02,
+      "loss": 1.7472,
+      "step": 724
+    },
+    {
+      "epoch": 0.07371913011426465,
+      "grad_norm": 0.1845703125,
+      "learning_rate": 0.02,
+      "loss": 1.8354,
+      "step": 725
+    },
+    {
+      "epoch": 0.07382081167304294,
+      "grad_norm": 0.1943359375,
+      "learning_rate": 0.02,
+      "loss": 1.7947,
+      "step": 726
+    },
+    {
+      "epoch": 0.07392249323182125,
+      "grad_norm": 0.1884765625,
+      "learning_rate": 0.02,
+      "loss": 1.8359,
+      "step": 727
+    },
+    {
+      "epoch": 0.07402417479059954,
+      "grad_norm": 0.1748046875,
+      "learning_rate": 0.02,
+      "loss": 1.7698,
+      "step": 728
+    },
+    {
+      "epoch": 0.07412585634937784,
+      "grad_norm": 0.1923828125,
+      "learning_rate": 0.02,
+      "loss": 1.8036,
+      "step": 729
+    },
+    {
+      "epoch": 0.07422753790815613,
+      "grad_norm": 0.197265625,
+      "learning_rate": 0.02,
+      "loss": 1.8761,
+      "step": 730
+    },
+    {
+      "epoch": 0.07432921946693442,
+      "grad_norm": 0.189453125,
+      "learning_rate": 0.02,
+      "loss": 1.8475,
+      "step": 731
+    },
+    {
+      "epoch": 0.07443090102571272,
+      "grad_norm": 0.1904296875,
+      "learning_rate": 0.02,
+      "loss": 1.819,
+      "step": 732
+    },
+    {
+      "epoch": 0.07453258258449103,
+      "grad_norm": 0.1884765625,
+      "learning_rate": 0.02,
+      "loss": 1.8036,
+      "step": 733
+    },
+    {
+      "epoch": 0.07463426414326932,
+      "grad_norm": 0.1962890625,
+      "learning_rate": 0.02,
+      "loss": 1.8351,
+      "step": 734
+    },
+    {
+      "epoch": 0.07473594570204761,
+      "grad_norm": 0.171875,
+      "learning_rate": 0.02,
+      "loss": 1.775,
+      "step": 735
+    },
+    {
+      "epoch": 0.0748376272608259,
+      "grad_norm": 0.1845703125,
+      "learning_rate": 0.02,
+      "loss": 1.8744,
+      "step": 736
+    },
+    {
+      "epoch": 0.0749393088196042,
+      "grad_norm": 0.1796875,
+      "learning_rate": 0.02,
+      "loss": 1.8185,
+      "step": 737
+    },
+    {
+      "epoch": 0.0750409903783825,
+      "grad_norm": 0.1806640625,
+      "learning_rate": 0.02,
+      "loss": 1.8726,
+      "step": 738
+    },
+    {
+      "epoch": 0.0751426719371608,
+      "grad_norm": 0.16796875,
+      "learning_rate": 0.02,
+      "loss": 1.7476,
+      "step": 739
+    },
+    {
+      "epoch": 0.0752443534959391,
+      "grad_norm": 0.1748046875,
+      "learning_rate": 0.02,
+      "loss": 1.8592,
+      "step": 740
+    },
+    {
+      "epoch": 0.07534603505471739,
+      "grad_norm": 0.16796875,
+      "learning_rate": 0.02,
+      "loss": 1.8928,
+      "step": 741
+    },
+    {
+      "epoch": 0.07544771661349568,
+      "grad_norm": 0.18359375,
+      "learning_rate": 0.02,
+      "loss": 1.8444,
+      "step": 742
+    },
+    {
+      "epoch": 0.07554939817227398,
+      "grad_norm": 0.1640625,
+      "learning_rate": 0.02,
+      "loss": 1.7664,
+      "step": 743
+    },
+    {
+      "epoch": 0.07565107973105227,
+      "grad_norm": 0.173828125,
+      "learning_rate": 0.02,
+      "loss": 1.8221,
+      "step": 744
+    },
+    {
+      "epoch": 0.07575276128983058,
+      "grad_norm": 0.1669921875,
+      "learning_rate": 0.02,
+      "loss": 1.8262,
+      "step": 745
+    },
+    {
+      "epoch": 0.07585444284860887,
+      "grad_norm": 0.171875,
+      "learning_rate": 0.02,
+      "loss": 1.896,
+      "step": 746
+    },
+    {
+      "epoch": 0.07595612440738717,
+      "grad_norm": 0.1689453125,
+      "learning_rate": 0.02,
+      "loss": 1.7998,
+      "step": 747
+    },
+    {
+      "epoch": 0.07605780596616546,
+      "grad_norm": 0.16796875,
+      "learning_rate": 0.02,
+      "loss": 1.8514,
+      "step": 748
+    },
+    {
+      "epoch": 0.07615948752494375,
+      "grad_norm": 0.1640625,
+      "learning_rate": 0.02,
+      "loss": 1.7885,
+      "step": 749
+    },
+    {
+      "epoch": 0.07626116908372205,
+      "grad_norm": 0.1806640625,
+      "learning_rate": 0.02,
+      "loss": 1.8944,
+      "step": 750
+    },
+    {
+      "epoch": 0.07636285064250035,
+      "grad_norm": 0.17578125,
+      "learning_rate": 0.02,
+      "loss": 1.8745,
+      "step": 751
+    },
+    {
+      "epoch": 0.07646453220127865,
+      "grad_norm": 0.1806640625,
+      "learning_rate": 0.02,
+      "loss": 1.8889,
+      "step": 752
+    },
+    {
+      "epoch": 0.07656621376005694,
+      "grad_norm": 0.1708984375,
+      "learning_rate": 0.02,
+      "loss": 1.7657,
+      "step": 753
+    },
+    {
+      "epoch": 0.07666789531883524,
+      "grad_norm": 0.1806640625,
+      "learning_rate": 0.02,
+      "loss": 1.783,
+      "step": 754
+    },
+    {
+      "epoch": 0.07676957687761353,
+      "grad_norm": 0.181640625,
+      "learning_rate": 0.02,
+      "loss": 1.8282,
+      "step": 755
+    },
+    {
+      "epoch": 0.07687125843639182,
+      "grad_norm": 0.1796875,
+      "learning_rate": 0.02,
+      "loss": 1.8033,
+      "step": 756
+    },
+    {
+      "epoch": 0.07697293999517013,
+      "grad_norm": 0.171875,
+      "learning_rate": 0.02,
+      "loss": 1.7795,
+      "step": 757
+    },
+    {
+      "epoch": 0.07707462155394842,
+      "grad_norm": 0.185546875,
+      "learning_rate": 0.02,
+      "loss": 1.8305,
+      "step": 758
+    },
+    {
+      "epoch": 0.07717630311272672,
+      "grad_norm": 0.1845703125,
+      "learning_rate": 0.02,
+      "loss": 1.8464,
+      "step": 759
+    },
+    {
+      "epoch": 0.07727798467150501,
+      "grad_norm": 0.1865234375,
+      "learning_rate": 0.02,
+      "loss": 1.8401,
+      "step": 760
+    },
+    {
+      "epoch": 0.0773796662302833,
+      "grad_norm": 0.1875,
+      "learning_rate": 0.02,
+      "loss": 1.8621,
+      "step": 761
+    },
+    {
+      "epoch": 0.0774813477890616,
+      "grad_norm": 0.189453125,
+      "learning_rate": 0.02,
+      "loss": 1.7909,
+      "step": 762
+    },
+    {
+      "epoch": 0.07758302934783991,
+      "grad_norm": 0.1845703125,
+      "learning_rate": 0.02,
+      "loss": 1.8341,
+      "step": 763
+    },
+    {
+      "epoch": 0.0776847109066182,
+      "grad_norm": 0.19140625,
+      "learning_rate": 0.02,
+      "loss": 1.871,
+      "step": 764
+    },
+    {
+      "epoch": 0.0777863924653965,
+      "grad_norm": 0.1943359375,
+      "learning_rate": 0.02,
+      "loss": 1.8292,
+      "step": 765
+    },
+    {
+      "epoch": 0.07788807402417479,
+      "grad_norm": 0.1962890625,
+      "learning_rate": 0.02,
+      "loss": 1.8056,
+      "step": 766
+    },
+    {
+      "epoch": 0.07798975558295308,
+      "grad_norm": 0.201171875,
+      "learning_rate": 0.02,
+      "loss": 1.7988,
+      "step": 767
+    },
+    {
+      "epoch": 0.07809143714173138,
+      "grad_norm": 0.193359375,
+      "learning_rate": 0.02,
+      "loss": 1.7788,
+      "step": 768
+    },
+    {
+      "epoch": 0.07819311870050968,
+      "grad_norm": 0.1953125,
+      "learning_rate": 0.02,
+      "loss": 1.8213,
+      "step": 769
+    },
+    {
+      "epoch": 0.07829480025928798,
+      "grad_norm": 0.177734375,
+      "learning_rate": 0.02,
+      "loss": 1.8004,
+      "step": 770
+    },
+    {
+      "epoch": 0.07839648181806627,
+      "grad_norm": 0.1708984375,
+      "learning_rate": 0.02,
+      "loss": 1.7865,
+      "step": 771
+    },
+    {
+      "epoch": 0.07849816337684457,
+      "grad_norm": 0.189453125,
+      "learning_rate": 0.02,
+      "loss": 1.8506,
+      "step": 772
+    },
+    {
+      "epoch": 0.07859984493562286,
+      "grad_norm": 0.1826171875,
+      "learning_rate": 0.02,
+      "loss": 1.8121,
+      "step": 773
+    },
+    {
+      "epoch": 0.07870152649440115,
+      "grad_norm": 0.1787109375,
+      "learning_rate": 0.02,
+      "loss": 1.7919,
+      "step": 774
+    },
+    {
+      "epoch": 0.07880320805317946,
+      "grad_norm": 0.1826171875,
+      "learning_rate": 0.02,
+      "loss": 1.8866,
+      "step": 775
+    },
+    {
+      "epoch": 0.07890488961195775,
+      "grad_norm": 0.185546875,
+      "learning_rate": 0.02,
+      "loss": 1.8767,
+      "step": 776
+    },
+    {
+      "epoch": 0.07900657117073605,
+      "grad_norm": 0.1806640625,
+      "learning_rate": 0.02,
+      "loss": 1.8389,
+      "step": 777
+    },
+    {
+      "epoch": 0.07910825272951434,
+      "grad_norm": 0.1806640625,
+      "learning_rate": 0.02,
+      "loss": 1.86,
+      "step": 778
+    },
+    {
+      "epoch": 0.07920993428829264,
+      "grad_norm": 0.1669921875,
+      "learning_rate": 0.02,
+      "loss": 1.7923,
+      "step": 779
+    },
+    {
+      "epoch": 0.07931161584707093,
+      "grad_norm": 0.1669921875,
+      "learning_rate": 0.02,
+      "loss": 1.7526,
+      "step": 780
+    },
+    {
+      "epoch": 0.07941329740584924,
+      "grad_norm": 0.177734375,
+      "learning_rate": 0.02,
+      "loss": 1.7722,
+      "step": 781
+    },
+    {
+      "epoch": 0.07951497896462753,
+      "grad_norm": 0.169921875,
+      "learning_rate": 0.02,
+      "loss": 1.8182,
+      "step": 782
+    },
+    {
+      "epoch": 0.07961666052340582,
+      "grad_norm": 0.1689453125,
+      "learning_rate": 0.02,
+      "loss": 1.852,
+      "step": 783
+    },
+    {
+      "epoch": 0.07971834208218412,
+      "grad_norm": 0.169921875,
+      "learning_rate": 0.02,
+      "loss": 1.8158,
+      "step": 784
+    },
+    {
+      "epoch": 0.07982002364096241,
+      "grad_norm": 0.1826171875,
+      "learning_rate": 0.02,
+      "loss": 1.8744,
+      "step": 785
+    },
+    {
+      "epoch": 0.0799217051997407,
+      "grad_norm": 0.171875,
+      "learning_rate": 0.02,
+      "loss": 1.8585,
+      "step": 786
+    },
+    {
+      "epoch": 0.08002338675851901,
+      "grad_norm": 0.1708984375,
+      "learning_rate": 0.02,
+      "loss": 1.8676,
+      "step": 787
+    },
+    {
+      "epoch": 0.08012506831729731,
+      "grad_norm": 0.17578125,
+      "learning_rate": 0.02,
+      "loss": 1.8289,
+      "step": 788
+    },
+    {
+      "epoch": 0.0802267498760756,
+      "grad_norm": 0.177734375,
+      "learning_rate": 0.02,
+      "loss": 1.8452,
+      "step": 789
+    },
+    {
+      "epoch": 0.0803284314348539,
+      "grad_norm": 0.1806640625,
+      "learning_rate": 0.02,
+      "loss": 1.8626,
+      "step": 790
+    },
+    {
+      "epoch": 0.08043011299363219,
+      "grad_norm": 0.1640625,
+      "learning_rate": 0.02,
+      "loss": 1.793,
+      "step": 791
+    },
+    {
+      "epoch": 0.08053179455241048,
+      "grad_norm": 0.1708984375,
+      "learning_rate": 0.02,
+      "loss": 1.8057,
+      "step": 792
+    },
+    {
+      "epoch": 0.08063347611118879,
+      "grad_norm": 0.1787109375,
+      "learning_rate": 0.02,
+      "loss": 1.8584,
+      "step": 793
+    },
+    {
+      "epoch": 0.08073515766996708,
+      "grad_norm": 0.177734375,
+      "learning_rate": 0.02,
+      "loss": 1.8268,
+      "step": 794
+    },
+    {
+      "epoch": 0.08083683922874538,
+      "grad_norm": 0.181640625,
+      "learning_rate": 0.02,
+      "loss": 1.773,
+      "step": 795
+    },
+    {
+      "epoch": 0.08093852078752367,
+      "grad_norm": 0.17578125,
+      "learning_rate": 0.02,
+      "loss": 1.7506,
+      "step": 796
+    },
+    {
+      "epoch": 0.08104020234630196,
+      "grad_norm": 0.1796875,
+      "learning_rate": 0.02,
+      "loss": 1.7988,
+      "step": 797
+    },
+    {
+      "epoch": 0.08114188390508026,
+      "grad_norm": 0.181640625,
+      "learning_rate": 0.02,
+      "loss": 1.9268,
+      "step": 798
+    },
+    {
+      "epoch": 0.08124356546385857,
+      "grad_norm": 0.1875,
+      "learning_rate": 0.02,
+      "loss": 1.8278,
+      "step": 799
+    },
+    {
+      "epoch": 0.08134524702263686,
+      "grad_norm": 0.1865234375,
+      "learning_rate": 0.02,
+      "loss": 1.8372,
+      "step": 800
+    },
+    {
+      "epoch": 0.08144692858141515,
+      "grad_norm": 0.197265625,
+      "learning_rate": 0.02,
+      "loss": 1.8037,
+      "step": 801
+    },
+    {
+      "epoch": 0.08154861014019345,
+      "grad_norm": 0.1875,
+      "learning_rate": 0.02,
+      "loss": 1.8436,
+      "step": 802
+    },
+    {
+      "epoch": 0.08165029169897174,
+      "grad_norm": 0.1865234375,
+      "learning_rate": 0.02,
+      "loss": 1.8158,
+      "step": 803
+    },
+    {
+      "epoch": 0.08175197325775004,
+      "grad_norm": 0.1826171875,
+      "learning_rate": 0.02,
+      "loss": 1.7991,
+      "step": 804
+    },
+    {
+      "epoch": 0.08185365481652834,
+      "grad_norm": 0.1904296875,
+      "learning_rate": 0.02,
+      "loss": 1.8109,
+      "step": 805
+    },
+    {
+      "epoch": 0.08195533637530664,
+      "grad_norm": 0.1787109375,
+      "learning_rate": 0.02,
+      "loss": 1.8185,
+      "step": 806
+    },
+    {
+      "epoch": 0.08205701793408493,
+      "grad_norm": 0.173828125,
+      "learning_rate": 0.02,
+      "loss": 1.7887,
+      "step": 807
+    },
+    {
+      "epoch": 0.08215869949286322,
+      "grad_norm": 0.193359375,
+      "learning_rate": 0.02,
+      "loss": 1.8596,
+      "step": 808
+    },
+    {
+      "epoch": 0.08226038105164152,
+      "grad_norm": 0.18359375,
+      "learning_rate": 0.02,
+      "loss": 1.8081,
+      "step": 809
+    },
+    {
+      "epoch": 0.08236206261041981,
+      "grad_norm": 0.1875,
+      "learning_rate": 0.02,
+      "loss": 1.7933,
+      "step": 810
+    },
+    {
+      "epoch": 0.08246374416919812,
+      "grad_norm": 0.1962890625,
+      "learning_rate": 0.02,
+      "loss": 1.8441,
+      "step": 811
+    },
+    {
+      "epoch": 0.08256542572797641,
+      "grad_norm": 0.197265625,
+      "learning_rate": 0.02,
+      "loss": 1.7854,
+      "step": 812
+    },
+    {
+      "epoch": 0.0826671072867547,
+      "grad_norm": 0.1875,
+      "learning_rate": 0.02,
+      "loss": 1.7749,
+      "step": 813
+    },
+    {
+      "epoch": 0.082768788845533,
+      "grad_norm": 0.1875,
+      "learning_rate": 0.02,
+      "loss": 1.7733,
+      "step": 814
+    },
+    {
+      "epoch": 0.0828704704043113,
+      "grad_norm": 0.1875,
+      "learning_rate": 0.02,
+      "loss": 1.8147,
+      "step": 815
+    },
+    {
+      "epoch": 0.08297215196308959,
+      "grad_norm": 0.1826171875,
+      "learning_rate": 0.02,
+      "loss": 1.8334,
+      "step": 816
+    },
+    {
+      "epoch": 0.0830738335218679,
+      "grad_norm": 0.18359375,
+      "learning_rate": 0.02,
+      "loss": 1.8312,
+      "step": 817
+    },
+    {
+      "epoch": 0.08317551508064619,
+      "grad_norm": 0.1806640625,
+      "learning_rate": 0.02,
+      "loss": 1.7844,
+      "step": 818
+    },
+    {
+      "epoch": 0.08327719663942448,
+      "grad_norm": 0.1826171875,
+      "learning_rate": 0.02,
+      "loss": 1.8148,
+      "step": 819
+    },
+    {
+      "epoch": 0.08337887819820278,
+      "grad_norm": 0.1787109375,
+      "learning_rate": 0.02,
+      "loss": 1.8021,
+      "step": 820
+    },
+    {
+      "epoch": 0.08348055975698107,
+      "grad_norm": 0.1982421875,
+      "learning_rate": 0.02,
+      "loss": 1.8545,
+      "step": 821
+    },
+    {
+      "epoch": 0.08358224131575936,
+      "grad_norm": 0.201171875,
+      "learning_rate": 0.02,
+      "loss": 1.8369,
+      "step": 822
+    },
+    {
+      "epoch": 0.08368392287453767,
+      "grad_norm": 0.185546875,
+      "learning_rate": 0.02,
+      "loss": 1.7756,
+      "step": 823
+    },
+    {
+      "epoch": 0.08378560443331597,
+      "grad_norm": 0.181640625,
+      "learning_rate": 0.02,
+      "loss": 1.8681,
+      "step": 824
+    },
+    {
+      "epoch": 0.08388728599209426,
+      "grad_norm": 0.1904296875,
+      "learning_rate": 0.02,
+      "loss": 1.8223,
+      "step": 825
+    },
+    {
+      "epoch": 0.08398896755087255,
+      "grad_norm": 0.1845703125,
+      "learning_rate": 0.02,
+      "loss": 1.8387,
+      "step": 826
+    },
+    {
+      "epoch": 0.08409064910965085,
+      "grad_norm": 0.1884765625,
+      "learning_rate": 0.02,
+      "loss": 1.7933,
+      "step": 827
+    },
+    {
+      "epoch": 0.08419233066842914,
+      "grad_norm": 0.1806640625,
+      "learning_rate": 0.02,
+      "loss": 1.7993,
+      "step": 828
+    },
+    {
+      "epoch": 0.08429401222720745,
+      "grad_norm": 0.1875,
+      "learning_rate": 0.02,
+      "loss": 1.8489,
+      "step": 829
+    },
+    {
+      "epoch": 0.08439569378598574,
+      "grad_norm": 0.1796875,
+      "learning_rate": 0.02,
+      "loss": 1.8037,
+      "step": 830
+    },
+    {
+      "epoch": 0.08449737534476404,
+      "grad_norm": 0.1875,
+      "learning_rate": 0.02,
+      "loss": 1.8211,
+      "step": 831
+    },
+    {
+      "epoch": 0.08459905690354233,
+      "grad_norm": 0.1787109375,
+      "learning_rate": 0.02,
+      "loss": 1.7969,
+      "step": 832
+    },
+    {
+      "epoch": 0.08470073846232062,
+      "grad_norm": 0.193359375,
+      "learning_rate": 0.02,
+      "loss": 1.8219,
+      "step": 833
+    },
+    {
+      "epoch": 0.08480242002109892,
+      "grad_norm": 0.1796875,
+      "learning_rate": 0.02,
+      "loss": 1.7981,
+      "step": 834
+    },
+    {
+      "epoch": 0.08490410157987723,
+      "grad_norm": 0.1796875,
+      "learning_rate": 0.02,
+      "loss": 1.8286,
+      "step": 835
+    },
+    {
+      "epoch": 0.08500578313865552,
+      "grad_norm": 0.185546875,
+      "learning_rate": 0.02,
+      "loss": 1.8327,
+      "step": 836
+    },
+    {
+      "epoch": 0.08510746469743381,
+      "grad_norm": 0.177734375,
+      "learning_rate": 0.02,
+      "loss": 1.7936,
+      "step": 837
+    },
+    {
+      "epoch": 0.0852091462562121,
+      "grad_norm": 0.1787109375,
+      "learning_rate": 0.02,
+      "loss": 1.8004,
+      "step": 838
+    },
+    {
+      "epoch": 0.0853108278149904,
+      "grad_norm": 0.169921875,
+      "learning_rate": 0.02,
+      "loss": 1.7659,
+      "step": 839
+    },
+    {
+      "epoch": 0.0854125093737687,
+      "grad_norm": 0.171875,
+      "learning_rate": 0.02,
+      "loss": 1.834,
+      "step": 840
+    },
+    {
+      "epoch": 0.085514190932547,
+      "grad_norm": 0.1845703125,
+      "learning_rate": 0.02,
+      "loss": 1.8353,
+      "step": 841
+    },
+    {
+      "epoch": 0.0856158724913253,
+      "grad_norm": 0.1806640625,
+      "learning_rate": 0.02,
+      "loss": 1.8302,
+      "step": 842
+    },
+    {
+      "epoch": 0.08571755405010359,
+      "grad_norm": 0.193359375,
+      "learning_rate": 0.02,
+      "loss": 1.797,
+      "step": 843
+    },
+    {
+      "epoch": 0.08581923560888188,
+      "grad_norm": 0.1796875,
+      "learning_rate": 0.02,
+      "loss": 1.787,
+      "step": 844
+    },
+    {
+      "epoch": 0.08592091716766018,
+      "grad_norm": 0.169921875,
+      "learning_rate": 0.02,
+      "loss": 1.7846,
+      "step": 845
+    },
+    {
+      "epoch": 0.08602259872643847,
+      "grad_norm": 0.1796875,
+      "learning_rate": 0.02,
+      "loss": 1.7675,
+      "step": 846
+    },
+    {
+      "epoch": 0.08612428028521678,
+      "grad_norm": 0.1708984375,
+      "learning_rate": 0.02,
+      "loss": 1.7912,
+      "step": 847
+    },
+    {
+      "epoch": 0.08622596184399507,
+      "grad_norm": 0.1904296875,
+      "learning_rate": 0.02,
+      "loss": 1.8381,
+      "step": 848
+    },
+    {
+      "epoch": 0.08632764340277337,
+      "grad_norm": 0.18359375,
+      "learning_rate": 0.02,
+      "loss": 1.8145,
+      "step": 849
+    },
+    {
+      "epoch": 0.08642932496155166,
+      "grad_norm": 0.1904296875,
+      "learning_rate": 0.02,
+      "loss": 1.8315,
+      "step": 850
+    },
+    {
+      "epoch": 0.08653100652032995,
+      "grad_norm": 0.1923828125,
+      "learning_rate": 0.02,
+      "loss": 1.844,
+      "step": 851
+    },
+    {
+      "epoch": 0.08663268807910825,
+      "grad_norm": 0.1748046875,
+      "learning_rate": 0.02,
+      "loss": 1.764,
+      "step": 852
+    },
+    {
+      "epoch": 0.08673436963788655,
+      "grad_norm": 0.173828125,
+      "learning_rate": 0.02,
+      "loss": 1.8211,
+      "step": 853
+    },
+    {
+      "epoch": 0.08683605119666485,
+      "grad_norm": 0.19921875,
+      "learning_rate": 0.02,
+      "loss": 1.8561,
+      "step": 854
+    },
+    {
+      "epoch": 0.08693773275544314,
+      "grad_norm": 0.1689453125,
+      "learning_rate": 0.02,
+      "loss": 1.7917,
+      "step": 855
+    },
+    {
+      "epoch": 0.08703941431422144,
+      "grad_norm": 0.169921875,
+      "learning_rate": 0.02,
+      "loss": 1.7871,
+      "step": 856
+    },
+    {
+      "epoch": 0.08714109587299973,
+      "grad_norm": 0.173828125,
+      "learning_rate": 0.02,
+      "loss": 1.8257,
+      "step": 857
+    },
+    {
+      "epoch": 0.08724277743177802,
+      "grad_norm": 0.1728515625,
+      "learning_rate": 0.02,
+      "loss": 1.8424,
+      "step": 858
+    },
+    {
+      "epoch": 0.08734445899055633,
+      "grad_norm": 0.28125,
+      "learning_rate": 0.02,
+      "loss": 1.819,
+      "step": 859
+    },
+    {
+      "epoch": 0.08744614054933462,
+      "grad_norm": 0.177734375,
+      "learning_rate": 0.02,
+      "loss": 1.8355,
+      "step": 860
+    },
+    {
+      "epoch": 0.08754782210811292,
+      "grad_norm": 0.173828125,
+      "learning_rate": 0.02,
+      "loss": 1.8782,
+      "step": 861
+    },
+    {
+      "epoch": 0.08764950366689121,
+      "grad_norm": 0.1669921875,
+      "learning_rate": 0.02,
+      "loss": 1.8362,
+      "step": 862
+    },
+    {
+      "epoch": 0.0877511852256695,
+      "grad_norm": 0.16015625,
+      "learning_rate": 0.02,
+      "loss": 1.819,
+      "step": 863
+    },
+    {
+      "epoch": 0.0878528667844478,
+      "grad_norm": 0.162109375,
+      "learning_rate": 0.02,
+      "loss": 1.8046,
+      "step": 864
+    },
+    {
+      "epoch": 0.08795454834322611,
+      "grad_norm": 0.166015625,
+      "learning_rate": 0.02,
+      "loss": 1.8154,
+      "step": 865
+    },
+    {
+      "epoch": 0.0880562299020044,
+      "grad_norm": 0.1708984375,
+      "learning_rate": 0.02,
+      "loss": 1.7993,
+      "step": 866
+    },
+    {
+      "epoch": 0.0881579114607827,
+      "grad_norm": 0.1689453125,
+      "learning_rate": 0.02,
+      "loss": 1.8413,
+      "step": 867
+    },
+    {
+      "epoch": 0.08825959301956099,
+      "grad_norm": 0.1748046875,
+      "learning_rate": 0.02,
+      "loss": 1.7567,
+      "step": 868
+    },
+    {
+      "epoch": 0.08836127457833928,
+      "grad_norm": 0.16796875,
+      "learning_rate": 0.02,
+      "loss": 1.8311,
+      "step": 869
+    },
+    {
+      "epoch": 0.08846295613711758,
+      "grad_norm": 0.1611328125,
+      "learning_rate": 0.02,
+      "loss": 1.7413,
+      "step": 870
+    },
+    {
+      "epoch": 0.08856463769589588,
+      "grad_norm": 0.1572265625,
+      "learning_rate": 0.02,
+      "loss": 1.7816,
+      "step": 871
+    },
+    {
+      "epoch": 0.08866631925467418,
+      "grad_norm": 0.158203125,
+      "learning_rate": 0.02,
+      "loss": 1.8152,
+      "step": 872
+    },
+    {
+      "epoch": 0.08876800081345247,
+      "grad_norm": 0.1640625,
+      "learning_rate": 0.02,
+      "loss": 1.8257,
+      "step": 873
+    },
+    {
+      "epoch": 0.08886968237223077,
+      "grad_norm": 0.1640625,
+      "learning_rate": 0.02,
+      "loss": 1.7654,
+      "step": 874
+    },
+    {
+      "epoch": 0.08897136393100906,
+      "grad_norm": 0.173828125,
+      "learning_rate": 0.02,
+      "loss": 1.8536,
+      "step": 875
+    },
+    {
+      "epoch": 0.08907304548978735,
+      "grad_norm": 0.1650390625,
+      "learning_rate": 0.02,
+      "loss": 1.8041,
+      "step": 876
+    },
+    {
+      "epoch": 0.08917472704856566,
+      "grad_norm": 0.162109375,
+      "learning_rate": 0.02,
+      "loss": 1.7858,
+      "step": 877
+    },
+    {
+      "epoch": 0.08927640860734395,
+      "grad_norm": 0.162109375,
+      "learning_rate": 0.02,
+      "loss": 1.7797,
+      "step": 878
+    },
+    {
+      "epoch": 0.08937809016612225,
+      "grad_norm": 0.1689453125,
+      "learning_rate": 0.02,
+      "loss": 1.8342,
+      "step": 879
+    },
+    {
+      "epoch": 0.08947977172490054,
+      "grad_norm": 0.1748046875,
+      "learning_rate": 0.02,
+      "loss": 1.8047,
+      "step": 880
+    },
+    {
+      "epoch": 0.08958145328367884,
+      "grad_norm": 0.16796875,
+      "learning_rate": 0.02,
+      "loss": 1.7864,
+      "step": 881
+    },
+    {
+      "epoch": 0.08968313484245713,
+      "grad_norm": 0.1650390625,
+      "learning_rate": 0.02,
+      "loss": 1.845,
+      "step": 882
+    },
+    {
+      "epoch": 0.08978481640123544,
+      "grad_norm": 0.1650390625,
+      "learning_rate": 0.02,
+      "loss": 1.7736,
+      "step": 883
+    },
+    {
+      "epoch": 0.08988649796001373,
+      "grad_norm": 0.16796875,
+      "learning_rate": 0.02,
+      "loss": 1.7616,
+      "step": 884
+    },
+    {
+      "epoch": 0.08998817951879202,
+      "grad_norm": 0.181640625,
+      "learning_rate": 0.02,
+      "loss": 1.8677,
+      "step": 885
+    },
+    {
+      "epoch": 0.09008986107757032,
+      "grad_norm": 0.1796875,
+      "learning_rate": 0.02,
+      "loss": 1.8485,
+      "step": 886
+    },
+    {
+      "epoch": 0.09019154263634861,
+      "grad_norm": 0.1708984375,
+      "learning_rate": 0.02,
+      "loss": 1.849,
+      "step": 887
+    },
+    {
+      "epoch": 0.0902932241951269,
+      "grad_norm": 0.185546875,
+      "learning_rate": 0.02,
+      "loss": 1.7924,
+      "step": 888
+    },
+    {
+      "epoch": 0.09039490575390521,
+      "grad_norm": 0.1748046875,
+      "learning_rate": 0.02,
+      "loss": 1.8337,
+      "step": 889
+    },
+    {
+      "epoch": 0.09049658731268351,
+      "grad_norm": 0.189453125,
+      "learning_rate": 0.02,
+      "loss": 1.8768,
+      "step": 890
+    },
+    {
+      "epoch": 0.0905982688714618,
+      "grad_norm": 0.1748046875,
+      "learning_rate": 0.02,
+      "loss": 1.7679,
+      "step": 891
+    },
+    {
+      "epoch": 0.0906999504302401,
+      "grad_norm": 0.1796875,
+      "learning_rate": 0.02,
+      "loss": 1.8219,
+      "step": 892
+    },
+    {
+      "epoch": 0.09080163198901839,
+      "grad_norm": 0.17578125,
+      "learning_rate": 0.02,
+      "loss": 1.7588,
+      "step": 893
+    },
+    {
+      "epoch": 0.09090331354779668,
+      "grad_norm": 0.185546875,
+      "learning_rate": 0.02,
+      "loss": 1.8212,
+      "step": 894
+    },
+    {
+      "epoch": 0.09100499510657499,
+      "grad_norm": 0.1904296875,
+      "learning_rate": 0.02,
+      "loss": 1.8164,
+      "step": 895
+    },
+    {
+      "epoch": 0.09110667666535328,
+      "grad_norm": 0.1875,
+      "learning_rate": 0.02,
+      "loss": 1.7975,
+      "step": 896
+    },
+    {
+      "epoch": 0.09120835822413158,
+      "grad_norm": 0.1982421875,
+      "learning_rate": 0.02,
+      "loss": 1.7958,
+      "step": 897
+    },
+    {
+      "epoch": 0.09131003978290987,
+      "grad_norm": 0.1943359375,
+      "learning_rate": 0.02,
+      "loss": 1.7718,
+      "step": 898
+    },
+    {
+      "epoch": 0.09141172134168817,
+      "grad_norm": 0.1708984375,
+      "learning_rate": 0.02,
+      "loss": 1.7032,
+      "step": 899
+    },
+    {
+      "epoch": 0.09151340290046646,
+      "grad_norm": 0.1728515625,
+      "learning_rate": 0.02,
+      "loss": 1.8431,
+      "step": 900
+    },
+    {
+      "epoch": 0.09161508445924477,
+      "grad_norm": 0.1904296875,
+      "learning_rate": 0.02,
+      "loss": 1.8235,
+      "step": 901
+    },
+    {
+      "epoch": 0.09171676601802306,
+      "grad_norm": 0.1796875,
+      "learning_rate": 0.02,
+      "loss": 1.8196,
+      "step": 902
+    },
+    {
+      "epoch": 0.09181844757680135,
+      "grad_norm": 0.171875,
+      "learning_rate": 0.02,
+      "loss": 1.7137,
+      "step": 903
+    },
+    {
+      "epoch": 0.09192012913557965,
+      "grad_norm": 0.1630859375,
+      "learning_rate": 0.02,
+      "loss": 1.7716,
+      "step": 904
+    },
+    {
+      "epoch": 0.09202181069435794,
+      "grad_norm": 0.173828125,
+      "learning_rate": 0.02,
+      "loss": 1.8181,
+      "step": 905
+    },
+    {
+      "epoch": 0.09212349225313624,
+      "grad_norm": 0.185546875,
+      "learning_rate": 0.02,
+      "loss": 1.8386,
+      "step": 906
+    },
+    {
+      "epoch": 0.09222517381191454,
+      "grad_norm": 0.177734375,
+      "learning_rate": 0.02,
+      "loss": 1.7804,
+      "step": 907
+    },
+    {
+      "epoch": 0.09232685537069284,
+      "grad_norm": 0.1669921875,
+      "learning_rate": 0.02,
+      "loss": 1.7931,
+      "step": 908
+    },
+    {
+      "epoch": 0.09242853692947113,
+      "grad_norm": 0.1630859375,
+      "learning_rate": 0.02,
+      "loss": 1.7825,
+      "step": 909
+    },
+    {
+      "epoch": 0.09253021848824942,
+      "grad_norm": 0.166015625,
+      "learning_rate": 0.02,
+      "loss": 1.7829,
+      "step": 910
+    },
+    {
+      "epoch": 0.09263190004702772,
+      "grad_norm": 0.171875,
+      "learning_rate": 0.02,
+      "loss": 1.7802,
+      "step": 911
+    },
+    {
+      "epoch": 0.09273358160580601,
+      "grad_norm": 0.1728515625,
+      "learning_rate": 0.02,
+      "loss": 1.788,
+      "step": 912
+    },
+    {
+      "epoch": 0.09283526316458432,
+      "grad_norm": 0.1640625,
+      "learning_rate": 0.02,
+      "loss": 1.7972,
+      "step": 913
+    },
+    {
+      "epoch": 0.09293694472336261,
+      "grad_norm": 0.1689453125,
+      "learning_rate": 0.02,
+      "loss": 1.8285,
+      "step": 914
+    },
+    {
+      "epoch": 0.09303862628214091,
+      "grad_norm": 0.1552734375,
+      "learning_rate": 0.02,
+      "loss": 1.7813,
+      "step": 915
+    },
+    {
+      "epoch": 0.0931403078409192,
+      "grad_norm": 0.173828125,
+      "learning_rate": 0.02,
+      "loss": 1.7964,
+      "step": 916
+    },
+    {
+      "epoch": 0.0932419893996975,
+      "grad_norm": 0.171875,
+      "learning_rate": 0.02,
+      "loss": 1.85,
+      "step": 917
+    },
+    {
+      "epoch": 0.09334367095847579,
+      "grad_norm": 0.1630859375,
+      "learning_rate": 0.02,
+      "loss": 1.8029,
+      "step": 918
+    },
+    {
+      "epoch": 0.0934453525172541,
+      "grad_norm": 0.162109375,
+      "learning_rate": 0.02,
+      "loss": 1.7945,
+      "step": 919
+    },
+    {
+      "epoch": 0.09354703407603239,
+      "grad_norm": 0.1630859375,
+      "learning_rate": 0.02,
+      "loss": 1.7884,
+      "step": 920
+    },
+    {
+      "epoch": 0.09364871563481068,
+      "grad_norm": 0.18359375,
+      "learning_rate": 0.02,
+      "loss": 1.8814,
+      "step": 921
+    },
+    {
+      "epoch": 0.09375039719358898,
+      "grad_norm": 0.1767578125,
+      "learning_rate": 0.02,
+      "loss": 1.8192,
+      "step": 922
+    },
+    {
+      "epoch": 0.09385207875236727,
+      "grad_norm": 0.1728515625,
+      "learning_rate": 0.02,
+      "loss": 1.8482,
+      "step": 923
+    },
+    {
+      "epoch": 0.09395376031114556,
+      "grad_norm": 0.1748046875,
+      "learning_rate": 0.02,
+      "loss": 1.8093,
+      "step": 924
+    },
+    {
+      "epoch": 0.09405544186992387,
+      "grad_norm": 0.1962890625,
+      "learning_rate": 0.02,
+      "loss": 1.7632,
+      "step": 925
+    },
+    {
+      "epoch": 0.09415712342870217,
+      "grad_norm": 0.181640625,
+      "learning_rate": 0.02,
+      "loss": 1.798,
+      "step": 926
+    },
+    {
+      "epoch": 0.09425880498748046,
+      "grad_norm": 0.1787109375,
+      "learning_rate": 0.02,
+      "loss": 1.7879,
+      "step": 927
+    },
+    {
+      "epoch": 0.09436048654625875,
+      "grad_norm": 0.1689453125,
+      "learning_rate": 0.02,
+      "loss": 1.761,
+      "step": 928
+    },
+    {
+      "epoch": 0.09446216810503705,
+      "grad_norm": 0.1767578125,
+      "learning_rate": 0.02,
+      "loss": 1.7625,
+      "step": 929
+    },
+    {
+      "epoch": 0.09456384966381534,
+      "grad_norm": 0.1669921875,
+      "learning_rate": 0.02,
+      "loss": 1.8176,
+      "step": 930
+    },
+    {
+      "epoch": 0.09466553122259365,
+      "grad_norm": 0.16796875,
+      "learning_rate": 0.02,
+      "loss": 1.7582,
+      "step": 931
+    },
+    {
+      "epoch": 0.09476721278137194,
+      "grad_norm": 0.158203125,
+      "learning_rate": 0.02,
+      "loss": 1.7841,
+      "step": 932
+    },
+    {
+      "epoch": 0.09486889434015024,
+      "grad_norm": 0.1689453125,
+      "learning_rate": 0.02,
+      "loss": 1.786,
+      "step": 933
+    },
+    {
+      "epoch": 0.09497057589892853,
+      "grad_norm": 0.1640625,
+      "learning_rate": 0.02,
+      "loss": 1.7297,
+      "step": 934
+    },
+    {
+      "epoch": 0.09507225745770682,
+      "grad_norm": 0.1728515625,
+      "learning_rate": 0.02,
+      "loss": 1.86,
+      "step": 935
+    },
+    {
+      "epoch": 0.09517393901648512,
+      "grad_norm": 0.1767578125,
+      "learning_rate": 0.02,
+      "loss": 1.8217,
+      "step": 936
+    },
+    {
+      "epoch": 0.09527562057526343,
+      "grad_norm": 0.1591796875,
+      "learning_rate": 0.02,
+      "loss": 1.7516,
+      "step": 937
+    },
+    {
+      "epoch": 0.09537730213404172,
+      "grad_norm": 0.1708984375,
+      "learning_rate": 0.02,
+      "loss": 1.8442,
+      "step": 938
+    },
+    {
+      "epoch": 0.09547898369282001,
+      "grad_norm": 0.1865234375,
+      "learning_rate": 0.02,
+      "loss": 1.8773,
+      "step": 939
+    },
+    {
+      "epoch": 0.0955806652515983,
+      "grad_norm": 0.171875,
+      "learning_rate": 0.02,
+      "loss": 1.7955,
+      "step": 940
+    },
+    {
+      "epoch": 0.0956823468103766,
+      "grad_norm": 0.173828125,
+      "learning_rate": 0.02,
+      "loss": 1.8266,
+      "step": 941
+    },
+    {
+      "epoch": 0.0957840283691549,
+      "grad_norm": 0.1630859375,
+      "learning_rate": 0.02,
+      "loss": 1.831,
+      "step": 942
+    },
+    {
+      "epoch": 0.0958857099279332,
+      "grad_norm": 0.16015625,
+      "learning_rate": 0.02,
+      "loss": 1.7877,
+      "step": 943
+    },
+    {
+      "epoch": 0.0959873914867115,
+      "grad_norm": 0.1591796875,
+      "learning_rate": 0.02,
+      "loss": 1.7276,
+      "step": 944
+    },
+    {
+      "epoch": 0.09608907304548979,
+      "grad_norm": 0.1689453125,
+      "learning_rate": 0.02,
+      "loss": 1.8482,
+      "step": 945
+    },
+    {
+      "epoch": 0.09619075460426808,
+      "grad_norm": 0.16796875,
+      "learning_rate": 0.02,
+      "loss": 1.7766,
+      "step": 946
+    },
+    {
+      "epoch": 0.09629243616304638,
+      "grad_norm": 0.1640625,
+      "learning_rate": 0.02,
+      "loss": 1.7882,
+      "step": 947
+    },
+    {
+      "epoch": 0.09639411772182467,
+      "grad_norm": 0.171875,
+      "learning_rate": 0.02,
+      "loss": 1.755,
+      "step": 948
+    },
+    {
+      "epoch": 0.09649579928060298,
+      "grad_norm": 0.1826171875,
+      "learning_rate": 0.02,
+      "loss": 1.7958,
+      "step": 949
+    },
+    {
+      "epoch": 0.09659748083938127,
+      "grad_norm": 0.19140625,
+      "learning_rate": 0.02,
+      "loss": 1.7857,
+      "step": 950
+    },
+    {
+      "epoch": 0.09669916239815957,
+      "grad_norm": 0.1669921875,
+      "learning_rate": 0.02,
+      "loss": 1.7638,
+      "step": 951
+    },
+    {
+      "epoch": 0.09680084395693786,
+      "grad_norm": 0.181640625,
+      "learning_rate": 0.02,
+      "loss": 1.789,
+      "step": 952
+    },
+    {
+      "epoch": 0.09690252551571615,
+      "grad_norm": 0.1728515625,
+      "learning_rate": 0.02,
+      "loss": 1.8508,
+      "step": 953
+    },
+    {
+      "epoch": 0.09700420707449445,
+      "grad_norm": 0.166015625,
+      "learning_rate": 0.02,
+      "loss": 1.793,
+      "step": 954
+    },
+    {
+      "epoch": 0.09710588863327275,
+      "grad_norm": 0.154296875,
+      "learning_rate": 0.02,
+      "loss": 1.7845,
+      "step": 955
+    },
+    {
+      "epoch": 0.09720757019205105,
+      "grad_norm": 0.17578125,
+      "learning_rate": 0.02,
+      "loss": 1.7691,
+      "step": 956
+    },
+    {
+      "epoch": 0.09730925175082934,
+      "grad_norm": 0.1669921875,
+      "learning_rate": 0.02,
+      "loss": 1.7955,
+      "step": 957
+    },
+    {
+      "epoch": 0.09741093330960764,
+      "grad_norm": 0.16015625,
+      "learning_rate": 0.02,
+      "loss": 1.8115,
+      "step": 958
+    },
+    {
+      "epoch": 0.09751261486838593,
+      "grad_norm": 0.162109375,
+      "learning_rate": 0.02,
+      "loss": 1.803,
+      "step": 959
+    },
+    {
+      "epoch": 0.09761429642716422,
+      "grad_norm": 0.1728515625,
+      "learning_rate": 0.02,
+      "loss": 1.8425,
+      "step": 960
+    },
+    {
+      "epoch": 0.09771597798594252,
+      "grad_norm": 0.16796875,
+      "learning_rate": 0.02,
+      "loss": 1.7651,
+      "step": 961
+    },
+    {
+      "epoch": 0.09781765954472083,
+      "grad_norm": 0.1787109375,
+      "learning_rate": 0.02,
+      "loss": 1.8492,
+      "step": 962
+    },
+    {
+      "epoch": 0.09791934110349912,
+      "grad_norm": 0.1728515625,
+      "learning_rate": 0.02,
+      "loss": 1.7465,
+      "step": 963
+    },
+    {
+      "epoch": 0.09802102266227741,
+      "grad_norm": 0.1728515625,
+      "learning_rate": 0.02,
+      "loss": 1.7946,
+      "step": 964
+    },
+    {
+      "epoch": 0.0981227042210557,
+      "grad_norm": 0.17578125,
+      "learning_rate": 0.02,
+      "loss": 1.8125,
+      "step": 965
+    },
+    {
+      "epoch": 0.098224385779834,
+      "grad_norm": 0.17578125,
+      "learning_rate": 0.02,
+      "loss": 1.7573,
+      "step": 966
+    },
+    {
+      "epoch": 0.0983260673386123,
+      "grad_norm": 0.173828125,
+      "learning_rate": 0.02,
+      "loss": 1.8205,
+      "step": 967
+    },
+    {
+      "epoch": 0.0984277488973906,
+      "grad_norm": 0.1865234375,
+      "learning_rate": 0.02,
+      "loss": 1.7724,
+      "step": 968
+    },
+    {
+      "epoch": 0.0985294304561689,
+      "grad_norm": 0.1875,
+      "learning_rate": 0.02,
+      "loss": 1.8132,
+      "step": 969
+    },
+    {
+      "epoch": 0.09863111201494719,
+      "grad_norm": 0.1962890625,
+      "learning_rate": 0.02,
+      "loss": 1.837,
+      "step": 970
+    },
+    {
+      "epoch": 0.09873279357372548,
+      "grad_norm": 0.18359375,
+      "learning_rate": 0.02,
+      "loss": 1.7598,
+      "step": 971
+    },
+    {
+      "epoch": 0.09883447513250378,
+      "grad_norm": 0.1767578125,
+      "learning_rate": 0.02,
+      "loss": 1.8047,
+      "step": 972
+    },
+    {
+      "epoch": 0.09893615669128207,
+      "grad_norm": 0.1796875,
+      "learning_rate": 0.02,
+      "loss": 1.8728,
+      "step": 973
+    },
+    {
+      "epoch": 0.09903783825006038,
+      "grad_norm": 0.185546875,
+      "learning_rate": 0.02,
+      "loss": 1.8184,
+      "step": 974
+    },
+    {
+      "epoch": 0.09913951980883867,
+      "grad_norm": 0.1904296875,
+      "learning_rate": 0.02,
+      "loss": 1.8127,
+      "step": 975
+    },
+    {
+      "epoch": 0.09924120136761697,
+      "grad_norm": 0.169921875,
+      "learning_rate": 0.02,
+      "loss": 1.7804,
+      "step": 976
+    },
+    {
+      "epoch": 0.09934288292639526,
+      "grad_norm": 0.181640625,
+      "learning_rate": 0.02,
+      "loss": 1.8445,
+      "step": 977
+    },
+    {
+      "epoch": 0.09944456448517355,
+      "grad_norm": 0.177734375,
+      "learning_rate": 0.02,
+      "loss": 1.8376,
+      "step": 978
+    },
+    {
+      "epoch": 0.09954624604395185,
+      "grad_norm": 0.16796875,
+      "learning_rate": 0.02,
+      "loss": 1.7915,
+      "step": 979
+    },
+    {
+      "epoch": 0.09964792760273015,
+      "grad_norm": 0.1767578125,
+      "learning_rate": 0.02,
+      "loss": 1.8519,
+      "step": 980
+    },
+    {
+      "epoch": 0.09974960916150845,
+      "grad_norm": 0.1669921875,
+      "learning_rate": 0.02,
+      "loss": 1.7622,
+      "step": 981
+    },
+    {
+      "epoch": 0.09985129072028674,
+      "grad_norm": 0.1796875,
+      "learning_rate": 0.02,
+      "loss": 1.83,
+      "step": 982
+    },
+    {
+      "epoch": 0.09995297227906504,
+      "grad_norm": 0.166015625,
+      "learning_rate": 0.02,
+      "loss": 1.7567,
+      "step": 983
+    },
+    {
+      "epoch": 0.10005465383784333,
+      "grad_norm": 0.1650390625,
+      "learning_rate": 0.02,
+      "loss": 1.8012,
+      "step": 984
+    },
+    {
+      "epoch": 0.10015633539662162,
+      "grad_norm": 0.1572265625,
+      "learning_rate": 0.02,
+      "loss": 1.7905,
+      "step": 985
+    },
+    {
+      "epoch": 0.10025801695539993,
+      "grad_norm": 0.171875,
+      "learning_rate": 0.02,
+      "loss": 1.7949,
+      "step": 986
+    },
+    {
+      "epoch": 0.10035969851417822,
+      "grad_norm": 0.1787109375,
+      "learning_rate": 0.02,
+      "loss": 1.7828,
+      "step": 987
+    },
+    {
+      "epoch": 0.10046138007295652,
+      "grad_norm": 0.1923828125,
+      "learning_rate": 0.02,
+      "loss": 1.8657,
+      "step": 988
+    },
+    {
+      "epoch": 0.10056306163173481,
+      "grad_norm": 0.193359375,
+      "learning_rate": 0.02,
+      "loss": 1.8223,
+      "step": 989
+    },
+    {
+      "epoch": 0.1006647431905131,
+      "grad_norm": 0.1865234375,
+      "learning_rate": 0.02,
+      "loss": 1.7826,
+      "step": 990
+    },
+    {
+      "epoch": 0.1007664247492914,
+      "grad_norm": 0.1826171875,
+      "learning_rate": 0.02,
+      "loss": 1.7462,
+      "step": 991
+    },
+    {
+      "epoch": 0.10086810630806971,
+      "grad_norm": 0.1826171875,
+      "learning_rate": 0.02,
+      "loss": 1.8202,
+      "step": 992
+    },
+    {
+      "epoch": 0.100969787866848,
+      "grad_norm": 0.169921875,
+      "learning_rate": 0.02,
+      "loss": 1.7925,
+      "step": 993
+    },
+    {
+      "epoch": 0.1010714694256263,
+      "grad_norm": 0.181640625,
+      "learning_rate": 0.02,
+      "loss": 1.8566,
+      "step": 994
+    },
+    {
+      "epoch": 0.10117315098440459,
+      "grad_norm": 0.1904296875,
+      "learning_rate": 0.02,
+      "loss": 1.8353,
+      "step": 995
+    },
+    {
+      "epoch": 0.10127483254318288,
+      "grad_norm": 0.193359375,
+      "learning_rate": 0.02,
+      "loss": 1.8391,
+      "step": 996
+    },
+    {
+      "epoch": 0.10137651410196118,
+      "grad_norm": 0.189453125,
+      "learning_rate": 0.02,
+      "loss": 1.8983,
+      "step": 997
+    },
+    {
+      "epoch": 0.10147819566073948,
+      "grad_norm": 0.1748046875,
+      "learning_rate": 0.02,
+      "loss": 1.8108,
+      "step": 998
+    },
+    {
+      "epoch": 0.10157987721951778,
+      "grad_norm": 0.1748046875,
+      "learning_rate": 0.02,
+      "loss": 1.7916,
+      "step": 999
+    },
+    {
+      "epoch": 0.10168155877829607,
+      "grad_norm": 0.1669921875,
+      "learning_rate": 0.02,
+      "loss": 1.7679,
+      "step": 1000
+    },
+    {
+      "epoch": 0.10178324033707437,
+      "grad_norm": 0.1708984375,
+      "learning_rate": 0.02,
+      "loss": 1.8105,
+      "step": 1001
+    },
+    {
+      "epoch": 0.10188492189585266,
+      "grad_norm": 0.169921875,
+      "learning_rate": 0.02,
+      "loss": 1.847,
+      "step": 1002
+    },
+    {
+      "epoch": 0.10198660345463095,
+      "grad_norm": 0.16796875,
+      "learning_rate": 0.02,
+      "loss": 1.7991,
+      "step": 1003
+    },
+    {
+      "epoch": 0.10208828501340926,
+      "grad_norm": 0.171875,
+      "learning_rate": 0.02,
+      "loss": 1.7445,
+      "step": 1004
+    },
+    {
+      "epoch": 0.10218996657218755,
+      "grad_norm": 0.1630859375,
+      "learning_rate": 0.02,
+      "loss": 1.7129,
+      "step": 1005
+    },
+    {
+      "epoch": 0.10229164813096585,
+      "grad_norm": 0.1708984375,
+      "learning_rate": 0.02,
+      "loss": 1.8329,
+      "step": 1006
+    },
+    {
+      "epoch": 0.10239332968974414,
+      "grad_norm": 0.1796875,
+      "learning_rate": 0.02,
+      "loss": 1.7607,
+      "step": 1007
+    },
+    {
+      "epoch": 0.10249501124852244,
+      "grad_norm": 0.1787109375,
+      "learning_rate": 0.02,
+      "loss": 1.8429,
+      "step": 1008
+    },
+    {
+      "epoch": 0.10259669280730073,
+      "grad_norm": 0.16015625,
+      "learning_rate": 0.02,
+      "loss": 1.7746,
+      "step": 1009
+    },
+    {
+      "epoch": 0.10269837436607904,
+      "grad_norm": 0.1728515625,
+      "learning_rate": 0.02,
+      "loss": 1.8397,
+      "step": 1010
+    },
+    {
+      "epoch": 0.10280005592485733,
+      "grad_norm": 0.16796875,
+      "learning_rate": 0.02,
+      "loss": 1.7599,
+      "step": 1011
+    },
+    {
+      "epoch": 0.10290173748363562,
+      "grad_norm": 0.14453125,
+      "learning_rate": 0.02,
+      "loss": 1.7751,
+      "step": 1012
+    },
+    {
+      "epoch": 0.10300341904241392,
+      "grad_norm": 0.15625,
+      "learning_rate": 0.02,
+      "loss": 1.772,
+      "step": 1013
+    },
+    {
+      "epoch": 0.10310510060119221,
+      "grad_norm": 0.16015625,
+      "learning_rate": 0.02,
+      "loss": 1.7809,
+      "step": 1014
+    },
+    {
+      "epoch": 0.1032067821599705,
+      "grad_norm": 0.158203125,
+      "learning_rate": 0.02,
+      "loss": 1.7961,
+      "step": 1015
+    },
+    {
+      "epoch": 0.10330846371874881,
+      "grad_norm": 0.1611328125,
+      "learning_rate": 0.02,
+      "loss": 1.781,
+      "step": 1016
+    },
+    {
+      "epoch": 0.10341014527752711,
+      "grad_norm": 0.158203125,
+      "learning_rate": 0.02,
+      "loss": 1.7537,
+      "step": 1017
+    },
+    {
+      "epoch": 0.1035118268363054,
+      "grad_norm": 0.15625,
+      "learning_rate": 0.02,
+      "loss": 1.7569,
+      "step": 1018
+    },
+    {
+      "epoch": 0.1036135083950837,
+      "grad_norm": 0.16015625,
+      "learning_rate": 0.02,
+      "loss": 1.7635,
+      "step": 1019
+    },
+    {
+      "epoch": 0.10371518995386199,
+      "grad_norm": 0.15625,
+      "learning_rate": 0.02,
+      "loss": 1.79,
+      "step": 1020
+    },
+    {
+      "epoch": 0.10381687151264028,
+      "grad_norm": 0.1591796875,
+      "learning_rate": 0.02,
+      "loss": 1.7823,
+      "step": 1021
+    },
+    {
+      "epoch": 0.10391855307141859,
+      "grad_norm": 0.1630859375,
+      "learning_rate": 0.02,
+      "loss": 1.7923,
+      "step": 1022
+    },
+    {
+      "epoch": 0.10402023463019688,
+      "grad_norm": 0.16015625,
+      "learning_rate": 0.02,
+      "loss": 1.7438,
+      "step": 1023
+    },
+    {
+      "epoch": 0.10412191618897518,
+      "grad_norm": 0.1552734375,
+      "learning_rate": 0.02,
+      "loss": 1.814,
+      "step": 1024
+    },
+    {
+      "epoch": 0.10422359774775347,
+      "grad_norm": 0.1513671875,
+      "learning_rate": 0.02,
+      "loss": 1.8191,
+      "step": 1025
+    },
+    {
+      "epoch": 0.10432527930653177,
+      "grad_norm": 0.1455078125,
+      "learning_rate": 0.02,
+      "loss": 1.8547,
+      "step": 1026
+    },
+    {
+      "epoch": 0.10442696086531006,
+      "grad_norm": 0.1455078125,
+      "learning_rate": 0.02,
+      "loss": 1.7693,
+      "step": 1027
+    },
+    {
+      "epoch": 0.10452864242408837,
+      "grad_norm": 0.14453125,
+      "learning_rate": 0.02,
+      "loss": 1.7325,
+      "step": 1028
+    },
+    {
+      "epoch": 0.10463032398286666,
+      "grad_norm": 0.1484375,
+      "learning_rate": 0.02,
+      "loss": 1.7932,
+      "step": 1029
+    },
+    {
+      "epoch": 0.10473200554164495,
+      "grad_norm": 0.1494140625,
+      "learning_rate": 0.02,
+      "loss": 1.7675,
+      "step": 1030
+    },
+    {
+      "epoch": 0.10483368710042325,
+      "grad_norm": 0.1591796875,
+      "learning_rate": 0.02,
+      "loss": 1.8031,
+      "step": 1031
+    },
+    {
+      "epoch": 0.10493536865920154,
+      "grad_norm": 0.154296875,
+      "learning_rate": 0.02,
+      "loss": 1.7804,
+      "step": 1032
+    },
+    {
+      "epoch": 0.10503705021797984,
+      "grad_norm": 0.1689453125,
+      "learning_rate": 0.02,
+      "loss": 1.7204,
+      "step": 1033
+    },
+    {
+      "epoch": 0.10513873177675814,
+      "grad_norm": 0.1650390625,
+      "learning_rate": 0.02,
+      "loss": 1.7715,
+      "step": 1034
+    },
+    {
+      "epoch": 0.10524041333553644,
+      "grad_norm": 0.17578125,
+      "learning_rate": 0.02,
+      "loss": 1.816,
+      "step": 1035
+    },
+    {
+      "epoch": 0.10534209489431473,
+      "grad_norm": 0.1728515625,
+      "learning_rate": 0.02,
+      "loss": 1.764,
+      "step": 1036
+    },
+    {
+      "epoch": 0.10544377645309302,
+      "grad_norm": 0.1640625,
+      "learning_rate": 0.02,
+      "loss": 1.7055,
+      "step": 1037
+    },
+    {
+      "epoch": 0.10554545801187132,
+      "grad_norm": 0.1484375,
+      "learning_rate": 0.02,
+      "loss": 1.7523,
+      "step": 1038
+    },
+    {
+      "epoch": 0.10564713957064961,
+      "grad_norm": 0.1708984375,
+      "learning_rate": 0.02,
+      "loss": 1.8196,
+      "step": 1039
+    },
+    {
+      "epoch": 0.10574882112942792,
+      "grad_norm": 0.1630859375,
+      "learning_rate": 0.02,
+      "loss": 1.83,
+      "step": 1040
+    },
+    {
+      "epoch": 0.10585050268820621,
+      "grad_norm": 0.16796875,
+      "learning_rate": 0.02,
+      "loss": 1.8039,
+      "step": 1041
+    },
+    {
+      "epoch": 0.10595218424698451,
+      "grad_norm": 0.15625,
+      "learning_rate": 0.02,
+      "loss": 1.7534,
+      "step": 1042
+    },
+    {
+      "epoch": 0.1060538658057628,
+      "grad_norm": 0.16796875,
+      "learning_rate": 0.02,
+      "loss": 1.7747,
+      "step": 1043
+    },
+    {
+      "epoch": 0.1061555473645411,
+      "grad_norm": 0.1689453125,
+      "learning_rate": 0.02,
+      "loss": 1.7919,
+      "step": 1044
+    },
+    {
+      "epoch": 0.10625722892331939,
+      "grad_norm": 0.1689453125,
+      "learning_rate": 0.02,
+      "loss": 1.7642,
+      "step": 1045
+    },
+    {
+      "epoch": 0.1063589104820977,
+      "grad_norm": 0.1552734375,
+      "learning_rate": 0.02,
+      "loss": 1.7742,
+      "step": 1046
+    },
+    {
+      "epoch": 0.10646059204087599,
+      "grad_norm": 0.16796875,
+      "learning_rate": 0.02,
+      "loss": 1.7643,
+      "step": 1047
+    },
+    {
+      "epoch": 0.10656227359965428,
+      "grad_norm": 0.1611328125,
+      "learning_rate": 0.02,
+      "loss": 1.724,
+      "step": 1048
+    },
+    {
+      "epoch": 0.10666395515843258,
+      "grad_norm": 0.1689453125,
+      "learning_rate": 0.02,
+      "loss": 1.7091,
+      "step": 1049
+    },
+    {
+      "epoch": 0.10676563671721087,
+      "grad_norm": 0.171875,
+      "learning_rate": 0.02,
+      "loss": 1.8569,
+      "step": 1050
+    },
+    {
+      "epoch": 0.10686731827598916,
+      "grad_norm": 0.16015625,
+      "learning_rate": 0.02,
+      "loss": 1.7687,
+      "step": 1051
+    },
+    {
+      "epoch": 0.10696899983476747,
+      "grad_norm": 0.1611328125,
+      "learning_rate": 0.02,
+      "loss": 1.7952,
+      "step": 1052
+    },
+    {
+      "epoch": 0.10707068139354577,
+      "grad_norm": 0.158203125,
+      "learning_rate": 0.02,
+      "loss": 1.7378,
+      "step": 1053
+    },
+    {
+      "epoch": 0.10717236295232406,
+      "grad_norm": 0.1640625,
+      "learning_rate": 0.02,
+      "loss": 1.769,
+      "step": 1054
+    },
+    {
+      "epoch": 0.10727404451110235,
+      "grad_norm": 0.162109375,
+      "learning_rate": 0.02,
+      "loss": 1.774,
+      "step": 1055
+    },
+    {
+      "epoch": 0.10737572606988065,
+      "grad_norm": 0.1611328125,
+      "learning_rate": 0.02,
+      "loss": 1.8556,
+      "step": 1056
+    },
+    {
+      "epoch": 0.10747740762865894,
+      "grad_norm": 0.158203125,
+      "learning_rate": 0.02,
+      "loss": 1.7691,
+      "step": 1057
+    },
+    {
+      "epoch": 0.10757908918743725,
+      "grad_norm": 0.1630859375,
+      "learning_rate": 0.02,
+      "loss": 1.7571,
+      "step": 1058
+    },
+    {
+      "epoch": 0.10768077074621554,
+      "grad_norm": 0.166015625,
+      "learning_rate": 0.02,
+      "loss": 1.7878,
+      "step": 1059
+    },
+    {
+      "epoch": 0.10778245230499384,
+      "grad_norm": 0.1650390625,
+      "learning_rate": 0.02,
+      "loss": 1.7837,
+      "step": 1060
+    },
+    {
+      "epoch": 0.10788413386377213,
+      "grad_norm": 0.1669921875,
+      "learning_rate": 0.02,
+      "loss": 1.8254,
+      "step": 1061
+    },
+    {
+      "epoch": 0.10798581542255042,
+      "grad_norm": 0.1669921875,
+      "learning_rate": 0.02,
+      "loss": 1.7935,
+      "step": 1062
+    },
+    {
+      "epoch": 0.10808749698132872,
+      "grad_norm": 0.1630859375,
+      "learning_rate": 0.02,
+      "loss": 1.8086,
+      "step": 1063
+    },
+    {
+      "epoch": 0.10818917854010703,
+      "grad_norm": 0.15625,
+      "learning_rate": 0.02,
+      "loss": 1.749,
+      "step": 1064
+    },
+    {
+      "epoch": 0.10829086009888532,
+      "grad_norm": 0.166015625,
+      "learning_rate": 0.02,
+      "loss": 1.7908,
+      "step": 1065
+    },
+    {
+      "epoch": 0.10839254165766361,
+      "grad_norm": 0.16796875,
+      "learning_rate": 0.02,
+      "loss": 1.7457,
+      "step": 1066
+    },
+    {
+      "epoch": 0.1084942232164419,
+      "grad_norm": 0.1640625,
+      "learning_rate": 0.02,
+      "loss": 1.8241,
+      "step": 1067
+    },
+    {
+      "epoch": 0.1085959047752202,
+      "grad_norm": 0.1611328125,
+      "learning_rate": 0.02,
+      "loss": 1.6965,
+      "step": 1068
+    },
+    {
+      "epoch": 0.1086975863339985,
+      "grad_norm": 0.1650390625,
+      "learning_rate": 0.02,
+      "loss": 1.72,
+      "step": 1069
+    },
+    {
+      "epoch": 0.1087992678927768,
+      "grad_norm": 0.1591796875,
+      "learning_rate": 0.02,
+      "loss": 1.8007,
+      "step": 1070
+    },
+    {
+      "epoch": 0.1089009494515551,
+      "grad_norm": 0.1650390625,
+      "learning_rate": 0.02,
+      "loss": 1.8191,
+      "step": 1071
+    },
+    {
+      "epoch": 0.10900263101033339,
+      "grad_norm": 0.16015625,
+      "learning_rate": 0.02,
+      "loss": 1.7594,
+      "step": 1072
+    },
+    {
+      "epoch": 0.10910431256911168,
+      "grad_norm": 0.1640625,
+      "learning_rate": 0.02,
+      "loss": 1.7726,
+      "step": 1073
+    },
+    {
+      "epoch": 0.10920599412788998,
+      "grad_norm": 0.158203125,
+      "learning_rate": 0.02,
+      "loss": 1.7647,
+      "step": 1074
+    },
+    {
+      "epoch": 0.10930767568666827,
+      "grad_norm": 0.1708984375,
+      "learning_rate": 0.02,
+      "loss": 1.7941,
+      "step": 1075
+    },
+    {
+      "epoch": 0.10940935724544658,
+      "grad_norm": 0.1611328125,
+      "learning_rate": 0.02,
+      "loss": 1.8022,
+      "step": 1076
+    },
+    {
+      "epoch": 0.10951103880422487,
+      "grad_norm": 0.1650390625,
+      "learning_rate": 0.02,
+      "loss": 1.7575,
+      "step": 1077
+    },
+    {
+      "epoch": 0.10961272036300317,
+      "grad_norm": 0.1630859375,
+      "learning_rate": 0.02,
+      "loss": 1.7482,
+      "step": 1078
+    },
+    {
+      "epoch": 0.10971440192178146,
+      "grad_norm": 0.162109375,
+      "learning_rate": 0.02,
+      "loss": 1.8132,
+      "step": 1079
+    },
+    {
+      "epoch": 0.10981608348055975,
+      "grad_norm": 0.1669921875,
+      "learning_rate": 0.02,
+      "loss": 1.7811,
+      "step": 1080
+    },
+    {
+      "epoch": 0.10991776503933805,
+      "grad_norm": 0.1708984375,
+      "learning_rate": 0.02,
+      "loss": 1.7758,
+      "step": 1081
+    },
+    {
+      "epoch": 0.11001944659811635,
+      "grad_norm": 0.181640625,
+      "learning_rate": 0.02,
+      "loss": 1.8112,
+      "step": 1082
+    },
+    {
+      "epoch": 0.11012112815689465,
+      "grad_norm": 0.1689453125,
+      "learning_rate": 0.02,
+      "loss": 1.8042,
+      "step": 1083
+    },
+    {
+      "epoch": 0.11022280971567294,
+      "grad_norm": 0.17578125,
+      "learning_rate": 0.02,
+      "loss": 1.7298,
+      "step": 1084
+    },
+    {
+      "epoch": 0.11032449127445124,
+      "grad_norm": 0.166015625,
+      "learning_rate": 0.02,
+      "loss": 1.7872,
+      "step": 1085
+    },
+    {
+      "epoch": 0.11042617283322953,
+      "grad_norm": 0.1669921875,
+      "learning_rate": 0.02,
+      "loss": 1.7755,
+      "step": 1086
+    },
+    {
+      "epoch": 0.11052785439200782,
+      "grad_norm": 0.1630859375,
+      "learning_rate": 0.02,
+      "loss": 1.7783,
+      "step": 1087
+    },
+    {
+      "epoch": 0.11062953595078613,
+      "grad_norm": 0.173828125,
+      "learning_rate": 0.02,
+      "loss": 1.8241,
+      "step": 1088
+    },
+    {
+      "epoch": 0.11073121750956443,
+      "grad_norm": 0.1787109375,
+      "learning_rate": 0.02,
+      "loss": 1.7515,
+      "step": 1089
+    },
+    {
+      "epoch": 0.11083289906834272,
+      "grad_norm": 0.1806640625,
+      "learning_rate": 0.02,
+      "loss": 1.7002,
+      "step": 1090
+    },
+    {
+      "epoch": 0.11093458062712101,
+      "grad_norm": 0.1875,
+      "learning_rate": 0.02,
+      "loss": 1.7775,
+      "step": 1091
+    },
+    {
+      "epoch": 0.1110362621858993,
+      "grad_norm": 0.1875,
+      "learning_rate": 0.02,
+      "loss": 1.8097,
+      "step": 1092
+    },
+    {
+      "epoch": 0.1111379437446776,
+      "grad_norm": 0.181640625,
+      "learning_rate": 0.02,
+      "loss": 1.7664,
+      "step": 1093
+    },
+    {
+      "epoch": 0.11123962530345591,
+      "grad_norm": 0.1708984375,
+      "learning_rate": 0.02,
+      "loss": 1.7309,
+      "step": 1094
+    },
+    {
+      "epoch": 0.1113413068622342,
+      "grad_norm": 0.1708984375,
+      "learning_rate": 0.02,
+      "loss": 1.7601,
+      "step": 1095
+    },
+    {
+      "epoch": 0.1114429884210125,
+      "grad_norm": 0.19140625,
+      "learning_rate": 0.02,
+      "loss": 1.7835,
+      "step": 1096
+    },
+    {
+      "epoch": 0.11154466997979079,
+      "grad_norm": 0.1904296875,
+      "learning_rate": 0.02,
+      "loss": 1.7941,
+      "step": 1097
+    },
+    {
+      "epoch": 0.11164635153856908,
+      "grad_norm": 0.173828125,
+      "learning_rate": 0.02,
+      "loss": 1.7781,
+      "step": 1098
+    },
+    {
+      "epoch": 0.11174803309734738,
+      "grad_norm": 0.1845703125,
+      "learning_rate": 0.02,
+      "loss": 1.7912,
+      "step": 1099
+    },
+    {
+      "epoch": 0.11184971465612568,
+      "grad_norm": 0.1875,
+      "learning_rate": 0.02,
+      "loss": 1.8193,
+      "step": 1100
+    },
+    {
+      "epoch": 0.11195139621490398,
+      "grad_norm": 0.1787109375,
+      "learning_rate": 0.02,
+      "loss": 1.7886,
+      "step": 1101
+    },
+    {
+      "epoch": 0.11205307777368227,
+      "grad_norm": 0.169921875,
+      "learning_rate": 0.02,
+      "loss": 1.6915,
+      "step": 1102
+    },
+    {
+      "epoch": 0.11215475933246057,
+      "grad_norm": 0.171875,
+      "learning_rate": 0.02,
+      "loss": 1.7768,
+      "step": 1103
+    },
+    {
+      "epoch": 0.11225644089123886,
+      "grad_norm": 0.1708984375,
+      "learning_rate": 0.02,
+      "loss": 1.7633,
+      "step": 1104
+    },
+    {
+      "epoch": 0.11235812245001715,
+      "grad_norm": 0.1728515625,
+      "learning_rate": 0.02,
+      "loss": 1.8038,
+      "step": 1105
+    },
+    {
+      "epoch": 0.11245980400879546,
+      "grad_norm": 0.16796875,
+      "learning_rate": 0.02,
+      "loss": 1.8038,
+      "step": 1106
+    },
+    {
+      "epoch": 0.11256148556757375,
+      "grad_norm": 0.1611328125,
+      "learning_rate": 0.02,
+      "loss": 1.7862,
+      "step": 1107
+    },
+    {
+      "epoch": 0.11266316712635205,
+      "grad_norm": 0.1689453125,
+      "learning_rate": 0.02,
+      "loss": 1.8266,
+      "step": 1108
+    },
+    {
+      "epoch": 0.11276484868513034,
+      "grad_norm": 0.1650390625,
+      "learning_rate": 0.02,
+      "loss": 1.7586,
+      "step": 1109
+    },
+    {
+      "epoch": 0.11286653024390864,
+      "grad_norm": 0.169921875,
+      "learning_rate": 0.02,
+      "loss": 1.8271,
+      "step": 1110
+    },
+    {
+      "epoch": 0.11296821180268693,
+      "grad_norm": 0.1708984375,
+      "learning_rate": 0.02,
+      "loss": 1.8184,
+      "step": 1111
+    },
+    {
+      "epoch": 0.11306989336146524,
+      "grad_norm": 0.1826171875,
+      "learning_rate": 0.02,
+      "loss": 1.8082,
+      "step": 1112
+    },
+    {
+      "epoch": 0.11317157492024353,
+      "grad_norm": 0.185546875,
+      "learning_rate": 0.02,
+      "loss": 1.8117,
+      "step": 1113
+    },
+    {
+      "epoch": 0.11327325647902182,
+      "grad_norm": 0.1875,
+      "learning_rate": 0.02,
+      "loss": 1.8365,
+      "step": 1114
+    },
+    {
+      "epoch": 0.11337493803780012,
+      "grad_norm": 0.181640625,
+      "learning_rate": 0.02,
+      "loss": 1.795,
+      "step": 1115
+    },
+    {
+      "epoch": 0.11347661959657841,
+      "grad_norm": 0.177734375,
+      "learning_rate": 0.02,
+      "loss": 1.7282,
+      "step": 1116
+    },
+    {
+      "epoch": 0.1135783011553567,
+      "grad_norm": 0.171875,
+      "learning_rate": 0.02,
+      "loss": 1.774,
+      "step": 1117
+    },
+    {
+      "epoch": 0.11367998271413501,
+      "grad_norm": 0.16015625,
+      "learning_rate": 0.02,
+      "loss": 1.7359,
+      "step": 1118
+    },
+    {
+      "epoch": 0.11378166427291331,
+      "grad_norm": 0.171875,
+      "learning_rate": 0.02,
+      "loss": 1.8277,
+      "step": 1119
+    },
+    {
+      "epoch": 0.1138833458316916,
+      "grad_norm": 0.16796875,
+      "learning_rate": 0.02,
+      "loss": 1.8568,
+      "step": 1120
+    },
+    {
+      "epoch": 0.1139850273904699,
+      "grad_norm": 0.1728515625,
+      "learning_rate": 0.02,
+      "loss": 1.8186,
+      "step": 1121
+    },
+    {
+      "epoch": 0.11408670894924819,
+      "grad_norm": 0.1533203125,
+      "learning_rate": 0.02,
+      "loss": 1.764,
+      "step": 1122
+    },
+    {
+      "epoch": 0.11418839050802648,
+      "grad_norm": 0.173828125,
+      "learning_rate": 0.02,
+      "loss": 1.772,
+      "step": 1123
+    },
+    {
+      "epoch": 0.11429007206680479,
+      "grad_norm": 0.1728515625,
+      "learning_rate": 0.02,
+      "loss": 1.8044,
+      "step": 1124
+    },
+    {
+      "epoch": 0.11439175362558308,
+      "grad_norm": 0.173828125,
+      "learning_rate": 0.02,
+      "loss": 1.8177,
+      "step": 1125
+    },
+    {
+      "epoch": 0.11449343518436138,
+      "grad_norm": 0.1689453125,
+      "learning_rate": 0.02,
+      "loss": 1.8175,
+      "step": 1126
+    },
+    {
+      "epoch": 0.11459511674313967,
+      "grad_norm": 0.1806640625,
+      "learning_rate": 0.02,
+      "loss": 1.8023,
+      "step": 1127
+    },
+    {
+      "epoch": 0.11469679830191797,
+      "grad_norm": 0.17578125,
+      "learning_rate": 0.02,
+      "loss": 1.7754,
+      "step": 1128
+    },
+    {
+      "epoch": 0.11479847986069626,
+      "grad_norm": 0.185546875,
+      "learning_rate": 0.02,
+      "loss": 1.7804,
+      "step": 1129
+    },
+    {
+      "epoch": 0.11490016141947457,
+      "grad_norm": 0.18359375,
+      "learning_rate": 0.02,
+      "loss": 1.8124,
+      "step": 1130
+    },
+    {
+      "epoch": 0.11500184297825286,
+      "grad_norm": 0.173828125,
+      "learning_rate": 0.02,
+      "loss": 1.7606,
+      "step": 1131
+    },
+    {
+      "epoch": 0.11510352453703115,
+      "grad_norm": 0.177734375,
+      "learning_rate": 0.02,
+      "loss": 1.8351,
+      "step": 1132
+    },
+    {
+      "epoch": 0.11520520609580945,
+      "grad_norm": 0.177734375,
+      "learning_rate": 0.02,
+      "loss": 1.7849,
+      "step": 1133
+    },
+    {
+      "epoch": 0.11530688765458774,
+      "grad_norm": 0.1689453125,
+      "learning_rate": 0.02,
+      "loss": 1.7514,
+      "step": 1134
+    },
+    {
+      "epoch": 0.11540856921336604,
+      "grad_norm": 0.1572265625,
+      "learning_rate": 0.02,
+      "loss": 1.7363,
+      "step": 1135
+    },
+    {
+      "epoch": 0.11551025077214434,
+      "grad_norm": 0.171875,
+      "learning_rate": 0.02,
+      "loss": 1.7347,
+      "step": 1136
+    },
+    {
+      "epoch": 0.11561193233092264,
+      "grad_norm": 0.16015625,
+      "learning_rate": 0.02,
+      "loss": 1.7926,
+      "step": 1137
+    },
+    {
+      "epoch": 0.11571361388970093,
+      "grad_norm": 0.173828125,
+      "learning_rate": 0.02,
+      "loss": 1.7948,
+      "step": 1138
+    },
+    {
+      "epoch": 0.11581529544847922,
+      "grad_norm": 0.16796875,
+      "learning_rate": 0.02,
+      "loss": 1.7986,
+      "step": 1139
+    },
+    {
+      "epoch": 0.11591697700725752,
+      "grad_norm": 0.169921875,
+      "learning_rate": 0.02,
+      "loss": 1.7657,
+      "step": 1140
+    },
+    {
+      "epoch": 0.11601865856603581,
+      "grad_norm": 0.1669921875,
+      "learning_rate": 0.02,
+      "loss": 1.752,
+      "step": 1141
+    },
+    {
+      "epoch": 0.11612034012481412,
+      "grad_norm": 0.1640625,
+      "learning_rate": 0.02,
+      "loss": 1.7653,
+      "step": 1142
+    },
+    {
+      "epoch": 0.11622202168359241,
+      "grad_norm": 0.1572265625,
+      "learning_rate": 0.02,
+      "loss": 1.801,
+      "step": 1143
+    },
+    {
+      "epoch": 0.11632370324237071,
+      "grad_norm": 0.171875,
+      "learning_rate": 0.02,
+      "loss": 1.8259,
+      "step": 1144
+    },
+    {
+      "epoch": 0.116425384801149,
+      "grad_norm": 0.1611328125,
+      "learning_rate": 0.02,
+      "loss": 1.6915,
+      "step": 1145
+    },
+    {
+      "epoch": 0.1165270663599273,
+      "grad_norm": 0.1611328125,
+      "learning_rate": 0.02,
+      "loss": 1.7973,
+      "step": 1146
+    },
+    {
+      "epoch": 0.11662874791870559,
+      "grad_norm": 0.166015625,
+      "learning_rate": 0.02,
+      "loss": 1.7581,
+      "step": 1147
+    },
+    {
+      "epoch": 0.1167304294774839,
+      "grad_norm": 0.15234375,
+      "learning_rate": 0.02,
+      "loss": 1.7106,
+      "step": 1148
+    },
+    {
+      "epoch": 0.11683211103626219,
+      "grad_norm": 0.1640625,
+      "learning_rate": 0.02,
+      "loss": 1.8247,
+      "step": 1149
+    },
+    {
+      "epoch": 0.11693379259504048,
+      "grad_norm": 0.1591796875,
+      "learning_rate": 0.02,
+      "loss": 1.7923,
+      "step": 1150
+    },
+    {
+      "epoch": 0.11703547415381878,
+      "grad_norm": 0.15625,
+      "learning_rate": 0.02,
+      "loss": 1.7563,
+      "step": 1151
+    },
+    {
+      "epoch": 0.11713715571259707,
+      "grad_norm": 0.1708984375,
+      "learning_rate": 0.02,
+      "loss": 1.8211,
+      "step": 1152
+    },
+    {
+      "epoch": 0.11723883727137537,
+      "grad_norm": 0.1728515625,
+      "learning_rate": 0.02,
+      "loss": 1.7994,
+      "step": 1153
+    },
+    {
+      "epoch": 0.11734051883015367,
+      "grad_norm": 0.1630859375,
+      "learning_rate": 0.02,
+      "loss": 1.7964,
+      "step": 1154
+    },
+    {
+      "epoch": 0.11744220038893197,
+      "grad_norm": 0.1640625,
+      "learning_rate": 0.02,
+      "loss": 1.7692,
+      "step": 1155
+    },
+    {
+      "epoch": 0.11754388194771026,
+      "grad_norm": 0.171875,
+      "learning_rate": 0.02,
+      "loss": 1.7474,
+      "step": 1156
+    },
+    {
+      "epoch": 0.11764556350648855,
+      "grad_norm": 0.1884765625,
+      "learning_rate": 0.02,
+      "loss": 1.8117,
+      "step": 1157
+    },
+    {
+      "epoch": 0.11774724506526685,
+      "grad_norm": 0.1748046875,
+      "learning_rate": 0.02,
+      "loss": 1.7315,
+      "step": 1158
+    },
+    {
+      "epoch": 0.11784892662404514,
+      "grad_norm": 0.1669921875,
+      "learning_rate": 0.02,
+      "loss": 1.7405,
+      "step": 1159
+    },
+    {
+      "epoch": 0.11795060818282345,
+      "grad_norm": 0.1708984375,
+      "learning_rate": 0.02,
+      "loss": 1.7342,
+      "step": 1160
+    },
+    {
+      "epoch": 0.11805228974160174,
+      "grad_norm": 0.185546875,
+      "learning_rate": 0.02,
+      "loss": 1.8057,
+      "step": 1161
+    },
+    {
+      "epoch": 0.11815397130038004,
+      "grad_norm": 0.18359375,
+      "learning_rate": 0.02,
+      "loss": 1.727,
+      "step": 1162
+    },
+    {
+      "epoch": 0.11825565285915833,
+      "grad_norm": 0.1689453125,
+      "learning_rate": 0.02,
+      "loss": 1.7222,
+      "step": 1163
+    },
+    {
+      "epoch": 0.11835733441793662,
+      "grad_norm": 0.177734375,
+      "learning_rate": 0.02,
+      "loss": 1.8001,
+      "step": 1164
+    },
+    {
+      "epoch": 0.11845901597671492,
+      "grad_norm": 0.1728515625,
+      "learning_rate": 0.02,
+      "loss": 1.719,
+      "step": 1165
+    },
+    {
+      "epoch": 0.11856069753549323,
+      "grad_norm": 0.16796875,
+      "learning_rate": 0.02,
+      "loss": 1.7512,
+      "step": 1166
+    },
+    {
+      "epoch": 0.11866237909427152,
+      "grad_norm": 0.166015625,
+      "learning_rate": 0.02,
+      "loss": 1.765,
+      "step": 1167
+    },
+    {
+      "epoch": 0.11876406065304981,
+      "grad_norm": 0.1591796875,
+      "learning_rate": 0.02,
+      "loss": 1.6882,
+      "step": 1168
+    },
+    {
+      "epoch": 0.11886574221182811,
+      "grad_norm": 0.158203125,
+      "learning_rate": 0.02,
+      "loss": 1.8208,
+      "step": 1169
+    },
+    {
+      "epoch": 0.1189674237706064,
+      "grad_norm": 0.162109375,
+      "learning_rate": 0.02,
+      "loss": 1.8008,
+      "step": 1170
+    },
+    {
+      "epoch": 0.1190691053293847,
+      "grad_norm": 0.15234375,
+      "learning_rate": 0.02,
+      "loss": 1.7341,
+      "step": 1171
+    },
+    {
+      "epoch": 0.119170786888163,
+      "grad_norm": 0.1533203125,
+      "learning_rate": 0.02,
+      "loss": 1.802,
+      "step": 1172
+    },
+    {
+      "epoch": 0.1192724684469413,
+      "grad_norm": 0.1630859375,
+      "learning_rate": 0.02,
+      "loss": 1.74,
+      "step": 1173
+    },
+    {
+      "epoch": 0.11937415000571959,
+      "grad_norm": 0.15234375,
+      "learning_rate": 0.02,
+      "loss": 1.8017,
+      "step": 1174
+    },
+    {
+      "epoch": 0.11947583156449788,
+      "grad_norm": 0.16015625,
+      "learning_rate": 0.02,
+      "loss": 1.7054,
+      "step": 1175
+    },
+    {
+      "epoch": 0.11957751312327618,
+      "grad_norm": 0.16015625,
+      "learning_rate": 0.02,
+      "loss": 1.765,
+      "step": 1176
+    },
+    {
+      "epoch": 0.11967919468205447,
+      "grad_norm": 0.1640625,
+      "learning_rate": 0.02,
+      "loss": 1.7601,
+      "step": 1177
+    },
+    {
+      "epoch": 0.11978087624083278,
+      "grad_norm": 0.1611328125,
+      "learning_rate": 0.02,
+      "loss": 1.7667,
+      "step": 1178
+    },
+    {
+      "epoch": 0.11988255779961107,
+      "grad_norm": 0.1708984375,
+      "learning_rate": 0.02,
+      "loss": 1.7817,
+      "step": 1179
+    },
+    {
+      "epoch": 0.11998423935838937,
+      "grad_norm": 0.1630859375,
+      "learning_rate": 0.02,
+      "loss": 1.7647,
+      "step": 1180
+    },
+    {
+      "epoch": 0.12008592091716766,
+      "grad_norm": 0.1630859375,
+      "learning_rate": 0.02,
+      "loss": 1.7701,
+      "step": 1181
+    },
+    {
+      "epoch": 0.12018760247594595,
+      "grad_norm": 0.1669921875,
+      "learning_rate": 0.02,
+      "loss": 1.7251,
+      "step": 1182
+    },
+    {
+      "epoch": 0.12028928403472425,
+      "grad_norm": 0.154296875,
+      "learning_rate": 0.02,
+      "loss": 1.7622,
+      "step": 1183
+    },
+    {
+      "epoch": 0.12039096559350256,
+      "grad_norm": 0.1611328125,
+      "learning_rate": 0.02,
+      "loss": 1.8365,
+      "step": 1184
+    },
+    {
+      "epoch": 0.12049264715228085,
+      "grad_norm": 0.1591796875,
+      "learning_rate": 0.02,
+      "loss": 1.812,
+      "step": 1185
+    },
+    {
+      "epoch": 0.12059432871105914,
+      "grad_norm": 0.146484375,
+      "learning_rate": 0.02,
+      "loss": 1.7464,
+      "step": 1186
+    },
+    {
+      "epoch": 0.12069601026983744,
+      "grad_norm": 0.142578125,
+      "learning_rate": 0.02,
+      "loss": 1.7302,
+      "step": 1187
+    },
+    {
+      "epoch": 0.12079769182861573,
+      "grad_norm": 0.1416015625,
+      "learning_rate": 0.02,
+      "loss": 1.733,
+      "step": 1188
+    },
+    {
+      "epoch": 0.12089937338739402,
+      "grad_norm": 0.14453125,
+      "learning_rate": 0.02,
+      "loss": 1.7314,
+      "step": 1189
+    },
+    {
+      "epoch": 0.12100105494617233,
+      "grad_norm": 0.1484375,
+      "learning_rate": 0.02,
+      "loss": 1.6839,
+      "step": 1190
+    },
+    {
+      "epoch": 0.12110273650495063,
+      "grad_norm": 0.1484375,
+      "learning_rate": 0.02,
+      "loss": 1.6929,
+      "step": 1191
+    },
+    {
+      "epoch": 0.12120441806372892,
+      "grad_norm": 0.1513671875,
+      "learning_rate": 0.02,
+      "loss": 1.7416,
+      "step": 1192
+    },
+    {
+      "epoch": 0.12130609962250721,
+      "grad_norm": 0.146484375,
+      "learning_rate": 0.02,
+      "loss": 1.6873,
+      "step": 1193
+    },
+    {
+      "epoch": 0.1214077811812855,
+      "grad_norm": 0.1572265625,
+      "learning_rate": 0.02,
+      "loss": 1.7268,
+      "step": 1194
+    },
+    {
+      "epoch": 0.1215094627400638,
+      "grad_norm": 0.162109375,
+      "learning_rate": 0.02,
+      "loss": 1.7338,
+      "step": 1195
+    },
+    {
+      "epoch": 0.1216111442988421,
+      "grad_norm": 0.171875,
+      "learning_rate": 0.02,
+      "loss": 1.836,
+      "step": 1196
+    },
+    {
+      "epoch": 0.1217128258576204,
+      "grad_norm": 0.1640625,
+      "learning_rate": 0.02,
+      "loss": 1.7663,
+      "step": 1197
+    },
+    {
+      "epoch": 0.1218145074163987,
+      "grad_norm": 0.1630859375,
+      "learning_rate": 0.02,
+      "loss": 1.807,
+      "step": 1198
+    },
+    {
+      "epoch": 0.12191618897517699,
+      "grad_norm": 0.1640625,
+      "learning_rate": 0.02,
+      "loss": 1.7463,
+      "step": 1199
+    },
+    {
+      "epoch": 0.12201787053395528,
+      "grad_norm": 0.1708984375,
+      "learning_rate": 0.02,
+      "loss": 1.8054,
+      "step": 1200
+    },
+    {
+      "epoch": 0.12211955209273358,
+      "grad_norm": 0.166015625,
+      "learning_rate": 0.02,
+      "loss": 1.7266,
+      "step": 1201
+    },
+    {
+      "epoch": 0.12222123365151187,
+      "grad_norm": 0.1669921875,
+      "learning_rate": 0.02,
+      "loss": 1.7968,
+      "step": 1202
+    },
+    {
+      "epoch": 0.12232291521029018,
+      "grad_norm": 0.1689453125,
+      "learning_rate": 0.02,
+      "loss": 1.7131,
+      "step": 1203
+    },
+    {
+      "epoch": 0.12242459676906847,
+      "grad_norm": 0.1611328125,
+      "learning_rate": 0.02,
+      "loss": 1.8452,
+      "step": 1204
+    },
+    {
+      "epoch": 0.12252627832784677,
+      "grad_norm": 0.1669921875,
+      "learning_rate": 0.02,
+      "loss": 1.8087,
+      "step": 1205
+    },
+    {
+      "epoch": 0.12262795988662506,
+      "grad_norm": 0.1640625,
+      "learning_rate": 0.02,
+      "loss": 1.7691,
+      "step": 1206
+    },
+    {
+      "epoch": 0.12272964144540335,
+      "grad_norm": 0.162109375,
+      "learning_rate": 0.02,
+      "loss": 1.7694,
+      "step": 1207
+    },
+    {
+      "epoch": 0.12283132300418165,
+      "grad_norm": 0.16015625,
+      "learning_rate": 0.02,
+      "loss": 1.7805,
+      "step": 1208
+    },
+    {
+      "epoch": 0.12293300456295995,
+      "grad_norm": 0.1630859375,
+      "learning_rate": 0.02,
+      "loss": 1.7803,
+      "step": 1209
+    },
+    {
+      "epoch": 0.12303468612173825,
+      "grad_norm": 0.1591796875,
+      "learning_rate": 0.02,
+      "loss": 1.7658,
+      "step": 1210
+    },
+    {
+      "epoch": 0.12313636768051654,
+      "grad_norm": 0.1650390625,
+      "learning_rate": 0.02,
+      "loss": 1.7674,
+      "step": 1211
+    },
+    {
+      "epoch": 0.12323804923929484,
+      "grad_norm": 0.1650390625,
+      "learning_rate": 0.02,
+      "loss": 1.7158,
+      "step": 1212
+    },
+    {
+      "epoch": 0.12333973079807313,
+      "grad_norm": 0.166015625,
+      "learning_rate": 0.02,
+      "loss": 1.7606,
+      "step": 1213
+    },
+    {
+      "epoch": 0.12344141235685142,
+      "grad_norm": 0.1640625,
+      "learning_rate": 0.02,
+      "loss": 1.7539,
+      "step": 1214
+    },
+    {
+      "epoch": 0.12354309391562973,
+      "grad_norm": 0.1611328125,
+      "learning_rate": 0.02,
+      "loss": 1.7732,
+      "step": 1215
+    },
+    {
+      "epoch": 0.12364477547440803,
+      "grad_norm": 0.1650390625,
+      "learning_rate": 0.02,
+      "loss": 1.7843,
+      "step": 1216
+    },
+    {
+      "epoch": 0.12374645703318632,
+      "grad_norm": 0.169921875,
+      "learning_rate": 0.02,
+      "loss": 1.8134,
+      "step": 1217
+    },
+    {
+      "epoch": 0.12384813859196461,
+      "grad_norm": 0.1640625,
+      "learning_rate": 0.02,
+      "loss": 1.7009,
+      "step": 1218
+    },
+    {
+      "epoch": 0.1239498201507429,
+      "grad_norm": 0.1748046875,
+      "learning_rate": 0.02,
+      "loss": 1.8168,
+      "step": 1219
+    },
+    {
+      "epoch": 0.1240515017095212,
+      "grad_norm": 0.171875,
+      "learning_rate": 0.02,
+      "loss": 1.7382,
+      "step": 1220
+    },
+    {
+      "epoch": 0.12415318326829951,
+      "grad_norm": 0.1650390625,
+      "learning_rate": 0.02,
+      "loss": 1.7442,
+      "step": 1221
+    },
+    {
+      "epoch": 0.1242548648270778,
+      "grad_norm": 0.177734375,
+      "learning_rate": 0.02,
+      "loss": 1.8461,
+      "step": 1222
+    },
+    {
+      "epoch": 0.1243565463858561,
+      "grad_norm": 0.171875,
+      "learning_rate": 0.02,
+      "loss": 1.7661,
+      "step": 1223
+    },
+    {
+      "epoch": 0.12445822794463439,
+      "grad_norm": 0.1708984375,
+      "learning_rate": 0.02,
+      "loss": 1.762,
+      "step": 1224
+    },
+    {
+      "epoch": 0.12455990950341268,
+      "grad_norm": 0.1650390625,
+      "learning_rate": 0.02,
+      "loss": 1.7302,
+      "step": 1225
+    },
+    {
+      "epoch": 0.12466159106219098,
+      "grad_norm": 0.169921875,
+      "learning_rate": 0.02,
+      "loss": 1.7769,
+      "step": 1226
+    },
+    {
+      "epoch": 0.12476327262096928,
+      "grad_norm": 0.1708984375,
+      "learning_rate": 0.02,
+      "loss": 1.7469,
+      "step": 1227
+    },
+    {
+      "epoch": 0.12486495417974758,
+      "grad_norm": 0.1640625,
+      "learning_rate": 0.02,
+      "loss": 1.7486,
+      "step": 1228
+    },
+    {
+      "epoch": 0.12496663573852587,
+      "grad_norm": 0.1630859375,
+      "learning_rate": 0.02,
+      "loss": 1.7501,
+      "step": 1229
+    },
+    {
+      "epoch": 0.12506831729730417,
+      "grad_norm": 0.1572265625,
+      "learning_rate": 0.02,
+      "loss": 1.7341,
+      "step": 1230
+    },
+    {
+      "epoch": 0.12516999885608246,
+      "grad_norm": 0.15625,
+      "learning_rate": 0.02,
+      "loss": 1.7453,
+      "step": 1231
+    },
+    {
+      "epoch": 0.12527168041486075,
+      "grad_norm": 0.1640625,
+      "learning_rate": 0.02,
+      "loss": 1.7882,
+      "step": 1232
+    },
+    {
+      "epoch": 0.12537336197363905,
+      "grad_norm": 0.16015625,
+      "learning_rate": 0.02,
+      "loss": 1.7361,
+      "step": 1233
+    },
+    {
+      "epoch": 0.12547504353241734,
+      "grad_norm": 0.1611328125,
+      "learning_rate": 0.02,
+      "loss": 1.7391,
+      "step": 1234
+    },
+    {
+      "epoch": 0.12557672509119563,
+      "grad_norm": 0.16015625,
+      "learning_rate": 0.02,
+      "loss": 1.7482,
+      "step": 1235
+    },
+    {
+      "epoch": 0.12567840664997396,
+      "grad_norm": 0.1650390625,
+      "learning_rate": 0.02,
+      "loss": 1.7512,
+      "step": 1236
+    },
+    {
+      "epoch": 0.12578008820875225,
+      "grad_norm": 0.16015625,
+      "learning_rate": 0.02,
+      "loss": 1.7942,
+      "step": 1237
+    },
+    {
+      "epoch": 0.12588176976753054,
+      "grad_norm": 0.1787109375,
+      "learning_rate": 0.02,
+      "loss": 1.8066,
+      "step": 1238
+    },
+    {
+      "epoch": 0.12598345132630884,
+      "grad_norm": 0.1708984375,
+      "learning_rate": 0.02,
+      "loss": 1.7552,
+      "step": 1239
+    },
+    {
+      "epoch": 0.12608513288508713,
+      "grad_norm": 0.1650390625,
+      "learning_rate": 0.02,
+      "loss": 1.6601,
+      "step": 1240
+    },
+    {
+      "epoch": 0.12618681444386542,
+      "grad_norm": 0.1591796875,
+      "learning_rate": 0.02,
+      "loss": 1.8199,
+      "step": 1241
+    },
+    {
+      "epoch": 0.12628849600264372,
+      "grad_norm": 0.171875,
+      "learning_rate": 0.02,
+      "loss": 1.8033,
+      "step": 1242
+    },
+    {
+      "epoch": 0.126390177561422,
+      "grad_norm": 0.154296875,
+      "learning_rate": 0.02,
+      "loss": 1.708,
+      "step": 1243
+    },
+    {
+      "epoch": 0.1264918591202003,
+      "grad_norm": 0.166015625,
+      "learning_rate": 0.02,
+      "loss": 1.7213,
+      "step": 1244
+    },
+    {
+      "epoch": 0.1265935406789786,
+      "grad_norm": 0.1669921875,
+      "learning_rate": 0.02,
+      "loss": 1.7336,
+      "step": 1245
+    },
+    {
+      "epoch": 0.1266952222377569,
+      "grad_norm": 0.1630859375,
+      "learning_rate": 0.02,
+      "loss": 1.7938,
+      "step": 1246
+    },
+    {
+      "epoch": 0.1267969037965352,
+      "grad_norm": 0.1611328125,
+      "learning_rate": 0.02,
+      "loss": 1.7526,
+      "step": 1247
+    },
+    {
+      "epoch": 0.1268985853553135,
+      "grad_norm": 0.16015625,
+      "learning_rate": 0.02,
+      "loss": 1.6914,
+      "step": 1248
+    },
+    {
+      "epoch": 0.1270002669140918,
+      "grad_norm": 0.169921875,
+      "learning_rate": 0.02,
+      "loss": 1.7773,
+      "step": 1249
+    },
+    {
+      "epoch": 0.1271019484728701,
+      "grad_norm": 0.1767578125,
+      "learning_rate": 0.02,
+      "loss": 1.7379,
+      "step": 1250
+    },
+    {
+      "epoch": 0.1272036300316484,
+      "grad_norm": 0.16796875,
+      "learning_rate": 0.02,
+      "loss": 1.756,
+      "step": 1251
+    },
+    {
+      "epoch": 0.12730531159042668,
+      "grad_norm": 0.1669921875,
+      "learning_rate": 0.02,
+      "loss": 1.7439,
+      "step": 1252
+    },
+    {
+      "epoch": 0.12740699314920498,
+      "grad_norm": 0.162109375,
+      "learning_rate": 0.02,
+      "loss": 1.7325,
+      "step": 1253
+    },
+    {
+      "epoch": 0.12750867470798327,
+      "grad_norm": 0.1708984375,
+      "learning_rate": 0.02,
+      "loss": 1.8304,
+      "step": 1254
+    },
+    {
+      "epoch": 0.12761035626676157,
+      "grad_norm": 0.1591796875,
+      "learning_rate": 0.02,
+      "loss": 1.8108,
+      "step": 1255
+    },
+    {
+      "epoch": 0.12771203782553986,
+      "grad_norm": 0.1640625,
+      "learning_rate": 0.02,
+      "loss": 1.7706,
+      "step": 1256
+    },
+    {
+      "epoch": 0.12781371938431815,
+      "grad_norm": 0.1845703125,
+      "learning_rate": 0.02,
+      "loss": 1.7721,
+      "step": 1257
+    },
+    {
+      "epoch": 0.12791540094309645,
+      "grad_norm": 0.1513671875,
+      "learning_rate": 0.02,
+      "loss": 1.716,
+      "step": 1258
+    },
+    {
+      "epoch": 0.12801708250187474,
+      "grad_norm": 0.158203125,
+      "learning_rate": 0.02,
+      "loss": 1.8283,
+      "step": 1259
+    },
+    {
+      "epoch": 0.12811876406065306,
+      "grad_norm": 0.171875,
+      "learning_rate": 0.02,
+      "loss": 1.7438,
+      "step": 1260
+    },
+    {
+      "epoch": 0.12822044561943136,
+      "grad_norm": 0.162109375,
+      "learning_rate": 0.02,
+      "loss": 1.7609,
+      "step": 1261
+    },
+    {
+      "epoch": 0.12832212717820965,
+      "grad_norm": 0.1708984375,
+      "learning_rate": 0.02,
+      "loss": 1.7445,
+      "step": 1262
+    },
+    {
+      "epoch": 0.12842380873698794,
+      "grad_norm": 0.171875,
+      "learning_rate": 0.02,
+      "loss": 1.7293,
+      "step": 1263
+    },
+    {
+      "epoch": 0.12852549029576624,
+      "grad_norm": 0.1796875,
+      "learning_rate": 0.02,
+      "loss": 1.7813,
+      "step": 1264
+    },
+    {
+      "epoch": 0.12862717185454453,
+      "grad_norm": 0.1748046875,
+      "learning_rate": 0.02,
+      "loss": 1.7922,
+      "step": 1265
+    },
+    {
+      "epoch": 0.12872885341332282,
+      "grad_norm": 0.1669921875,
+      "learning_rate": 0.02,
+      "loss": 1.7412,
+      "step": 1266
+    },
+    {
+      "epoch": 0.12883053497210112,
+      "grad_norm": 0.1533203125,
+      "learning_rate": 0.02,
+      "loss": 1.7378,
+      "step": 1267
+    },
+    {
+      "epoch": 0.1289322165308794,
+      "grad_norm": 0.1708984375,
+      "learning_rate": 0.02,
+      "loss": 1.7551,
+      "step": 1268
+    },
+    {
+      "epoch": 0.1290338980896577,
+      "grad_norm": 0.1728515625,
+      "learning_rate": 0.02,
+      "loss": 1.779,
+      "step": 1269
+    },
+    {
+      "epoch": 0.129135579648436,
+      "grad_norm": 0.17578125,
+      "learning_rate": 0.02,
+      "loss": 1.7811,
+      "step": 1270
+    },
+    {
+      "epoch": 0.1292372612072143,
+      "grad_norm": 0.1640625,
+      "learning_rate": 0.02,
+      "loss": 1.7487,
+      "step": 1271
+    },
+    {
+      "epoch": 0.12933894276599261,
+      "grad_norm": 0.1669921875,
+      "learning_rate": 0.02,
+      "loss": 1.6837,
+      "step": 1272
+    },
+    {
+      "epoch": 0.1294406243247709,
+      "grad_norm": 0.1767578125,
+      "learning_rate": 0.02,
+      "loss": 1.7726,
+      "step": 1273
+    },
+    {
+      "epoch": 0.1295423058835492,
+      "grad_norm": 0.1787109375,
+      "learning_rate": 0.02,
+      "loss": 1.7636,
+      "step": 1274
+    },
+    {
+      "epoch": 0.1296439874423275,
+      "grad_norm": 0.1748046875,
+      "learning_rate": 0.02,
+      "loss": 1.7639,
+      "step": 1275
+    },
+    {
+      "epoch": 0.1297456690011058,
+      "grad_norm": 0.166015625,
+      "learning_rate": 0.02,
+      "loss": 1.817,
+      "step": 1276
+    },
+    {
+      "epoch": 0.12984735055988408,
+      "grad_norm": 0.1630859375,
+      "learning_rate": 0.02,
+      "loss": 1.763,
+      "step": 1277
+    },
+    {
+      "epoch": 0.12994903211866238,
+      "grad_norm": 0.1611328125,
+      "learning_rate": 0.02,
+      "loss": 1.7636,
+      "step": 1278
+    },
+    {
+      "epoch": 0.13005071367744067,
+      "grad_norm": 0.16015625,
+      "learning_rate": 0.02,
+      "loss": 1.7346,
+      "step": 1279
+    },
+    {
+      "epoch": 0.13015239523621897,
+      "grad_norm": 0.154296875,
+      "learning_rate": 0.02,
+      "loss": 1.7621,
+      "step": 1280
+    },
+    {
+      "epoch": 0.13025407679499726,
+      "grad_norm": 0.1611328125,
+      "learning_rate": 0.02,
+      "loss": 1.7249,
+      "step": 1281
+    },
+    {
+      "epoch": 0.13035575835377555,
+      "grad_norm": 0.1552734375,
+      "learning_rate": 0.02,
+      "loss": 1.7408,
+      "step": 1282
+    },
+    {
+      "epoch": 0.13045743991255385,
+      "grad_norm": 0.154296875,
+      "learning_rate": 0.02,
+      "loss": 1.7338,
+      "step": 1283
+    },
+    {
+      "epoch": 0.13055912147133217,
+      "grad_norm": 0.1552734375,
+      "learning_rate": 0.02,
+      "loss": 1.7501,
+      "step": 1284
+    },
+    {
+      "epoch": 0.13066080303011046,
+      "grad_norm": 0.1572265625,
+      "learning_rate": 0.02,
+      "loss": 1.7303,
+      "step": 1285
+    },
+    {
+      "epoch": 0.13076248458888876,
+      "grad_norm": 0.1533203125,
+      "learning_rate": 0.02,
+      "loss": 1.7866,
+      "step": 1286
+    },
+    {
+      "epoch": 0.13086416614766705,
+      "grad_norm": 0.1650390625,
+      "learning_rate": 0.02,
+      "loss": 1.7747,
+      "step": 1287
+    },
+    {
+      "epoch": 0.13096584770644534,
+      "grad_norm": 0.15625,
+      "learning_rate": 0.02,
+      "loss": 1.8164,
+      "step": 1288
+    },
+    {
+      "epoch": 0.13106752926522364,
+      "grad_norm": 0.1591796875,
+      "learning_rate": 0.02,
+      "loss": 1.7105,
+      "step": 1289
+    },
+    {
+      "epoch": 0.13116921082400193,
+      "grad_norm": 0.1533203125,
+      "learning_rate": 0.02,
+      "loss": 1.7104,
+      "step": 1290
+    },
+    {
+      "epoch": 0.13127089238278022,
+      "grad_norm": 0.154296875,
+      "learning_rate": 0.02,
+      "loss": 1.7551,
+      "step": 1291
+    },
+    {
+      "epoch": 0.13137257394155852,
+      "grad_norm": 0.1494140625,
+      "learning_rate": 0.02,
+      "loss": 1.8149,
+      "step": 1292
+    },
+    {
+      "epoch": 0.1314742555003368,
+      "grad_norm": 0.1591796875,
+      "learning_rate": 0.02,
+      "loss": 1.7947,
+      "step": 1293
+    },
+    {
+      "epoch": 0.1315759370591151,
+      "grad_norm": 0.1591796875,
+      "learning_rate": 0.02,
+      "loss": 1.8442,
+      "step": 1294
+    },
+    {
+      "epoch": 0.1316776186178934,
+      "grad_norm": 0.154296875,
+      "learning_rate": 0.02,
+      "loss": 1.7244,
+      "step": 1295
+    },
+    {
+      "epoch": 0.13177930017667172,
+      "grad_norm": 0.16015625,
+      "learning_rate": 0.02,
+      "loss": 1.7775,
+      "step": 1296
+    },
+    {
+      "epoch": 0.13188098173545001,
+      "grad_norm": 0.1611328125,
+      "learning_rate": 0.02,
+      "loss": 1.7756,
+      "step": 1297
+    },
+    {
+      "epoch": 0.1319826632942283,
+      "grad_norm": 0.1650390625,
+      "learning_rate": 0.02,
+      "loss": 1.7467,
+      "step": 1298
+    },
+    {
+      "epoch": 0.1320843448530066,
+      "grad_norm": 0.1611328125,
+      "learning_rate": 0.02,
+      "loss": 1.7429,
+      "step": 1299
+    },
+    {
+      "epoch": 0.1321860264117849,
+      "grad_norm": 0.1650390625,
+      "learning_rate": 0.02,
+      "loss": 1.7654,
+      "step": 1300
+    },
+    {
+      "epoch": 0.1322877079705632,
+      "grad_norm": 0.15234375,
+      "learning_rate": 0.02,
+      "loss": 1.7014,
+      "step": 1301
+    },
+    {
+      "epoch": 0.13238938952934148,
+      "grad_norm": 0.1650390625,
+      "learning_rate": 0.02,
+      "loss": 1.7339,
+      "step": 1302
+    },
+    {
+      "epoch": 0.13249107108811978,
+      "grad_norm": 0.1572265625,
+      "learning_rate": 0.02,
+      "loss": 1.7738,
+      "step": 1303
+    },
+    {
+      "epoch": 0.13259275264689807,
+      "grad_norm": 0.166015625,
+      "learning_rate": 0.02,
+      "loss": 1.7056,
+      "step": 1304
+    },
+    {
+      "epoch": 0.13269443420567636,
+      "grad_norm": 0.162109375,
+      "learning_rate": 0.02,
+      "loss": 1.7472,
+      "step": 1305
+    },
+    {
+      "epoch": 0.13279611576445466,
+      "grad_norm": 0.1748046875,
+      "learning_rate": 0.02,
+      "loss": 1.7511,
+      "step": 1306
+    },
+    {
+      "epoch": 0.13289779732323295,
+      "grad_norm": 0.1806640625,
+      "learning_rate": 0.02,
+      "loss": 1.8341,
+      "step": 1307
+    },
+    {
+      "epoch": 0.13299947888201127,
+      "grad_norm": 0.17578125,
+      "learning_rate": 0.02,
+      "loss": 1.7422,
+      "step": 1308
+    },
+    {
+      "epoch": 0.13310116044078957,
+      "grad_norm": 0.1552734375,
+      "learning_rate": 0.02,
+      "loss": 1.745,
+      "step": 1309
+    },
+    {
+      "epoch": 0.13320284199956786,
+      "grad_norm": 0.1689453125,
+      "learning_rate": 0.02,
+      "loss": 1.7862,
+      "step": 1310
+    },
+    {
+      "epoch": 0.13330452355834616,
+      "grad_norm": 0.1728515625,
+      "learning_rate": 0.02,
+      "loss": 1.7328,
+      "step": 1311
+    },
+    {
+      "epoch": 0.13340620511712445,
+      "grad_norm": 0.158203125,
+      "learning_rate": 0.02,
+      "loss": 1.7351,
+      "step": 1312
+    },
+    {
+      "epoch": 0.13350788667590274,
+      "grad_norm": 0.169921875,
+      "learning_rate": 0.02,
+      "loss": 1.7622,
+      "step": 1313
+    },
+    {
+      "epoch": 0.13360956823468104,
+      "grad_norm": 0.162109375,
+      "learning_rate": 0.02,
+      "loss": 1.7401,
+      "step": 1314
+    },
+    {
+      "epoch": 0.13371124979345933,
+      "grad_norm": 0.1552734375,
+      "learning_rate": 0.02,
+      "loss": 1.7696,
+      "step": 1315
+    },
+    {
+      "epoch": 0.13381293135223762,
+      "grad_norm": 0.1611328125,
+      "learning_rate": 0.02,
+      "loss": 1.7736,
+      "step": 1316
+    },
+    {
+      "epoch": 0.13391461291101592,
+      "grad_norm": 0.1728515625,
+      "learning_rate": 0.02,
+      "loss": 1.7818,
+      "step": 1317
+    },
+    {
+      "epoch": 0.1340162944697942,
+      "grad_norm": 0.158203125,
+      "learning_rate": 0.02,
+      "loss": 1.7763,
+      "step": 1318
+    },
+    {
+      "epoch": 0.1341179760285725,
+      "grad_norm": 0.1474609375,
+      "learning_rate": 0.02,
+      "loss": 1.7045,
+      "step": 1319
+    },
+    {
+      "epoch": 0.13421965758735083,
+      "grad_norm": 0.1474609375,
+      "learning_rate": 0.02,
+      "loss": 1.7131,
+      "step": 1320
+    },
+    {
+      "epoch": 0.13432133914612912,
+      "grad_norm": 0.162109375,
+      "learning_rate": 0.02,
+      "loss": 1.7626,
+      "step": 1321
+    },
+    {
+      "epoch": 0.13442302070490741,
+      "grad_norm": 0.1689453125,
+      "learning_rate": 0.02,
+      "loss": 1.7848,
+      "step": 1322
+    },
+    {
+      "epoch": 0.1345247022636857,
+      "grad_norm": 0.1591796875,
+      "learning_rate": 0.02,
+      "loss": 1.7532,
+      "step": 1323
+    },
+    {
+      "epoch": 0.134626383822464,
+      "grad_norm": 0.1630859375,
+      "learning_rate": 0.02,
+      "loss": 1.7635,
+      "step": 1324
+    },
+    {
+      "epoch": 0.1347280653812423,
+      "grad_norm": 0.1689453125,
+      "learning_rate": 0.02,
+      "loss": 1.7558,
+      "step": 1325
+    },
+    {
+      "epoch": 0.1348297469400206,
+      "grad_norm": 0.166015625,
+      "learning_rate": 0.02,
+      "loss": 1.8283,
+      "step": 1326
+    },
+    {
+      "epoch": 0.13493142849879888,
+      "grad_norm": 0.169921875,
+      "learning_rate": 0.02,
+      "loss": 1.7598,
+      "step": 1327
+    },
+    {
+      "epoch": 0.13503311005757718,
+      "grad_norm": 0.1572265625,
+      "learning_rate": 0.02,
+      "loss": 1.7181,
+      "step": 1328
+    },
+    {
+      "epoch": 0.13513479161635547,
+      "grad_norm": 0.16796875,
+      "learning_rate": 0.02,
+      "loss": 1.7498,
+      "step": 1329
+    },
+    {
+      "epoch": 0.13523647317513376,
+      "grad_norm": 0.166015625,
+      "learning_rate": 0.02,
+      "loss": 1.7518,
+      "step": 1330
+    },
+    {
+      "epoch": 0.13533815473391206,
+      "grad_norm": 0.15234375,
+      "learning_rate": 0.02,
+      "loss": 1.7221,
+      "step": 1331
+    },
+    {
+      "epoch": 0.13543983629269038,
+      "grad_norm": 0.158203125,
+      "learning_rate": 0.02,
+      "loss": 1.7365,
+      "step": 1332
+    },
+    {
+      "epoch": 0.13554151785146867,
+      "grad_norm": 0.1513671875,
+      "learning_rate": 0.02,
+      "loss": 1.7975,
+      "step": 1333
+    },
+    {
+      "epoch": 0.13564319941024697,
+      "grad_norm": 0.1416015625,
+      "learning_rate": 0.02,
+      "loss": 1.7172,
+      "step": 1334
+    },
+    {
+      "epoch": 0.13574488096902526,
+      "grad_norm": 0.154296875,
+      "learning_rate": 0.02,
+      "loss": 1.7796,
+      "step": 1335
+    },
+    {
+      "epoch": 0.13584656252780355,
+      "grad_norm": 0.1494140625,
+      "learning_rate": 0.02,
+      "loss": 1.753,
+      "step": 1336
+    },
+    {
+      "epoch": 0.13594824408658185,
+      "grad_norm": 0.1533203125,
+      "learning_rate": 0.02,
+      "loss": 1.7518,
+      "step": 1337
+    },
+    {
+      "epoch": 0.13604992564536014,
+      "grad_norm": 0.1513671875,
+      "learning_rate": 0.02,
+      "loss": 1.7312,
+      "step": 1338
+    },
+    {
+      "epoch": 0.13615160720413844,
+      "grad_norm": 0.154296875,
+      "learning_rate": 0.02,
+      "loss": 1.7118,
+      "step": 1339
+    },
+    {
+      "epoch": 0.13625328876291673,
+      "grad_norm": 0.1494140625,
+      "learning_rate": 0.02,
+      "loss": 1.7147,
+      "step": 1340
+    },
+    {
+      "epoch": 0.13635497032169502,
+      "grad_norm": 0.166015625,
+      "learning_rate": 0.02,
+      "loss": 1.7535,
+      "step": 1341
+    },
+    {
+      "epoch": 0.13645665188047332,
+      "grad_norm": 0.1640625,
+      "learning_rate": 0.02,
+      "loss": 1.7651,
+      "step": 1342
+    },
+    {
+      "epoch": 0.1365583334392516,
+      "grad_norm": 0.1552734375,
+      "learning_rate": 0.02,
+      "loss": 1.7713,
+      "step": 1343
+    },
+    {
+      "epoch": 0.13666001499802993,
+      "grad_norm": 0.1650390625,
+      "learning_rate": 0.02,
+      "loss": 1.7759,
+      "step": 1344
+    },
+    {
+      "epoch": 0.13676169655680823,
+      "grad_norm": 0.1611328125,
+      "learning_rate": 0.02,
+      "loss": 1.7579,
+      "step": 1345
+    },
+    {
+      "epoch": 0.13686337811558652,
+      "grad_norm": 0.15625,
+      "learning_rate": 0.02,
+      "loss": 1.7188,
+      "step": 1346
+    },
+    {
+      "epoch": 0.13696505967436481,
+      "grad_norm": 0.1611328125,
+      "learning_rate": 0.02,
+      "loss": 1.7917,
+      "step": 1347
+    },
+    {
+      "epoch": 0.1370667412331431,
+      "grad_norm": 0.1630859375,
+      "learning_rate": 0.02,
+      "loss": 1.7419,
+      "step": 1348
+    },
+    {
+      "epoch": 0.1371684227919214,
+      "grad_norm": 0.1669921875,
+      "learning_rate": 0.02,
+      "loss": 1.7777,
+      "step": 1349
+    },
+    {
+      "epoch": 0.1372701043506997,
+      "grad_norm": 0.166015625,
+      "learning_rate": 0.02,
+      "loss": 1.7106,
+      "step": 1350
+    },
+    {
+      "epoch": 0.137371785909478,
+      "grad_norm": 0.1572265625,
+      "learning_rate": 0.02,
+      "loss": 1.7276,
+      "step": 1351
+    },
+    {
+      "epoch": 0.13747346746825628,
+      "grad_norm": 0.17578125,
+      "learning_rate": 0.02,
+      "loss": 1.6742,
+      "step": 1352
+    },
+    {
+      "epoch": 0.13757514902703458,
+      "grad_norm": 0.1640625,
+      "learning_rate": 0.02,
+      "loss": 1.7615,
+      "step": 1353
+    },
+    {
+      "epoch": 0.13767683058581287,
+      "grad_norm": 0.16015625,
+      "learning_rate": 0.02,
+      "loss": 1.7383,
+      "step": 1354
+    },
+    {
+      "epoch": 0.13777851214459116,
+      "grad_norm": 0.1572265625,
+      "learning_rate": 0.02,
+      "loss": 1.6922,
+      "step": 1355
+    },
+    {
+      "epoch": 0.13788019370336949,
+      "grad_norm": 0.16015625,
+      "learning_rate": 0.02,
+      "loss": 1.7494,
+      "step": 1356
+    },
+    {
+      "epoch": 0.13798187526214778,
+      "grad_norm": 0.1572265625,
+      "learning_rate": 0.02,
+      "loss": 1.7874,
+      "step": 1357
+    },
+    {
+      "epoch": 0.13808355682092607,
+      "grad_norm": 0.158203125,
+      "learning_rate": 0.02,
+      "loss": 1.7756,
+      "step": 1358
+    },
+    {
+      "epoch": 0.13818523837970437,
+      "grad_norm": 0.154296875,
+      "learning_rate": 0.02,
+      "loss": 1.7477,
+      "step": 1359
+    },
+    {
+      "epoch": 0.13828691993848266,
+      "grad_norm": 0.158203125,
+      "learning_rate": 0.02,
+      "loss": 1.7848,
+      "step": 1360
+    },
+    {
+      "epoch": 0.13838860149726095,
+      "grad_norm": 0.1552734375,
+      "learning_rate": 0.02,
+      "loss": 1.782,
+      "step": 1361
+    },
+    {
+      "epoch": 0.13849028305603925,
+      "grad_norm": 0.15625,
+      "learning_rate": 0.02,
+      "loss": 1.8124,
+      "step": 1362
+    },
+    {
+      "epoch": 0.13859196461481754,
+      "grad_norm": 0.1533203125,
+      "learning_rate": 0.02,
+      "loss": 1.737,
+      "step": 1363
+    },
+    {
+      "epoch": 0.13869364617359584,
+      "grad_norm": 0.154296875,
+      "learning_rate": 0.02,
+      "loss": 1.7795,
+      "step": 1364
+    },
+    {
+      "epoch": 0.13879532773237413,
+      "grad_norm": 0.1611328125,
+      "learning_rate": 0.02,
+      "loss": 1.801,
+      "step": 1365
+    },
+    {
+      "epoch": 0.13889700929115242,
+      "grad_norm": 0.1484375,
+      "learning_rate": 0.02,
+      "loss": 1.7043,
+      "step": 1366
+    },
+    {
+      "epoch": 0.13899869084993072,
+      "grad_norm": 0.1533203125,
+      "learning_rate": 0.02,
+      "loss": 1.8057,
+      "step": 1367
+    },
+    {
+      "epoch": 0.13910037240870904,
+      "grad_norm": 0.146484375,
+      "learning_rate": 0.02,
+      "loss": 1.7576,
+      "step": 1368
+    },
+    {
+      "epoch": 0.13920205396748733,
+      "grad_norm": 0.16015625,
+      "learning_rate": 0.02,
+      "loss": 1.7519,
+      "step": 1369
+    },
+    {
+      "epoch": 0.13930373552626563,
+      "grad_norm": 0.1611328125,
+      "learning_rate": 0.02,
+      "loss": 1.7847,
+      "step": 1370
+    },
+    {
+      "epoch": 0.13940541708504392,
+      "grad_norm": 0.1630859375,
+      "learning_rate": 0.02,
+      "loss": 1.7554,
+      "step": 1371
+    },
+    {
+      "epoch": 0.1395070986438222,
+      "grad_norm": 0.16015625,
+      "learning_rate": 0.02,
+      "loss": 1.7547,
+      "step": 1372
+    },
+    {
+      "epoch": 0.1396087802026005,
+      "grad_norm": 0.1767578125,
+      "learning_rate": 0.02,
+      "loss": 1.8331,
+      "step": 1373
+    },
+    {
+      "epoch": 0.1397104617613788,
+      "grad_norm": 0.1689453125,
+      "learning_rate": 0.02,
+      "loss": 1.7868,
+      "step": 1374
+    },
+    {
+      "epoch": 0.1398121433201571,
+      "grad_norm": 0.162109375,
+      "learning_rate": 0.02,
+      "loss": 1.7622,
+      "step": 1375
+    },
+    {
+      "epoch": 0.1399138248789354,
+      "grad_norm": 0.16015625,
+      "learning_rate": 0.02,
+      "loss": 1.7615,
+      "step": 1376
+    },
+    {
+      "epoch": 0.14001550643771368,
+      "grad_norm": 0.1630859375,
+      "learning_rate": 0.02,
+      "loss": 1.768,
+      "step": 1377
+    },
+    {
+      "epoch": 0.14011718799649198,
+      "grad_norm": 0.16015625,
+      "learning_rate": 0.02,
+      "loss": 1.7781,
+      "step": 1378
+    },
+    {
+      "epoch": 0.14021886955527027,
+      "grad_norm": 0.15625,
+      "learning_rate": 0.02,
+      "loss": 1.6745,
+      "step": 1379
+    },
+    {
+      "epoch": 0.1403205511140486,
+      "grad_norm": 0.1611328125,
+      "learning_rate": 0.02,
+      "loss": 1.7431,
+      "step": 1380
+    },
+    {
+      "epoch": 0.14042223267282689,
+      "grad_norm": 0.1552734375,
+      "learning_rate": 0.02,
+      "loss": 1.7573,
+      "step": 1381
+    },
+    {
+      "epoch": 0.14052391423160518,
+      "grad_norm": 0.16796875,
+      "learning_rate": 0.02,
+      "loss": 1.7759,
+      "step": 1382
+    },
+    {
+      "epoch": 0.14062559579038347,
+      "grad_norm": 0.16796875,
+      "learning_rate": 0.02,
+      "loss": 1.7539,
+      "step": 1383
+    },
+    {
+      "epoch": 0.14072727734916177,
+      "grad_norm": 0.162109375,
+      "learning_rate": 0.02,
+      "loss": 1.7877,
+      "step": 1384
+    },
+    {
+      "epoch": 0.14082895890794006,
+      "grad_norm": 0.1533203125,
+      "learning_rate": 0.02,
+      "loss": 1.6706,
+      "step": 1385
+    },
+    {
+      "epoch": 0.14093064046671835,
+      "grad_norm": 0.1640625,
+      "learning_rate": 0.02,
+      "loss": 1.7903,
+      "step": 1386
+    },
+    {
+      "epoch": 0.14103232202549665,
+      "grad_norm": 0.16015625,
+      "learning_rate": 0.02,
+      "loss": 1.7173,
+      "step": 1387
+    },
+    {
+      "epoch": 0.14113400358427494,
+      "grad_norm": 0.15234375,
+      "learning_rate": 0.02,
+      "loss": 1.7371,
+      "step": 1388
+    },
+    {
+      "epoch": 0.14123568514305324,
+      "grad_norm": 0.154296875,
+      "learning_rate": 0.02,
+      "loss": 1.7518,
+      "step": 1389
+    },
+    {
+      "epoch": 0.14133736670183153,
+      "grad_norm": 0.16015625,
+      "learning_rate": 0.02,
+      "loss": 1.718,
+      "step": 1390
+    },
+    {
+      "epoch": 0.14143904826060982,
+      "grad_norm": 0.1484375,
+      "learning_rate": 0.02,
+      "loss": 1.7683,
+      "step": 1391
+    },
+    {
+      "epoch": 0.14154072981938814,
+      "grad_norm": 0.162109375,
+      "learning_rate": 0.02,
+      "loss": 1.7493,
+      "step": 1392
+    },
+    {
+      "epoch": 0.14164241137816644,
+      "grad_norm": 0.1513671875,
+      "learning_rate": 0.02,
+      "loss": 1.7558,
+      "step": 1393
+    },
+    {
+      "epoch": 0.14174409293694473,
+      "grad_norm": 0.15625,
+      "learning_rate": 0.02,
+      "loss": 1.7258,
+      "step": 1394
+    },
+    {
+      "epoch": 0.14184577449572303,
+      "grad_norm": 0.158203125,
+      "learning_rate": 0.02,
+      "loss": 1.7299,
+      "step": 1395
+    },
+    {
+      "epoch": 0.14194745605450132,
+      "grad_norm": 0.15625,
+      "learning_rate": 0.02,
+      "loss": 1.7058,
+      "step": 1396
+    },
+    {
+      "epoch": 0.1420491376132796,
+      "grad_norm": 0.162109375,
+      "learning_rate": 0.02,
+      "loss": 1.7874,
+      "step": 1397
+    },
+    {
+      "epoch": 0.1421508191720579,
+      "grad_norm": 0.1494140625,
+      "learning_rate": 0.02,
+      "loss": 1.7452,
+      "step": 1398
+    },
+    {
+      "epoch": 0.1422525007308362,
+      "grad_norm": 0.1484375,
+      "learning_rate": 0.02,
+      "loss": 1.7673,
+      "step": 1399
+    },
+    {
+      "epoch": 0.1423541822896145,
+      "grad_norm": 0.146484375,
+      "learning_rate": 0.02,
+      "loss": 1.7199,
+      "step": 1400
+    },
+    {
+      "epoch": 0.1424558638483928,
+      "grad_norm": 0.15234375,
+      "learning_rate": 0.02,
+      "loss": 1.7077,
+      "step": 1401
+    },
+    {
+      "epoch": 0.14255754540717108,
+      "grad_norm": 0.1494140625,
+      "learning_rate": 0.02,
+      "loss": 1.7797,
+      "step": 1402
+    },
+    {
+      "epoch": 0.14265922696594938,
+      "grad_norm": 0.1435546875,
+      "learning_rate": 0.02,
+      "loss": 1.7858,
+      "step": 1403
+    },
+    {
+      "epoch": 0.1427609085247277,
+      "grad_norm": 0.13671875,
+      "learning_rate": 0.02,
+      "loss": 1.7234,
+      "step": 1404
+    },
+    {
+      "epoch": 0.142862590083506,
+      "grad_norm": 0.1396484375,
+      "learning_rate": 0.02,
+      "loss": 1.736,
+      "step": 1405
+    },
+    {
+      "epoch": 0.14296427164228429,
+      "grad_norm": 0.162109375,
+      "learning_rate": 0.02,
+      "loss": 1.7419,
+      "step": 1406
+    },
+    {
+      "epoch": 0.14306595320106258,
+      "grad_norm": 0.1533203125,
+      "learning_rate": 0.02,
+      "loss": 1.7145,
+      "step": 1407
+    },
+    {
+      "epoch": 0.14316763475984087,
+      "grad_norm": 0.1533203125,
+      "learning_rate": 0.02,
+      "loss": 1.7836,
+      "step": 1408
+    },
+    {
+      "epoch": 0.14326931631861917,
+      "grad_norm": 0.150390625,
+      "learning_rate": 0.02,
+      "loss": 1.7112,
+      "step": 1409
+    },
+    {
+      "epoch": 0.14337099787739746,
+      "grad_norm": 0.146484375,
+      "learning_rate": 0.02,
+      "loss": 1.7446,
+      "step": 1410
+    },
+    {
+      "epoch": 0.14347267943617575,
+      "grad_norm": 0.1474609375,
+      "learning_rate": 0.02,
+      "loss": 1.7952,
+      "step": 1411
+    },
+    {
+      "epoch": 0.14357436099495405,
+      "grad_norm": 0.1484375,
+      "learning_rate": 0.02,
+      "loss": 1.7623,
+      "step": 1412
+    },
+    {
+      "epoch": 0.14367604255373234,
+      "grad_norm": 0.1533203125,
+      "learning_rate": 0.02,
+      "loss": 1.7319,
+      "step": 1413
+    },
+    {
+      "epoch": 0.14377772411251064,
+      "grad_norm": 0.158203125,
+      "learning_rate": 0.02,
+      "loss": 1.793,
+      "step": 1414
+    },
+    {
+      "epoch": 0.14387940567128893,
+      "grad_norm": 0.1650390625,
+      "learning_rate": 0.02,
+      "loss": 1.7762,
+      "step": 1415
+    },
+    {
+      "epoch": 0.14398108723006725,
+      "grad_norm": 0.158203125,
+      "learning_rate": 0.02,
+      "loss": 1.8109,
+      "step": 1416
+    },
+    {
+      "epoch": 0.14408276878884554,
+      "grad_norm": 0.1669921875,
+      "learning_rate": 0.02,
+      "loss": 1.7015,
+      "step": 1417
+    },
+    {
+      "epoch": 0.14418445034762384,
+      "grad_norm": 0.1630859375,
+      "learning_rate": 0.02,
+      "loss": 1.7708,
+      "step": 1418
+    },
+    {
+      "epoch": 0.14428613190640213,
+      "grad_norm": 0.16796875,
+      "learning_rate": 0.02,
+      "loss": 1.713,
+      "step": 1419
+    },
+    {
+      "epoch": 0.14438781346518043,
+      "grad_norm": 0.16796875,
+      "learning_rate": 0.02,
+      "loss": 1.7238,
+      "step": 1420
+    },
+    {
+      "epoch": 0.14448949502395872,
+      "grad_norm": 0.1806640625,
+      "learning_rate": 0.02,
+      "loss": 1.77,
+      "step": 1421
+    },
+    {
+      "epoch": 0.144591176582737,
+      "grad_norm": 0.1875,
+      "learning_rate": 0.02,
+      "loss": 1.7768,
+      "step": 1422
+    },
+    {
+      "epoch": 0.1446928581415153,
+      "grad_norm": 0.1728515625,
+      "learning_rate": 0.02,
+      "loss": 1.7715,
+      "step": 1423
+    },
+    {
+      "epoch": 0.1447945397002936,
+      "grad_norm": 0.1865234375,
+      "learning_rate": 0.02,
+      "loss": 1.8237,
+      "step": 1424
+    },
+    {
+      "epoch": 0.1448962212590719,
+      "grad_norm": 0.177734375,
+      "learning_rate": 0.02,
+      "loss": 1.7804,
+      "step": 1425
+    },
+    {
+      "epoch": 0.1449979028178502,
+      "grad_norm": 0.169921875,
+      "learning_rate": 0.02,
+      "loss": 1.7198,
+      "step": 1426
+    },
+    {
+      "epoch": 0.14509958437662848,
+      "grad_norm": 0.1689453125,
+      "learning_rate": 0.02,
+      "loss": 1.8006,
+      "step": 1427
+    },
+    {
+      "epoch": 0.1452012659354068,
+      "grad_norm": 0.15625,
+      "learning_rate": 0.02,
+      "loss": 1.7845,
+      "step": 1428
+    },
+    {
+      "epoch": 0.1453029474941851,
+      "grad_norm": 0.1611328125,
+      "learning_rate": 0.02,
+      "loss": 1.775,
+      "step": 1429
+    },
+    {
+      "epoch": 0.1454046290529634,
+      "grad_norm": 0.154296875,
+      "learning_rate": 0.02,
+      "loss": 1.7629,
+      "step": 1430
+    },
+    {
+      "epoch": 0.14550631061174168,
+      "grad_norm": 0.1630859375,
+      "learning_rate": 0.02,
+      "loss": 1.7838,
+      "step": 1431
+    },
+    {
+      "epoch": 0.14560799217051998,
+      "grad_norm": 0.15234375,
+      "learning_rate": 0.02,
+      "loss": 1.7217,
+      "step": 1432
+    },
+    {
+      "epoch": 0.14570967372929827,
+      "grad_norm": 0.1552734375,
+      "learning_rate": 0.02,
+      "loss": 1.8087,
+      "step": 1433
+    },
+    {
+      "epoch": 0.14581135528807657,
+      "grad_norm": 0.1572265625,
+      "learning_rate": 0.02,
+      "loss": 1.7876,
+      "step": 1434
+    },
+    {
+      "epoch": 0.14591303684685486,
+      "grad_norm": 0.1611328125,
+      "learning_rate": 0.02,
+      "loss": 1.8058,
+      "step": 1435
+    },
+    {
+      "epoch": 0.14601471840563315,
+      "grad_norm": 0.1474609375,
+      "learning_rate": 0.02,
+      "loss": 1.8048,
+      "step": 1436
+    },
+    {
+      "epoch": 0.14611639996441145,
+      "grad_norm": 0.158203125,
+      "learning_rate": 0.02,
+      "loss": 1.727,
+      "step": 1437
+    },
+    {
+      "epoch": 0.14621808152318974,
+      "grad_norm": 0.1494140625,
+      "learning_rate": 0.02,
+      "loss": 1.8061,
+      "step": 1438
+    },
+    {
+      "epoch": 0.14631976308196804,
+      "grad_norm": 0.158203125,
+      "learning_rate": 0.02,
+      "loss": 1.7914,
+      "step": 1439
+    },
+    {
+      "epoch": 0.14642144464074633,
+      "grad_norm": 0.1494140625,
+      "learning_rate": 0.02,
+      "loss": 1.7213,
+      "step": 1440
+    },
+    {
+      "epoch": 0.14652312619952465,
+      "grad_norm": 0.1708984375,
+      "learning_rate": 0.02,
+      "loss": 1.7535,
+      "step": 1441
+    },
+    {
+      "epoch": 0.14662480775830294,
+      "grad_norm": 0.166015625,
+      "learning_rate": 0.02,
+      "loss": 1.7393,
+      "step": 1442
+    },
+    {
+      "epoch": 0.14672648931708124,
+      "grad_norm": 0.1640625,
+      "learning_rate": 0.02,
+      "loss": 1.7575,
+      "step": 1443
+    },
+    {
+      "epoch": 0.14682817087585953,
+      "grad_norm": 0.1533203125,
+      "learning_rate": 0.02,
+      "loss": 1.7674,
+      "step": 1444
+    },
+    {
+      "epoch": 0.14692985243463783,
+      "grad_norm": 0.169921875,
+      "learning_rate": 0.02,
+      "loss": 1.7349,
+      "step": 1445
+    },
+    {
+      "epoch": 0.14703153399341612,
+      "grad_norm": 0.1630859375,
+      "learning_rate": 0.02,
+      "loss": 1.707,
+      "step": 1446
+    },
+    {
+      "epoch": 0.1471332155521944,
+      "grad_norm": 0.1630859375,
+      "learning_rate": 0.02,
+      "loss": 1.8016,
+      "step": 1447
+    },
+    {
+      "epoch": 0.1472348971109727,
+      "grad_norm": 0.1640625,
+      "learning_rate": 0.02,
+      "loss": 1.7914,
+      "step": 1448
+    },
+    {
+      "epoch": 0.147336578669751,
+      "grad_norm": 0.1611328125,
+      "learning_rate": 0.02,
+      "loss": 1.762,
+      "step": 1449
+    },
+    {
+      "epoch": 0.1474382602285293,
+      "grad_norm": 0.166015625,
+      "learning_rate": 0.02,
+      "loss": 1.7791,
+      "step": 1450
+    },
+    {
+      "epoch": 0.1475399417873076,
+      "grad_norm": 0.17578125,
+      "learning_rate": 0.02,
+      "loss": 1.847,
+      "step": 1451
+    },
+    {
+      "epoch": 0.14764162334608588,
+      "grad_norm": 0.1708984375,
+      "learning_rate": 0.02,
+      "loss": 1.8126,
+      "step": 1452
+    },
+    {
+      "epoch": 0.1477433049048642,
+      "grad_norm": 0.15625,
+      "learning_rate": 0.02,
+      "loss": 1.7861,
+      "step": 1453
+    },
+    {
+      "epoch": 0.1478449864636425,
+      "grad_norm": 0.1591796875,
+      "learning_rate": 0.02,
+      "loss": 1.708,
+      "step": 1454
+    },
+    {
+      "epoch": 0.1479466680224208,
+      "grad_norm": 0.15625,
+      "learning_rate": 0.02,
+      "loss": 1.7236,
+      "step": 1455
+    },
+    {
+      "epoch": 0.14804834958119908,
+      "grad_norm": 0.1552734375,
+      "learning_rate": 0.02,
+      "loss": 1.6702,
+      "step": 1456
+    },
+    {
+      "epoch": 0.14815003113997738,
+      "grad_norm": 0.15625,
+      "learning_rate": 0.02,
+      "loss": 1.835,
+      "step": 1457
+    },
+    {
+      "epoch": 0.14825171269875567,
+      "grad_norm": 0.1484375,
+      "learning_rate": 0.02,
+      "loss": 1.7007,
+      "step": 1458
+    },
+    {
+      "epoch": 0.14835339425753397,
+      "grad_norm": 0.1484375,
+      "learning_rate": 0.02,
+      "loss": 1.724,
+      "step": 1459
+    },
+    {
+      "epoch": 0.14845507581631226,
+      "grad_norm": 0.16015625,
+      "learning_rate": 0.02,
+      "loss": 1.6749,
+      "step": 1460
+    },
+    {
+      "epoch": 0.14855675737509055,
+      "grad_norm": 0.1484375,
+      "learning_rate": 0.02,
+      "loss": 1.7015,
+      "step": 1461
+    },
+    {
+      "epoch": 0.14865843893386885,
+      "grad_norm": 0.1640625,
+      "learning_rate": 0.02,
+      "loss": 1.7596,
+      "step": 1462
+    },
+    {
+      "epoch": 0.14876012049264714,
+      "grad_norm": 0.1572265625,
+      "learning_rate": 0.02,
+      "loss": 1.7954,
+      "step": 1463
+    },
+    {
+      "epoch": 0.14886180205142543,
+      "grad_norm": 0.1552734375,
+      "learning_rate": 0.02,
+      "loss": 1.7359,
+      "step": 1464
+    },
+    {
+      "epoch": 0.14896348361020376,
+      "grad_norm": 0.1552734375,
+      "learning_rate": 0.02,
+      "loss": 1.7783,
+      "step": 1465
+    },
+    {
+      "epoch": 0.14906516516898205,
+      "grad_norm": 0.1640625,
+      "learning_rate": 0.02,
+      "loss": 1.7015,
+      "step": 1466
+    },
+    {
+      "epoch": 0.14916684672776034,
+      "grad_norm": 0.169921875,
+      "learning_rate": 0.02,
+      "loss": 1.8073,
+      "step": 1467
+    },
+    {
+      "epoch": 0.14926852828653864,
+      "grad_norm": 0.1572265625,
+      "learning_rate": 0.02,
+      "loss": 1.7495,
+      "step": 1468
+    },
+    {
+      "epoch": 0.14937020984531693,
+      "grad_norm": 0.16015625,
+      "learning_rate": 0.02,
+      "loss": 1.7353,
+      "step": 1469
+    },
+    {
+      "epoch": 0.14947189140409523,
+      "grad_norm": 0.1611328125,
+      "learning_rate": 0.02,
+      "loss": 1.8099,
+      "step": 1470
+    },
+    {
+      "epoch": 0.14957357296287352,
+      "grad_norm": 0.1591796875,
+      "learning_rate": 0.02,
+      "loss": 1.7412,
+      "step": 1471
+    },
+    {
+      "epoch": 0.1496752545216518,
+      "grad_norm": 0.1484375,
+      "learning_rate": 0.02,
+      "loss": 1.7316,
+      "step": 1472
+    },
+    {
+      "epoch": 0.1497769360804301,
+      "grad_norm": 0.15234375,
+      "learning_rate": 0.02,
+      "loss": 1.7171,
+      "step": 1473
+    },
+    {
+      "epoch": 0.1498786176392084,
+      "grad_norm": 0.158203125,
+      "learning_rate": 0.02,
+      "loss": 1.7919,
+      "step": 1474
+    },
+    {
+      "epoch": 0.1499802991979867,
+      "grad_norm": 0.1572265625,
+      "learning_rate": 0.02,
+      "loss": 1.7501,
+      "step": 1475
+    },
+    {
+      "epoch": 0.150081980756765,
+      "grad_norm": 0.15234375,
+      "learning_rate": 0.02,
+      "loss": 1.7415,
+      "step": 1476
+    },
+    {
+      "epoch": 0.13673552045362494,
+      "grad_norm": 0.146484375,
+      "learning_rate": 0.02,
+      "loss": 1.7336,
+      "step": 1477
+    },
+    {
+      "epoch": 0.136828096973905,
+      "grad_norm": 0.1650390625,
+      "learning_rate": 0.02,
+      "loss": 1.7345,
+      "step": 1478
+    },
+    {
+      "epoch": 0.13692067349418505,
+      "grad_norm": 0.1767578125,
+      "learning_rate": 0.02,
+      "loss": 1.7464,
+      "step": 1479
+    },
+    {
+      "epoch": 0.13701325001446507,
+      "grad_norm": 0.1748046875,
+      "learning_rate": 0.02,
+      "loss": 1.7494,
+      "step": 1480
+    },
+    {
+      "epoch": 0.13710582653474512,
+      "grad_norm": 0.173828125,
+      "learning_rate": 0.02,
+      "loss": 1.7619,
+      "step": 1481
+    },
+    {
+      "epoch": 0.13719840305502518,
+      "grad_norm": 0.1923828125,
+      "learning_rate": 0.02,
+      "loss": 1.7448,
+      "step": 1482
+    },
+    {
+      "epoch": 0.1372909795753052,
+      "grad_norm": 0.1796875,
+      "learning_rate": 0.02,
+      "loss": 1.7474,
+      "step": 1483
+    },
+    {
+      "epoch": 0.13738355609558525,
+      "grad_norm": 0.1787109375,
+      "learning_rate": 0.02,
+      "loss": 1.7293,
+      "step": 1484
+    },
+    {
+      "epoch": 0.1374761326158653,
+      "grad_norm": 0.1669921875,
+      "learning_rate": 0.02,
+      "loss": 1.7892,
+      "step": 1485
+    },
+    {
+      "epoch": 0.13756870913614536,
+      "grad_norm": 0.173828125,
+      "learning_rate": 0.02,
+      "loss": 1.7649,
+      "step": 1486
+    },
+    {
+      "epoch": 0.13766128565642538,
+      "grad_norm": 0.166015625,
+      "learning_rate": 0.02,
+      "loss": 1.7535,
+      "step": 1487
+    },
+    {
+      "epoch": 0.13775386217670543,
+      "grad_norm": 0.1591796875,
+      "learning_rate": 0.02,
+      "loss": 1.7285,
+      "step": 1488
+    },
+    {
+      "epoch": 0.13784643869698548,
+      "grad_norm": 0.171875,
+      "learning_rate": 0.02,
+      "loss": 1.7585,
+      "step": 1489
+    },
+    {
+      "epoch": 0.1379390152172655,
+      "grad_norm": 0.166015625,
+      "learning_rate": 0.02,
+      "loss": 1.7321,
+      "step": 1490
+    },
+    {
+      "epoch": 0.13803159173754556,
+      "grad_norm": 0.15234375,
+      "learning_rate": 0.02,
+      "loss": 1.702,
+      "step": 1491
+    },
+    {
+      "epoch": 0.1381241682578256,
+      "grad_norm": 0.162109375,
+      "learning_rate": 0.02,
+      "loss": 1.7497,
+      "step": 1492
+    },
+    {
+      "epoch": 0.13821674477810567,
+      "grad_norm": 0.16796875,
+      "learning_rate": 0.02,
+      "loss": 1.6854,
+      "step": 1493
+    },
+    {
+      "epoch": 0.1383093212983857,
+      "grad_norm": 0.158203125,
+      "learning_rate": 0.02,
+      "loss": 1.6824,
+      "step": 1494
+    },
+    {
+      "epoch": 0.13840189781866574,
+      "grad_norm": 0.17578125,
+      "learning_rate": 0.02,
+      "loss": 1.7125,
+      "step": 1495
+    },
+    {
+      "epoch": 0.1384944743389458,
+      "grad_norm": 0.16015625,
+      "learning_rate": 0.02,
+      "loss": 1.7885,
+      "step": 1496
+    },
+    {
+      "epoch": 0.13858705085922582,
+      "grad_norm": 0.1591796875,
+      "learning_rate": 0.02,
+      "loss": 1.7029,
+      "step": 1497
+    },
+    {
+      "epoch": 0.13867962737950587,
+      "grad_norm": 0.1552734375,
+      "learning_rate": 0.02,
+      "loss": 1.7193,
+      "step": 1498
+    },
+    {
+      "epoch": 0.13877220389978592,
+      "grad_norm": 0.162109375,
+      "learning_rate": 0.02,
+      "loss": 1.7591,
+      "step": 1499
+    },
+    {
+      "epoch": 0.13886478042006595,
+      "grad_norm": 0.17578125,
+      "learning_rate": 0.02,
+      "loss": 1.7179,
+      "step": 1500
+    },
+    {
+      "epoch": 0.138957356940346,
+      "grad_norm": 0.1875,
+      "learning_rate": 0.02,
+      "loss": 1.7505,
+      "step": 1501
+    },
+    {
+      "epoch": 0.13904993346062605,
+      "grad_norm": 0.1826171875,
+      "learning_rate": 0.02,
+      "loss": 1.8072,
+      "step": 1502
+    },
+    {
+      "epoch": 0.1391425099809061,
+      "grad_norm": 0.17578125,
+      "learning_rate": 0.02,
+      "loss": 1.71,
+      "step": 1503
+    },
+    {
+      "epoch": 0.13923508650118613,
+      "grad_norm": 0.185546875,
+      "learning_rate": 0.02,
+      "loss": 1.7634,
+      "step": 1504
+    },
+    {
+      "epoch": 0.13932766302146618,
+      "grad_norm": 0.166015625,
+      "learning_rate": 0.02,
+      "loss": 1.7296,
+      "step": 1505
+    },
+    {
+      "epoch": 0.13942023954174623,
+      "grad_norm": 0.1748046875,
+      "learning_rate": 0.02,
+      "loss": 1.7251,
+      "step": 1506
+    },
+    {
+      "epoch": 0.13951281606202626,
+      "grad_norm": 0.16796875,
+      "learning_rate": 0.02,
+      "loss": 1.736,
+      "step": 1507
+    },
+    {
+      "epoch": 0.1396053925823063,
+      "grad_norm": 0.171875,
+      "learning_rate": 0.02,
+      "loss": 1.7416,
+      "step": 1508
+    },
+    {
+      "epoch": 0.13969796910258636,
+      "grad_norm": 0.173828125,
+      "learning_rate": 0.02,
+      "loss": 1.7921,
+      "step": 1509
+    },
+    {
+      "epoch": 0.1397905456228664,
+      "grad_norm": 0.1708984375,
+      "learning_rate": 0.02,
+      "loss": 1.7154,
+      "step": 1510
+    },
+    {
+      "epoch": 0.13988312214314644,
+      "grad_norm": 0.169921875,
+      "learning_rate": 0.02,
+      "loss": 1.7204,
+      "step": 1511
+    },
+    {
+      "epoch": 0.1399756986634265,
+      "grad_norm": 0.16796875,
+      "learning_rate": 0.02,
+      "loss": 1.6484,
+      "step": 1512
+    },
+    {
+      "epoch": 0.14006827518370654,
+      "grad_norm": 0.171875,
+      "learning_rate": 0.02,
+      "loss": 1.764,
+      "step": 1513
+    },
+    {
+      "epoch": 0.14016085170398657,
+      "grad_norm": 0.197265625,
+      "learning_rate": 0.02,
+      "loss": 1.7804,
+      "step": 1514
+    },
+    {
+      "epoch": 0.14025342822426662,
+      "grad_norm": 0.162109375,
+      "learning_rate": 0.02,
+      "loss": 1.7415,
+      "step": 1515
+    },
+    {
+      "epoch": 0.14034600474454667,
+      "grad_norm": 0.1728515625,
+      "learning_rate": 0.02,
+      "loss": 1.7211,
+      "step": 1516
+    },
+    {
+      "epoch": 0.1404385812648267,
+      "grad_norm": 0.1767578125,
+      "learning_rate": 0.02,
+      "loss": 1.7944,
+      "step": 1517
+    },
+    {
+      "epoch": 0.14053115778510675,
+      "grad_norm": 0.1748046875,
+      "learning_rate": 0.02,
+      "loss": 1.7671,
+      "step": 1518
+    },
+    {
+      "epoch": 0.1406237343053868,
+      "grad_norm": 0.1748046875,
+      "learning_rate": 0.02,
+      "loss": 1.7775,
+      "step": 1519
+    },
+    {
+      "epoch": 0.14071631082566685,
+      "grad_norm": 0.1796875,
+      "learning_rate": 0.02,
+      "loss": 1.7762,
+      "step": 1520
+    },
+    {
+      "epoch": 0.14080888734594688,
+      "grad_norm": 0.17578125,
+      "learning_rate": 0.02,
+      "loss": 1.7523,
+      "step": 1521
+    },
+    {
+      "epoch": 0.14090146386622693,
+      "grad_norm": 0.16015625,
+      "learning_rate": 0.02,
+      "loss": 1.6953,
+      "step": 1522
+    },
+    {
+      "epoch": 0.14099404038650698,
+      "grad_norm": 0.1640625,
+      "learning_rate": 0.02,
+      "loss": 1.7305,
+      "step": 1523
+    },
+    {
+      "epoch": 0.141086616906787,
+      "grad_norm": 0.1572265625,
+      "learning_rate": 0.02,
+      "loss": 1.7339,
+      "step": 1524
+    },
+    {
+      "epoch": 0.14117919342706706,
+      "grad_norm": 0.1513671875,
+      "learning_rate": 0.02,
+      "loss": 1.7074,
+      "step": 1525
+    },
+    {
+      "epoch": 0.1412717699473471,
+      "grad_norm": 0.154296875,
+      "learning_rate": 0.02,
+      "loss": 1.6777,
+      "step": 1526
+    },
+    {
+      "epoch": 0.14136434646762716,
+      "grad_norm": 0.1572265625,
+      "learning_rate": 0.02,
+      "loss": 1.7668,
+      "step": 1527
+    },
+    {
+      "epoch": 0.14145692298790719,
+      "grad_norm": 0.1630859375,
+      "learning_rate": 0.02,
+      "loss": 1.7179,
+      "step": 1528
+    },
+    {
+      "epoch": 0.14154949950818724,
+      "grad_norm": 0.158203125,
+      "learning_rate": 0.02,
+      "loss": 1.7723,
+      "step": 1529
+    },
+    {
+      "epoch": 0.1416420760284673,
+      "grad_norm": 0.1611328125,
+      "learning_rate": 0.02,
+      "loss": 1.7511,
+      "step": 1530
+    },
+    {
+      "epoch": 0.14173465254874731,
+      "grad_norm": 0.15625,
+      "learning_rate": 0.02,
+      "loss": 1.7889,
+      "step": 1531
+    },
+    {
+      "epoch": 0.14182722906902737,
+      "grad_norm": 0.1708984375,
+      "learning_rate": 0.02,
+      "loss": 1.7614,
+      "step": 1532
+    },
+    {
+      "epoch": 0.14191980558930742,
+      "grad_norm": 0.16015625,
+      "learning_rate": 0.02,
+      "loss": 1.6925,
+      "step": 1533
+    },
+    {
+      "epoch": 0.14201238210958744,
+      "grad_norm": 0.1630859375,
+      "learning_rate": 0.02,
+      "loss": 1.7316,
+      "step": 1534
+    },
+    {
+      "epoch": 0.1421049586298675,
+      "grad_norm": 0.15234375,
+      "learning_rate": 0.02,
+      "loss": 1.7882,
+      "step": 1535
+    },
+    {
+      "epoch": 0.14219753515014755,
+      "grad_norm": 0.1591796875,
+      "learning_rate": 0.02,
+      "loss": 1.7725,
+      "step": 1536
+    },
+    {
+      "epoch": 0.1422901116704276,
+      "grad_norm": 0.1611328125,
+      "learning_rate": 0.02,
+      "loss": 1.6861,
+      "step": 1537
+    },
+    {
+      "epoch": 0.14238268819070762,
+      "grad_norm": 0.1611328125,
+      "learning_rate": 0.02,
+      "loss": 1.749,
+      "step": 1538
+    },
+    {
+      "epoch": 0.14247526471098768,
+      "grad_norm": 0.16015625,
+      "learning_rate": 0.02,
+      "loss": 1.7974,
+      "step": 1539
+    },
+    {
+      "epoch": 0.14256784123126773,
+      "grad_norm": 0.1533203125,
+      "learning_rate": 0.02,
+      "loss": 1.7325,
+      "step": 1540
+    },
+    {
+      "epoch": 0.14266041775154775,
+      "grad_norm": 0.154296875,
+      "learning_rate": 0.02,
+      "loss": 1.6767,
+      "step": 1541
+    },
+    {
+      "epoch": 0.1427529942718278,
+      "grad_norm": 0.158203125,
+      "learning_rate": 0.02,
+      "loss": 1.7416,
+      "step": 1542
+    },
+    {
+      "epoch": 0.14284557079210786,
+      "grad_norm": 0.1708984375,
+      "learning_rate": 0.02,
+      "loss": 1.7867,
+      "step": 1543
+    },
+    {
+      "epoch": 0.1429381473123879,
+      "grad_norm": 0.158203125,
+      "learning_rate": 0.02,
+      "loss": 1.7044,
+      "step": 1544
+    },
+    {
+      "epoch": 0.14303072383266793,
+      "grad_norm": 0.15625,
+      "learning_rate": 0.02,
+      "loss": 1.697,
+      "step": 1545
+    },
+    {
+      "epoch": 0.14312330035294799,
+      "grad_norm": 0.1630859375,
+      "learning_rate": 0.02,
+      "loss": 1.7754,
+      "step": 1546
+    },
+    {
+      "epoch": 0.14321587687322804,
+      "grad_norm": 0.177734375,
+      "learning_rate": 0.02,
+      "loss": 1.7834,
+      "step": 1547
+    },
+    {
+      "epoch": 0.14330845339350806,
+      "grad_norm": 0.1767578125,
+      "learning_rate": 0.02,
+      "loss": 1.8335,
+      "step": 1548
+    },
+    {
+      "epoch": 0.14340102991378811,
+      "grad_norm": 0.169921875,
+      "learning_rate": 0.02,
+      "loss": 1.755,
+      "step": 1549
+    },
+    {
+      "epoch": 0.14349360643406817,
+      "grad_norm": 0.16796875,
+      "learning_rate": 0.02,
+      "loss": 1.7561,
+      "step": 1550
+    },
+    {
+      "epoch": 0.1435861829543482,
+      "grad_norm": 0.1826171875,
+      "learning_rate": 0.02,
+      "loss": 1.7053,
+      "step": 1551
+    },
+    {
+      "epoch": 0.14367875947462824,
+      "grad_norm": 0.1806640625,
+      "learning_rate": 0.02,
+      "loss": 1.7926,
+      "step": 1552
+    },
+    {
+      "epoch": 0.1437713359949083,
+      "grad_norm": 0.1630859375,
+      "learning_rate": 0.02,
+      "loss": 1.7589,
+      "step": 1553
+    },
+    {
+      "epoch": 0.14386391251518835,
+      "grad_norm": 0.181640625,
+      "learning_rate": 0.02,
+      "loss": 1.7877,
+      "step": 1554
+    },
+    {
+      "epoch": 0.14395648903546837,
+      "grad_norm": 0.1728515625,
+      "learning_rate": 0.02,
+      "loss": 1.7595,
+      "step": 1555
+    },
+    {
+      "epoch": 0.14404906555574842,
+      "grad_norm": 0.173828125,
+      "learning_rate": 0.02,
+      "loss": 1.7671,
+      "step": 1556
+    },
+    {
+      "epoch": 0.14414164207602848,
+      "grad_norm": 0.1708984375,
+      "learning_rate": 0.02,
+      "loss": 1.7392,
+      "step": 1557
+    },
+    {
+      "epoch": 0.1442342185963085,
+      "grad_norm": 0.1923828125,
+      "learning_rate": 0.02,
+      "loss": 1.7876,
+      "step": 1558
+    },
+    {
+      "epoch": 0.14432679511658855,
+      "grad_norm": 0.173828125,
+      "learning_rate": 0.02,
+      "loss": 1.7606,
+      "step": 1559
+    },
+    {
+      "epoch": 0.1444193716368686,
+      "grad_norm": 0.177734375,
+      "learning_rate": 0.02,
+      "loss": 1.6797,
+      "step": 1560
+    },
+    {
+      "epoch": 0.14451194815714866,
+      "grad_norm": 0.1552734375,
+      "learning_rate": 0.02,
+      "loss": 1.6944,
+      "step": 1561
+    },
+    {
+      "epoch": 0.14460452467742868,
+      "grad_norm": 0.1796875,
+      "learning_rate": 0.02,
+      "loss": 1.7753,
+      "step": 1562
+    },
+    {
+      "epoch": 0.14469710119770873,
+      "grad_norm": 0.1728515625,
+      "learning_rate": 0.02,
+      "loss": 1.7208,
+      "step": 1563
+    },
+    {
+      "epoch": 0.14478967771798879,
+      "grad_norm": 0.1767578125,
+      "learning_rate": 0.02,
+      "loss": 1.7764,
+      "step": 1564
+    },
+    {
+      "epoch": 0.1448822542382688,
+      "grad_norm": 0.16796875,
+      "learning_rate": 0.02,
+      "loss": 1.77,
+      "step": 1565
+    },
+    {
+      "epoch": 0.14497483075854886,
+      "grad_norm": 0.1767578125,
+      "learning_rate": 0.02,
+      "loss": 1.7693,
+      "step": 1566
+    },
+    {
+      "epoch": 0.1450674072788289,
+      "grad_norm": 0.166015625,
+      "learning_rate": 0.02,
+      "loss": 1.7288,
+      "step": 1567
+    },
+    {
+      "epoch": 0.14515998379910894,
+      "grad_norm": 0.166015625,
+      "learning_rate": 0.02,
+      "loss": 1.6845,
+      "step": 1568
+    },
+    {
+      "epoch": 0.145252560319389,
+      "grad_norm": 0.1572265625,
+      "learning_rate": 0.02,
+      "loss": 1.6983,
+      "step": 1569
+    },
+    {
+      "epoch": 0.14534513683966904,
+      "grad_norm": 0.1513671875,
+      "learning_rate": 0.02,
+      "loss": 1.6766,
+      "step": 1570
+    },
+    {
+      "epoch": 0.1454377133599491,
+      "grad_norm": 0.15625,
+      "learning_rate": 0.02,
+      "loss": 1.7793,
+      "step": 1571
+    },
+    {
+      "epoch": 0.14553028988022912,
+      "grad_norm": 0.158203125,
+      "learning_rate": 0.02,
+      "loss": 1.6795,
+      "step": 1572
+    },
+    {
+      "epoch": 0.14562286640050917,
+      "grad_norm": 0.1552734375,
+      "learning_rate": 0.02,
+      "loss": 1.8063,
+      "step": 1573
+    },
+    {
+      "epoch": 0.14571544292078922,
+      "grad_norm": 0.1494140625,
+      "learning_rate": 0.02,
+      "loss": 1.72,
+      "step": 1574
+    },
+    {
+      "epoch": 0.14580801944106925,
+      "grad_norm": 0.1552734375,
+      "learning_rate": 0.02,
+      "loss": 1.818,
+      "step": 1575
+    },
+    {
+      "epoch": 0.1459005959613493,
+      "grad_norm": 0.1552734375,
+      "learning_rate": 0.02,
+      "loss": 1.724,
+      "step": 1576
+    },
+    {
+      "epoch": 0.14599317248162935,
+      "grad_norm": 0.1611328125,
+      "learning_rate": 0.02,
+      "loss": 1.7855,
+      "step": 1577
+    },
+    {
+      "epoch": 0.1460857490019094,
+      "grad_norm": 0.158203125,
+      "learning_rate": 0.02,
+      "loss": 1.7232,
+      "step": 1578
+    },
+    {
+      "epoch": 0.14617832552218943,
+      "grad_norm": 0.158203125,
+      "learning_rate": 0.02,
+      "loss": 1.7863,
+      "step": 1579
+    },
+    {
+      "epoch": 0.14627090204246948,
+      "grad_norm": 0.1533203125,
+      "learning_rate": 0.02,
+      "loss": 1.7437,
+      "step": 1580
+    },
+    {
+      "epoch": 0.14636347856274953,
+      "grad_norm": 0.169921875,
+      "learning_rate": 0.02,
+      "loss": 1.7387,
+      "step": 1581
+    },
+    {
+      "epoch": 0.14645605508302956,
+      "grad_norm": 0.1533203125,
+      "learning_rate": 0.02,
+      "loss": 1.7742,
+      "step": 1582
+    },
+    {
+      "epoch": 0.1465486316033096,
+      "grad_norm": 0.1474609375,
+      "learning_rate": 0.02,
+      "loss": 1.7453,
+      "step": 1583
+    },
+    {
+      "epoch": 0.14664120812358966,
+      "grad_norm": 0.15625,
+      "learning_rate": 0.02,
+      "loss": 1.6541,
+      "step": 1584
+    },
+    {
+      "epoch": 0.14673378464386969,
+      "grad_norm": 0.158203125,
+      "learning_rate": 0.02,
+      "loss": 1.7074,
+      "step": 1585
+    },
+    {
+      "epoch": 0.14682636116414974,
+      "grad_norm": 0.1591796875,
+      "learning_rate": 0.02,
+      "loss": 1.7341,
+      "step": 1586
+    },
+    {
+      "epoch": 0.1469189376844298,
+      "grad_norm": 0.1708984375,
+      "learning_rate": 0.02,
+      "loss": 1.787,
+      "step": 1587
+    },
+    {
+      "epoch": 0.14701151420470984,
+      "grad_norm": 0.1708984375,
+      "learning_rate": 0.02,
+      "loss": 1.7208,
+      "step": 1588
+    },
+    {
+      "epoch": 0.14710409072498987,
+      "grad_norm": 0.1728515625,
+      "learning_rate": 0.02,
+      "loss": 1.7693,
+      "step": 1589
+    },
+    {
+      "epoch": 0.14719666724526992,
+      "grad_norm": 0.1748046875,
+      "learning_rate": 0.02,
+      "loss": 1.7527,
+      "step": 1590
+    },
+    {
+      "epoch": 0.14728924376554997,
+      "grad_norm": 0.1572265625,
+      "learning_rate": 0.02,
+      "loss": 1.6955,
+      "step": 1591
+    },
+    {
+      "epoch": 0.14738182028583,
+      "grad_norm": 0.1513671875,
+      "learning_rate": 0.02,
+      "loss": 1.6725,
+      "step": 1592
+    },
+    {
+      "epoch": 0.14747439680611005,
+      "grad_norm": 0.1591796875,
+      "learning_rate": 0.02,
+      "loss": 1.722,
+      "step": 1593
+    },
+    {
+      "epoch": 0.1475669733263901,
+      "grad_norm": 0.166015625,
+      "learning_rate": 0.02,
+      "loss": 1.7813,
+      "step": 1594
+    },
+    {
+      "epoch": 0.14765954984667015,
+      "grad_norm": 0.1552734375,
+      "learning_rate": 0.02,
+      "loss": 1.7214,
+      "step": 1595
+    },
+    {
+      "epoch": 0.14775212636695018,
+      "grad_norm": 0.158203125,
+      "learning_rate": 0.02,
+      "loss": 1.6779,
+      "step": 1596
+    },
+    {
+      "epoch": 0.14784470288723023,
+      "grad_norm": 0.1484375,
+      "learning_rate": 0.02,
+      "loss": 1.6628,
+      "step": 1597
+    },
+    {
+      "epoch": 0.14793727940751028,
+      "grad_norm": 0.146484375,
+      "learning_rate": 0.02,
+      "loss": 1.7532,
+      "step": 1598
+    },
+    {
+      "epoch": 0.1480298559277903,
+      "grad_norm": 0.1533203125,
+      "learning_rate": 0.02,
+      "loss": 1.7023,
+      "step": 1599
+    },
+    {
+      "epoch": 0.14812243244807036,
+      "grad_norm": 0.1513671875,
+      "learning_rate": 0.02,
+      "loss": 1.7006,
+      "step": 1600
+    },
+    {
+      "epoch": 0.1482150089683504,
+      "grad_norm": 0.1611328125,
+      "learning_rate": 0.02,
+      "loss": 1.7344,
+      "step": 1601
+    },
+    {
+      "epoch": 0.14830758548863043,
+      "grad_norm": 0.1591796875,
+      "learning_rate": 0.02,
+      "loss": 1.7522,
+      "step": 1602
+    },
+    {
+      "epoch": 0.14840016200891049,
+      "grad_norm": 0.1591796875,
+      "learning_rate": 0.02,
+      "loss": 1.6943,
+      "step": 1603
+    },
+    {
+      "epoch": 0.14849273852919054,
+      "grad_norm": 0.1533203125,
+      "learning_rate": 0.02,
+      "loss": 1.7198,
+      "step": 1604
+    },
+    {
+      "epoch": 0.1485853150494706,
+      "grad_norm": 0.1455078125,
+      "learning_rate": 0.02,
+      "loss": 1.7337,
+      "step": 1605
+    },
+    {
+      "epoch": 0.14867789156975061,
+      "grad_norm": 0.158203125,
+      "learning_rate": 0.02,
+      "loss": 1.6912,
+      "step": 1606
+    },
+    {
+      "epoch": 0.14877046809003067,
+      "grad_norm": 0.162109375,
+      "learning_rate": 0.02,
+      "loss": 1.764,
+      "step": 1607
+    },
+    {
+      "epoch": 0.14886304461031072,
+      "grad_norm": 0.17578125,
+      "learning_rate": 0.02,
+      "loss": 1.8474,
+      "step": 1608
+    },
+    {
+      "epoch": 0.14895562113059074,
+      "grad_norm": 0.1552734375,
+      "learning_rate": 0.02,
+      "loss": 1.6748,
+      "step": 1609
+    },
+    {
+      "epoch": 0.1490481976508708,
+      "grad_norm": 0.16796875,
+      "learning_rate": 0.02,
+      "loss": 1.7134,
+      "step": 1610
+    },
+    {
+      "epoch": 0.14914077417115085,
+      "grad_norm": 0.1552734375,
+      "learning_rate": 0.02,
+      "loss": 1.6912,
+      "step": 1611
+    },
+    {
+      "epoch": 0.1492333506914309,
+      "grad_norm": 0.16015625,
+      "learning_rate": 0.02,
+      "loss": 1.7664,
+      "step": 1612
+    },
+    {
+      "epoch": 0.14932592721171092,
+      "grad_norm": 0.1748046875,
+      "learning_rate": 0.02,
+      "loss": 1.7777,
+      "step": 1613
+    },
+    {
+      "epoch": 0.14941850373199098,
+      "grad_norm": 0.1591796875,
+      "learning_rate": 0.02,
+      "loss": 1.7536,
+      "step": 1614
+    },
+    {
+      "epoch": 0.14951108025227103,
+      "grad_norm": 0.1669921875,
+      "learning_rate": 0.02,
+      "loss": 1.7571,
+      "step": 1615
+    },
+    {
+      "epoch": 0.14960365677255105,
+      "grad_norm": 0.1630859375,
+      "learning_rate": 0.02,
+      "loss": 1.6507,
+      "step": 1616
+    },
+    {
+      "epoch": 0.1496962332928311,
+      "grad_norm": 0.158203125,
+      "learning_rate": 0.02,
+      "loss": 1.7261,
+      "step": 1617
+    },
+    {
+      "epoch": 0.14978880981311116,
+      "grad_norm": 0.16015625,
+      "learning_rate": 0.02,
+      "loss": 1.7554,
+      "step": 1618
+    },
+    {
+      "epoch": 0.14988138633339118,
+      "grad_norm": 0.1708984375,
+      "learning_rate": 0.02,
+      "loss": 1.6885,
+      "step": 1619
+    },
+    {
+      "epoch": 0.14997396285367123,
+      "grad_norm": 0.1552734375,
+      "learning_rate": 0.02,
+      "loss": 1.6819,
+      "step": 1620
+    },
+    {
+      "epoch": 0.15006653937395129,
+      "grad_norm": 0.1611328125,
+      "learning_rate": 0.02,
+      "loss": 1.7647,
+      "step": 1621
+    },
+    {
+      "epoch": 0.15015911589423134,
+      "grad_norm": 0.1435546875,
+      "learning_rate": 0.02,
+      "loss": 1.7588,
+      "step": 1622
+    },
+    {
+      "epoch": 0.15025169241451136,
+      "grad_norm": 0.1357421875,
+      "learning_rate": 0.02,
+      "loss": 1.6873,
+      "step": 1623
+    },
+    {
+      "epoch": 0.15034426893479141,
+      "grad_norm": 0.1484375,
+      "learning_rate": 0.02,
+      "loss": 1.6251,
+      "step": 1624
+    },
+    {
+      "epoch": 0.15043684545507147,
+      "grad_norm": 0.150390625,
+      "learning_rate": 0.02,
+      "loss": 1.7115,
+      "step": 1625
+    },
+    {
+      "epoch": 0.1505294219753515,
+      "grad_norm": 0.162109375,
+      "learning_rate": 0.02,
+      "loss": 1.7399,
+      "step": 1626
+    },
+    {
+      "epoch": 0.15062199849563154,
+      "grad_norm": 0.1689453125,
+      "learning_rate": 0.02,
+      "loss": 1.7085,
+      "step": 1627
+    },
+    {
+      "epoch": 0.1507145750159116,
+      "grad_norm": 0.1630859375,
+      "learning_rate": 0.02,
+      "loss": 1.6791,
+      "step": 1628
+    },
+    {
+      "epoch": 0.15080715153619165,
+      "grad_norm": 0.1611328125,
+      "learning_rate": 0.02,
+      "loss": 1.6827,
+      "step": 1629
+    },
+    {
+      "epoch": 0.15089972805647167,
+      "grad_norm": 0.1748046875,
+      "learning_rate": 0.02,
+      "loss": 1.7669,
+      "step": 1630
+    },
+    {
+      "epoch": 0.15099230457675172,
+      "grad_norm": 0.1630859375,
+      "learning_rate": 0.02,
+      "loss": 1.7021,
+      "step": 1631
+    },
+    {
+      "epoch": 0.15108488109703178,
+      "grad_norm": 0.1650390625,
+      "learning_rate": 0.02,
+      "loss": 1.705,
+      "step": 1632
+    },
+    {
+      "epoch": 0.1511774576173118,
+      "grad_norm": 0.171875,
+      "learning_rate": 0.02,
+      "loss": 1.7375,
+      "step": 1633
+    },
+    {
+      "epoch": 0.15127003413759185,
+      "grad_norm": 0.169921875,
+      "learning_rate": 0.02,
+      "loss": 1.8,
+      "step": 1634
+    },
+    {
+      "epoch": 0.1513626106578719,
+      "grad_norm": 0.1650390625,
+      "learning_rate": 0.02,
+      "loss": 1.7563,
+      "step": 1635
+    },
+    {
+      "epoch": 0.15145518717815193,
+      "grad_norm": 0.16796875,
+      "learning_rate": 0.02,
+      "loss": 1.79,
+      "step": 1636
+    },
+    {
+      "epoch": 0.15154776369843198,
+      "grad_norm": 0.1708984375,
+      "learning_rate": 0.02,
+      "loss": 1.7276,
+      "step": 1637
+    },
+    {
+      "epoch": 0.15164034021871203,
+      "grad_norm": 0.1689453125,
+      "learning_rate": 0.02,
+      "loss": 1.7292,
+      "step": 1638
+    },
+    {
+      "epoch": 0.15173291673899209,
+      "grad_norm": 0.1748046875,
+      "learning_rate": 0.02,
+      "loss": 1.7643,
+      "step": 1639
+    },
+    {
+      "epoch": 0.1518254932592721,
+      "grad_norm": 0.16796875,
+      "learning_rate": 0.02,
+      "loss": 1.6778,
+      "step": 1640
+    },
+    {
+      "epoch": 0.15191806977955216,
+      "grad_norm": 0.171875,
+      "learning_rate": 0.02,
+      "loss": 1.7134,
+      "step": 1641
+    },
+    {
+      "epoch": 0.15201064629983221,
+      "grad_norm": 0.1748046875,
+      "learning_rate": 0.02,
+      "loss": 1.7491,
+      "step": 1642
+    },
+    {
+      "epoch": 0.15210322282011224,
+      "grad_norm": 0.1640625,
+      "learning_rate": 0.02,
+      "loss": 1.7118,
+      "step": 1643
+    },
+    {
+      "epoch": 0.1521957993403923,
+      "grad_norm": 0.1689453125,
+      "learning_rate": 0.02,
+      "loss": 1.7222,
+      "step": 1644
+    },
+    {
+      "epoch": 0.15228837586067234,
+      "grad_norm": 0.16015625,
+      "learning_rate": 0.02,
+      "loss": 1.667,
+      "step": 1645
+    },
+    {
+      "epoch": 0.1523809523809524,
+      "grad_norm": 0.158203125,
+      "learning_rate": 0.02,
+      "loss": 1.6846,
+      "step": 1646
+    },
+    {
+      "epoch": 0.15247352890123242,
+      "grad_norm": 0.169921875,
+      "learning_rate": 0.02,
+      "loss": 1.733,
+      "step": 1647
+    },
+    {
+      "epoch": 0.15256610542151247,
+      "grad_norm": 0.154296875,
+      "learning_rate": 0.02,
+      "loss": 1.6514,
+      "step": 1648
+    },
+    {
+      "epoch": 0.15265868194179252,
+      "grad_norm": 0.146484375,
+      "learning_rate": 0.02,
+      "loss": 1.7198,
+      "step": 1649
+    },
+    {
+      "epoch": 0.15275125846207255,
+      "grad_norm": 0.1396484375,
+      "learning_rate": 0.02,
+      "loss": 1.6607,
+      "step": 1650
+    },
+    {
+      "epoch": 0.1528438349823526,
+      "grad_norm": 0.15234375,
+      "learning_rate": 0.02,
+      "loss": 1.7451,
+      "step": 1651
+    },
+    {
+      "epoch": 0.15293641150263265,
+      "grad_norm": 0.1650390625,
+      "learning_rate": 0.02,
+      "loss": 1.7399,
+      "step": 1652
+    },
+    {
+      "epoch": 0.15302898802291268,
+      "grad_norm": 0.162109375,
+      "learning_rate": 0.02,
+      "loss": 1.7182,
+      "step": 1653
+    },
+    {
+      "epoch": 0.15312156454319273,
+      "grad_norm": 0.158203125,
+      "learning_rate": 0.02,
+      "loss": 1.7885,
+      "step": 1654
+    },
+    {
+      "epoch": 0.15321414106347278,
+      "grad_norm": 0.15625,
+      "learning_rate": 0.02,
+      "loss": 1.7413,
+      "step": 1655
+    },
+    {
+      "epoch": 0.15330671758375283,
+      "grad_norm": 0.1474609375,
+      "learning_rate": 0.02,
+      "loss": 1.6885,
+      "step": 1656
+    },
+    {
+      "epoch": 0.15339929410403286,
+      "grad_norm": 0.15625,
+      "learning_rate": 0.02,
+      "loss": 1.7265,
+      "step": 1657
+    },
+    {
+      "epoch": 0.1534918706243129,
+      "grad_norm": 0.1533203125,
+      "learning_rate": 0.02,
+      "loss": 1.7435,
+      "step": 1658
+    },
+    {
+      "epoch": 0.15358444714459296,
+      "grad_norm": 0.1650390625,
+      "learning_rate": 0.02,
+      "loss": 1.7849,
+      "step": 1659
+    },
+    {
+      "epoch": 0.15367702366487299,
+      "grad_norm": 0.1640625,
+      "learning_rate": 0.02,
+      "loss": 1.673,
+      "step": 1660
+    },
+    {
+      "epoch": 0.15376960018515304,
+      "grad_norm": 0.15625,
+      "learning_rate": 0.02,
+      "loss": 1.7819,
+      "step": 1661
+    },
+    {
+      "epoch": 0.1538621767054331,
+      "grad_norm": 0.1552734375,
+      "learning_rate": 0.02,
+      "loss": 1.767,
+      "step": 1662
+    },
+    {
+      "epoch": 0.15395475322571314,
+      "grad_norm": 0.158203125,
+      "learning_rate": 0.02,
+      "loss": 1.8164,
+      "step": 1663
+    },
+    {
+      "epoch": 0.15404732974599317,
+      "grad_norm": 0.150390625,
+      "learning_rate": 0.02,
+      "loss": 1.7146,
+      "step": 1664
+    },
+    {
+      "epoch": 0.15413990626627322,
+      "grad_norm": 0.1474609375,
+      "learning_rate": 0.02,
+      "loss": 1.729,
+      "step": 1665
+    },
+    {
+      "epoch": 0.15423248278655327,
+      "grad_norm": 0.1474609375,
+      "learning_rate": 0.02,
+      "loss": 1.6952,
+      "step": 1666
+    },
+    {
+      "epoch": 0.1543250593068333,
+      "grad_norm": 0.166015625,
+      "learning_rate": 0.02,
+      "loss": 1.7781,
+      "step": 1667
+    },
+    {
+      "epoch": 0.15441763582711335,
+      "grad_norm": 0.1513671875,
+      "learning_rate": 0.02,
+      "loss": 1.6934,
+      "step": 1668
+    },
+    {
+      "epoch": 0.1545102123473934,
+      "grad_norm": 0.1513671875,
+      "learning_rate": 0.02,
+      "loss": 1.7041,
+      "step": 1669
+    },
+    {
+      "epoch": 0.15460278886767342,
+      "grad_norm": 0.1533203125,
+      "learning_rate": 0.02,
+      "loss": 1.7173,
+      "step": 1670
+    },
+    {
+      "epoch": 0.15469536538795348,
+      "grad_norm": 0.1572265625,
+      "learning_rate": 0.02,
+      "loss": 1.7129,
+      "step": 1671
+    },
+    {
+      "epoch": 0.15478794190823353,
+      "grad_norm": 0.1533203125,
+      "learning_rate": 0.02,
+      "loss": 1.7837,
+      "step": 1672
+    },
+    {
+      "epoch": 0.15488051842851358,
+      "grad_norm": 0.1611328125,
+      "learning_rate": 0.02,
+      "loss": 1.6999,
+      "step": 1673
+    },
+    {
+      "epoch": 0.1549730949487936,
+      "grad_norm": 0.1767578125,
+      "learning_rate": 0.02,
+      "loss": 1.6898,
+      "step": 1674
+    },
+    {
+      "epoch": 0.15506567146907366,
+      "grad_norm": 0.1669921875,
+      "learning_rate": 0.02,
+      "loss": 1.7481,
+      "step": 1675
+    },
+    {
+      "epoch": 0.1551582479893537,
+      "grad_norm": 0.150390625,
+      "learning_rate": 0.02,
+      "loss": 1.7452,
+      "step": 1676
+    },
+    {
+      "epoch": 0.15525082450963373,
+      "grad_norm": 0.1552734375,
+      "learning_rate": 0.02,
+      "loss": 1.6663,
+      "step": 1677
+    },
+    {
+      "epoch": 0.15534340102991379,
+      "grad_norm": 0.1591796875,
+      "learning_rate": 0.02,
+      "loss": 1.7347,
+      "step": 1678
+    },
+    {
+      "epoch": 0.15543597755019384,
+      "grad_norm": 0.16015625,
+      "learning_rate": 0.02,
+      "loss": 1.6629,
+      "step": 1679
+    },
+    {
+      "epoch": 0.15552855407047386,
+      "grad_norm": 0.166015625,
+      "learning_rate": 0.02,
+      "loss": 1.7273,
+      "step": 1680
+    },
+    {
+      "epoch": 0.15562113059075391,
+      "grad_norm": 0.171875,
+      "learning_rate": 0.02,
+      "loss": 1.6701,
+      "step": 1681
+    },
+    {
+      "epoch": 0.15571370711103397,
+      "grad_norm": 0.1611328125,
+      "learning_rate": 0.02,
+      "loss": 1.6781,
+      "step": 1682
+    },
+    {
+      "epoch": 0.15580628363131402,
+      "grad_norm": 0.162109375,
+      "learning_rate": 0.02,
+      "loss": 1.7793,
+      "step": 1683
+    },
+    {
+      "epoch": 0.15589886015159404,
+      "grad_norm": 0.171875,
+      "learning_rate": 0.02,
+      "loss": 1.7159,
+      "step": 1684
+    },
+    {
+      "epoch": 0.1559914366718741,
+      "grad_norm": 0.1640625,
+      "learning_rate": 0.02,
+      "loss": 1.7607,
+      "step": 1685
+    },
+    {
+      "epoch": 0.15608401319215415,
+      "grad_norm": 0.1640625,
+      "learning_rate": 0.02,
+      "loss": 1.77,
+      "step": 1686
+    },
+    {
+      "epoch": 0.15617658971243417,
+      "grad_norm": 0.162109375,
+      "learning_rate": 0.02,
+      "loss": 1.7372,
+      "step": 1687
+    },
+    {
+      "epoch": 0.15626916623271422,
+      "grad_norm": 0.150390625,
+      "learning_rate": 0.02,
+      "loss": 1.7087,
+      "step": 1688
+    },
+    {
+      "epoch": 0.15636174275299428,
+      "grad_norm": 0.1494140625,
+      "learning_rate": 0.02,
+      "loss": 1.6929,
+      "step": 1689
+    },
+    {
+      "epoch": 0.15645431927327433,
+      "grad_norm": 0.1640625,
+      "learning_rate": 0.02,
+      "loss": 1.7713,
+      "step": 1690
+    },
+    {
+      "epoch": 0.15654689579355435,
+      "grad_norm": 0.1650390625,
+      "learning_rate": 0.02,
+      "loss": 1.7394,
+      "step": 1691
+    },
+    {
+      "epoch": 0.1566394723138344,
+      "grad_norm": 0.15625,
+      "learning_rate": 0.02,
+      "loss": 1.6709,
+      "step": 1692
+    },
+    {
+      "epoch": 0.15673204883411446,
+      "grad_norm": 0.1572265625,
+      "learning_rate": 0.02,
+      "loss": 1.7308,
+      "step": 1693
+    },
+    {
+      "epoch": 0.15682462535439448,
+      "grad_norm": 0.1552734375,
+      "learning_rate": 0.02,
+      "loss": 1.6811,
+      "step": 1694
+    },
+    {
+      "epoch": 0.15691720187467453,
+      "grad_norm": 0.1572265625,
+      "learning_rate": 0.02,
+      "loss": 1.7271,
+      "step": 1695
+    },
+    {
+      "epoch": 0.15700977839495459,
+      "grad_norm": 0.1630859375,
+      "learning_rate": 0.02,
+      "loss": 1.7058,
+      "step": 1696
+    },
+    {
+      "epoch": 0.1571023549152346,
+      "grad_norm": 0.1640625,
+      "learning_rate": 0.02,
+      "loss": 1.7717,
+      "step": 1697
+    },
+    {
+      "epoch": 0.15719493143551466,
+      "grad_norm": 0.17578125,
+      "learning_rate": 0.02,
+      "loss": 1.6923,
+      "step": 1698
+    },
+    {
+      "epoch": 0.15728750795579471,
+      "grad_norm": 0.1494140625,
+      "learning_rate": 0.02,
+      "loss": 1.7289,
+      "step": 1699
+    },
+    {
+      "epoch": 0.15738008447607477,
+      "grad_norm": 0.1572265625,
+      "learning_rate": 0.02,
+      "loss": 1.73,
+      "step": 1700
+    },
+    {
+      "epoch": 0.1574726609963548,
+      "grad_norm": 0.1484375,
+      "learning_rate": 0.02,
+      "loss": 1.6899,
+      "step": 1701
+    },
+    {
+      "epoch": 0.15756523751663484,
+      "grad_norm": 0.150390625,
+      "learning_rate": 0.02,
+      "loss": 1.7319,
+      "step": 1702
+    },
+    {
+      "epoch": 0.1576578140369149,
+      "grad_norm": 0.16015625,
+      "learning_rate": 0.02,
+      "loss": 1.7535,
+      "step": 1703
+    },
+    {
+      "epoch": 0.15775039055719492,
+      "grad_norm": 0.158203125,
+      "learning_rate": 0.02,
+      "loss": 1.6829,
+      "step": 1704
+    },
+    {
+      "epoch": 0.15784296707747497,
+      "grad_norm": 0.16796875,
+      "learning_rate": 0.02,
+      "loss": 1.7318,
+      "step": 1705
+    },
+    {
+      "epoch": 0.15793554359775502,
+      "grad_norm": 0.1513671875,
+      "learning_rate": 0.02,
+      "loss": 1.7082,
+      "step": 1706
+    },
+    {
+      "epoch": 0.15802812011803508,
+      "grad_norm": 0.1572265625,
+      "learning_rate": 0.02,
+      "loss": 1.6716,
+      "step": 1707
+    },
+    {
+      "epoch": 0.1581206966383151,
+      "grad_norm": 0.15625,
+      "learning_rate": 0.02,
+      "loss": 1.7782,
+      "step": 1708
+    },
+    {
+      "epoch": 0.15821327315859515,
+      "grad_norm": 0.158203125,
+      "learning_rate": 0.02,
+      "loss": 1.7259,
+      "step": 1709
+    },
+    {
+      "epoch": 0.1583058496788752,
+      "grad_norm": 0.1591796875,
+      "learning_rate": 0.02,
+      "loss": 1.7213,
+      "step": 1710
+    },
+    {
+      "epoch": 0.15839842619915523,
+      "grad_norm": 0.1689453125,
+      "learning_rate": 0.02,
+      "loss": 1.7482,
+      "step": 1711
+    },
+    {
+      "epoch": 0.15849100271943528,
+      "grad_norm": 0.1533203125,
+      "learning_rate": 0.02,
+      "loss": 1.668,
+      "step": 1712
+    },
+    {
+      "epoch": 0.15858357923971533,
+      "grad_norm": 0.1552734375,
+      "learning_rate": 0.02,
+      "loss": 1.7365,
+      "step": 1713
+    },
+    {
+      "epoch": 0.15867615575999536,
+      "grad_norm": 0.1640625,
+      "learning_rate": 0.02,
+      "loss": 1.6991,
+      "step": 1714
+    },
+    {
+      "epoch": 0.1587687322802754,
+      "grad_norm": 0.1640625,
+      "learning_rate": 0.02,
+      "loss": 1.7288,
+      "step": 1715
+    },
+    {
+      "epoch": 0.15886130880055546,
+      "grad_norm": 0.1572265625,
+      "learning_rate": 0.02,
+      "loss": 1.7081,
+      "step": 1716
+    },
+    {
+      "epoch": 0.15895388532083551,
+      "grad_norm": 0.1552734375,
+      "learning_rate": 0.02,
+      "loss": 1.7055,
+      "step": 1717
+    },
+    {
+      "epoch": 0.15904646184111554,
+      "grad_norm": 0.154296875,
+      "learning_rate": 0.02,
+      "loss": 1.7039,
+      "step": 1718
+    },
+    {
+      "epoch": 0.1591390383613956,
+      "grad_norm": 0.1494140625,
+      "learning_rate": 0.02,
+      "loss": 1.6755,
+      "step": 1719
+    },
+    {
+      "epoch": 0.15923161488167564,
+      "grad_norm": 0.1572265625,
+      "learning_rate": 0.02,
+      "loss": 1.7025,
+      "step": 1720
+    },
+    {
+      "epoch": 0.15932419140195567,
+      "grad_norm": 0.16015625,
+      "learning_rate": 0.02,
+      "loss": 1.7182,
+      "step": 1721
+    },
+    {
+      "epoch": 0.15941676792223572,
+      "grad_norm": 0.169921875,
+      "learning_rate": 0.02,
+      "loss": 1.7508,
+      "step": 1722
+    },
+    {
+      "epoch": 0.15950934444251577,
+      "grad_norm": 0.1650390625,
+      "learning_rate": 0.02,
+      "loss": 1.6867,
+      "step": 1723
+    },
+    {
+      "epoch": 0.15960192096279582,
+      "grad_norm": 0.16015625,
+      "learning_rate": 0.02,
+      "loss": 1.7096,
+      "step": 1724
+    },
+    {
+      "epoch": 0.15969449748307585,
+      "grad_norm": 0.154296875,
+      "learning_rate": 0.02,
+      "loss": 1.7029,
+      "step": 1725
+    },
+    {
+      "epoch": 0.1597870740033559,
+      "grad_norm": 0.1611328125,
+      "learning_rate": 0.02,
+      "loss": 1.7493,
+      "step": 1726
+    },
+    {
+      "epoch": 0.15987965052363595,
+      "grad_norm": 0.1494140625,
+      "learning_rate": 0.02,
+      "loss": 1.6887,
+      "step": 1727
+    },
+    {
+      "epoch": 0.15997222704391598,
+      "grad_norm": 0.1552734375,
+      "learning_rate": 0.02,
+      "loss": 1.7744,
+      "step": 1728
+    },
+    {
+      "epoch": 0.16006480356419603,
+      "grad_norm": 0.146484375,
+      "learning_rate": 0.02,
+      "loss": 1.7936,
+      "step": 1729
+    },
+    {
+      "epoch": 0.16015738008447608,
+      "grad_norm": 0.1591796875,
+      "learning_rate": 0.02,
+      "loss": 1.7217,
+      "step": 1730
+    },
+    {
+      "epoch": 0.1602499566047561,
+      "grad_norm": 0.5859375,
+      "learning_rate": 0.02,
+      "loss": 1.6995,
+      "step": 1731
+    },
+    {
+      "epoch": 0.16034253312503616,
+      "grad_norm": 0.1298828125,
+      "learning_rate": 0.02,
+      "loss": 1.6918,
+      "step": 1732
+    },
+    {
+      "epoch": 0.1604351096453162,
+      "grad_norm": 0.150390625,
+      "learning_rate": 0.02,
+      "loss": 1.7681,
+      "step": 1733
+    },
+    {
+      "epoch": 0.16052768616559626,
+      "grad_norm": 0.1533203125,
+      "learning_rate": 0.02,
+      "loss": 1.6322,
+      "step": 1734
+    },
+    {
+      "epoch": 0.16062026268587629,
+      "grad_norm": 0.162109375,
+      "learning_rate": 0.02,
+      "loss": 1.6792,
+      "step": 1735
+    },
+    {
+      "epoch": 0.16071283920615634,
+      "grad_norm": 0.1767578125,
+      "learning_rate": 0.02,
+      "loss": 1.7489,
+      "step": 1736
+    },
+    {
+      "epoch": 0.1608054157264364,
+      "grad_norm": 0.1787109375,
+      "learning_rate": 0.02,
+      "loss": 1.7204,
+      "step": 1737
+    },
+    {
+      "epoch": 0.16089799224671641,
+      "grad_norm": 0.1708984375,
+      "learning_rate": 0.02,
+      "loss": 1.6787,
+      "step": 1738
+    },
+    {
+      "epoch": 0.16099056876699647,
+      "grad_norm": 0.1689453125,
+      "learning_rate": 0.02,
+      "loss": 1.6703,
+      "step": 1739
+    },
+    {
+      "epoch": 0.16108314528727652,
+      "grad_norm": 0.16015625,
+      "learning_rate": 0.02,
+      "loss": 1.7162,
+      "step": 1740
+    },
+    {
+      "epoch": 0.16117572180755657,
+      "grad_norm": 0.158203125,
+      "learning_rate": 0.02,
+      "loss": 1.6905,
+      "step": 1741
+    },
+    {
+      "epoch": 0.1612682983278366,
+      "grad_norm": 0.177734375,
+      "learning_rate": 0.02,
+      "loss": 1.7373,
+      "step": 1742
+    },
+    {
+      "epoch": 0.16136087484811665,
+      "grad_norm": 0.158203125,
+      "learning_rate": 0.02,
+      "loss": 1.6403,
+      "step": 1743
+    },
+    {
+      "epoch": 0.1614534513683967,
+      "grad_norm": 0.158203125,
+      "learning_rate": 0.02,
+      "loss": 1.7284,
+      "step": 1744
+    },
+    {
+      "epoch": 0.16154602788867672,
+      "grad_norm": 0.1689453125,
+      "learning_rate": 0.02,
+      "loss": 1.7715,
+      "step": 1745
+    },
+    {
+      "epoch": 0.16163860440895678,
+      "grad_norm": 0.1650390625,
+      "learning_rate": 0.02,
+      "loss": 1.7437,
+      "step": 1746
+    },
+    {
+      "epoch": 0.16173118092923683,
+      "grad_norm": 0.1611328125,
+      "learning_rate": 0.02,
+      "loss": 1.6859,
+      "step": 1747
+    },
+    {
+      "epoch": 0.16182375744951685,
+      "grad_norm": 0.1611328125,
+      "learning_rate": 0.02,
+      "loss": 1.7414,
+      "step": 1748
+    },
+    {
+      "epoch": 0.1619163339697969,
+      "grad_norm": 0.1572265625,
+      "learning_rate": 0.02,
+      "loss": 1.6924,
+      "step": 1749
+    },
+    {
+      "epoch": 0.16200891049007696,
+      "grad_norm": 0.16796875,
+      "learning_rate": 0.02,
+      "loss": 1.7354,
+      "step": 1750
+    },
+    {
+      "epoch": 0.162101487010357,
+      "grad_norm": 0.1630859375,
+      "learning_rate": 0.02,
+      "loss": 1.7759,
+      "step": 1751
+    },
+    {
+      "epoch": 0.16219406353063703,
+      "grad_norm": 0.1572265625,
+      "learning_rate": 0.02,
+      "loss": 1.68,
+      "step": 1752
+    },
+    {
+      "epoch": 0.16228664005091709,
+      "grad_norm": 0.1572265625,
+      "learning_rate": 0.02,
+      "loss": 1.7465,
+      "step": 1753
+    },
+    {
+      "epoch": 0.16237921657119714,
+      "grad_norm": 0.1689453125,
+      "learning_rate": 0.02,
+      "loss": 1.7368,
+      "step": 1754
+    },
+    {
+      "epoch": 0.16247179309147716,
+      "grad_norm": 0.1708984375,
+      "learning_rate": 0.02,
+      "loss": 1.7238,
+      "step": 1755
+    },
+    {
+      "epoch": 0.16256436961175721,
+      "grad_norm": 0.1708984375,
+      "learning_rate": 0.02,
+      "loss": 1.7112,
+      "step": 1756
+    },
+    {
+      "epoch": 0.16265694613203727,
+      "grad_norm": 0.1708984375,
+      "learning_rate": 0.02,
+      "loss": 1.7656,
+      "step": 1757
+    },
+    {
+      "epoch": 0.16274952265231732,
+      "grad_norm": 0.16015625,
+      "learning_rate": 0.02,
+      "loss": 1.7135,
+      "step": 1758
+    },
+    {
+      "epoch": 0.16284209917259734,
+      "grad_norm": 0.162109375,
+      "learning_rate": 0.02,
+      "loss": 1.6978,
+      "step": 1759
+    },
+    {
+      "epoch": 0.1629346756928774,
+      "grad_norm": 0.15234375,
+      "learning_rate": 0.02,
+      "loss": 1.6596,
+      "step": 1760
+    },
+    {
+      "epoch": 0.16302725221315745,
+      "grad_norm": 0.1484375,
+      "learning_rate": 0.02,
+      "loss": 1.7043,
+      "step": 1761
+    },
+    {
+      "epoch": 0.16311982873343747,
+      "grad_norm": 0.1572265625,
+      "learning_rate": 0.02,
+      "loss": 1.7538,
+      "step": 1762
+    },
+    {
+      "epoch": 0.16321240525371752,
+      "grad_norm": 0.162109375,
+      "learning_rate": 0.02,
+      "loss": 1.7601,
+      "step": 1763
+    },
+    {
+      "epoch": 0.16330498177399758,
+      "grad_norm": 0.1533203125,
+      "learning_rate": 0.02,
+      "loss": 1.7431,
+      "step": 1764
+    },
+    {
+      "epoch": 0.1633975582942776,
+      "grad_norm": 0.154296875,
+      "learning_rate": 0.02,
+      "loss": 1.6536,
+      "step": 1765
+    },
+    {
+      "epoch": 0.16349013481455765,
+      "grad_norm": 0.1552734375,
+      "learning_rate": 0.02,
+      "loss": 1.6967,
+      "step": 1766
+    },
+    {
+      "epoch": 0.1635827113348377,
+      "grad_norm": 0.1513671875,
+      "learning_rate": 0.02,
+      "loss": 1.7011,
+      "step": 1767
+    },
+    {
+      "epoch": 0.16367528785511776,
+      "grad_norm": 0.1572265625,
+      "learning_rate": 0.02,
+      "loss": 1.7324,
+      "step": 1768
+    },
+    {
+      "epoch": 0.16376786437539778,
+      "grad_norm": 0.166015625,
+      "learning_rate": 0.02,
+      "loss": 1.7582,
+      "step": 1769
+    },
+    {
+      "epoch": 0.16386044089567783,
+      "grad_norm": 0.1533203125,
+      "learning_rate": 0.02,
+      "loss": 1.6755,
+      "step": 1770
+    },
+    {
+      "epoch": 0.16395301741595789,
+      "grad_norm": 0.150390625,
+      "learning_rate": 0.02,
+      "loss": 1.7652,
+      "step": 1771
+    },
+    {
+      "epoch": 0.1640455939362379,
+      "grad_norm": 0.16015625,
+      "learning_rate": 0.02,
+      "loss": 1.6623,
+      "step": 1772
+    },
+    {
+      "epoch": 0.16413817045651796,
+      "grad_norm": 0.158203125,
+      "learning_rate": 0.02,
+      "loss": 1.717,
+      "step": 1773
+    },
+    {
+      "epoch": 0.16423074697679801,
+      "grad_norm": 0.16015625,
+      "learning_rate": 0.02,
+      "loss": 1.7195,
+      "step": 1774
+    },
+    {
+      "epoch": 0.16432332349707807,
+      "grad_norm": 0.1591796875,
+      "learning_rate": 0.02,
+      "loss": 1.6993,
+      "step": 1775
+    },
+    {
+      "epoch": 0.1644159000173581,
+      "grad_norm": 0.1455078125,
+      "learning_rate": 0.02,
+      "loss": 1.6974,
+      "step": 1776
+    },
+    {
+      "epoch": 0.16450847653763814,
+      "grad_norm": 0.15234375,
+      "learning_rate": 0.02,
+      "loss": 1.7481,
+      "step": 1777
+    },
+    {
+      "epoch": 0.1646010530579182,
+      "grad_norm": 0.150390625,
+      "learning_rate": 0.02,
+      "loss": 1.7608,
+      "step": 1778
+    },
+    {
+      "epoch": 0.16469362957819822,
+      "grad_norm": 0.158203125,
+      "learning_rate": 0.02,
+      "loss": 1.7714,
+      "step": 1779
+    },
+    {
+      "epoch": 0.16478620609847827,
+      "grad_norm": 0.158203125,
+      "learning_rate": 0.02,
+      "loss": 1.7118,
+      "step": 1780
+    },
+    {
+      "epoch": 0.16487878261875832,
+      "grad_norm": 0.1474609375,
+      "learning_rate": 0.02,
+      "loss": 1.6788,
+      "step": 1781
+    },
+    {
+      "epoch": 0.16497135913903835,
+      "grad_norm": 0.142578125,
+      "learning_rate": 0.02,
+      "loss": 1.7028,
+      "step": 1782
+    },
+    {
+      "epoch": 0.1650639356593184,
+      "grad_norm": 0.162109375,
+      "learning_rate": 0.02,
+      "loss": 1.7656,
+      "step": 1783
+    },
+    {
+      "epoch": 0.16515651217959845,
+      "grad_norm": 0.1533203125,
+      "learning_rate": 0.02,
+      "loss": 1.6312,
+      "step": 1784
+    },
+    {
+      "epoch": 0.1652490886998785,
+      "grad_norm": 0.1591796875,
+      "learning_rate": 0.02,
+      "loss": 1.7758,
+      "step": 1785
+    },
+    {
+      "epoch": 0.16534166522015853,
+      "grad_norm": 0.15625,
+      "learning_rate": 0.02,
+      "loss": 1.6692,
+      "step": 1786
+    },
+    {
+      "epoch": 0.16543424174043858,
+      "grad_norm": 0.1689453125,
+      "learning_rate": 0.02,
+      "loss": 1.762,
+      "step": 1787
+    },
+    {
+      "epoch": 0.16552681826071863,
+      "grad_norm": 0.1640625,
+      "learning_rate": 0.02,
+      "loss": 1.6304,
+      "step": 1788
+    },
+    {
+      "epoch": 0.16561939478099866,
+      "grad_norm": 0.1689453125,
+      "learning_rate": 0.02,
+      "loss": 1.6829,
+      "step": 1789
+    },
+    {
+      "epoch": 0.1657119713012787,
+      "grad_norm": 0.162109375,
+      "learning_rate": 0.02,
+      "loss": 1.7535,
+      "step": 1790
+    },
+    {
+      "epoch": 0.16580454782155876,
+      "grad_norm": 0.15625,
+      "learning_rate": 0.02,
+      "loss": 1.6422,
+      "step": 1791
+    },
+    {
+      "epoch": 0.16589712434183881,
+      "grad_norm": 0.162109375,
+      "learning_rate": 0.02,
+      "loss": 1.6806,
+      "step": 1792
+    },
+    {
+      "epoch": 0.16598970086211884,
+      "grad_norm": 0.162109375,
+      "learning_rate": 0.02,
+      "loss": 1.7147,
+      "step": 1793
+    },
+    {
+      "epoch": 0.1660822773823989,
+      "grad_norm": 0.162109375,
+      "learning_rate": 0.02,
+      "loss": 1.6755,
+      "step": 1794
+    },
+    {
+      "epoch": 0.16617485390267894,
+      "grad_norm": 0.1591796875,
+      "learning_rate": 0.02,
+      "loss": 1.6934,
+      "step": 1795
+    },
+    {
+      "epoch": 0.16626743042295897,
+      "grad_norm": 0.169921875,
+      "learning_rate": 0.02,
+      "loss": 1.7774,
+      "step": 1796
+    },
+    {
+      "epoch": 0.16636000694323902,
+      "grad_norm": 0.158203125,
+      "learning_rate": 0.02,
+      "loss": 1.7597,
+      "step": 1797
+    },
+    {
+      "epoch": 0.16645258346351907,
+      "grad_norm": 0.1640625,
+      "learning_rate": 0.02,
+      "loss": 1.6785,
+      "step": 1798
+    },
+    {
+      "epoch": 0.1665451599837991,
+      "grad_norm": 0.1533203125,
+      "learning_rate": 0.02,
+      "loss": 1.7124,
+      "step": 1799
+    },
+    {
+      "epoch": 0.16663773650407915,
+      "grad_norm": 0.1572265625,
+      "learning_rate": 0.02,
+      "loss": 1.7191,
+      "step": 1800
+    },
+    {
+      "epoch": 0.1667303130243592,
+      "grad_norm": 0.1572265625,
+      "learning_rate": 0.02,
+      "loss": 1.7283,
+      "step": 1801
+    },
+    {
+      "epoch": 0.16682288954463925,
+      "grad_norm": 0.1611328125,
+      "learning_rate": 0.02,
+      "loss": 1.7171,
+      "step": 1802
+    },
+    {
+      "epoch": 0.16691546606491928,
+      "grad_norm": 0.154296875,
+      "learning_rate": 0.02,
+      "loss": 1.6692,
+      "step": 1803
+    },
+    {
+      "epoch": 0.16700804258519933,
+      "grad_norm": 0.1533203125,
+      "learning_rate": 0.02,
+      "loss": 1.7535,
+      "step": 1804
+    },
+    {
+      "epoch": 0.16710061910547938,
+      "grad_norm": 0.1533203125,
+      "learning_rate": 0.02,
+      "loss": 1.7359,
+      "step": 1805
+    },
+    {
+      "epoch": 0.1671931956257594,
+      "grad_norm": 0.1572265625,
+      "learning_rate": 0.02,
+      "loss": 1.672,
+      "step": 1806
+    },
+    {
+      "epoch": 0.16728577214603946,
+      "grad_norm": 0.1484375,
+      "learning_rate": 0.02,
+      "loss": 1.7069,
+      "step": 1807
+    },
+    {
+      "epoch": 0.1673783486663195,
+      "grad_norm": 0.15234375,
+      "learning_rate": 0.02,
+      "loss": 1.6823,
+      "step": 1808
+    },
+    {
+      "epoch": 0.16747092518659956,
+      "grad_norm": 0.15234375,
+      "learning_rate": 0.02,
+      "loss": 1.7021,
+      "step": 1809
+    },
+    {
+      "epoch": 0.1675635017068796,
+      "grad_norm": 0.1552734375,
+      "learning_rate": 0.02,
+      "loss": 1.7331,
+      "step": 1810
+    },
+    {
+      "epoch": 0.16765607822715964,
+      "grad_norm": 0.1611328125,
+      "learning_rate": 0.02,
+      "loss": 1.6734,
+      "step": 1811
+    },
+    {
+      "epoch": 0.1677486547474397,
+      "grad_norm": 0.1611328125,
+      "learning_rate": 0.02,
+      "loss": 1.6964,
+      "step": 1812
+    },
+    {
+      "epoch": 0.16784123126771971,
+      "grad_norm": 0.1708984375,
+      "learning_rate": 0.02,
+      "loss": 1.7706,
+      "step": 1813
+    },
+    {
+      "epoch": 0.16793380778799977,
+      "grad_norm": 0.1630859375,
+      "learning_rate": 0.02,
+      "loss": 1.7134,
+      "step": 1814
+    },
+    {
+      "epoch": 0.16802638430827982,
+      "grad_norm": 0.162109375,
+      "learning_rate": 0.02,
+      "loss": 1.6997,
+      "step": 1815
+    },
+    {
+      "epoch": 0.16811896082855984,
+      "grad_norm": 0.1728515625,
+      "learning_rate": 0.02,
+      "loss": 1.745,
+      "step": 1816
+    },
+    {
+      "epoch": 0.1682115373488399,
+      "grad_norm": 0.1708984375,
+      "learning_rate": 0.02,
+      "loss": 1.7253,
+      "step": 1817
+    },
+    {
+      "epoch": 0.16830411386911995,
+      "grad_norm": 0.162109375,
+      "learning_rate": 0.02,
+      "loss": 1.6337,
+      "step": 1818
+    },
+    {
+      "epoch": 0.1683966903894,
+      "grad_norm": 0.154296875,
+      "learning_rate": 0.02,
+      "loss": 1.663,
+      "step": 1819
+    },
+    {
+      "epoch": 0.16848926690968002,
+      "grad_norm": 0.1630859375,
+      "learning_rate": 0.02,
+      "loss": 1.6866,
+      "step": 1820
+    },
+    {
+      "epoch": 0.16858184342996008,
+      "grad_norm": 0.16015625,
+      "learning_rate": 0.02,
+      "loss": 1.7706,
+      "step": 1821
+    },
+    {
+      "epoch": 0.16867441995024013,
+      "grad_norm": 0.1552734375,
+      "learning_rate": 0.02,
+      "loss": 1.7335,
+      "step": 1822
+    },
+    {
+      "epoch": 0.16876699647052015,
+      "grad_norm": 0.1513671875,
+      "learning_rate": 0.02,
+      "loss": 1.6794,
+      "step": 1823
+    },
+    {
+      "epoch": 0.1688595729908002,
+      "grad_norm": 0.1455078125,
+      "learning_rate": 0.02,
+      "loss": 1.6891,
+      "step": 1824
+    },
+    {
+      "epoch": 0.16895214951108026,
+      "grad_norm": 0.1533203125,
+      "learning_rate": 0.02,
+      "loss": 1.6981,
+      "step": 1825
+    },
+    {
+      "epoch": 0.1690447260313603,
+      "grad_norm": 0.158203125,
+      "learning_rate": 0.02,
+      "loss": 1.752,
+      "step": 1826
+    },
+    {
+      "epoch": 0.16913730255164033,
+      "grad_norm": 0.166015625,
+      "learning_rate": 0.02,
+      "loss": 1.7738,
+      "step": 1827
+    },
+    {
+      "epoch": 0.16922987907192039,
+      "grad_norm": 0.16015625,
+      "learning_rate": 0.02,
+      "loss": 1.7045,
+      "step": 1828
+    },
+    {
+      "epoch": 0.16932245559220044,
+      "grad_norm": 0.158203125,
+      "learning_rate": 0.02,
+      "loss": 1.7015,
+      "step": 1829
+    },
+    {
+      "epoch": 0.16941503211248046,
+      "grad_norm": 0.15234375,
+      "learning_rate": 0.02,
+      "loss": 1.7152,
+      "step": 1830
+    },
+    {
+      "epoch": 0.16950760863276051,
+      "grad_norm": 0.158203125,
+      "learning_rate": 0.02,
+      "loss": 1.7155,
+      "step": 1831
+    },
+    {
+      "epoch": 0.16960018515304057,
+      "grad_norm": 0.150390625,
+      "learning_rate": 0.02,
+      "loss": 1.7375,
+      "step": 1832
+    },
+    {
+      "epoch": 0.1696927616733206,
+      "grad_norm": 0.1689453125,
+      "learning_rate": 0.02,
+      "loss": 1.742,
+      "step": 1833
+    },
+    {
+      "epoch": 0.16978533819360064,
+      "grad_norm": 0.158203125,
+      "learning_rate": 0.02,
+      "loss": 1.7514,
+      "step": 1834
+    },
+    {
+      "epoch": 0.1698779147138807,
+      "grad_norm": 0.15234375,
+      "learning_rate": 0.02,
+      "loss": 1.7095,
+      "step": 1835
+    },
+    {
+      "epoch": 0.16997049123416075,
+      "grad_norm": 0.150390625,
+      "learning_rate": 0.02,
+      "loss": 1.7184,
+      "step": 1836
+    },
+    {
+      "epoch": 0.17006306775444077,
+      "grad_norm": 0.158203125,
+      "learning_rate": 0.02,
+      "loss": 1.7822,
+      "step": 1837
+    },
+    {
+      "epoch": 0.17015564427472082,
+      "grad_norm": 0.15234375,
+      "learning_rate": 0.02,
+      "loss": 1.7563,
+      "step": 1838
+    },
+    {
+      "epoch": 0.17024822079500088,
+      "grad_norm": 0.1533203125,
+      "learning_rate": 0.02,
+      "loss": 1.787,
+      "step": 1839
+    },
+    {
+      "epoch": 0.1703407973152809,
+      "grad_norm": 0.1474609375,
+      "learning_rate": 0.02,
+      "loss": 1.7484,
+      "step": 1840
+    },
+    {
+      "epoch": 0.17043337383556095,
+      "grad_norm": 0.150390625,
+      "learning_rate": 0.02,
+      "loss": 1.7808,
+      "step": 1841
+    },
+    {
+      "epoch": 0.170525950355841,
+      "grad_norm": 0.154296875,
+      "learning_rate": 0.02,
+      "loss": 1.7258,
+      "step": 1842
+    },
+    {
+      "epoch": 0.17061852687612106,
+      "grad_norm": 0.1455078125,
+      "learning_rate": 0.02,
+      "loss": 1.6565,
+      "step": 1843
+    },
+    {
+      "epoch": 0.17071110339640108,
+      "grad_norm": 0.1513671875,
+      "learning_rate": 0.02,
+      "loss": 1.7133,
+      "step": 1844
+    },
+    {
+      "epoch": 0.17080367991668113,
+      "grad_norm": 0.1376953125,
+      "learning_rate": 0.02,
+      "loss": 1.6565,
+      "step": 1845
+    },
+    {
+      "epoch": 0.17089625643696119,
+      "grad_norm": 0.15234375,
+      "learning_rate": 0.02,
+      "loss": 1.7222,
+      "step": 1846
+    },
+    {
+      "epoch": 0.1709888329572412,
+      "grad_norm": 0.1484375,
+      "learning_rate": 0.02,
+      "loss": 1.6759,
+      "step": 1847
+    },
+    {
+      "epoch": 0.17108140947752126,
+      "grad_norm": 0.15234375,
+      "learning_rate": 0.02,
+      "loss": 1.702,
+      "step": 1848
+    },
+    {
+      "epoch": 0.17117398599780131,
+      "grad_norm": 0.1484375,
+      "learning_rate": 0.02,
+      "loss": 1.7478,
+      "step": 1849
+    },
+    {
+      "epoch": 0.17126656251808134,
+      "grad_norm": 0.1474609375,
+      "learning_rate": 0.02,
+      "loss": 1.6986,
+      "step": 1850
+    },
+    {
+      "epoch": 0.1713591390383614,
+      "grad_norm": 0.15234375,
+      "learning_rate": 0.02,
+      "loss": 1.7241,
+      "step": 1851
+    },
+    {
+      "epoch": 0.17145171555864144,
+      "grad_norm": 0.158203125,
+      "learning_rate": 0.02,
+      "loss": 1.7057,
+      "step": 1852
+    },
+    {
+      "epoch": 0.1715442920789215,
+      "grad_norm": 0.1474609375,
+      "learning_rate": 0.02,
+      "loss": 1.7188,
+      "step": 1853
+    },
+    {
+      "epoch": 0.17163686859920152,
+      "grad_norm": 0.15625,
+      "learning_rate": 0.02,
+      "loss": 1.7214,
+      "step": 1854
+    },
+    {
+      "epoch": 0.17172944511948157,
+      "grad_norm": 0.1533203125,
+      "learning_rate": 0.02,
+      "loss": 1.7221,
+      "step": 1855
+    },
+    {
+      "epoch": 0.17182202163976162,
+      "grad_norm": 0.15234375,
+      "learning_rate": 0.02,
+      "loss": 1.6769,
+      "step": 1856
+    },
+    {
+      "epoch": 0.17191459816004165,
+      "grad_norm": 0.1630859375,
+      "learning_rate": 0.02,
+      "loss": 1.7411,
+      "step": 1857
+    },
+    {
+      "epoch": 0.1720071746803217,
+      "grad_norm": 0.171875,
+      "learning_rate": 0.02,
+      "loss": 1.7384,
+      "step": 1858
+    },
+    {
+      "epoch": 0.17209975120060175,
+      "grad_norm": 0.16015625,
+      "learning_rate": 0.02,
+      "loss": 1.7282,
+      "step": 1859
+    },
+    {
+      "epoch": 0.1721923277208818,
+      "grad_norm": 0.173828125,
+      "learning_rate": 0.02,
+      "loss": 1.7134,
+      "step": 1860
+    },
+    {
+      "epoch": 0.17228490424116183,
+      "grad_norm": 0.166015625,
+      "learning_rate": 0.02,
+      "loss": 1.6696,
+      "step": 1861
+    },
+    {
+      "epoch": 0.17237748076144188,
+      "grad_norm": 0.177734375,
+      "learning_rate": 0.02,
+      "loss": 1.709,
+      "step": 1862
+    },
+    {
+      "epoch": 0.17247005728172193,
+      "grad_norm": 0.1669921875,
+      "learning_rate": 0.02,
+      "loss": 1.7327,
+      "step": 1863
+    },
+    {
+      "epoch": 0.17256263380200196,
+      "grad_norm": 0.1689453125,
+      "learning_rate": 0.02,
+      "loss": 1.6961,
+      "step": 1864
+    },
+    {
+      "epoch": 0.172655210322282,
+      "grad_norm": 0.1630859375,
+      "learning_rate": 0.02,
+      "loss": 1.6818,
+      "step": 1865
+    },
+    {
+      "epoch": 0.17274778684256206,
+      "grad_norm": 0.1796875,
+      "learning_rate": 0.02,
+      "loss": 1.7705,
+      "step": 1866
+    },
+    {
+      "epoch": 0.1728403633628421,
+      "grad_norm": 0.173828125,
+      "learning_rate": 0.02,
+      "loss": 1.7219,
+      "step": 1867
+    },
+    {
+      "epoch": 0.17293293988312214,
+      "grad_norm": 0.154296875,
+      "learning_rate": 0.02,
+      "loss": 1.7791,
+      "step": 1868
+    },
+    {
+      "epoch": 0.1730255164034022,
+      "grad_norm": 0.158203125,
+      "learning_rate": 0.02,
+      "loss": 1.7107,
+      "step": 1869
+    },
+    {
+      "epoch": 0.17311809292368224,
+      "grad_norm": 0.1640625,
+      "learning_rate": 0.02,
+      "loss": 1.7457,
+      "step": 1870
+    },
+    {
+      "epoch": 0.17321066944396227,
+      "grad_norm": 0.1630859375,
+      "learning_rate": 0.02,
+      "loss": 1.7197,
+      "step": 1871
+    },
+    {
+      "epoch": 0.17330324596424232,
+      "grad_norm": 0.16015625,
+      "learning_rate": 0.02,
+      "loss": 1.7191,
+      "step": 1872
+    },
+    {
+      "epoch": 0.17339582248452237,
+      "grad_norm": 0.1640625,
+      "learning_rate": 0.02,
+      "loss": 1.7699,
+      "step": 1873
+    },
+    {
+      "epoch": 0.1734883990048024,
+      "grad_norm": 0.1689453125,
+      "learning_rate": 0.02,
+      "loss": 1.7782,
+      "step": 1874
+    },
+    {
+      "epoch": 0.17358097552508245,
+      "grad_norm": 0.1640625,
+      "learning_rate": 0.02,
+      "loss": 1.7174,
+      "step": 1875
+    },
+    {
+      "epoch": 0.1736735520453625,
+      "grad_norm": 0.1806640625,
+      "learning_rate": 0.02,
+      "loss": 1.7879,
+      "step": 1876
+    },
+    {
+      "epoch": 0.17376612856564255,
+      "grad_norm": 0.154296875,
+      "learning_rate": 0.02,
+      "loss": 1.7067,
+      "step": 1877
+    },
+    {
+      "epoch": 0.17385870508592258,
+      "grad_norm": 0.14453125,
+      "learning_rate": 0.02,
+      "loss": 1.7108,
+      "step": 1878
+    },
+    {
+      "epoch": 0.17395128160620263,
+      "grad_norm": 0.1494140625,
+      "learning_rate": 0.02,
+      "loss": 1.6946,
+      "step": 1879
+    },
+    {
+      "epoch": 0.17404385812648268,
+      "grad_norm": 0.154296875,
+      "learning_rate": 0.02,
+      "loss": 1.7228,
+      "step": 1880
+    },
+    {
+      "epoch": 0.1741364346467627,
+      "grad_norm": 0.1533203125,
+      "learning_rate": 0.02,
+      "loss": 1.7238,
+      "step": 1881
+    },
+    {
+      "epoch": 0.17422901116704276,
+      "grad_norm": 0.1484375,
+      "learning_rate": 0.02,
+      "loss": 1.737,
+      "step": 1882
+    },
+    {
+      "epoch": 0.1743215876873228,
+      "grad_norm": 0.1357421875,
+      "learning_rate": 0.02,
+      "loss": 1.6407,
+      "step": 1883
+    },
+    {
+      "epoch": 0.17441416420760283,
+      "grad_norm": 0.1513671875,
+      "learning_rate": 0.02,
+      "loss": 1.7359,
+      "step": 1884
+    },
+    {
+      "epoch": 0.1745067407278829,
+      "grad_norm": 0.1630859375,
+      "learning_rate": 0.02,
+      "loss": 1.7058,
+      "step": 1885
+    },
+    {
+      "epoch": 0.17459931724816294,
+      "grad_norm": 0.1572265625,
+      "learning_rate": 0.02,
+      "loss": 1.8002,
+      "step": 1886
+    },
+    {
+      "epoch": 0.174691893768443,
+      "grad_norm": 0.1552734375,
+      "learning_rate": 0.02,
+      "loss": 1.7126,
+      "step": 1887
+    },
+    {
+      "epoch": 0.17478447028872302,
+      "grad_norm": 0.17578125,
+      "learning_rate": 0.02,
+      "loss": 1.8146,
+      "step": 1888
+    },
+    {
+      "epoch": 0.17487704680900307,
+      "grad_norm": 0.169921875,
+      "learning_rate": 0.02,
+      "loss": 1.6902,
+      "step": 1889
+    },
+    {
+      "epoch": 0.17496962332928312,
+      "grad_norm": 0.173828125,
+      "learning_rate": 0.02,
+      "loss": 1.6664,
+      "step": 1890
+    },
+    {
+      "epoch": 0.17506219984956314,
+      "grad_norm": 0.162109375,
+      "learning_rate": 0.02,
+      "loss": 1.7447,
+      "step": 1891
+    },
+    {
+      "epoch": 0.1751547763698432,
+      "grad_norm": 0.1796875,
+      "learning_rate": 0.02,
+      "loss": 1.6768,
+      "step": 1892
+    },
+    {
+      "epoch": 0.17524735289012325,
+      "grad_norm": 0.1611328125,
+      "learning_rate": 0.02,
+      "loss": 1.7119,
+      "step": 1893
+    },
+    {
+      "epoch": 0.1753399294104033,
+      "grad_norm": 0.1640625,
+      "learning_rate": 0.02,
+      "loss": 1.7018,
+      "step": 1894
+    },
+    {
+      "epoch": 0.17543250593068332,
+      "grad_norm": 0.1552734375,
+      "learning_rate": 0.02,
+      "loss": 1.7114,
+      "step": 1895
+    },
+    {
+      "epoch": 0.17552508245096338,
+      "grad_norm": 0.1591796875,
+      "learning_rate": 0.02,
+      "loss": 1.7443,
+      "step": 1896
+    },
+    {
+      "epoch": 0.17561765897124343,
+      "grad_norm": 0.1572265625,
+      "learning_rate": 0.02,
+      "loss": 1.7091,
+      "step": 1897
+    },
+    {
+      "epoch": 0.17571023549152345,
+      "grad_norm": 0.1572265625,
+      "learning_rate": 0.02,
+      "loss": 1.7352,
+      "step": 1898
+    },
+    {
+      "epoch": 0.1758028120118035,
+      "grad_norm": 0.162109375,
+      "learning_rate": 0.02,
+      "loss": 1.7409,
+      "step": 1899
+    },
+    {
+      "epoch": 0.17589538853208356,
+      "grad_norm": 0.1513671875,
+      "learning_rate": 0.02,
+      "loss": 1.6562,
+      "step": 1900
+    },
+    {
+      "epoch": 0.17598796505236358,
+      "grad_norm": 0.15234375,
+      "learning_rate": 0.02,
+      "loss": 1.7162,
+      "step": 1901
+    },
+    {
+      "epoch": 0.17608054157264363,
+      "grad_norm": 0.1689453125,
+      "learning_rate": 0.02,
+      "loss": 1.731,
+      "step": 1902
+    },
+    {
+      "epoch": 0.1761731180929237,
+      "grad_norm": 0.1748046875,
+      "learning_rate": 0.02,
+      "loss": 1.7849,
+      "step": 1903
+    },
+    {
+      "epoch": 0.17626569461320374,
+      "grad_norm": 0.1708984375,
+      "learning_rate": 0.02,
+      "loss": 1.6457,
+      "step": 1904
+    },
+    {
+      "epoch": 0.17635827113348376,
+      "grad_norm": 0.17578125,
+      "learning_rate": 0.02,
+      "loss": 1.7128,
+      "step": 1905
+    },
+    {
+      "epoch": 0.17645084765376381,
+      "grad_norm": 0.1650390625,
+      "learning_rate": 0.02,
+      "loss": 1.7143,
+      "step": 1906
+    },
+    {
+      "epoch": 0.17654342417404387,
+      "grad_norm": 0.1640625,
+      "learning_rate": 0.02,
+      "loss": 1.7145,
+      "step": 1907
+    },
+    {
+      "epoch": 0.1766360006943239,
+      "grad_norm": 0.1669921875,
+      "learning_rate": 0.02,
+      "loss": 1.7816,
+      "step": 1908
+    },
+    {
+      "epoch": 0.17672857721460394,
+      "grad_norm": 0.154296875,
+      "learning_rate": 0.02,
+      "loss": 1.6194,
+      "step": 1909
+    },
+    {
+      "epoch": 0.176821153734884,
+      "grad_norm": 0.1708984375,
+      "learning_rate": 0.02,
+      "loss": 1.7553,
+      "step": 1910
+    },
+    {
+      "epoch": 0.17691373025516405,
+      "grad_norm": 0.1630859375,
+      "learning_rate": 0.02,
+      "loss": 1.676,
+      "step": 1911
+    },
+    {
+      "epoch": 0.17700630677544407,
+      "grad_norm": 0.166015625,
+      "learning_rate": 0.02,
+      "loss": 1.7514,
+      "step": 1912
+    },
+    {
+      "epoch": 0.17709888329572412,
+      "grad_norm": 0.16796875,
+      "learning_rate": 0.02,
+      "loss": 1.756,
+      "step": 1913
+    },
+    {
+      "epoch": 0.17719145981600418,
+      "grad_norm": 0.1552734375,
+      "learning_rate": 0.02,
+      "loss": 1.682,
+      "step": 1914
+    },
+    {
+      "epoch": 0.1772840363362842,
+      "grad_norm": 0.150390625,
+      "learning_rate": 0.02,
+      "loss": 1.6527,
+      "step": 1915
+    },
+    {
+      "epoch": 0.17737661285656425,
+      "grad_norm": 0.1572265625,
+      "learning_rate": 0.02,
+      "loss": 1.7208,
+      "step": 1916
+    },
+    {
+      "epoch": 0.1774691893768443,
+      "grad_norm": 0.15625,
+      "learning_rate": 0.02,
+      "loss": 1.7489,
+      "step": 1917
+    },
+    {
+      "epoch": 0.17756176589712433,
+      "grad_norm": 0.1513671875,
+      "learning_rate": 0.02,
+      "loss": 1.7205,
+      "step": 1918
+    },
+    {
+      "epoch": 0.17765434241740438,
+      "grad_norm": 0.1474609375,
+      "learning_rate": 0.02,
+      "loss": 1.6896,
+      "step": 1919
+    },
+    {
+      "epoch": 0.17774691893768443,
+      "grad_norm": 0.1533203125,
+      "learning_rate": 0.02,
+      "loss": 1.7096,
+      "step": 1920
+    },
+    {
+      "epoch": 0.17783949545796449,
+      "grad_norm": 0.140625,
+      "learning_rate": 0.02,
+      "loss": 1.6364,
+      "step": 1921
+    },
+    {
+      "epoch": 0.1779320719782445,
+      "grad_norm": 0.14453125,
+      "learning_rate": 0.02,
+      "loss": 1.6905,
+      "step": 1922
+    },
+    {
+      "epoch": 0.17802464849852456,
+      "grad_norm": 0.1591796875,
+      "learning_rate": 0.02,
+      "loss": 1.6875,
+      "step": 1923
+    },
+    {
+      "epoch": 0.17811722501880461,
+      "grad_norm": 0.1552734375,
+      "learning_rate": 0.02,
+      "loss": 1.7383,
+      "step": 1924
+    },
+    {
+      "epoch": 0.17820980153908464,
+      "grad_norm": 0.1572265625,
+      "learning_rate": 0.02,
+      "loss": 1.7046,
+      "step": 1925
+    },
+    {
+      "epoch": 0.1783023780593647,
+      "grad_norm": 0.1572265625,
+      "learning_rate": 0.02,
+      "loss": 1.7725,
+      "step": 1926
+    },
+    {
+      "epoch": 0.17839495457964474,
+      "grad_norm": 0.1552734375,
+      "learning_rate": 0.02,
+      "loss": 1.7222,
+      "step": 1927
+    },
+    {
+      "epoch": 0.17848753109992477,
+      "grad_norm": 0.1640625,
+      "learning_rate": 0.02,
+      "loss": 1.7845,
+      "step": 1928
+    },
+    {
+      "epoch": 0.17858010762020482,
+      "grad_norm": 0.150390625,
+      "learning_rate": 0.02,
+      "loss": 1.7074,
+      "step": 1929
+    },
+    {
+      "epoch": 0.17867268414048487,
+      "grad_norm": 0.1513671875,
+      "learning_rate": 0.02,
+      "loss": 1.6691,
+      "step": 1930
+    },
+    {
+      "epoch": 0.17876526066076492,
+      "grad_norm": 0.1552734375,
+      "learning_rate": 0.02,
+      "loss": 1.7014,
+      "step": 1931
+    },
+    {
+      "epoch": 0.17885783718104495,
+      "grad_norm": 0.1494140625,
+      "learning_rate": 0.02,
+      "loss": 1.6983,
+      "step": 1932
+    },
+    {
+      "epoch": 0.178950413701325,
+      "grad_norm": 0.1591796875,
+      "learning_rate": 0.02,
+      "loss": 1.6964,
+      "step": 1933
+    },
+    {
+      "epoch": 0.17904299022160505,
+      "grad_norm": 0.1669921875,
+      "learning_rate": 0.02,
+      "loss": 1.7239,
+      "step": 1934
+    },
+    {
+      "epoch": 0.17913556674188508,
+      "grad_norm": 0.171875,
+      "learning_rate": 0.02,
+      "loss": 1.7461,
+      "step": 1935
+    },
+    {
+      "epoch": 0.17922814326216513,
+      "grad_norm": 0.1650390625,
+      "learning_rate": 0.02,
+      "loss": 1.7426,
+      "step": 1936
+    },
+    {
+      "epoch": 0.17932071978244518,
+      "grad_norm": 0.1630859375,
+      "learning_rate": 0.02,
+      "loss": 1.6839,
+      "step": 1937
+    },
+    {
+      "epoch": 0.17941329630272523,
+      "grad_norm": 0.15625,
+      "learning_rate": 0.02,
+      "loss": 1.6973,
+      "step": 1938
+    },
+    {
+      "epoch": 0.17950587282300526,
+      "grad_norm": 0.16796875,
+      "learning_rate": 0.02,
+      "loss": 1.7558,
+      "step": 1939
+    },
+    {
+      "epoch": 0.1795984493432853,
+      "grad_norm": 0.150390625,
+      "learning_rate": 0.02,
+      "loss": 1.6441,
+      "step": 1940
+    },
+    {
+      "epoch": 0.17969102586356536,
+      "grad_norm": 0.1630859375,
+      "learning_rate": 0.02,
+      "loss": 1.6938,
+      "step": 1941
+    },
+    {
+      "epoch": 0.1797836023838454,
+      "grad_norm": 0.1552734375,
+      "learning_rate": 0.02,
+      "loss": 1.7175,
+      "step": 1942
+    },
+    {
+      "epoch": 0.17987617890412544,
+      "grad_norm": 0.16796875,
+      "learning_rate": 0.02,
+      "loss": 1.7586,
+      "step": 1943
+    },
+    {
+      "epoch": 0.1799687554244055,
+      "grad_norm": 0.154296875,
+      "learning_rate": 0.02,
+      "loss": 1.7102,
+      "step": 1944
+    },
+    {
+      "epoch": 0.18006133194468552,
+      "grad_norm": 0.16015625,
+      "learning_rate": 0.02,
+      "loss": 1.7301,
+      "step": 1945
+    },
+    {
+      "epoch": 0.18015390846496557,
+      "grad_norm": 0.162109375,
+      "learning_rate": 0.02,
+      "loss": 1.698,
+      "step": 1946
+    },
+    {
+      "epoch": 0.18024648498524562,
+      "grad_norm": 0.1650390625,
+      "learning_rate": 0.02,
+      "loss": 1.7026,
+      "step": 1947
+    },
+    {
+      "epoch": 0.18033906150552567,
+      "grad_norm": 0.166015625,
+      "learning_rate": 0.02,
+      "loss": 1.72,
+      "step": 1948
+    },
+    {
+      "epoch": 0.1804316380258057,
+      "grad_norm": 0.169921875,
+      "learning_rate": 0.02,
+      "loss": 1.7046,
+      "step": 1949
+    },
+    {
+      "epoch": 0.18052421454608575,
+      "grad_norm": 0.1630859375,
+      "learning_rate": 0.02,
+      "loss": 1.6986,
+      "step": 1950
+    },
+    {
+      "epoch": 0.1806167910663658,
+      "grad_norm": 0.1640625,
+      "learning_rate": 0.02,
+      "loss": 1.6592,
+      "step": 1951
+    },
+    {
+      "epoch": 0.18070936758664582,
+      "grad_norm": 0.1591796875,
+      "learning_rate": 0.02,
+      "loss": 1.7238,
+      "step": 1952
+    },
+    {
+      "epoch": 0.18080194410692588,
+      "grad_norm": 0.1533203125,
+      "learning_rate": 0.02,
+      "loss": 1.685,
+      "step": 1953
+    },
+    {
+      "epoch": 0.18089452062720593,
+      "grad_norm": 0.146484375,
+      "learning_rate": 0.02,
+      "loss": 1.6963,
+      "step": 1954
+    },
+    {
+      "epoch": 0.18098709714748598,
+      "grad_norm": 0.1474609375,
+      "learning_rate": 0.02,
+      "loss": 1.6632,
+      "step": 1955
+    },
+    {
+      "epoch": 0.181079673667766,
+      "grad_norm": 0.1494140625,
+      "learning_rate": 0.02,
+      "loss": 1.6768,
+      "step": 1956
+    },
+    {
+      "epoch": 0.18117225018804606,
+      "grad_norm": 0.1484375,
+      "learning_rate": 0.02,
+      "loss": 1.6919,
+      "step": 1957
+    },
+    {
+      "epoch": 0.1812648267083261,
+      "grad_norm": 0.1640625,
+      "learning_rate": 0.02,
+      "loss": 1.7618,
+      "step": 1958
+    },
+    {
+      "epoch": 0.18135740322860613,
+      "grad_norm": 0.1484375,
+      "learning_rate": 0.02,
+      "loss": 1.6658,
+      "step": 1959
+    },
+    {
+      "epoch": 0.1814499797488862,
+      "grad_norm": 0.1513671875,
+      "learning_rate": 0.02,
+      "loss": 1.7078,
+      "step": 1960
+    },
+    {
+      "epoch": 0.18154255626916624,
+      "grad_norm": 0.1630859375,
+      "learning_rate": 0.02,
+      "loss": 1.6669,
+      "step": 1961
+    },
+    {
+      "epoch": 0.18163513278944626,
+      "grad_norm": 0.166015625,
+      "learning_rate": 0.02,
+      "loss": 1.6942,
+      "step": 1962
+    },
+    {
+      "epoch": 0.18172770930972632,
+      "grad_norm": 0.1533203125,
+      "learning_rate": 0.02,
+      "loss": 1.6821,
+      "step": 1963
+    },
+    {
+      "epoch": 0.18182028583000637,
+      "grad_norm": 0.150390625,
+      "learning_rate": 0.02,
+      "loss": 1.6622,
+      "step": 1964
+    },
+    {
+      "epoch": 0.18191286235028642,
+      "grad_norm": 0.15625,
+      "learning_rate": 0.02,
+      "loss": 1.6843,
+      "step": 1965
+    },
+    {
+      "epoch": 0.18200543887056644,
+      "grad_norm": 0.15625,
+      "learning_rate": 0.02,
+      "loss": 1.6453,
+      "step": 1966
+    },
+    {
+      "epoch": 0.1820980153908465,
+      "grad_norm": 0.1533203125,
+      "learning_rate": 0.02,
+      "loss": 1.6997,
+      "step": 1967
+    },
+    {
+      "epoch": 0.18219059191112655,
+      "grad_norm": 0.15234375,
+      "learning_rate": 0.02,
+      "loss": 1.6407,
+      "step": 1968
+    },
+    {
+      "epoch": 0.18228316843140657,
+      "grad_norm": 0.166015625,
+      "learning_rate": 0.02,
+      "loss": 1.7293,
+      "step": 1969
+    },
+    {
+      "epoch": 0.18237574495168662,
+      "grad_norm": 0.1640625,
+      "learning_rate": 0.02,
+      "loss": 1.7052,
+      "step": 1970
+    },
+    {
+      "epoch": 0.18246832147196668,
+      "grad_norm": 0.1708984375,
+      "learning_rate": 0.02,
+      "loss": 1.7328,
+      "step": 1971
+    },
+    {
+      "epoch": 0.18256089799224673,
+      "grad_norm": 0.16796875,
+      "learning_rate": 0.02,
+      "loss": 1.6089,
+      "step": 1972
+    },
+    {
+      "epoch": 0.18265347451252675,
+      "grad_norm": 0.1630859375,
+      "learning_rate": 0.02,
+      "loss": 1.6946,
+      "step": 1973
+    },
+    {
+      "epoch": 0.1827460510328068,
+      "grad_norm": 0.1640625,
+      "learning_rate": 0.02,
+      "loss": 1.6757,
+      "step": 1974
+    },
+    {
+      "epoch": 0.18283862755308686,
+      "grad_norm": 0.1572265625,
+      "learning_rate": 0.02,
+      "loss": 1.709,
+      "step": 1975
+    },
+    {
+      "epoch": 0.18293120407336688,
+      "grad_norm": 0.1591796875,
+      "learning_rate": 0.02,
+      "loss": 1.7105,
+      "step": 1976
+    },
+    {
+      "epoch": 0.18302378059364693,
+      "grad_norm": 0.1611328125,
+      "learning_rate": 0.02,
+      "loss": 1.6856,
+      "step": 1977
+    },
+    {
+      "epoch": 0.183116357113927,
+      "grad_norm": 0.15625,
+      "learning_rate": 0.02,
+      "loss": 1.6847,
+      "step": 1978
+    },
+    {
+      "epoch": 0.183208933634207,
+      "grad_norm": 0.15234375,
+      "learning_rate": 0.02,
+      "loss": 1.6579,
+      "step": 1979
+    },
+    {
+      "epoch": 0.18330151015448706,
+      "grad_norm": 0.1640625,
+      "learning_rate": 0.02,
+      "loss": 1.7462,
+      "step": 1980
+    },
+    {
+      "epoch": 0.18339408667476712,
+      "grad_norm": 0.1591796875,
+      "learning_rate": 0.02,
+      "loss": 1.7133,
+      "step": 1981
+    },
+    {
+      "epoch": 0.18348666319504717,
+      "grad_norm": 0.142578125,
+      "learning_rate": 0.02,
+      "loss": 1.6938,
+      "step": 1982
+    },
+    {
+      "epoch": 0.1835792397153272,
+      "grad_norm": 0.146484375,
+      "learning_rate": 0.02,
+      "loss": 1.7614,
+      "step": 1983
+    },
+    {
+      "epoch": 0.18367181623560724,
+      "grad_norm": 0.15234375,
+      "learning_rate": 0.02,
+      "loss": 1.6767,
+      "step": 1984
+    },
+    {
+      "epoch": 0.1837643927558873,
+      "grad_norm": 0.1552734375,
+      "learning_rate": 0.02,
+      "loss": 1.6718,
+      "step": 1985
+    },
+    {
+      "epoch": 0.18385696927616732,
+      "grad_norm": 0.1494140625,
+      "learning_rate": 0.02,
+      "loss": 1.685,
+      "step": 1986
+    },
+    {
+      "epoch": 0.18394954579644737,
+      "grad_norm": 0.1533203125,
+      "learning_rate": 0.02,
+      "loss": 1.7522,
+      "step": 1987
+    },
+    {
+      "epoch": 0.18404212231672742,
+      "grad_norm": 0.150390625,
+      "learning_rate": 0.02,
+      "loss": 1.6803,
+      "step": 1988
+    },
+    {
+      "epoch": 0.18413469883700748,
+      "grad_norm": 0.1416015625,
+      "learning_rate": 0.02,
+      "loss": 1.6179,
+      "step": 1989
+    },
+    {
+      "epoch": 0.1842272753572875,
+      "grad_norm": 0.1416015625,
+      "learning_rate": 0.02,
+      "loss": 1.6848,
+      "step": 1990
+    },
+    {
+      "epoch": 0.18431985187756755,
+      "grad_norm": 0.1455078125,
+      "learning_rate": 0.02,
+      "loss": 1.7282,
+      "step": 1991
+    },
+    {
+      "epoch": 0.1844124283978476,
+      "grad_norm": 0.1416015625,
+      "learning_rate": 0.02,
+      "loss": 1.6806,
+      "step": 1992
+    },
+    {
+      "epoch": 0.18450500491812763,
+      "grad_norm": 0.15234375,
+      "learning_rate": 0.02,
+      "loss": 1.7804,
+      "step": 1993
+    },
+    {
+      "epoch": 0.18459758143840768,
+      "grad_norm": 0.158203125,
+      "learning_rate": 0.02,
+      "loss": 1.7467,
+      "step": 1994
+    },
+    {
+      "epoch": 0.18469015795868773,
+      "grad_norm": 0.154296875,
+      "learning_rate": 0.02,
+      "loss": 1.6922,
+      "step": 1995
+    },
+    {
+      "epoch": 0.18478273447896776,
+      "grad_norm": 0.1533203125,
+      "learning_rate": 0.02,
+      "loss": 1.6524,
+      "step": 1996
+    },
+    {
+      "epoch": 0.1848753109992478,
+      "grad_norm": 0.154296875,
+      "learning_rate": 0.02,
+      "loss": 1.6851,
+      "step": 1997
+    },
+    {
+      "epoch": 0.18496788751952786,
+      "grad_norm": 0.1552734375,
+      "learning_rate": 0.02,
+      "loss": 1.713,
+      "step": 1998
+    },
+    {
+      "epoch": 0.18506046403980791,
+      "grad_norm": 0.1484375,
+      "learning_rate": 0.02,
+      "loss": 1.6675,
+      "step": 1999
+    },
+    {
+      "epoch": 0.18515304056008794,
+      "grad_norm": 0.1689453125,
+      "learning_rate": 0.02,
+      "loss": 1.7356,
+      "step": 2000
+    },
+    {
+      "epoch": 0.185245617080368,
+      "grad_norm": 0.1650390625,
+      "learning_rate": 0.02,
+      "loss": 1.7156,
+      "step": 2001
+    },
+    {
+      "epoch": 0.18533819360064804,
+      "grad_norm": 0.1552734375,
+      "learning_rate": 0.02,
+      "loss": 1.6602,
+      "step": 2002
+    },
+    {
+      "epoch": 0.18543077012092807,
+      "grad_norm": 0.1611328125,
+      "learning_rate": 0.02,
+      "loss": 1.7262,
+      "step": 2003
+    },
+    {
+      "epoch": 0.18552334664120812,
+      "grad_norm": 0.1533203125,
+      "learning_rate": 0.02,
+      "loss": 1.6697,
+      "step": 2004
+    },
+    {
+      "epoch": 0.18561592316148817,
+      "grad_norm": 0.146484375,
+      "learning_rate": 0.02,
+      "loss": 1.7074,
+      "step": 2005
+    },
+    {
+      "epoch": 0.18570849968176822,
+      "grad_norm": 0.1572265625,
+      "learning_rate": 0.02,
+      "loss": 1.7156,
+      "step": 2006
+    },
+    {
+      "epoch": 0.18580107620204825,
+      "grad_norm": 0.16015625,
+      "learning_rate": 0.02,
+      "loss": 1.7394,
+      "step": 2007
+    },
+    {
+      "epoch": 0.1858936527223283,
+      "grad_norm": 0.16015625,
+      "learning_rate": 0.02,
+      "loss": 1.71,
+      "step": 2008
+    },
+    {
+      "epoch": 0.18598622924260835,
+      "grad_norm": 0.15234375,
+      "learning_rate": 0.02,
+      "loss": 1.7212,
+      "step": 2009
+    },
+    {
+      "epoch": 0.18607880576288838,
+      "grad_norm": 0.166015625,
+      "learning_rate": 0.02,
+      "loss": 1.6844,
+      "step": 2010
+    },
+    {
+      "epoch": 0.18617138228316843,
+      "grad_norm": 0.1513671875,
+      "learning_rate": 0.02,
+      "loss": 1.6101,
+      "step": 2011
+    },
+    {
+      "epoch": 0.18626395880344848,
+      "grad_norm": 0.1552734375,
+      "learning_rate": 0.02,
+      "loss": 1.7118,
+      "step": 2012
+    },
+    {
+      "epoch": 0.1863565353237285,
+      "grad_norm": 0.171875,
+      "learning_rate": 0.02,
+      "loss": 1.7459,
+      "step": 2013
+    },
+    {
+      "epoch": 0.18644911184400856,
+      "grad_norm": 0.166015625,
+      "learning_rate": 0.02,
+      "loss": 1.7813,
+      "step": 2014
+    },
+    {
+      "epoch": 0.1865416883642886,
+      "grad_norm": 0.166015625,
+      "learning_rate": 0.02,
+      "loss": 1.6308,
+      "step": 2015
+    },
+    {
+      "epoch": 0.18663426488456866,
+      "grad_norm": 0.1611328125,
+      "learning_rate": 0.02,
+      "loss": 1.7353,
+      "step": 2016
+    },
+    {
+      "epoch": 0.1867268414048487,
+      "grad_norm": 0.158203125,
+      "learning_rate": 0.02,
+      "loss": 1.7122,
+      "step": 2017
+    },
+    {
+      "epoch": 0.18681941792512874,
+      "grad_norm": 0.1640625,
+      "learning_rate": 0.02,
+      "loss": 1.6964,
+      "step": 2018
+    },
+    {
+      "epoch": 0.1869119944454088,
+      "grad_norm": 0.154296875,
+      "learning_rate": 0.02,
+      "loss": 1.6373,
+      "step": 2019
+    },
+    {
+      "epoch": 0.18700457096568882,
+      "grad_norm": 0.1552734375,
+      "learning_rate": 0.02,
+      "loss": 1.7122,
+      "step": 2020
+    },
+    {
+      "epoch": 0.18709714748596887,
+      "grad_norm": 0.154296875,
+      "learning_rate": 0.02,
+      "loss": 1.644,
+      "step": 2021
+    },
+    {
+      "epoch": 0.18718972400624892,
+      "grad_norm": 0.1533203125,
+      "learning_rate": 0.02,
+      "loss": 1.7326,
+      "step": 2022
+    },
+    {
+      "epoch": 0.18728230052652897,
+      "grad_norm": 0.15625,
+      "learning_rate": 0.02,
+      "loss": 1.7191,
+      "step": 2023
+    },
+    {
+      "epoch": 0.187374877046809,
+      "grad_norm": 0.1650390625,
+      "learning_rate": 0.02,
+      "loss": 1.7629,
+      "step": 2024
+    },
+    {
+      "epoch": 0.18746745356708905,
+      "grad_norm": 0.1640625,
+      "learning_rate": 0.02,
+      "loss": 1.7203,
+      "step": 2025
+    },
+    {
+      "epoch": 0.1875600300873691,
+      "grad_norm": 0.1669921875,
+      "learning_rate": 0.02,
+      "loss": 1.7729,
+      "step": 2026
+    },
+    {
+      "epoch": 0.18765260660764913,
+      "grad_norm": 0.1533203125,
+      "learning_rate": 0.02,
+      "loss": 1.6184,
+      "step": 2027
+    },
+    {
+      "epoch": 0.18774518312792918,
+      "grad_norm": 0.154296875,
+      "learning_rate": 0.02,
+      "loss": 1.6972,
+      "step": 2028
+    },
+    {
+      "epoch": 0.18783775964820923,
+      "grad_norm": 0.1416015625,
+      "learning_rate": 0.02,
+      "loss": 1.636,
+      "step": 2029
+    },
+    {
+      "epoch": 0.18793033616848925,
+      "grad_norm": 0.154296875,
+      "learning_rate": 0.02,
+      "loss": 1.7116,
+      "step": 2030
+    },
+    {
+      "epoch": 0.1880229126887693,
+      "grad_norm": 0.166015625,
+      "learning_rate": 0.02,
+      "loss": 1.6986,
+      "step": 2031
+    },
+    {
+      "epoch": 0.18811548920904936,
+      "grad_norm": 0.16796875,
+      "learning_rate": 0.02,
+      "loss": 1.6926,
+      "step": 2032
+    },
+    {
+      "epoch": 0.1882080657293294,
+      "grad_norm": 0.146484375,
+      "learning_rate": 0.02,
+      "loss": 1.632,
+      "step": 2033
+    },
+    {
+      "epoch": 0.18830064224960943,
+      "grad_norm": 0.1533203125,
+      "learning_rate": 0.02,
+      "loss": 1.7037,
+      "step": 2034
+    },
+    {
+      "epoch": 0.1883932187698895,
+      "grad_norm": 0.1494140625,
+      "learning_rate": 0.02,
+      "loss": 1.7188,
+      "step": 2035
+    },
+    {
+      "epoch": 0.18848579529016954,
+      "grad_norm": 0.15625,
+      "learning_rate": 0.02,
+      "loss": 1.6342,
+      "step": 2036
+    },
+    {
+      "epoch": 0.18857837181044956,
+      "grad_norm": 0.154296875,
+      "learning_rate": 0.02,
+      "loss": 1.7019,
+      "step": 2037
+    },
+    {
+      "epoch": 0.18867094833072962,
+      "grad_norm": 0.16015625,
+      "learning_rate": 0.02,
+      "loss": 1.7128,
+      "step": 2038
+    },
+    {
+      "epoch": 0.18876352485100967,
+      "grad_norm": 0.1640625,
+      "learning_rate": 0.02,
+      "loss": 1.6879,
+      "step": 2039
+    },
+    {
+      "epoch": 0.18885610137128972,
+      "grad_norm": 0.1728515625,
+      "learning_rate": 0.02,
+      "loss": 1.7407,
+      "step": 2040
+    },
+    {
+      "epoch": 0.18894867789156974,
+      "grad_norm": 0.171875,
+      "learning_rate": 0.02,
+      "loss": 1.7686,
+      "step": 2041
+    },
+    {
+      "epoch": 0.1890412544118498,
+      "grad_norm": 0.15625,
+      "learning_rate": 0.02,
+      "loss": 1.7048,
+      "step": 2042
+    },
+    {
+      "epoch": 0.18913383093212985,
+      "grad_norm": 0.15625,
+      "learning_rate": 0.02,
+      "loss": 1.7504,
+      "step": 2043
+    },
+    {
+      "epoch": 0.18922640745240987,
+      "grad_norm": 0.1533203125,
+      "learning_rate": 0.02,
+      "loss": 1.6752,
+      "step": 2044
+    },
+    {
+      "epoch": 0.18931898397268992,
+      "grad_norm": 0.1435546875,
+      "learning_rate": 0.02,
+      "loss": 1.6606,
+      "step": 2045
+    },
+    {
+      "epoch": 0.18941156049296998,
+      "grad_norm": 0.15625,
+      "learning_rate": 0.02,
+      "loss": 1.7154,
+      "step": 2046
+    },
+    {
+      "epoch": 0.18950413701325,
+      "grad_norm": 0.14453125,
+      "learning_rate": 0.02,
+      "loss": 1.6755,
+      "step": 2047
+    },
+    {
+      "epoch": 0.18959671353353005,
+      "grad_norm": 0.146484375,
+      "learning_rate": 0.02,
+      "loss": 1.6865,
+      "step": 2048
+    },
+    {
+      "epoch": 0.1896892900538101,
+      "grad_norm": 0.1494140625,
+      "learning_rate": 0.02,
+      "loss": 1.671,
+      "step": 2049
+    },
+    {
+      "epoch": 0.18978186657409016,
+      "grad_norm": 0.1611328125,
+      "learning_rate": 0.02,
+      "loss": 1.7964,
+      "step": 2050
+    },
+    {
+      "epoch": 0.18987444309437018,
+      "grad_norm": 0.1552734375,
+      "learning_rate": 0.02,
+      "loss": 1.7116,
+      "step": 2051
+    },
+    {
+      "epoch": 0.18996701961465023,
+      "grad_norm": 0.15234375,
+      "learning_rate": 0.02,
+      "loss": 1.7394,
+      "step": 2052
+    },
+    {
+      "epoch": 0.1900595961349303,
+      "grad_norm": 0.1611328125,
+      "learning_rate": 0.02,
+      "loss": 1.7398,
+      "step": 2053
+    },
+    {
+      "epoch": 0.1901521726552103,
+      "grad_norm": 0.1552734375,
+      "learning_rate": 0.02,
+      "loss": 1.72,
+      "step": 2054
+    },
+    {
+      "epoch": 0.19024474917549036,
+      "grad_norm": 0.1572265625,
+      "learning_rate": 0.02,
+      "loss": 1.6506,
+      "step": 2055
+    },
+    {
+      "epoch": 0.19033732569577042,
+      "grad_norm": 0.1416015625,
+      "learning_rate": 0.02,
+      "loss": 1.6708,
+      "step": 2056
+    },
+    {
+      "epoch": 0.19042990221605047,
+      "grad_norm": 0.1533203125,
+      "learning_rate": 0.02,
+      "loss": 1.6983,
+      "step": 2057
+    },
+    {
+      "epoch": 0.1905224787363305,
+      "grad_norm": 0.16015625,
+      "learning_rate": 0.02,
+      "loss": 1.6789,
+      "step": 2058
+    },
+    {
+      "epoch": 0.19061505525661054,
+      "grad_norm": 0.154296875,
+      "learning_rate": 0.02,
+      "loss": 1.7309,
+      "step": 2059
+    },
+    {
+      "epoch": 0.1907076317768906,
+      "grad_norm": 0.15234375,
+      "learning_rate": 0.02,
+      "loss": 1.7047,
+      "step": 2060
+    },
+    {
+      "epoch": 0.19080020829717062,
+      "grad_norm": 0.154296875,
+      "learning_rate": 0.02,
+      "loss": 1.6947,
+      "step": 2061
+    },
+    {
+      "epoch": 0.19089278481745067,
+      "grad_norm": 0.150390625,
+      "learning_rate": 0.02,
+      "loss": 1.7503,
+      "step": 2062
+    },
+    {
+      "epoch": 0.19098536133773072,
+      "grad_norm": 0.1591796875,
+      "learning_rate": 0.02,
+      "loss": 1.7178,
+      "step": 2063
+    },
+    {
+      "epoch": 0.19107793785801075,
+      "grad_norm": 0.154296875,
+      "learning_rate": 0.02,
+      "loss": 1.6914,
+      "step": 2064
+    },
+    {
+      "epoch": 0.1911705143782908,
+      "grad_norm": 0.1650390625,
+      "learning_rate": 0.02,
+      "loss": 1.7948,
+      "step": 2065
+    },
+    {
+      "epoch": 0.19126309089857085,
+      "grad_norm": 0.1572265625,
+      "learning_rate": 0.02,
+      "loss": 1.7001,
+      "step": 2066
+    },
+    {
+      "epoch": 0.1913556674188509,
+      "grad_norm": 0.16015625,
+      "learning_rate": 0.02,
+      "loss": 1.6825,
+      "step": 2067
+    },
+    {
+      "epoch": 0.19144824393913093,
+      "grad_norm": 0.1728515625,
+      "learning_rate": 0.02,
+      "loss": 1.6968,
+      "step": 2068
+    },
+    {
+      "epoch": 0.19154082045941098,
+      "grad_norm": 0.1611328125,
+      "learning_rate": 0.02,
+      "loss": 1.7039,
+      "step": 2069
+    },
+    {
+      "epoch": 0.19163339697969103,
+      "grad_norm": 0.162109375,
+      "learning_rate": 0.02,
+      "loss": 1.6617,
+      "step": 2070
+    },
+    {
+      "epoch": 0.19172597349997106,
+      "grad_norm": 0.177734375,
+      "learning_rate": 0.02,
+      "loss": 1.7288,
+      "step": 2071
+    },
+    {
+      "epoch": 0.1918185500202511,
+      "grad_norm": 0.15234375,
+      "learning_rate": 0.02,
+      "loss": 1.6488,
+      "step": 2072
+    },
+    {
+      "epoch": 0.19191112654053116,
+      "grad_norm": 0.16015625,
+      "learning_rate": 0.02,
+      "loss": 1.6898,
+      "step": 2073
+    },
+    {
+      "epoch": 0.19200370306081121,
+      "grad_norm": 0.16015625,
+      "learning_rate": 0.02,
+      "loss": 1.7528,
+      "step": 2074
+    },
+    {
+      "epoch": 0.19209627958109124,
+      "grad_norm": 0.16015625,
+      "learning_rate": 0.02,
+      "loss": 1.672,
+      "step": 2075
+    },
+    {
+      "epoch": 0.1921888561013713,
+      "grad_norm": 0.1552734375,
+      "learning_rate": 0.02,
+      "loss": 1.6858,
+      "step": 2076
+    },
+    {
+      "epoch": 0.19228143262165134,
+      "grad_norm": 0.16015625,
+      "learning_rate": 0.02,
+      "loss": 1.6184,
+      "step": 2077
+    },
+    {
+      "epoch": 0.19237400914193137,
+      "grad_norm": 0.16015625,
+      "learning_rate": 0.02,
+      "loss": 1.7549,
+      "step": 2078
+    },
+    {
+      "epoch": 0.19246658566221142,
+      "grad_norm": 0.1572265625,
+      "learning_rate": 0.02,
+      "loss": 1.7248,
+      "step": 2079
+    },
+    {
+      "epoch": 0.19255916218249147,
+      "grad_norm": 0.1640625,
+      "learning_rate": 0.02,
+      "loss": 1.7108,
+      "step": 2080
+    },
+    {
+      "epoch": 0.1926517387027715,
+      "grad_norm": 0.16015625,
+      "learning_rate": 0.02,
+      "loss": 1.6804,
+      "step": 2081
+    },
+    {
+      "epoch": 0.19274431522305155,
+      "grad_norm": 0.15625,
+      "learning_rate": 0.02,
+      "loss": 1.6753,
+      "step": 2082
+    },
+    {
+      "epoch": 0.1928368917433316,
+      "grad_norm": 0.1494140625,
+      "learning_rate": 0.02,
+      "loss": 1.7203,
+      "step": 2083
+    },
+    {
+      "epoch": 0.19292946826361165,
+      "grad_norm": 0.150390625,
+      "learning_rate": 0.02,
+      "loss": 1.6809,
+      "step": 2084
+    },
+    {
+      "epoch": 0.19302204478389168,
+      "grad_norm": 0.15234375,
+      "learning_rate": 0.02,
+      "loss": 1.6447,
+      "step": 2085
+    },
+    {
+      "epoch": 0.19311462130417173,
+      "grad_norm": 0.1591796875,
+      "learning_rate": 0.02,
+      "loss": 1.7363,
+      "step": 2086
+    },
+    {
+      "epoch": 0.19320719782445178,
+      "grad_norm": 0.1416015625,
+      "learning_rate": 0.02,
+      "loss": 1.7091,
+      "step": 2087
+    },
+    {
+      "epoch": 0.1932997743447318,
+      "grad_norm": 0.14453125,
+      "learning_rate": 0.02,
+      "loss": 1.7135,
+      "step": 2088
+    },
+    {
+      "epoch": 0.19339235086501186,
+      "grad_norm": 0.1552734375,
+      "learning_rate": 0.02,
+      "loss": 1.7697,
+      "step": 2089
+    },
+    {
+      "epoch": 0.1934849273852919,
+      "grad_norm": 0.150390625,
+      "learning_rate": 0.02,
+      "loss": 1.6556,
+      "step": 2090
+    },
+    {
+      "epoch": 0.19357750390557196,
+      "grad_norm": 0.146484375,
+      "learning_rate": 0.02,
+      "loss": 1.6341,
+      "step": 2091
+    },
+    {
+      "epoch": 0.193670080425852,
+      "grad_norm": 0.1533203125,
+      "learning_rate": 0.02,
+      "loss": 1.6572,
+      "step": 2092
+    },
+    {
+      "epoch": 0.19376265694613204,
+      "grad_norm": 0.1728515625,
+      "learning_rate": 0.02,
+      "loss": 1.761,
+      "step": 2093
+    },
+    {
+      "epoch": 0.1938552334664121,
+      "grad_norm": 0.1611328125,
+      "learning_rate": 0.02,
+      "loss": 1.7115,
+      "step": 2094
+    },
+    {
+      "epoch": 0.19394780998669212,
+      "grad_norm": 0.16015625,
+      "learning_rate": 0.02,
+      "loss": 1.7205,
+      "step": 2095
+    },
+    {
+      "epoch": 0.19404038650697217,
+      "grad_norm": 0.15234375,
+      "learning_rate": 0.02,
+      "loss": 1.6502,
+      "step": 2096
+    },
+    {
+      "epoch": 0.19413296302725222,
+      "grad_norm": 0.1552734375,
+      "learning_rate": 0.02,
+      "loss": 1.7324,
+      "step": 2097
+    },
+    {
+      "epoch": 0.19422553954753224,
+      "grad_norm": 0.1513671875,
+      "learning_rate": 0.02,
+      "loss": 1.669,
+      "step": 2098
+    },
+    {
+      "epoch": 0.1943181160678123,
+      "grad_norm": 0.162109375,
+      "learning_rate": 0.02,
+      "loss": 1.7379,
+      "step": 2099
+    },
+    {
+      "epoch": 0.19441069258809235,
+      "grad_norm": 0.15625,
+      "learning_rate": 0.02,
+      "loss": 1.6664,
+      "step": 2100
+    },
+    {
+      "epoch": 0.1945032691083724,
+      "grad_norm": 0.1455078125,
+      "learning_rate": 0.02,
+      "loss": 1.5892,
+      "step": 2101
+    },
+    {
+      "epoch": 0.19459584562865243,
+      "grad_norm": 0.1455078125,
+      "learning_rate": 0.02,
+      "loss": 1.7528,
+      "step": 2102
+    },
+    {
+      "epoch": 0.19468842214893248,
+      "grad_norm": 0.1455078125,
+      "learning_rate": 0.02,
+      "loss": 1.6901,
+      "step": 2103
+    },
+    {
+      "epoch": 0.19478099866921253,
+      "grad_norm": 0.1513671875,
+      "learning_rate": 0.02,
+      "loss": 1.6666,
+      "step": 2104
+    },
+    {
+      "epoch": 0.19487357518949255,
+      "grad_norm": 0.1455078125,
+      "learning_rate": 0.02,
+      "loss": 1.7132,
+      "step": 2105
+    },
+    {
+      "epoch": 0.1949661517097726,
+      "grad_norm": 0.142578125,
+      "learning_rate": 0.02,
+      "loss": 1.6642,
+      "step": 2106
+    },
+    {
+      "epoch": 0.19505872823005266,
+      "grad_norm": 0.16015625,
+      "learning_rate": 0.02,
+      "loss": 1.7328,
+      "step": 2107
+    },
+    {
+      "epoch": 0.1951513047503327,
+      "grad_norm": 0.150390625,
+      "learning_rate": 0.02,
+      "loss": 1.744,
+      "step": 2108
+    },
+    {
+      "epoch": 0.19524388127061273,
+      "grad_norm": 0.1474609375,
+      "learning_rate": 0.02,
+      "loss": 1.7135,
+      "step": 2109
+    },
+    {
+      "epoch": 0.1953364577908928,
+      "grad_norm": 0.1416015625,
+      "learning_rate": 0.02,
+      "loss": 1.7284,
+      "step": 2110
+    },
+    {
+      "epoch": 0.19542903431117284,
+      "grad_norm": 0.1455078125,
+      "learning_rate": 0.02,
+      "loss": 1.6999,
+      "step": 2111
+    },
+    {
+      "epoch": 0.19552161083145286,
+      "grad_norm": 0.15234375,
+      "learning_rate": 0.02,
+      "loss": 1.7638,
+      "step": 2112
+    },
+    {
+      "epoch": 0.19561418735173292,
+      "grad_norm": 0.154296875,
+      "learning_rate": 0.02,
+      "loss": 1.7332,
+      "step": 2113
+    },
+    {
+      "epoch": 0.19570676387201297,
+      "grad_norm": 0.14453125,
+      "learning_rate": 0.02,
+      "loss": 1.7082,
+      "step": 2114
+    },
+    {
+      "epoch": 0.195799340392293,
+      "grad_norm": 0.1611328125,
+      "learning_rate": 0.02,
+      "loss": 1.7836,
+      "step": 2115
+    },
+    {
+      "epoch": 0.19589191691257304,
+      "grad_norm": 0.1611328125,
+      "learning_rate": 0.02,
+      "loss": 1.6755,
+      "step": 2116
+    },
+    {
+      "epoch": 0.1959844934328531,
+      "grad_norm": 0.1591796875,
+      "learning_rate": 0.02,
+      "loss": 1.6913,
+      "step": 2117
+    },
+    {
+      "epoch": 0.19607706995313315,
+      "grad_norm": 0.16015625,
+      "learning_rate": 0.02,
+      "loss": 1.729,
+      "step": 2118
+    },
+    {
+      "epoch": 0.19616964647341317,
+      "grad_norm": 0.1630859375,
+      "learning_rate": 0.02,
+      "loss": 1.6792,
+      "step": 2119
+    },
+    {
+      "epoch": 0.19626222299369322,
+      "grad_norm": 0.150390625,
+      "learning_rate": 0.02,
+      "loss": 1.7056,
+      "step": 2120
+    },
+    {
+      "epoch": 0.19635479951397328,
+      "grad_norm": 0.166015625,
+      "learning_rate": 0.02,
+      "loss": 1.7487,
+      "step": 2121
+    },
+    {
+      "epoch": 0.1964473760342533,
+      "grad_norm": 0.1689453125,
+      "learning_rate": 0.02,
+      "loss": 1.7112,
+      "step": 2122
+    },
+    {
+      "epoch": 0.19653995255453335,
+      "grad_norm": 0.18359375,
+      "learning_rate": 0.02,
+      "loss": 1.7427,
+      "step": 2123
+    },
+    {
+      "epoch": 0.1966325290748134,
+      "grad_norm": 0.1533203125,
+      "learning_rate": 0.02,
+      "loss": 1.6512,
+      "step": 2124
+    },
+    {
+      "epoch": 0.19672510559509346,
+      "grad_norm": 0.1484375,
+      "learning_rate": 0.02,
+      "loss": 1.6859,
+      "step": 2125
+    },
+    {
+      "epoch": 0.19681768211537348,
+      "grad_norm": 0.158203125,
+      "learning_rate": 0.02,
+      "loss": 1.7452,
+      "step": 2126
+    },
+    {
+      "epoch": 0.19691025863565353,
+      "grad_norm": 0.1640625,
+      "learning_rate": 0.02,
+      "loss": 1.7576,
+      "step": 2127
+    },
+    {
+      "epoch": 0.1970028351559336,
+      "grad_norm": 0.1484375,
+      "learning_rate": 0.02,
+      "loss": 1.6885,
+      "step": 2128
+    },
+    {
+      "epoch": 0.1970954116762136,
+      "grad_norm": 0.15234375,
+      "learning_rate": 0.02,
+      "loss": 1.6833,
+      "step": 2129
+    },
+    {
+      "epoch": 0.19718798819649366,
+      "grad_norm": 0.158203125,
+      "learning_rate": 0.02,
+      "loss": 1.7472,
+      "step": 2130
+    },
+    {
+      "epoch": 0.19728056471677372,
+      "grad_norm": 0.166015625,
+      "learning_rate": 0.02,
+      "loss": 1.6968,
+      "step": 2131
+    },
+    {
+      "epoch": 0.19737314123705374,
+      "grad_norm": 0.1552734375,
+      "learning_rate": 0.02,
+      "loss": 1.7243,
+      "step": 2132
+    },
+    {
+      "epoch": 0.1974657177573338,
+      "grad_norm": 0.1552734375,
+      "learning_rate": 0.02,
+      "loss": 1.7443,
+      "step": 2133
+    },
+    {
+      "epoch": 0.19755829427761384,
+      "grad_norm": 0.1416015625,
+      "learning_rate": 0.02,
+      "loss": 1.7289,
+      "step": 2134
+    },
+    {
+      "epoch": 0.1976508707978939,
+      "grad_norm": 0.15625,
+      "learning_rate": 0.02,
+      "loss": 1.6307,
+      "step": 2135
+    },
+    {
+      "epoch": 0.19774344731817392,
+      "grad_norm": 0.1611328125,
+      "learning_rate": 0.02,
+      "loss": 1.6646,
+      "step": 2136
+    },
+    {
+      "epoch": 0.19783602383845397,
+      "grad_norm": 0.1455078125,
+      "learning_rate": 0.02,
+      "loss": 1.6737,
+      "step": 2137
+    },
+    {
+      "epoch": 0.19792860035873402,
+      "grad_norm": 0.140625,
+      "learning_rate": 0.02,
+      "loss": 1.6805,
+      "step": 2138
+    },
+    {
+      "epoch": 0.19802117687901405,
+      "grad_norm": 0.1533203125,
+      "learning_rate": 0.02,
+      "loss": 1.66,
+      "step": 2139
+    },
+    {
+      "epoch": 0.1981137533992941,
+      "grad_norm": 0.1455078125,
+      "learning_rate": 0.02,
+      "loss": 1.6444,
+      "step": 2140
+    },
+    {
+      "epoch": 0.19820632991957415,
+      "grad_norm": 0.1611328125,
+      "learning_rate": 0.02,
+      "loss": 1.7332,
+      "step": 2141
+    },
+    {
+      "epoch": 0.1982989064398542,
+      "grad_norm": 0.16015625,
+      "learning_rate": 0.02,
+      "loss": 1.6902,
+      "step": 2142
+    },
+    {
+      "epoch": 0.19839148296013423,
+      "grad_norm": 0.1591796875,
+      "learning_rate": 0.02,
+      "loss": 1.7389,
+      "step": 2143
+    },
+    {
+      "epoch": 0.19848405948041428,
+      "grad_norm": 2.71875,
+      "learning_rate": 0.02,
+      "loss": 1.742,
+      "step": 2144
+    },
+    {
+      "epoch": 0.19857663600069433,
+      "grad_norm": 0.1416015625,
+      "learning_rate": 0.02,
+      "loss": 1.6547,
+      "step": 2145
+    },
+    {
+      "epoch": 0.19866921252097436,
+      "grad_norm": 0.177734375,
+      "learning_rate": 0.02,
+      "loss": 1.7349,
+      "step": 2146
+    },
+    {
+      "epoch": 0.1987617890412544,
+      "grad_norm": 0.1474609375,
+      "learning_rate": 0.02,
+      "loss": 1.6321,
+      "step": 2147
+    },
+    {
+      "epoch": 0.19885436556153446,
+      "grad_norm": 0.1630859375,
+      "learning_rate": 0.02,
+      "loss": 1.6685,
+      "step": 2148
+    },
+    {
+      "epoch": 0.1989469420818145,
+      "grad_norm": 0.154296875,
+      "learning_rate": 0.02,
+      "loss": 1.67,
+      "step": 2149
+    },
+    {
+      "epoch": 0.19903951860209454,
+      "grad_norm": 0.1591796875,
+      "learning_rate": 0.02,
+      "loss": 1.7364,
+      "step": 2150
+    },
+    {
+      "epoch": 0.1991320951223746,
+      "grad_norm": 0.1474609375,
+      "learning_rate": 0.02,
+      "loss": 1.6153,
+      "step": 2151
+    },
+    {
+      "epoch": 0.19922467164265464,
+      "grad_norm": 0.1572265625,
+      "learning_rate": 0.02,
+      "loss": 1.6903,
+      "step": 2152
+    },
+    {
+      "epoch": 0.19931724816293467,
+      "grad_norm": 0.16015625,
+      "learning_rate": 0.02,
+      "loss": 1.6518,
+      "step": 2153
+    },
+    {
+      "epoch": 0.19940982468321472,
+      "grad_norm": 0.1630859375,
+      "learning_rate": 0.02,
+      "loss": 1.7578,
+      "step": 2154
+    },
+    {
+      "epoch": 0.19950240120349477,
+      "grad_norm": 0.1572265625,
+      "learning_rate": 0.02,
+      "loss": 1.6931,
+      "step": 2155
+    },
+    {
+      "epoch": 0.1995949777237748,
+      "grad_norm": 0.1591796875,
+      "learning_rate": 0.02,
+      "loss": 1.6951,
+      "step": 2156
+    },
+    {
+      "epoch": 0.19968755424405485,
+      "grad_norm": 0.1455078125,
+      "learning_rate": 0.02,
+      "loss": 1.6124,
+      "step": 2157
+    },
+    {
+      "epoch": 0.1997801307643349,
+      "grad_norm": 0.14453125,
+      "learning_rate": 0.02,
+      "loss": 1.6833,
+      "step": 2158
+    },
+    {
+      "epoch": 0.19987270728461495,
+      "grad_norm": 0.1513671875,
+      "learning_rate": 0.02,
+      "loss": 1.7332,
+      "step": 2159
+    },
+    {
+      "epoch": 0.19996528380489498,
+      "grad_norm": 0.1591796875,
+      "learning_rate": 0.02,
+      "loss": 1.6938,
+      "step": 2160
+    },
+    {
+      "epoch": 0.20005786032517503,
+      "grad_norm": 0.1630859375,
+      "learning_rate": 0.02,
+      "loss": 1.7023,
+      "step": 2161
+    },
+    {
+      "epoch": 0.20015043684545508,
+      "grad_norm": 0.146484375,
+      "learning_rate": 0.02,
+      "loss": 1.679,
+      "step": 2162
+    },
+    {
+      "epoch": 0.2002430133657351,
+      "grad_norm": 0.15625,
+      "learning_rate": 0.02,
+      "loss": 1.7224,
+      "step": 2163
+    },
+    {
+      "epoch": 0.20033558988601516,
+      "grad_norm": 0.1474609375,
+      "learning_rate": 0.02,
+      "loss": 1.695,
+      "step": 2164
+    },
+    {
+      "epoch": 0.2004281664062952,
+      "grad_norm": 0.154296875,
+      "learning_rate": 0.02,
+      "loss": 1.6812,
+      "step": 2165
+    },
+    {
+      "epoch": 0.20052074292657524,
+      "grad_norm": 0.1533203125,
+      "learning_rate": 0.02,
+      "loss": 1.7512,
+      "step": 2166
+    },
+    {
+      "epoch": 0.2006133194468553,
+      "grad_norm": 0.150390625,
+      "learning_rate": 0.02,
+      "loss": 1.693,
+      "step": 2167
+    },
+    {
+      "epoch": 0.20070589596713534,
+      "grad_norm": 0.1533203125,
+      "learning_rate": 0.02,
+      "loss": 1.7483,
+      "step": 2168
+    },
+    {
+      "epoch": 0.2007984724874154,
+      "grad_norm": 0.171875,
+      "learning_rate": 0.02,
+      "loss": 1.7088,
+      "step": 2169
+    },
+    {
+      "epoch": 0.20089104900769542,
+      "grad_norm": 0.162109375,
+      "learning_rate": 0.02,
+      "loss": 1.7216,
+      "step": 2170
+    },
+    {
+      "epoch": 0.20098362552797547,
+      "grad_norm": 0.1591796875,
+      "learning_rate": 0.02,
+      "loss": 1.6827,
+      "step": 2171
+    },
+    {
+      "epoch": 0.20107620204825552,
+      "grad_norm": 0.1552734375,
+      "learning_rate": 0.02,
+      "loss": 1.6509,
+      "step": 2172
+    },
+    {
+      "epoch": 0.20116877856853554,
+      "grad_norm": 0.158203125,
+      "learning_rate": 0.02,
+      "loss": 1.7076,
+      "step": 2173
+    },
+    {
+      "epoch": 0.2012613550888156,
+      "grad_norm": 0.1484375,
+      "learning_rate": 0.02,
+      "loss": 1.678,
+      "step": 2174
+    },
+    {
+      "epoch": 0.20135393160909565,
+      "grad_norm": 0.1533203125,
+      "learning_rate": 0.02,
+      "loss": 1.6815,
+      "step": 2175
+    },
+    {
+      "epoch": 0.2014465081293757,
+      "grad_norm": 0.154296875,
+      "learning_rate": 0.02,
+      "loss": 1.6827,
+      "step": 2176
+    },
+    {
+      "epoch": 0.20153908464965573,
+      "grad_norm": 0.154296875,
+      "learning_rate": 0.02,
+      "loss": 1.6789,
+      "step": 2177
+    },
+    {
+      "epoch": 0.20163166116993578,
+      "grad_norm": 0.1474609375,
+      "learning_rate": 0.02,
+      "loss": 1.6442,
+      "step": 2178
+    },
+    {
+      "epoch": 0.20172423769021583,
+      "grad_norm": 0.150390625,
+      "learning_rate": 0.02,
+      "loss": 1.6823,
+      "step": 2179
+    },
+    {
+      "epoch": 0.20181681421049585,
+      "grad_norm": 0.1513671875,
+      "learning_rate": 0.02,
+      "loss": 1.6828,
+      "step": 2180
+    },
+    {
+      "epoch": 0.2019093907307759,
+      "grad_norm": 0.1474609375,
+      "learning_rate": 0.02,
+      "loss": 1.657,
+      "step": 2181
+    },
+    {
+      "epoch": 0.20200196725105596,
+      "grad_norm": 0.1455078125,
+      "learning_rate": 0.02,
+      "loss": 1.744,
+      "step": 2182
+    },
+    {
+      "epoch": 0.20209454377133598,
+      "grad_norm": 0.1416015625,
+      "learning_rate": 0.02,
+      "loss": 1.6591,
+      "step": 2183
+    },
+    {
+      "epoch": 0.20218712029161603,
+      "grad_norm": 0.1396484375,
+      "learning_rate": 0.02,
+      "loss": 1.6668,
+      "step": 2184
+    },
+    {
+      "epoch": 0.2022796968118961,
+      "grad_norm": 0.1435546875,
+      "learning_rate": 0.02,
+      "loss": 1.6198,
+      "step": 2185
+    },
+    {
+      "epoch": 0.20237227333217614,
+      "grad_norm": 0.1494140625,
+      "learning_rate": 0.02,
+      "loss": 1.6713,
+      "step": 2186
+    },
+    {
+      "epoch": 0.20246484985245616,
+      "grad_norm": 0.1611328125,
+      "learning_rate": 0.02,
+      "loss": 1.7345,
+      "step": 2187
+    },
+    {
+      "epoch": 0.20255742637273622,
+      "grad_norm": 0.1484375,
+      "learning_rate": 0.02,
+      "loss": 1.6319,
+      "step": 2188
+    },
+    {
+      "epoch": 0.20265000289301627,
+      "grad_norm": 0.1552734375,
+      "learning_rate": 0.02,
+      "loss": 1.6334,
+      "step": 2189
+    },
+    {
+      "epoch": 0.2027425794132963,
+      "grad_norm": 0.158203125,
+      "learning_rate": 0.02,
+      "loss": 1.6899,
+      "step": 2190
+    },
+    {
+      "epoch": 0.20283515593357634,
+      "grad_norm": 0.1435546875,
+      "learning_rate": 0.02,
+      "loss": 1.5566,
+      "step": 2191
+    },
+    {
+      "epoch": 0.2029277324538564,
+      "grad_norm": 0.150390625,
+      "learning_rate": 0.02,
+      "loss": 1.7674,
+      "step": 2192
+    },
+    {
+      "epoch": 0.20302030897413642,
+      "grad_norm": 0.166015625,
+      "learning_rate": 0.02,
+      "loss": 1.7312,
+      "step": 2193
+    },
+    {
+      "epoch": 0.20311288549441647,
+      "grad_norm": 0.150390625,
+      "learning_rate": 0.02,
+      "loss": 1.6664,
+      "step": 2194
+    },
+    {
+      "epoch": 0.20320546201469653,
+      "grad_norm": 0.1357421875,
+      "learning_rate": 0.02,
+      "loss": 1.6446,
+      "step": 2195
+    },
+    {
+      "epoch": 0.20329803853497658,
+      "grad_norm": 0.1494140625,
+      "learning_rate": 0.02,
+      "loss": 1.6689,
+      "step": 2196
+    },
+    {
+      "epoch": 0.2033906150552566,
+      "grad_norm": 0.1474609375,
+      "learning_rate": 0.02,
+      "loss": 1.6863,
+      "step": 2197
+    },
+    {
+      "epoch": 0.20348319157553665,
+      "grad_norm": 0.146484375,
+      "learning_rate": 0.02,
+      "loss": 1.6405,
+      "step": 2198
+    },
+    {
+      "epoch": 0.2035757680958167,
+      "grad_norm": 0.1533203125,
+      "learning_rate": 0.02,
+      "loss": 1.6826,
+      "step": 2199
+    },
+    {
+      "epoch": 0.20366834461609673,
+      "grad_norm": 0.1572265625,
+      "learning_rate": 0.02,
+      "loss": 1.7133,
+      "step": 2200
+    },
+    {
+      "epoch": 0.20376092113637678,
+      "grad_norm": 0.1689453125,
+      "learning_rate": 0.02,
+      "loss": 1.7858,
+      "step": 2201
+    },
+    {
+      "epoch": 0.20385349765665683,
+      "grad_norm": 0.1494140625,
+      "learning_rate": 0.02,
+      "loss": 1.6514,
+      "step": 2202
+    },
+    {
+      "epoch": 0.2039460741769369,
+      "grad_norm": 0.1474609375,
+      "learning_rate": 0.02,
+      "loss": 1.6532,
+      "step": 2203
+    },
+    {
+      "epoch": 0.2040386506972169,
+      "grad_norm": 0.162109375,
+      "learning_rate": 0.02,
+      "loss": 1.7098,
+      "step": 2204
+    },
+    {
+      "epoch": 0.20413122721749696,
+      "grad_norm": 0.1689453125,
+      "learning_rate": 0.02,
+      "loss": 1.749,
+      "step": 2205
+    },
+    {
+      "epoch": 0.20422380373777702,
+      "grad_norm": 0.158203125,
+      "learning_rate": 0.02,
+      "loss": 1.7086,
+      "step": 2206
+    },
+    {
+      "epoch": 0.20431638025805704,
+      "grad_norm": 0.1650390625,
+      "learning_rate": 0.02,
+      "loss": 1.7345,
+      "step": 2207
+    },
+    {
+      "epoch": 0.2044089567783371,
+      "grad_norm": 0.166015625,
+      "learning_rate": 0.02,
+      "loss": 1.6735,
+      "step": 2208
+    },
+    {
+      "epoch": 0.20450153329861714,
+      "grad_norm": 0.154296875,
+      "learning_rate": 0.02,
+      "loss": 1.6244,
+      "step": 2209
+    },
+    {
+      "epoch": 0.20459410981889717,
+      "grad_norm": 0.1494140625,
+      "learning_rate": 0.02,
+      "loss": 1.6507,
+      "step": 2210
+    },
+    {
+      "epoch": 0.20468668633917722,
+      "grad_norm": 0.142578125,
+      "learning_rate": 0.02,
+      "loss": 1.5844,
+      "step": 2211
+    },
+    {
+      "epoch": 0.20477926285945727,
+      "grad_norm": 0.1484375,
+      "learning_rate": 0.02,
+      "loss": 1.6182,
+      "step": 2212
+    },
+    {
+      "epoch": 0.20487183937973732,
+      "grad_norm": 0.1513671875,
+      "learning_rate": 0.02,
+      "loss": 1.7161,
+      "step": 2213
+    },
+    {
+      "epoch": 0.20496441590001735,
+      "grad_norm": 0.15625,
+      "learning_rate": 0.02,
+      "loss": 1.7438,
+      "step": 2214
+    },
+    {
+      "epoch": 0.2050569924202974,
+      "grad_norm": 0.14453125,
+      "learning_rate": 0.02,
+      "loss": 1.5888,
+      "step": 2215
+    },
+    {
+      "epoch": 0.20514956894057745,
+      "grad_norm": 0.140625,
+      "learning_rate": 0.02,
+      "loss": 1.7291,
+      "step": 2216
+    },
+    {
+      "epoch": 0.20524214546085748,
+      "grad_norm": 0.142578125,
+      "learning_rate": 0.02,
+      "loss": 1.6678,
+      "step": 2217
+    },
+    {
+      "epoch": 0.20533472198113753,
+      "grad_norm": 0.1435546875,
+      "learning_rate": 0.02,
+      "loss": 1.6798,
+      "step": 2218
+    },
+    {
+      "epoch": 0.20542729850141758,
+      "grad_norm": 0.14453125,
+      "learning_rate": 0.02,
+      "loss": 1.6693,
+      "step": 2219
+    },
+    {
+      "epoch": 0.20551987502169763,
+      "grad_norm": 0.1455078125,
+      "learning_rate": 0.02,
+      "loss": 1.6952,
+      "step": 2220
+    },
+    {
+      "epoch": 0.20561245154197766,
+      "grad_norm": 0.1494140625,
+      "learning_rate": 0.02,
+      "loss": 1.6962,
+      "step": 2221
+    },
+    {
+      "epoch": 0.2057050280622577,
+      "grad_norm": 0.1484375,
+      "learning_rate": 0.02,
+      "loss": 1.7078,
+      "step": 2222
+    },
+    {
+      "epoch": 0.20579760458253776,
+      "grad_norm": 0.1484375,
+      "learning_rate": 0.02,
+      "loss": 1.6842,
+      "step": 2223
+    },
+    {
+      "epoch": 0.2058901811028178,
+      "grad_norm": 0.14453125,
+      "learning_rate": 0.02,
+      "loss": 1.6564,
+      "step": 2224
+    },
+    {
+      "epoch": 0.20598275762309784,
+      "grad_norm": 0.138671875,
+      "learning_rate": 0.02,
+      "loss": 1.6397,
+      "step": 2225
+    },
+    {
+      "epoch": 0.2060753341433779,
+      "grad_norm": 0.1494140625,
+      "learning_rate": 0.02,
+      "loss": 1.6697,
+      "step": 2226
+    },
+    {
+      "epoch": 0.20616791066365792,
+      "grad_norm": 0.15625,
+      "learning_rate": 0.02,
+      "loss": 1.7181,
+      "step": 2227
+    },
+    {
+      "epoch": 0.20626048718393797,
+      "grad_norm": 0.154296875,
+      "learning_rate": 0.02,
+      "loss": 1.7259,
+      "step": 2228
+    },
+    {
+      "epoch": 0.20635306370421802,
+      "grad_norm": 0.1533203125,
+      "learning_rate": 0.02,
+      "loss": 1.6679,
+      "step": 2229
+    },
+    {
+      "epoch": 0.20644564022449807,
+      "grad_norm": 0.1552734375,
+      "learning_rate": 0.02,
+      "loss": 1.6626,
+      "step": 2230
+    },
+    {
+      "epoch": 0.2065382167447781,
+      "grad_norm": 0.15625,
+      "learning_rate": 0.02,
+      "loss": 1.6985,
+      "step": 2231
+    },
+    {
+      "epoch": 0.20663079326505815,
+      "grad_norm": 0.1494140625,
+      "learning_rate": 0.02,
+      "loss": 1.6018,
+      "step": 2232
+    },
+    {
+      "epoch": 0.2067233697853382,
+      "grad_norm": 0.1533203125,
+      "learning_rate": 0.02,
+      "loss": 1.6282,
+      "step": 2233
+    },
+    {
+      "epoch": 0.20681594630561823,
+      "grad_norm": 0.1552734375,
+      "learning_rate": 0.02,
+      "loss": 1.6331,
+      "step": 2234
+    },
+    {
+      "epoch": 0.20690852282589828,
+      "grad_norm": 0.15234375,
+      "learning_rate": 0.02,
+      "loss": 1.6835,
+      "step": 2235
+    },
+    {
+      "epoch": 0.20700109934617833,
+      "grad_norm": 0.1640625,
+      "learning_rate": 0.02,
+      "loss": 1.6722,
+      "step": 2236
+    },
+    {
+      "epoch": 0.20709367586645838,
+      "grad_norm": 0.1669921875,
+      "learning_rate": 0.02,
+      "loss": 1.6795,
+      "step": 2237
+    },
+    {
+      "epoch": 0.2071862523867384,
+      "grad_norm": 0.162109375,
+      "learning_rate": 0.02,
+      "loss": 1.7243,
+      "step": 2238
+    },
+    {
+      "epoch": 0.20727882890701846,
+      "grad_norm": 0.158203125,
+      "learning_rate": 0.02,
+      "loss": 1.6204,
+      "step": 2239
+    },
+    {
+      "epoch": 0.2073714054272985,
+      "grad_norm": 0.1591796875,
+      "learning_rate": 0.02,
+      "loss": 1.676,
+      "step": 2240
+    },
+    {
+      "epoch": 0.20746398194757854,
+      "grad_norm": 0.1728515625,
+      "learning_rate": 0.02,
+      "loss": 1.6498,
+      "step": 2241
+    },
+    {
+      "epoch": 0.2075565584678586,
+      "grad_norm": 0.1689453125,
+      "learning_rate": 0.02,
+      "loss": 1.6584,
+      "step": 2242
+    },
+    {
+      "epoch": 0.20764913498813864,
+      "grad_norm": 0.1640625,
+      "learning_rate": 0.02,
+      "loss": 1.6253,
+      "step": 2243
+    },
+    {
+      "epoch": 0.20774171150841866,
+      "grad_norm": 0.154296875,
+      "learning_rate": 0.02,
+      "loss": 1.6988,
+      "step": 2244
+    },
+    {
+      "epoch": 0.20783428802869872,
+      "grad_norm": 0.1572265625,
+      "learning_rate": 0.02,
+      "loss": 1.6934,
+      "step": 2245
+    },
+    {
+      "epoch": 0.20792686454897877,
+      "grad_norm": 0.1552734375,
+      "learning_rate": 0.02,
+      "loss": 1.6959,
+      "step": 2246
+    },
+    {
+      "epoch": 0.20801944106925882,
+      "grad_norm": 0.1630859375,
+      "learning_rate": 0.02,
+      "loss": 1.7298,
+      "step": 2247
+    },
+    {
+      "epoch": 0.20811201758953884,
+      "grad_norm": 0.1591796875,
+      "learning_rate": 0.02,
+      "loss": 1.7086,
+      "step": 2248
+    },
+    {
+      "epoch": 0.2082045941098189,
+      "grad_norm": 0.16015625,
+      "learning_rate": 0.02,
+      "loss": 1.6933,
+      "step": 2249
+    },
+    {
+      "epoch": 0.20829717063009895,
+      "grad_norm": 0.1484375,
+      "learning_rate": 0.02,
+      "loss": 1.6184,
+      "step": 2250
+    },
+    {
+      "epoch": 0.20838974715037897,
+      "grad_norm": 0.1474609375,
+      "learning_rate": 0.02,
+      "loss": 1.6969,
+      "step": 2251
+    },
+    {
+      "epoch": 0.20848232367065903,
+      "grad_norm": 0.1552734375,
+      "learning_rate": 0.02,
+      "loss": 1.7096,
+      "step": 2252
+    },
+    {
+      "epoch": 0.20857490019093908,
+      "grad_norm": 0.1513671875,
+      "learning_rate": 0.02,
+      "loss": 1.6467,
+      "step": 2253
+    },
+    {
+      "epoch": 0.20866747671121913,
+      "grad_norm": 0.1494140625,
+      "learning_rate": 0.02,
+      "loss": 1.6679,
+      "step": 2254
+    },
+    {
+      "epoch": 0.20876005323149915,
+      "grad_norm": 0.1494140625,
+      "learning_rate": 0.02,
+      "loss": 1.742,
+      "step": 2255
+    },
+    {
+      "epoch": 0.2088526297517792,
+      "grad_norm": 0.1474609375,
+      "learning_rate": 0.02,
+      "loss": 1.6675,
+      "step": 2256
+    },
+    {
+      "epoch": 0.20894520627205926,
+      "grad_norm": 0.1611328125,
+      "learning_rate": 0.02,
+      "loss": 1.7369,
+      "step": 2257
+    },
+    {
+      "epoch": 0.20903778279233928,
+      "grad_norm": 0.1689453125,
+      "learning_rate": 0.02,
+      "loss": 1.7149,
+      "step": 2258
+    },
+    {
+      "epoch": 0.20913035931261933,
+      "grad_norm": 0.1552734375,
+      "learning_rate": 0.02,
+      "loss": 1.6748,
+      "step": 2259
+    },
+    {
+      "epoch": 0.2092229358328994,
+      "grad_norm": 0.154296875,
+      "learning_rate": 0.02,
+      "loss": 1.6358,
+      "step": 2260
+    },
+    {
+      "epoch": 0.2093155123531794,
+      "grad_norm": 0.14453125,
+      "learning_rate": 0.02,
+      "loss": 1.6819,
+      "step": 2261
+    },
+    {
+      "epoch": 0.20940808887345946,
+      "grad_norm": 0.1494140625,
+      "learning_rate": 0.02,
+      "loss": 1.7656,
+      "step": 2262
+    },
+    {
+      "epoch": 0.20950066539373952,
+      "grad_norm": 0.1533203125,
+      "learning_rate": 0.02,
+      "loss": 1.7314,
+      "step": 2263
+    },
+    {
+      "epoch": 0.20959324191401957,
+      "grad_norm": 0.14453125,
+      "learning_rate": 0.02,
+      "loss": 1.6883,
+      "step": 2264
+    },
+    {
+      "epoch": 0.2096858184342996,
+      "grad_norm": 0.1416015625,
+      "learning_rate": 0.02,
+      "loss": 1.6098,
+      "step": 2265
+    },
+    {
+      "epoch": 0.20977839495457964,
+      "grad_norm": 0.1484375,
+      "learning_rate": 0.02,
+      "loss": 1.6629,
+      "step": 2266
+    },
+    {
+      "epoch": 0.2098709714748597,
+      "grad_norm": 0.1630859375,
+      "learning_rate": 0.02,
+      "loss": 1.6758,
+      "step": 2267
+    },
+    {
+      "epoch": 0.20996354799513972,
+      "grad_norm": 0.14453125,
+      "learning_rate": 0.02,
+      "loss": 1.6963,
+      "step": 2268
+    },
+    {
+      "epoch": 0.21005612451541977,
+      "grad_norm": 0.1435546875,
+      "learning_rate": 0.02,
+      "loss": 1.6706,
+      "step": 2269
+    },
+    {
+      "epoch": 0.21014870103569983,
+      "grad_norm": 0.13671875,
+      "learning_rate": 0.02,
+      "loss": 1.65,
+      "step": 2270
+    },
+    {
+      "epoch": 0.21024127755597988,
+      "grad_norm": 0.1552734375,
+      "learning_rate": 0.02,
+      "loss": 1.6881,
+      "step": 2271
+    },
+    {
+      "epoch": 0.2103338540762599,
+      "grad_norm": 0.15234375,
+      "learning_rate": 0.02,
+      "loss": 1.707,
+      "step": 2272
+    },
+    {
+      "epoch": 0.21042643059653995,
+      "grad_norm": 0.1513671875,
+      "learning_rate": 0.02,
+      "loss": 1.667,
+      "step": 2273
+    },
+    {
+      "epoch": 0.21051900711682,
+      "grad_norm": 0.1630859375,
+      "learning_rate": 0.02,
+      "loss": 1.707,
+      "step": 2274
+    },
+    {
+      "epoch": 0.21061158363710003,
+      "grad_norm": 0.162109375,
+      "learning_rate": 0.02,
+      "loss": 1.6709,
+      "step": 2275
+    },
+    {
+      "epoch": 0.21070416015738008,
+      "grad_norm": 0.1533203125,
+      "learning_rate": 0.02,
+      "loss": 1.6751,
+      "step": 2276
+    },
+    {
+      "epoch": 0.21079673667766013,
+      "grad_norm": 0.1591796875,
+      "learning_rate": 0.02,
+      "loss": 1.7352,
+      "step": 2277
+    },
+    {
+      "epoch": 0.21088931319794016,
+      "grad_norm": 0.15625,
+      "learning_rate": 0.02,
+      "loss": 1.7086,
+      "step": 2278
+    },
+    {
+      "epoch": 0.2109818897182202,
+      "grad_norm": 0.1533203125,
+      "learning_rate": 0.02,
+      "loss": 1.7229,
+      "step": 2279
+    },
+    {
+      "epoch": 0.21107446623850026,
+      "grad_norm": 0.1474609375,
+      "learning_rate": 0.02,
+      "loss": 1.6865,
+      "step": 2280
+    },
+    {
+      "epoch": 0.21116704275878032,
+      "grad_norm": 0.1494140625,
+      "learning_rate": 0.02,
+      "loss": 1.7141,
+      "step": 2281
+    },
+    {
+      "epoch": 0.21125961927906034,
+      "grad_norm": 0.146484375,
+      "learning_rate": 0.02,
+      "loss": 1.6777,
+      "step": 2282
+    },
+    {
+      "epoch": 0.2113521957993404,
+      "grad_norm": 0.140625,
+      "learning_rate": 0.02,
+      "loss": 1.638,
+      "step": 2283
+    },
+    {
+      "epoch": 0.21144477231962044,
+      "grad_norm": 0.15234375,
+      "learning_rate": 0.02,
+      "loss": 1.715,
+      "step": 2284
+    },
+    {
+      "epoch": 0.21153734883990047,
+      "grad_norm": 0.1591796875,
+      "learning_rate": 0.02,
+      "loss": 1.7129,
+      "step": 2285
+    },
+    {
+      "epoch": 0.21162992536018052,
+      "grad_norm": 0.1513671875,
+      "learning_rate": 0.02,
+      "loss": 1.7031,
+      "step": 2286
+    },
+    {
+      "epoch": 0.21172250188046057,
+      "grad_norm": 0.1640625,
+      "learning_rate": 0.02,
+      "loss": 1.662,
+      "step": 2287
+    },
+    {
+      "epoch": 0.21181507840074063,
+      "grad_norm": 0.169921875,
+      "learning_rate": 0.02,
+      "loss": 1.735,
+      "step": 2288
+    },
+    {
+      "epoch": 0.21190765492102065,
+      "grad_norm": 0.1669921875,
+      "learning_rate": 0.02,
+      "loss": 1.7036,
+      "step": 2289
+    },
+    {
+      "epoch": 0.2120002314413007,
+      "grad_norm": 0.1533203125,
+      "learning_rate": 0.02,
+      "loss": 1.6343,
+      "step": 2290
+    },
+    {
+      "epoch": 0.21209280796158075,
+      "grad_norm": 0.1552734375,
+      "learning_rate": 0.02,
+      "loss": 1.6716,
+      "step": 2291
+    },
+    {
+      "epoch": 0.21218538448186078,
+      "grad_norm": 0.166015625,
+      "learning_rate": 0.02,
+      "loss": 1.6504,
+      "step": 2292
+    },
+    {
+      "epoch": 0.21227796100214083,
+      "grad_norm": 0.1474609375,
+      "learning_rate": 0.02,
+      "loss": 1.7147,
+      "step": 2293
+    },
+    {
+      "epoch": 0.21237053752242088,
+      "grad_norm": 0.1591796875,
+      "learning_rate": 0.02,
+      "loss": 1.7561,
+      "step": 2294
+    },
+    {
+      "epoch": 0.2124631140427009,
+      "grad_norm": 0.15625,
+      "learning_rate": 0.02,
+      "loss": 1.7232,
+      "step": 2295
+    },
+    {
+      "epoch": 0.21255569056298096,
+      "grad_norm": 0.1513671875,
+      "learning_rate": 0.02,
+      "loss": 1.7352,
+      "step": 2296
+    },
+    {
+      "epoch": 0.212648267083261,
+      "grad_norm": 0.146484375,
+      "learning_rate": 0.02,
+      "loss": 1.6457,
+      "step": 2297
+    },
+    {
+      "epoch": 0.21274084360354106,
+      "grad_norm": 0.15234375,
+      "learning_rate": 0.02,
+      "loss": 1.7242,
+      "step": 2298
+    },
+    {
+      "epoch": 0.2128334201238211,
+      "grad_norm": 0.1591796875,
+      "learning_rate": 0.02,
+      "loss": 1.7565,
+      "step": 2299
+    },
+    {
+      "epoch": 0.21292599664410114,
+      "grad_norm": 0.1611328125,
+      "learning_rate": 0.02,
+      "loss": 1.7135,
+      "step": 2300
+    },
+    {
+      "epoch": 0.2130185731643812,
+      "grad_norm": 0.15234375,
+      "learning_rate": 0.02,
+      "loss": 1.6772,
+      "step": 2301
+    },
+    {
+      "epoch": 0.21311114968466122,
+      "grad_norm": 0.1591796875,
+      "learning_rate": 0.02,
+      "loss": 1.7056,
+      "step": 2302
+    },
+    {
+      "epoch": 0.21320372620494127,
+      "grad_norm": 0.154296875,
+      "learning_rate": 0.02,
+      "loss": 1.7291,
+      "step": 2303
+    },
+    {
+      "epoch": 0.21329630272522132,
+      "grad_norm": 0.1572265625,
+      "learning_rate": 0.02,
+      "loss": 1.72,
+      "step": 2304
+    },
+    {
+      "epoch": 0.21338887924550137,
+      "grad_norm": 0.150390625,
+      "learning_rate": 0.02,
+      "loss": 1.6572,
+      "step": 2305
+    },
+    {
+      "epoch": 0.2134814557657814,
+      "grad_norm": 0.1611328125,
+      "learning_rate": 0.02,
+      "loss": 1.6714,
+      "step": 2306
+    },
+    {
+      "epoch": 0.21357403228606145,
+      "grad_norm": 0.162109375,
+      "learning_rate": 0.02,
+      "loss": 1.6786,
+      "step": 2307
+    },
+    {
+      "epoch": 0.2136666088063415,
+      "grad_norm": 0.166015625,
+      "learning_rate": 0.02,
+      "loss": 1.6361,
+      "step": 2308
+    },
+    {
+      "epoch": 0.21375918532662153,
+      "grad_norm": 0.1630859375,
+      "learning_rate": 0.02,
+      "loss": 1.6881,
+      "step": 2309
+    },
+    {
+      "epoch": 0.21385176184690158,
+      "grad_norm": 0.1669921875,
+      "learning_rate": 0.02,
+      "loss": 1.6372,
+      "step": 2310
+    },
+    {
+      "epoch": 0.21394433836718163,
+      "grad_norm": 0.1669921875,
+      "learning_rate": 0.02,
+      "loss": 1.6519,
+      "step": 2311
+    },
+    {
+      "epoch": 0.21403691488746165,
+      "grad_norm": 0.15234375,
+      "learning_rate": 0.02,
+      "loss": 1.6172,
+      "step": 2312
+    },
+    {
+      "epoch": 0.2141294914077417,
+      "grad_norm": 0.162109375,
+      "learning_rate": 0.02,
+      "loss": 1.7293,
+      "step": 2313
+    },
+    {
+      "epoch": 0.21422206792802176,
+      "grad_norm": 0.1591796875,
+      "learning_rate": 0.02,
+      "loss": 1.7064,
+      "step": 2314
+    },
+    {
+      "epoch": 0.2143146444483018,
+      "grad_norm": 0.154296875,
+      "learning_rate": 0.02,
+      "loss": 1.6752,
+      "step": 2315
+    },
+    {
+      "epoch": 0.21440722096858184,
+      "grad_norm": 0.1572265625,
+      "learning_rate": 0.02,
+      "loss": 1.7071,
+      "step": 2316
+    },
+    {
+      "epoch": 0.2144997974888619,
+      "grad_norm": 0.15625,
+      "learning_rate": 0.02,
+      "loss": 1.6869,
+      "step": 2317
+    },
+    {
+      "epoch": 0.21459237400914194,
+      "grad_norm": 0.1611328125,
+      "learning_rate": 0.02,
+      "loss": 1.6465,
+      "step": 2318
+    },
+    {
+      "epoch": 0.21468495052942196,
+      "grad_norm": 0.1533203125,
+      "learning_rate": 0.02,
+      "loss": 1.6479,
+      "step": 2319
+    },
+    {
+      "epoch": 0.21477752704970202,
+      "grad_norm": 0.1533203125,
+      "learning_rate": 0.02,
+      "loss": 1.6983,
+      "step": 2320
+    },
+    {
+      "epoch": 0.21487010356998207,
+      "grad_norm": 0.1494140625,
+      "learning_rate": 0.02,
+      "loss": 1.6588,
+      "step": 2321
+    },
+    {
+      "epoch": 0.21496268009026212,
+      "grad_norm": 0.1640625,
+      "learning_rate": 0.02,
+      "loss": 1.7237,
+      "step": 2322
+    },
+    {
+      "epoch": 0.21505525661054214,
+      "grad_norm": 0.150390625,
+      "learning_rate": 0.02,
+      "loss": 1.6046,
+      "step": 2323
+    },
+    {
+      "epoch": 0.2151478331308222,
+      "grad_norm": 0.162109375,
+      "learning_rate": 0.02,
+      "loss": 1.6883,
+      "step": 2324
+    },
+    {
+      "epoch": 0.21524040965110225,
+      "grad_norm": 0.15625,
+      "learning_rate": 0.02,
+      "loss": 1.7044,
+      "step": 2325
+    },
+    {
+      "epoch": 0.21533298617138227,
+      "grad_norm": 0.1552734375,
+      "learning_rate": 0.02,
+      "loss": 1.6342,
+      "step": 2326
+    },
+    {
+      "epoch": 0.21542556269166233,
+      "grad_norm": 0.158203125,
+      "learning_rate": 0.02,
+      "loss": 1.6834,
+      "step": 2327
+    },
+    {
+      "epoch": 0.21551813921194238,
+      "grad_norm": 0.166015625,
+      "learning_rate": 0.02,
+      "loss": 1.687,
+      "step": 2328
+    },
+    {
+      "epoch": 0.2156107157322224,
+      "grad_norm": 0.1650390625,
+      "learning_rate": 0.02,
+      "loss": 1.6894,
+      "step": 2329
+    },
+    {
+      "epoch": 0.21570329225250245,
+      "grad_norm": 0.158203125,
+      "learning_rate": 0.02,
+      "loss": 1.6541,
+      "step": 2330
+    },
+    {
+      "epoch": 0.2157958687727825,
+      "grad_norm": 0.1591796875,
+      "learning_rate": 0.02,
+      "loss": 1.622,
+      "step": 2331
+    },
+    {
+      "epoch": 0.21588844529306256,
+      "grad_norm": 0.1591796875,
+      "learning_rate": 0.02,
+      "loss": 1.7154,
+      "step": 2332
+    },
+    {
+      "epoch": 0.21598102181334258,
+      "grad_norm": 0.1640625,
+      "learning_rate": 0.02,
+      "loss": 1.7682,
+      "step": 2333
+    },
+    {
+      "epoch": 0.21607359833362264,
+      "grad_norm": 0.15625,
+      "learning_rate": 0.02,
+      "loss": 1.7001,
+      "step": 2334
+    },
+    {
+      "epoch": 0.2161661748539027,
+      "grad_norm": 0.1484375,
+      "learning_rate": 0.02,
+      "loss": 1.6272,
+      "step": 2335
+    },
+    {
+      "epoch": 0.2162587513741827,
+      "grad_norm": 0.1435546875,
+      "learning_rate": 0.02,
+      "loss": 1.604,
+      "step": 2336
+    },
+    {
+      "epoch": 0.21635132789446276,
+      "grad_norm": 0.1494140625,
+      "learning_rate": 0.02,
+      "loss": 1.7081,
+      "step": 2337
+    },
+    {
+      "epoch": 0.21644390441474282,
+      "grad_norm": 0.146484375,
+      "learning_rate": 0.02,
+      "loss": 1.6456,
+      "step": 2338
+    },
+    {
+      "epoch": 0.21653648093502287,
+      "grad_norm": 0.15234375,
+      "learning_rate": 0.02,
+      "loss": 1.7205,
+      "step": 2339
+    },
+    {
+      "epoch": 0.2166290574553029,
+      "grad_norm": 0.138671875,
+      "learning_rate": 0.02,
+      "loss": 1.6662,
+      "step": 2340
+    },
+    {
+      "epoch": 0.21672163397558294,
+      "grad_norm": 0.14453125,
+      "learning_rate": 0.02,
+      "loss": 1.6588,
+      "step": 2341
+    },
+    {
+      "epoch": 0.216814210495863,
+      "grad_norm": 0.1474609375,
+      "learning_rate": 0.02,
+      "loss": 1.6844,
+      "step": 2342
+    },
+    {
+      "epoch": 0.21690678701614302,
+      "grad_norm": 0.1494140625,
+      "learning_rate": 0.02,
+      "loss": 1.7136,
+      "step": 2343
+    },
+    {
+      "epoch": 0.21699936353642307,
+      "grad_norm": 0.1435546875,
+      "learning_rate": 0.02,
+      "loss": 1.6815,
+      "step": 2344
+    },
+    {
+      "epoch": 0.21709194005670313,
+      "grad_norm": 0.1474609375,
+      "learning_rate": 0.02,
+      "loss": 1.7222,
+      "step": 2345
+    },
+    {
+      "epoch": 0.21718451657698315,
+      "grad_norm": 0.1484375,
+      "learning_rate": 0.02,
+      "loss": 1.6899,
+      "step": 2346
+    },
+    {
+      "epoch": 0.2172770930972632,
+      "grad_norm": 0.1435546875,
+      "learning_rate": 0.02,
+      "loss": 1.7076,
+      "step": 2347
+    },
+    {
+      "epoch": 0.21736966961754325,
+      "grad_norm": 0.1376953125,
+      "learning_rate": 0.02,
+      "loss": 1.7344,
+      "step": 2348
+    },
+    {
+      "epoch": 0.2174622461378233,
+      "grad_norm": 0.140625,
+      "learning_rate": 0.02,
+      "loss": 1.6791,
+      "step": 2349
+    },
+    {
+      "epoch": 0.21755482265810333,
+      "grad_norm": 0.1435546875,
+      "learning_rate": 0.02,
+      "loss": 1.6833,
+      "step": 2350
+    },
+    {
+      "epoch": 0.21764739917838338,
+      "grad_norm": 0.142578125,
+      "learning_rate": 0.02,
+      "loss": 1.7055,
+      "step": 2351
+    },
+    {
+      "epoch": 0.21773997569866343,
+      "grad_norm": 0.1455078125,
+      "learning_rate": 0.02,
+      "loss": 1.689,
+      "step": 2352
+    },
+    {
+      "epoch": 0.21783255221894346,
+      "grad_norm": 0.1337890625,
+      "learning_rate": 0.02,
+      "loss": 1.641,
+      "step": 2353
+    },
+    {
+      "epoch": 0.2179251287392235,
+      "grad_norm": 0.1435546875,
+      "learning_rate": 0.02,
+      "loss": 1.6432,
+      "step": 2354
+    },
+    {
+      "epoch": 0.21801770525950356,
+      "grad_norm": 0.1396484375,
+      "learning_rate": 0.02,
+      "loss": 1.683,
+      "step": 2355
+    },
+    {
+      "epoch": 0.21811028177978362,
+      "grad_norm": 0.1416015625,
+      "learning_rate": 0.02,
+      "loss": 1.6819,
+      "step": 2356
+    },
+    {
+      "epoch": 0.21820285830006364,
+      "grad_norm": 0.1533203125,
+      "learning_rate": 0.02,
+      "loss": 1.7525,
+      "step": 2357
+    },
+    {
+      "epoch": 0.2182954348203437,
+      "grad_norm": 0.1533203125,
+      "learning_rate": 0.02,
+      "loss": 1.6519,
+      "step": 2358
+    },
+    {
+      "epoch": 0.21838801134062374,
+      "grad_norm": 0.1337890625,
+      "learning_rate": 0.02,
+      "loss": 1.5901,
+      "step": 2359
+    },
+    {
+      "epoch": 0.21848058786090377,
+      "grad_norm": 0.154296875,
+      "learning_rate": 0.02,
+      "loss": 1.6748,
+      "step": 2360
+    },
+    {
+      "epoch": 0.21857316438118382,
+      "grad_norm": 0.15234375,
+      "learning_rate": 0.02,
+      "loss": 1.669,
+      "step": 2361
+    },
+    {
+      "epoch": 0.21866574090146387,
+      "grad_norm": 0.1513671875,
+      "learning_rate": 0.02,
+      "loss": 1.6975,
+      "step": 2362
+    },
+    {
+      "epoch": 0.2187583174217439,
+      "grad_norm": 0.142578125,
+      "learning_rate": 0.02,
+      "loss": 1.6691,
+      "step": 2363
+    },
+    {
+      "epoch": 0.21885089394202395,
+      "grad_norm": 0.1552734375,
+      "learning_rate": 0.02,
+      "loss": 1.718,
+      "step": 2364
+    },
+    {
+      "epoch": 0.218943470462304,
+      "grad_norm": 0.146484375,
+      "learning_rate": 0.02,
+      "loss": 1.5778,
+      "step": 2365
+    },
+    {
+      "epoch": 0.21903604698258405,
+      "grad_norm": 0.158203125,
+      "learning_rate": 0.02,
+      "loss": 1.6972,
+      "step": 2366
+    },
+    {
+      "epoch": 0.21912862350286408,
+      "grad_norm": 0.138671875,
+      "learning_rate": 0.02,
+      "loss": 1.6269,
+      "step": 2367
+    },
+    {
+      "epoch": 0.21922120002314413,
+      "grad_norm": 0.1484375,
+      "learning_rate": 0.02,
+      "loss": 1.6289,
+      "step": 2368
+    },
+    {
+      "epoch": 0.21931377654342418,
+      "grad_norm": 0.142578125,
+      "learning_rate": 0.02,
+      "loss": 1.5948,
+      "step": 2369
+    },
+    {
+      "epoch": 0.2194063530637042,
+      "grad_norm": 0.1552734375,
+      "learning_rate": 0.02,
+      "loss": 1.7188,
+      "step": 2370
+    },
+    {
+      "epoch": 0.21949892958398426,
+      "grad_norm": 0.15234375,
+      "learning_rate": 0.02,
+      "loss": 1.613,
+      "step": 2371
+    },
+    {
+      "epoch": 0.2195915061042643,
+      "grad_norm": 0.1552734375,
+      "learning_rate": 0.02,
+      "loss": 1.6879,
+      "step": 2372
+    },
+    {
+      "epoch": 0.21968408262454436,
+      "grad_norm": 0.15625,
+      "learning_rate": 0.02,
+      "loss": 1.675,
+      "step": 2373
+    },
+    {
+      "epoch": 0.2197766591448244,
+      "grad_norm": 0.1552734375,
+      "learning_rate": 0.02,
+      "loss": 1.6235,
+      "step": 2374
+    },
+    {
+      "epoch": 0.21986923566510444,
+      "grad_norm": 0.15625,
+      "learning_rate": 0.02,
+      "loss": 1.6524,
+      "step": 2375
+    },
+    {
+      "epoch": 0.2199618121853845,
+      "grad_norm": 0.173828125,
+      "learning_rate": 0.02,
+      "loss": 1.6443,
+      "step": 2376
+    },
+    {
+      "epoch": 0.22005438870566452,
+      "grad_norm": 0.171875,
+      "learning_rate": 0.02,
+      "loss": 1.6429,
+      "step": 2377
+    },
+    {
+      "epoch": 0.22014696522594457,
+      "grad_norm": 0.166015625,
+      "learning_rate": 0.02,
+      "loss": 1.6696,
+      "step": 2378
+    },
+    {
+      "epoch": 0.22023954174622462,
+      "grad_norm": 0.16015625,
+      "learning_rate": 0.02,
+      "loss": 1.6902,
+      "step": 2379
+    },
+    {
+      "epoch": 0.22033211826650465,
+      "grad_norm": 0.166015625,
+      "learning_rate": 0.02,
+      "loss": 1.7101,
+      "step": 2380
+    },
+    {
+      "epoch": 0.2204246947867847,
+      "grad_norm": 0.1552734375,
+      "learning_rate": 0.02,
+      "loss": 1.7581,
+      "step": 2381
+    },
+    {
+      "epoch": 0.22051727130706475,
+      "grad_norm": 0.15234375,
+      "learning_rate": 0.02,
+      "loss": 1.6344,
+      "step": 2382
+    },
+    {
+      "epoch": 0.2206098478273448,
+      "grad_norm": 0.15234375,
+      "learning_rate": 0.02,
+      "loss": 1.663,
+      "step": 2383
+    },
+    {
+      "epoch": 0.22070242434762483,
+      "grad_norm": 0.150390625,
+      "learning_rate": 0.02,
+      "loss": 1.6737,
+      "step": 2384
+    },
+    {
+      "epoch": 0.22079500086790488,
+      "grad_norm": 0.1435546875,
+      "learning_rate": 0.02,
+      "loss": 1.7072,
+      "step": 2385
+    },
+    {
+      "epoch": 0.22088757738818493,
+      "grad_norm": 0.1455078125,
+      "learning_rate": 0.02,
+      "loss": 1.6284,
+      "step": 2386
+    },
+    {
+      "epoch": 0.22098015390846495,
+      "grad_norm": 0.142578125,
+      "learning_rate": 0.02,
+      "loss": 1.6288,
+      "step": 2387
+    },
+    {
+      "epoch": 0.221072730428745,
+      "grad_norm": 0.146484375,
+      "learning_rate": 0.02,
+      "loss": 1.6524,
+      "step": 2388
+    },
+    {
+      "epoch": 0.22116530694902506,
+      "grad_norm": 0.150390625,
+      "learning_rate": 0.02,
+      "loss": 1.6679,
+      "step": 2389
+    },
+    {
+      "epoch": 0.2212578834693051,
+      "grad_norm": 0.1484375,
+      "learning_rate": 0.02,
+      "loss": 1.6388,
+      "step": 2390
+    },
+    {
+      "epoch": 0.22135045998958514,
+      "grad_norm": 0.14453125,
+      "learning_rate": 0.02,
+      "loss": 1.6882,
+      "step": 2391
+    },
+    {
+      "epoch": 0.2214430365098652,
+      "grad_norm": 0.15234375,
+      "learning_rate": 0.02,
+      "loss": 1.6947,
+      "step": 2392
+    },
+    {
+      "epoch": 0.22153561303014524,
+      "grad_norm": 0.154296875,
+      "learning_rate": 0.02,
+      "loss": 1.6905,
+      "step": 2393
+    },
+    {
+      "epoch": 0.22162818955042526,
+      "grad_norm": 0.142578125,
+      "learning_rate": 0.02,
+      "loss": 1.6473,
+      "step": 2394
+    },
+    {
+      "epoch": 0.22172076607070532,
+      "grad_norm": 0.146484375,
+      "learning_rate": 0.02,
+      "loss": 1.6751,
+      "step": 2395
+    },
+    {
+      "epoch": 0.22181334259098537,
+      "grad_norm": 0.1484375,
+      "learning_rate": 0.02,
+      "loss": 1.6913,
+      "step": 2396
+    },
+    {
+      "epoch": 0.2219059191112654,
+      "grad_norm": 0.1533203125,
+      "learning_rate": 0.02,
+      "loss": 1.6832,
+      "step": 2397
+    },
+    {
+      "epoch": 0.22199849563154544,
+      "grad_norm": 0.1455078125,
+      "learning_rate": 0.02,
+      "loss": 1.6759,
+      "step": 2398
+    },
+    {
+      "epoch": 0.2220910721518255,
+      "grad_norm": 0.1650390625,
+      "learning_rate": 0.02,
+      "loss": 1.679,
+      "step": 2399
+    },
+    {
+      "epoch": 0.22218364867210555,
+      "grad_norm": 0.166015625,
+      "learning_rate": 0.02,
+      "loss": 1.7043,
+      "step": 2400
+    },
+    {
+      "epoch": 0.22227622519238557,
+      "grad_norm": 0.1572265625,
+      "learning_rate": 0.02,
+      "loss": 1.6588,
+      "step": 2401
+    },
+    {
+      "epoch": 0.22236880171266563,
+      "grad_norm": 0.1591796875,
+      "learning_rate": 0.02,
+      "loss": 1.6382,
+      "step": 2402
+    },
+    {
+      "epoch": 0.22246137823294568,
+      "grad_norm": 0.1640625,
+      "learning_rate": 0.02,
+      "loss": 1.7331,
+      "step": 2403
+    },
+    {
+      "epoch": 0.2225539547532257,
+      "grad_norm": 0.166015625,
+      "learning_rate": 0.02,
+      "loss": 1.7345,
+      "step": 2404
+    },
+    {
+      "epoch": 0.22264653127350575,
+      "grad_norm": 0.171875,
+      "learning_rate": 0.02,
+      "loss": 1.711,
+      "step": 2405
+    },
+    {
+      "epoch": 0.2227391077937858,
+      "grad_norm": 0.16796875,
+      "learning_rate": 0.02,
+      "loss": 1.723,
+      "step": 2406
+    },
+    {
+      "epoch": 0.22283168431406586,
+      "grad_norm": 0.1650390625,
+      "learning_rate": 0.02,
+      "loss": 1.6605,
+      "step": 2407
+    },
+    {
+      "epoch": 0.22292426083434588,
+      "grad_norm": 0.166015625,
+      "learning_rate": 0.02,
+      "loss": 1.7576,
+      "step": 2408
+    },
+    {
+      "epoch": 0.22301683735462594,
+      "grad_norm": 0.1552734375,
+      "learning_rate": 0.02,
+      "loss": 1.6733,
+      "step": 2409
+    },
+    {
+      "epoch": 0.223109413874906,
+      "grad_norm": 0.150390625,
+      "learning_rate": 0.02,
+      "loss": 1.688,
+      "step": 2410
+    },
+    {
+      "epoch": 0.223201990395186,
+      "grad_norm": 0.166015625,
+      "learning_rate": 0.02,
+      "loss": 1.7182,
+      "step": 2411
+    },
+    {
+      "epoch": 0.22329456691546606,
+      "grad_norm": 0.171875,
+      "learning_rate": 0.02,
+      "loss": 1.724,
+      "step": 2412
+    },
+    {
+      "epoch": 0.22338714343574612,
+      "grad_norm": 0.15625,
+      "learning_rate": 0.02,
+      "loss": 1.761,
+      "step": 2413
+    },
+    {
+      "epoch": 0.22347971995602614,
+      "grad_norm": 0.166015625,
+      "learning_rate": 0.02,
+      "loss": 1.6487,
+      "step": 2414
+    },
+    {
+      "epoch": 0.2235722964763062,
+      "grad_norm": 0.154296875,
+      "learning_rate": 0.02,
+      "loss": 1.6773,
+      "step": 2415
+    },
+    {
+      "epoch": 0.22366487299658624,
+      "grad_norm": 0.1513671875,
+      "learning_rate": 0.02,
+      "loss": 1.6829,
+      "step": 2416
+    },
+    {
+      "epoch": 0.2237574495168663,
+      "grad_norm": 0.1533203125,
+      "learning_rate": 0.02,
+      "loss": 1.6672,
+      "step": 2417
+    },
+    {
+      "epoch": 0.22385002603714632,
+      "grad_norm": 0.15625,
+      "learning_rate": 0.02,
+      "loss": 1.6825,
+      "step": 2418
+    },
+    {
+      "epoch": 0.22394260255742637,
+      "grad_norm": 0.1533203125,
+      "learning_rate": 0.02,
+      "loss": 1.6647,
+      "step": 2419
+    },
+    {
+      "epoch": 0.22403517907770643,
+      "grad_norm": 0.166015625,
+      "learning_rate": 0.02,
+      "loss": 1.6496,
+      "step": 2420
+    },
+    {
+      "epoch": 0.22412775559798645,
+      "grad_norm": 0.1484375,
+      "learning_rate": 0.02,
+      "loss": 1.683,
+      "step": 2421
+    },
+    {
+      "epoch": 0.2242203321182665,
+      "grad_norm": 0.1572265625,
+      "learning_rate": 0.02,
+      "loss": 1.6988,
+      "step": 2422
+    },
+    {
+      "epoch": 0.22431290863854655,
+      "grad_norm": 0.15625,
+      "learning_rate": 0.02,
+      "loss": 1.643,
+      "step": 2423
+    },
+    {
+      "epoch": 0.2244054851588266,
+      "grad_norm": 0.162109375,
+      "learning_rate": 0.02,
+      "loss": 1.7374,
+      "step": 2424
+    },
+    {
+      "epoch": 0.22449806167910663,
+      "grad_norm": 0.1533203125,
+      "learning_rate": 0.02,
+      "loss": 1.7249,
+      "step": 2425
+    },
+    {
+      "epoch": 0.22459063819938668,
+      "grad_norm": 0.15625,
+      "learning_rate": 0.02,
+      "loss": 1.6691,
+      "step": 2426
+    },
+    {
+      "epoch": 0.22468321471966674,
+      "grad_norm": 0.154296875,
+      "learning_rate": 0.02,
+      "loss": 1.6134,
+      "step": 2427
+    },
+    {
+      "epoch": 0.22477579123994676,
+      "grad_norm": 0.150390625,
+      "learning_rate": 0.02,
+      "loss": 1.6605,
+      "step": 2428
+    },
+    {
+      "epoch": 0.2248683677602268,
+      "grad_norm": 0.1494140625,
+      "learning_rate": 0.02,
+      "loss": 1.6566,
+      "step": 2429
+    },
+    {
+      "epoch": 0.22496094428050686,
+      "grad_norm": 0.14453125,
+      "learning_rate": 0.02,
+      "loss": 1.63,
+      "step": 2430
+    },
+    {
+      "epoch": 0.2250535208007869,
+      "grad_norm": 0.16015625,
+      "learning_rate": 0.02,
+      "loss": 1.6774,
+      "step": 2431
+    },
+    {
+      "epoch": 0.22514609732106694,
+      "grad_norm": 0.1474609375,
+      "learning_rate": 0.02,
+      "loss": 1.615,
+      "step": 2432
+    },
+    {
+      "epoch": 0.225238673841347,
+      "grad_norm": 0.1552734375,
+      "learning_rate": 0.02,
+      "loss": 1.7399,
+      "step": 2433
+    },
+    {
+      "epoch": 0.22533125036162704,
+      "grad_norm": 0.140625,
+      "learning_rate": 0.02,
+      "loss": 1.6869,
+      "step": 2434
+    },
+    {
+      "epoch": 0.22542382688190707,
+      "grad_norm": 0.1494140625,
+      "learning_rate": 0.02,
+      "loss": 1.6816,
+      "step": 2435
+    },
+    {
+      "epoch": 0.22551640340218712,
+      "grad_norm": 0.15625,
+      "learning_rate": 0.02,
+      "loss": 1.5897,
+      "step": 2436
+    },
+    {
+      "epoch": 0.22560897992246717,
+      "grad_norm": 0.15234375,
+      "learning_rate": 0.02,
+      "loss": 1.7538,
+      "step": 2437
+    },
+    {
+      "epoch": 0.2257015564427472,
+      "grad_norm": 0.154296875,
+      "learning_rate": 0.02,
+      "loss": 1.7187,
+      "step": 2438
+    },
+    {
+      "epoch": 0.22579413296302725,
+      "grad_norm": 0.1552734375,
+      "learning_rate": 0.02,
+      "loss": 1.7201,
+      "step": 2439
+    },
+    {
+      "epoch": 0.2258867094833073,
+      "grad_norm": 0.1494140625,
+      "learning_rate": 0.02,
+      "loss": 1.644,
+      "step": 2440
+    },
+    {
+      "epoch": 0.22597928600358733,
+      "grad_norm": 0.1484375,
+      "learning_rate": 0.02,
+      "loss": 1.6601,
+      "step": 2441
+    },
+    {
+      "epoch": 0.22607186252386738,
+      "grad_norm": 0.142578125,
+      "learning_rate": 0.02,
+      "loss": 1.6402,
+      "step": 2442
+    },
+    {
+      "epoch": 0.22616443904414743,
+      "grad_norm": 0.1533203125,
+      "learning_rate": 0.02,
+      "loss": 1.6923,
+      "step": 2443
+    },
+    {
+      "epoch": 0.22625701556442748,
+      "grad_norm": 0.1650390625,
+      "learning_rate": 0.02,
+      "loss": 1.7219,
+      "step": 2444
+    },
+    {
+      "epoch": 0.2263495920847075,
+      "grad_norm": 0.1533203125,
+      "learning_rate": 0.02,
+      "loss": 1.6885,
+      "step": 2445
+    },
+    {
+      "epoch": 0.22644216860498756,
+      "grad_norm": 0.1484375,
+      "learning_rate": 0.02,
+      "loss": 1.714,
+      "step": 2446
+    },
+    {
+      "epoch": 0.2265347451252676,
+      "grad_norm": 0.16015625,
+      "learning_rate": 0.02,
+      "loss": 1.6631,
+      "step": 2447
+    },
+    {
+      "epoch": 0.22662732164554764,
+      "grad_norm": 0.16015625,
+      "learning_rate": 0.02,
+      "loss": 1.6479,
+      "step": 2448
+    },
+    {
+      "epoch": 0.2267198981658277,
+      "grad_norm": 0.150390625,
+      "learning_rate": 0.02,
+      "loss": 1.6514,
+      "step": 2449
+    },
+    {
+      "epoch": 0.22681247468610774,
+      "grad_norm": 0.150390625,
+      "learning_rate": 0.02,
+      "loss": 1.6651,
+      "step": 2450
+    },
+    {
+      "epoch": 0.2269050512063878,
+      "grad_norm": 0.1591796875,
+      "learning_rate": 0.02,
+      "loss": 1.6721,
+      "step": 2451
+    },
+    {
+      "epoch": 0.22699762772666782,
+      "grad_norm": 0.1552734375,
+      "learning_rate": 0.02,
+      "loss": 1.6544,
+      "step": 2452
+    },
+    {
+      "epoch": 0.22709020424694787,
+      "grad_norm": 0.1455078125,
+      "learning_rate": 0.02,
+      "loss": 1.6817,
+      "step": 2453
+    },
+    {
+      "epoch": 0.22718278076722792,
+      "grad_norm": 0.14453125,
+      "learning_rate": 0.02,
+      "loss": 1.6717,
+      "step": 2454
+    },
+    {
+      "epoch": 0.22727535728750795,
+      "grad_norm": 0.1552734375,
+      "learning_rate": 0.02,
+      "loss": 1.6491,
+      "step": 2455
+    },
+    {
+      "epoch": 0.227367933807788,
+      "grad_norm": 0.1494140625,
+      "learning_rate": 0.02,
+      "loss": 1.6324,
+      "step": 2456
+    },
+    {
+      "epoch": 0.22746051032806805,
+      "grad_norm": 0.140625,
+      "learning_rate": 0.02,
+      "loss": 1.6846,
+      "step": 2457
+    },
+    {
+      "epoch": 0.22755308684834807,
+      "grad_norm": 0.1455078125,
+      "learning_rate": 0.02,
+      "loss": 1.6214,
+      "step": 2458
+    },
+    {
+      "epoch": 0.22764566336862813,
+      "grad_norm": 0.1484375,
+      "learning_rate": 0.02,
+      "loss": 1.6544,
+      "step": 2459
+    },
+    {
+      "epoch": 0.22773823988890818,
+      "grad_norm": 0.1552734375,
+      "learning_rate": 0.02,
+      "loss": 1.6362,
+      "step": 2460
+    },
+    {
+      "epoch": 0.22783081640918823,
+      "grad_norm": 0.15234375,
+      "learning_rate": 0.02,
+      "loss": 1.6988,
+      "step": 2461
+    },
+    {
+      "epoch": 0.22792339292946825,
+      "grad_norm": 0.1484375,
+      "learning_rate": 0.02,
+      "loss": 1.7047,
+      "step": 2462
+    },
+    {
+      "epoch": 0.2280159694497483,
+      "grad_norm": 0.1494140625,
+      "learning_rate": 0.02,
+      "loss": 1.7028,
+      "step": 2463
+    },
+    {
+      "epoch": 0.22810854597002836,
+      "grad_norm": 0.1494140625,
+      "learning_rate": 0.02,
+      "loss": 1.6961,
+      "step": 2464
+    },
+    {
+      "epoch": 0.22820112249030838,
+      "grad_norm": 0.1533203125,
+      "learning_rate": 0.02,
+      "loss": 1.659,
+      "step": 2465
+    },
+    {
+      "epoch": 0.22829369901058844,
+      "grad_norm": 0.1591796875,
+      "learning_rate": 0.02,
+      "loss": 1.6269,
+      "step": 2466
+    },
+    {
+      "epoch": 0.2283862755308685,
+      "grad_norm": 0.146484375,
+      "learning_rate": 0.02,
+      "loss": 1.68,
+      "step": 2467
+    },
+    {
+      "epoch": 0.22847885205114854,
+      "grad_norm": 0.1494140625,
+      "learning_rate": 0.02,
+      "loss": 1.6858,
+      "step": 2468
+    },
+    {
+      "epoch": 0.22857142857142856,
+      "grad_norm": 0.1416015625,
+      "learning_rate": 0.02,
+      "loss": 1.6022,
+      "step": 2469
+    },
+    {
+      "epoch": 0.22866400509170862,
+      "grad_norm": 0.1494140625,
+      "learning_rate": 0.02,
+      "loss": 1.7036,
+      "step": 2470
+    },
+    {
+      "epoch": 0.22875658161198867,
+      "grad_norm": 0.1455078125,
+      "learning_rate": 0.02,
+      "loss": 1.644,
+      "step": 2471
+    },
+    {
+      "epoch": 0.2288491581322687,
+      "grad_norm": 0.1494140625,
+      "learning_rate": 0.02,
+      "loss": 1.6868,
+      "step": 2472
+    },
+    {
+      "epoch": 0.22894173465254875,
+      "grad_norm": 0.1494140625,
+      "learning_rate": 0.02,
+      "loss": 1.7016,
+      "step": 2473
+    },
+    {
+      "epoch": 0.2290343111728288,
+      "grad_norm": 0.1484375,
+      "learning_rate": 0.02,
+      "loss": 1.614,
+      "step": 2474
+    },
+    {
+      "epoch": 0.22912688769310882,
+      "grad_norm": 0.1474609375,
+      "learning_rate": 0.02,
+      "loss": 1.6701,
+      "step": 2475
+    },
+    {
+      "epoch": 0.22921946421338887,
+      "grad_norm": 0.1552734375,
+      "learning_rate": 0.02,
+      "loss": 1.6503,
+      "step": 2476
+    },
+    {
+      "epoch": 0.22931204073366893,
+      "grad_norm": 0.1611328125,
+      "learning_rate": 0.02,
+      "loss": 1.774,
+      "step": 2477
+    },
+    {
+      "epoch": 0.22940461725394898,
+      "grad_norm": 0.1513671875,
+      "learning_rate": 0.02,
+      "loss": 1.6693,
+      "step": 2478
+    },
+    {
+      "epoch": 0.229497193774229,
+      "grad_norm": 0.14453125,
+      "learning_rate": 0.02,
+      "loss": 1.6432,
+      "step": 2479
+    },
+    {
+      "epoch": 0.22958977029450905,
+      "grad_norm": 0.154296875,
+      "learning_rate": 0.02,
+      "loss": 1.7218,
+      "step": 2480
+    },
+    {
+      "epoch": 0.2296823468147891,
+      "grad_norm": 0.150390625,
+      "learning_rate": 0.02,
+      "loss": 1.628,
+      "step": 2481
+    },
+    {
+      "epoch": 0.22977492333506913,
+      "grad_norm": 0.1484375,
+      "learning_rate": 0.02,
+      "loss": 1.653,
+      "step": 2482
+    },
+    {
+      "epoch": 0.22986749985534918,
+      "grad_norm": 0.1572265625,
+      "learning_rate": 0.02,
+      "loss": 1.6877,
+      "step": 2483
+    },
+    {
+      "epoch": 0.22996007637562924,
+      "grad_norm": 0.1630859375,
+      "learning_rate": 0.02,
+      "loss": 1.7259,
+      "step": 2484
+    },
+    {
+      "epoch": 0.2300526528959093,
+      "grad_norm": 0.1611328125,
+      "learning_rate": 0.02,
+      "loss": 1.6571,
+      "step": 2485
+    },
+    {
+      "epoch": 0.2301452294161893,
+      "grad_norm": 0.1474609375,
+      "learning_rate": 0.02,
+      "loss": 1.6584,
+      "step": 2486
+    },
+    {
+      "epoch": 0.23023780593646936,
+      "grad_norm": 0.1474609375,
+      "learning_rate": 0.02,
+      "loss": 1.6268,
+      "step": 2487
+    },
+    {
+      "epoch": 0.23033038245674942,
+      "grad_norm": 0.150390625,
+      "learning_rate": 0.02,
+      "loss": 1.7213,
+      "step": 2488
+    },
+    {
+      "epoch": 0.23042295897702944,
+      "grad_norm": 0.15234375,
+      "learning_rate": 0.02,
+      "loss": 1.6607,
+      "step": 2489
+    },
+    {
+      "epoch": 0.2305155354973095,
+      "grad_norm": 0.15234375,
+      "learning_rate": 0.02,
+      "loss": 1.6339,
+      "step": 2490
+    },
+    {
+      "epoch": 0.23060811201758954,
+      "grad_norm": 0.1572265625,
+      "learning_rate": 0.02,
+      "loss": 1.6952,
+      "step": 2491
+    },
+    {
+      "epoch": 0.23070068853786957,
+      "grad_norm": 0.1533203125,
+      "learning_rate": 0.02,
+      "loss": 1.6764,
+      "step": 2492
+    },
+    {
+      "epoch": 0.23079326505814962,
+      "grad_norm": 0.1474609375,
+      "learning_rate": 0.02,
+      "loss": 1.7212,
+      "step": 2493
+    },
+    {
+      "epoch": 0.23088584157842967,
+      "grad_norm": 0.138671875,
+      "learning_rate": 0.02,
+      "loss": 1.6924,
+      "step": 2494
+    },
+    {
+      "epoch": 0.23097841809870973,
+      "grad_norm": 0.140625,
+      "learning_rate": 0.02,
+      "loss": 1.6027,
+      "step": 2495
+    },
+    {
+      "epoch": 0.23107099461898975,
+      "grad_norm": 0.142578125,
+      "learning_rate": 0.02,
+      "loss": 1.6437,
+      "step": 2496
+    },
+    {
+      "epoch": 0.2311635711392698,
+      "grad_norm": 0.1484375,
+      "learning_rate": 0.02,
+      "loss": 1.6949,
+      "step": 2497
+    },
+    {
+      "epoch": 0.23125614765954985,
+      "grad_norm": 0.1357421875,
+      "learning_rate": 0.02,
+      "loss": 1.676,
+      "step": 2498
+    },
+    {
+      "epoch": 0.23134872417982988,
+      "grad_norm": 0.13671875,
+      "learning_rate": 0.02,
+      "loss": 1.6361,
+      "step": 2499
+    },
+    {
+      "epoch": 0.23144130070010993,
+      "grad_norm": 0.146484375,
+      "learning_rate": 0.02,
+      "loss": 1.7479,
+      "step": 2500
+    },
+    {
+      "epoch": 0.23153387722038998,
+      "grad_norm": 0.15234375,
+      "learning_rate": 0.02,
+      "loss": 1.7123,
+      "step": 2501
+    },
+    {
+      "epoch": 0.23162645374067004,
+      "grad_norm": 0.1455078125,
+      "learning_rate": 0.02,
+      "loss": 1.6494,
+      "step": 2502
+    },
+    {
+      "epoch": 0.23171903026095006,
+      "grad_norm": 0.15234375,
+      "learning_rate": 0.02,
+      "loss": 1.7023,
+      "step": 2503
+    },
+    {
+      "epoch": 0.2318116067812301,
+      "grad_norm": 0.1455078125,
+      "learning_rate": 0.02,
+      "loss": 1.6683,
+      "step": 2504
+    },
+    {
+      "epoch": 0.23190418330151016,
+      "grad_norm": 0.1357421875,
+      "learning_rate": 0.02,
+      "loss": 1.689,
+      "step": 2505
+    },
+    {
+      "epoch": 0.2319967598217902,
+      "grad_norm": 0.138671875,
+      "learning_rate": 0.02,
+      "loss": 1.6452,
+      "step": 2506
+    },
+    {
+      "epoch": 0.23208933634207024,
+      "grad_norm": 0.1513671875,
+      "learning_rate": 0.02,
+      "loss": 1.6936,
+      "step": 2507
+    },
+    {
+      "epoch": 0.2321819128623503,
+      "grad_norm": 0.1572265625,
+      "learning_rate": 0.02,
+      "loss": 1.7269,
+      "step": 2508
+    },
+    {
+      "epoch": 0.23227448938263032,
+      "grad_norm": 0.1513671875,
+      "learning_rate": 0.02,
+      "loss": 1.7082,
+      "step": 2509
+    },
+    {
+      "epoch": 0.23236706590291037,
+      "grad_norm": 0.1494140625,
+      "learning_rate": 0.02,
+      "loss": 1.7202,
+      "step": 2510
+    },
+    {
+      "epoch": 0.23245964242319042,
+      "grad_norm": 0.1533203125,
+      "learning_rate": 0.02,
+      "loss": 1.6258,
+      "step": 2511
+    },
+    {
+      "epoch": 0.23255221894347047,
+      "grad_norm": 0.158203125,
+      "learning_rate": 0.02,
+      "loss": 1.612,
+      "step": 2512
+    },
+    {
+      "epoch": 0.2326447954637505,
+      "grad_norm": 0.15625,
+      "learning_rate": 0.02,
+      "loss": 1.7478,
+      "step": 2513
+    },
+    {
+      "epoch": 0.23273737198403055,
+      "grad_norm": 0.1552734375,
+      "learning_rate": 0.02,
+      "loss": 1.7372,
+      "step": 2514
+    },
+    {
+      "epoch": 0.2328299485043106,
+      "grad_norm": 0.1474609375,
+      "learning_rate": 0.02,
+      "loss": 1.6854,
+      "step": 2515
+    },
+    {
+      "epoch": 0.23292252502459063,
+      "grad_norm": 0.142578125,
+      "learning_rate": 0.02,
+      "loss": 1.6367,
+      "step": 2516
+    },
+    {
+      "epoch": 0.23301510154487068,
+      "grad_norm": 0.14453125,
+      "learning_rate": 0.02,
+      "loss": 1.6622,
+      "step": 2517
+    },
+    {
+      "epoch": 0.23310767806515073,
+      "grad_norm": 0.142578125,
+      "learning_rate": 0.02,
+      "loss": 1.6445,
+      "step": 2518
+    },
+    {
+      "epoch": 0.23320025458543078,
+      "grad_norm": 0.1484375,
+      "learning_rate": 0.02,
+      "loss": 1.6398,
+      "step": 2519
+    },
+    {
+      "epoch": 0.2332928311057108,
+      "grad_norm": 0.1484375,
+      "learning_rate": 0.02,
+      "loss": 1.6635,
+      "step": 2520
+    },
+    {
+      "epoch": 0.23338540762599086,
+      "grad_norm": 0.1455078125,
+      "learning_rate": 0.02,
+      "loss": 1.6919,
+      "step": 2521
+    },
+    {
+      "epoch": 0.2334779841462709,
+      "grad_norm": 0.16015625,
+      "learning_rate": 0.02,
+      "loss": 1.6623,
+      "step": 2522
+    },
+    {
+      "epoch": 0.23357056066655094,
+      "grad_norm": 0.15625,
+      "learning_rate": 0.02,
+      "loss": 1.7046,
+      "step": 2523
+    },
+    {
+      "epoch": 0.233663137186831,
+      "grad_norm": 0.1552734375,
+      "learning_rate": 0.02,
+      "loss": 1.689,
+      "step": 2524
+    },
+    {
+      "epoch": 0.23375571370711104,
+      "grad_norm": 0.14453125,
+      "learning_rate": 0.02,
+      "loss": 1.713,
+      "step": 2525
+    },
+    {
+      "epoch": 0.23384829022739106,
+      "grad_norm": 0.1416015625,
+      "learning_rate": 0.02,
+      "loss": 1.6724,
+      "step": 2526
+    },
+    {
+      "epoch": 0.23394086674767112,
+      "grad_norm": 0.1455078125,
+      "learning_rate": 0.02,
+      "loss": 1.701,
+      "step": 2527
+    },
+    {
+      "epoch": 0.23403344326795117,
+      "grad_norm": 0.15625,
+      "learning_rate": 0.02,
+      "loss": 1.707,
+      "step": 2528
+    },
+    {
+      "epoch": 0.23412601978823122,
+      "grad_norm": 0.154296875,
+      "learning_rate": 0.02,
+      "loss": 1.596,
+      "step": 2529
+    },
+    {
+      "epoch": 0.23421859630851125,
+      "grad_norm": 0.14453125,
+      "learning_rate": 0.02,
+      "loss": 1.6777,
+      "step": 2530
+    },
+    {
+      "epoch": 0.2343111728287913,
+      "grad_norm": 0.154296875,
+      "learning_rate": 0.02,
+      "loss": 1.6366,
+      "step": 2531
+    },
+    {
+      "epoch": 0.23440374934907135,
+      "grad_norm": 0.1513671875,
+      "learning_rate": 0.02,
+      "loss": 1.6342,
+      "step": 2532
+    },
+    {
+      "epoch": 0.23449632586935137,
+      "grad_norm": 0.158203125,
+      "learning_rate": 0.02,
+      "loss": 1.611,
+      "step": 2533
+    },
+    {
+      "epoch": 0.23458890238963143,
+      "grad_norm": 0.154296875,
+      "learning_rate": 0.02,
+      "loss": 1.7006,
+      "step": 2534
+    },
+    {
+      "epoch": 0.23468147890991148,
+      "grad_norm": 0.1650390625,
+      "learning_rate": 0.02,
+      "loss": 1.6867,
+      "step": 2535
+    },
+    {
+      "epoch": 0.23477405543019153,
+      "grad_norm": 0.1572265625,
+      "learning_rate": 0.02,
+      "loss": 1.753,
+      "step": 2536
+    },
+    {
+      "epoch": 0.23486663195047155,
+      "grad_norm": 0.16015625,
+      "learning_rate": 0.02,
+      "loss": 1.6279,
+      "step": 2537
+    },
+    {
+      "epoch": 0.2349592084707516,
+      "grad_norm": 0.150390625,
+      "learning_rate": 0.02,
+      "loss": 1.7052,
+      "step": 2538
+    },
+    {
+      "epoch": 0.23505178499103166,
+      "grad_norm": 0.1650390625,
+      "learning_rate": 0.02,
+      "loss": 1.621,
+      "step": 2539
+    },
+    {
+      "epoch": 0.23514436151131168,
+      "grad_norm": 0.1552734375,
+      "learning_rate": 0.02,
+      "loss": 1.6222,
+      "step": 2540
+    },
+    {
+      "epoch": 0.23523693803159174,
+      "grad_norm": 0.1591796875,
+      "learning_rate": 0.02,
+      "loss": 1.6234,
+      "step": 2541
+    },
+    {
+      "epoch": 0.2353295145518718,
+      "grad_norm": 0.1494140625,
+      "learning_rate": 0.02,
+      "loss": 1.7017,
+      "step": 2542
+    },
+    {
+      "epoch": 0.2354220910721518,
+      "grad_norm": 0.1572265625,
+      "learning_rate": 0.02,
+      "loss": 1.675,
+      "step": 2543
+    },
+    {
+      "epoch": 0.23551466759243186,
+      "grad_norm": 0.158203125,
+      "learning_rate": 0.02,
+      "loss": 1.7356,
+      "step": 2544
+    },
+    {
+      "epoch": 0.23560724411271192,
+      "grad_norm": 0.166015625,
+      "learning_rate": 0.02,
+      "loss": 1.6933,
+      "step": 2545
+    },
+    {
+      "epoch": 0.23569982063299197,
+      "grad_norm": 0.1650390625,
+      "learning_rate": 0.02,
+      "loss": 1.6727,
+      "step": 2546
+    },
+    {
+      "epoch": 0.235792397153272,
+      "grad_norm": 0.1630859375,
+      "learning_rate": 0.02,
+      "loss": 1.6447,
+      "step": 2547
+    },
+    {
+      "epoch": 0.23588497367355205,
+      "grad_norm": 0.1572265625,
+      "learning_rate": 0.02,
+      "loss": 1.6778,
+      "step": 2548
+    },
+    {
+      "epoch": 0.2359775501938321,
+      "grad_norm": 0.16796875,
+      "learning_rate": 0.02,
+      "loss": 1.6902,
+      "step": 2549
+    },
+    {
+      "epoch": 0.23607012671411212,
+      "grad_norm": 0.1630859375,
+      "learning_rate": 0.02,
+      "loss": 1.6965,
+      "step": 2550
+    },
+    {
+      "epoch": 0.23616270323439217,
+      "grad_norm": 0.1533203125,
+      "learning_rate": 0.02,
+      "loss": 1.6387,
+      "step": 2551
+    },
+    {
+      "epoch": 0.23625527975467223,
+      "grad_norm": 0.15234375,
+      "learning_rate": 0.02,
+      "loss": 1.6252,
+      "step": 2552
+    },
+    {
+      "epoch": 0.23634785627495228,
+      "grad_norm": 0.154296875,
+      "learning_rate": 0.02,
+      "loss": 1.5882,
+      "step": 2553
+    },
+    {
+      "epoch": 0.2364404327952323,
+      "grad_norm": 0.162109375,
+      "learning_rate": 0.02,
+      "loss": 1.6347,
+      "step": 2554
+    },
+    {
+      "epoch": 0.23653300931551235,
+      "grad_norm": 0.154296875,
+      "learning_rate": 0.02,
+      "loss": 1.648,
+      "step": 2555
+    },
+    {
+      "epoch": 0.2366255858357924,
+      "grad_norm": 0.1513671875,
+      "learning_rate": 0.02,
+      "loss": 1.6711,
+      "step": 2556
+    },
+    {
+      "epoch": 0.23671816235607243,
+      "grad_norm": 0.16015625,
+      "learning_rate": 0.02,
+      "loss": 1.653,
+      "step": 2557
+    },
+    {
+      "epoch": 0.23681073887635248,
+      "grad_norm": 0.1455078125,
+      "learning_rate": 0.02,
+      "loss": 1.6545,
+      "step": 2558
+    },
+    {
+      "epoch": 0.23690331539663254,
+      "grad_norm": 0.1533203125,
+      "learning_rate": 0.02,
+      "loss": 1.6898,
+      "step": 2559
+    },
+    {
+      "epoch": 0.23699589191691256,
+      "grad_norm": 0.16015625,
+      "learning_rate": 0.02,
+      "loss": 1.6893,
+      "step": 2560
+    },
+    {
+      "epoch": 0.2370884684371926,
+      "grad_norm": 0.154296875,
+      "learning_rate": 0.02,
+      "loss": 1.6841,
+      "step": 2561
+    },
+    {
+      "epoch": 0.23718104495747266,
+      "grad_norm": 0.150390625,
+      "learning_rate": 0.02,
+      "loss": 1.7139,
+      "step": 2562
+    },
+    {
+      "epoch": 0.23727362147775272,
+      "grad_norm": 0.1591796875,
+      "learning_rate": 0.02,
+      "loss": 1.6801,
+      "step": 2563
+    },
+    {
+      "epoch": 0.23736619799803274,
+      "grad_norm": 0.16015625,
+      "learning_rate": 0.02,
+      "loss": 1.6869,
+      "step": 2564
+    },
+    {
+      "epoch": 0.2374587745183128,
+      "grad_norm": 0.15234375,
+      "learning_rate": 0.02,
+      "loss": 1.6134,
+      "step": 2565
+    },
+    {
+      "epoch": 0.23755135103859285,
+      "grad_norm": 0.1396484375,
+      "learning_rate": 0.02,
+      "loss": 1.6441,
+      "step": 2566
+    },
+    {
+      "epoch": 0.23764392755887287,
+      "grad_norm": 0.1533203125,
+      "learning_rate": 0.02,
+      "loss": 1.6278,
+      "step": 2567
+    },
+    {
+      "epoch": 0.23773650407915292,
+      "grad_norm": 0.1455078125,
+      "learning_rate": 0.02,
+      "loss": 1.6582,
+      "step": 2568
+    },
+    {
+      "epoch": 0.23782908059943297,
+      "grad_norm": 0.158203125,
+      "learning_rate": 0.02,
+      "loss": 1.7179,
+      "step": 2569
+    },
+    {
+      "epoch": 0.23792165711971303,
+      "grad_norm": 0.15625,
+      "learning_rate": 0.02,
+      "loss": 1.6585,
+      "step": 2570
+    },
+    {
+      "epoch": 0.23801423363999305,
+      "grad_norm": 0.1474609375,
+      "learning_rate": 0.02,
+      "loss": 1.6518,
+      "step": 2571
+    },
+    {
+      "epoch": 0.2381068101602731,
+      "grad_norm": 0.1416015625,
+      "learning_rate": 0.02,
+      "loss": 1.6047,
+      "step": 2572
+    },
+    {
+      "epoch": 0.23819938668055315,
+      "grad_norm": 0.1533203125,
+      "learning_rate": 0.02,
+      "loss": 1.7011,
+      "step": 2573
+    },
+    {
+      "epoch": 0.23829196320083318,
+      "grad_norm": 0.1484375,
+      "learning_rate": 0.02,
+      "loss": 1.6548,
+      "step": 2574
+    },
+    {
+      "epoch": 0.23838453972111323,
+      "grad_norm": 0.1455078125,
+      "learning_rate": 0.02,
+      "loss": 1.6359,
+      "step": 2575
+    },
+    {
+      "epoch": 0.23847711624139328,
+      "grad_norm": 0.1376953125,
+      "learning_rate": 0.02,
+      "loss": 1.6733,
+      "step": 2576
+    },
+    {
+      "epoch": 0.2385696927616733,
+      "grad_norm": 0.146484375,
+      "learning_rate": 0.02,
+      "loss": 1.6377,
+      "step": 2577
+    },
+    {
+      "epoch": 0.23866226928195336,
+      "grad_norm": 0.1396484375,
+      "learning_rate": 0.02,
+      "loss": 1.6935,
+      "step": 2578
+    },
+    {
+      "epoch": 0.2387548458022334,
+      "grad_norm": 0.1591796875,
+      "learning_rate": 0.02,
+      "loss": 1.6977,
+      "step": 2579
+    },
+    {
+      "epoch": 0.23884742232251346,
+      "grad_norm": 0.1435546875,
+      "learning_rate": 0.02,
+      "loss": 1.5989,
+      "step": 2580
+    },
+    {
+      "epoch": 0.2389399988427935,
+      "grad_norm": 0.1513671875,
+      "learning_rate": 0.02,
+      "loss": 1.7097,
+      "step": 2581
+    },
+    {
+      "epoch": 0.23903257536307354,
+      "grad_norm": 0.1396484375,
+      "learning_rate": 0.02,
+      "loss": 1.6746,
+      "step": 2582
+    },
+    {
+      "epoch": 0.2391251518833536,
+      "grad_norm": 0.1435546875,
+      "learning_rate": 0.02,
+      "loss": 1.6428,
+      "step": 2583
+    },
+    {
+      "epoch": 0.23921772840363362,
+      "grad_norm": 0.14453125,
+      "learning_rate": 0.02,
+      "loss": 1.6383,
+      "step": 2584
+    },
+    {
+      "epoch": 0.23931030492391367,
+      "grad_norm": 0.1494140625,
+      "learning_rate": 0.02,
+      "loss": 1.6793,
+      "step": 2585
+    },
+    {
+      "epoch": 0.23940288144419372,
+      "grad_norm": 0.140625,
+      "learning_rate": 0.02,
+      "loss": 1.6865,
+      "step": 2586
+    },
+    {
+      "epoch": 0.23949545796447377,
+      "grad_norm": 0.1484375,
+      "learning_rate": 0.02,
+      "loss": 1.6283,
+      "step": 2587
+    },
+    {
+      "epoch": 0.2395880344847538,
+      "grad_norm": 0.146484375,
+      "learning_rate": 0.02,
+      "loss": 1.7432,
+      "step": 2588
+    },
+    {
+      "epoch": 0.23968061100503385,
+      "grad_norm": 0.1474609375,
+      "learning_rate": 0.02,
+      "loss": 1.6889,
+      "step": 2589
+    },
+    {
+      "epoch": 0.2397731875253139,
+      "grad_norm": 0.1416015625,
+      "learning_rate": 0.02,
+      "loss": 1.7378,
+      "step": 2590
+    },
+    {
+      "epoch": 0.23986576404559393,
+      "grad_norm": 0.150390625,
+      "learning_rate": 0.02,
+      "loss": 1.58,
+      "step": 2591
+    },
+    {
+      "epoch": 0.23995834056587398,
+      "grad_norm": 0.1552734375,
+      "learning_rate": 0.02,
+      "loss": 1.6611,
+      "step": 2592
+    },
+    {
+      "epoch": 0.24005091708615403,
+      "grad_norm": 0.162109375,
+      "learning_rate": 0.02,
+      "loss": 1.6326,
+      "step": 2593
+    },
+    {
+      "epoch": 0.24014349360643406,
+      "grad_norm": 0.162109375,
+      "learning_rate": 0.02,
+      "loss": 1.646,
+      "step": 2594
+    },
+    {
+      "epoch": 0.2402360701267141,
+      "grad_norm": 0.1572265625,
+      "learning_rate": 0.02,
+      "loss": 1.641,
+      "step": 2595
+    },
+    {
+      "epoch": 0.24032864664699416,
+      "grad_norm": 0.15234375,
+      "learning_rate": 0.02,
+      "loss": 1.6926,
+      "step": 2596
+    },
+    {
+      "epoch": 0.2404212231672742,
+      "grad_norm": 0.1689453125,
+      "learning_rate": 0.02,
+      "loss": 1.6752,
+      "step": 2597
+    },
+    {
+      "epoch": 0.24051379968755424,
+      "grad_norm": 0.1630859375,
+      "learning_rate": 0.02,
+      "loss": 1.7213,
+      "step": 2598
+    },
+    {
+      "epoch": 0.2406063762078343,
+      "grad_norm": 0.16015625,
+      "learning_rate": 0.02,
+      "loss": 1.7003,
+      "step": 2599
+    },
+    {
+      "epoch": 0.24069895272811434,
+      "grad_norm": 0.14453125,
+      "learning_rate": 0.02,
+      "loss": 1.6978,
+      "step": 2600
+    },
+    {
+      "epoch": 0.24079152924839436,
+      "grad_norm": 0.1474609375,
+      "learning_rate": 0.02,
+      "loss": 1.6217,
+      "step": 2601
+    },
+    {
+      "epoch": 0.24088410576867442,
+      "grad_norm": 0.1513671875,
+      "learning_rate": 0.02,
+      "loss": 1.6908,
+      "step": 2602
+    },
+    {
+      "epoch": 0.24097668228895447,
+      "grad_norm": 0.1611328125,
+      "learning_rate": 0.02,
+      "loss": 1.6978,
+      "step": 2603
+    },
+    {
+      "epoch": 0.24106925880923452,
+      "grad_norm": 0.1494140625,
+      "learning_rate": 0.02,
+      "loss": 1.6323,
+      "step": 2604
+    },
+    {
+      "epoch": 0.24116183532951455,
+      "grad_norm": 0.14453125,
+      "learning_rate": 0.02,
+      "loss": 1.7372,
+      "step": 2605
+    },
+    {
+      "epoch": 0.2412544118497946,
+      "grad_norm": 0.1484375,
+      "learning_rate": 0.02,
+      "loss": 1.6353,
+      "step": 2606
+    },
+    {
+      "epoch": 0.24134698837007465,
+      "grad_norm": 0.15234375,
+      "learning_rate": 0.02,
+      "loss": 1.6863,
+      "step": 2607
+    },
+    {
+      "epoch": 0.24143956489035467,
+      "grad_norm": 0.146484375,
+      "learning_rate": 0.02,
+      "loss": 1.6135,
+      "step": 2608
+    },
+    {
+      "epoch": 0.24153214141063473,
+      "grad_norm": 0.154296875,
+      "learning_rate": 0.02,
+      "loss": 1.6775,
+      "step": 2609
+    },
+    {
+      "epoch": 0.24162471793091478,
+      "grad_norm": 0.14453125,
+      "learning_rate": 0.02,
+      "loss": 1.6705,
+      "step": 2610
+    },
+    {
+      "epoch": 0.2417172944511948,
+      "grad_norm": 0.14453125,
+      "learning_rate": 0.02,
+      "loss": 1.6209,
+      "step": 2611
+    },
+    {
+      "epoch": 0.24180987097147486,
+      "grad_norm": 0.134765625,
+      "learning_rate": 0.02,
+      "loss": 1.6261,
+      "step": 2612
+    },
+    {
+      "epoch": 0.2419024474917549,
+      "grad_norm": 0.1376953125,
+      "learning_rate": 0.02,
+      "loss": 1.6419,
+      "step": 2613
+    },
+    {
+      "epoch": 0.24199502401203496,
+      "grad_norm": 0.15234375,
+      "learning_rate": 0.02,
+      "loss": 1.6701,
+      "step": 2614
+    },
+    {
+      "epoch": 0.24208760053231498,
+      "grad_norm": 0.142578125,
+      "learning_rate": 0.02,
+      "loss": 1.687,
+      "step": 2615
+    },
+    {
+      "epoch": 0.24218017705259504,
+      "grad_norm": 0.1435546875,
+      "learning_rate": 0.02,
+      "loss": 1.6425,
+      "step": 2616
+    },
+    {
+      "epoch": 0.2422727535728751,
+      "grad_norm": 0.1484375,
+      "learning_rate": 0.02,
+      "loss": 1.6786,
+      "step": 2617
+    },
+    {
+      "epoch": 0.2423653300931551,
+      "grad_norm": 0.1513671875,
+      "learning_rate": 0.02,
+      "loss": 1.5871,
+      "step": 2618
+    },
+    {
+      "epoch": 0.24245790661343516,
+      "grad_norm": 0.1455078125,
+      "learning_rate": 0.02,
+      "loss": 1.603,
+      "step": 2619
+    },
+    {
+      "epoch": 0.24255048313371522,
+      "grad_norm": 0.15625,
+      "learning_rate": 0.02,
+      "loss": 1.6614,
+      "step": 2620
+    },
+    {
+      "epoch": 0.24264305965399527,
+      "grad_norm": 0.1533203125,
+      "learning_rate": 0.02,
+      "loss": 1.6613,
+      "step": 2621
+    },
+    {
+      "epoch": 0.2427356361742753,
+      "grad_norm": 0.15625,
+      "learning_rate": 0.02,
+      "loss": 1.6533,
+      "step": 2622
+    },
+    {
+      "epoch": 0.24282821269455535,
+      "grad_norm": 0.154296875,
+      "learning_rate": 0.02,
+      "loss": 1.6639,
+      "step": 2623
+    },
+    {
+      "epoch": 0.2429207892148354,
+      "grad_norm": 0.16015625,
+      "learning_rate": 0.02,
+      "loss": 1.6339,
+      "step": 2624
+    },
+    {
+      "epoch": 0.24301336573511542,
+      "grad_norm": 0.154296875,
+      "learning_rate": 0.02,
+      "loss": 1.6194,
+      "step": 2625
+    },
+    {
+      "epoch": 0.24310594225539547,
+      "grad_norm": 0.15234375,
+      "learning_rate": 0.02,
+      "loss": 1.6777,
+      "step": 2626
+    },
+    {
+      "epoch": 0.24319851877567553,
+      "grad_norm": 0.1533203125,
+      "learning_rate": 0.02,
+      "loss": 1.694,
+      "step": 2627
+    },
+    {
+      "epoch": 0.24329109529595555,
+      "grad_norm": 0.14453125,
+      "learning_rate": 0.02,
+      "loss": 1.666,
+      "step": 2628
+    },
+    {
+      "epoch": 0.2433836718162356,
+      "grad_norm": 0.1494140625,
+      "learning_rate": 0.02,
+      "loss": 1.6598,
+      "step": 2629
+    },
+    {
+      "epoch": 0.24347624833651565,
+      "grad_norm": 0.1474609375,
+      "learning_rate": 0.02,
+      "loss": 1.6678,
+      "step": 2630
+    },
+    {
+      "epoch": 0.2435688248567957,
+      "grad_norm": 0.1513671875,
+      "learning_rate": 0.02,
+      "loss": 1.6775,
+      "step": 2631
+    },
+    {
+      "epoch": 0.24366140137707573,
+      "grad_norm": 0.1494140625,
+      "learning_rate": 0.02,
+      "loss": 1.6848,
+      "step": 2632
+    },
+    {
+      "epoch": 0.24375397789735578,
+      "grad_norm": 0.1494140625,
+      "learning_rate": 0.02,
+      "loss": 1.6924,
+      "step": 2633
+    },
+    {
+      "epoch": 0.24384655441763584,
+      "grad_norm": 0.1357421875,
+      "learning_rate": 0.02,
+      "loss": 1.5874,
+      "step": 2634
+    },
+    {
+      "epoch": 0.24393913093791586,
+      "grad_norm": 0.1416015625,
+      "learning_rate": 0.02,
+      "loss": 1.6152,
+      "step": 2635
+    },
+    {
+      "epoch": 0.2440317074581959,
+      "grad_norm": 0.1357421875,
+      "learning_rate": 0.02,
+      "loss": 1.6561,
+      "step": 2636
+    },
+    {
+      "epoch": 0.24412428397847596,
+      "grad_norm": 0.150390625,
+      "learning_rate": 0.02,
+      "loss": 1.6867,
+      "step": 2637
+    },
+    {
+      "epoch": 0.24421686049875602,
+      "grad_norm": 0.16015625,
+      "learning_rate": 0.02,
+      "loss": 1.7139,
+      "step": 2638
+    },
+    {
+      "epoch": 0.24430943701903604,
+      "grad_norm": 0.154296875,
+      "learning_rate": 0.02,
+      "loss": 1.6819,
+      "step": 2639
+    },
+    {
+      "epoch": 0.2444020135393161,
+      "grad_norm": 0.146484375,
+      "learning_rate": 0.02,
+      "loss": 1.6436,
+      "step": 2640
+    },
+    {
+      "epoch": 0.24449459005959615,
+      "grad_norm": 0.1513671875,
+      "learning_rate": 0.02,
+      "loss": 1.6243,
+      "step": 2641
+    },
+    {
+      "epoch": 0.24458716657987617,
+      "grad_norm": 0.1416015625,
+      "learning_rate": 0.02,
+      "loss": 1.6036,
+      "step": 2642
+    },
+    {
+      "epoch": 0.24467974310015622,
+      "grad_norm": 0.150390625,
+      "learning_rate": 0.02,
+      "loss": 1.6353,
+      "step": 2643
+    },
+    {
+      "epoch": 0.24477231962043627,
+      "grad_norm": 0.1494140625,
+      "learning_rate": 0.02,
+      "loss": 1.5888,
+      "step": 2644
+    },
+    {
+      "epoch": 0.2448648961407163,
+      "grad_norm": 0.1455078125,
+      "learning_rate": 0.02,
+      "loss": 1.6343,
+      "step": 2645
+    },
+    {
+      "epoch": 0.24495747266099635,
+      "grad_norm": 0.1396484375,
+      "learning_rate": 0.02,
+      "loss": 1.6279,
+      "step": 2646
+    },
+    {
+      "epoch": 0.2450500491812764,
+      "grad_norm": 0.1435546875,
+      "learning_rate": 0.02,
+      "loss": 1.703,
+      "step": 2647
+    },
+    {
+      "epoch": 0.24514262570155645,
+      "grad_norm": 0.142578125,
+      "learning_rate": 0.02,
+      "loss": 1.6384,
+      "step": 2648
+    },
+    {
+      "epoch": 0.24523520222183648,
+      "grad_norm": 0.1474609375,
+      "learning_rate": 0.02,
+      "loss": 1.6152,
+      "step": 2649
+    },
+    {
+      "epoch": 0.24532777874211653,
+      "grad_norm": 0.150390625,
+      "learning_rate": 0.02,
+      "loss": 1.6887,
+      "step": 2650
+    },
+    {
+      "epoch": 0.24542035526239658,
+      "grad_norm": 0.1494140625,
+      "learning_rate": 0.02,
+      "loss": 1.6758,
+      "step": 2651
+    },
+    {
+      "epoch": 0.2455129317826766,
+      "grad_norm": 0.1494140625,
+      "learning_rate": 0.02,
+      "loss": 1.6623,
+      "step": 2652
+    },
+    {
+      "epoch": 0.24560550830295666,
+      "grad_norm": 0.154296875,
+      "learning_rate": 0.02,
+      "loss": 1.706,
+      "step": 2653
+    },
+    {
+      "epoch": 0.2456980848232367,
+      "grad_norm": 0.1669921875,
+      "learning_rate": 0.02,
+      "loss": 1.7213,
+      "step": 2654
+    },
+    {
+      "epoch": 0.24579066134351676,
+      "grad_norm": 0.1669921875,
+      "learning_rate": 0.02,
+      "loss": 1.7687,
+      "step": 2655
+    },
+    {
+      "epoch": 0.2458832378637968,
+      "grad_norm": 0.1611328125,
+      "learning_rate": 0.02,
+      "loss": 1.7282,
+      "step": 2656
+    },
+    {
+      "epoch": 0.24597581438407684,
+      "grad_norm": 0.1611328125,
+      "learning_rate": 0.02,
+      "loss": 1.7056,
+      "step": 2657
+    },
+    {
+      "epoch": 0.2460683909043569,
+      "grad_norm": 0.15234375,
+      "learning_rate": 0.02,
+      "loss": 1.6426,
+      "step": 2658
+    },
+    {
+      "epoch": 0.24616096742463692,
+      "grad_norm": 0.1552734375,
+      "learning_rate": 0.02,
+      "loss": 1.6115,
+      "step": 2659
+    },
+    {
+      "epoch": 0.24625354394491697,
+      "grad_norm": 0.154296875,
+      "learning_rate": 0.02,
+      "loss": 1.6938,
+      "step": 2660
+    },
+    {
+      "epoch": 0.24634612046519702,
+      "grad_norm": 0.1591796875,
+      "learning_rate": 0.02,
+      "loss": 1.6401,
+      "step": 2661
+    },
+    {
+      "epoch": 0.24643869698547705,
+      "grad_norm": 0.1484375,
+      "learning_rate": 0.02,
+      "loss": 1.6226,
+      "step": 2662
+    },
+    {
+      "epoch": 0.2465312735057571,
+      "grad_norm": 0.158203125,
+      "learning_rate": 0.02,
+      "loss": 1.6953,
+      "step": 2663
+    },
+    {
+      "epoch": 0.24662385002603715,
+      "grad_norm": 0.158203125,
+      "learning_rate": 0.02,
+      "loss": 1.6921,
+      "step": 2664
+    },
+    {
+      "epoch": 0.2467164265463172,
+      "grad_norm": 0.1484375,
+      "learning_rate": 0.02,
+      "loss": 1.6004,
+      "step": 2665
+    },
+    {
+      "epoch": 0.24680900306659723,
+      "grad_norm": 0.1396484375,
+      "learning_rate": 0.02,
+      "loss": 1.6302,
+      "step": 2666
+    },
+    {
+      "epoch": 0.24690157958687728,
+      "grad_norm": 0.1435546875,
+      "learning_rate": 0.02,
+      "loss": 1.604,
+      "step": 2667
+    },
+    {
+      "epoch": 0.24699415610715733,
+      "grad_norm": 0.1494140625,
+      "learning_rate": 0.02,
+      "loss": 1.5945,
+      "step": 2668
+    },
+    {
+      "epoch": 0.24708673262743736,
+      "grad_norm": 0.1630859375,
+      "learning_rate": 0.02,
+      "loss": 1.6713,
+      "step": 2669
+    },
+    {
+      "epoch": 0.2471793091477174,
+      "grad_norm": 0.158203125,
+      "learning_rate": 0.02,
+      "loss": 1.6453,
+      "step": 2670
+    },
+    {
+      "epoch": 0.24727188566799746,
+      "grad_norm": 0.15234375,
+      "learning_rate": 0.02,
+      "loss": 1.6212,
+      "step": 2671
+    },
+    {
+      "epoch": 0.2473644621882775,
+      "grad_norm": 0.1552734375,
+      "learning_rate": 0.02,
+      "loss": 1.6691,
+      "step": 2672
+    },
+    {
+      "epoch": 0.24745703870855754,
+      "grad_norm": 0.16015625,
+      "learning_rate": 0.02,
+      "loss": 1.6895,
+      "step": 2673
+    },
+    {
+      "epoch": 0.2475496152288376,
+      "grad_norm": 0.1494140625,
+      "learning_rate": 0.02,
+      "loss": 1.6548,
+      "step": 2674
+    },
+    {
+      "epoch": 0.24764219174911764,
+      "grad_norm": 0.15625,
+      "learning_rate": 0.02,
+      "loss": 1.6663,
+      "step": 2675
+    },
+    {
+      "epoch": 0.24773476826939766,
+      "grad_norm": 0.15234375,
+      "learning_rate": 0.02,
+      "loss": 1.6238,
+      "step": 2676
+    },
+    {
+      "epoch": 0.24782734478967772,
+      "grad_norm": 0.16015625,
+      "learning_rate": 0.02,
+      "loss": 1.6859,
+      "step": 2677
+    },
+    {
+      "epoch": 0.24791992130995777,
+      "grad_norm": 0.1630859375,
+      "learning_rate": 0.02,
+      "loss": 1.5579,
+      "step": 2678
+    },
+    {
+      "epoch": 0.2480124978302378,
+      "grad_norm": 0.1572265625,
+      "learning_rate": 0.02,
+      "loss": 1.6192,
+      "step": 2679
+    },
+    {
+      "epoch": 0.24810507435051785,
+      "grad_norm": 0.1552734375,
+      "learning_rate": 0.02,
+      "loss": 1.6615,
+      "step": 2680
+    },
+    {
+      "epoch": 0.2481976508707979,
+      "grad_norm": 0.1552734375,
+      "learning_rate": 0.02,
+      "loss": 1.6472,
+      "step": 2681
+    },
+    {
+      "epoch": 0.24829022739107795,
+      "grad_norm": 0.1484375,
+      "learning_rate": 0.02,
+      "loss": 1.6743,
+      "step": 2682
+    },
+    {
+      "epoch": 0.24838280391135797,
+      "grad_norm": 0.150390625,
+      "learning_rate": 0.02,
+      "loss": 1.6743,
+      "step": 2683
+    },
+    {
+      "epoch": 0.24847538043163803,
+      "grad_norm": 0.140625,
+      "learning_rate": 0.02,
+      "loss": 1.676,
+      "step": 2684
+    },
+    {
+      "epoch": 0.24856795695191808,
+      "grad_norm": 0.1396484375,
+      "learning_rate": 0.02,
+      "loss": 1.6005,
+      "step": 2685
+    },
+    {
+      "epoch": 0.2486605334721981,
+      "grad_norm": 0.13671875,
+      "learning_rate": 0.02,
+      "loss": 1.6041,
+      "step": 2686
+    },
+    {
+      "epoch": 0.24875310999247816,
+      "grad_norm": 0.1455078125,
+      "learning_rate": 0.02,
+      "loss": 1.6173,
+      "step": 2687
+    },
+    {
+      "epoch": 0.2488456865127582,
+      "grad_norm": 0.13671875,
+      "learning_rate": 0.02,
+      "loss": 1.6514,
+      "step": 2688
+    },
+    {
+      "epoch": 0.24893826303303826,
+      "grad_norm": 0.1376953125,
+      "learning_rate": 0.02,
+      "loss": 1.6773,
+      "step": 2689
+    },
+    {
+      "epoch": 0.24903083955331828,
+      "grad_norm": 0.14453125,
+      "learning_rate": 0.02,
+      "loss": 1.6496,
+      "step": 2690
+    },
+    {
+      "epoch": 0.24912341607359834,
+      "grad_norm": 0.1572265625,
+      "learning_rate": 0.02,
+      "loss": 1.6966,
+      "step": 2691
+    },
+    {
+      "epoch": 0.2492159925938784,
+      "grad_norm": 0.1513671875,
+      "learning_rate": 0.02,
+      "loss": 1.6741,
+      "step": 2692
+    },
+    {
+      "epoch": 0.2493085691141584,
+      "grad_norm": 0.166015625,
+      "learning_rate": 0.02,
+      "loss": 1.6864,
+      "step": 2693
+    },
+    {
+      "epoch": 0.24940114563443846,
+      "grad_norm": 0.1572265625,
+      "learning_rate": 0.02,
+      "loss": 1.6526,
+      "step": 2694
+    },
+    {
+      "epoch": 0.24949372215471852,
+      "grad_norm": 0.1611328125,
+      "learning_rate": 0.02,
+      "loss": 1.7017,
+      "step": 2695
+    },
+    {
+      "epoch": 0.24958629867499854,
+      "grad_norm": 0.1533203125,
+      "learning_rate": 0.02,
+      "loss": 1.6501,
+      "step": 2696
+    },
+    {
+      "epoch": 0.2496788751952786,
+      "grad_norm": 0.16015625,
+      "learning_rate": 0.02,
+      "loss": 1.6995,
+      "step": 2697
+    },
+    {
+      "epoch": 0.24977145171555865,
+      "grad_norm": 0.1552734375,
+      "learning_rate": 0.02,
+      "loss": 1.6778,
+      "step": 2698
+    },
+    {
+      "epoch": 0.2498640282358387,
+      "grad_norm": 0.1552734375,
+      "learning_rate": 0.02,
+      "loss": 1.6304,
+      "step": 2699
+    },
+    {
+      "epoch": 0.24995660475611872,
+      "grad_norm": 0.142578125,
+      "learning_rate": 0.02,
+      "loss": 1.7184,
+      "step": 2700
+    },
+    {
+      "epoch": 0.25004918127639875,
+      "grad_norm": 0.16015625,
+      "learning_rate": 0.02,
+      "loss": 1.7369,
+      "step": 2701
+    },
+    {
+      "epoch": 0.2501417577966788,
+      "grad_norm": 0.1484375,
+      "learning_rate": 0.02,
+      "loss": 1.6094,
+      "step": 2702
+    },
+    {
+      "epoch": 0.25023433431695885,
+      "grad_norm": 0.1533203125,
+      "learning_rate": 0.02,
+      "loss": 1.6358,
+      "step": 2703
+    },
+    {
+      "epoch": 0.2503269108372389,
+      "grad_norm": 0.15234375,
+      "learning_rate": 0.02,
+      "loss": 1.6927,
+      "step": 2704
+    },
+    {
+      "epoch": 0.25041948735751896,
+      "grad_norm": 0.1572265625,
+      "learning_rate": 0.02,
+      "loss": 1.6453,
+      "step": 2705
+    },
+    {
+      "epoch": 0.250512063877799,
+      "grad_norm": 0.1728515625,
+      "learning_rate": 0.02,
+      "loss": 1.6397,
+      "step": 2706
+    },
+    {
+      "epoch": 0.25060464039807906,
+      "grad_norm": 0.1533203125,
+      "learning_rate": 0.02,
+      "loss": 1.6373,
+      "step": 2707
+    },
+    {
+      "epoch": 0.25069721691835906,
+      "grad_norm": 0.1572265625,
+      "learning_rate": 0.02,
+      "loss": 1.6574,
+      "step": 2708
+    },
+    {
+      "epoch": 0.2507897934386391,
+      "grad_norm": 0.1669921875,
+      "learning_rate": 0.02,
+      "loss": 1.683,
+      "step": 2709
+    },
+    {
+      "epoch": 0.25088236995891916,
+      "grad_norm": 0.1640625,
+      "learning_rate": 0.02,
+      "loss": 1.7074,
+      "step": 2710
+    },
+    {
+      "epoch": 0.2509749464791992,
+      "grad_norm": 0.158203125,
+      "learning_rate": 0.02,
+      "loss": 1.6745,
+      "step": 2711
+    },
+    {
+      "epoch": 0.25106752299947926,
+      "grad_norm": 0.1552734375,
+      "learning_rate": 0.02,
+      "loss": 1.6705,
+      "step": 2712
+    },
+    {
+      "epoch": 0.2511600995197593,
+      "grad_norm": 0.162109375,
+      "learning_rate": 0.02,
+      "loss": 1.6188,
+      "step": 2713
+    },
+    {
+      "epoch": 0.25125267604003937,
+      "grad_norm": 0.1474609375,
+      "learning_rate": 0.02,
+      "loss": 1.6323,
+      "step": 2714
+    },
+    {
+      "epoch": 0.25134525256031937,
+      "grad_norm": 0.158203125,
+      "learning_rate": 0.02,
+      "loss": 1.6792,
+      "step": 2715
+    },
+    {
+      "epoch": 0.2514378290805994,
+      "grad_norm": 0.1552734375,
+      "learning_rate": 0.02,
+      "loss": 1.7093,
+      "step": 2716
+    },
+    {
+      "epoch": 0.25153040560087947,
+      "grad_norm": 0.158203125,
+      "learning_rate": 0.02,
+      "loss": 1.6941,
+      "step": 2717
+    },
+    {
+      "epoch": 0.2516229821211595,
+      "grad_norm": 0.1474609375,
+      "learning_rate": 0.02,
+      "loss": 1.6991,
+      "step": 2718
+    },
+    {
+      "epoch": 0.2517155586414396,
+      "grad_norm": 0.150390625,
+      "learning_rate": 0.02,
+      "loss": 1.6604,
+      "step": 2719
+    },
+    {
+      "epoch": 0.2518081351617196,
+      "grad_norm": 0.142578125,
+      "learning_rate": 0.02,
+      "loss": 1.7023,
+      "step": 2720
+    },
+    {
+      "epoch": 0.2519007116819997,
+      "grad_norm": 0.1552734375,
+      "learning_rate": 0.02,
+      "loss": 1.72,
+      "step": 2721
+    },
+    {
+      "epoch": 0.2519932882022797,
+      "grad_norm": 0.1474609375,
+      "learning_rate": 0.02,
+      "loss": 1.6754,
+      "step": 2722
+    },
+    {
+      "epoch": 0.2520858647225597,
+      "grad_norm": 0.146484375,
+      "learning_rate": 0.02,
+      "loss": 1.713,
+      "step": 2723
+    },
+    {
+      "epoch": 0.2521784412428398,
+      "grad_norm": 0.150390625,
+      "learning_rate": 0.02,
+      "loss": 1.728,
+      "step": 2724
+    },
+    {
+      "epoch": 0.25227101776311983,
+      "grad_norm": 0.1533203125,
+      "learning_rate": 0.02,
+      "loss": 1.6826,
+      "step": 2725
+    },
+    {
+      "epoch": 0.2523635942833999,
+      "grad_norm": 0.158203125,
+      "learning_rate": 0.02,
+      "loss": 1.6465,
+      "step": 2726
+    },
+    {
+      "epoch": 0.25245617080367994,
+      "grad_norm": 0.1533203125,
+      "learning_rate": 0.02,
+      "loss": 1.6399,
+      "step": 2727
+    },
+    {
+      "epoch": 0.25254874732396,
+      "grad_norm": 0.1435546875,
+      "learning_rate": 0.02,
+      "loss": 1.6052,
+      "step": 2728
+    },
+    {
+      "epoch": 0.25264132384424,
+      "grad_norm": 0.1416015625,
+      "learning_rate": 0.02,
+      "loss": 1.5989,
+      "step": 2729
+    },
+    {
+      "epoch": 0.25273390036452004,
+      "grad_norm": 0.14453125,
+      "learning_rate": 0.02,
+      "loss": 1.6454,
+      "step": 2730
+    },
+    {
+      "epoch": 0.2528264768848001,
+      "grad_norm": 0.1533203125,
+      "learning_rate": 0.02,
+      "loss": 1.6814,
+      "step": 2731
+    },
+    {
+      "epoch": 0.25291905340508014,
+      "grad_norm": 0.166015625,
+      "learning_rate": 0.02,
+      "loss": 1.6407,
+      "step": 2732
+    },
+    {
+      "epoch": 0.2530116299253602,
+      "grad_norm": 0.1484375,
+      "learning_rate": 0.02,
+      "loss": 1.6849,
+      "step": 2733
+    },
+    {
+      "epoch": 0.25310420644564025,
+      "grad_norm": 0.1552734375,
+      "learning_rate": 0.02,
+      "loss": 1.5967,
+      "step": 2734
+    },
+    {
+      "epoch": 0.25319678296592024,
+      "grad_norm": 0.1640625,
+      "learning_rate": 0.02,
+      "loss": 1.6596,
+      "step": 2735
+    },
+    {
+      "epoch": 0.2532893594862003,
+      "grad_norm": 0.15234375,
+      "learning_rate": 0.02,
+      "loss": 1.6836,
+      "step": 2736
+    },
+    {
+      "epoch": 0.25338193600648035,
+      "grad_norm": 0.146484375,
+      "learning_rate": 0.02,
+      "loss": 1.689,
+      "step": 2737
+    },
+    {
+      "epoch": 0.2534745125267604,
+      "grad_norm": 0.150390625,
+      "learning_rate": 0.02,
+      "loss": 1.7011,
+      "step": 2738
+    },
+    {
+      "epoch": 0.25356708904704045,
+      "grad_norm": 0.140625,
+      "learning_rate": 0.02,
+      "loss": 1.6348,
+      "step": 2739
+    },
+    {
+      "epoch": 0.2536596655673205,
+      "grad_norm": 0.142578125,
+      "learning_rate": 0.02,
+      "loss": 1.6331,
+      "step": 2740
+    },
+    {
+      "epoch": 0.25375224208760055,
+      "grad_norm": 0.1484375,
+      "learning_rate": 0.02,
+      "loss": 1.62,
+      "step": 2741
+    },
+    {
+      "epoch": 0.25384481860788055,
+      "grad_norm": 0.146484375,
+      "learning_rate": 0.02,
+      "loss": 1.6445,
+      "step": 2742
+    },
+    {
+      "epoch": 0.2539373951281606,
+      "grad_norm": 0.1416015625,
+      "learning_rate": 0.02,
+      "loss": 1.6724,
+      "step": 2743
+    },
+    {
+      "epoch": 0.25402997164844066,
+      "grad_norm": 0.150390625,
+      "learning_rate": 0.02,
+      "loss": 1.6466,
+      "step": 2744
+    },
+    {
+      "epoch": 0.2541225481687207,
+      "grad_norm": 0.146484375,
+      "learning_rate": 0.02,
+      "loss": 1.689,
+      "step": 2745
+    },
+    {
+      "epoch": 0.25421512468900076,
+      "grad_norm": 0.1494140625,
+      "learning_rate": 0.02,
+      "loss": 1.6311,
+      "step": 2746
+    },
+    {
+      "epoch": 0.2543077012092808,
+      "grad_norm": 0.158203125,
+      "learning_rate": 0.02,
+      "loss": 1.6765,
+      "step": 2747
+    },
+    {
+      "epoch": 0.25440027772956086,
+      "grad_norm": 0.1611328125,
+      "learning_rate": 0.02,
+      "loss": 1.6515,
+      "step": 2748
+    },
+    {
+      "epoch": 0.25449285424984086,
+      "grad_norm": 0.15234375,
+      "learning_rate": 0.02,
+      "loss": 1.6626,
+      "step": 2749
+    },
+    {
+      "epoch": 0.2545854307701209,
+      "grad_norm": 0.158203125,
+      "learning_rate": 0.02,
+      "loss": 1.7178,
+      "step": 2750
+    },
+    {
+      "epoch": 0.25467800729040097,
+      "grad_norm": 0.1513671875,
+      "learning_rate": 0.02,
+      "loss": 1.6839,
+      "step": 2751
+    },
+    {
+      "epoch": 0.254770583810681,
+      "grad_norm": 0.1572265625,
+      "learning_rate": 0.02,
+      "loss": 1.6863,
+      "step": 2752
+    },
+    {
+      "epoch": 0.25486316033096107,
+      "grad_norm": 0.1552734375,
+      "learning_rate": 0.02,
+      "loss": 1.6898,
+      "step": 2753
+    },
+    {
+      "epoch": 0.2549557368512411,
+      "grad_norm": 0.1630859375,
+      "learning_rate": 0.02,
+      "loss": 1.609,
+      "step": 2754
+    },
+    {
+      "epoch": 0.2550483133715212,
+      "grad_norm": 0.1533203125,
+      "learning_rate": 0.02,
+      "loss": 1.616,
+      "step": 2755
+    },
+    {
+      "epoch": 0.25514088989180117,
+      "grad_norm": 0.158203125,
+      "learning_rate": 0.02,
+      "loss": 1.6102,
+      "step": 2756
+    },
+    {
+      "epoch": 0.2552334664120812,
+      "grad_norm": 0.1572265625,
+      "learning_rate": 0.02,
+      "loss": 1.6667,
+      "step": 2757
+    },
+    {
+      "epoch": 0.2553260429323613,
+      "grad_norm": 0.154296875,
+      "learning_rate": 0.02,
+      "loss": 1.7157,
+      "step": 2758
+    },
+    {
+      "epoch": 0.2554186194526413,
+      "grad_norm": 0.150390625,
+      "learning_rate": 0.02,
+      "loss": 1.6555,
+      "step": 2759
+    },
+    {
+      "epoch": 0.2555111959729214,
+      "grad_norm": 0.142578125,
+      "learning_rate": 0.02,
+      "loss": 1.6451,
+      "step": 2760
+    },
+    {
+      "epoch": 0.25560377249320143,
+      "grad_norm": 0.154296875,
+      "learning_rate": 0.02,
+      "loss": 1.5886,
+      "step": 2761
+    },
+    {
+      "epoch": 0.2556963490134815,
+      "grad_norm": 0.1416015625,
+      "learning_rate": 0.02,
+      "loss": 1.6771,
+      "step": 2762
+    },
+    {
+      "epoch": 0.2557889255337615,
+      "grad_norm": 0.1572265625,
+      "learning_rate": 0.02,
+      "loss": 1.6772,
+      "step": 2763
+    },
+    {
+      "epoch": 0.25588150205404153,
+      "grad_norm": 0.1591796875,
+      "learning_rate": 0.02,
+      "loss": 1.6507,
+      "step": 2764
+    },
+    {
+      "epoch": 0.2559740785743216,
+      "grad_norm": 0.15234375,
+      "learning_rate": 0.02,
+      "loss": 1.6707,
+      "step": 2765
+    },
+    {
+      "epoch": 0.25606665509460164,
+      "grad_norm": 0.14453125,
+      "learning_rate": 0.02,
+      "loss": 1.624,
+      "step": 2766
+    },
+    {
+      "epoch": 0.2561592316148817,
+      "grad_norm": 0.140625,
+      "learning_rate": 0.02,
+      "loss": 1.6516,
+      "step": 2767
+    },
+    {
+      "epoch": 0.25625180813516174,
+      "grad_norm": 0.1513671875,
+      "learning_rate": 0.02,
+      "loss": 1.6816,
+      "step": 2768
+    },
+    {
+      "epoch": 0.25634438465544174,
+      "grad_norm": 0.15625,
+      "learning_rate": 0.02,
+      "loss": 1.6997,
+      "step": 2769
+    },
+    {
+      "epoch": 0.2564369611757218,
+      "grad_norm": 0.142578125,
+      "learning_rate": 0.02,
+      "loss": 1.6429,
+      "step": 2770
+    },
+    {
+      "epoch": 0.25652953769600184,
+      "grad_norm": 0.1572265625,
+      "learning_rate": 0.02,
+      "loss": 1.6529,
+      "step": 2771
+    },
+    {
+      "epoch": 0.2566221142162819,
+      "grad_norm": 0.1572265625,
+      "learning_rate": 0.02,
+      "loss": 1.6278,
+      "step": 2772
+    },
+    {
+      "epoch": 0.25671469073656195,
+      "grad_norm": 0.16015625,
+      "learning_rate": 0.02,
+      "loss": 1.6645,
+      "step": 2773
+    },
+    {
+      "epoch": 0.256807267256842,
+      "grad_norm": 0.1513671875,
+      "learning_rate": 0.02,
+      "loss": 1.6599,
+      "step": 2774
+    },
+    {
+      "epoch": 0.25689984377712205,
+      "grad_norm": 0.154296875,
+      "learning_rate": 0.02,
+      "loss": 1.6418,
+      "step": 2775
+    },
+    {
+      "epoch": 0.25699242029740205,
+      "grad_norm": 0.1484375,
+      "learning_rate": 0.02,
+      "loss": 1.6788,
+      "step": 2776
+    },
+    {
+      "epoch": 0.2570849968176821,
+      "grad_norm": 0.154296875,
+      "learning_rate": 0.02,
+      "loss": 1.638,
+      "step": 2777
+    },
+    {
+      "epoch": 0.25717757333796215,
+      "grad_norm": 0.1484375,
+      "learning_rate": 0.02,
+      "loss": 1.6118,
+      "step": 2778
+    },
+    {
+      "epoch": 0.2572701498582422,
+      "grad_norm": 0.1494140625,
+      "learning_rate": 0.02,
+      "loss": 1.6631,
+      "step": 2779
+    },
+    {
+      "epoch": 0.25736272637852226,
+      "grad_norm": 0.15234375,
+      "learning_rate": 0.02,
+      "loss": 1.6593,
+      "step": 2780
+    },
+    {
+      "epoch": 0.2574553028988023,
+      "grad_norm": 0.1572265625,
+      "learning_rate": 0.02,
+      "loss": 1.7023,
+      "step": 2781
+    },
+    {
+      "epoch": 0.25754787941908236,
+      "grad_norm": 0.1484375,
+      "learning_rate": 0.02,
+      "loss": 1.6811,
+      "step": 2782
+    },
+    {
+      "epoch": 0.25764045593936236,
+      "grad_norm": 0.142578125,
+      "learning_rate": 0.02,
+      "loss": 1.6432,
+      "step": 2783
+    },
+    {
+      "epoch": 0.2577330324596424,
+      "grad_norm": 0.150390625,
+      "learning_rate": 0.02,
+      "loss": 1.663,
+      "step": 2784
+    },
+    {
+      "epoch": 0.25782560897992246,
+      "grad_norm": 0.142578125,
+      "learning_rate": 0.02,
+      "loss": 1.5793,
+      "step": 2785
+    },
+    {
+      "epoch": 0.2579181855002025,
+      "grad_norm": 0.1494140625,
+      "learning_rate": 0.02,
+      "loss": 1.6421,
+      "step": 2786
+    },
+    {
+      "epoch": 0.25801076202048256,
+      "grad_norm": 0.150390625,
+      "learning_rate": 0.02,
+      "loss": 1.6715,
+      "step": 2787
+    },
+    {
+      "epoch": 0.2581033385407626,
+      "grad_norm": 0.1396484375,
+      "learning_rate": 0.02,
+      "loss": 1.7028,
+      "step": 2788
+    },
+    {
+      "epoch": 0.25819591506104267,
+      "grad_norm": 0.150390625,
+      "learning_rate": 0.02,
+      "loss": 1.5634,
+      "step": 2789
+    },
+    {
+      "epoch": 0.25828849158132267,
+      "grad_norm": 0.154296875,
+      "learning_rate": 0.02,
+      "loss": 1.7299,
+      "step": 2790
+    },
+    {
+      "epoch": 0.2583810681016027,
+      "grad_norm": 0.1455078125,
+      "learning_rate": 0.02,
+      "loss": 1.6353,
+      "step": 2791
+    },
+    {
+      "epoch": 0.25847364462188277,
+      "grad_norm": 0.1396484375,
+      "learning_rate": 0.02,
+      "loss": 1.6583,
+      "step": 2792
+    },
+    {
+      "epoch": 0.2585662211421628,
+      "grad_norm": 0.13671875,
+      "learning_rate": 0.02,
+      "loss": 1.6632,
+      "step": 2793
+    },
+    {
+      "epoch": 0.2586587976624429,
+      "grad_norm": 0.1474609375,
+      "learning_rate": 0.02,
+      "loss": 1.7041,
+      "step": 2794
+    },
+    {
+      "epoch": 0.2587513741827229,
+      "grad_norm": 0.1474609375,
+      "learning_rate": 0.02,
+      "loss": 1.6851,
+      "step": 2795
+    },
+    {
+      "epoch": 0.258843950703003,
+      "grad_norm": 0.1416015625,
+      "learning_rate": 0.02,
+      "loss": 1.6089,
+      "step": 2796
+    },
+    {
+      "epoch": 0.258936527223283,
+      "grad_norm": 0.154296875,
+      "learning_rate": 0.02,
+      "loss": 1.6764,
+      "step": 2797
+    },
+    {
+      "epoch": 0.259029103743563,
+      "grad_norm": 0.1455078125,
+      "learning_rate": 0.02,
+      "loss": 1.5725,
+      "step": 2798
+    },
+    {
+      "epoch": 0.2591216802638431,
+      "grad_norm": 0.1513671875,
+      "learning_rate": 0.02,
+      "loss": 1.6125,
+      "step": 2799
+    },
+    {
+      "epoch": 0.25921425678412313,
+      "grad_norm": 0.154296875,
+      "learning_rate": 0.02,
+      "loss": 1.6711,
+      "step": 2800
+    },
+    {
+      "epoch": 0.2593068333044032,
+      "grad_norm": 0.154296875,
+      "learning_rate": 0.02,
+      "loss": 1.6687,
+      "step": 2801
+    },
+    {
+      "epoch": 0.25939940982468324,
+      "grad_norm": 0.154296875,
+      "learning_rate": 0.02,
+      "loss": 1.6241,
+      "step": 2802
+    },
+    {
+      "epoch": 0.25949198634496323,
+      "grad_norm": 0.1484375,
+      "learning_rate": 0.02,
+      "loss": 1.6478,
+      "step": 2803
+    },
+    {
+      "epoch": 0.2595845628652433,
+      "grad_norm": 0.1455078125,
+      "learning_rate": 0.02,
+      "loss": 1.6141,
+      "step": 2804
+    },
+    {
+      "epoch": 0.25967713938552334,
+      "grad_norm": 0.1513671875,
+      "learning_rate": 0.02,
+      "loss": 1.6976,
+      "step": 2805
+    },
+    {
+      "epoch": 0.2597697159058034,
+      "grad_norm": 0.1484375,
+      "learning_rate": 0.02,
+      "loss": 1.6385,
+      "step": 2806
+    },
+    {
+      "epoch": 0.25986229242608344,
+      "grad_norm": 0.15234375,
+      "learning_rate": 0.02,
+      "loss": 1.6946,
+      "step": 2807
+    },
+    {
+      "epoch": 0.2599548689463635,
+      "grad_norm": 0.1484375,
+      "learning_rate": 0.02,
+      "loss": 1.6303,
+      "step": 2808
+    },
+    {
+      "epoch": 0.26004744546664355,
+      "grad_norm": 0.1591796875,
+      "learning_rate": 0.02,
+      "loss": 1.7219,
+      "step": 2809
+    },
+    {
+      "epoch": 0.26014002198692354,
+      "grad_norm": 0.158203125,
+      "learning_rate": 0.02,
+      "loss": 1.6885,
+      "step": 2810
+    },
+    {
+      "epoch": 0.2602325985072036,
+      "grad_norm": 0.1513671875,
+      "learning_rate": 0.02,
+      "loss": 1.6945,
+      "step": 2811
+    },
+    {
+      "epoch": 0.26032517502748365,
+      "grad_norm": 0.1435546875,
+      "learning_rate": 0.02,
+      "loss": 1.6883,
+      "step": 2812
+    },
+    {
+      "epoch": 0.2604177515477637,
+      "grad_norm": 0.15625,
+      "learning_rate": 0.02,
+      "loss": 1.7024,
+      "step": 2813
+    },
+    {
+      "epoch": 0.26051032806804375,
+      "grad_norm": 0.154296875,
+      "learning_rate": 0.02,
+      "loss": 1.7098,
+      "step": 2814
+    },
+    {
+      "epoch": 0.2606029045883238,
+      "grad_norm": 0.1455078125,
+      "learning_rate": 0.02,
+      "loss": 1.6763,
+      "step": 2815
+    },
+    {
+      "epoch": 0.26069548110860385,
+      "grad_norm": 0.146484375,
+      "learning_rate": 0.02,
+      "loss": 1.6422,
+      "step": 2816
+    },
+    {
+      "epoch": 0.26078805762888385,
+      "grad_norm": 0.140625,
+      "learning_rate": 0.02,
+      "loss": 1.6549,
+      "step": 2817
+    },
+    {
+      "epoch": 0.2608806341491639,
+      "grad_norm": 0.1474609375,
+      "learning_rate": 0.02,
+      "loss": 1.6597,
+      "step": 2818
+    },
+    {
+      "epoch": 0.26097321066944396,
+      "grad_norm": 0.1435546875,
+      "learning_rate": 0.02,
+      "loss": 1.632,
+      "step": 2819
+    },
+    {
+      "epoch": 0.261065787189724,
+      "grad_norm": 0.1494140625,
+      "learning_rate": 0.02,
+      "loss": 1.7053,
+      "step": 2820
+    },
+    {
+      "epoch": 0.26115836371000406,
+      "grad_norm": 0.140625,
+      "learning_rate": 0.02,
+      "loss": 1.6298,
+      "step": 2821
+    },
+    {
+      "epoch": 0.2612509402302841,
+      "grad_norm": 0.1435546875,
+      "learning_rate": 0.02,
+      "loss": 1.6617,
+      "step": 2822
+    },
+    {
+      "epoch": 0.26134351675056416,
+      "grad_norm": 0.1455078125,
+      "learning_rate": 0.02,
+      "loss": 1.6142,
+      "step": 2823
+    },
+    {
+      "epoch": 0.26143609327084416,
+      "grad_norm": 0.146484375,
+      "learning_rate": 0.02,
+      "loss": 1.6591,
+      "step": 2824
+    },
+    {
+      "epoch": 0.2615286697911242,
+      "grad_norm": 0.1396484375,
+      "learning_rate": 0.02,
+      "loss": 1.6943,
+      "step": 2825
+    },
+    {
+      "epoch": 0.26162124631140427,
+      "grad_norm": 0.14453125,
+      "learning_rate": 0.02,
+      "loss": 1.6459,
+      "step": 2826
+    },
+    {
+      "epoch": 0.2617138228316843,
+      "grad_norm": 0.1494140625,
+      "learning_rate": 0.02,
+      "loss": 1.7123,
+      "step": 2827
+    },
+    {
+      "epoch": 0.26180639935196437,
+      "grad_norm": 0.1552734375,
+      "learning_rate": 0.02,
+      "loss": 1.7169,
+      "step": 2828
+    },
+    {
+      "epoch": 0.2618989758722444,
+      "grad_norm": 0.1435546875,
+      "learning_rate": 0.02,
+      "loss": 1.6423,
+      "step": 2829
+    },
+    {
+      "epoch": 0.2619915523925244,
+      "grad_norm": 0.134765625,
+      "learning_rate": 0.02,
+      "loss": 1.6434,
+      "step": 2830
+    },
+    {
+      "epoch": 0.26208412891280447,
+      "grad_norm": 0.1494140625,
+      "learning_rate": 0.02,
+      "loss": 1.6682,
+      "step": 2831
+    },
+    {
+      "epoch": 0.2621767054330845,
+      "grad_norm": 0.1552734375,
+      "learning_rate": 0.02,
+      "loss": 1.6019,
+      "step": 2832
+    },
+    {
+      "epoch": 0.2622692819533646,
+      "grad_norm": 0.1552734375,
+      "learning_rate": 0.02,
+      "loss": 1.6233,
+      "step": 2833
+    },
+    {
+      "epoch": 0.2623618584736446,
+      "grad_norm": 0.1552734375,
+      "learning_rate": 0.02,
+      "loss": 1.6967,
+      "step": 2834
+    },
+    {
+      "epoch": 0.2624544349939247,
+      "grad_norm": 0.158203125,
+      "learning_rate": 0.02,
+      "loss": 1.6257,
+      "step": 2835
+    },
+    {
+      "epoch": 0.26254701151420473,
+      "grad_norm": 0.15234375,
+      "learning_rate": 0.02,
+      "loss": 1.6696,
+      "step": 2836
+    },
+    {
+      "epoch": 0.26263958803448473,
+      "grad_norm": 0.158203125,
+      "learning_rate": 0.02,
+      "loss": 1.6953,
+      "step": 2837
+    },
+    {
+      "epoch": 0.2627321645547648,
+      "grad_norm": 0.1474609375,
+      "learning_rate": 0.02,
+      "loss": 1.6865,
+      "step": 2838
+    },
+    {
+      "epoch": 0.26282474107504483,
+      "grad_norm": 0.1513671875,
+      "learning_rate": 0.02,
+      "loss": 1.6295,
+      "step": 2839
+    },
+    {
+      "epoch": 0.2629173175953249,
+      "grad_norm": 0.1591796875,
+      "learning_rate": 0.02,
+      "loss": 1.6418,
+      "step": 2840
+    },
+    {
+      "epoch": 0.26300989411560494,
+      "grad_norm": 0.1572265625,
+      "learning_rate": 0.02,
+      "loss": 1.6713,
+      "step": 2841
+    },
+    {
+      "epoch": 0.263102470635885,
+      "grad_norm": 0.1474609375,
+      "learning_rate": 0.02,
+      "loss": 1.6951,
+      "step": 2842
+    },
+    {
+      "epoch": 0.26319504715616504,
+      "grad_norm": 0.1474609375,
+      "learning_rate": 0.02,
+      "loss": 1.6727,
+      "step": 2843
+    },
+    {
+      "epoch": 0.26328762367644504,
+      "grad_norm": 0.146484375,
+      "learning_rate": 0.02,
+      "loss": 1.5912,
+      "step": 2844
+    },
+    {
+      "epoch": 0.2633802001967251,
+      "grad_norm": 0.1533203125,
+      "learning_rate": 0.02,
+      "loss": 1.6671,
+      "step": 2845
+    },
+    {
+      "epoch": 0.26347277671700514,
+      "grad_norm": 0.1533203125,
+      "learning_rate": 0.02,
+      "loss": 1.6898,
+      "step": 2846
+    },
+    {
+      "epoch": 0.2635653532372852,
+      "grad_norm": 0.1591796875,
+      "learning_rate": 0.02,
+      "loss": 1.6778,
+      "step": 2847
+    },
+    {
+      "epoch": 0.26365792975756525,
+      "grad_norm": 0.1484375,
+      "learning_rate": 0.02,
+      "loss": 1.6392,
+      "step": 2848
+    },
+    {
+      "epoch": 0.2637505062778453,
+      "grad_norm": 0.1591796875,
+      "learning_rate": 0.02,
+      "loss": 1.6483,
+      "step": 2849
+    },
+    {
+      "epoch": 0.26384308279812535,
+      "grad_norm": 0.154296875,
+      "learning_rate": 0.02,
+      "loss": 1.6028,
+      "step": 2850
+    },
+    {
+      "epoch": 0.26393565931840535,
+      "grad_norm": 0.16015625,
+      "learning_rate": 0.02,
+      "loss": 1.5759,
+      "step": 2851
+    },
+    {
+      "epoch": 0.2640282358386854,
+      "grad_norm": 0.16015625,
+      "learning_rate": 0.02,
+      "loss": 1.6332,
+      "step": 2852
+    },
+    {
+      "epoch": 0.26412081235896545,
+      "grad_norm": 0.15625,
+      "learning_rate": 0.02,
+      "loss": 1.6672,
+      "step": 2853
+    },
+    {
+      "epoch": 0.2642133888792455,
+      "grad_norm": 0.1611328125,
+      "learning_rate": 0.02,
+      "loss": 1.6469,
+      "step": 2854
+    },
+    {
+      "epoch": 0.26430596539952556,
+      "grad_norm": 0.1513671875,
+      "learning_rate": 0.02,
+      "loss": 1.7041,
+      "step": 2855
+    },
+    {
+      "epoch": 0.2643985419198056,
+      "grad_norm": 0.1494140625,
+      "learning_rate": 0.02,
+      "loss": 1.6714,
+      "step": 2856
+    },
+    {
+      "epoch": 0.26449111844008566,
+      "grad_norm": 0.142578125,
+      "learning_rate": 0.02,
+      "loss": 1.5934,
+      "step": 2857
+    },
+    {
+      "epoch": 0.26458369496036566,
+      "grad_norm": 0.15234375,
+      "learning_rate": 0.02,
+      "loss": 1.6739,
+      "step": 2858
+    },
+    {
+      "epoch": 0.2646762714806457,
+      "grad_norm": 0.1474609375,
+      "learning_rate": 0.02,
+      "loss": 1.6576,
+      "step": 2859
+    },
+    {
+      "epoch": 0.26476884800092576,
+      "grad_norm": 0.154296875,
+      "learning_rate": 0.02,
+      "loss": 1.6805,
+      "step": 2860
+    },
+    {
+      "epoch": 0.2648614245212058,
+      "grad_norm": 0.1416015625,
+      "learning_rate": 0.02,
+      "loss": 1.6645,
+      "step": 2861
+    },
+    {
+      "epoch": 0.26495400104148586,
+      "grad_norm": 0.15234375,
+      "learning_rate": 0.02,
+      "loss": 1.6594,
+      "step": 2862
+    },
+    {
+      "epoch": 0.2650465775617659,
+      "grad_norm": 0.146484375,
+      "learning_rate": 0.02,
+      "loss": 1.6264,
+      "step": 2863
+    },
+    {
+      "epoch": 0.2651391540820459,
+      "grad_norm": 0.1455078125,
+      "learning_rate": 0.02,
+      "loss": 1.6439,
+      "step": 2864
+    },
+    {
+      "epoch": 0.26523173060232597,
+      "grad_norm": 0.1396484375,
+      "learning_rate": 0.02,
+      "loss": 1.6258,
+      "step": 2865
+    },
+    {
+      "epoch": 0.265324307122606,
+      "grad_norm": 0.150390625,
+      "learning_rate": 0.02,
+      "loss": 1.6453,
+      "step": 2866
+    },
+    {
+      "epoch": 0.26541688364288607,
+      "grad_norm": 0.150390625,
+      "learning_rate": 0.02,
+      "loss": 1.643,
+      "step": 2867
+    },
+    {
+      "epoch": 0.2655094601631661,
+      "grad_norm": 0.1474609375,
+      "learning_rate": 0.02,
+      "loss": 1.7105,
+      "step": 2868
+    },
+    {
+      "epoch": 0.2656020366834462,
+      "grad_norm": 0.1591796875,
+      "learning_rate": 0.02,
+      "loss": 1.6835,
+      "step": 2869
+    },
+    {
+      "epoch": 0.2656946132037262,
+      "grad_norm": 0.1484375,
+      "learning_rate": 0.02,
+      "loss": 1.6739,
+      "step": 2870
+    },
+    {
+      "epoch": 0.2657871897240062,
+      "grad_norm": 0.154296875,
+      "learning_rate": 0.02,
+      "loss": 1.6796,
+      "step": 2871
+    },
+    {
+      "epoch": 0.2658797662442863,
+      "grad_norm": 0.146484375,
+      "learning_rate": 0.02,
+      "loss": 1.6223,
+      "step": 2872
+    },
+    {
+      "epoch": 0.2659723427645663,
+      "grad_norm": 0.1455078125,
+      "learning_rate": 0.02,
+      "loss": 1.6662,
+      "step": 2873
+    },
+    {
+      "epoch": 0.2660649192848464,
+      "grad_norm": 0.1455078125,
+      "learning_rate": 0.02,
+      "loss": 1.6816,
+      "step": 2874
+    },
+    {
+      "epoch": 0.26615749580512643,
+      "grad_norm": 0.1513671875,
+      "learning_rate": 0.02,
+      "loss": 1.6126,
+      "step": 2875
+    },
+    {
+      "epoch": 0.2662500723254065,
+      "grad_norm": 0.1572265625,
+      "learning_rate": 0.02,
+      "loss": 1.719,
+      "step": 2876
+    },
+    {
+      "epoch": 0.26634264884568654,
+      "grad_norm": 0.15234375,
+      "learning_rate": 0.02,
+      "loss": 1.639,
+      "step": 2877
+    },
+    {
+      "epoch": 0.26643522536596653,
+      "grad_norm": 0.154296875,
+      "learning_rate": 0.02,
+      "loss": 1.6367,
+      "step": 2878
+    },
+    {
+      "epoch": 0.2665278018862466,
+      "grad_norm": 0.1494140625,
+      "learning_rate": 0.02,
+      "loss": 1.5886,
+      "step": 2879
+    },
+    {
+      "epoch": 0.26662037840652664,
+      "grad_norm": 0.154296875,
+      "learning_rate": 0.02,
+      "loss": 1.5996,
+      "step": 2880
+    },
+    {
+      "epoch": 0.2667129549268067,
+      "grad_norm": 0.1474609375,
+      "learning_rate": 0.02,
+      "loss": 1.6856,
+      "step": 2881
+    },
+    {
+      "epoch": 0.26680553144708674,
+      "grad_norm": 0.16015625,
+      "learning_rate": 0.02,
+      "loss": 1.7201,
+      "step": 2882
+    },
+    {
+      "epoch": 0.2668981079673668,
+      "grad_norm": 0.1474609375,
+      "learning_rate": 0.02,
+      "loss": 1.6602,
+      "step": 2883
+    },
+    {
+      "epoch": 0.26699068448764685,
+      "grad_norm": 0.1611328125,
+      "learning_rate": 0.02,
+      "loss": 1.7035,
+      "step": 2884
+    },
+    {
+      "epoch": 0.26708326100792684,
+      "grad_norm": 0.15625,
+      "learning_rate": 0.02,
+      "loss": 1.6811,
+      "step": 2885
+    },
+    {
+      "epoch": 0.2671758375282069,
+      "grad_norm": 0.16015625,
+      "learning_rate": 0.02,
+      "loss": 1.6841,
+      "step": 2886
+    },
+    {
+      "epoch": 0.26726841404848695,
+      "grad_norm": 0.1435546875,
+      "learning_rate": 0.02,
+      "loss": 1.6108,
+      "step": 2887
+    },
+    {
+      "epoch": 0.267360990568767,
+      "grad_norm": 0.15625,
+      "learning_rate": 0.02,
+      "loss": 1.7174,
+      "step": 2888
+    },
+    {
+      "epoch": 0.26745356708904705,
+      "grad_norm": 0.1435546875,
+      "learning_rate": 0.02,
+      "loss": 1.5924,
+      "step": 2889
+    },
+    {
+      "epoch": 0.2675461436093271,
+      "grad_norm": 0.1513671875,
+      "learning_rate": 0.02,
+      "loss": 1.6595,
+      "step": 2890
+    },
+    {
+      "epoch": 0.26763872012960715,
+      "grad_norm": 0.1513671875,
+      "learning_rate": 0.02,
+      "loss": 1.6713,
+      "step": 2891
+    },
+    {
+      "epoch": 0.26773129664988715,
+      "grad_norm": 0.162109375,
+      "learning_rate": 0.02,
+      "loss": 1.7279,
+      "step": 2892
+    },
+    {
+      "epoch": 0.2678238731701672,
+      "grad_norm": 0.1455078125,
+      "learning_rate": 0.02,
+      "loss": 1.6218,
+      "step": 2893
+    },
+    {
+      "epoch": 0.26791644969044726,
+      "grad_norm": 0.158203125,
+      "learning_rate": 0.02,
+      "loss": 1.6724,
+      "step": 2894
+    },
+    {
+      "epoch": 0.2680090262107273,
+      "grad_norm": 0.1474609375,
+      "learning_rate": 0.02,
+      "loss": 1.7069,
+      "step": 2895
+    },
+    {
+      "epoch": 0.26810160273100736,
+      "grad_norm": 0.1494140625,
+      "learning_rate": 0.02,
+      "loss": 1.6208,
+      "step": 2896
+    },
+    {
+      "epoch": 0.2681941792512874,
+      "grad_norm": 0.1396484375,
+      "learning_rate": 0.02,
+      "loss": 1.691,
+      "step": 2897
+    },
+    {
+      "epoch": 0.2682867557715674,
+      "grad_norm": 0.1591796875,
+      "learning_rate": 0.02,
+      "loss": 1.7264,
+      "step": 2898
+    },
+    {
+      "epoch": 0.26837933229184746,
+      "grad_norm": 0.1455078125,
+      "learning_rate": 0.02,
+      "loss": 1.6605,
+      "step": 2899
+    },
+    {
+      "epoch": 0.2684719088121275,
+      "grad_norm": 0.150390625,
+      "learning_rate": 0.02,
+      "loss": 1.6441,
+      "step": 2900
+    },
+    {
+      "epoch": 0.26856448533240757,
+      "grad_norm": 0.1494140625,
+      "learning_rate": 0.02,
+      "loss": 1.6617,
+      "step": 2901
+    },
+    {
+      "epoch": 0.2686570618526876,
+      "grad_norm": 0.1552734375,
+      "learning_rate": 0.02,
+      "loss": 1.6914,
+      "step": 2902
+    },
+    {
+      "epoch": 0.26874963837296767,
+      "grad_norm": 0.1513671875,
+      "learning_rate": 0.02,
+      "loss": 1.6822,
+      "step": 2903
+    },
+    {
+      "epoch": 0.2688422148932477,
+      "grad_norm": 0.1376953125,
+      "learning_rate": 0.02,
+      "loss": 1.5771,
+      "step": 2904
+    },
+    {
+      "epoch": 0.2689347914135277,
+      "grad_norm": 0.146484375,
+      "learning_rate": 0.02,
+      "loss": 1.6515,
+      "step": 2905
+    },
+    {
+      "epoch": 0.26902736793380777,
+      "grad_norm": 0.1494140625,
+      "learning_rate": 0.02,
+      "loss": 1.6529,
+      "step": 2906
+    },
+    {
+      "epoch": 0.2691199444540878,
+      "grad_norm": 0.1474609375,
+      "learning_rate": 0.02,
+      "loss": 1.639,
+      "step": 2907
+    },
+    {
+      "epoch": 0.2692125209743679,
+      "grad_norm": 0.1435546875,
+      "learning_rate": 0.02,
+      "loss": 1.6697,
+      "step": 2908
+    },
+    {
+      "epoch": 0.2693050974946479,
+      "grad_norm": 0.1396484375,
+      "learning_rate": 0.02,
+      "loss": 1.6447,
+      "step": 2909
+    },
+    {
+      "epoch": 0.269397674014928,
+      "grad_norm": 0.142578125,
+      "learning_rate": 0.02,
+      "loss": 1.6552,
+      "step": 2910
+    },
+    {
+      "epoch": 0.26949025053520803,
+      "grad_norm": 0.1494140625,
+      "learning_rate": 0.02,
+      "loss": 1.7034,
+      "step": 2911
+    },
+    {
+      "epoch": 0.26958282705548803,
+      "grad_norm": 0.1484375,
+      "learning_rate": 0.02,
+      "loss": 1.7098,
+      "step": 2912
+    },
+    {
+      "epoch": 0.2696754035757681,
+      "grad_norm": 0.1494140625,
+      "learning_rate": 0.02,
+      "loss": 1.7295,
+      "step": 2913
+    },
+    {
+      "epoch": 0.26976798009604813,
+      "grad_norm": 0.1474609375,
+      "learning_rate": 0.02,
+      "loss": 1.6588,
+      "step": 2914
+    },
+    {
+      "epoch": 0.2698605566163282,
+      "grad_norm": 0.138671875,
+      "learning_rate": 0.02,
+      "loss": 1.6701,
+      "step": 2915
+    },
+    {
+      "epoch": 0.26995313313660824,
+      "grad_norm": 0.154296875,
+      "learning_rate": 0.02,
+      "loss": 1.6751,
+      "step": 2916
+    },
+    {
+      "epoch": 0.2700457096568883,
+      "grad_norm": 0.1513671875,
+      "learning_rate": 0.02,
+      "loss": 1.6817,
+      "step": 2917
+    },
+    {
+      "epoch": 0.27013828617716834,
+      "grad_norm": 0.1494140625,
+      "learning_rate": 0.02,
+      "loss": 1.7306,
+      "step": 2918
+    },
+    {
+      "epoch": 0.27023086269744834,
+      "grad_norm": 0.142578125,
+      "learning_rate": 0.02,
+      "loss": 1.6562,
+      "step": 2919
+    },
+    {
+      "epoch": 0.2703234392177284,
+      "grad_norm": 0.142578125,
+      "learning_rate": 0.02,
+      "loss": 1.5626,
+      "step": 2920
+    },
+    {
+      "epoch": 0.27041601573800844,
+      "grad_norm": 0.1416015625,
+      "learning_rate": 0.02,
+      "loss": 1.6434,
+      "step": 2921
+    },
+    {
+      "epoch": 0.2705085922582885,
+      "grad_norm": 0.1513671875,
+      "learning_rate": 0.02,
+      "loss": 1.6845,
+      "step": 2922
+    },
+    {
+      "epoch": 0.27060116877856855,
+      "grad_norm": 0.1435546875,
+      "learning_rate": 0.02,
+      "loss": 1.6769,
+      "step": 2923
+    },
+    {
+      "epoch": 0.2706937452988486,
+      "grad_norm": 0.146484375,
+      "learning_rate": 0.02,
+      "loss": 1.6294,
+      "step": 2924
+    },
+    {
+      "epoch": 0.27078632181912865,
+      "grad_norm": 0.1435546875,
+      "learning_rate": 0.02,
+      "loss": 1.6379,
+      "step": 2925
+    },
+    {
+      "epoch": 0.27087889833940865,
+      "grad_norm": 0.142578125,
+      "learning_rate": 0.02,
+      "loss": 1.6709,
+      "step": 2926
+    },
+    {
+      "epoch": 0.2709714748596887,
+      "grad_norm": 0.1357421875,
+      "learning_rate": 0.02,
+      "loss": 1.586,
+      "step": 2927
+    },
+    {
+      "epoch": 0.27106405137996875,
+      "grad_norm": 0.14453125,
+      "learning_rate": 0.02,
+      "loss": 1.5888,
+      "step": 2928
+    },
+    {
+      "epoch": 0.2711566279002488,
+      "grad_norm": 0.142578125,
+      "learning_rate": 0.02,
+      "loss": 1.587,
+      "step": 2929
+    },
+    {
+      "epoch": 0.27124920442052886,
+      "grad_norm": 0.146484375,
+      "learning_rate": 0.02,
+      "loss": 1.6487,
+      "step": 2930
+    },
+    {
+      "epoch": 0.2713417809408089,
+      "grad_norm": 0.1474609375,
+      "learning_rate": 0.02,
+      "loss": 1.6153,
+      "step": 2931
+    },
+    {
+      "epoch": 0.2714343574610889,
+      "grad_norm": 0.16015625,
+      "learning_rate": 0.02,
+      "loss": 1.6689,
+      "step": 2932
+    },
+    {
+      "epoch": 0.27152693398136896,
+      "grad_norm": 0.1669921875,
+      "learning_rate": 0.02,
+      "loss": 1.6482,
+      "step": 2933
+    },
+    {
+      "epoch": 0.271619510501649,
+      "grad_norm": 0.150390625,
+      "learning_rate": 0.02,
+      "loss": 1.7232,
+      "step": 2934
+    },
+    {
+      "epoch": 0.27171208702192906,
+      "grad_norm": 0.14453125,
+      "learning_rate": 0.02,
+      "loss": 1.6562,
+      "step": 2935
+    },
+    {
+      "epoch": 0.2718046635422091,
+      "grad_norm": 0.1552734375,
+      "learning_rate": 0.02,
+      "loss": 1.6476,
+      "step": 2936
+    },
+    {
+      "epoch": 0.27189724006248916,
+      "grad_norm": 0.16015625,
+      "learning_rate": 0.02,
+      "loss": 1.6312,
+      "step": 2937
+    },
+    {
+      "epoch": 0.2719898165827692,
+      "grad_norm": 0.15234375,
+      "learning_rate": 0.02,
+      "loss": 1.7164,
+      "step": 2938
+    },
+    {
+      "epoch": 0.2720823931030492,
+      "grad_norm": 0.1533203125,
+      "learning_rate": 0.02,
+      "loss": 1.6385,
+      "step": 2939
+    },
+    {
+      "epoch": 0.27217496962332927,
+      "grad_norm": 0.1455078125,
+      "learning_rate": 0.02,
+      "loss": 1.6087,
+      "step": 2940
+    },
+    {
+      "epoch": 0.2722675461436093,
+      "grad_norm": 0.1494140625,
+      "learning_rate": 0.02,
+      "loss": 1.6188,
+      "step": 2941
+    },
+    {
+      "epoch": 0.27236012266388937,
+      "grad_norm": 0.142578125,
+      "learning_rate": 0.02,
+      "loss": 1.6092,
+      "step": 2942
+    },
+    {
+      "epoch": 0.2724526991841694,
+      "grad_norm": 0.14453125,
+      "learning_rate": 0.02,
+      "loss": 1.6976,
+      "step": 2943
+    },
+    {
+      "epoch": 0.2725452757044495,
+      "grad_norm": 0.1416015625,
+      "learning_rate": 0.02,
+      "loss": 1.6015,
+      "step": 2944
+    },
+    {
+      "epoch": 0.2726378522247295,
+      "grad_norm": 0.154296875,
+      "learning_rate": 0.02,
+      "loss": 1.6333,
+      "step": 2945
+    },
+    {
+      "epoch": 0.2727304287450095,
+      "grad_norm": 0.158203125,
+      "learning_rate": 0.02,
+      "loss": 1.7133,
+      "step": 2946
+    },
+    {
+      "epoch": 0.2728230052652896,
+      "grad_norm": 0.1533203125,
+      "learning_rate": 0.02,
+      "loss": 1.6359,
+      "step": 2947
+    },
+    {
+      "epoch": 0.2729155817855696,
+      "grad_norm": 0.15625,
+      "learning_rate": 0.02,
+      "loss": 1.5624,
+      "step": 2948
+    },
+    {
+      "epoch": 0.2730081583058497,
+      "grad_norm": 0.154296875,
+      "learning_rate": 0.02,
+      "loss": 1.6915,
+      "step": 2949
+    },
+    {
+      "epoch": 0.27310073482612973,
+      "grad_norm": 0.154296875,
+      "learning_rate": 0.02,
+      "loss": 1.647,
+      "step": 2950
+    },
+    {
+      "epoch": 0.2731933113464098,
+      "grad_norm": 0.158203125,
+      "learning_rate": 0.02,
+      "loss": 1.7084,
+      "step": 2951
+    },
+    {
+      "epoch": 0.2731933113464098,
+      "eval_loss": 1.6467796564102173,
+      "eval_runtime": 63.8421,
+      "eval_samples_per_second": 23.871,
+      "eval_steps_per_second": 5.968,
+      "step": 2951
+    },
+    {
+      "epoch": 0.27328588786668984,
+      "grad_norm": 0.1552734375,
+      "learning_rate": 0.02,
+      "loss": 1.6319,
+      "step": 2952
+    },
+    {
+      "epoch": 0.27337846438696983,
+      "grad_norm": 0.1494140625,
+      "learning_rate": 0.02,
+      "loss": 1.6934,
+      "step": 2953
+    },
+    {
+      "epoch": 0.2734710409072499,
+      "grad_norm": 0.1494140625,
+      "learning_rate": 0.02,
+      "loss": 1.6256,
+      "step": 2954
+    },
+    {
+      "epoch": 0.27356361742752994,
+      "grad_norm": 0.1376953125,
+      "learning_rate": 0.02,
+      "loss": 1.5788,
+      "step": 2955
+    },
+    {
+      "epoch": 0.27365619394781,
+      "grad_norm": 0.1484375,
+      "learning_rate": 0.02,
+      "loss": 1.6994,
+      "step": 2956
+    },
+    {
+      "epoch": 0.27374877046809004,
+      "grad_norm": 0.15234375,
+      "learning_rate": 0.02,
+      "loss": 1.7362,
+      "step": 2957
+    },
+    {
+      "epoch": 0.2738413469883701,
+      "grad_norm": 0.150390625,
+      "learning_rate": 0.02,
+      "loss": 1.6726,
+      "step": 2958
+    },
+    {
+      "epoch": 0.27393392350865015,
+      "grad_norm": 0.1494140625,
+      "learning_rate": 0.02,
+      "loss": 1.6046,
+      "step": 2959
+    },
+    {
+      "epoch": 0.27402650002893014,
+      "grad_norm": 0.1494140625,
+      "learning_rate": 0.02,
+      "loss": 1.6937,
+      "step": 2960
+    },
+    {
+      "epoch": 0.2741190765492102,
+      "grad_norm": 0.1513671875,
+      "learning_rate": 0.02,
+      "loss": 1.6249,
+      "step": 2961
+    },
+    {
+      "epoch": 0.27421165306949025,
+      "grad_norm": 0.154296875,
+      "learning_rate": 0.02,
+      "loss": 1.6794,
+      "step": 2962
+    },
+    {
+      "epoch": 0.2743042295897703,
+      "grad_norm": 0.1416015625,
+      "learning_rate": 0.02,
+      "loss": 1.6063,
+      "step": 2963
+    },
+    {
+      "epoch": 0.27439680611005035,
+      "grad_norm": 0.1484375,
+      "learning_rate": 0.02,
+      "loss": 1.6664,
+      "step": 2964
+    },
+    {
+      "epoch": 0.2744893826303304,
+      "grad_norm": 0.1474609375,
+      "learning_rate": 0.02,
+      "loss": 1.6308,
+      "step": 2965
+    },
+    {
+      "epoch": 0.2745819591506104,
+      "grad_norm": 0.146484375,
+      "learning_rate": 0.02,
+      "loss": 1.6903,
+      "step": 2966
+    },
+    {
+      "epoch": 0.27467453567089045,
+      "grad_norm": 0.1416015625,
+      "learning_rate": 0.02,
+      "loss": 1.5371,
+      "step": 2967
+    },
+    {
+      "epoch": 0.2747671121911705,
+      "grad_norm": 0.1552734375,
+      "learning_rate": 0.02,
+      "loss": 1.7016,
+      "step": 2968
+    },
+    {
+      "epoch": 0.27485968871145056,
+      "grad_norm": 0.1533203125,
+      "learning_rate": 0.02,
+      "loss": 1.6229,
+      "step": 2969
+    },
+    {
+      "epoch": 0.2749522652317306,
+      "grad_norm": 0.1513671875,
+      "learning_rate": 0.02,
+      "loss": 1.6649,
+      "step": 2970
+    },
+    {
+      "epoch": 0.27504484175201066,
+      "grad_norm": 0.158203125,
+      "learning_rate": 0.02,
+      "loss": 1.672,
+      "step": 2971
+    },
+    {
+      "epoch": 0.2751374182722907,
+      "grad_norm": 0.15234375,
+      "learning_rate": 0.02,
+      "loss": 1.6233,
+      "step": 2972
+    },
+    {
+      "epoch": 0.2752299947925707,
+      "grad_norm": 0.142578125,
+      "learning_rate": 0.02,
+      "loss": 1.6205,
+      "step": 2973
+    },
+    {
+      "epoch": 0.27532257131285076,
+      "grad_norm": 0.1494140625,
+      "learning_rate": 0.02,
+      "loss": 1.582,
+      "step": 2974
+    },
+    {
+      "epoch": 0.2754151478331308,
+      "grad_norm": 0.146484375,
+      "learning_rate": 0.02,
+      "loss": 1.6705,
+      "step": 2975
+    },
+    {
+      "epoch": 0.27550772435341087,
+      "grad_norm": 0.15625,
+      "learning_rate": 0.02,
+      "loss": 1.6853,
+      "step": 2976
+    },
+    {
+      "epoch": 0.2756003008736909,
+      "grad_norm": 0.158203125,
+      "learning_rate": 0.02,
+      "loss": 1.6507,
+      "step": 2977
+    },
+    {
+      "epoch": 0.27569287739397097,
+      "grad_norm": 0.1484375,
+      "learning_rate": 0.02,
+      "loss": 1.6084,
+      "step": 2978
+    },
+    {
+      "epoch": 0.275785453914251,
+      "grad_norm": 0.146484375,
+      "learning_rate": 0.02,
+      "loss": 1.639,
+      "step": 2979
+    },
+    {
+      "epoch": 0.275878030434531,
+      "grad_norm": 0.154296875,
+      "learning_rate": 0.02,
+      "loss": 1.6382,
+      "step": 2980
+    },
+    {
+      "epoch": 0.27597060695481107,
+      "grad_norm": 0.1416015625,
+      "learning_rate": 0.02,
+      "loss": 1.5851,
+      "step": 2981
+    },
+    {
+      "epoch": 0.2760631834750911,
+      "grad_norm": 0.16015625,
+      "learning_rate": 0.02,
+      "loss": 1.6638,
+      "step": 2982
+    },
+    {
+      "epoch": 0.2761557599953712,
+      "grad_norm": 0.14453125,
+      "learning_rate": 0.02,
+      "loss": 1.5929,
+      "step": 2983
+    },
+    {
+      "epoch": 0.2762483365156512,
+      "grad_norm": 0.154296875,
+      "learning_rate": 0.02,
+      "loss": 1.6806,
+      "step": 2984
+    },
+    {
+      "epoch": 0.2763409130359313,
+      "grad_norm": 0.154296875,
+      "learning_rate": 0.02,
+      "loss": 1.6414,
+      "step": 2985
+    },
+    {
+      "epoch": 0.27643348955621133,
+      "grad_norm": 0.1552734375,
+      "learning_rate": 0.02,
+      "loss": 1.6779,
+      "step": 2986
+    },
+    {
+      "epoch": 0.27652606607649133,
+      "grad_norm": 0.15234375,
+      "learning_rate": 0.02,
+      "loss": 1.6147,
+      "step": 2987
+    },
+    {
+      "epoch": 0.2766186425967714,
+      "grad_norm": 0.169921875,
+      "learning_rate": 0.02,
+      "loss": 1.6612,
+      "step": 2988
+    },
+    {
+      "epoch": 0.27671121911705143,
+      "grad_norm": 0.1689453125,
+      "learning_rate": 0.02,
+      "loss": 1.6038,
+      "step": 2989
+    },
+    {
+      "epoch": 0.2768037956373315,
+      "grad_norm": 0.1572265625,
+      "learning_rate": 0.02,
+      "loss": 1.6406,
+      "step": 2990
+    },
+    {
+      "epoch": 0.27689637215761154,
+      "grad_norm": 0.1474609375,
+      "learning_rate": 0.02,
+      "loss": 1.6711,
+      "step": 2991
+    },
+    {
+      "epoch": 0.2769889486778916,
+      "grad_norm": 0.1552734375,
+      "learning_rate": 0.02,
+      "loss": 1.6493,
+      "step": 2992
+    },
+    {
+      "epoch": 0.27708152519817164,
+      "grad_norm": 0.15234375,
+      "learning_rate": 0.02,
+      "loss": 1.598,
+      "step": 2993
+    },
+    {
+      "epoch": 0.27717410171845164,
+      "grad_norm": 0.150390625,
+      "learning_rate": 0.02,
+      "loss": 1.6276,
+      "step": 2994
+    },
+    {
+      "epoch": 0.2772666782387317,
+      "grad_norm": 0.142578125,
+      "learning_rate": 0.02,
+      "loss": 1.6419,
+      "step": 2995
+    },
+    {
+      "epoch": 0.27735925475901174,
+      "grad_norm": 0.146484375,
+      "learning_rate": 0.02,
+      "loss": 1.6313,
+      "step": 2996
+    },
+    {
+      "epoch": 0.2774518312792918,
+      "grad_norm": 0.154296875,
+      "learning_rate": 0.02,
+      "loss": 1.6995,
+      "step": 2997
+    },
+    {
+      "epoch": 0.27754440779957185,
+      "grad_norm": 0.1591796875,
+      "learning_rate": 0.02,
+      "loss": 1.6983,
+      "step": 2998
+    },
+    {
+      "epoch": 0.2776369843198519,
+      "grad_norm": 0.146484375,
+      "learning_rate": 0.02,
+      "loss": 1.6808,
+      "step": 2999
+    },
+    {
+      "epoch": 0.2777295608401319,
+      "grad_norm": 0.1552734375,
+      "learning_rate": 0.02,
+      "loss": 1.6811,
+      "step": 3000
+    },
+    {
+      "epoch": 0.27782213736041195,
+      "grad_norm": 0.16015625,
+      "learning_rate": 0.02,
+      "loss": 1.6762,
+      "step": 3001
+    },
+    {
+      "epoch": 0.277914713880692,
+      "grad_norm": 0.1416015625,
+      "learning_rate": 0.02,
+      "loss": 1.668,
+      "step": 3002
+    },
+    {
+      "epoch": 0.27800729040097205,
+      "grad_norm": 0.1455078125,
+      "learning_rate": 0.02,
+      "loss": 1.6282,
+      "step": 3003
+    },
+    {
+      "epoch": 0.2780998669212521,
+      "grad_norm": 0.13671875,
+      "learning_rate": 0.02,
+      "loss": 1.5678,
+      "step": 3004
+    },
+    {
+      "epoch": 0.27819244344153216,
+      "grad_norm": 0.1484375,
+      "learning_rate": 0.02,
+      "loss": 1.6592,
+      "step": 3005
+    },
+    {
+      "epoch": 0.2782850199618122,
+      "grad_norm": 0.1455078125,
+      "learning_rate": 0.02,
+      "loss": 1.6368,
+      "step": 3006
+    },
+    {
+      "epoch": 0.2783775964820922,
+      "grad_norm": 0.15234375,
+      "learning_rate": 0.02,
+      "loss": 1.5944,
+      "step": 3007
+    },
+    {
+      "epoch": 0.27847017300237226,
+      "grad_norm": 0.1474609375,
+      "learning_rate": 0.02,
+      "loss": 1.6043,
+      "step": 3008
+    },
+    {
+      "epoch": 0.2785627495226523,
+      "grad_norm": 0.154296875,
+      "learning_rate": 0.02,
+      "loss": 1.6701,
+      "step": 3009
+    },
+    {
+      "epoch": 0.27865532604293236,
+      "grad_norm": 0.1552734375,
+      "learning_rate": 0.02,
+      "loss": 1.6936,
+      "step": 3010
+    },
+    {
+      "epoch": 0.2787479025632124,
+      "grad_norm": 0.1591796875,
+      "learning_rate": 0.02,
+      "loss": 1.7021,
+      "step": 3011
+    },
+    {
+      "epoch": 0.27884047908349247,
+      "grad_norm": 0.150390625,
+      "learning_rate": 0.02,
+      "loss": 1.6176,
+      "step": 3012
+    },
+    {
+      "epoch": 0.2789330556037725,
+      "grad_norm": 0.1533203125,
+      "learning_rate": 0.02,
+      "loss": 1.6846,
+      "step": 3013
+    },
+    {
+      "epoch": 0.2790256321240525,
+      "grad_norm": 0.138671875,
+      "learning_rate": 0.02,
+      "loss": 1.6148,
+      "step": 3014
+    },
+    {
+      "epoch": 0.27911820864433257,
+      "grad_norm": 0.14453125,
+      "learning_rate": 0.02,
+      "loss": 1.605,
+      "step": 3015
+    },
+    {
+      "epoch": 0.2792107851646126,
+      "grad_norm": 0.1474609375,
+      "learning_rate": 0.02,
+      "loss": 1.6793,
+      "step": 3016
+    },
+    {
+      "epoch": 0.27930336168489267,
+      "grad_norm": 0.1455078125,
+      "learning_rate": 0.02,
+      "loss": 1.696,
+      "step": 3017
+    },
+    {
+      "epoch": 0.2793959382051727,
+      "grad_norm": 0.14453125,
+      "learning_rate": 0.02,
+      "loss": 1.6339,
+      "step": 3018
+    },
+    {
+      "epoch": 0.2794885147254528,
+      "grad_norm": 0.1435546875,
+      "learning_rate": 0.02,
+      "loss": 1.67,
+      "step": 3019
+    },
+    {
+      "epoch": 0.2795810912457328,
+      "grad_norm": 0.1494140625,
+      "learning_rate": 0.02,
+      "loss": 1.6844,
+      "step": 3020
+    },
+    {
+      "epoch": 0.2796736677660128,
+      "grad_norm": 0.1533203125,
+      "learning_rate": 0.02,
+      "loss": 1.6333,
+      "step": 3021
+    },
+    {
+      "epoch": 0.2797662442862929,
+      "grad_norm": 0.15234375,
+      "learning_rate": 0.02,
+      "loss": 1.7006,
+      "step": 3022
+    },
+    {
+      "epoch": 0.2798588208065729,
+      "grad_norm": 0.154296875,
+      "learning_rate": 0.02,
+      "loss": 1.651,
+      "step": 3023
+    },
+    {
+      "epoch": 0.279951397326853,
+      "grad_norm": 0.1416015625,
+      "learning_rate": 0.02,
+      "loss": 1.647,
+      "step": 3024
+    },
+    {
+      "epoch": 0.28004397384713303,
+      "grad_norm": 0.1533203125,
+      "learning_rate": 0.02,
+      "loss": 1.6122,
+      "step": 3025
+    },
+    {
+      "epoch": 0.2801365503674131,
+      "grad_norm": 0.1435546875,
+      "learning_rate": 0.02,
+      "loss": 1.6672,
+      "step": 3026
+    },
+    {
+      "epoch": 0.28022912688769314,
+      "grad_norm": 0.150390625,
+      "learning_rate": 0.02,
+      "loss": 1.661,
+      "step": 3027
+    },
+    {
+      "epoch": 0.28032170340797313,
+      "grad_norm": 0.150390625,
+      "learning_rate": 0.02,
+      "loss": 1.6573,
+      "step": 3028
+    },
+    {
+      "epoch": 0.2804142799282532,
+      "grad_norm": 0.15234375,
+      "learning_rate": 0.02,
+      "loss": 1.6587,
+      "step": 3029
+    },
+    {
+      "epoch": 0.28050685644853324,
+      "grad_norm": 0.1337890625,
+      "learning_rate": 0.02,
+      "loss": 1.5925,
+      "step": 3030
+    },
+    {
+      "epoch": 0.2805994329688133,
+      "grad_norm": 0.134765625,
+      "learning_rate": 0.02,
+      "loss": 1.6513,
+      "step": 3031
+    },
+    {
+      "epoch": 0.28069200948909334,
+      "grad_norm": 0.140625,
+      "learning_rate": 0.02,
+      "loss": 1.6461,
+      "step": 3032
+    },
+    {
+      "epoch": 0.2807845860093734,
+      "grad_norm": 0.140625,
+      "learning_rate": 0.02,
+      "loss": 1.6294,
+      "step": 3033
+    },
+    {
+      "epoch": 0.2808771625296534,
+      "grad_norm": 0.1357421875,
+      "learning_rate": 0.02,
+      "loss": 1.5613,
+      "step": 3034
+    },
+    {
+      "epoch": 0.28096973904993344,
+      "grad_norm": 0.1396484375,
+      "learning_rate": 0.02,
+      "loss": 1.6022,
+      "step": 3035
+    },
+    {
+      "epoch": 0.2810623155702135,
+      "grad_norm": 0.1484375,
+      "learning_rate": 0.02,
+      "loss": 1.6439,
+      "step": 3036
+    },
+    {
+      "epoch": 0.28115489209049355,
+      "grad_norm": 0.154296875,
+      "learning_rate": 0.02,
+      "loss": 1.7109,
+      "step": 3037
+    },
+    {
+      "epoch": 0.2812474686107736,
+      "grad_norm": 0.142578125,
+      "learning_rate": 0.02,
+      "loss": 1.7053,
+      "step": 3038
+    },
+    {
+      "epoch": 0.28134004513105365,
+      "grad_norm": 0.1455078125,
+      "learning_rate": 0.02,
+      "loss": 1.6403,
+      "step": 3039
+    },
+    {
+      "epoch": 0.2814326216513337,
+      "grad_norm": 0.1513671875,
+      "learning_rate": 0.02,
+      "loss": 1.6558,
+      "step": 3040
+    },
+    {
+      "epoch": 0.2815251981716137,
+      "grad_norm": 0.146484375,
+      "learning_rate": 0.02,
+      "loss": 1.661,
+      "step": 3041
+    },
+    {
+      "epoch": 0.28161777469189375,
+      "grad_norm": 0.140625,
+      "learning_rate": 0.02,
+      "loss": 1.6084,
+      "step": 3042
+    },
+    {
+      "epoch": 0.2817103512121738,
+      "grad_norm": 0.140625,
+      "learning_rate": 0.02,
+      "loss": 1.6369,
+      "step": 3043
+    },
+    {
+      "epoch": 0.28180292773245386,
+      "grad_norm": 0.1416015625,
+      "learning_rate": 0.02,
+      "loss": 1.7114,
+      "step": 3044
+    },
+    {
+      "epoch": 0.2818955042527339,
+      "grad_norm": 0.1474609375,
+      "learning_rate": 0.02,
+      "loss": 1.7034,
+      "step": 3045
+    },
+    {
+      "epoch": 0.28198808077301396,
+      "grad_norm": 0.1513671875,
+      "learning_rate": 0.02,
+      "loss": 1.6974,
+      "step": 3046
+    },
+    {
+      "epoch": 0.282080657293294,
+      "grad_norm": 0.1572265625,
+      "learning_rate": 0.02,
+      "loss": 1.5933,
+      "step": 3047
+    },
+    {
+      "epoch": 0.282173233813574,
+      "grad_norm": 0.1474609375,
+      "learning_rate": 0.02,
+      "loss": 1.5932,
+      "step": 3048
+    },
+    {
+      "epoch": 0.28226581033385406,
+      "grad_norm": 0.146484375,
+      "learning_rate": 0.02,
+      "loss": 1.6345,
+      "step": 3049
+    },
+    {
+      "epoch": 0.2823583868541341,
+      "grad_norm": 0.1474609375,
+      "learning_rate": 0.02,
+      "loss": 1.607,
+      "step": 3050
+    },
+    {
+      "epoch": 0.28245096337441417,
+      "grad_norm": 0.1572265625,
+      "learning_rate": 0.02,
+      "loss": 1.7024,
+      "step": 3051
+    },
+    {
+      "epoch": 0.2825435398946942,
+      "grad_norm": 0.15625,
+      "learning_rate": 0.02,
+      "loss": 1.637,
+      "step": 3052
+    },
+    {
+      "epoch": 0.28263611641497427,
+      "grad_norm": 0.15625,
+      "learning_rate": 0.02,
+      "loss": 1.6633,
+      "step": 3053
+    },
+    {
+      "epoch": 0.2827286929352543,
+      "grad_norm": 0.1474609375,
+      "learning_rate": 0.02,
+      "loss": 1.6791,
+      "step": 3054
+    },
+    {
+      "epoch": 0.2828212694555343,
+      "grad_norm": 0.1435546875,
+      "learning_rate": 0.02,
+      "loss": 1.6469,
+      "step": 3055
+    },
+    {
+      "epoch": 0.28291384597581437,
+      "grad_norm": 0.15234375,
+      "learning_rate": 0.02,
+      "loss": 1.6856,
+      "step": 3056
+    },
+    {
+      "epoch": 0.2830064224960944,
+      "grad_norm": 0.154296875,
+      "learning_rate": 0.02,
+      "loss": 1.647,
+      "step": 3057
+    },
+    {
+      "epoch": 0.2830989990163745,
+      "grad_norm": 0.154296875,
+      "learning_rate": 0.02,
+      "loss": 1.6691,
+      "step": 3058
+    },
+    {
+      "epoch": 0.2831915755366545,
+      "grad_norm": 0.1435546875,
+      "learning_rate": 0.02,
+      "loss": 1.5618,
+      "step": 3059
+    },
+    {
+      "epoch": 0.2832841520569346,
+      "grad_norm": 0.15234375,
+      "learning_rate": 0.02,
+      "loss": 1.7262,
+      "step": 3060
+    },
+    {
+      "epoch": 0.28337672857721463,
+      "grad_norm": 0.154296875,
+      "learning_rate": 0.02,
+      "loss": 1.651,
+      "step": 3061
+    },
+    {
+      "epoch": 0.28346930509749463,
+      "grad_norm": 0.1474609375,
+      "learning_rate": 0.02,
+      "loss": 1.6853,
+      "step": 3062
+    },
+    {
+      "epoch": 0.2835618816177747,
+      "grad_norm": 0.154296875,
+      "learning_rate": 0.02,
+      "loss": 1.7013,
+      "step": 3063
+    },
+    {
+      "epoch": 0.28365445813805473,
+      "grad_norm": 0.154296875,
+      "learning_rate": 0.02,
+      "loss": 1.6417,
+      "step": 3064
+    },
+    {
+      "epoch": 0.2837470346583348,
+      "grad_norm": 0.1416015625,
+      "learning_rate": 0.02,
+      "loss": 1.645,
+      "step": 3065
+    },
+    {
+      "epoch": 0.28383961117861484,
+      "grad_norm": 0.1484375,
+      "learning_rate": 0.02,
+      "loss": 1.688,
+      "step": 3066
+    },
+    {
+      "epoch": 0.2839321876988949,
+      "grad_norm": 0.1513671875,
+      "learning_rate": 0.02,
+      "loss": 1.6017,
+      "step": 3067
+    },
+    {
+      "epoch": 0.2840247642191749,
+      "grad_norm": 0.1435546875,
+      "learning_rate": 0.02,
+      "loss": 1.7023,
+      "step": 3068
+    },
+    {
+      "epoch": 0.28411734073945494,
+      "grad_norm": 0.1455078125,
+      "learning_rate": 0.02,
+      "loss": 1.6343,
+      "step": 3069
+    },
+    {
+      "epoch": 0.284209917259735,
+      "grad_norm": 0.16015625,
+      "learning_rate": 0.02,
+      "loss": 1.7089,
+      "step": 3070
+    },
+    {
+      "epoch": 0.28430249378001504,
+      "grad_norm": 0.15625,
+      "learning_rate": 0.02,
+      "loss": 1.608,
+      "step": 3071
+    },
+    {
+      "epoch": 0.2843950703002951,
+      "grad_norm": 0.16015625,
+      "learning_rate": 0.02,
+      "loss": 1.5908,
+      "step": 3072
+    },
+    {
+      "epoch": 0.28448764682057515,
+      "grad_norm": 0.162109375,
+      "learning_rate": 0.02,
+      "loss": 1.6105,
+      "step": 3073
+    },
+    {
+      "epoch": 0.2845802233408552,
+      "grad_norm": 0.1591796875,
+      "learning_rate": 0.02,
+      "loss": 1.6281,
+      "step": 3074
+    },
+    {
+      "epoch": 0.2846727998611352,
+      "grad_norm": 0.1689453125,
+      "learning_rate": 0.02,
+      "loss": 1.6848,
+      "step": 3075
+    },
+    {
+      "epoch": 0.28476537638141525,
+      "grad_norm": 0.16796875,
+      "learning_rate": 0.02,
+      "loss": 1.6712,
+      "step": 3076
+    },
+    {
+      "epoch": 0.2848579529016953,
+      "grad_norm": 0.158203125,
+      "learning_rate": 0.02,
+      "loss": 1.6009,
+      "step": 3077
+    },
+    {
+      "epoch": 0.28495052942197535,
+      "grad_norm": 0.162109375,
+      "learning_rate": 0.02,
+      "loss": 1.706,
+      "step": 3078
+    },
+    {
+      "epoch": 0.2850431059422554,
+      "grad_norm": 0.15625,
+      "learning_rate": 0.02,
+      "loss": 1.639,
+      "step": 3079
+    },
+    {
+      "epoch": 0.28513568246253546,
+      "grad_norm": 0.1513671875,
+      "learning_rate": 0.02,
+      "loss": 1.6045,
+      "step": 3080
+    },
+    {
+      "epoch": 0.2852282589828155,
+      "grad_norm": 0.14453125,
+      "learning_rate": 0.02,
+      "loss": 1.5797,
+      "step": 3081
+    },
+    {
+      "epoch": 0.2853208355030955,
+      "grad_norm": 0.158203125,
+      "learning_rate": 0.02,
+      "loss": 1.666,
+      "step": 3082
+    },
+    {
+      "epoch": 0.28541341202337556,
+      "grad_norm": 0.1611328125,
+      "learning_rate": 0.02,
+      "loss": 1.6524,
+      "step": 3083
+    },
+    {
+      "epoch": 0.2855059885436556,
+      "grad_norm": 0.154296875,
+      "learning_rate": 0.02,
+      "loss": 1.6488,
+      "step": 3084
+    },
+    {
+      "epoch": 0.28559856506393566,
+      "grad_norm": 0.1552734375,
+      "learning_rate": 0.02,
+      "loss": 1.6292,
+      "step": 3085
+    },
+    {
+      "epoch": 0.2856911415842157,
+      "grad_norm": 0.1494140625,
+      "learning_rate": 0.02,
+      "loss": 1.6197,
+      "step": 3086
+    },
+    {
+      "epoch": 0.28578371810449577,
+      "grad_norm": 0.1533203125,
+      "learning_rate": 0.02,
+      "loss": 1.6344,
+      "step": 3087
+    },
+    {
+      "epoch": 0.2858762946247758,
+      "grad_norm": 0.166015625,
+      "learning_rate": 0.02,
+      "loss": 1.6809,
+      "step": 3088
+    },
+    {
+      "epoch": 0.2859688711450558,
+      "grad_norm": 0.1591796875,
+      "learning_rate": 0.02,
+      "loss": 1.6324,
+      "step": 3089
+    },
+    {
+      "epoch": 0.28606144766533587,
+      "grad_norm": 0.1416015625,
+      "learning_rate": 0.02,
+      "loss": 1.6053,
+      "step": 3090
+    },
+    {
+      "epoch": 0.2861540241856159,
+      "grad_norm": 0.1572265625,
+      "learning_rate": 0.02,
+      "loss": 1.6379,
+      "step": 3091
+    },
+    {
+      "epoch": 0.28624660070589597,
+      "grad_norm": 0.1396484375,
+      "learning_rate": 0.02,
+      "loss": 1.5912,
+      "step": 3092
+    },
+    {
+      "epoch": 0.286339177226176,
+      "grad_norm": 0.1494140625,
+      "learning_rate": 0.02,
+      "loss": 1.6259,
+      "step": 3093
+    },
+    {
+      "epoch": 0.2864317537464561,
+      "grad_norm": 0.1435546875,
+      "learning_rate": 0.02,
+      "loss": 1.6026,
+      "step": 3094
+    },
+    {
+      "epoch": 0.28652433026673607,
+      "grad_norm": 0.1484375,
+      "learning_rate": 0.02,
+      "loss": 1.6866,
+      "step": 3095
+    },
+    {
+      "epoch": 0.2866169067870161,
+      "grad_norm": 0.166015625,
+      "learning_rate": 0.02,
+      "loss": 1.5995,
+      "step": 3096
+    },
+    {
+      "epoch": 0.2867094833072962,
+      "grad_norm": 0.154296875,
+      "learning_rate": 0.02,
+      "loss": 1.6426,
+      "step": 3097
+    },
+    {
+      "epoch": 0.28680205982757623,
+      "grad_norm": 0.15234375,
+      "learning_rate": 0.02,
+      "loss": 1.6621,
+      "step": 3098
+    },
+    {
+      "epoch": 0.2868946363478563,
+      "grad_norm": 0.1572265625,
+      "learning_rate": 0.02,
+      "loss": 1.5918,
+      "step": 3099
+    },
+    {
+      "epoch": 0.28698721286813633,
+      "grad_norm": 0.1416015625,
+      "learning_rate": 0.02,
+      "loss": 1.585,
+      "step": 3100
+    },
+    {
+      "epoch": 0.2870797893884164,
+      "grad_norm": 0.1484375,
+      "learning_rate": 0.02,
+      "loss": 1.6172,
+      "step": 3101
+    },
+    {
+      "epoch": 0.2871723659086964,
+      "grad_norm": 0.1630859375,
+      "learning_rate": 0.02,
+      "loss": 1.7021,
+      "step": 3102
+    },
+    {
+      "epoch": 0.28726494242897643,
+      "grad_norm": 0.1513671875,
+      "learning_rate": 0.02,
+      "loss": 1.6727,
+      "step": 3103
+    },
+    {
+      "epoch": 0.2873575189492565,
+      "grad_norm": 0.138671875,
+      "learning_rate": 0.02,
+      "loss": 1.5781,
+      "step": 3104
+    },
+    {
+      "epoch": 0.28745009546953654,
+      "grad_norm": 0.138671875,
+      "learning_rate": 0.02,
+      "loss": 1.6349,
+      "step": 3105
+    },
+    {
+      "epoch": 0.2875426719898166,
+      "grad_norm": 0.1474609375,
+      "learning_rate": 0.02,
+      "loss": 1.6374,
+      "step": 3106
+    },
+    {
+      "epoch": 0.28763524851009664,
+      "grad_norm": 0.142578125,
+      "learning_rate": 0.02,
+      "loss": 1.5979,
+      "step": 3107
+    },
+    {
+      "epoch": 0.2877278250303767,
+      "grad_norm": 0.15234375,
+      "learning_rate": 0.02,
+      "loss": 1.6611,
+      "step": 3108
+    },
+    {
+      "epoch": 0.2878204015506567,
+      "grad_norm": 0.14453125,
+      "learning_rate": 0.02,
+      "loss": 1.6233,
+      "step": 3109
+    },
+    {
+      "epoch": 0.28791297807093674,
+      "grad_norm": 0.1435546875,
+      "learning_rate": 0.02,
+      "loss": 1.634,
+      "step": 3110
+    },
+    {
+      "epoch": 0.2880055545912168,
+      "grad_norm": 0.1435546875,
+      "learning_rate": 0.02,
+      "loss": 1.658,
+      "step": 3111
+    },
+    {
+      "epoch": 0.28809813111149685,
+      "grad_norm": 0.1396484375,
+      "learning_rate": 0.02,
+      "loss": 1.6028,
+      "step": 3112
+    },
+    {
+      "epoch": 0.2881907076317769,
+      "grad_norm": 0.1357421875,
+      "learning_rate": 0.02,
+      "loss": 1.5755,
+      "step": 3113
+    },
+    {
+      "epoch": 0.28828328415205695,
+      "grad_norm": 0.150390625,
+      "learning_rate": 0.02,
+      "loss": 1.6498,
+      "step": 3114
+    },
+    {
+      "epoch": 0.288375860672337,
+      "grad_norm": 0.15234375,
+      "learning_rate": 0.02,
+      "loss": 1.6294,
+      "step": 3115
+    },
+    {
+      "epoch": 0.288468437192617,
+      "grad_norm": 0.1474609375,
+      "learning_rate": 0.02,
+      "loss": 1.6725,
+      "step": 3116
+    },
+    {
+      "epoch": 0.28856101371289705,
+      "grad_norm": 0.142578125,
+      "learning_rate": 0.02,
+      "loss": 1.5937,
+      "step": 3117
+    },
+    {
+      "epoch": 0.2886535902331771,
+      "grad_norm": 0.1396484375,
+      "learning_rate": 0.02,
+      "loss": 1.6173,
+      "step": 3118
+    },
+    {
+      "epoch": 0.28874616675345716,
+      "grad_norm": 0.1474609375,
+      "learning_rate": 0.02,
+      "loss": 1.7205,
+      "step": 3119
+    },
+    {
+      "epoch": 0.2888387432737372,
+      "grad_norm": 0.1455078125,
+      "learning_rate": 0.02,
+      "loss": 1.6108,
+      "step": 3120
+    },
+    {
+      "epoch": 0.28893131979401726,
+      "grad_norm": 0.1357421875,
+      "learning_rate": 0.02,
+      "loss": 1.5598,
+      "step": 3121
+    },
+    {
+      "epoch": 0.2890238963142973,
+      "grad_norm": 0.138671875,
+      "learning_rate": 0.02,
+      "loss": 1.6216,
+      "step": 3122
+    },
+    {
+      "epoch": 0.2891164728345773,
+      "grad_norm": 0.1494140625,
+      "learning_rate": 0.02,
+      "loss": 1.6773,
+      "step": 3123
+    },
+    {
+      "epoch": 0.28920904935485736,
+      "grad_norm": 0.1484375,
+      "learning_rate": 0.02,
+      "loss": 1.7219,
+      "step": 3124
+    },
+    {
+      "epoch": 0.2893016258751374,
+      "grad_norm": 0.146484375,
+      "learning_rate": 0.02,
+      "loss": 1.664,
+      "step": 3125
+    },
+    {
+      "epoch": 0.28939420239541747,
+      "grad_norm": 0.1455078125,
+      "learning_rate": 0.02,
+      "loss": 1.5667,
+      "step": 3126
+    },
+    {
+      "epoch": 0.2894867789156975,
+      "grad_norm": 0.140625,
+      "learning_rate": 0.02,
+      "loss": 1.6766,
+      "step": 3127
+    },
+    {
+      "epoch": 0.28957935543597757,
+      "grad_norm": 0.1337890625,
+      "learning_rate": 0.02,
+      "loss": 1.6073,
+      "step": 3128
+    },
+    {
+      "epoch": 0.28967193195625757,
+      "grad_norm": 0.142578125,
+      "learning_rate": 0.02,
+      "loss": 1.6359,
+      "step": 3129
+    },
+    {
+      "epoch": 0.2897645084765376,
+      "grad_norm": 0.1435546875,
+      "learning_rate": 0.02,
+      "loss": 1.6336,
+      "step": 3130
+    },
+    {
+      "epoch": 0.28985708499681767,
+      "grad_norm": 0.1484375,
+      "learning_rate": 0.02,
+      "loss": 1.6282,
+      "step": 3131
+    },
+    {
+      "epoch": 0.2899496615170977,
+      "grad_norm": 0.1484375,
+      "learning_rate": 0.02,
+      "loss": 1.6412,
+      "step": 3132
+    },
+    {
+      "epoch": 0.2900422380373778,
+      "grad_norm": 0.1640625,
+      "learning_rate": 0.02,
+      "loss": 1.5947,
+      "step": 3133
+    },
+    {
+      "epoch": 0.2901348145576578,
+      "grad_norm": 0.162109375,
+      "learning_rate": 0.02,
+      "loss": 1.6327,
+      "step": 3134
+    },
+    {
+      "epoch": 0.2902273910779379,
+      "grad_norm": 0.16796875,
+      "learning_rate": 0.02,
+      "loss": 1.6913,
+      "step": 3135
+    },
+    {
+      "epoch": 0.2903199675982179,
+      "grad_norm": 0.150390625,
+      "learning_rate": 0.02,
+      "loss": 1.6276,
+      "step": 3136
+    },
+    {
+      "epoch": 0.29041254411849793,
+      "grad_norm": 0.1611328125,
+      "learning_rate": 0.02,
+      "loss": 1.6735,
+      "step": 3137
+    },
+    {
+      "epoch": 0.290505120638778,
+      "grad_norm": 0.15625,
+      "learning_rate": 0.02,
+      "loss": 1.6666,
+      "step": 3138
+    },
+    {
+      "epoch": 0.29059769715905803,
+      "grad_norm": 0.158203125,
+      "learning_rate": 0.02,
+      "loss": 1.6725,
+      "step": 3139
+    },
+    {
+      "epoch": 0.2906902736793381,
+      "grad_norm": 0.15234375,
+      "learning_rate": 0.02,
+      "loss": 1.655,
+      "step": 3140
+    },
+    {
+      "epoch": 0.29078285019961814,
+      "grad_norm": 0.16796875,
+      "learning_rate": 0.02,
+      "loss": 1.6524,
+      "step": 3141
+    },
+    {
+      "epoch": 0.2908754267198982,
+      "grad_norm": 0.162109375,
+      "learning_rate": 0.02,
+      "loss": 1.6581,
+      "step": 3142
+    },
+    {
+      "epoch": 0.2909680032401782,
+      "grad_norm": 0.1611328125,
+      "learning_rate": 0.02,
+      "loss": 1.6657,
+      "step": 3143
+    },
+    {
+      "epoch": 0.29106057976045824,
+      "grad_norm": 0.1572265625,
+      "learning_rate": 0.02,
+      "loss": 1.6058,
+      "step": 3144
+    },
+    {
+      "epoch": 0.2911531562807383,
+      "grad_norm": 0.150390625,
+      "learning_rate": 0.02,
+      "loss": 1.6796,
+      "step": 3145
+    },
+    {
+      "epoch": 0.29124573280101834,
+      "grad_norm": 0.1435546875,
+      "learning_rate": 0.02,
+      "loss": 1.6344,
+      "step": 3146
+    },
+    {
+      "epoch": 0.2913383093212984,
+      "grad_norm": 0.1572265625,
+      "learning_rate": 0.02,
+      "loss": 1.65,
+      "step": 3147
+    },
+    {
+      "epoch": 0.29143088584157845,
+      "grad_norm": 0.14453125,
+      "learning_rate": 0.02,
+      "loss": 1.6311,
+      "step": 3148
+    },
+    {
+      "epoch": 0.2915234623618585,
+      "grad_norm": 0.142578125,
+      "learning_rate": 0.02,
+      "loss": 1.6832,
+      "step": 3149
+    },
+    {
+      "epoch": 0.2916160388821385,
+      "grad_norm": 0.1318359375,
+      "learning_rate": 0.02,
+      "loss": 1.5998,
+      "step": 3150
+    },
+    {
+      "epoch": 0.29170861540241855,
+      "grad_norm": 0.142578125,
+      "learning_rate": 0.02,
+      "loss": 1.6324,
+      "step": 3151
+    },
+    {
+      "epoch": 0.2918011919226986,
+      "grad_norm": 0.14453125,
+      "learning_rate": 0.02,
+      "loss": 1.7262,
+      "step": 3152
+    },
+    {
+      "epoch": 0.29189376844297865,
+      "grad_norm": 0.1435546875,
+      "learning_rate": 0.02,
+      "loss": 1.6261,
+      "step": 3153
+    },
+    {
+      "epoch": 0.2919863449632587,
+      "grad_norm": 0.1494140625,
+      "learning_rate": 0.02,
+      "loss": 1.6936,
+      "step": 3154
+    },
+    {
+      "epoch": 0.29207892148353876,
+      "grad_norm": 0.1484375,
+      "learning_rate": 0.02,
+      "loss": 1.7054,
+      "step": 3155
+    },
+    {
+      "epoch": 0.2921714980038188,
+      "grad_norm": 0.1435546875,
+      "learning_rate": 0.02,
+      "loss": 1.5872,
+      "step": 3156
+    },
+    {
+      "epoch": 0.2922640745240988,
+      "grad_norm": 0.146484375,
+      "learning_rate": 0.02,
+      "loss": 1.6888,
+      "step": 3157
+    },
+    {
+      "epoch": 0.29235665104437886,
+      "grad_norm": 0.1494140625,
+      "learning_rate": 0.02,
+      "loss": 1.6674,
+      "step": 3158
+    },
+    {
+      "epoch": 0.2924492275646589,
+      "grad_norm": 0.142578125,
+      "learning_rate": 0.02,
+      "loss": 1.6291,
+      "step": 3159
+    },
+    {
+      "epoch": 0.29254180408493896,
+      "grad_norm": 0.14453125,
+      "learning_rate": 0.02,
+      "loss": 1.5985,
+      "step": 3160
+    },
+    {
+      "epoch": 0.292634380605219,
+      "grad_norm": 0.150390625,
+      "learning_rate": 0.02,
+      "loss": 1.6188,
+      "step": 3161
+    },
+    {
+      "epoch": 0.29272695712549907,
+      "grad_norm": 0.1494140625,
+      "learning_rate": 0.02,
+      "loss": 1.6746,
+      "step": 3162
+    },
+    {
+      "epoch": 0.29281953364577906,
+      "grad_norm": 0.1455078125,
+      "learning_rate": 0.02,
+      "loss": 1.5703,
+      "step": 3163
+    },
+    {
+      "epoch": 0.2929121101660591,
+      "grad_norm": 0.1435546875,
+      "learning_rate": 0.02,
+      "loss": 1.6448,
+      "step": 3164
+    },
+    {
+      "epoch": 0.29300468668633917,
+      "grad_norm": 0.146484375,
+      "learning_rate": 0.02,
+      "loss": 1.6263,
+      "step": 3165
+    },
+    {
+      "epoch": 0.2930972632066192,
+      "grad_norm": 0.142578125,
+      "learning_rate": 0.02,
+      "loss": 1.6676,
+      "step": 3166
+    },
+    {
+      "epoch": 0.29318983972689927,
+      "grad_norm": 0.1474609375,
+      "learning_rate": 0.02,
+      "loss": 1.6766,
+      "step": 3167
+    },
+    {
+      "epoch": 0.2932824162471793,
+      "grad_norm": 0.1357421875,
+      "learning_rate": 0.02,
+      "loss": 1.648,
+      "step": 3168
+    },
+    {
+      "epoch": 0.2933749927674594,
+      "grad_norm": 0.1435546875,
+      "learning_rate": 0.02,
+      "loss": 1.6572,
+      "step": 3169
+    },
+    {
+      "epoch": 0.29346756928773937,
+      "grad_norm": 0.1376953125,
+      "learning_rate": 0.02,
+      "loss": 1.5613,
+      "step": 3170
+    },
+    {
+      "epoch": 0.2935601458080194,
+      "grad_norm": 0.140625,
+      "learning_rate": 0.02,
+      "loss": 1.6178,
+      "step": 3171
+    },
+    {
+      "epoch": 0.2936527223282995,
+      "grad_norm": 0.142578125,
+      "learning_rate": 0.02,
+      "loss": 1.6233,
+      "step": 3172
+    },
+    {
+      "epoch": 0.29374529884857953,
+      "grad_norm": 0.142578125,
+      "learning_rate": 0.02,
+      "loss": 1.6083,
+      "step": 3173
+    },
+    {
+      "epoch": 0.2938378753688596,
+      "grad_norm": 0.146484375,
+      "learning_rate": 0.02,
+      "loss": 1.6699,
+      "step": 3174
+    },
+    {
+      "epoch": 0.29393045188913963,
+      "grad_norm": 0.1455078125,
+      "learning_rate": 0.02,
+      "loss": 1.6405,
+      "step": 3175
+    },
+    {
+      "epoch": 0.2940230284094197,
+      "grad_norm": 0.1533203125,
+      "learning_rate": 0.02,
+      "loss": 1.6556,
+      "step": 3176
+    },
+    {
+      "epoch": 0.2941156049296997,
+      "grad_norm": 0.1396484375,
+      "learning_rate": 0.02,
+      "loss": 1.6218,
+      "step": 3177
+    },
+    {
+      "epoch": 0.29420818144997973,
+      "grad_norm": 0.1376953125,
+      "learning_rate": 0.02,
+      "loss": 1.6585,
+      "step": 3178
+    },
+    {
+      "epoch": 0.2943007579702598,
+      "grad_norm": 0.1474609375,
+      "learning_rate": 0.02,
+      "loss": 1.624,
+      "step": 3179
+    },
+    {
+      "epoch": 0.29439333449053984,
+      "grad_norm": 0.1484375,
+      "learning_rate": 0.02,
+      "loss": 1.6089,
+      "step": 3180
+    },
+    {
+      "epoch": 0.2944859110108199,
+      "grad_norm": 0.1455078125,
+      "learning_rate": 0.02,
+      "loss": 1.596,
+      "step": 3181
+    },
+    {
+      "epoch": 0.29457848753109994,
+      "grad_norm": 0.1455078125,
+      "learning_rate": 0.02,
+      "loss": 1.641,
+      "step": 3182
+    },
+    {
+      "epoch": 0.29467106405138,
+      "grad_norm": 0.1572265625,
+      "learning_rate": 0.02,
+      "loss": 1.6858,
+      "step": 3183
+    },
+    {
+      "epoch": 0.29476364057166,
+      "grad_norm": 0.150390625,
+      "learning_rate": 0.02,
+      "loss": 1.6303,
+      "step": 3184
+    },
+    {
+      "epoch": 0.29485621709194004,
+      "grad_norm": 0.15625,
+      "learning_rate": 0.02,
+      "loss": 1.6397,
+      "step": 3185
+    },
+    {
+      "epoch": 0.2949487936122201,
+      "grad_norm": 0.1484375,
+      "learning_rate": 0.02,
+      "loss": 1.5597,
+      "step": 3186
+    },
+    {
+      "epoch": 0.29504137013250015,
+      "grad_norm": 0.15234375,
+      "learning_rate": 0.02,
+      "loss": 1.6552,
+      "step": 3187
+    },
+    {
+      "epoch": 0.2951339466527802,
+      "grad_norm": 0.1435546875,
+      "learning_rate": 0.02,
+      "loss": 1.6316,
+      "step": 3188
+    },
+    {
+      "epoch": 0.29522652317306025,
+      "grad_norm": 0.1494140625,
+      "learning_rate": 0.02,
+      "loss": 1.6607,
+      "step": 3189
+    },
+    {
+      "epoch": 0.2953190996933403,
+      "grad_norm": 0.150390625,
+      "learning_rate": 0.02,
+      "loss": 1.6265,
+      "step": 3190
+    },
+    {
+      "epoch": 0.2954116762136203,
+      "grad_norm": 0.146484375,
+      "learning_rate": 0.02,
+      "loss": 1.6501,
+      "step": 3191
+    },
+    {
+      "epoch": 0.29550425273390035,
+      "grad_norm": 0.138671875,
+      "learning_rate": 0.02,
+      "loss": 1.6192,
+      "step": 3192
+    },
+    {
+      "epoch": 0.2955968292541804,
+      "grad_norm": 0.1455078125,
+      "learning_rate": 0.02,
+      "loss": 1.6023,
+      "step": 3193
+    },
+    {
+      "epoch": 0.29568940577446046,
+      "grad_norm": 0.1416015625,
+      "learning_rate": 0.02,
+      "loss": 1.5745,
+      "step": 3194
+    },
+    {
+      "epoch": 0.2957819822947405,
+      "grad_norm": 0.158203125,
+      "learning_rate": 0.02,
+      "loss": 1.6576,
+      "step": 3195
+    },
+    {
+      "epoch": 0.29587455881502056,
+      "grad_norm": 0.1533203125,
+      "learning_rate": 0.02,
+      "loss": 1.6259,
+      "step": 3196
+    },
+    {
+      "epoch": 0.29596713533530056,
+      "grad_norm": 0.1435546875,
+      "learning_rate": 0.02,
+      "loss": 1.611,
+      "step": 3197
+    },
+    {
+      "epoch": 0.2960597118555806,
+      "grad_norm": 0.134765625,
+      "learning_rate": 0.02,
+      "loss": 1.6503,
+      "step": 3198
+    },
+    {
+      "epoch": 0.29615228837586066,
+      "grad_norm": 0.146484375,
+      "learning_rate": 0.02,
+      "loss": 1.6313,
+      "step": 3199
+    },
+    {
+      "epoch": 0.2962448648961407,
+      "grad_norm": 0.1396484375,
+      "learning_rate": 0.02,
+      "loss": 1.6734,
+      "step": 3200
+    },
+    {
+      "epoch": 0.29633744141642077,
+      "grad_norm": 0.1416015625,
+      "learning_rate": 0.02,
+      "loss": 1.659,
+      "step": 3201
+    },
+    {
+      "epoch": 0.2964300179367008,
+      "grad_norm": 0.146484375,
+      "learning_rate": 0.02,
+      "loss": 1.6526,
+      "step": 3202
+    },
+    {
+      "epoch": 0.29652259445698087,
+      "grad_norm": 0.142578125,
+      "learning_rate": 0.02,
+      "loss": 1.5782,
+      "step": 3203
+    },
+    {
+      "epoch": 0.29661517097726087,
+      "grad_norm": 0.138671875,
+      "learning_rate": 0.02,
+      "loss": 1.6747,
+      "step": 3204
+    },
+    {
+      "epoch": 0.2967077474975409,
+      "grad_norm": 0.1396484375,
+      "learning_rate": 0.02,
+      "loss": 1.6502,
+      "step": 3205
+    },
+    {
+      "epoch": 0.29680032401782097,
+      "grad_norm": 0.142578125,
+      "learning_rate": 0.02,
+      "loss": 1.6704,
+      "step": 3206
+    },
+    {
+      "epoch": 0.296892900538101,
+      "grad_norm": 0.142578125,
+      "learning_rate": 0.02,
+      "loss": 1.6721,
+      "step": 3207
+    },
+    {
+      "epoch": 0.2969854770583811,
+      "grad_norm": 0.142578125,
+      "learning_rate": 0.02,
+      "loss": 1.6302,
+      "step": 3208
+    },
+    {
+      "epoch": 0.2970780535786611,
+      "grad_norm": 0.14453125,
+      "learning_rate": 0.02,
+      "loss": 1.6616,
+      "step": 3209
+    },
+    {
+      "epoch": 0.2971706300989412,
+      "grad_norm": 0.1484375,
+      "learning_rate": 0.02,
+      "loss": 1.6429,
+      "step": 3210
+    },
+    {
+      "epoch": 0.2972632066192212,
+      "grad_norm": 0.142578125,
+      "learning_rate": 0.02,
+      "loss": 1.6255,
+      "step": 3211
+    },
+    {
+      "epoch": 0.29735578313950123,
+      "grad_norm": 0.1591796875,
+      "learning_rate": 0.02,
+      "loss": 1.6812,
+      "step": 3212
+    },
+    {
+      "epoch": 0.2974483596597813,
+      "grad_norm": 0.1455078125,
+      "learning_rate": 0.02,
+      "loss": 1.6323,
+      "step": 3213
+    },
+    {
+      "epoch": 0.29754093618006133,
+      "grad_norm": 0.1630859375,
+      "learning_rate": 0.02,
+      "loss": 1.696,
+      "step": 3214
+    },
+    {
+      "epoch": 0.2976335127003414,
+      "grad_norm": 0.138671875,
+      "learning_rate": 0.02,
+      "loss": 1.6188,
+      "step": 3215
+    },
+    {
+      "epoch": 0.29772608922062144,
+      "grad_norm": 0.1396484375,
+      "learning_rate": 0.02,
+      "loss": 1.5786,
+      "step": 3216
+    },
+    {
+      "epoch": 0.2978186657409015,
+      "grad_norm": 0.1337890625,
+      "learning_rate": 0.02,
+      "loss": 1.6151,
+      "step": 3217
+    },
+    {
+      "epoch": 0.2979112422611815,
+      "grad_norm": 0.1416015625,
+      "learning_rate": 0.02,
+      "loss": 1.636,
+      "step": 3218
+    },
+    {
+      "epoch": 0.29800381878146154,
+      "grad_norm": 0.1318359375,
+      "learning_rate": 0.02,
+      "loss": 1.628,
+      "step": 3219
+    },
+    {
+      "epoch": 0.2980963953017416,
+      "grad_norm": 0.1416015625,
+      "learning_rate": 0.02,
+      "loss": 1.6125,
+      "step": 3220
+    },
+    {
+      "epoch": 0.29818897182202164,
+      "grad_norm": 0.1298828125,
+      "learning_rate": 0.02,
+      "loss": 1.639,
+      "step": 3221
+    },
+    {
+      "epoch": 0.2982815483423017,
+      "grad_norm": 0.1474609375,
+      "learning_rate": 0.02,
+      "loss": 1.6691,
+      "step": 3222
+    },
+    {
+      "epoch": 0.29837412486258175,
+      "grad_norm": 0.1357421875,
+      "learning_rate": 0.02,
+      "loss": 1.6394,
+      "step": 3223
+    },
+    {
+      "epoch": 0.2984667013828618,
+      "grad_norm": 0.1455078125,
+      "learning_rate": 0.02,
+      "loss": 1.6618,
+      "step": 3224
+    },
+    {
+      "epoch": 0.2985592779031418,
+      "grad_norm": 0.1396484375,
+      "learning_rate": 0.02,
+      "loss": 1.611,
+      "step": 3225
+    },
+    {
+      "epoch": 0.29865185442342185,
+      "grad_norm": 0.1396484375,
+      "learning_rate": 0.02,
+      "loss": 1.6264,
+      "step": 3226
+    },
+    {
+      "epoch": 0.2987444309437019,
+      "grad_norm": 0.14453125,
+      "learning_rate": 0.02,
+      "loss": 1.6753,
+      "step": 3227
+    },
+    {
+      "epoch": 0.29883700746398195,
+      "grad_norm": 0.1494140625,
+      "learning_rate": 0.02,
+      "loss": 1.5888,
+      "step": 3228
+    },
+    {
+      "epoch": 0.298929583984262,
+      "grad_norm": 0.14453125,
+      "learning_rate": 0.02,
+      "loss": 1.6223,
+      "step": 3229
+    },
+    {
+      "epoch": 0.29902216050454206,
+      "grad_norm": 0.14453125,
+      "learning_rate": 0.02,
+      "loss": 1.6428,
+      "step": 3230
+    },
+    {
+      "epoch": 0.29911473702482205,
+      "grad_norm": 0.150390625,
+      "learning_rate": 0.02,
+      "loss": 1.655,
+      "step": 3231
+    },
+    {
+      "epoch": 0.2992073135451021,
+      "grad_norm": 0.1396484375,
+      "learning_rate": 0.02,
+      "loss": 1.6157,
+      "step": 3232
+    },
+    {
+      "epoch": 0.29929989006538216,
+      "grad_norm": 0.146484375,
+      "learning_rate": 0.02,
+      "loss": 1.6688,
+      "step": 3233
+    },
+    {
+      "epoch": 0.2993924665856622,
+      "grad_norm": 0.1552734375,
+      "learning_rate": 0.02,
+      "loss": 1.7371,
+      "step": 3234
+    },
+    {
+      "epoch": 0.29948504310594226,
+      "grad_norm": 0.15625,
+      "learning_rate": 0.02,
+      "loss": 1.6679,
+      "step": 3235
+    },
+    {
+      "epoch": 0.2995776196262223,
+      "grad_norm": 0.15234375,
+      "learning_rate": 0.02,
+      "loss": 1.6308,
+      "step": 3236
+    },
+    {
+      "epoch": 0.29967019614650237,
+      "grad_norm": 0.1611328125,
+      "learning_rate": 0.02,
+      "loss": 1.653,
+      "step": 3237
+    },
+    {
+      "epoch": 0.29976277266678236,
+      "grad_norm": 0.1455078125,
+      "learning_rate": 0.02,
+      "loss": 1.6206,
+      "step": 3238
+    },
+    {
+      "epoch": 0.2998553491870624,
+      "grad_norm": 0.16015625,
+      "learning_rate": 0.02,
+      "loss": 1.6326,
+      "step": 3239
+    },
+    {
+      "epoch": 0.29994792570734247,
+      "grad_norm": 0.1484375,
+      "learning_rate": 0.02,
+      "loss": 1.6291,
+      "step": 3240
+    },
+    {
+      "epoch": 0.3000405022276225,
+      "grad_norm": 0.1552734375,
+      "learning_rate": 0.02,
+      "loss": 1.6518,
+      "step": 3241
+    },
+    {
+      "epoch": 0.30013307874790257,
+      "grad_norm": 0.1474609375,
+      "learning_rate": 0.02,
+      "loss": 1.6397,
+      "step": 3242
+    },
+    {
+      "epoch": 0.3002256552681826,
+      "grad_norm": 0.1533203125,
+      "learning_rate": 0.02,
+      "loss": 1.6372,
+      "step": 3243
+    },
+    {
+      "epoch": 0.3003182317884627,
+      "grad_norm": 0.134765625,
+      "learning_rate": 0.02,
+      "loss": 1.5555,
+      "step": 3244
+    },
+    {
+      "epoch": 0.30041080830874267,
+      "grad_norm": 0.14453125,
+      "learning_rate": 0.02,
+      "loss": 1.6611,
+      "step": 3245
+    },
+    {
+      "epoch": 0.3005033848290227,
+      "grad_norm": 0.138671875,
+      "learning_rate": 0.02,
+      "loss": 1.631,
+      "step": 3246
+    },
+    {
+      "epoch": 0.3005959613493028,
+      "grad_norm": 0.1513671875,
+      "learning_rate": 0.02,
+      "loss": 1.6936,
+      "step": 3247
+    },
+    {
+      "epoch": 0.30068853786958283,
+      "grad_norm": 0.140625,
+      "learning_rate": 0.02,
+      "loss": 1.5454,
+      "step": 3248
+    },
+    {
+      "epoch": 0.3007811143898629,
+      "grad_norm": 0.154296875,
+      "learning_rate": 0.02,
+      "loss": 1.7128,
+      "step": 3249
+    },
+    {
+      "epoch": 0.30087369091014293,
+      "grad_norm": 0.146484375,
+      "learning_rate": 0.02,
+      "loss": 1.6363,
+      "step": 3250
+    },
+    {
+      "epoch": 0.300966267430423,
+      "grad_norm": 0.1494140625,
+      "learning_rate": 0.02,
+      "loss": 1.6148,
+      "step": 3251
+    },
+    {
+      "epoch": 0.301058843950703,
+      "grad_norm": 0.1484375,
+      "learning_rate": 0.02,
+      "loss": 1.6271,
+      "step": 3252
+    },
+    {
+      "epoch": 0.30115142047098303,
+      "grad_norm": 0.15234375,
+      "learning_rate": 0.02,
+      "loss": 1.6397,
+      "step": 3253
+    },
+    {
+      "epoch": 0.3012439969912631,
+      "grad_norm": 0.1455078125,
+      "learning_rate": 0.02,
+      "loss": 1.6242,
+      "step": 3254
+    },
+    {
+      "epoch": 0.30133657351154314,
+      "grad_norm": 0.162109375,
+      "learning_rate": 0.02,
+      "loss": 1.5787,
+      "step": 3255
+    },
+    {
+      "epoch": 0.3014291500318232,
+      "grad_norm": 0.1396484375,
+      "learning_rate": 0.02,
+      "loss": 1.6559,
+      "step": 3256
+    },
+    {
+      "epoch": 0.30152172655210324,
+      "grad_norm": 0.1494140625,
+      "learning_rate": 0.02,
+      "loss": 1.5733,
+      "step": 3257
+    },
+    {
+      "epoch": 0.3016143030723833,
+      "grad_norm": 0.1484375,
+      "learning_rate": 0.02,
+      "loss": 1.6118,
+      "step": 3258
+    },
+    {
+      "epoch": 0.3017068795926633,
+      "grad_norm": 0.1611328125,
+      "learning_rate": 0.02,
+      "loss": 1.6765,
+      "step": 3259
+    },
+    {
+      "epoch": 0.30179945611294334,
+      "grad_norm": 0.1455078125,
+      "learning_rate": 0.02,
+      "loss": 1.6529,
+      "step": 3260
+    },
+    {
+      "epoch": 0.3018920326332234,
+      "grad_norm": 0.1357421875,
+      "learning_rate": 0.02,
+      "loss": 1.5974,
+      "step": 3261
+    },
+    {
+      "epoch": 0.30198460915350345,
+      "grad_norm": 0.1494140625,
+      "learning_rate": 0.02,
+      "loss": 1.6656,
+      "step": 3262
+    },
+    {
+      "epoch": 0.3020771856737835,
+      "grad_norm": 0.1435546875,
+      "learning_rate": 0.02,
+      "loss": 1.6544,
+      "step": 3263
+    },
+    {
+      "epoch": 0.30216976219406355,
+      "grad_norm": 0.14453125,
+      "learning_rate": 0.02,
+      "loss": 1.6942,
+      "step": 3264
+    },
+    {
+      "epoch": 0.30226233871434355,
+      "grad_norm": 0.154296875,
+      "learning_rate": 0.02,
+      "loss": 1.6032,
+      "step": 3265
+    },
+    {
+      "epoch": 0.3023549152346236,
+      "grad_norm": 0.1416015625,
+      "learning_rate": 0.02,
+      "loss": 1.5938,
+      "step": 3266
+    },
+    {
+      "epoch": 0.30244749175490365,
+      "grad_norm": 0.15234375,
+      "learning_rate": 0.02,
+      "loss": 1.6453,
+      "step": 3267
+    },
+    {
+      "epoch": 0.3025400682751837,
+      "grad_norm": 0.1474609375,
+      "learning_rate": 0.02,
+      "loss": 1.6159,
+      "step": 3268
+    },
+    {
+      "epoch": 0.30263264479546376,
+      "grad_norm": 0.1572265625,
+      "learning_rate": 0.02,
+      "loss": 1.6854,
+      "step": 3269
+    },
+    {
+      "epoch": 0.3027252213157438,
+      "grad_norm": 0.1494140625,
+      "learning_rate": 0.02,
+      "loss": 1.6284,
+      "step": 3270
+    },
+    {
+      "epoch": 0.30281779783602386,
+      "grad_norm": 0.146484375,
+      "learning_rate": 0.02,
+      "loss": 1.6025,
+      "step": 3271
+    },
+    {
+      "epoch": 0.30291037435630386,
+      "grad_norm": 0.142578125,
+      "learning_rate": 0.02,
+      "loss": 1.651,
+      "step": 3272
+    },
+    {
+      "epoch": 0.3030029508765839,
+      "grad_norm": 0.142578125,
+      "learning_rate": 0.02,
+      "loss": 1.6148,
+      "step": 3273
+    },
+    {
+      "epoch": 0.30309552739686396,
+      "grad_norm": 0.1513671875,
+      "learning_rate": 0.02,
+      "loss": 1.6254,
+      "step": 3274
+    },
+    {
+      "epoch": 0.303188103917144,
+      "grad_norm": 0.154296875,
+      "learning_rate": 0.02,
+      "loss": 1.6396,
+      "step": 3275
+    },
+    {
+      "epoch": 0.30328068043742407,
+      "grad_norm": 0.1474609375,
+      "learning_rate": 0.02,
+      "loss": 1.5783,
+      "step": 3276
+    },
+    {
+      "epoch": 0.3033732569577041,
+      "grad_norm": 0.1455078125,
+      "learning_rate": 0.02,
+      "loss": 1.6279,
+      "step": 3277
+    },
+    {
+      "epoch": 0.30346583347798417,
+      "grad_norm": 0.1513671875,
+      "learning_rate": 0.02,
+      "loss": 1.6059,
+      "step": 3278
+    },
+    {
+      "epoch": 0.30355840999826417,
+      "grad_norm": 0.150390625,
+      "learning_rate": 0.02,
+      "loss": 1.6968,
+      "step": 3279
+    },
+    {
+      "epoch": 0.3036509865185442,
+      "grad_norm": 0.1494140625,
+      "learning_rate": 0.02,
+      "loss": 1.647,
+      "step": 3280
+    },
+    {
+      "epoch": 0.30374356303882427,
+      "grad_norm": 0.15234375,
+      "learning_rate": 0.02,
+      "loss": 1.6193,
+      "step": 3281
+    },
+    {
+      "epoch": 0.3038361395591043,
+      "grad_norm": 0.15625,
+      "learning_rate": 0.02,
+      "loss": 1.7335,
+      "step": 3282
+    },
+    {
+      "epoch": 0.3039287160793844,
+      "grad_norm": 0.1494140625,
+      "learning_rate": 0.02,
+      "loss": 1.6726,
+      "step": 3283
+    },
+    {
+      "epoch": 0.30402129259966443,
+      "grad_norm": 0.16015625,
+      "learning_rate": 0.02,
+      "loss": 1.636,
+      "step": 3284
+    },
+    {
+      "epoch": 0.3041138691199445,
+      "grad_norm": 0.166015625,
+      "learning_rate": 0.02,
+      "loss": 1.6922,
+      "step": 3285
+    },
+    {
+      "epoch": 0.3042064456402245,
+      "grad_norm": 0.16015625,
+      "learning_rate": 0.02,
+      "loss": 1.6413,
+      "step": 3286
+    },
+    {
+      "epoch": 0.30429902216050453,
+      "grad_norm": 0.14453125,
+      "learning_rate": 0.02,
+      "loss": 1.5575,
+      "step": 3287
+    },
+    {
+      "epoch": 0.3043915986807846,
+      "grad_norm": 0.150390625,
+      "learning_rate": 0.02,
+      "loss": 1.6456,
+      "step": 3288
+    },
+    {
+      "epoch": 0.30448417520106463,
+      "grad_norm": 0.15234375,
+      "learning_rate": 0.02,
+      "loss": 1.6726,
+      "step": 3289
+    },
+    {
+      "epoch": 0.3045767517213447,
+      "grad_norm": 0.1640625,
+      "learning_rate": 0.02,
+      "loss": 1.7019,
+      "step": 3290
+    },
+    {
+      "epoch": 0.30466932824162474,
+      "grad_norm": 0.142578125,
+      "learning_rate": 0.02,
+      "loss": 1.617,
+      "step": 3291
+    },
+    {
+      "epoch": 0.3047619047619048,
+      "grad_norm": 0.15234375,
+      "learning_rate": 0.02,
+      "loss": 1.7183,
+      "step": 3292
+    },
+    {
+      "epoch": 0.3048544812821848,
+      "grad_norm": 0.154296875,
+      "learning_rate": 0.02,
+      "loss": 1.6508,
+      "step": 3293
+    },
+    {
+      "epoch": 0.30494705780246484,
+      "grad_norm": 0.158203125,
+      "learning_rate": 0.02,
+      "loss": 1.7104,
+      "step": 3294
+    },
+    {
+      "epoch": 0.3050396343227449,
+      "grad_norm": 0.1513671875,
+      "learning_rate": 0.02,
+      "loss": 1.6996,
+      "step": 3295
+    },
+    {
+      "epoch": 0.30513221084302494,
+      "grad_norm": 0.1435546875,
+      "learning_rate": 0.02,
+      "loss": 1.5764,
+      "step": 3296
+    },
+    {
+      "epoch": 0.305224787363305,
+      "grad_norm": 0.142578125,
+      "learning_rate": 0.02,
+      "loss": 1.6836,
+      "step": 3297
+    },
+    {
+      "epoch": 0.30531736388358505,
+      "grad_norm": 0.13671875,
+      "learning_rate": 0.02,
+      "loss": 1.5789,
+      "step": 3298
+    },
+    {
+      "epoch": 0.30540994040386504,
+      "grad_norm": 0.1376953125,
+      "learning_rate": 0.02,
+      "loss": 1.6352,
+      "step": 3299
+    },
+    {
+      "epoch": 0.3055025169241451,
+      "grad_norm": 0.1337890625,
+      "learning_rate": 0.02,
+      "loss": 1.592,
+      "step": 3300
+    },
+    {
+      "epoch": 0.30559509344442515,
+      "grad_norm": 0.1455078125,
+      "learning_rate": 0.02,
+      "loss": 1.6561,
+      "step": 3301
+    },
+    {
+      "epoch": 0.3056876699647052,
+      "grad_norm": 0.1396484375,
+      "learning_rate": 0.02,
+      "loss": 1.6149,
+      "step": 3302
+    },
+    {
+      "epoch": 0.30578024648498525,
+      "grad_norm": 0.14453125,
+      "learning_rate": 0.02,
+      "loss": 1.6394,
+      "step": 3303
+    },
+    {
+      "epoch": 0.3058728230052653,
+      "grad_norm": 0.1494140625,
+      "learning_rate": 0.02,
+      "loss": 1.6654,
+      "step": 3304
+    },
+    {
+      "epoch": 0.30596539952554536,
+      "grad_norm": 0.1494140625,
+      "learning_rate": 0.02,
+      "loss": 1.6923,
+      "step": 3305
+    },
+    {
+      "epoch": 0.30605797604582535,
+      "grad_norm": 0.1455078125,
+      "learning_rate": 0.02,
+      "loss": 1.5938,
+      "step": 3306
+    },
+    {
+      "epoch": 0.3061505525661054,
+      "grad_norm": 0.134765625,
+      "learning_rate": 0.02,
+      "loss": 1.5949,
+      "step": 3307
+    },
+    {
+      "epoch": 0.30624312908638546,
+      "grad_norm": 0.146484375,
+      "learning_rate": 0.02,
+      "loss": 1.5991,
+      "step": 3308
+    },
+    {
+      "epoch": 0.3063357056066655,
+      "grad_norm": 0.146484375,
+      "learning_rate": 0.02,
+      "loss": 1.5976,
+      "step": 3309
+    },
+    {
+      "epoch": 0.30642828212694556,
+      "grad_norm": 0.1494140625,
+      "learning_rate": 0.02,
+      "loss": 1.6423,
+      "step": 3310
+    },
+    {
+      "epoch": 0.3065208586472256,
+      "grad_norm": 0.1376953125,
+      "learning_rate": 0.02,
+      "loss": 1.6318,
+      "step": 3311
+    },
+    {
+      "epoch": 0.30661343516750567,
+      "grad_norm": 0.14453125,
+      "learning_rate": 0.02,
+      "loss": 1.6725,
+      "step": 3312
+    },
+    {
+      "epoch": 0.30670601168778566,
+      "grad_norm": 0.14453125,
+      "learning_rate": 0.02,
+      "loss": 1.6234,
+      "step": 3313
+    },
+    {
+      "epoch": 0.3067985882080657,
+      "grad_norm": 0.140625,
+      "learning_rate": 0.02,
+      "loss": 1.569,
+      "step": 3314
+    },
+    {
+      "epoch": 0.30689116472834577,
+      "grad_norm": 0.140625,
+      "learning_rate": 0.02,
+      "loss": 1.6449,
+      "step": 3315
+    },
+    {
+      "epoch": 0.3069837412486258,
+      "grad_norm": 0.1416015625,
+      "learning_rate": 0.02,
+      "loss": 1.6356,
+      "step": 3316
+    },
+    {
+      "epoch": 0.30707631776890587,
+      "grad_norm": 0.1435546875,
+      "learning_rate": 0.02,
+      "loss": 1.6449,
+      "step": 3317
+    },
+    {
+      "epoch": 0.3071688942891859,
+      "grad_norm": 0.142578125,
+      "learning_rate": 0.02,
+      "loss": 1.5654,
+      "step": 3318
+    },
+    {
+      "epoch": 0.307261470809466,
+      "grad_norm": 0.140625,
+      "learning_rate": 0.02,
+      "loss": 1.616,
+      "step": 3319
+    },
+    {
+      "epoch": 0.30735404732974597,
+      "grad_norm": 0.1533203125,
+      "learning_rate": 0.02,
+      "loss": 1.623,
+      "step": 3320
+    },
+    {
+      "epoch": 0.307446623850026,
+      "grad_norm": 0.1455078125,
+      "learning_rate": 0.02,
+      "loss": 1.7176,
+      "step": 3321
+    },
+    {
+      "epoch": 0.3075392003703061,
+      "grad_norm": 0.1494140625,
+      "learning_rate": 0.02,
+      "loss": 1.6575,
+      "step": 3322
+    },
+    {
+      "epoch": 0.30763177689058613,
+      "grad_norm": 0.142578125,
+      "learning_rate": 0.02,
+      "loss": 1.5491,
+      "step": 3323
+    },
+    {
+      "epoch": 0.3077243534108662,
+      "grad_norm": 0.1474609375,
+      "learning_rate": 0.02,
+      "loss": 1.5852,
+      "step": 3324
+    },
+    {
+      "epoch": 0.30781692993114623,
+      "grad_norm": 0.1474609375,
+      "learning_rate": 0.02,
+      "loss": 1.6261,
+      "step": 3325
+    },
+    {
+      "epoch": 0.3079095064514263,
+      "grad_norm": 0.140625,
+      "learning_rate": 0.02,
+      "loss": 1.622,
+      "step": 3326
+    },
+    {
+      "epoch": 0.3080020829717063,
+      "grad_norm": 0.1435546875,
+      "learning_rate": 0.02,
+      "loss": 1.6383,
+      "step": 3327
+    },
+    {
+      "epoch": 0.30809465949198633,
+      "grad_norm": 0.1416015625,
+      "learning_rate": 0.02,
+      "loss": 1.6769,
+      "step": 3328
+    },
+    {
+      "epoch": 0.3081872360122664,
+      "grad_norm": 0.1474609375,
+      "learning_rate": 0.02,
+      "loss": 1.6652,
+      "step": 3329
+    },
+    {
+      "epoch": 0.30827981253254644,
+      "grad_norm": 0.1513671875,
+      "learning_rate": 0.02,
+      "loss": 1.6284,
+      "step": 3330
+    },
+    {
+      "epoch": 0.3083723890528265,
+      "grad_norm": 0.14453125,
+      "learning_rate": 0.02,
+      "loss": 1.6718,
+      "step": 3331
+    },
+    {
+      "epoch": 0.30846496557310654,
+      "grad_norm": 0.138671875,
+      "learning_rate": 0.02,
+      "loss": 1.669,
+      "step": 3332
+    },
+    {
+      "epoch": 0.30855754209338654,
+      "grad_norm": 0.140625,
+      "learning_rate": 0.02,
+      "loss": 1.583,
+      "step": 3333
+    },
+    {
+      "epoch": 0.3086501186136666,
+      "grad_norm": 0.1396484375,
+      "learning_rate": 0.02,
+      "loss": 1.6543,
+      "step": 3334
+    },
+    {
+      "epoch": 0.30874269513394664,
+      "grad_norm": 0.1435546875,
+      "learning_rate": 0.02,
+      "loss": 1.5555,
+      "step": 3335
+    },
+    {
+      "epoch": 0.3088352716542267,
+      "grad_norm": 0.142578125,
+      "learning_rate": 0.02,
+      "loss": 1.6438,
+      "step": 3336
+    },
+    {
+      "epoch": 0.30892784817450675,
+      "grad_norm": 0.1357421875,
+      "learning_rate": 0.02,
+      "loss": 1.5781,
+      "step": 3337
+    },
+    {
+      "epoch": 0.3090204246947868,
+      "grad_norm": 0.142578125,
+      "learning_rate": 0.02,
+      "loss": 1.6306,
+      "step": 3338
+    },
+    {
+      "epoch": 0.30911300121506685,
+      "grad_norm": 0.15234375,
+      "learning_rate": 0.02,
+      "loss": 1.6629,
+      "step": 3339
+    },
+    {
+      "epoch": 0.30920557773534685,
+      "grad_norm": 0.14453125,
+      "learning_rate": 0.02,
+      "loss": 1.6566,
+      "step": 3340
+    },
+    {
+      "epoch": 0.3092981542556269,
+      "grad_norm": 0.1435546875,
+      "learning_rate": 0.02,
+      "loss": 1.6189,
+      "step": 3341
+    },
+    {
+      "epoch": 0.30939073077590695,
+      "grad_norm": 0.15625,
+      "learning_rate": 0.02,
+      "loss": 1.61,
+      "step": 3342
+    },
+    {
+      "epoch": 0.309483307296187,
+      "grad_norm": 0.1669921875,
+      "learning_rate": 0.02,
+      "loss": 1.6632,
+      "step": 3343
+    },
+    {
+      "epoch": 0.30957588381646706,
+      "grad_norm": 0.1533203125,
+      "learning_rate": 0.02,
+      "loss": 1.5994,
+      "step": 3344
+    },
+    {
+      "epoch": 0.3096684603367471,
+      "grad_norm": 0.1416015625,
+      "learning_rate": 0.02,
+      "loss": 1.6619,
+      "step": 3345
+    },
+    {
+      "epoch": 0.30976103685702716,
+      "grad_norm": 0.1455078125,
+      "learning_rate": 0.02,
+      "loss": 1.6666,
+      "step": 3346
+    },
+    {
+      "epoch": 0.30985361337730716,
+      "grad_norm": 0.1513671875,
+      "learning_rate": 0.02,
+      "loss": 1.6267,
+      "step": 3347
+    },
+    {
+      "epoch": 0.3099461898975872,
+      "grad_norm": 0.1474609375,
+      "learning_rate": 0.02,
+      "loss": 1.6621,
+      "step": 3348
+    },
+    {
+      "epoch": 0.31003876641786726,
+      "grad_norm": 0.1533203125,
+      "learning_rate": 0.02,
+      "loss": 1.5985,
+      "step": 3349
+    },
+    {
+      "epoch": 0.3101313429381473,
+      "grad_norm": 0.1513671875,
+      "learning_rate": 0.02,
+      "loss": 1.6961,
+      "step": 3350
+    },
+    {
+      "epoch": 0.31022391945842737,
+      "grad_norm": 0.146484375,
+      "learning_rate": 0.02,
+      "loss": 1.6537,
+      "step": 3351
+    },
+    {
+      "epoch": 0.3103164959787074,
+      "grad_norm": 0.1416015625,
+      "learning_rate": 0.02,
+      "loss": 1.6215,
+      "step": 3352
+    },
+    {
+      "epoch": 0.31040907249898747,
+      "grad_norm": 0.154296875,
+      "learning_rate": 0.02,
+      "loss": 1.6499,
+      "step": 3353
+    },
+    {
+      "epoch": 0.31050164901926747,
+      "grad_norm": 0.1455078125,
+      "learning_rate": 0.02,
+      "loss": 1.603,
+      "step": 3354
+    },
+    {
+      "epoch": 0.3105942255395475,
+      "grad_norm": 0.1494140625,
+      "learning_rate": 0.02,
+      "loss": 1.589,
+      "step": 3355
+    },
+    {
+      "epoch": 0.31068680205982757,
+      "grad_norm": 0.146484375,
+      "learning_rate": 0.02,
+      "loss": 1.64,
+      "step": 3356
+    },
+    {
+      "epoch": 0.3107793785801076,
+      "grad_norm": 0.1572265625,
+      "learning_rate": 0.02,
+      "loss": 1.6171,
+      "step": 3357
+    },
+    {
+      "epoch": 0.3108719551003877,
+      "grad_norm": 0.1484375,
+      "learning_rate": 0.02,
+      "loss": 1.5694,
+      "step": 3358
+    },
+    {
+      "epoch": 0.31096453162066773,
+      "grad_norm": 0.1455078125,
+      "learning_rate": 0.02,
+      "loss": 1.6401,
+      "step": 3359
+    },
+    {
+      "epoch": 0.3110571081409477,
+      "grad_norm": 0.1494140625,
+      "learning_rate": 0.02,
+      "loss": 1.6859,
+      "step": 3360
+    },
+    {
+      "epoch": 0.3111496846612278,
+      "grad_norm": 0.13671875,
+      "learning_rate": 0.02,
+      "loss": 1.6235,
+      "step": 3361
+    },
+    {
+      "epoch": 0.31124226118150783,
+      "grad_norm": 0.1416015625,
+      "learning_rate": 0.02,
+      "loss": 1.6298,
+      "step": 3362
+    },
+    {
+      "epoch": 0.3113348377017879,
+      "grad_norm": 0.1435546875,
+      "learning_rate": 0.02,
+      "loss": 1.5688,
+      "step": 3363
+    },
+    {
+      "epoch": 0.31142741422206793,
+      "grad_norm": 0.1396484375,
+      "learning_rate": 0.02,
+      "loss": 1.6668,
+      "step": 3364
+    },
+    {
+      "epoch": 0.311519990742348,
+      "grad_norm": 0.13671875,
+      "learning_rate": 0.02,
+      "loss": 1.6217,
+      "step": 3365
+    },
+    {
+      "epoch": 0.31161256726262804,
+      "grad_norm": 0.14453125,
+      "learning_rate": 0.02,
+      "loss": 1.6841,
+      "step": 3366
+    },
+    {
+      "epoch": 0.31170514378290803,
+      "grad_norm": 0.1435546875,
+      "learning_rate": 0.02,
+      "loss": 1.6522,
+      "step": 3367
+    },
+    {
+      "epoch": 0.3117977203031881,
+      "grad_norm": 0.1328125,
+      "learning_rate": 0.02,
+      "loss": 1.6177,
+      "step": 3368
+    },
+    {
+      "epoch": 0.31189029682346814,
+      "grad_norm": 0.1416015625,
+      "learning_rate": 0.02,
+      "loss": 1.6097,
+      "step": 3369
+    },
+    {
+      "epoch": 0.3119828733437482,
+      "grad_norm": 0.1455078125,
+      "learning_rate": 0.02,
+      "loss": 1.6387,
+      "step": 3370
+    },
+    {
+      "epoch": 0.31207544986402824,
+      "grad_norm": 0.142578125,
+      "learning_rate": 0.02,
+      "loss": 1.6277,
+      "step": 3371
+    },
+    {
+      "epoch": 0.3121680263843083,
+      "grad_norm": 0.1416015625,
+      "learning_rate": 0.02,
+      "loss": 1.6893,
+      "step": 3372
+    },
+    {
+      "epoch": 0.31226060290458835,
+      "grad_norm": 0.1396484375,
+      "learning_rate": 0.02,
+      "loss": 1.5554,
+      "step": 3373
+    },
+    {
+      "epoch": 0.31235317942486834,
+      "grad_norm": 0.140625,
+      "learning_rate": 0.02,
+      "loss": 1.5794,
+      "step": 3374
+    },
+    {
+      "epoch": 0.3124457559451484,
+      "grad_norm": 0.1435546875,
+      "learning_rate": 0.02,
+      "loss": 1.5752,
+      "step": 3375
+    },
+    {
+      "epoch": 0.31253833246542845,
+      "grad_norm": 0.146484375,
+      "learning_rate": 0.02,
+      "loss": 1.6252,
+      "step": 3376
+    },
+    {
+      "epoch": 0.3126309089857085,
+      "grad_norm": 0.1494140625,
+      "learning_rate": 0.02,
+      "loss": 1.6201,
+      "step": 3377
+    },
+    {
+      "epoch": 0.31272348550598855,
+      "grad_norm": 0.15234375,
+      "learning_rate": 0.02,
+      "loss": 1.672,
+      "step": 3378
+    },
+    {
+      "epoch": 0.3128160620262686,
+      "grad_norm": 0.1591796875,
+      "learning_rate": 0.02,
+      "loss": 1.6504,
+      "step": 3379
+    },
+    {
+      "epoch": 0.31290863854654866,
+      "grad_norm": 0.162109375,
+      "learning_rate": 0.02,
+      "loss": 1.5595,
+      "step": 3380
+    },
+    {
+      "epoch": 0.31300121506682865,
+      "grad_norm": 0.16015625,
+      "learning_rate": 0.02,
+      "loss": 1.6157,
+      "step": 3381
+    },
+    {
+      "epoch": 0.3130937915871087,
+      "grad_norm": 0.166015625,
+      "learning_rate": 0.02,
+      "loss": 1.7263,
+      "step": 3382
+    },
+    {
+      "epoch": 0.31318636810738876,
+      "grad_norm": 0.1650390625,
+      "learning_rate": 0.02,
+      "loss": 1.6506,
+      "step": 3383
+    },
+    {
+      "epoch": 0.3132789446276688,
+      "grad_norm": 0.1640625,
+      "learning_rate": 0.02,
+      "loss": 1.6336,
+      "step": 3384
+    },
+    {
+      "epoch": 0.31337152114794886,
+      "grad_norm": 0.1572265625,
+      "learning_rate": 0.02,
+      "loss": 1.6702,
+      "step": 3385
+    },
+    {
+      "epoch": 0.3134640976682289,
+      "grad_norm": 0.1669921875,
+      "learning_rate": 0.02,
+      "loss": 1.69,
+      "step": 3386
+    },
+    {
+      "epoch": 0.31355667418850897,
+      "grad_norm": 0.1513671875,
+      "learning_rate": 0.02,
+      "loss": 1.6635,
+      "step": 3387
+    },
+    {
+      "epoch": 0.31364925070878896,
+      "grad_norm": 0.15234375,
+      "learning_rate": 0.02,
+      "loss": 1.673,
+      "step": 3388
+    },
+    {
+      "epoch": 0.313741827229069,
+      "grad_norm": 0.1484375,
+      "learning_rate": 0.02,
+      "loss": 1.6639,
+      "step": 3389
+    },
+    {
+      "epoch": 0.31383440374934907,
+      "grad_norm": 0.1552734375,
+      "learning_rate": 0.02,
+      "loss": 1.5989,
+      "step": 3390
+    },
+    {
+      "epoch": 0.3139269802696291,
+      "grad_norm": 0.1416015625,
+      "learning_rate": 0.02,
+      "loss": 1.6657,
+      "step": 3391
+    },
+    {
+      "epoch": 0.31401955678990917,
+      "grad_norm": 0.158203125,
+      "learning_rate": 0.02,
+      "loss": 1.646,
+      "step": 3392
+    },
+    {
+      "epoch": 0.3141121333101892,
+      "grad_norm": 0.15625,
+      "learning_rate": 0.02,
+      "loss": 1.6502,
+      "step": 3393
+    },
+    {
+      "epoch": 0.3142047098304692,
+      "grad_norm": 0.1513671875,
+      "learning_rate": 0.02,
+      "loss": 1.5719,
+      "step": 3394
+    },
+    {
+      "epoch": 0.31429728635074927,
+      "grad_norm": 0.1552734375,
+      "learning_rate": 0.02,
+      "loss": 1.6369,
+      "step": 3395
+    },
+    {
+      "epoch": 0.3143898628710293,
+      "grad_norm": 0.1474609375,
+      "learning_rate": 0.02,
+      "loss": 1.5653,
+      "step": 3396
+    },
+    {
+      "epoch": 0.3144824393913094,
+      "grad_norm": 0.1474609375,
+      "learning_rate": 0.02,
+      "loss": 1.6197,
+      "step": 3397
+    },
+    {
+      "epoch": 0.31457501591158943,
+      "grad_norm": 0.14453125,
+      "learning_rate": 0.02,
+      "loss": 1.5871,
+      "step": 3398
+    },
+    {
+      "epoch": 0.3146675924318695,
+      "grad_norm": 0.1533203125,
+      "learning_rate": 0.02,
+      "loss": 1.633,
+      "step": 3399
+    },
+    {
+      "epoch": 0.31476016895214953,
+      "grad_norm": 0.146484375,
+      "learning_rate": 0.02,
+      "loss": 1.6056,
+      "step": 3400
+    },
+    {
+      "epoch": 0.31485274547242953,
+      "grad_norm": 0.1484375,
+      "learning_rate": 0.02,
+      "loss": 1.5768,
+      "step": 3401
+    },
+    {
+      "epoch": 0.3149453219927096,
+      "grad_norm": 0.1494140625,
+      "learning_rate": 0.02,
+      "loss": 1.6338,
+      "step": 3402
+    },
+    {
+      "epoch": 0.31503789851298963,
+      "grad_norm": 0.1474609375,
+      "learning_rate": 0.02,
+      "loss": 1.6571,
+      "step": 3403
+    },
+    {
+      "epoch": 0.3151304750332697,
+      "grad_norm": 0.1494140625,
+      "learning_rate": 0.02,
+      "loss": 1.6316,
+      "step": 3404
+    },
+    {
+      "epoch": 0.31522305155354974,
+      "grad_norm": 0.16015625,
+      "learning_rate": 0.02,
+      "loss": 1.6412,
+      "step": 3405
+    },
+    {
+      "epoch": 0.3153156280738298,
+      "grad_norm": 0.1572265625,
+      "learning_rate": 0.02,
+      "loss": 1.6724,
+      "step": 3406
+    },
+    {
+      "epoch": 0.31540820459410984,
+      "grad_norm": 0.1513671875,
+      "learning_rate": 0.02,
+      "loss": 1.6106,
+      "step": 3407
+    },
+    {
+      "epoch": 0.31550078111438984,
+      "grad_norm": 0.1533203125,
+      "learning_rate": 0.02,
+      "loss": 1.704,
+      "step": 3408
+    },
+    {
+      "epoch": 0.3155933576346699,
+      "grad_norm": 0.1533203125,
+      "learning_rate": 0.02,
+      "loss": 1.6396,
+      "step": 3409
+    },
+    {
+      "epoch": 0.31568593415494994,
+      "grad_norm": 0.1455078125,
+      "learning_rate": 0.02,
+      "loss": 1.621,
+      "step": 3410
+    },
+    {
+      "epoch": 0.31577851067523,
+      "grad_norm": 0.150390625,
+      "learning_rate": 0.02,
+      "loss": 1.6133,
+      "step": 3411
+    },
+    {
+      "epoch": 0.31587108719551005,
+      "grad_norm": 0.1484375,
+      "learning_rate": 0.02,
+      "loss": 1.6351,
+      "step": 3412
+    },
+    {
+      "epoch": 0.3159636637157901,
+      "grad_norm": 0.150390625,
+      "learning_rate": 0.02,
+      "loss": 1.6477,
+      "step": 3413
+    },
+    {
+      "epoch": 0.31605624023607015,
+      "grad_norm": 0.1474609375,
+      "learning_rate": 0.02,
+      "loss": 1.6309,
+      "step": 3414
+    },
+    {
+      "epoch": 0.31614881675635015,
+      "grad_norm": 0.140625,
+      "learning_rate": 0.02,
+      "loss": 1.6207,
+      "step": 3415
+    },
+    {
+      "epoch": 0.3162413932766302,
+      "grad_norm": 0.1376953125,
+      "learning_rate": 0.02,
+      "loss": 1.624,
+      "step": 3416
+    },
+    {
+      "epoch": 0.31633396979691025,
+      "grad_norm": 0.1474609375,
+      "learning_rate": 0.02,
+      "loss": 1.6112,
+      "step": 3417
+    },
+    {
+      "epoch": 0.3164265463171903,
+      "grad_norm": 0.1455078125,
+      "learning_rate": 0.02,
+      "loss": 1.6216,
+      "step": 3418
+    },
+    {
+      "epoch": 0.31651912283747036,
+      "grad_norm": 0.1435546875,
+      "learning_rate": 0.02,
+      "loss": 1.6124,
+      "step": 3419
+    },
+    {
+      "epoch": 0.3166116993577504,
+      "grad_norm": 0.1455078125,
+      "learning_rate": 0.02,
+      "loss": 1.6625,
+      "step": 3420
+    },
+    {
+      "epoch": 0.31670427587803046,
+      "grad_norm": 0.15234375,
+      "learning_rate": 0.02,
+      "loss": 1.6671,
+      "step": 3421
+    },
+    {
+      "epoch": 0.31679685239831046,
+      "grad_norm": 0.146484375,
+      "learning_rate": 0.02,
+      "loss": 1.5982,
+      "step": 3422
+    },
+    {
+      "epoch": 0.3168894289185905,
+      "grad_norm": 0.1533203125,
+      "learning_rate": 0.02,
+      "loss": 1.5931,
+      "step": 3423
+    },
+    {
+      "epoch": 0.31698200543887056,
+      "grad_norm": 0.154296875,
+      "learning_rate": 0.02,
+      "loss": 1.5872,
+      "step": 3424
+    },
+    {
+      "epoch": 0.3170745819591506,
+      "grad_norm": 0.15234375,
+      "learning_rate": 0.02,
+      "loss": 1.6526,
+      "step": 3425
+    },
+    {
+      "epoch": 0.31716715847943067,
+      "grad_norm": 0.1494140625,
+      "learning_rate": 0.02,
+      "loss": 1.577,
+      "step": 3426
+    },
+    {
+      "epoch": 0.3172597349997107,
+      "grad_norm": 0.1435546875,
+      "learning_rate": 0.02,
+      "loss": 1.6219,
+      "step": 3427
+    },
+    {
+      "epoch": 0.3173523115199907,
+      "grad_norm": 0.1474609375,
+      "learning_rate": 0.02,
+      "loss": 1.609,
+      "step": 3428
+    },
+    {
+      "epoch": 0.31744488804027077,
+      "grad_norm": 0.1513671875,
+      "learning_rate": 0.02,
+      "loss": 1.6502,
+      "step": 3429
+    },
+    {
+      "epoch": 0.3175374645605508,
+      "grad_norm": 0.1572265625,
+      "learning_rate": 0.02,
+      "loss": 1.6508,
+      "step": 3430
+    },
+    {
+      "epoch": 0.31763004108083087,
+      "grad_norm": 0.1474609375,
+      "learning_rate": 0.02,
+      "loss": 1.6173,
+      "step": 3431
+    },
+    {
+      "epoch": 0.3177226176011109,
+      "grad_norm": 0.1513671875,
+      "learning_rate": 0.02,
+      "loss": 1.6465,
+      "step": 3432
+    },
+    {
+      "epoch": 0.317815194121391,
+      "grad_norm": 0.146484375,
+      "learning_rate": 0.02,
+      "loss": 1.5968,
+      "step": 3433
+    },
+    {
+      "epoch": 0.31790777064167103,
+      "grad_norm": 0.1474609375,
+      "learning_rate": 0.02,
+      "loss": 1.6194,
+      "step": 3434
+    },
+    {
+      "epoch": 0.318000347161951,
+      "grad_norm": 0.138671875,
+      "learning_rate": 0.02,
+      "loss": 1.6008,
+      "step": 3435
+    },
+    {
+      "epoch": 0.3180929236822311,
+      "grad_norm": 0.15625,
+      "learning_rate": 0.02,
+      "loss": 1.6365,
+      "step": 3436
+    },
+    {
+      "epoch": 0.31818550020251113,
+      "grad_norm": 0.1513671875,
+      "learning_rate": 0.02,
+      "loss": 1.6135,
+      "step": 3437
+    },
+    {
+      "epoch": 0.3182780767227912,
+      "grad_norm": 0.1533203125,
+      "learning_rate": 0.02,
+      "loss": 1.6052,
+      "step": 3438
+    },
+    {
+      "epoch": 0.31837065324307123,
+      "grad_norm": 0.150390625,
+      "learning_rate": 0.02,
+      "loss": 1.5808,
+      "step": 3439
+    },
+    {
+      "epoch": 0.3184632297633513,
+      "grad_norm": 0.1533203125,
+      "learning_rate": 0.02,
+      "loss": 1.6795,
+      "step": 3440
+    },
+    {
+      "epoch": 0.31855580628363134,
+      "grad_norm": 0.1484375,
+      "learning_rate": 0.02,
+      "loss": 1.6343,
+      "step": 3441
+    },
+    {
+      "epoch": 0.31864838280391133,
+      "grad_norm": 0.1513671875,
+      "learning_rate": 0.02,
+      "loss": 1.6497,
+      "step": 3442
+    },
+    {
+      "epoch": 0.3187409593241914,
+      "grad_norm": 0.1474609375,
+      "learning_rate": 0.02,
+      "loss": 1.6385,
+      "step": 3443
+    },
+    {
+      "epoch": 0.31883353584447144,
+      "grad_norm": 0.1455078125,
+      "learning_rate": 0.02,
+      "loss": 1.5686,
+      "step": 3444
+    },
+    {
+      "epoch": 0.3189261123647515,
+      "grad_norm": 0.1494140625,
+      "learning_rate": 0.02,
+      "loss": 1.6257,
+      "step": 3445
+    },
+    {
+      "epoch": 0.31901868888503154,
+      "grad_norm": 0.15234375,
+      "learning_rate": 0.02,
+      "loss": 1.6117,
+      "step": 3446
+    },
+    {
+      "epoch": 0.3191112654053116,
+      "grad_norm": 0.1572265625,
+      "learning_rate": 0.02,
+      "loss": 1.6358,
+      "step": 3447
+    },
+    {
+      "epoch": 0.31920384192559165,
+      "grad_norm": 0.16015625,
+      "learning_rate": 0.02,
+      "loss": 1.6852,
+      "step": 3448
+    },
+    {
+      "epoch": 0.31929641844587164,
+      "grad_norm": 0.1474609375,
+      "learning_rate": 0.02,
+      "loss": 1.6323,
+      "step": 3449
+    },
+    {
+      "epoch": 0.3193889949661517,
+      "grad_norm": 0.14453125,
+      "learning_rate": 0.02,
+      "loss": 1.6015,
+      "step": 3450
+    },
+    {
+      "epoch": 0.31948157148643175,
+      "grad_norm": 0.150390625,
+      "learning_rate": 0.02,
+      "loss": 1.6646,
+      "step": 3451
+    },
+    {
+      "epoch": 0.3195741480067118,
+      "grad_norm": 0.1513671875,
+      "learning_rate": 0.02,
+      "loss": 1.6323,
+      "step": 3452
+    },
+    {
+      "epoch": 0.31966672452699185,
+      "grad_norm": 0.1484375,
+      "learning_rate": 0.02,
+      "loss": 1.6481,
+      "step": 3453
+    },
+    {
+      "epoch": 0.3197593010472719,
+      "grad_norm": 0.15234375,
+      "learning_rate": 0.02,
+      "loss": 1.5606,
+      "step": 3454
+    },
+    {
+      "epoch": 0.31985187756755196,
+      "grad_norm": 0.150390625,
+      "learning_rate": 0.02,
+      "loss": 1.6365,
+      "step": 3455
+    },
+    {
+      "epoch": 0.31994445408783195,
+      "grad_norm": 0.154296875,
+      "learning_rate": 0.02,
+      "loss": 1.5948,
+      "step": 3456
+    },
+    {
+      "epoch": 0.320037030608112,
+      "grad_norm": 0.150390625,
+      "learning_rate": 0.02,
+      "loss": 1.6112,
+      "step": 3457
+    },
+    {
+      "epoch": 0.32012960712839206,
+      "grad_norm": 0.150390625,
+      "learning_rate": 0.02,
+      "loss": 1.6261,
+      "step": 3458
+    },
+    {
+      "epoch": 0.3202221836486721,
+      "grad_norm": 0.1533203125,
+      "learning_rate": 0.02,
+      "loss": 1.6163,
+      "step": 3459
+    },
+    {
+      "epoch": 0.32031476016895216,
+      "grad_norm": 0.1474609375,
+      "learning_rate": 0.02,
+      "loss": 1.6378,
+      "step": 3460
+    },
+    {
+      "epoch": 0.3204073366892322,
+      "grad_norm": 0.1513671875,
+      "learning_rate": 0.02,
+      "loss": 1.6249,
+      "step": 3461
+    },
+    {
+      "epoch": 0.3204999132095122,
+      "grad_norm": 0.138671875,
+      "learning_rate": 0.02,
+      "loss": 1.616,
+      "step": 3462
+    },
+    {
+      "epoch": 0.32059248972979226,
+      "grad_norm": 0.1416015625,
+      "learning_rate": 0.02,
+      "loss": 1.6481,
+      "step": 3463
+    },
+    {
+      "epoch": 0.3206850662500723,
+      "grad_norm": 0.1337890625,
+      "learning_rate": 0.02,
+      "loss": 1.6064,
+      "step": 3464
+    },
+    {
+      "epoch": 0.32077764277035237,
+      "grad_norm": 0.146484375,
+      "learning_rate": 0.02,
+      "loss": 1.6112,
+      "step": 3465
+    },
+    {
+      "epoch": 0.3208702192906324,
+      "grad_norm": 0.1435546875,
+      "learning_rate": 0.02,
+      "loss": 1.5945,
+      "step": 3466
+    },
+    {
+      "epoch": 0.32096279581091247,
+      "grad_norm": 0.1474609375,
+      "learning_rate": 0.02,
+      "loss": 1.5334,
+      "step": 3467
+    },
+    {
+      "epoch": 0.3210553723311925,
+      "grad_norm": 0.1474609375,
+      "learning_rate": 0.02,
+      "loss": 1.5626,
+      "step": 3468
+    },
+    {
+      "epoch": 0.3211479488514725,
+      "grad_norm": 0.1650390625,
+      "learning_rate": 0.02,
+      "loss": 1.6743,
+      "step": 3469
+    },
+    {
+      "epoch": 0.32124052537175257,
+      "grad_norm": 0.1455078125,
+      "learning_rate": 0.02,
+      "loss": 1.6488,
+      "step": 3470
+    },
+    {
+      "epoch": 0.3213331018920326,
+      "grad_norm": 0.1513671875,
+      "learning_rate": 0.02,
+      "loss": 1.625,
+      "step": 3471
+    },
+    {
+      "epoch": 0.3214256784123127,
+      "grad_norm": 0.15234375,
+      "learning_rate": 0.02,
+      "loss": 1.6444,
+      "step": 3472
+    },
+    {
+      "epoch": 0.32151825493259273,
+      "grad_norm": 0.150390625,
+      "learning_rate": 0.02,
+      "loss": 1.6116,
+      "step": 3473
+    },
+    {
+      "epoch": 0.3216108314528728,
+      "grad_norm": 0.14453125,
+      "learning_rate": 0.02,
+      "loss": 1.6154,
+      "step": 3474
+    },
+    {
+      "epoch": 0.32170340797315283,
+      "grad_norm": 0.15625,
+      "learning_rate": 0.02,
+      "loss": 1.6158,
+      "step": 3475
+    },
+    {
+      "epoch": 0.32179598449343283,
+      "grad_norm": 0.1484375,
+      "learning_rate": 0.02,
+      "loss": 1.6474,
+      "step": 3476
+    },
+    {
+      "epoch": 0.3218885610137129,
+      "grad_norm": 0.154296875,
+      "learning_rate": 0.02,
+      "loss": 1.6261,
+      "step": 3477
+    },
+    {
+      "epoch": 0.32198113753399293,
+      "grad_norm": 0.1494140625,
+      "learning_rate": 0.02,
+      "loss": 1.64,
+      "step": 3478
+    },
+    {
+      "epoch": 0.322073714054273,
+      "grad_norm": 0.146484375,
+      "learning_rate": 0.02,
+      "loss": 1.5817,
+      "step": 3479
+    },
+    {
+      "epoch": 0.32216629057455304,
+      "grad_norm": 0.1494140625,
+      "learning_rate": 0.02,
+      "loss": 1.6878,
+      "step": 3480
+    },
+    {
+      "epoch": 0.3222588670948331,
+      "grad_norm": 0.16015625,
+      "learning_rate": 0.02,
+      "loss": 1.7114,
+      "step": 3481
+    },
+    {
+      "epoch": 0.32235144361511314,
+      "grad_norm": 0.154296875,
+      "learning_rate": 0.02,
+      "loss": 1.6221,
+      "step": 3482
+    },
+    {
+      "epoch": 0.32244402013539314,
+      "grad_norm": 0.15625,
+      "learning_rate": 0.02,
+      "loss": 1.6737,
+      "step": 3483
+    },
+    {
+      "epoch": 0.3225365966556732,
+      "grad_norm": 0.15625,
+      "learning_rate": 0.02,
+      "loss": 1.6348,
+      "step": 3484
+    },
+    {
+      "epoch": 0.32262917317595324,
+      "grad_norm": 0.15234375,
+      "learning_rate": 0.02,
+      "loss": 1.6341,
+      "step": 3485
+    },
+    {
+      "epoch": 0.3227217496962333,
+      "grad_norm": 0.1591796875,
+      "learning_rate": 0.02,
+      "loss": 1.6601,
+      "step": 3486
+    },
+    {
+      "epoch": 0.32281432621651335,
+      "grad_norm": 0.14453125,
+      "learning_rate": 0.02,
+      "loss": 1.7011,
+      "step": 3487
+    },
+    {
+      "epoch": 0.3229069027367934,
+      "grad_norm": 0.1513671875,
+      "learning_rate": 0.02,
+      "loss": 1.6072,
+      "step": 3488
+    },
+    {
+      "epoch": 0.32299947925707345,
+      "grad_norm": 0.15234375,
+      "learning_rate": 0.02,
+      "loss": 1.6238,
+      "step": 3489
+    },
+    {
+      "epoch": 0.32309205577735345,
+      "grad_norm": 0.1513671875,
+      "learning_rate": 0.02,
+      "loss": 1.6359,
+      "step": 3490
+    },
+    {
+      "epoch": 0.3231846322976335,
+      "grad_norm": 0.1435546875,
+      "learning_rate": 0.02,
+      "loss": 1.6213,
+      "step": 3491
+    },
+    {
+      "epoch": 0.32327720881791355,
+      "grad_norm": 0.162109375,
+      "learning_rate": 0.02,
+      "loss": 1.6555,
+      "step": 3492
+    },
+    {
+      "epoch": 0.3233697853381936,
+      "grad_norm": 0.1455078125,
+      "learning_rate": 0.02,
+      "loss": 1.6087,
+      "step": 3493
+    },
+    {
+      "epoch": 0.32346236185847366,
+      "grad_norm": 0.1494140625,
+      "learning_rate": 0.02,
+      "loss": 1.6881,
+      "step": 3494
+    },
+    {
+      "epoch": 0.3235549383787537,
+      "grad_norm": 0.1484375,
+      "learning_rate": 0.02,
+      "loss": 1.6306,
+      "step": 3495
+    },
+    {
+      "epoch": 0.3236475148990337,
+      "grad_norm": 0.134765625,
+      "learning_rate": 0.02,
+      "loss": 1.6663,
+      "step": 3496
+    },
+    {
+      "epoch": 0.32374009141931376,
+      "grad_norm": 0.1484375,
+      "learning_rate": 0.02,
+      "loss": 1.6182,
+      "step": 3497
+    },
+    {
+      "epoch": 0.3238326679395938,
+      "grad_norm": 0.1484375,
+      "learning_rate": 0.02,
+      "loss": 1.6129,
+      "step": 3498
+    },
+    {
+      "epoch": 0.32392524445987386,
+      "grad_norm": 0.1396484375,
+      "learning_rate": 0.02,
+      "loss": 1.532,
+      "step": 3499
+    },
+    {
+      "epoch": 0.3240178209801539,
+      "grad_norm": 0.1455078125,
+      "learning_rate": 0.02,
+      "loss": 1.6481,
+      "step": 3500
+    },
+    {
+      "epoch": 0.32411039750043397,
+      "grad_norm": 0.1513671875,
+      "learning_rate": 0.02,
+      "loss": 1.583,
+      "step": 3501
+    },
+    {
+      "epoch": 0.324202974020714,
+      "grad_norm": 0.13671875,
+      "learning_rate": 0.02,
+      "loss": 1.6264,
+      "step": 3502
+    },
+    {
+      "epoch": 0.324295550540994,
+      "grad_norm": 0.140625,
+      "learning_rate": 0.02,
+      "loss": 1.653,
+      "step": 3503
+    },
+    {
+      "epoch": 0.32438812706127407,
+      "grad_norm": 0.1494140625,
+      "learning_rate": 0.02,
+      "loss": 1.6324,
+      "step": 3504
+    },
+    {
+      "epoch": 0.3244807035815541,
+      "grad_norm": 0.140625,
+      "learning_rate": 0.02,
+      "loss": 1.6921,
+      "step": 3505
+    },
+    {
+      "epoch": 0.32457328010183417,
+      "grad_norm": 0.142578125,
+      "learning_rate": 0.02,
+      "loss": 1.6196,
+      "step": 3506
+    },
+    {
+      "epoch": 0.3246658566221142,
+      "grad_norm": 0.1357421875,
+      "learning_rate": 0.02,
+      "loss": 1.6684,
+      "step": 3507
+    },
+    {
+      "epoch": 0.3247584331423943,
+      "grad_norm": 0.1376953125,
+      "learning_rate": 0.02,
+      "loss": 1.516,
+      "step": 3508
+    },
+    {
+      "epoch": 0.32485100966267433,
+      "grad_norm": 0.1357421875,
+      "learning_rate": 0.02,
+      "loss": 1.5396,
+      "step": 3509
+    },
+    {
+      "epoch": 0.3249435861829543,
+      "grad_norm": 0.1416015625,
+      "learning_rate": 0.02,
+      "loss": 1.662,
+      "step": 3510
+    },
+    {
+      "epoch": 0.3250361627032344,
+      "grad_norm": 0.146484375,
+      "learning_rate": 0.02,
+      "loss": 1.6509,
+      "step": 3511
+    },
+    {
+      "epoch": 0.32512873922351443,
+      "grad_norm": 0.1435546875,
+      "learning_rate": 0.02,
+      "loss": 1.5799,
+      "step": 3512
+    },
+    {
+      "epoch": 0.3252213157437945,
+      "grad_norm": 0.1396484375,
+      "learning_rate": 0.02,
+      "loss": 1.6464,
+      "step": 3513
+    },
+    {
+      "epoch": 0.32531389226407453,
+      "grad_norm": 0.142578125,
+      "learning_rate": 0.02,
+      "loss": 1.6565,
+      "step": 3514
+    },
+    {
+      "epoch": 0.3254064687843546,
+      "grad_norm": 0.1484375,
+      "learning_rate": 0.02,
+      "loss": 1.653,
+      "step": 3515
+    },
+    {
+      "epoch": 0.32549904530463464,
+      "grad_norm": 0.1435546875,
+      "learning_rate": 0.02,
+      "loss": 1.6026,
+      "step": 3516
+    },
+    {
+      "epoch": 0.32559162182491463,
+      "grad_norm": 0.1533203125,
+      "learning_rate": 0.02,
+      "loss": 1.6132,
+      "step": 3517
+    },
+    {
+      "epoch": 0.3256841983451947,
+      "grad_norm": 0.1591796875,
+      "learning_rate": 0.02,
+      "loss": 1.6225,
+      "step": 3518
+    },
+    {
+      "epoch": 0.32577677486547474,
+      "grad_norm": 0.1552734375,
+      "learning_rate": 0.02,
+      "loss": 1.6452,
+      "step": 3519
+    },
+    {
+      "epoch": 0.3258693513857548,
+      "grad_norm": 0.1513671875,
+      "learning_rate": 0.02,
+      "loss": 1.6392,
+      "step": 3520
+    },
+    {
+      "epoch": 0.32596192790603484,
+      "grad_norm": 0.1474609375,
+      "learning_rate": 0.02,
+      "loss": 1.6246,
+      "step": 3521
+    },
+    {
+      "epoch": 0.3260545044263149,
+      "grad_norm": 0.1494140625,
+      "learning_rate": 0.02,
+      "loss": 1.6611,
+      "step": 3522
+    },
+    {
+      "epoch": 0.32614708094659495,
+      "grad_norm": 0.1474609375,
+      "learning_rate": 0.02,
+      "loss": 1.6143,
+      "step": 3523
+    },
+    {
+      "epoch": 0.32623965746687494,
+      "grad_norm": 0.1416015625,
+      "learning_rate": 0.02,
+      "loss": 1.6248,
+      "step": 3524
+    },
+    {
+      "epoch": 0.326332233987155,
+      "grad_norm": 0.154296875,
+      "learning_rate": 0.02,
+      "loss": 1.6398,
+      "step": 3525
+    },
+    {
+      "epoch": 0.32642481050743505,
+      "grad_norm": 0.140625,
+      "learning_rate": 0.02,
+      "loss": 1.5938,
+      "step": 3526
+    },
+    {
+      "epoch": 0.3265173870277151,
+      "grad_norm": 0.1435546875,
+      "learning_rate": 0.02,
+      "loss": 1.5937,
+      "step": 3527
+    },
+    {
+      "epoch": 0.32660996354799515,
+      "grad_norm": 0.1416015625,
+      "learning_rate": 0.02,
+      "loss": 1.611,
+      "step": 3528
+    },
+    {
+      "epoch": 0.3267025400682752,
+      "grad_norm": 0.14453125,
+      "learning_rate": 0.02,
+      "loss": 1.6384,
+      "step": 3529
+    },
+    {
+      "epoch": 0.3267951165885552,
+      "grad_norm": 0.1376953125,
+      "learning_rate": 0.02,
+      "loss": 1.6031,
+      "step": 3530
+    },
+    {
+      "epoch": 0.32688769310883525,
+      "grad_norm": 0.140625,
+      "learning_rate": 0.02,
+      "loss": 1.6222,
+      "step": 3531
+    },
+    {
+      "epoch": 0.3269802696291153,
+      "grad_norm": 0.1474609375,
+      "learning_rate": 0.02,
+      "loss": 1.6251,
+      "step": 3532
+    },
+    {
+      "epoch": 0.32707284614939536,
+      "grad_norm": 0.1484375,
+      "learning_rate": 0.02,
+      "loss": 1.6246,
+      "step": 3533
+    },
+    {
+      "epoch": 0.3271654226696754,
+      "grad_norm": 0.142578125,
+      "learning_rate": 0.02,
+      "loss": 1.6047,
+      "step": 3534
+    },
+    {
+      "epoch": 0.32725799918995546,
+      "grad_norm": 0.1396484375,
+      "learning_rate": 0.02,
+      "loss": 1.6059,
+      "step": 3535
+    },
+    {
+      "epoch": 0.3273505757102355,
+      "grad_norm": 0.1376953125,
+      "learning_rate": 0.02,
+      "loss": 1.5769,
+      "step": 3536
+    },
+    {
+      "epoch": 0.3274431522305155,
+      "grad_norm": 0.1474609375,
+      "learning_rate": 0.02,
+      "loss": 1.6778,
+      "step": 3537
+    },
+    {
+      "epoch": 0.32753572875079556,
+      "grad_norm": 0.1513671875,
+      "learning_rate": 0.02,
+      "loss": 1.6636,
+      "step": 3538
+    },
+    {
+      "epoch": 0.3276283052710756,
+      "grad_norm": 0.1484375,
+      "learning_rate": 0.02,
+      "loss": 1.6217,
+      "step": 3539
+    },
+    {
+      "epoch": 0.32772088179135567,
+      "grad_norm": 0.150390625,
+      "learning_rate": 0.02,
+      "loss": 1.6622,
+      "step": 3540
+    },
+    {
+      "epoch": 0.3278134583116357,
+      "grad_norm": 0.1298828125,
+      "learning_rate": 0.02,
+      "loss": 1.594,
+      "step": 3541
+    },
+    {
+      "epoch": 0.32790603483191577,
+      "grad_norm": 0.1416015625,
+      "learning_rate": 0.02,
+      "loss": 1.5466,
+      "step": 3542
+    },
+    {
+      "epoch": 0.3279986113521958,
+      "grad_norm": 0.150390625,
+      "learning_rate": 0.02,
+      "loss": 1.6406,
+      "step": 3543
+    },
+    {
+      "epoch": 0.3280911878724758,
+      "grad_norm": 0.158203125,
+      "learning_rate": 0.02,
+      "loss": 1.7219,
+      "step": 3544
+    },
+    {
+      "epoch": 0.3281837643927559,
+      "grad_norm": 0.14453125,
+      "learning_rate": 0.02,
+      "loss": 1.5939,
+      "step": 3545
+    },
+    {
+      "epoch": 0.3282763409130359,
+      "grad_norm": 0.1513671875,
+      "learning_rate": 0.02,
+      "loss": 1.5639,
+      "step": 3546
+    },
+    {
+      "epoch": 0.328368917433316,
+      "grad_norm": 0.15234375,
+      "learning_rate": 0.02,
+      "loss": 1.5972,
+      "step": 3547
+    },
+    {
+      "epoch": 0.32846149395359603,
+      "grad_norm": 0.1552734375,
+      "learning_rate": 0.02,
+      "loss": 1.5744,
+      "step": 3548
+    },
+    {
+      "epoch": 0.3285540704738761,
+      "grad_norm": 0.1513671875,
+      "learning_rate": 0.02,
+      "loss": 1.6138,
+      "step": 3549
+    },
+    {
+      "epoch": 0.32864664699415613,
+      "grad_norm": 0.1572265625,
+      "learning_rate": 0.02,
+      "loss": 1.6863,
+      "step": 3550
+    },
+    {
+      "epoch": 0.32873922351443613,
+      "grad_norm": 0.16015625,
+      "learning_rate": 0.02,
+      "loss": 1.6147,
+      "step": 3551
+    },
+    {
+      "epoch": 0.3288318000347162,
+      "grad_norm": 0.16015625,
+      "learning_rate": 0.02,
+      "loss": 1.6539,
+      "step": 3552
+    },
+    {
+      "epoch": 0.32892437655499623,
+      "grad_norm": 0.1435546875,
+      "learning_rate": 0.02,
+      "loss": 1.6315,
+      "step": 3553
+    },
+    {
+      "epoch": 0.3290169530752763,
+      "grad_norm": 0.15234375,
+      "learning_rate": 0.02,
+      "loss": 1.6432,
+      "step": 3554
+    },
+    {
+      "epoch": 0.32910952959555634,
+      "grad_norm": 0.1533203125,
+      "learning_rate": 0.02,
+      "loss": 1.6377,
+      "step": 3555
+    },
+    {
+      "epoch": 0.3292021061158364,
+      "grad_norm": 0.150390625,
+      "learning_rate": 0.02,
+      "loss": 1.6561,
+      "step": 3556
+    },
+    {
+      "epoch": 0.32929468263611644,
+      "grad_norm": 0.150390625,
+      "learning_rate": 0.02,
+      "loss": 1.6354,
+      "step": 3557
+    },
+    {
+      "epoch": 0.32938725915639644,
+      "grad_norm": 0.15234375,
+      "learning_rate": 0.02,
+      "loss": 1.6712,
+      "step": 3558
+    },
+    {
+      "epoch": 0.3294798356766765,
+      "grad_norm": 0.14453125,
+      "learning_rate": 0.02,
+      "loss": 1.6444,
+      "step": 3559
+    },
+    {
+      "epoch": 0.32957241219695654,
+      "grad_norm": 0.134765625,
+      "learning_rate": 0.02,
+      "loss": 1.6125,
+      "step": 3560
+    },
+    {
+      "epoch": 0.3296649887172366,
+      "grad_norm": 0.140625,
+      "learning_rate": 0.02,
+      "loss": 1.6423,
+      "step": 3561
+    },
+    {
+      "epoch": 0.32975756523751665,
+      "grad_norm": 0.138671875,
+      "learning_rate": 0.02,
+      "loss": 1.6035,
+      "step": 3562
+    },
+    {
+      "epoch": 0.3298501417577967,
+      "grad_norm": 0.1416015625,
+      "learning_rate": 0.02,
+      "loss": 1.6691,
+      "step": 3563
+    },
+    {
+      "epoch": 0.3299427182780767,
+      "grad_norm": 0.1455078125,
+      "learning_rate": 0.02,
+      "loss": 1.6464,
+      "step": 3564
+    },
+    {
+      "epoch": 0.33003529479835675,
+      "grad_norm": 0.1494140625,
+      "learning_rate": 0.02,
+      "loss": 1.6471,
+      "step": 3565
+    },
+    {
+      "epoch": 0.3301278713186368,
+      "grad_norm": 0.134765625,
+      "learning_rate": 0.02,
+      "loss": 1.6549,
+      "step": 3566
+    },
+    {
+      "epoch": 0.33022044783891685,
+      "grad_norm": 0.146484375,
+      "learning_rate": 0.02,
+      "loss": 1.5738,
+      "step": 3567
+    },
+    {
+      "epoch": 0.3303130243591969,
+      "grad_norm": 0.14453125,
+      "learning_rate": 0.02,
+      "loss": 1.6771,
+      "step": 3568
+    },
+    {
+      "epoch": 0.33040560087947696,
+      "grad_norm": 0.138671875,
+      "learning_rate": 0.02,
+      "loss": 1.6506,
+      "step": 3569
+    },
+    {
+      "epoch": 0.330498177399757,
+      "grad_norm": 0.138671875,
+      "learning_rate": 0.02,
+      "loss": 1.575,
+      "step": 3570
+    },
+    {
+      "epoch": 0.330590753920037,
+      "grad_norm": 0.150390625,
+      "learning_rate": 0.02,
+      "loss": 1.639,
+      "step": 3571
+    },
+    {
+      "epoch": 0.33068333044031706,
+      "grad_norm": 0.150390625,
+      "learning_rate": 0.02,
+      "loss": 1.631,
+      "step": 3572
+    },
+    {
+      "epoch": 0.3307759069605971,
+      "grad_norm": 0.15234375,
+      "learning_rate": 0.02,
+      "loss": 1.6146,
+      "step": 3573
+    },
+    {
+      "epoch": 0.33086848348087716,
+      "grad_norm": 0.15625,
+      "learning_rate": 0.02,
+      "loss": 1.6657,
+      "step": 3574
+    },
+    {
+      "epoch": 0.3309610600011572,
+      "grad_norm": 0.16015625,
+      "learning_rate": 0.02,
+      "loss": 1.664,
+      "step": 3575
+    },
+    {
+      "epoch": 0.33105363652143727,
+      "grad_norm": 0.158203125,
+      "learning_rate": 0.02,
+      "loss": 1.5513,
+      "step": 3576
+    },
+    {
+      "epoch": 0.3311462130417173,
+      "grad_norm": 0.162109375,
+      "learning_rate": 0.02,
+      "loss": 1.5895,
+      "step": 3577
+    },
+    {
+      "epoch": 0.3312387895619973,
+      "grad_norm": 0.1708984375,
+      "learning_rate": 0.02,
+      "loss": 1.639,
+      "step": 3578
+    },
+    {
+      "epoch": 0.33133136608227737,
+      "grad_norm": 0.1533203125,
+      "learning_rate": 0.02,
+      "loss": 1.6016,
+      "step": 3579
+    },
+    {
+      "epoch": 0.3314239426025574,
+      "grad_norm": 0.15625,
+      "learning_rate": 0.02,
+      "loss": 1.6136,
+      "step": 3580
+    },
+    {
+      "epoch": 0.33151651912283747,
+      "grad_norm": 0.1484375,
+      "learning_rate": 0.02,
+      "loss": 1.639,
+      "step": 3581
+    },
+    {
+      "epoch": 0.3316090956431175,
+      "grad_norm": 0.1484375,
+      "learning_rate": 0.02,
+      "loss": 1.5776,
+      "step": 3582
+    },
+    {
+      "epoch": 0.3317016721633976,
+      "grad_norm": 0.150390625,
+      "learning_rate": 0.02,
+      "loss": 1.6496,
+      "step": 3583
+    },
+    {
+      "epoch": 0.33179424868367763,
+      "grad_norm": 0.16015625,
+      "learning_rate": 0.02,
+      "loss": 1.584,
+      "step": 3584
+    },
+    {
+      "epoch": 0.3318868252039576,
+      "grad_norm": 0.1455078125,
+      "learning_rate": 0.02,
+      "loss": 1.6763,
+      "step": 3585
+    },
+    {
+      "epoch": 0.3319794017242377,
+      "grad_norm": 0.1630859375,
+      "learning_rate": 0.02,
+      "loss": 1.6906,
+      "step": 3586
+    },
+    {
+      "epoch": 0.33207197824451773,
+      "grad_norm": 0.154296875,
+      "learning_rate": 0.02,
+      "loss": 1.6719,
+      "step": 3587
+    },
+    {
+      "epoch": 0.3321645547647978,
+      "grad_norm": 0.15234375,
+      "learning_rate": 0.02,
+      "loss": 1.6538,
+      "step": 3588
+    },
+    {
+      "epoch": 0.33225713128507783,
+      "grad_norm": 0.15625,
+      "learning_rate": 0.02,
+      "loss": 1.7027,
+      "step": 3589
+    },
+    {
+      "epoch": 0.3323497078053579,
+      "grad_norm": 0.1630859375,
+      "learning_rate": 0.02,
+      "loss": 1.6914,
+      "step": 3590
+    },
+    {
+      "epoch": 0.3324422843256379,
+      "grad_norm": 0.16015625,
+      "learning_rate": 0.02,
+      "loss": 1.6118,
+      "step": 3591
+    },
+    {
+      "epoch": 0.33253486084591793,
+      "grad_norm": 0.1435546875,
+      "learning_rate": 0.02,
+      "loss": 1.6219,
+      "step": 3592
+    },
+    {
+      "epoch": 0.332627437366198,
+      "grad_norm": 0.1416015625,
+      "learning_rate": 0.02,
+      "loss": 1.5402,
+      "step": 3593
+    },
+    {
+      "epoch": 0.33272001388647804,
+      "grad_norm": 0.1416015625,
+      "learning_rate": 0.02,
+      "loss": 1.6113,
+      "step": 3594
+    },
+    {
+      "epoch": 0.3328125904067581,
+      "grad_norm": 0.14453125,
+      "learning_rate": 0.02,
+      "loss": 1.6098,
+      "step": 3595
+    },
+    {
+      "epoch": 0.33290516692703814,
+      "grad_norm": 0.1484375,
+      "learning_rate": 0.02,
+      "loss": 1.6616,
+      "step": 3596
+    },
+    {
+      "epoch": 0.3329977434473182,
+      "grad_norm": 0.1533203125,
+      "learning_rate": 0.02,
+      "loss": 1.6963,
+      "step": 3597
+    },
+    {
+      "epoch": 0.3330903199675982,
+      "grad_norm": 0.1474609375,
+      "learning_rate": 0.02,
+      "loss": 1.5985,
+      "step": 3598
+    },
+    {
+      "epoch": 0.33318289648787824,
+      "grad_norm": 0.146484375,
+      "learning_rate": 0.02,
+      "loss": 1.6088,
+      "step": 3599
+    },
+    {
+      "epoch": 0.3332754730081583,
+      "grad_norm": 0.1396484375,
+      "learning_rate": 0.02,
+      "loss": 1.6495,
+      "step": 3600
+    },
+    {
+      "epoch": 0.33336804952843835,
+      "grad_norm": 0.1376953125,
+      "learning_rate": 0.02,
+      "loss": 1.5908,
+      "step": 3601
+    },
+    {
+      "epoch": 0.3334606260487184,
+      "grad_norm": 0.140625,
+      "learning_rate": 0.02,
+      "loss": 1.5552,
+      "step": 3602
+    },
+    {
+      "epoch": 0.33355320256899845,
+      "grad_norm": 0.1435546875,
+      "learning_rate": 0.02,
+      "loss": 1.5637,
+      "step": 3603
+    },
+    {
+      "epoch": 0.3336457790892785,
+      "grad_norm": 0.146484375,
+      "learning_rate": 0.02,
+      "loss": 1.5979,
+      "step": 3604
+    },
+    {
+      "epoch": 0.3337383556095585,
+      "grad_norm": 0.1484375,
+      "learning_rate": 0.02,
+      "loss": 1.6072,
+      "step": 3605
+    },
+    {
+      "epoch": 0.33383093212983855,
+      "grad_norm": 0.1455078125,
+      "learning_rate": 0.02,
+      "loss": 1.6743,
+      "step": 3606
+    },
+    {
+      "epoch": 0.3339235086501186,
+      "grad_norm": 0.1533203125,
+      "learning_rate": 0.02,
+      "loss": 1.6041,
+      "step": 3607
+    },
+    {
+      "epoch": 0.33401608517039866,
+      "grad_norm": 0.15625,
+      "learning_rate": 0.02,
+      "loss": 1.6683,
+      "step": 3608
+    },
+    {
+      "epoch": 0.3341086616906787,
+      "grad_norm": 0.1494140625,
+      "learning_rate": 0.02,
+      "loss": 1.6248,
+      "step": 3609
+    },
+    {
+      "epoch": 0.33420123821095876,
+      "grad_norm": 0.1455078125,
+      "learning_rate": 0.02,
+      "loss": 1.5878,
+      "step": 3610
+    },
+    {
+      "epoch": 0.3342938147312388,
+      "grad_norm": 0.150390625,
+      "learning_rate": 0.02,
+      "loss": 1.6758,
+      "step": 3611
+    },
+    {
+      "epoch": 0.3343863912515188,
+      "grad_norm": 0.158203125,
+      "learning_rate": 0.02,
+      "loss": 1.6826,
+      "step": 3612
+    },
+    {
+      "epoch": 0.33447896777179886,
+      "grad_norm": 0.1416015625,
+      "learning_rate": 0.02,
+      "loss": 1.6429,
+      "step": 3613
+    },
+    {
+      "epoch": 0.3345715442920789,
+      "grad_norm": 0.1455078125,
+      "learning_rate": 0.02,
+      "loss": 1.6361,
+      "step": 3614
+    },
+    {
+      "epoch": 0.33466412081235897,
+      "grad_norm": 0.146484375,
+      "learning_rate": 0.02,
+      "loss": 1.6642,
+      "step": 3615
+    },
+    {
+      "epoch": 0.334756697332639,
+      "grad_norm": 0.1474609375,
+      "learning_rate": 0.02,
+      "loss": 1.6747,
+      "step": 3616
+    },
+    {
+      "epoch": 0.33484927385291907,
+      "grad_norm": 0.14453125,
+      "learning_rate": 0.02,
+      "loss": 1.6602,
+      "step": 3617
+    },
+    {
+      "epoch": 0.3349418503731991,
+      "grad_norm": 0.1572265625,
+      "learning_rate": 0.02,
+      "loss": 1.7148,
+      "step": 3618
+    },
+    {
+      "epoch": 0.3350344268934791,
+      "grad_norm": 0.1455078125,
+      "learning_rate": 0.02,
+      "loss": 1.6145,
+      "step": 3619
+    },
+    {
+      "epoch": 0.3351270034137592,
+      "grad_norm": 0.140625,
+      "learning_rate": 0.02,
+      "loss": 1.6358,
+      "step": 3620
+    },
+    {
+      "epoch": 0.3352195799340392,
+      "grad_norm": 0.1474609375,
+      "learning_rate": 0.02,
+      "loss": 1.6383,
+      "step": 3621
+    },
+    {
+      "epoch": 0.3353121564543193,
+      "grad_norm": 0.1474609375,
+      "learning_rate": 0.02,
+      "loss": 1.6064,
+      "step": 3622
+    },
+    {
+      "epoch": 0.33540473297459933,
+      "grad_norm": 0.14453125,
+      "learning_rate": 0.02,
+      "loss": 1.6287,
+      "step": 3623
+    },
+    {
+      "epoch": 0.3354973094948794,
+      "grad_norm": 0.1484375,
+      "learning_rate": 0.02,
+      "loss": 1.6054,
+      "step": 3624
+    },
+    {
+      "epoch": 0.3355898860151594,
+      "grad_norm": 0.1474609375,
+      "learning_rate": 0.02,
+      "loss": 1.6386,
+      "step": 3625
+    },
+    {
+      "epoch": 0.33568246253543943,
+      "grad_norm": 0.14453125,
+      "learning_rate": 0.02,
+      "loss": 1.5893,
+      "step": 3626
+    },
+    {
+      "epoch": 0.3357750390557195,
+      "grad_norm": 0.140625,
+      "learning_rate": 0.02,
+      "loss": 1.5905,
+      "step": 3627
+    },
+    {
+      "epoch": 0.33586761557599953,
+      "grad_norm": 0.1357421875,
+      "learning_rate": 0.02,
+      "loss": 1.5678,
+      "step": 3628
+    },
+    {
+      "epoch": 0.3359601920962796,
+      "grad_norm": 0.1474609375,
+      "learning_rate": 0.02,
+      "loss": 1.6292,
+      "step": 3629
+    },
+    {
+      "epoch": 0.33605276861655964,
+      "grad_norm": 0.1435546875,
+      "learning_rate": 0.02,
+      "loss": 1.6397,
+      "step": 3630
+    },
+    {
+      "epoch": 0.3361453451368397,
+      "grad_norm": 0.1396484375,
+      "learning_rate": 0.02,
+      "loss": 1.6446,
+      "step": 3631
+    },
+    {
+      "epoch": 0.3362379216571197,
+      "grad_norm": 0.1513671875,
+      "learning_rate": 0.02,
+      "loss": 1.6185,
+      "step": 3632
+    },
+    {
+      "epoch": 0.33633049817739974,
+      "grad_norm": 0.15234375,
+      "learning_rate": 0.02,
+      "loss": 1.6329,
+      "step": 3633
+    },
+    {
+      "epoch": 0.3364230746976798,
+      "grad_norm": 0.154296875,
+      "learning_rate": 0.02,
+      "loss": 1.7175,
+      "step": 3634
+    },
+    {
+      "epoch": 0.33651565121795984,
+      "grad_norm": 0.1494140625,
+      "learning_rate": 0.02,
+      "loss": 1.5675,
+      "step": 3635
+    },
+    {
+      "epoch": 0.3366082277382399,
+      "grad_norm": 0.15234375,
+      "learning_rate": 0.02,
+      "loss": 1.6986,
+      "step": 3636
+    },
+    {
+      "epoch": 0.33670080425851995,
+      "grad_norm": 0.138671875,
+      "learning_rate": 0.02,
+      "loss": 1.5875,
+      "step": 3637
+    },
+    {
+      "epoch": 0.3367933807788,
+      "grad_norm": 0.1474609375,
+      "learning_rate": 0.02,
+      "loss": 1.6244,
+      "step": 3638
+    },
+    {
+      "epoch": 0.33688595729908,
+      "grad_norm": 0.13671875,
+      "learning_rate": 0.02,
+      "loss": 1.6303,
+      "step": 3639
+    },
+    {
+      "epoch": 0.33697853381936005,
+      "grad_norm": 0.140625,
+      "learning_rate": 0.02,
+      "loss": 1.7024,
+      "step": 3640
+    },
+    {
+      "epoch": 0.3370711103396401,
+      "grad_norm": 0.1240234375,
+      "learning_rate": 0.02,
+      "loss": 1.5659,
+      "step": 3641
+    },
+    {
+      "epoch": 0.33716368685992015,
+      "grad_norm": 0.1298828125,
+      "learning_rate": 0.02,
+      "loss": 1.6211,
+      "step": 3642
+    },
+    {
+      "epoch": 0.3372562633802002,
+      "grad_norm": 0.130859375,
+      "learning_rate": 0.02,
+      "loss": 1.6761,
+      "step": 3643
+    },
+    {
+      "epoch": 0.33734883990048026,
+      "grad_norm": 0.1435546875,
+      "learning_rate": 0.02,
+      "loss": 1.6796,
+      "step": 3644
+    },
+    {
+      "epoch": 0.3374414164207603,
+      "grad_norm": 0.1298828125,
+      "learning_rate": 0.02,
+      "loss": 1.5816,
+      "step": 3645
+    },
+    {
+      "epoch": 0.3375339929410403,
+      "grad_norm": 0.1337890625,
+      "learning_rate": 0.02,
+      "loss": 1.5651,
+      "step": 3646
+    },
+    {
+      "epoch": 0.33762656946132036,
+      "grad_norm": 0.1396484375,
+      "learning_rate": 0.02,
+      "loss": 1.5543,
+      "step": 3647
+    },
+    {
+      "epoch": 0.3377191459816004,
+      "grad_norm": 0.142578125,
+      "learning_rate": 0.02,
+      "loss": 1.6741,
+      "step": 3648
+    },
+    {
+      "epoch": 0.33781172250188046,
+      "grad_norm": 0.1494140625,
+      "learning_rate": 0.02,
+      "loss": 1.7472,
+      "step": 3649
+    },
+    {
+      "epoch": 0.3379042990221605,
+      "grad_norm": 0.154296875,
+      "learning_rate": 0.02,
+      "loss": 1.578,
+      "step": 3650
+    },
+    {
+      "epoch": 0.33799687554244057,
+      "grad_norm": 0.154296875,
+      "learning_rate": 0.02,
+      "loss": 1.5945,
+      "step": 3651
+    },
+    {
+      "epoch": 0.3380894520627206,
+      "grad_norm": 0.1455078125,
+      "learning_rate": 0.02,
+      "loss": 1.6126,
+      "step": 3652
+    },
+    {
+      "epoch": 0.3381820285830006,
+      "grad_norm": 0.146484375,
+      "learning_rate": 0.02,
+      "loss": 1.6367,
+      "step": 3653
+    },
+    {
+      "epoch": 0.33827460510328067,
+      "grad_norm": 0.1455078125,
+      "learning_rate": 0.02,
+      "loss": 1.5793,
+      "step": 3654
+    },
+    {
+      "epoch": 0.3383671816235607,
+      "grad_norm": 0.1611328125,
+      "learning_rate": 0.02,
+      "loss": 1.6068,
+      "step": 3655
+    },
+    {
+      "epoch": 0.33845975814384077,
+      "grad_norm": 0.146484375,
+      "learning_rate": 0.02,
+      "loss": 1.6714,
+      "step": 3656
+    },
+    {
+      "epoch": 0.3385523346641208,
+      "grad_norm": 0.1611328125,
+      "learning_rate": 0.02,
+      "loss": 1.588,
+      "step": 3657
+    },
+    {
+      "epoch": 0.3386449111844009,
+      "grad_norm": 0.140625,
+      "learning_rate": 0.02,
+      "loss": 1.6489,
+      "step": 3658
+    },
+    {
+      "epoch": 0.3387374877046809,
+      "grad_norm": 0.1474609375,
+      "learning_rate": 0.02,
+      "loss": 1.6349,
+      "step": 3659
+    },
+    {
+      "epoch": 0.3388300642249609,
+      "grad_norm": 0.1533203125,
+      "learning_rate": 0.02,
+      "loss": 1.6639,
+      "step": 3660
+    },
+    {
+      "epoch": 0.338922640745241,
+      "grad_norm": 0.1640625,
+      "learning_rate": 0.02,
+      "loss": 1.6153,
+      "step": 3661
+    },
+    {
+      "epoch": 0.33901521726552103,
+      "grad_norm": 0.15234375,
+      "learning_rate": 0.02,
+      "loss": 1.6417,
+      "step": 3662
+    },
+    {
+      "epoch": 0.3391077937858011,
+      "grad_norm": 0.158203125,
+      "learning_rate": 0.02,
+      "loss": 1.5968,
+      "step": 3663
+    },
+    {
+      "epoch": 0.33920037030608113,
+      "grad_norm": 0.1552734375,
+      "learning_rate": 0.02,
+      "loss": 1.6297,
+      "step": 3664
+    },
+    {
+      "epoch": 0.3392929468263612,
+      "grad_norm": 0.158203125,
+      "learning_rate": 0.02,
+      "loss": 1.6368,
+      "step": 3665
+    },
+    {
+      "epoch": 0.3393855233466412,
+      "grad_norm": 0.1513671875,
+      "learning_rate": 0.02,
+      "loss": 1.6357,
+      "step": 3666
+    },
+    {
+      "epoch": 0.33947809986692123,
+      "grad_norm": 0.1494140625,
+      "learning_rate": 0.02,
+      "loss": 1.6479,
+      "step": 3667
+    },
+    {
+      "epoch": 0.3395706763872013,
+      "grad_norm": 0.1591796875,
+      "learning_rate": 0.02,
+      "loss": 1.7271,
+      "step": 3668
+    },
+    {
+      "epoch": 0.33966325290748134,
+      "grad_norm": 0.14453125,
+      "learning_rate": 0.02,
+      "loss": 1.5947,
+      "step": 3669
+    },
+    {
+      "epoch": 0.3397558294277614,
+      "grad_norm": 0.1474609375,
+      "learning_rate": 0.02,
+      "loss": 1.6121,
+      "step": 3670
+    },
+    {
+      "epoch": 0.33984840594804144,
+      "grad_norm": 0.1435546875,
+      "learning_rate": 0.02,
+      "loss": 1.6151,
+      "step": 3671
+    },
+    {
+      "epoch": 0.3399409824683215,
+      "grad_norm": 0.1513671875,
+      "learning_rate": 0.02,
+      "loss": 1.6224,
+      "step": 3672
+    },
+    {
+      "epoch": 0.3400335589886015,
+      "grad_norm": 0.146484375,
+      "learning_rate": 0.02,
+      "loss": 1.6478,
+      "step": 3673
+    },
+    {
+      "epoch": 0.34012613550888154,
+      "grad_norm": 0.1513671875,
+      "learning_rate": 0.02,
+      "loss": 1.5428,
+      "step": 3674
+    },
+    {
+      "epoch": 0.3402187120291616,
+      "grad_norm": 0.1484375,
+      "learning_rate": 0.02,
+      "loss": 1.6599,
+      "step": 3675
+    },
+    {
+      "epoch": 0.34031128854944165,
+      "grad_norm": 0.1494140625,
+      "learning_rate": 0.02,
+      "loss": 1.6393,
+      "step": 3676
+    },
+    {
+      "epoch": 0.3404038650697217,
+      "grad_norm": 0.1513671875,
+      "learning_rate": 0.02,
+      "loss": 1.5852,
+      "step": 3677
+    },
+    {
+      "epoch": 0.34049644159000175,
+      "grad_norm": 0.1591796875,
+      "learning_rate": 0.02,
+      "loss": 1.5561,
+      "step": 3678
+    },
+    {
+      "epoch": 0.3405890181102818,
+      "grad_norm": 0.1474609375,
+      "learning_rate": 0.02,
+      "loss": 1.6326,
+      "step": 3679
+    },
+    {
+      "epoch": 0.3406815946305618,
+      "grad_norm": 0.1455078125,
+      "learning_rate": 0.02,
+      "loss": 1.5907,
+      "step": 3680
+    },
+    {
+      "epoch": 0.34077417115084185,
+      "grad_norm": 0.1591796875,
+      "learning_rate": 0.02,
+      "loss": 1.6777,
+      "step": 3681
+    },
+    {
+      "epoch": 0.3408667476711219,
+      "grad_norm": 0.1552734375,
+      "learning_rate": 0.02,
+      "loss": 1.6489,
+      "step": 3682
+    },
+    {
+      "epoch": 0.34095932419140196,
+      "grad_norm": 0.1533203125,
+      "learning_rate": 0.02,
+      "loss": 1.6372,
+      "step": 3683
+    },
+    {
+      "epoch": 0.341051900711682,
+      "grad_norm": 0.1474609375,
+      "learning_rate": 0.02,
+      "loss": 1.6049,
+      "step": 3684
+    },
+    {
+      "epoch": 0.34114447723196206,
+      "grad_norm": 0.1484375,
+      "learning_rate": 0.02,
+      "loss": 1.5568,
+      "step": 3685
+    },
+    {
+      "epoch": 0.3412370537522421,
+      "grad_norm": 0.1484375,
+      "learning_rate": 0.02,
+      "loss": 1.583,
+      "step": 3686
+    },
+    {
+      "epoch": 0.3413296302725221,
+      "grad_norm": 0.1650390625,
+      "learning_rate": 0.02,
+      "loss": 1.6499,
+      "step": 3687
+    },
+    {
+      "epoch": 0.34142220679280216,
+      "grad_norm": 0.150390625,
+      "learning_rate": 0.02,
+      "loss": 1.6281,
+      "step": 3688
+    },
+    {
+      "epoch": 0.3415147833130822,
+      "grad_norm": 0.154296875,
+      "learning_rate": 0.02,
+      "loss": 1.6792,
+      "step": 3689
+    },
+    {
+      "epoch": 0.34160735983336227,
+      "grad_norm": 0.14453125,
+      "learning_rate": 0.02,
+      "loss": 1.6357,
+      "step": 3690
+    },
+    {
+      "epoch": 0.3416999363536423,
+      "grad_norm": 0.1572265625,
+      "learning_rate": 0.02,
+      "loss": 1.6354,
+      "step": 3691
+    },
+    {
+      "epoch": 0.34179251287392237,
+      "grad_norm": 0.1455078125,
+      "learning_rate": 0.02,
+      "loss": 1.5725,
+      "step": 3692
+    },
+    {
+      "epoch": 0.34188508939420237,
+      "grad_norm": 0.15234375,
+      "learning_rate": 0.02,
+      "loss": 1.6339,
+      "step": 3693
+    },
+    {
+      "epoch": 0.3419776659144824,
+      "grad_norm": 0.12890625,
+      "learning_rate": 0.02,
+      "loss": 1.6186,
+      "step": 3694
+    },
+    {
+      "epoch": 0.3420702424347625,
+      "grad_norm": 0.140625,
+      "learning_rate": 0.02,
+      "loss": 1.6146,
+      "step": 3695
+    },
+    {
+      "epoch": 0.3421628189550425,
+      "grad_norm": 0.140625,
+      "learning_rate": 0.02,
+      "loss": 1.6581,
+      "step": 3696
+    },
+    {
+      "epoch": 0.3422553954753226,
+      "grad_norm": 0.1474609375,
+      "learning_rate": 0.02,
+      "loss": 1.6254,
+      "step": 3697
+    },
+    {
+      "epoch": 0.34234797199560263,
+      "grad_norm": 0.146484375,
+      "learning_rate": 0.02,
+      "loss": 1.6768,
+      "step": 3698
+    },
+    {
+      "epoch": 0.3424405485158827,
+      "grad_norm": 0.142578125,
+      "learning_rate": 0.02,
+      "loss": 1.5825,
+      "step": 3699
+    },
+    {
+      "epoch": 0.3425331250361627,
+      "grad_norm": 0.1484375,
+      "learning_rate": 0.02,
+      "loss": 1.639,
+      "step": 3700
+    },
+    {
+      "epoch": 0.34262570155644273,
+      "grad_norm": 0.1484375,
+      "learning_rate": 0.02,
+      "loss": 1.5902,
+      "step": 3701
+    },
+    {
+      "epoch": 0.3427182780767228,
+      "grad_norm": 0.1494140625,
+      "learning_rate": 0.02,
+      "loss": 1.6316,
+      "step": 3702
+    },
+    {
+      "epoch": 0.34281085459700283,
+      "grad_norm": 0.1455078125,
+      "learning_rate": 0.02,
+      "loss": 1.6078,
+      "step": 3703
+    },
+    {
+      "epoch": 0.3429034311172829,
+      "grad_norm": 0.1484375,
+      "learning_rate": 0.02,
+      "loss": 1.6179,
+      "step": 3704
+    },
+    {
+      "epoch": 0.34299600763756294,
+      "grad_norm": 0.1484375,
+      "learning_rate": 0.02,
+      "loss": 1.5844,
+      "step": 3705
+    },
+    {
+      "epoch": 0.343088584157843,
+      "grad_norm": 0.1552734375,
+      "learning_rate": 0.02,
+      "loss": 1.6676,
+      "step": 3706
+    },
+    {
+      "epoch": 0.343181160678123,
+      "grad_norm": 0.154296875,
+      "learning_rate": 0.02,
+      "loss": 1.6073,
+      "step": 3707
+    },
+    {
+      "epoch": 0.34327373719840304,
+      "grad_norm": 0.1484375,
+      "learning_rate": 0.02,
+      "loss": 1.6562,
+      "step": 3708
+    },
+    {
+      "epoch": 0.3433663137186831,
+      "grad_norm": 0.1513671875,
+      "learning_rate": 0.02,
+      "loss": 1.6269,
+      "step": 3709
+    },
+    {
+      "epoch": 0.34345889023896314,
+      "grad_norm": 0.1533203125,
+      "learning_rate": 0.02,
+      "loss": 1.6232,
+      "step": 3710
+    },
+    {
+      "epoch": 0.3435514667592432,
+      "grad_norm": 0.1484375,
+      "learning_rate": 0.02,
+      "loss": 1.6839,
+      "step": 3711
+    },
+    {
+      "epoch": 0.34364404327952325,
+      "grad_norm": 0.15625,
+      "learning_rate": 0.02,
+      "loss": 1.6347,
+      "step": 3712
+    },
+    {
+      "epoch": 0.3437366197998033,
+      "grad_norm": 0.1484375,
+      "learning_rate": 0.02,
+      "loss": 1.5939,
+      "step": 3713
+    },
+    {
+      "epoch": 0.3438291963200833,
+      "grad_norm": 0.1455078125,
+      "learning_rate": 0.02,
+      "loss": 1.5895,
+      "step": 3714
+    },
+    {
+      "epoch": 0.34392177284036335,
+      "grad_norm": 0.1552734375,
+      "learning_rate": 0.02,
+      "loss": 1.7031,
+      "step": 3715
+    },
+    {
+      "epoch": 0.3440143493606434,
+      "grad_norm": 0.1474609375,
+      "learning_rate": 0.02,
+      "loss": 1.5668,
+      "step": 3716
+    },
+    {
+      "epoch": 0.34410692588092345,
+      "grad_norm": 0.1494140625,
+      "learning_rate": 0.02,
+      "loss": 1.653,
+      "step": 3717
+    },
+    {
+      "epoch": 0.3441995024012035,
+      "grad_norm": 0.15625,
+      "learning_rate": 0.02,
+      "loss": 1.6684,
+      "step": 3718
+    },
+    {
+      "epoch": 0.34429207892148356,
+      "grad_norm": 0.15234375,
+      "learning_rate": 0.02,
+      "loss": 1.6562,
+      "step": 3719
+    },
+    {
+      "epoch": 0.3443846554417636,
+      "grad_norm": 0.142578125,
+      "learning_rate": 0.02,
+      "loss": 1.5942,
+      "step": 3720
+    },
+    {
+      "epoch": 0.3444772319620436,
+      "grad_norm": 0.150390625,
+      "learning_rate": 0.02,
+      "loss": 1.6845,
+      "step": 3721
+    },
+    {
+      "epoch": 0.34456980848232366,
+      "grad_norm": 0.1484375,
+      "learning_rate": 0.02,
+      "loss": 1.5711,
+      "step": 3722
+    },
+    {
+      "epoch": 0.3446623850026037,
+      "grad_norm": 0.1513671875,
+      "learning_rate": 0.02,
+      "loss": 1.6205,
+      "step": 3723
+    },
+    {
+      "epoch": 0.34475496152288376,
+      "grad_norm": 0.1474609375,
+      "learning_rate": 0.02,
+      "loss": 1.5916,
+      "step": 3724
+    },
+    {
+      "epoch": 0.3448475380431638,
+      "grad_norm": 0.1435546875,
+      "learning_rate": 0.02,
+      "loss": 1.6276,
+      "step": 3725
+    },
+    {
+      "epoch": 0.34494011456344387,
+      "grad_norm": 0.1474609375,
+      "learning_rate": 0.02,
+      "loss": 1.66,
+      "step": 3726
+    },
+    {
+      "epoch": 0.34503269108372386,
+      "grad_norm": 0.1396484375,
+      "learning_rate": 0.02,
+      "loss": 1.5579,
+      "step": 3727
+    },
+    {
+      "epoch": 0.3451252676040039,
+      "grad_norm": 0.1376953125,
+      "learning_rate": 0.02,
+      "loss": 1.6133,
+      "step": 3728
+    },
+    {
+      "epoch": 0.34521784412428397,
+      "grad_norm": 0.1376953125,
+      "learning_rate": 0.02,
+      "loss": 1.6028,
+      "step": 3729
+    },
+    {
+      "epoch": 0.345310420644564,
+      "grad_norm": 0.146484375,
+      "learning_rate": 0.02,
+      "loss": 1.5933,
+      "step": 3730
+    },
+    {
+      "epoch": 0.34540299716484407,
+      "grad_norm": 0.138671875,
+      "learning_rate": 0.02,
+      "loss": 1.6603,
+      "step": 3731
+    },
+    {
+      "epoch": 0.3454955736851241,
+      "grad_norm": 0.1435546875,
+      "learning_rate": 0.02,
+      "loss": 1.5888,
+      "step": 3732
+    },
+    {
+      "epoch": 0.3455881502054042,
+      "grad_norm": 0.1318359375,
+      "learning_rate": 0.02,
+      "loss": 1.6662,
+      "step": 3733
+    },
+    {
+      "epoch": 0.3456807267256842,
+      "grad_norm": 0.134765625,
+      "learning_rate": 0.02,
+      "loss": 1.5963,
+      "step": 3734
+    },
+    {
+      "epoch": 0.3457733032459642,
+      "grad_norm": 0.1455078125,
+      "learning_rate": 0.02,
+      "loss": 1.6505,
+      "step": 3735
+    },
+    {
+      "epoch": 0.3458658797662443,
+      "grad_norm": 0.138671875,
+      "learning_rate": 0.02,
+      "loss": 1.5623,
+      "step": 3736
+    },
+    {
+      "epoch": 0.34595845628652433,
+      "grad_norm": 0.15234375,
+      "learning_rate": 0.02,
+      "loss": 1.6733,
+      "step": 3737
+    },
+    {
+      "epoch": 0.3460510328068044,
+      "grad_norm": 0.140625,
+      "learning_rate": 0.02,
+      "loss": 1.7053,
+      "step": 3738
+    },
+    {
+      "epoch": 0.34614360932708443,
+      "grad_norm": 0.1484375,
+      "learning_rate": 0.02,
+      "loss": 1.5926,
+      "step": 3739
+    },
+    {
+      "epoch": 0.3462361858473645,
+      "grad_norm": 0.158203125,
+      "learning_rate": 0.02,
+      "loss": 1.5331,
+      "step": 3740
+    },
+    {
+      "epoch": 0.3463287623676445,
+      "grad_norm": 0.14453125,
+      "learning_rate": 0.02,
+      "loss": 1.6089,
+      "step": 3741
+    },
+    {
+      "epoch": 0.34642133888792453,
+      "grad_norm": 0.1435546875,
+      "learning_rate": 0.02,
+      "loss": 1.6237,
+      "step": 3742
+    },
+    {
+      "epoch": 0.3465139154082046,
+      "grad_norm": 0.1474609375,
+      "learning_rate": 0.02,
+      "loss": 1.6297,
+      "step": 3743
+    },
+    {
+      "epoch": 0.34660649192848464,
+      "grad_norm": 0.1484375,
+      "learning_rate": 0.02,
+      "loss": 1.6532,
+      "step": 3744
+    },
+    {
+      "epoch": 0.3466990684487647,
+      "grad_norm": 0.1611328125,
+      "learning_rate": 0.02,
+      "loss": 1.5554,
+      "step": 3745
+    },
+    {
+      "epoch": 0.34679164496904474,
+      "grad_norm": 0.154296875,
+      "learning_rate": 0.02,
+      "loss": 1.5998,
+      "step": 3746
+    },
+    {
+      "epoch": 0.3468842214893248,
+      "grad_norm": 0.1455078125,
+      "learning_rate": 0.02,
+      "loss": 1.5607,
+      "step": 3747
+    },
+    {
+      "epoch": 0.3469767980096048,
+      "grad_norm": 0.1494140625,
+      "learning_rate": 0.02,
+      "loss": 1.6011,
+      "step": 3748
+    },
+    {
+      "epoch": 0.34706937452988484,
+      "grad_norm": 0.1552734375,
+      "learning_rate": 0.02,
+      "loss": 1.6114,
+      "step": 3749
+    },
+    {
+      "epoch": 0.3471619510501649,
+      "grad_norm": 0.1474609375,
+      "learning_rate": 0.02,
+      "loss": 1.6484,
+      "step": 3750
+    },
+    {
+      "epoch": 0.34725452757044495,
+      "grad_norm": 0.158203125,
+      "learning_rate": 0.02,
+      "loss": 1.6572,
+      "step": 3751
+    },
+    {
+      "epoch": 0.347347104090725,
+      "grad_norm": 0.1494140625,
+      "learning_rate": 0.02,
+      "loss": 1.6082,
+      "step": 3752
+    },
+    {
+      "epoch": 0.34743968061100505,
+      "grad_norm": 0.1484375,
+      "learning_rate": 0.02,
+      "loss": 1.5558,
+      "step": 3753
+    },
+    {
+      "epoch": 0.3475322571312851,
+      "grad_norm": 0.14453125,
+      "learning_rate": 0.02,
+      "loss": 1.5755,
+      "step": 3754
+    },
+    {
+      "epoch": 0.3476248336515651,
+      "grad_norm": 0.1494140625,
+      "learning_rate": 0.02,
+      "loss": 1.6208,
+      "step": 3755
+    },
+    {
+      "epoch": 0.34771741017184515,
+      "grad_norm": 0.1435546875,
+      "learning_rate": 0.02,
+      "loss": 1.5578,
+      "step": 3756
+    },
+    {
+      "epoch": 0.3478099866921252,
+      "grad_norm": 0.1513671875,
+      "learning_rate": 0.02,
+      "loss": 1.6016,
+      "step": 3757
+    },
+    {
+      "epoch": 0.34790256321240526,
+      "grad_norm": 0.1455078125,
+      "learning_rate": 0.02,
+      "loss": 1.6373,
+      "step": 3758
+    },
+    {
+      "epoch": 0.3479951397326853,
+      "grad_norm": 0.15234375,
+      "learning_rate": 0.02,
+      "loss": 1.6537,
+      "step": 3759
+    },
+    {
+      "epoch": 0.34808771625296536,
+      "grad_norm": 0.1396484375,
+      "learning_rate": 0.02,
+      "loss": 1.6472,
+      "step": 3760
+    },
+    {
+      "epoch": 0.34818029277324536,
+      "grad_norm": 0.146484375,
+      "learning_rate": 0.02,
+      "loss": 1.6512,
+      "step": 3761
+    },
+    {
+      "epoch": 0.3482728692935254,
+      "grad_norm": 0.1396484375,
+      "learning_rate": 0.02,
+      "loss": 1.6488,
+      "step": 3762
+    },
+    {
+      "epoch": 0.34836544581380546,
+      "grad_norm": 0.1396484375,
+      "learning_rate": 0.02,
+      "loss": 1.6077,
+      "step": 3763
+    },
+    {
+      "epoch": 0.3484580223340855,
+      "grad_norm": 0.1337890625,
+      "learning_rate": 0.02,
+      "loss": 1.6731,
+      "step": 3764
+    },
+    {
+      "epoch": 0.34855059885436557,
+      "grad_norm": 0.142578125,
+      "learning_rate": 0.02,
+      "loss": 1.6217,
+      "step": 3765
+    },
+    {
+      "epoch": 0.3486431753746456,
+      "grad_norm": 0.1337890625,
+      "learning_rate": 0.02,
+      "loss": 1.6024,
+      "step": 3766
+    },
+    {
+      "epoch": 0.34873575189492567,
+      "grad_norm": 0.15234375,
+      "learning_rate": 0.02,
+      "loss": 1.6348,
+      "step": 3767
+    },
+    {
+      "epoch": 0.34882832841520567,
+      "grad_norm": 0.146484375,
+      "learning_rate": 0.02,
+      "loss": 1.6067,
+      "step": 3768
+    },
+    {
+      "epoch": 0.3489209049354857,
+      "grad_norm": 0.14453125,
+      "learning_rate": 0.02,
+      "loss": 1.6543,
+      "step": 3769
+    },
+    {
+      "epoch": 0.3490134814557658,
+      "grad_norm": 0.150390625,
+      "learning_rate": 0.02,
+      "loss": 1.6383,
+      "step": 3770
+    },
+    {
+      "epoch": 0.3491060579760458,
+      "grad_norm": 0.1376953125,
+      "learning_rate": 0.02,
+      "loss": 1.6562,
+      "step": 3771
+    },
+    {
+      "epoch": 0.3491986344963259,
+      "grad_norm": 0.1396484375,
+      "learning_rate": 0.02,
+      "loss": 1.635,
+      "step": 3772
+    },
+    {
+      "epoch": 0.34929121101660593,
+      "grad_norm": 0.140625,
+      "learning_rate": 0.02,
+      "loss": 1.6485,
+      "step": 3773
+    },
+    {
+      "epoch": 0.349383787536886,
+      "grad_norm": 0.1533203125,
+      "learning_rate": 0.02,
+      "loss": 1.646,
+      "step": 3774
+    },
+    {
+      "epoch": 0.349476364057166,
+      "grad_norm": 0.150390625,
+      "learning_rate": 0.02,
+      "loss": 1.6126,
+      "step": 3775
+    },
+    {
+      "epoch": 0.34956894057744603,
+      "grad_norm": 0.1474609375,
+      "learning_rate": 0.02,
+      "loss": 1.5677,
+      "step": 3776
+    },
+    {
+      "epoch": 0.3496615170977261,
+      "grad_norm": 0.1484375,
+      "learning_rate": 0.02,
+      "loss": 1.6085,
+      "step": 3777
+    },
+    {
+      "epoch": 0.34975409361800613,
+      "grad_norm": 0.1474609375,
+      "learning_rate": 0.02,
+      "loss": 1.614,
+      "step": 3778
+    },
+    {
+      "epoch": 0.3498466701382862,
+      "grad_norm": 0.1376953125,
+      "learning_rate": 0.02,
+      "loss": 1.648,
+      "step": 3779
+    },
+    {
+      "epoch": 0.34993924665856624,
+      "grad_norm": 0.1435546875,
+      "learning_rate": 0.02,
+      "loss": 1.6374,
+      "step": 3780
+    },
+    {
+      "epoch": 0.3500318231788463,
+      "grad_norm": 0.1533203125,
+      "learning_rate": 0.02,
+      "loss": 1.5644,
+      "step": 3781
+    },
+    {
+      "epoch": 0.3501243996991263,
+      "grad_norm": 0.154296875,
+      "learning_rate": 0.02,
+      "loss": 1.6079,
+      "step": 3782
+    },
+    {
+      "epoch": 0.35021697621940634,
+      "grad_norm": 0.1533203125,
+      "learning_rate": 0.02,
+      "loss": 1.6498,
+      "step": 3783
+    },
+    {
+      "epoch": 0.3503095527396864,
+      "grad_norm": 0.1533203125,
+      "learning_rate": 0.02,
+      "loss": 1.5614,
+      "step": 3784
+    },
+    {
+      "epoch": 0.35040212925996644,
+      "grad_norm": 0.1513671875,
+      "learning_rate": 0.02,
+      "loss": 1.6637,
+      "step": 3785
+    },
+    {
+      "epoch": 0.3504947057802465,
+      "grad_norm": 0.142578125,
+      "learning_rate": 0.02,
+      "loss": 1.612,
+      "step": 3786
+    },
+    {
+      "epoch": 0.35058728230052655,
+      "grad_norm": 0.140625,
+      "learning_rate": 0.02,
+      "loss": 1.5839,
+      "step": 3787
+    },
+    {
+      "epoch": 0.3506798588208066,
+      "grad_norm": 0.142578125,
+      "learning_rate": 0.02,
+      "loss": 1.5602,
+      "step": 3788
+    },
+    {
+      "epoch": 0.3507724353410866,
+      "grad_norm": 0.1416015625,
+      "learning_rate": 0.02,
+      "loss": 1.5884,
+      "step": 3789
+    },
+    {
+      "epoch": 0.35086501186136665,
+      "grad_norm": 0.134765625,
+      "learning_rate": 0.02,
+      "loss": 1.6025,
+      "step": 3790
+    },
+    {
+      "epoch": 0.3509575883816467,
+      "grad_norm": 0.1376953125,
+      "learning_rate": 0.02,
+      "loss": 1.5644,
+      "step": 3791
+    },
+    {
+      "epoch": 0.35105016490192675,
+      "grad_norm": 0.142578125,
+      "learning_rate": 0.02,
+      "loss": 1.6227,
+      "step": 3792
+    },
+    {
+      "epoch": 0.3511427414222068,
+      "grad_norm": 0.146484375,
+      "learning_rate": 0.02,
+      "loss": 1.6218,
+      "step": 3793
+    },
+    {
+      "epoch": 0.35123531794248686,
+      "grad_norm": 0.140625,
+      "learning_rate": 0.02,
+      "loss": 1.6603,
+      "step": 3794
+    },
+    {
+      "epoch": 0.35132789446276685,
+      "grad_norm": 0.142578125,
+      "learning_rate": 0.02,
+      "loss": 1.6046,
+      "step": 3795
+    },
+    {
+      "epoch": 0.3514204709830469,
+      "grad_norm": 0.1494140625,
+      "learning_rate": 0.02,
+      "loss": 1.6306,
+      "step": 3796
+    },
+    {
+      "epoch": 0.35151304750332696,
+      "grad_norm": 0.1455078125,
+      "learning_rate": 0.02,
+      "loss": 1.6375,
+      "step": 3797
+    },
+    {
+      "epoch": 0.351605624023607,
+      "grad_norm": 0.1455078125,
+      "learning_rate": 0.02,
+      "loss": 1.6258,
+      "step": 3798
+    },
+    {
+      "epoch": 0.35169820054388706,
+      "grad_norm": 0.1455078125,
+      "learning_rate": 0.02,
+      "loss": 1.6587,
+      "step": 3799
+    },
+    {
+      "epoch": 0.3517907770641671,
+      "grad_norm": 0.140625,
+      "learning_rate": 0.02,
+      "loss": 1.637,
+      "step": 3800
+    },
+    {
+      "epoch": 0.35188335358444717,
+      "grad_norm": 0.1572265625,
+      "learning_rate": 0.02,
+      "loss": 1.6579,
+      "step": 3801
+    },
+    {
+      "epoch": 0.35197593010472716,
+      "grad_norm": 0.154296875,
+      "learning_rate": 0.02,
+      "loss": 1.5959,
+      "step": 3802
+    },
+    {
+      "epoch": 0.3520685066250072,
+      "grad_norm": 0.1591796875,
+      "learning_rate": 0.02,
+      "loss": 1.6087,
+      "step": 3803
+    },
+    {
+      "epoch": 0.35216108314528727,
+      "grad_norm": 0.1396484375,
+      "learning_rate": 0.02,
+      "loss": 1.6423,
+      "step": 3804
+    },
+    {
+      "epoch": 0.3522536596655673,
+      "grad_norm": 0.1513671875,
+      "learning_rate": 0.02,
+      "loss": 1.5693,
+      "step": 3805
+    },
+    {
+      "epoch": 0.3523462361858474,
+      "grad_norm": 0.142578125,
+      "learning_rate": 0.02,
+      "loss": 1.6441,
+      "step": 3806
+    },
+    {
+      "epoch": 0.3524388127061274,
+      "grad_norm": 0.1591796875,
+      "learning_rate": 0.02,
+      "loss": 1.6331,
+      "step": 3807
+    },
+    {
+      "epoch": 0.3525313892264075,
+      "grad_norm": 0.1494140625,
+      "learning_rate": 0.02,
+      "loss": 1.6329,
+      "step": 3808
+    },
+    {
+      "epoch": 0.3526239657466875,
+      "grad_norm": 0.1640625,
+      "learning_rate": 0.02,
+      "loss": 1.6701,
+      "step": 3809
+    },
+    {
+      "epoch": 0.3527165422669675,
+      "grad_norm": 0.1552734375,
+      "learning_rate": 0.02,
+      "loss": 1.6705,
+      "step": 3810
+    },
+    {
+      "epoch": 0.3528091187872476,
+      "grad_norm": 0.1533203125,
+      "learning_rate": 0.02,
+      "loss": 1.6354,
+      "step": 3811
+    },
+    {
+      "epoch": 0.35290169530752763,
+      "grad_norm": 0.146484375,
+      "learning_rate": 0.02,
+      "loss": 1.6349,
+      "step": 3812
+    },
+    {
+      "epoch": 0.3529942718278077,
+      "grad_norm": 0.1669921875,
+      "learning_rate": 0.02,
+      "loss": 1.6589,
+      "step": 3813
+    },
+    {
+      "epoch": 0.35308684834808773,
+      "grad_norm": 0.1513671875,
+      "learning_rate": 0.02,
+      "loss": 1.6048,
+      "step": 3814
+    },
+    {
+      "epoch": 0.3531794248683678,
+      "grad_norm": 0.1552734375,
+      "learning_rate": 0.02,
+      "loss": 1.634,
+      "step": 3815
+    },
+    {
+      "epoch": 0.3532720013886478,
+      "grad_norm": 0.1494140625,
+      "learning_rate": 0.02,
+      "loss": 1.6434,
+      "step": 3816
+    },
+    {
+      "epoch": 0.35336457790892783,
+      "grad_norm": 0.1455078125,
+      "learning_rate": 0.02,
+      "loss": 1.6586,
+      "step": 3817
+    },
+    {
+      "epoch": 0.3534571544292079,
+      "grad_norm": 0.1416015625,
+      "learning_rate": 0.02,
+      "loss": 1.6455,
+      "step": 3818
+    },
+    {
+      "epoch": 0.35354973094948794,
+      "grad_norm": 0.1484375,
+      "learning_rate": 0.02,
+      "loss": 1.6226,
+      "step": 3819
+    },
+    {
+      "epoch": 0.353642307469768,
+      "grad_norm": 0.1416015625,
+      "learning_rate": 0.02,
+      "loss": 1.6149,
+      "step": 3820
+    },
+    {
+      "epoch": 0.35373488399004804,
+      "grad_norm": 0.1484375,
+      "learning_rate": 0.02,
+      "loss": 1.6497,
+      "step": 3821
+    },
+    {
+      "epoch": 0.3538274605103281,
+      "grad_norm": 0.146484375,
+      "learning_rate": 0.02,
+      "loss": 1.6285,
+      "step": 3822
+    },
+    {
+      "epoch": 0.3539200370306081,
+      "grad_norm": 0.1494140625,
+      "learning_rate": 0.02,
+      "loss": 1.6433,
+      "step": 3823
+    },
+    {
+      "epoch": 0.35401261355088814,
+      "grad_norm": 0.142578125,
+      "learning_rate": 0.02,
+      "loss": 1.5357,
+      "step": 3824
+    },
+    {
+      "epoch": 0.3541051900711682,
+      "grad_norm": 0.1435546875,
+      "learning_rate": 0.02,
+      "loss": 1.6336,
+      "step": 3825
+    },
+    {
+      "epoch": 0.35419776659144825,
+      "grad_norm": 0.1552734375,
+      "learning_rate": 0.02,
+      "loss": 1.672,
+      "step": 3826
+    },
+    {
+      "epoch": 0.3542903431117283,
+      "grad_norm": 0.15234375,
+      "learning_rate": 0.02,
+      "loss": 1.6421,
+      "step": 3827
+    },
+    {
+      "epoch": 0.35438291963200835,
+      "grad_norm": 0.15234375,
+      "learning_rate": 0.02,
+      "loss": 1.5841,
+      "step": 3828
+    },
+    {
+      "epoch": 0.35447549615228835,
+      "grad_norm": 0.1416015625,
+      "learning_rate": 0.02,
+      "loss": 1.6408,
+      "step": 3829
+    },
+    {
+      "epoch": 0.3545680726725684,
+      "grad_norm": 0.1376953125,
+      "learning_rate": 0.02,
+      "loss": 1.5927,
+      "step": 3830
+    },
+    {
+      "epoch": 0.35466064919284845,
+      "grad_norm": 0.1435546875,
+      "learning_rate": 0.02,
+      "loss": 1.6408,
+      "step": 3831
+    },
+    {
+      "epoch": 0.3547532257131285,
+      "grad_norm": 0.1494140625,
+      "learning_rate": 0.02,
+      "loss": 1.6348,
+      "step": 3832
+    },
+    {
+      "epoch": 0.35484580223340856,
+      "grad_norm": 0.1455078125,
+      "learning_rate": 0.02,
+      "loss": 1.5923,
+      "step": 3833
+    },
+    {
+      "epoch": 0.3549383787536886,
+      "grad_norm": 0.1474609375,
+      "learning_rate": 0.02,
+      "loss": 1.5471,
+      "step": 3834
+    },
+    {
+      "epoch": 0.35503095527396866,
+      "grad_norm": 0.15234375,
+      "learning_rate": 0.02,
+      "loss": 1.5771,
+      "step": 3835
+    },
+    {
+      "epoch": 0.35512353179424866,
+      "grad_norm": 0.14453125,
+      "learning_rate": 0.02,
+      "loss": 1.5083,
+      "step": 3836
+    },
+    {
+      "epoch": 0.3552161083145287,
+      "grad_norm": 0.1484375,
+      "learning_rate": 0.02,
+      "loss": 1.6237,
+      "step": 3837
+    },
+    {
+      "epoch": 0.35530868483480876,
+      "grad_norm": 0.1533203125,
+      "learning_rate": 0.02,
+      "loss": 1.6219,
+      "step": 3838
+    },
+    {
+      "epoch": 0.3554012613550888,
+      "grad_norm": 0.1474609375,
+      "learning_rate": 0.02,
+      "loss": 1.666,
+      "step": 3839
+    },
+    {
+      "epoch": 0.35549383787536887,
+      "grad_norm": 0.1416015625,
+      "learning_rate": 0.02,
+      "loss": 1.6001,
+      "step": 3840
+    },
+    {
+      "epoch": 0.3555864143956489,
+      "grad_norm": 0.14453125,
+      "learning_rate": 0.02,
+      "loss": 1.6452,
+      "step": 3841
+    },
+    {
+      "epoch": 0.35567899091592897,
+      "grad_norm": 0.1376953125,
+      "learning_rate": 0.02,
+      "loss": 1.583,
+      "step": 3842
+    },
+    {
+      "epoch": 0.35577156743620897,
+      "grad_norm": 0.150390625,
+      "learning_rate": 0.02,
+      "loss": 1.6709,
+      "step": 3843
+    },
+    {
+      "epoch": 0.355864143956489,
+      "grad_norm": 0.1328125,
+      "learning_rate": 0.02,
+      "loss": 1.5817,
+      "step": 3844
+    },
+    {
+      "epoch": 0.3559567204767691,
+      "grad_norm": 0.1484375,
+      "learning_rate": 0.02,
+      "loss": 1.5625,
+      "step": 3845
+    },
+    {
+      "epoch": 0.3560492969970491,
+      "grad_norm": 0.142578125,
+      "learning_rate": 0.02,
+      "loss": 1.5845,
+      "step": 3846
+    },
+    {
+      "epoch": 0.3561418735173292,
+      "grad_norm": 0.14453125,
+      "learning_rate": 0.02,
+      "loss": 1.572,
+      "step": 3847
+    },
+    {
+      "epoch": 0.35623445003760923,
+      "grad_norm": 0.1474609375,
+      "learning_rate": 0.02,
+      "loss": 1.6522,
+      "step": 3848
+    },
+    {
+      "epoch": 0.3563270265578893,
+      "grad_norm": 0.1494140625,
+      "learning_rate": 0.02,
+      "loss": 1.5321,
+      "step": 3849
+    },
+    {
+      "epoch": 0.3564196030781693,
+      "grad_norm": 0.1484375,
+      "learning_rate": 0.02,
+      "loss": 1.7081,
+      "step": 3850
+    },
+    {
+      "epoch": 0.35651217959844933,
+      "grad_norm": 0.142578125,
+      "learning_rate": 0.02,
+      "loss": 1.6225,
+      "step": 3851
+    },
+    {
+      "epoch": 0.3566047561187294,
+      "grad_norm": 0.142578125,
+      "learning_rate": 0.02,
+      "loss": 1.67,
+      "step": 3852
+    },
+    {
+      "epoch": 0.35669733263900943,
+      "grad_norm": 0.138671875,
+      "learning_rate": 0.02,
+      "loss": 1.5372,
+      "step": 3853
+    },
+    {
+      "epoch": 0.3567899091592895,
+      "grad_norm": 0.1416015625,
+      "learning_rate": 0.02,
+      "loss": 1.592,
+      "step": 3854
+    },
+    {
+      "epoch": 0.35688248567956954,
+      "grad_norm": 0.1396484375,
+      "learning_rate": 0.02,
+      "loss": 1.594,
+      "step": 3855
+    },
+    {
+      "epoch": 0.35697506219984954,
+      "grad_norm": 0.142578125,
+      "learning_rate": 0.02,
+      "loss": 1.627,
+      "step": 3856
+    },
+    {
+      "epoch": 0.3570676387201296,
+      "grad_norm": 0.146484375,
+      "learning_rate": 0.02,
+      "loss": 1.6161,
+      "step": 3857
+    },
+    {
+      "epoch": 0.35716021524040964,
+      "grad_norm": 0.158203125,
+      "learning_rate": 0.02,
+      "loss": 1.6321,
+      "step": 3858
+    },
+    {
+      "epoch": 0.3572527917606897,
+      "grad_norm": 0.15234375,
+      "learning_rate": 0.02,
+      "loss": 1.6147,
+      "step": 3859
+    },
+    {
+      "epoch": 0.35734536828096974,
+      "grad_norm": 0.154296875,
+      "learning_rate": 0.02,
+      "loss": 1.653,
+      "step": 3860
+    },
+    {
+      "epoch": 0.3574379448012498,
+      "grad_norm": 0.1494140625,
+      "learning_rate": 0.02,
+      "loss": 1.5721,
+      "step": 3861
+    },
+    {
+      "epoch": 0.35753052132152985,
+      "grad_norm": 0.1513671875,
+      "learning_rate": 0.02,
+      "loss": 1.6282,
+      "step": 3862
+    },
+    {
+      "epoch": 0.35762309784180984,
+      "grad_norm": 0.1533203125,
+      "learning_rate": 0.02,
+      "loss": 1.5964,
+      "step": 3863
+    },
+    {
+      "epoch": 0.3577156743620899,
+      "grad_norm": 0.1435546875,
+      "learning_rate": 0.02,
+      "loss": 1.5671,
+      "step": 3864
+    },
+    {
+      "epoch": 0.35780825088236995,
+      "grad_norm": 0.14453125,
+      "learning_rate": 0.02,
+      "loss": 1.5996,
+      "step": 3865
+    },
+    {
+      "epoch": 0.35790082740265,
+      "grad_norm": 0.158203125,
+      "learning_rate": 0.02,
+      "loss": 1.644,
+      "step": 3866
+    },
+    {
+      "epoch": 0.35799340392293005,
+      "grad_norm": 0.146484375,
+      "learning_rate": 0.02,
+      "loss": 1.7005,
+      "step": 3867
+    },
+    {
+      "epoch": 0.3580859804432101,
+      "grad_norm": 0.1494140625,
+      "learning_rate": 0.02,
+      "loss": 1.6199,
+      "step": 3868
+    },
+    {
+      "epoch": 0.35817855696349016,
+      "grad_norm": 0.1357421875,
+      "learning_rate": 0.02,
+      "loss": 1.6139,
+      "step": 3869
+    },
+    {
+      "epoch": 0.35827113348377015,
+      "grad_norm": 0.1318359375,
+      "learning_rate": 0.02,
+      "loss": 1.5307,
+      "step": 3870
+    },
+    {
+      "epoch": 0.3583637100040502,
+      "grad_norm": 0.1396484375,
+      "learning_rate": 0.02,
+      "loss": 1.606,
+      "step": 3871
+    },
+    {
+      "epoch": 0.35845628652433026,
+      "grad_norm": 0.146484375,
+      "learning_rate": 0.02,
+      "loss": 1.648,
+      "step": 3872
+    },
+    {
+      "epoch": 0.3585488630446103,
+      "grad_norm": 0.138671875,
+      "learning_rate": 0.02,
+      "loss": 1.5962,
+      "step": 3873
+    },
+    {
+      "epoch": 0.35864143956489036,
+      "grad_norm": 0.142578125,
+      "learning_rate": 0.02,
+      "loss": 1.6171,
+      "step": 3874
+    },
+    {
+      "epoch": 0.3587340160851704,
+      "grad_norm": 0.140625,
+      "learning_rate": 0.02,
+      "loss": 1.6289,
+      "step": 3875
+    },
+    {
+      "epoch": 0.35882659260545047,
+      "grad_norm": 0.1533203125,
+      "learning_rate": 0.02,
+      "loss": 1.6223,
+      "step": 3876
+    },
+    {
+      "epoch": 0.35891916912573046,
+      "grad_norm": 0.150390625,
+      "learning_rate": 0.02,
+      "loss": 1.6237,
+      "step": 3877
+    },
+    {
+      "epoch": 0.3590117456460105,
+      "grad_norm": 0.1552734375,
+      "learning_rate": 0.02,
+      "loss": 1.6947,
+      "step": 3878
+    },
+    {
+      "epoch": 0.35910432216629057,
+      "grad_norm": 0.1533203125,
+      "learning_rate": 0.02,
+      "loss": 1.6143,
+      "step": 3879
+    },
+    {
+      "epoch": 0.3591968986865706,
+      "grad_norm": 0.158203125,
+      "learning_rate": 0.02,
+      "loss": 1.7056,
+      "step": 3880
+    },
+    {
+      "epoch": 0.3592894752068507,
+      "grad_norm": 0.154296875,
+      "learning_rate": 0.02,
+      "loss": 1.6103,
+      "step": 3881
+    },
+    {
+      "epoch": 0.3593820517271307,
+      "grad_norm": 0.1494140625,
+      "learning_rate": 0.02,
+      "loss": 1.6573,
+      "step": 3882
+    },
+    {
+      "epoch": 0.3594746282474108,
+      "grad_norm": 0.1611328125,
+      "learning_rate": 0.02,
+      "loss": 1.6183,
+      "step": 3883
+    },
+    {
+      "epoch": 0.3595672047676908,
+      "grad_norm": 0.1591796875,
+      "learning_rate": 0.02,
+      "loss": 1.655,
+      "step": 3884
+    },
+    {
+      "epoch": 0.3596597812879708,
+      "grad_norm": 0.15234375,
+      "learning_rate": 0.02,
+      "loss": 1.6329,
+      "step": 3885
+    },
+    {
+      "epoch": 0.3597523578082509,
+      "grad_norm": 0.1513671875,
+      "learning_rate": 0.02,
+      "loss": 1.6464,
+      "step": 3886
+    },
+    {
+      "epoch": 0.35984493432853093,
+      "grad_norm": 0.1474609375,
+      "learning_rate": 0.02,
+      "loss": 1.635,
+      "step": 3887
+    },
+    {
+      "epoch": 0.359937510848811,
+      "grad_norm": 0.1396484375,
+      "learning_rate": 0.02,
+      "loss": 1.5642,
+      "step": 3888
+    },
+    {
+      "epoch": 0.36003008736909103,
+      "grad_norm": 0.1455078125,
+      "learning_rate": 0.02,
+      "loss": 1.5528,
+      "step": 3889
+    },
+    {
+      "epoch": 0.36012266388937103,
+      "grad_norm": 0.16015625,
+      "learning_rate": 0.02,
+      "loss": 1.6013,
+      "step": 3890
+    },
+    {
+      "epoch": 0.3602152404096511,
+      "grad_norm": 0.154296875,
+      "learning_rate": 0.02,
+      "loss": 1.5972,
+      "step": 3891
+    },
+    {
+      "epoch": 0.36030781692993114,
+      "grad_norm": 0.1572265625,
+      "learning_rate": 0.02,
+      "loss": 1.6172,
+      "step": 3892
+    },
+    {
+      "epoch": 0.3604003934502112,
+      "grad_norm": 0.150390625,
+      "learning_rate": 0.02,
+      "loss": 1.6589,
+      "step": 3893
+    },
+    {
+      "epoch": 0.36049296997049124,
+      "grad_norm": 0.142578125,
+      "learning_rate": 0.02,
+      "loss": 1.5219,
+      "step": 3894
+    },
+    {
+      "epoch": 0.3605855464907713,
+      "grad_norm": 0.138671875,
+      "learning_rate": 0.02,
+      "loss": 1.6024,
+      "step": 3895
+    },
+    {
+      "epoch": 0.36067812301105134,
+      "grad_norm": 0.1435546875,
+      "learning_rate": 0.02,
+      "loss": 1.6213,
+      "step": 3896
+    },
+    {
+      "epoch": 0.36077069953133134,
+      "grad_norm": 0.142578125,
+      "learning_rate": 0.02,
+      "loss": 1.6123,
+      "step": 3897
+    },
+    {
+      "epoch": 0.3608632760516114,
+      "grad_norm": 0.1455078125,
+      "learning_rate": 0.02,
+      "loss": 1.6611,
+      "step": 3898
+    },
+    {
+      "epoch": 0.36095585257189144,
+      "grad_norm": 0.1474609375,
+      "learning_rate": 0.02,
+      "loss": 1.6032,
+      "step": 3899
+    },
+    {
+      "epoch": 0.3610484290921715,
+      "grad_norm": 0.140625,
+      "learning_rate": 0.02,
+      "loss": 1.6415,
+      "step": 3900
+    },
+    {
+      "epoch": 0.36114100561245155,
+      "grad_norm": 0.154296875,
+      "learning_rate": 0.02,
+      "loss": 1.627,
+      "step": 3901
+    },
+    {
+      "epoch": 0.3612335821327316,
+      "grad_norm": 0.1591796875,
+      "learning_rate": 0.02,
+      "loss": 1.6804,
+      "step": 3902
+    },
+    {
+      "epoch": 0.36132615865301165,
+      "grad_norm": 0.1572265625,
+      "learning_rate": 0.02,
+      "loss": 1.635,
+      "step": 3903
+    },
+    {
+      "epoch": 0.36141873517329165,
+      "grad_norm": 0.1484375,
+      "learning_rate": 0.02,
+      "loss": 1.6479,
+      "step": 3904
+    },
+    {
+      "epoch": 0.3615113116935717,
+      "grad_norm": 0.150390625,
+      "learning_rate": 0.02,
+      "loss": 1.6098,
+      "step": 3905
+    },
+    {
+      "epoch": 0.36160388821385175,
+      "grad_norm": 0.1494140625,
+      "learning_rate": 0.02,
+      "loss": 1.6453,
+      "step": 3906
+    },
+    {
+      "epoch": 0.3616964647341318,
+      "grad_norm": 0.1474609375,
+      "learning_rate": 0.02,
+      "loss": 1.6487,
+      "step": 3907
+    },
+    {
+      "epoch": 0.36178904125441186,
+      "grad_norm": 0.1494140625,
+      "learning_rate": 0.02,
+      "loss": 1.6017,
+      "step": 3908
+    },
+    {
+      "epoch": 0.3618816177746919,
+      "grad_norm": 0.140625,
+      "learning_rate": 0.02,
+      "loss": 1.5603,
+      "step": 3909
+    },
+    {
+      "epoch": 0.36197419429497196,
+      "grad_norm": 0.1650390625,
+      "learning_rate": 0.02,
+      "loss": 1.6123,
+      "step": 3910
+    },
+    {
+      "epoch": 0.36206677081525196,
+      "grad_norm": 0.142578125,
+      "learning_rate": 0.02,
+      "loss": 1.6488,
+      "step": 3911
+    },
+    {
+      "epoch": 0.362159347335532,
+      "grad_norm": 0.1513671875,
+      "learning_rate": 0.02,
+      "loss": 1.6025,
+      "step": 3912
+    },
+    {
+      "epoch": 0.36225192385581206,
+      "grad_norm": 0.138671875,
+      "learning_rate": 0.02,
+      "loss": 1.6486,
+      "step": 3913
+    },
+    {
+      "epoch": 0.3623445003760921,
+      "grad_norm": 0.1435546875,
+      "learning_rate": 0.02,
+      "loss": 1.5401,
+      "step": 3914
+    },
+    {
+      "epoch": 0.36243707689637217,
+      "grad_norm": 0.150390625,
+      "learning_rate": 0.02,
+      "loss": 1.5429,
+      "step": 3915
+    },
+    {
+      "epoch": 0.3625296534166522,
+      "grad_norm": 0.1552734375,
+      "learning_rate": 0.02,
+      "loss": 1.6584,
+      "step": 3916
+    },
+    {
+      "epoch": 0.36262222993693227,
+      "grad_norm": 0.142578125,
+      "learning_rate": 0.02,
+      "loss": 1.6181,
+      "step": 3917
+    },
+    {
+      "epoch": 0.36271480645721227,
+      "grad_norm": 0.154296875,
+      "learning_rate": 0.02,
+      "loss": 1.5855,
+      "step": 3918
+    },
+    {
+      "epoch": 0.3628073829774923,
+      "grad_norm": 0.146484375,
+      "learning_rate": 0.02,
+      "loss": 1.6158,
+      "step": 3919
+    },
+    {
+      "epoch": 0.3628999594977724,
+      "grad_norm": 0.1572265625,
+      "learning_rate": 0.02,
+      "loss": 1.6084,
+      "step": 3920
+    },
+    {
+      "epoch": 0.3629925360180524,
+      "grad_norm": 0.140625,
+      "learning_rate": 0.02,
+      "loss": 1.5693,
+      "step": 3921
+    },
+    {
+      "epoch": 0.3630851125383325,
+      "grad_norm": 0.14453125,
+      "learning_rate": 0.02,
+      "loss": 1.53,
+      "step": 3922
+    },
+    {
+      "epoch": 0.36317768905861253,
+      "grad_norm": 0.1357421875,
+      "learning_rate": 0.02,
+      "loss": 1.5726,
+      "step": 3923
+    },
+    {
+      "epoch": 0.3632702655788925,
+      "grad_norm": 0.142578125,
+      "learning_rate": 0.02,
+      "loss": 1.6285,
+      "step": 3924
+    },
+    {
+      "epoch": 0.3633628420991726,
+      "grad_norm": 0.150390625,
+      "learning_rate": 0.02,
+      "loss": 1.637,
+      "step": 3925
+    },
+    {
+      "epoch": 0.36345541861945263,
+      "grad_norm": 0.1484375,
+      "learning_rate": 0.02,
+      "loss": 1.5843,
+      "step": 3926
+    },
+    {
+      "epoch": 0.3635479951397327,
+      "grad_norm": 0.146484375,
+      "learning_rate": 0.02,
+      "loss": 1.5602,
+      "step": 3927
+    },
+    {
+      "epoch": 0.36364057166001273,
+      "grad_norm": 0.1474609375,
+      "learning_rate": 0.02,
+      "loss": 1.6251,
+      "step": 3928
+    },
+    {
+      "epoch": 0.3637331481802928,
+      "grad_norm": 0.162109375,
+      "learning_rate": 0.02,
+      "loss": 1.6077,
+      "step": 3929
+    },
+    {
+      "epoch": 0.36382572470057284,
+      "grad_norm": 0.150390625,
+      "learning_rate": 0.02,
+      "loss": 1.6236,
+      "step": 3930
+    },
+    {
+      "epoch": 0.36391830122085284,
+      "grad_norm": 0.16015625,
+      "learning_rate": 0.02,
+      "loss": 1.6879,
+      "step": 3931
+    },
+    {
+      "epoch": 0.3640108777411329,
+      "grad_norm": 0.15625,
+      "learning_rate": 0.02,
+      "loss": 1.5462,
+      "step": 3932
+    },
+    {
+      "epoch": 0.36410345426141294,
+      "grad_norm": 0.154296875,
+      "learning_rate": 0.02,
+      "loss": 1.5955,
+      "step": 3933
+    },
+    {
+      "epoch": 0.364196030781693,
+      "grad_norm": 0.14453125,
+      "learning_rate": 0.02,
+      "loss": 1.5524,
+      "step": 3934
+    },
+    {
+      "epoch": 0.36428860730197304,
+      "grad_norm": 0.150390625,
+      "learning_rate": 0.02,
+      "loss": 1.6648,
+      "step": 3935
+    },
+    {
+      "epoch": 0.3643811838222531,
+      "grad_norm": 0.1474609375,
+      "learning_rate": 0.02,
+      "loss": 1.6245,
+      "step": 3936
+    },
+    {
+      "epoch": 0.36447376034253315,
+      "grad_norm": 0.1494140625,
+      "learning_rate": 0.02,
+      "loss": 1.5954,
+      "step": 3937
+    },
+    {
+      "epoch": 0.36456633686281315,
+      "grad_norm": 0.1494140625,
+      "learning_rate": 0.02,
+      "loss": 1.641,
+      "step": 3938
+    },
+    {
+      "epoch": 0.3646589133830932,
+      "grad_norm": 0.1591796875,
+      "learning_rate": 0.02,
+      "loss": 1.5892,
+      "step": 3939
+    },
+    {
+      "epoch": 0.36475148990337325,
+      "grad_norm": 0.14453125,
+      "learning_rate": 0.02,
+      "loss": 1.609,
+      "step": 3940
+    },
+    {
+      "epoch": 0.3648440664236533,
+      "grad_norm": 0.138671875,
+      "learning_rate": 0.02,
+      "loss": 1.5894,
+      "step": 3941
+    },
+    {
+      "epoch": 0.36493664294393335,
+      "grad_norm": 0.1416015625,
+      "learning_rate": 0.02,
+      "loss": 1.5245,
+      "step": 3942
+    },
+    {
+      "epoch": 0.3650292194642134,
+      "grad_norm": 0.1474609375,
+      "learning_rate": 0.02,
+      "loss": 1.6495,
+      "step": 3943
+    },
+    {
+      "epoch": 0.36512179598449346,
+      "grad_norm": 0.1435546875,
+      "learning_rate": 0.02,
+      "loss": 1.6082,
+      "step": 3944
+    },
+    {
+      "epoch": 0.36521437250477345,
+      "grad_norm": 0.1455078125,
+      "learning_rate": 0.02,
+      "loss": 1.5968,
+      "step": 3945
+    },
+    {
+      "epoch": 0.3653069490250535,
+      "grad_norm": 0.14453125,
+      "learning_rate": 0.02,
+      "loss": 1.6733,
+      "step": 3946
+    },
+    {
+      "epoch": 0.36539952554533356,
+      "grad_norm": 0.1416015625,
+      "learning_rate": 0.02,
+      "loss": 1.6568,
+      "step": 3947
+    },
+    {
+      "epoch": 0.3654921020656136,
+      "grad_norm": 0.1416015625,
+      "learning_rate": 0.02,
+      "loss": 1.6908,
+      "step": 3948
+    },
+    {
+      "epoch": 0.36558467858589366,
+      "grad_norm": 0.1474609375,
+      "learning_rate": 0.02,
+      "loss": 1.6909,
+      "step": 3949
+    },
+    {
+      "epoch": 0.3656772551061737,
+      "grad_norm": 0.1337890625,
+      "learning_rate": 0.02,
+      "loss": 1.6146,
+      "step": 3950
+    },
+    {
+      "epoch": 0.36576983162645377,
+      "grad_norm": 0.138671875,
+      "learning_rate": 0.02,
+      "loss": 1.5759,
+      "step": 3951
+    },
+    {
+      "epoch": 0.36586240814673376,
+      "grad_norm": 0.1396484375,
+      "learning_rate": 0.02,
+      "loss": 1.6181,
+      "step": 3952
+    },
+    {
+      "epoch": 0.3659549846670138,
+      "grad_norm": 0.1533203125,
+      "learning_rate": 0.02,
+      "loss": 1.587,
+      "step": 3953
+    },
+    {
+      "epoch": 0.36604756118729387,
+      "grad_norm": 0.146484375,
+      "learning_rate": 0.02,
+      "loss": 1.6717,
+      "step": 3954
+    },
+    {
+      "epoch": 0.3661401377075739,
+      "grad_norm": 0.15625,
+      "learning_rate": 0.02,
+      "loss": 1.6436,
+      "step": 3955
+    },
+    {
+      "epoch": 0.366232714227854,
+      "grad_norm": 0.13671875,
+      "learning_rate": 0.02,
+      "loss": 1.6008,
+      "step": 3956
+    },
+    {
+      "epoch": 0.366325290748134,
+      "grad_norm": 0.1533203125,
+      "learning_rate": 0.02,
+      "loss": 1.6127,
+      "step": 3957
+    },
+    {
+      "epoch": 0.366417867268414,
+      "grad_norm": 0.150390625,
+      "learning_rate": 0.02,
+      "loss": 1.606,
+      "step": 3958
+    },
+    {
+      "epoch": 0.3665104437886941,
+      "grad_norm": 0.162109375,
+      "learning_rate": 0.02,
+      "loss": 1.6458,
+      "step": 3959
+    },
+    {
+      "epoch": 0.3666030203089741,
+      "grad_norm": 0.1484375,
+      "learning_rate": 0.02,
+      "loss": 1.5833,
+      "step": 3960
+    },
+    {
+      "epoch": 0.3666955968292542,
+      "grad_norm": 0.150390625,
+      "learning_rate": 0.02,
+      "loss": 1.616,
+      "step": 3961
+    },
+    {
+      "epoch": 0.36678817334953423,
+      "grad_norm": 0.1513671875,
+      "learning_rate": 0.02,
+      "loss": 1.562,
+      "step": 3962
+    },
+    {
+      "epoch": 0.3668807498698143,
+      "grad_norm": 0.1435546875,
+      "learning_rate": 0.02,
+      "loss": 1.5867,
+      "step": 3963
+    },
+    {
+      "epoch": 0.36697332639009433,
+      "grad_norm": 0.1416015625,
+      "learning_rate": 0.02,
+      "loss": 1.5319,
+      "step": 3964
+    },
+    {
+      "epoch": 0.36706590291037433,
+      "grad_norm": 0.1494140625,
+      "learning_rate": 0.02,
+      "loss": 1.5863,
+      "step": 3965
+    },
+    {
+      "epoch": 0.3671584794306544,
+      "grad_norm": 0.1513671875,
+      "learning_rate": 0.02,
+      "loss": 1.5301,
+      "step": 3966
+    },
+    {
+      "epoch": 0.36725105595093444,
+      "grad_norm": 0.1513671875,
+      "learning_rate": 0.02,
+      "loss": 1.5862,
+      "step": 3967
+    },
+    {
+      "epoch": 0.3673436324712145,
+      "grad_norm": 0.1591796875,
+      "learning_rate": 0.02,
+      "loss": 1.6369,
+      "step": 3968
+    },
+    {
+      "epoch": 0.36743620899149454,
+      "grad_norm": 0.1591796875,
+      "learning_rate": 0.02,
+      "loss": 1.6175,
+      "step": 3969
+    },
+    {
+      "epoch": 0.3675287855117746,
+      "grad_norm": 0.154296875,
+      "learning_rate": 0.02,
+      "loss": 1.6166,
+      "step": 3970
+    },
+    {
+      "epoch": 0.36762136203205464,
+      "grad_norm": 0.15234375,
+      "learning_rate": 0.02,
+      "loss": 1.5819,
+      "step": 3971
+    },
+    {
+      "epoch": 0.36771393855233464,
+      "grad_norm": 0.1669921875,
+      "learning_rate": 0.02,
+      "loss": 1.6271,
+      "step": 3972
+    },
+    {
+      "epoch": 0.3678065150726147,
+      "grad_norm": 0.15625,
+      "learning_rate": 0.02,
+      "loss": 1.6427,
+      "step": 3973
+    },
+    {
+      "epoch": 0.36789909159289474,
+      "grad_norm": 0.158203125,
+      "learning_rate": 0.02,
+      "loss": 1.6029,
+      "step": 3974
+    },
+    {
+      "epoch": 0.3679916681131748,
+      "grad_norm": 0.1416015625,
+      "learning_rate": 0.02,
+      "loss": 1.6067,
+      "step": 3975
+    },
+    {
+      "epoch": 0.36808424463345485,
+      "grad_norm": 0.1533203125,
+      "learning_rate": 0.02,
+      "loss": 1.6145,
+      "step": 3976
+    },
+    {
+      "epoch": 0.3681768211537349,
+      "grad_norm": 0.1416015625,
+      "learning_rate": 0.02,
+      "loss": 1.6037,
+      "step": 3977
+    },
+    {
+      "epoch": 0.36826939767401495,
+      "grad_norm": 0.146484375,
+      "learning_rate": 0.02,
+      "loss": 1.6308,
+      "step": 3978
+    },
+    {
+      "epoch": 0.36836197419429495,
+      "grad_norm": 0.1455078125,
+      "learning_rate": 0.02,
+      "loss": 1.5922,
+      "step": 3979
+    },
+    {
+      "epoch": 0.368454550714575,
+      "grad_norm": 0.142578125,
+      "learning_rate": 0.02,
+      "loss": 1.5804,
+      "step": 3980
+    },
+    {
+      "epoch": 0.36854712723485505,
+      "grad_norm": 0.1416015625,
+      "learning_rate": 0.02,
+      "loss": 1.5504,
+      "step": 3981
+    },
+    {
+      "epoch": 0.3686397037551351,
+      "grad_norm": 0.1484375,
+      "learning_rate": 0.02,
+      "loss": 1.5425,
+      "step": 3982
+    },
+    {
+      "epoch": 0.36873228027541516,
+      "grad_norm": 0.1474609375,
+      "learning_rate": 0.02,
+      "loss": 1.5728,
+      "step": 3983
+    },
+    {
+      "epoch": 0.3688248567956952,
+      "grad_norm": 0.140625,
+      "learning_rate": 0.02,
+      "loss": 1.5863,
+      "step": 3984
+    },
+    {
+      "epoch": 0.36891743331597526,
+      "grad_norm": 0.14453125,
+      "learning_rate": 0.02,
+      "loss": 1.5823,
+      "step": 3985
+    },
+    {
+      "epoch": 0.36901000983625526,
+      "grad_norm": 0.150390625,
+      "learning_rate": 0.02,
+      "loss": 1.6161,
+      "step": 3986
+    },
+    {
+      "epoch": 0.3691025863565353,
+      "grad_norm": 0.1435546875,
+      "learning_rate": 0.02,
+      "loss": 1.5945,
+      "step": 3987
+    },
+    {
+      "epoch": 0.36919516287681536,
+      "grad_norm": 0.138671875,
+      "learning_rate": 0.02,
+      "loss": 1.5958,
+      "step": 3988
+    },
+    {
+      "epoch": 0.3692877393970954,
+      "grad_norm": 0.1474609375,
+      "learning_rate": 0.02,
+      "loss": 1.5926,
+      "step": 3989
+    },
+    {
+      "epoch": 0.36938031591737547,
+      "grad_norm": 0.1396484375,
+      "learning_rate": 0.02,
+      "loss": 1.6403,
+      "step": 3990
+    },
+    {
+      "epoch": 0.3694728924376555,
+      "grad_norm": 0.15234375,
+      "learning_rate": 0.02,
+      "loss": 1.6067,
+      "step": 3991
+    },
+    {
+      "epoch": 0.3695654689579355,
+      "grad_norm": 0.134765625,
+      "learning_rate": 0.02,
+      "loss": 1.6001,
+      "step": 3992
+    },
+    {
+      "epoch": 0.36965804547821557,
+      "grad_norm": 0.1396484375,
+      "learning_rate": 0.02,
+      "loss": 1.5701,
+      "step": 3993
+    },
+    {
+      "epoch": 0.3697506219984956,
+      "grad_norm": 0.1396484375,
+      "learning_rate": 0.02,
+      "loss": 1.6214,
+      "step": 3994
+    },
+    {
+      "epoch": 0.3698431985187757,
+      "grad_norm": 0.1484375,
+      "learning_rate": 0.02,
+      "loss": 1.6074,
+      "step": 3995
+    },
+    {
+      "epoch": 0.3699357750390557,
+      "grad_norm": 0.146484375,
+      "learning_rate": 0.02,
+      "loss": 1.5829,
+      "step": 3996
+    },
+    {
+      "epoch": 0.3700283515593358,
+      "grad_norm": 0.154296875,
+      "learning_rate": 0.02,
+      "loss": 1.6616,
+      "step": 3997
+    },
+    {
+      "epoch": 0.37012092807961583,
+      "grad_norm": 0.1513671875,
+      "learning_rate": 0.02,
+      "loss": 1.6001,
+      "step": 3998
+    },
+    {
+      "epoch": 0.3702135045998958,
+      "grad_norm": 0.1474609375,
+      "learning_rate": 0.02,
+      "loss": 1.5932,
+      "step": 3999
+    },
+    {
+      "epoch": 0.3703060811201759,
+      "grad_norm": 0.1484375,
+      "learning_rate": 0.02,
+      "loss": 1.6599,
+      "step": 4000
+    },
+    {
+      "epoch": 0.37039865764045593,
+      "grad_norm": 0.1533203125,
+      "learning_rate": 0.02,
+      "loss": 1.6349,
+      "step": 4001
+    },
+    {
+      "epoch": 0.370491234160736,
+      "grad_norm": 0.15234375,
+      "learning_rate": 0.02,
+      "loss": 1.6165,
+      "step": 4002
+    },
+    {
+      "epoch": 0.37058381068101603,
+      "grad_norm": 0.142578125,
+      "learning_rate": 0.02,
+      "loss": 1.5804,
+      "step": 4003
+    },
+    {
+      "epoch": 0.3706763872012961,
+      "grad_norm": 0.1416015625,
+      "learning_rate": 0.02,
+      "loss": 1.5717,
+      "step": 4004
+    },
+    {
+      "epoch": 0.37076896372157614,
+      "grad_norm": 0.1455078125,
+      "learning_rate": 0.02,
+      "loss": 1.5796,
+      "step": 4005
+    },
+    {
+      "epoch": 0.37086154024185614,
+      "grad_norm": 0.1435546875,
+      "learning_rate": 0.02,
+      "loss": 1.6141,
+      "step": 4006
+    },
+    {
+      "epoch": 0.3709541167621362,
+      "grad_norm": 0.1513671875,
+      "learning_rate": 0.02,
+      "loss": 1.6844,
+      "step": 4007
+    },
+    {
+      "epoch": 0.37104669328241624,
+      "grad_norm": 0.1513671875,
+      "learning_rate": 0.02,
+      "loss": 1.5828,
+      "step": 4008
+    },
+    {
+      "epoch": 0.3711392698026963,
+      "grad_norm": 0.1611328125,
+      "learning_rate": 0.02,
+      "loss": 1.6121,
+      "step": 4009
+    },
+    {
+      "epoch": 0.37123184632297634,
+      "grad_norm": 0.154296875,
+      "learning_rate": 0.02,
+      "loss": 1.6915,
+      "step": 4010
+    },
+    {
+      "epoch": 0.3713244228432564,
+      "grad_norm": 0.140625,
+      "learning_rate": 0.02,
+      "loss": 1.6615,
+      "step": 4011
+    },
+    {
+      "epoch": 0.37141699936353645,
+      "grad_norm": 0.1337890625,
+      "learning_rate": 0.02,
+      "loss": 1.5642,
+      "step": 4012
+    },
+    {
+      "epoch": 0.37150957588381645,
+      "grad_norm": 0.1435546875,
+      "learning_rate": 0.02,
+      "loss": 1.5622,
+      "step": 4013
+    },
+    {
+      "epoch": 0.3716021524040965,
+      "grad_norm": 0.140625,
+      "learning_rate": 0.02,
+      "loss": 1.6465,
+      "step": 4014
+    },
+    {
+      "epoch": 0.37169472892437655,
+      "grad_norm": 0.1435546875,
+      "learning_rate": 0.02,
+      "loss": 1.6447,
+      "step": 4015
+    },
+    {
+      "epoch": 0.3717873054446566,
+      "grad_norm": 0.1357421875,
+      "learning_rate": 0.02,
+      "loss": 1.6123,
+      "step": 4016
+    },
+    {
+      "epoch": 0.37187988196493665,
+      "grad_norm": 0.138671875,
+      "learning_rate": 0.02,
+      "loss": 1.6093,
+      "step": 4017
+    },
+    {
+      "epoch": 0.3719724584852167,
+      "grad_norm": 0.1416015625,
+      "learning_rate": 0.02,
+      "loss": 1.6216,
+      "step": 4018
+    },
+    {
+      "epoch": 0.37206503500549676,
+      "grad_norm": 0.138671875,
+      "learning_rate": 0.02,
+      "loss": 1.6126,
+      "step": 4019
+    },
+    {
+      "epoch": 0.37215761152577675,
+      "grad_norm": 0.140625,
+      "learning_rate": 0.02,
+      "loss": 1.5888,
+      "step": 4020
+    },
+    {
+      "epoch": 0.3722501880460568,
+      "grad_norm": 0.140625,
+      "learning_rate": 0.02,
+      "loss": 1.6132,
+      "step": 4021
+    },
+    {
+      "epoch": 0.37234276456633686,
+      "grad_norm": 0.14453125,
+      "learning_rate": 0.02,
+      "loss": 1.5446,
+      "step": 4022
+    },
+    {
+      "epoch": 0.3724353410866169,
+      "grad_norm": 0.1396484375,
+      "learning_rate": 0.02,
+      "loss": 1.6131,
+      "step": 4023
+    },
+    {
+      "epoch": 0.37252791760689696,
+      "grad_norm": 0.142578125,
+      "learning_rate": 0.02,
+      "loss": 1.5855,
+      "step": 4024
+    },
+    {
+      "epoch": 0.372620494127177,
+      "grad_norm": 0.140625,
+      "learning_rate": 0.02,
+      "loss": 1.6256,
+      "step": 4025
+    },
+    {
+      "epoch": 0.372713070647457,
+      "grad_norm": 0.1435546875,
+      "learning_rate": 0.02,
+      "loss": 1.6451,
+      "step": 4026
+    },
+    {
+      "epoch": 0.37280564716773706,
+      "grad_norm": 0.142578125,
+      "learning_rate": 0.02,
+      "loss": 1.6151,
+      "step": 4027
+    },
+    {
+      "epoch": 0.3728982236880171,
+      "grad_norm": 0.146484375,
+      "learning_rate": 0.02,
+      "loss": 1.6087,
+      "step": 4028
+    },
+    {
+      "epoch": 0.37299080020829717,
+      "grad_norm": 0.1552734375,
+      "learning_rate": 0.02,
+      "loss": 1.6046,
+      "step": 4029
+    },
+    {
+      "epoch": 0.3730833767285772,
+      "grad_norm": 0.1455078125,
+      "learning_rate": 0.02,
+      "loss": 1.6161,
+      "step": 4030
+    },
+    {
+      "epoch": 0.3731759532488573,
+      "grad_norm": 0.1640625,
+      "learning_rate": 0.02,
+      "loss": 1.6405,
+      "step": 4031
+    },
+    {
+      "epoch": 0.3732685297691373,
+      "grad_norm": 0.154296875,
+      "learning_rate": 0.02,
+      "loss": 1.5586,
+      "step": 4032
+    },
+    {
+      "epoch": 0.3733611062894173,
+      "grad_norm": 0.16015625,
+      "learning_rate": 0.02,
+      "loss": 1.6651,
+      "step": 4033
+    },
+    {
+      "epoch": 0.3734536828096974,
+      "grad_norm": 0.1474609375,
+      "learning_rate": 0.02,
+      "loss": 1.5984,
+      "step": 4034
+    },
+    {
+      "epoch": 0.3735462593299774,
+      "grad_norm": 0.1572265625,
+      "learning_rate": 0.02,
+      "loss": 1.5842,
+      "step": 4035
+    },
+    {
+      "epoch": 0.3736388358502575,
+      "grad_norm": 0.1513671875,
+      "learning_rate": 0.02,
+      "loss": 1.6177,
+      "step": 4036
+    },
+    {
+      "epoch": 0.37373141237053753,
+      "grad_norm": 0.1455078125,
+      "learning_rate": 0.02,
+      "loss": 1.6245,
+      "step": 4037
+    },
+    {
+      "epoch": 0.3738239888908176,
+      "grad_norm": 0.1513671875,
+      "learning_rate": 0.02,
+      "loss": 1.6941,
+      "step": 4038
+    },
+    {
+      "epoch": 0.37391656541109763,
+      "grad_norm": 0.1484375,
+      "learning_rate": 0.02,
+      "loss": 1.6252,
+      "step": 4039
+    },
+    {
+      "epoch": 0.37400914193137763,
+      "grad_norm": 0.142578125,
+      "learning_rate": 0.02,
+      "loss": 1.508,
+      "step": 4040
+    },
+    {
+      "epoch": 0.3741017184516577,
+      "grad_norm": 0.142578125,
+      "learning_rate": 0.02,
+      "loss": 1.646,
+      "step": 4041
+    },
+    {
+      "epoch": 0.37419429497193774,
+      "grad_norm": 0.15234375,
+      "learning_rate": 0.02,
+      "loss": 1.6139,
+      "step": 4042
+    },
+    {
+      "epoch": 0.3742868714922178,
+      "grad_norm": 0.142578125,
+      "learning_rate": 0.02,
+      "loss": 1.6028,
+      "step": 4043
+    },
+    {
+      "epoch": 0.37437944801249784,
+      "grad_norm": 0.1533203125,
+      "learning_rate": 0.02,
+      "loss": 1.6157,
+      "step": 4044
+    },
+    {
+      "epoch": 0.3744720245327779,
+      "grad_norm": 0.146484375,
+      "learning_rate": 0.02,
+      "loss": 1.6583,
+      "step": 4045
+    },
+    {
+      "epoch": 0.37456460105305794,
+      "grad_norm": 0.150390625,
+      "learning_rate": 0.02,
+      "loss": 1.5949,
+      "step": 4046
+    },
+    {
+      "epoch": 0.37465717757333794,
+      "grad_norm": 0.1396484375,
+      "learning_rate": 0.02,
+      "loss": 1.6439,
+      "step": 4047
+    },
+    {
+      "epoch": 0.374749754093618,
+      "grad_norm": 0.15625,
+      "learning_rate": 0.02,
+      "loss": 1.6939,
+      "step": 4048
+    },
+    {
+      "epoch": 0.37484233061389804,
+      "grad_norm": 0.1572265625,
+      "learning_rate": 0.02,
+      "loss": 1.6058,
+      "step": 4049
+    },
+    {
+      "epoch": 0.3749349071341781,
+      "grad_norm": 0.1435546875,
+      "learning_rate": 0.02,
+      "loss": 1.5668,
+      "step": 4050
+    },
+    {
+      "epoch": 0.37502748365445815,
+      "grad_norm": 0.142578125,
+      "learning_rate": 0.02,
+      "loss": 1.6346,
+      "step": 4051
+    },
+    {
+      "epoch": 0.3751200601747382,
+      "grad_norm": 0.1533203125,
+      "learning_rate": 0.02,
+      "loss": 1.5838,
+      "step": 4052
+    },
+    {
+      "epoch": 0.37521263669501825,
+      "grad_norm": 0.1474609375,
+      "learning_rate": 0.02,
+      "loss": 1.5954,
+      "step": 4053
+    },
+    {
+      "epoch": 0.37530521321529825,
+      "grad_norm": 0.1416015625,
+      "learning_rate": 0.02,
+      "loss": 1.5734,
+      "step": 4054
+    },
+    {
+      "epoch": 0.3753977897355783,
+      "grad_norm": 0.1416015625,
+      "learning_rate": 0.02,
+      "loss": 1.5643,
+      "step": 4055
+    },
+    {
+      "epoch": 0.37549036625585835,
+      "grad_norm": 0.14453125,
+      "learning_rate": 0.02,
+      "loss": 1.6497,
+      "step": 4056
+    },
+    {
+      "epoch": 0.3755829427761384,
+      "grad_norm": 0.1484375,
+      "learning_rate": 0.02,
+      "loss": 1.599,
+      "step": 4057
+    },
+    {
+      "epoch": 0.37567551929641846,
+      "grad_norm": 0.14453125,
+      "learning_rate": 0.02,
+      "loss": 1.5963,
+      "step": 4058
+    },
+    {
+      "epoch": 0.3757680958166985,
+      "grad_norm": 0.1533203125,
+      "learning_rate": 0.02,
+      "loss": 1.5897,
+      "step": 4059
+    },
+    {
+      "epoch": 0.3758606723369785,
+      "grad_norm": 0.14453125,
+      "learning_rate": 0.02,
+      "loss": 1.5644,
+      "step": 4060
+    },
+    {
+      "epoch": 0.37595324885725856,
+      "grad_norm": 0.1484375,
+      "learning_rate": 0.02,
+      "loss": 1.6111,
+      "step": 4061
+    },
+    {
+      "epoch": 0.3760458253775386,
+      "grad_norm": 0.1416015625,
+      "learning_rate": 0.02,
+      "loss": 1.6527,
+      "step": 4062
+    },
+    {
+      "epoch": 0.37613840189781866,
+      "grad_norm": 0.1474609375,
+      "learning_rate": 0.02,
+      "loss": 1.633,
+      "step": 4063
+    },
+    {
+      "epoch": 0.3762309784180987,
+      "grad_norm": 0.1396484375,
+      "learning_rate": 0.02,
+      "loss": 1.6249,
+      "step": 4064
+    },
+    {
+      "epoch": 0.37632355493837877,
+      "grad_norm": 0.1494140625,
+      "learning_rate": 0.02,
+      "loss": 1.5536,
+      "step": 4065
+    },
+    {
+      "epoch": 0.3764161314586588,
+      "grad_norm": 0.13671875,
+      "learning_rate": 0.02,
+      "loss": 1.5869,
+      "step": 4066
+    },
+    {
+      "epoch": 0.3765087079789388,
+      "grad_norm": 0.1376953125,
+      "learning_rate": 0.02,
+      "loss": 1.5885,
+      "step": 4067
+    },
+    {
+      "epoch": 0.37660128449921887,
+      "grad_norm": 0.13671875,
+      "learning_rate": 0.02,
+      "loss": 1.5618,
+      "step": 4068
+    },
+    {
+      "epoch": 0.3766938610194989,
+      "grad_norm": 0.1396484375,
+      "learning_rate": 0.02,
+      "loss": 1.5974,
+      "step": 4069
+    },
+    {
+      "epoch": 0.376786437539779,
+      "grad_norm": 0.1435546875,
+      "learning_rate": 0.02,
+      "loss": 1.5592,
+      "step": 4070
+    },
+    {
+      "epoch": 0.376879014060059,
+      "grad_norm": 0.1513671875,
+      "learning_rate": 0.02,
+      "loss": 1.6602,
+      "step": 4071
+    },
+    {
+      "epoch": 0.3769715905803391,
+      "grad_norm": 0.146484375,
+      "learning_rate": 0.02,
+      "loss": 1.5945,
+      "step": 4072
+    },
+    {
+      "epoch": 0.37706416710061913,
+      "grad_norm": 0.1484375,
+      "learning_rate": 0.02,
+      "loss": 1.555,
+      "step": 4073
+    },
+    {
+      "epoch": 0.3771567436208991,
+      "grad_norm": 0.1484375,
+      "learning_rate": 0.02,
+      "loss": 1.6416,
+      "step": 4074
+    },
+    {
+      "epoch": 0.3772493201411792,
+      "grad_norm": 0.1484375,
+      "learning_rate": 0.02,
+      "loss": 1.591,
+      "step": 4075
+    },
+    {
+      "epoch": 0.37734189666145923,
+      "grad_norm": 0.150390625,
+      "learning_rate": 0.02,
+      "loss": 1.6398,
+      "step": 4076
+    },
+    {
+      "epoch": 0.3774344731817393,
+      "grad_norm": 0.1650390625,
+      "learning_rate": 0.02,
+      "loss": 1.6582,
+      "step": 4077
+    },
+    {
+      "epoch": 0.37752704970201933,
+      "grad_norm": 0.146484375,
+      "learning_rate": 0.02,
+      "loss": 1.5782,
+      "step": 4078
+    },
+    {
+      "epoch": 0.3776196262222994,
+      "grad_norm": 0.15625,
+      "learning_rate": 0.02,
+      "loss": 1.5382,
+      "step": 4079
+    },
+    {
+      "epoch": 0.37771220274257944,
+      "grad_norm": 0.1484375,
+      "learning_rate": 0.02,
+      "loss": 1.5922,
+      "step": 4080
+    },
+    {
+      "epoch": 0.37780477926285944,
+      "grad_norm": 0.1494140625,
+      "learning_rate": 0.02,
+      "loss": 1.5861,
+      "step": 4081
+    },
+    {
+      "epoch": 0.3778973557831395,
+      "grad_norm": 0.1669921875,
+      "learning_rate": 0.02,
+      "loss": 1.6978,
+      "step": 4082
+    },
+    {
+      "epoch": 0.37798993230341954,
+      "grad_norm": 0.1513671875,
+      "learning_rate": 0.02,
+      "loss": 1.5812,
+      "step": 4083
+    },
+    {
+      "epoch": 0.3780825088236996,
+      "grad_norm": 0.15234375,
+      "learning_rate": 0.02,
+      "loss": 1.6042,
+      "step": 4084
+    },
+    {
+      "epoch": 0.37817508534397964,
+      "grad_norm": 0.142578125,
+      "learning_rate": 0.02,
+      "loss": 1.6433,
+      "step": 4085
+    },
+    {
+      "epoch": 0.3782676618642597,
+      "grad_norm": 0.16015625,
+      "learning_rate": 0.02,
+      "loss": 1.6058,
+      "step": 4086
+    },
+    {
+      "epoch": 0.37836023838453975,
+      "grad_norm": 0.146484375,
+      "learning_rate": 0.02,
+      "loss": 1.5878,
+      "step": 4087
+    },
+    {
+      "epoch": 0.37845281490481975,
+      "grad_norm": 0.1455078125,
+      "learning_rate": 0.02,
+      "loss": 1.6515,
+      "step": 4088
+    },
+    {
+      "epoch": 0.3785453914250998,
+      "grad_norm": 0.1611328125,
+      "learning_rate": 0.02,
+      "loss": 1.609,
+      "step": 4089
+    },
+    {
+      "epoch": 0.37863796794537985,
+      "grad_norm": 0.162109375,
+      "learning_rate": 0.02,
+      "loss": 1.6488,
+      "step": 4090
+    },
+    {
+      "epoch": 0.3787305444656599,
+      "grad_norm": 0.1513671875,
+      "learning_rate": 0.02,
+      "loss": 1.561,
+      "step": 4091
+    },
+    {
+      "epoch": 0.37882312098593995,
+      "grad_norm": 0.1533203125,
+      "learning_rate": 0.02,
+      "loss": 1.5352,
+      "step": 4092
+    },
+    {
+      "epoch": 0.37891569750622,
+      "grad_norm": 0.1455078125,
+      "learning_rate": 0.02,
+      "loss": 1.5793,
+      "step": 4093
+    },
+    {
+      "epoch": 0.3790082740265,
+      "grad_norm": 0.146484375,
+      "learning_rate": 0.02,
+      "loss": 1.6268,
+      "step": 4094
+    },
+    {
+      "epoch": 0.37910085054678005,
+      "grad_norm": 0.146484375,
+      "learning_rate": 0.02,
+      "loss": 1.6011,
+      "step": 4095
+    },
+    {
+      "epoch": 0.3791934270670601,
+      "grad_norm": 0.142578125,
+      "learning_rate": 0.02,
+      "loss": 1.6456,
+      "step": 4096
+    },
+    {
+      "epoch": 0.37928600358734016,
+      "grad_norm": 0.146484375,
+      "learning_rate": 0.02,
+      "loss": 1.6555,
+      "step": 4097
+    },
+    {
+      "epoch": 0.3793785801076202,
+      "grad_norm": 0.140625,
+      "learning_rate": 0.02,
+      "loss": 1.6045,
+      "step": 4098
+    },
+    {
+      "epoch": 0.37947115662790026,
+      "grad_norm": 0.130859375,
+      "learning_rate": 0.02,
+      "loss": 1.5133,
+      "step": 4099
+    },
+    {
+      "epoch": 0.3795637331481803,
+      "grad_norm": 0.1318359375,
+      "learning_rate": 0.02,
+      "loss": 1.609,
+      "step": 4100
+    },
+    {
+      "epoch": 0.3796563096684603,
+      "grad_norm": 0.13671875,
+      "learning_rate": 0.02,
+      "loss": 1.5401,
+      "step": 4101
+    },
+    {
+      "epoch": 0.37974888618874036,
+      "grad_norm": 0.1376953125,
+      "learning_rate": 0.02,
+      "loss": 1.5756,
+      "step": 4102
+    },
+    {
+      "epoch": 0.3798414627090204,
+      "grad_norm": 0.1513671875,
+      "learning_rate": 0.02,
+      "loss": 1.543,
+      "step": 4103
+    },
+    {
+      "epoch": 0.37993403922930047,
+      "grad_norm": 0.1611328125,
+      "learning_rate": 0.02,
+      "loss": 1.6058,
+      "step": 4104
+    },
+    {
+      "epoch": 0.3800266157495805,
+      "grad_norm": 0.166015625,
+      "learning_rate": 0.02,
+      "loss": 1.6035,
+      "step": 4105
+    },
+    {
+      "epoch": 0.3801191922698606,
+      "grad_norm": 0.154296875,
+      "learning_rate": 0.02,
+      "loss": 1.5653,
+      "step": 4106
+    },
+    {
+      "epoch": 0.3802117687901406,
+      "grad_norm": 0.1552734375,
+      "learning_rate": 0.02,
+      "loss": 1.5872,
+      "step": 4107
+    },
+    {
+      "epoch": 0.3803043453104206,
+      "grad_norm": 0.1669921875,
+      "learning_rate": 0.02,
+      "loss": 1.5941,
+      "step": 4108
+    },
+    {
+      "epoch": 0.3803969218307007,
+      "grad_norm": 0.162109375,
+      "learning_rate": 0.02,
+      "loss": 1.6186,
+      "step": 4109
+    },
+    {
+      "epoch": 0.3804894983509807,
+      "grad_norm": 0.150390625,
+      "learning_rate": 0.02,
+      "loss": 1.5742,
+      "step": 4110
+    },
+    {
+      "epoch": 0.3805820748712608,
+      "grad_norm": 0.142578125,
+      "learning_rate": 0.02,
+      "loss": 1.5936,
+      "step": 4111
+    },
+    {
+      "epoch": 0.38067465139154083,
+      "grad_norm": 0.1484375,
+      "learning_rate": 0.02,
+      "loss": 1.56,
+      "step": 4112
+    },
+    {
+      "epoch": 0.3807672279118209,
+      "grad_norm": 0.150390625,
+      "learning_rate": 0.02,
+      "loss": 1.6227,
+      "step": 4113
+    },
+    {
+      "epoch": 0.38085980443210093,
+      "grad_norm": 0.1533203125,
+      "learning_rate": 0.02,
+      "loss": 1.6092,
+      "step": 4114
+    },
+    {
+      "epoch": 0.38095238095238093,
+      "grad_norm": 0.140625,
+      "learning_rate": 0.02,
+      "loss": 1.6056,
+      "step": 4115
+    },
+    {
+      "epoch": 0.381044957472661,
+      "grad_norm": 0.1474609375,
+      "learning_rate": 0.02,
+      "loss": 1.6413,
+      "step": 4116
+    },
+    {
+      "epoch": 0.38113753399294104,
+      "grad_norm": 0.138671875,
+      "learning_rate": 0.02,
+      "loss": 1.5748,
+      "step": 4117
+    },
+    {
+      "epoch": 0.3812301105132211,
+      "grad_norm": 0.1416015625,
+      "learning_rate": 0.02,
+      "loss": 1.5694,
+      "step": 4118
+    },
+    {
+      "epoch": 0.38132268703350114,
+      "grad_norm": 0.150390625,
+      "learning_rate": 0.02,
+      "loss": 1.6314,
+      "step": 4119
+    },
+    {
+      "epoch": 0.3814152635537812,
+      "grad_norm": 0.142578125,
+      "learning_rate": 0.02,
+      "loss": 1.5968,
+      "step": 4120
+    },
+    {
+      "epoch": 0.3815078400740612,
+      "grad_norm": 0.142578125,
+      "learning_rate": 0.02,
+      "loss": 1.647,
+      "step": 4121
+    },
+    {
+      "epoch": 0.38160041659434124,
+      "grad_norm": 0.1435546875,
+      "learning_rate": 0.02,
+      "loss": 1.5674,
+      "step": 4122
+    },
+    {
+      "epoch": 0.3816929931146213,
+      "grad_norm": 0.1552734375,
+      "learning_rate": 0.02,
+      "loss": 1.646,
+      "step": 4123
+    },
+    {
+      "epoch": 0.38178556963490135,
+      "grad_norm": 0.1552734375,
+      "learning_rate": 0.02,
+      "loss": 1.6189,
+      "step": 4124
+    },
+    {
+      "epoch": 0.3818781461551814,
+      "grad_norm": 0.1650390625,
+      "learning_rate": 0.02,
+      "loss": 1.5681,
+      "step": 4125
+    },
+    {
+      "epoch": 0.38197072267546145,
+      "grad_norm": 0.1533203125,
+      "learning_rate": 0.02,
+      "loss": 1.5854,
+      "step": 4126
+    },
+    {
+      "epoch": 0.3820632991957415,
+      "grad_norm": 0.162109375,
+      "learning_rate": 0.02,
+      "loss": 1.634,
+      "step": 4127
+    },
+    {
+      "epoch": 0.3821558757160215,
+      "grad_norm": 0.150390625,
+      "learning_rate": 0.02,
+      "loss": 1.712,
+      "step": 4128
+    },
+    {
+      "epoch": 0.38224845223630155,
+      "grad_norm": 0.1572265625,
+      "learning_rate": 0.02,
+      "loss": 1.6649,
+      "step": 4129
+    },
+    {
+      "epoch": 0.3823410287565816,
+      "grad_norm": 0.14453125,
+      "learning_rate": 0.02,
+      "loss": 1.6005,
+      "step": 4130
+    },
+    {
+      "epoch": 0.38243360527686165,
+      "grad_norm": 0.15234375,
+      "learning_rate": 0.02,
+      "loss": 1.5706,
+      "step": 4131
+    },
+    {
+      "epoch": 0.3825261817971417,
+      "grad_norm": 0.134765625,
+      "learning_rate": 0.02,
+      "loss": 1.5867,
+      "step": 4132
+    },
+    {
+      "epoch": 0.38261875831742176,
+      "grad_norm": 0.1484375,
+      "learning_rate": 0.02,
+      "loss": 1.5997,
+      "step": 4133
+    },
+    {
+      "epoch": 0.3827113348377018,
+      "grad_norm": 0.134765625,
+      "learning_rate": 0.02,
+      "loss": 1.5512,
+      "step": 4134
+    },
+    {
+      "epoch": 0.3828039113579818,
+      "grad_norm": 0.1435546875,
+      "learning_rate": 0.02,
+      "loss": 1.611,
+      "step": 4135
+    },
+    {
+      "epoch": 0.38289648787826186,
+      "grad_norm": 0.150390625,
+      "learning_rate": 0.02,
+      "loss": 1.5931,
+      "step": 4136
+    },
+    {
+      "epoch": 0.3829890643985419,
+      "grad_norm": 0.154296875,
+      "learning_rate": 0.02,
+      "loss": 1.561,
+      "step": 4137
+    },
+    {
+      "epoch": 0.38308164091882196,
+      "grad_norm": 0.1455078125,
+      "learning_rate": 0.02,
+      "loss": 1.5728,
+      "step": 4138
+    },
+    {
+      "epoch": 0.383174217439102,
+      "grad_norm": 0.142578125,
+      "learning_rate": 0.02,
+      "loss": 1.5945,
+      "step": 4139
+    },
+    {
+      "epoch": 0.38326679395938207,
+      "grad_norm": 0.15234375,
+      "learning_rate": 0.02,
+      "loss": 1.5759,
+      "step": 4140
+    },
+    {
+      "epoch": 0.3833593704796621,
+      "grad_norm": 0.1455078125,
+      "learning_rate": 0.02,
+      "loss": 1.6014,
+      "step": 4141
+    },
+    {
+      "epoch": 0.3834519469999421,
+      "grad_norm": 0.14453125,
+      "learning_rate": 0.02,
+      "loss": 1.5781,
+      "step": 4142
+    },
+    {
+      "epoch": 0.38354452352022217,
+      "grad_norm": 0.14453125,
+      "learning_rate": 0.02,
+      "loss": 1.6056,
+      "step": 4143
+    },
+    {
+      "epoch": 0.3836371000405022,
+      "grad_norm": 0.150390625,
+      "learning_rate": 0.02,
+      "loss": 1.611,
+      "step": 4144
+    },
+    {
+      "epoch": 0.3837296765607823,
+      "grad_norm": 0.134765625,
+      "learning_rate": 0.02,
+      "loss": 1.514,
+      "step": 4145
+    },
+    {
+      "epoch": 0.3838222530810623,
+      "grad_norm": 0.1484375,
+      "learning_rate": 0.02,
+      "loss": 1.5825,
+      "step": 4146
+    },
+    {
+      "epoch": 0.3839148296013424,
+      "grad_norm": 0.138671875,
+      "learning_rate": 0.02,
+      "loss": 1.6236,
+      "step": 4147
+    },
+    {
+      "epoch": 0.38400740612162243,
+      "grad_norm": 0.140625,
+      "learning_rate": 0.02,
+      "loss": 1.5875,
+      "step": 4148
+    },
+    {
+      "epoch": 0.3840999826419024,
+      "grad_norm": 0.1474609375,
+      "learning_rate": 0.02,
+      "loss": 1.5974,
+      "step": 4149
+    },
+    {
+      "epoch": 0.3841925591621825,
+      "grad_norm": 0.146484375,
+      "learning_rate": 0.02,
+      "loss": 1.5513,
+      "step": 4150
+    },
+    {
+      "epoch": 0.38428513568246253,
+      "grad_norm": 0.1328125,
+      "learning_rate": 0.02,
+      "loss": 1.5998,
+      "step": 4151
+    },
+    {
+      "epoch": 0.3843777122027426,
+      "grad_norm": 0.1513671875,
+      "learning_rate": 0.02,
+      "loss": 1.592,
+      "step": 4152
+    },
+    {
+      "epoch": 0.38447028872302264,
+      "grad_norm": 0.1513671875,
+      "learning_rate": 0.02,
+      "loss": 1.6064,
+      "step": 4153
+    },
+    {
+      "epoch": 0.3845628652433027,
+      "grad_norm": 0.1533203125,
+      "learning_rate": 0.02,
+      "loss": 1.6101,
+      "step": 4154
+    },
+    {
+      "epoch": 0.3846554417635827,
+      "grad_norm": 0.1669921875,
+      "learning_rate": 0.02,
+      "loss": 1.6752,
+      "step": 4155
+    },
+    {
+      "epoch": 0.38474801828386274,
+      "grad_norm": 0.14453125,
+      "learning_rate": 0.02,
+      "loss": 1.5853,
+      "step": 4156
+    },
+    {
+      "epoch": 0.3848405948041428,
+      "grad_norm": 0.154296875,
+      "learning_rate": 0.02,
+      "loss": 1.6166,
+      "step": 4157
+    },
+    {
+      "epoch": 0.38493317132442284,
+      "grad_norm": 0.13671875,
+      "learning_rate": 0.02,
+      "loss": 1.6698,
+      "step": 4158
+    },
+    {
+      "epoch": 0.3850257478447029,
+      "grad_norm": 0.14453125,
+      "learning_rate": 0.02,
+      "loss": 1.6147,
+      "step": 4159
+    },
+    {
+      "epoch": 0.38511832436498294,
+      "grad_norm": 0.1416015625,
+      "learning_rate": 0.02,
+      "loss": 1.5945,
+      "step": 4160
+    },
+    {
+      "epoch": 0.385210900885263,
+      "grad_norm": 0.1416015625,
+      "learning_rate": 0.02,
+      "loss": 1.6361,
+      "step": 4161
+    },
+    {
+      "epoch": 0.385303477405543,
+      "grad_norm": 0.138671875,
+      "learning_rate": 0.02,
+      "loss": 1.5931,
+      "step": 4162
+    },
+    {
+      "epoch": 0.38539605392582305,
+      "grad_norm": 0.14453125,
+      "learning_rate": 0.02,
+      "loss": 1.5974,
+      "step": 4163
+    },
+    {
+      "epoch": 0.3854886304461031,
+      "grad_norm": 0.146484375,
+      "learning_rate": 0.02,
+      "loss": 1.6292,
+      "step": 4164
+    },
+    {
+      "epoch": 0.38558120696638315,
+      "grad_norm": 0.15625,
+      "learning_rate": 0.02,
+      "loss": 1.577,
+      "step": 4165
+    },
+    {
+      "epoch": 0.3856737834866632,
+      "grad_norm": 0.1396484375,
+      "learning_rate": 0.02,
+      "loss": 1.5449,
+      "step": 4166
+    },
+    {
+      "epoch": 0.38576636000694325,
+      "grad_norm": 0.1591796875,
+      "learning_rate": 0.02,
+      "loss": 1.5932,
+      "step": 4167
+    },
+    {
+      "epoch": 0.3858589365272233,
+      "grad_norm": 0.13671875,
+      "learning_rate": 0.02,
+      "loss": 1.5603,
+      "step": 4168
+    },
+    {
+      "epoch": 0.3859515130475033,
+      "grad_norm": 0.146484375,
+      "learning_rate": 0.02,
+      "loss": 1.5479,
+      "step": 4169
+    },
+    {
+      "epoch": 0.38604408956778336,
+      "grad_norm": 0.142578125,
+      "learning_rate": 0.02,
+      "loss": 1.6323,
+      "step": 4170
+    },
+    {
+      "epoch": 0.3861366660880634,
+      "grad_norm": 0.15625,
+      "learning_rate": 0.02,
+      "loss": 1.6388,
+      "step": 4171
+    },
+    {
+      "epoch": 0.38622924260834346,
+      "grad_norm": 0.1435546875,
+      "learning_rate": 0.02,
+      "loss": 1.5816,
+      "step": 4172
+    },
+    {
+      "epoch": 0.3863218191286235,
+      "grad_norm": 0.1494140625,
+      "learning_rate": 0.02,
+      "loss": 1.6271,
+      "step": 4173
+    },
+    {
+      "epoch": 0.38641439564890356,
+      "grad_norm": 0.1455078125,
+      "learning_rate": 0.02,
+      "loss": 1.5756,
+      "step": 4174
+    },
+    {
+      "epoch": 0.3865069721691836,
+      "grad_norm": 0.150390625,
+      "learning_rate": 0.02,
+      "loss": 1.5889,
+      "step": 4175
+    },
+    {
+      "epoch": 0.3865995486894636,
+      "grad_norm": 0.1435546875,
+      "learning_rate": 0.02,
+      "loss": 1.6069,
+      "step": 4176
+    },
+    {
+      "epoch": 0.38669212520974366,
+      "grad_norm": 0.1474609375,
+      "learning_rate": 0.02,
+      "loss": 1.6513,
+      "step": 4177
+    },
+    {
+      "epoch": 0.3867847017300237,
+      "grad_norm": 0.140625,
+      "learning_rate": 0.02,
+      "loss": 1.5952,
+      "step": 4178
+    },
+    {
+      "epoch": 0.38687727825030377,
+      "grad_norm": 0.14453125,
+      "learning_rate": 0.02,
+      "loss": 1.6068,
+      "step": 4179
+    },
+    {
+      "epoch": 0.3869698547705838,
+      "grad_norm": 0.1435546875,
+      "learning_rate": 0.02,
+      "loss": 1.6604,
+      "step": 4180
+    },
+    {
+      "epoch": 0.3870624312908639,
+      "grad_norm": 0.150390625,
+      "learning_rate": 0.02,
+      "loss": 1.5652,
+      "step": 4181
+    },
+    {
+      "epoch": 0.3871550078111439,
+      "grad_norm": 0.1572265625,
+      "learning_rate": 0.02,
+      "loss": 1.5416,
+      "step": 4182
+    },
+    {
+      "epoch": 0.3872475843314239,
+      "grad_norm": 0.1630859375,
+      "learning_rate": 0.02,
+      "loss": 1.6489,
+      "step": 4183
+    },
+    {
+      "epoch": 0.387340160851704,
+      "grad_norm": 0.1513671875,
+      "learning_rate": 0.02,
+      "loss": 1.6367,
+      "step": 4184
+    },
+    {
+      "epoch": 0.387432737371984,
+      "grad_norm": 0.1591796875,
+      "learning_rate": 0.02,
+      "loss": 1.6441,
+      "step": 4185
+    },
+    {
+      "epoch": 0.3875253138922641,
+      "grad_norm": 0.1455078125,
+      "learning_rate": 0.02,
+      "loss": 1.5737,
+      "step": 4186
+    },
+    {
+      "epoch": 0.38761789041254413,
+      "grad_norm": 0.1533203125,
+      "learning_rate": 0.02,
+      "loss": 1.6161,
+      "step": 4187
+    },
+    {
+      "epoch": 0.3877104669328242,
+      "grad_norm": 0.1337890625,
+      "learning_rate": 0.02,
+      "loss": 1.5753,
+      "step": 4188
+    },
+    {
+      "epoch": 0.3878030434531042,
+      "grad_norm": 0.1455078125,
+      "learning_rate": 0.02,
+      "loss": 1.6292,
+      "step": 4189
+    },
+    {
+      "epoch": 0.38789561997338423,
+      "grad_norm": 0.1435546875,
+      "learning_rate": 0.02,
+      "loss": 1.5591,
+      "step": 4190
+    },
+    {
+      "epoch": 0.3879881964936643,
+      "grad_norm": 0.1640625,
+      "learning_rate": 0.02,
+      "loss": 1.6344,
+      "step": 4191
+    },
+    {
+      "epoch": 0.38808077301394434,
+      "grad_norm": 0.1533203125,
+      "learning_rate": 0.02,
+      "loss": 1.6642,
+      "step": 4192
+    },
+    {
+      "epoch": 0.3881733495342244,
+      "grad_norm": 0.1572265625,
+      "learning_rate": 0.02,
+      "loss": 1.5711,
+      "step": 4193
+    },
+    {
+      "epoch": 0.38826592605450444,
+      "grad_norm": 0.134765625,
+      "learning_rate": 0.02,
+      "loss": 1.5871,
+      "step": 4194
+    },
+    {
+      "epoch": 0.3883585025747845,
+      "grad_norm": 0.1435546875,
+      "learning_rate": 0.02,
+      "loss": 1.5651,
+      "step": 4195
+    },
+    {
+      "epoch": 0.3884510790950645,
+      "grad_norm": 0.138671875,
+      "learning_rate": 0.02,
+      "loss": 1.5437,
+      "step": 4196
+    },
+    {
+      "epoch": 0.38854365561534454,
+      "grad_norm": 0.1396484375,
+      "learning_rate": 0.02,
+      "loss": 1.6441,
+      "step": 4197
+    },
+    {
+      "epoch": 0.3886362321356246,
+      "grad_norm": 0.1474609375,
+      "learning_rate": 0.02,
+      "loss": 1.6114,
+      "step": 4198
+    },
+    {
+      "epoch": 0.38872880865590465,
+      "grad_norm": 0.1513671875,
+      "learning_rate": 0.02,
+      "loss": 1.572,
+      "step": 4199
+    },
+    {
+      "epoch": 0.3888213851761847,
+      "grad_norm": 0.150390625,
+      "learning_rate": 0.02,
+      "loss": 1.5845,
+      "step": 4200
+    },
+    {
+      "epoch": 0.38891396169646475,
+      "grad_norm": 0.1484375,
+      "learning_rate": 0.02,
+      "loss": 1.6258,
+      "step": 4201
+    },
+    {
+      "epoch": 0.3890065382167448,
+      "grad_norm": 0.158203125,
+      "learning_rate": 0.02,
+      "loss": 1.583,
+      "step": 4202
+    },
+    {
+      "epoch": 0.3890991147370248,
+      "grad_norm": 0.1494140625,
+      "learning_rate": 0.02,
+      "loss": 1.6492,
+      "step": 4203
+    },
+    {
+      "epoch": 0.38919169125730485,
+      "grad_norm": 0.158203125,
+      "learning_rate": 0.02,
+      "loss": 1.558,
+      "step": 4204
+    },
+    {
+      "epoch": 0.3892842677775849,
+      "grad_norm": 0.1474609375,
+      "learning_rate": 0.02,
+      "loss": 1.7135,
+      "step": 4205
+    },
+    {
+      "epoch": 0.38937684429786495,
+      "grad_norm": 0.15234375,
+      "learning_rate": 0.02,
+      "loss": 1.5927,
+      "step": 4206
+    },
+    {
+      "epoch": 0.389469420818145,
+      "grad_norm": 0.1455078125,
+      "learning_rate": 0.02,
+      "loss": 1.577,
+      "step": 4207
+    },
+    {
+      "epoch": 0.38956199733842506,
+      "grad_norm": 0.134765625,
+      "learning_rate": 0.02,
+      "loss": 1.5624,
+      "step": 4208
+    },
+    {
+      "epoch": 0.3896545738587051,
+      "grad_norm": 0.1396484375,
+      "learning_rate": 0.02,
+      "loss": 1.6452,
+      "step": 4209
+    },
+    {
+      "epoch": 0.3897471503789851,
+      "grad_norm": 0.13671875,
+      "learning_rate": 0.02,
+      "loss": 1.6489,
+      "step": 4210
+    },
+    {
+      "epoch": 0.38983972689926516,
+      "grad_norm": 0.138671875,
+      "learning_rate": 0.02,
+      "loss": 1.5723,
+      "step": 4211
+    },
+    {
+      "epoch": 0.3899323034195452,
+      "grad_norm": 0.13671875,
+      "learning_rate": 0.02,
+      "loss": 1.609,
+      "step": 4212
+    },
+    {
+      "epoch": 0.39002487993982526,
+      "grad_norm": 0.142578125,
+      "learning_rate": 0.02,
+      "loss": 1.6052,
+      "step": 4213
+    },
+    {
+      "epoch": 0.3901174564601053,
+      "grad_norm": 0.140625,
+      "learning_rate": 0.02,
+      "loss": 1.5764,
+      "step": 4214
+    },
+    {
+      "epoch": 0.39021003298038537,
+      "grad_norm": 0.140625,
+      "learning_rate": 0.02,
+      "loss": 1.6119,
+      "step": 4215
+    },
+    {
+      "epoch": 0.3903026095006654,
+      "grad_norm": 0.1416015625,
+      "learning_rate": 0.02,
+      "loss": 1.5814,
+      "step": 4216
+    },
+    {
+      "epoch": 0.3903951860209454,
+      "grad_norm": 0.1552734375,
+      "learning_rate": 0.02,
+      "loss": 1.5487,
+      "step": 4217
+    },
+    {
+      "epoch": 0.39048776254122547,
+      "grad_norm": 0.1484375,
+      "learning_rate": 0.02,
+      "loss": 1.5588,
+      "step": 4218
+    },
+    {
+      "epoch": 0.3905803390615055,
+      "grad_norm": 0.154296875,
+      "learning_rate": 0.02,
+      "loss": 1.6287,
+      "step": 4219
+    },
+    {
+      "epoch": 0.3906729155817856,
+      "grad_norm": 0.140625,
+      "learning_rate": 0.02,
+      "loss": 1.5852,
+      "step": 4220
+    },
+    {
+      "epoch": 0.3907654921020656,
+      "grad_norm": 0.158203125,
+      "learning_rate": 0.02,
+      "loss": 1.6163,
+      "step": 4221
+    },
+    {
+      "epoch": 0.3908580686223457,
+      "grad_norm": 0.1513671875,
+      "learning_rate": 0.02,
+      "loss": 1.5547,
+      "step": 4222
+    },
+    {
+      "epoch": 0.3909506451426257,
+      "grad_norm": 0.154296875,
+      "learning_rate": 0.02,
+      "loss": 1.5735,
+      "step": 4223
+    },
+    {
+      "epoch": 0.3910432216629057,
+      "grad_norm": 0.1455078125,
+      "learning_rate": 0.02,
+      "loss": 1.5658,
+      "step": 4224
+    },
+    {
+      "epoch": 0.3911357981831858,
+      "grad_norm": 0.138671875,
+      "learning_rate": 0.02,
+      "loss": 1.5676,
+      "step": 4225
+    },
+    {
+      "epoch": 0.39122837470346583,
+      "grad_norm": 0.14453125,
+      "learning_rate": 0.02,
+      "loss": 1.6048,
+      "step": 4226
+    },
+    {
+      "epoch": 0.3913209512237459,
+      "grad_norm": 0.1455078125,
+      "learning_rate": 0.02,
+      "loss": 1.5977,
+      "step": 4227
+    },
+    {
+      "epoch": 0.39141352774402594,
+      "grad_norm": 0.1435546875,
+      "learning_rate": 0.02,
+      "loss": 1.6514,
+      "step": 4228
+    },
+    {
+      "epoch": 0.391506104264306,
+      "grad_norm": 0.140625,
+      "learning_rate": 0.02,
+      "loss": 1.5914,
+      "step": 4229
+    },
+    {
+      "epoch": 0.391598680784586,
+      "grad_norm": 0.1484375,
+      "learning_rate": 0.02,
+      "loss": 1.5828,
+      "step": 4230
+    },
+    {
+      "epoch": 0.39169125730486604,
+      "grad_norm": 0.1552734375,
+      "learning_rate": 0.02,
+      "loss": 1.6153,
+      "step": 4231
+    },
+    {
+      "epoch": 0.3917838338251461,
+      "grad_norm": 0.1484375,
+      "learning_rate": 0.02,
+      "loss": 1.5978,
+      "step": 4232
+    },
+    {
+      "epoch": 0.39187641034542614,
+      "grad_norm": 0.1455078125,
+      "learning_rate": 0.02,
+      "loss": 1.588,
+      "step": 4233
+    },
+    {
+      "epoch": 0.3919689868657062,
+      "grad_norm": 0.1435546875,
+      "learning_rate": 0.02,
+      "loss": 1.5525,
+      "step": 4234
+    },
+    {
+      "epoch": 0.39206156338598624,
+      "grad_norm": 0.1513671875,
+      "learning_rate": 0.02,
+      "loss": 1.6438,
+      "step": 4235
+    },
+    {
+      "epoch": 0.3921541399062663,
+      "grad_norm": 0.1572265625,
+      "learning_rate": 0.02,
+      "loss": 1.6077,
+      "step": 4236
+    },
+    {
+      "epoch": 0.3922467164265463,
+      "grad_norm": 0.15234375,
+      "learning_rate": 0.02,
+      "loss": 1.6195,
+      "step": 4237
+    },
+    {
+      "epoch": 0.39233929294682635,
+      "grad_norm": 0.1416015625,
+      "learning_rate": 0.02,
+      "loss": 1.6483,
+      "step": 4238
+    },
+    {
+      "epoch": 0.3924318694671064,
+      "grad_norm": 0.1513671875,
+      "learning_rate": 0.02,
+      "loss": 1.6536,
+      "step": 4239
+    },
+    {
+      "epoch": 0.39252444598738645,
+      "grad_norm": 0.140625,
+      "learning_rate": 0.02,
+      "loss": 1.6264,
+      "step": 4240
+    },
+    {
+      "epoch": 0.3926170225076665,
+      "grad_norm": 0.142578125,
+      "learning_rate": 0.02,
+      "loss": 1.6642,
+      "step": 4241
+    },
+    {
+      "epoch": 0.39270959902794655,
+      "grad_norm": 0.13671875,
+      "learning_rate": 0.02,
+      "loss": 1.5683,
+      "step": 4242
+    },
+    {
+      "epoch": 0.3928021755482266,
+      "grad_norm": 0.140625,
+      "learning_rate": 0.02,
+      "loss": 1.5813,
+      "step": 4243
+    },
+    {
+      "epoch": 0.3928947520685066,
+      "grad_norm": 0.1376953125,
+      "learning_rate": 0.02,
+      "loss": 1.5724,
+      "step": 4244
+    },
+    {
+      "epoch": 0.39298732858878666,
+      "grad_norm": 0.146484375,
+      "learning_rate": 0.02,
+      "loss": 1.594,
+      "step": 4245
+    },
+    {
+      "epoch": 0.3930799051090667,
+      "grad_norm": 0.146484375,
+      "learning_rate": 0.02,
+      "loss": 1.5493,
+      "step": 4246
+    },
+    {
+      "epoch": 0.39317248162934676,
+      "grad_norm": 0.1474609375,
+      "learning_rate": 0.02,
+      "loss": 1.67,
+      "step": 4247
+    },
+    {
+      "epoch": 0.3932650581496268,
+      "grad_norm": 0.1533203125,
+      "learning_rate": 0.02,
+      "loss": 1.6045,
+      "step": 4248
+    },
+    {
+      "epoch": 0.39335763466990686,
+      "grad_norm": 0.150390625,
+      "learning_rate": 0.02,
+      "loss": 1.6172,
+      "step": 4249
+    },
+    {
+      "epoch": 0.3934502111901869,
+      "grad_norm": 0.1455078125,
+      "learning_rate": 0.02,
+      "loss": 1.5652,
+      "step": 4250
+    },
+    {
+      "epoch": 0.3935427877104669,
+      "grad_norm": 0.142578125,
+      "learning_rate": 0.02,
+      "loss": 1.582,
+      "step": 4251
+    },
+    {
+      "epoch": 0.39363536423074696,
+      "grad_norm": 0.1533203125,
+      "learning_rate": 0.02,
+      "loss": 1.6105,
+      "step": 4252
+    },
+    {
+      "epoch": 0.393727940751027,
+      "grad_norm": 0.1435546875,
+      "learning_rate": 0.02,
+      "loss": 1.5323,
+      "step": 4253
+    },
+    {
+      "epoch": 0.39382051727130707,
+      "grad_norm": 0.1513671875,
+      "learning_rate": 0.02,
+      "loss": 1.6132,
+      "step": 4254
+    },
+    {
+      "epoch": 0.3939130937915871,
+      "grad_norm": 0.1494140625,
+      "learning_rate": 0.02,
+      "loss": 1.5509,
+      "step": 4255
+    },
+    {
+      "epoch": 0.3940056703118672,
+      "grad_norm": 0.154296875,
+      "learning_rate": 0.02,
+      "loss": 1.5851,
+      "step": 4256
+    },
+    {
+      "epoch": 0.39409824683214717,
+      "grad_norm": 0.150390625,
+      "learning_rate": 0.02,
+      "loss": 1.5867,
+      "step": 4257
+    },
+    {
+      "epoch": 0.3941908233524272,
+      "grad_norm": 0.1435546875,
+      "learning_rate": 0.02,
+      "loss": 1.6072,
+      "step": 4258
+    },
+    {
+      "epoch": 0.3942833998727073,
+      "grad_norm": 0.1494140625,
+      "learning_rate": 0.02,
+      "loss": 1.5707,
+      "step": 4259
+    },
+    {
+      "epoch": 0.3943759763929873,
+      "grad_norm": 0.1435546875,
+      "learning_rate": 0.02,
+      "loss": 1.6016,
+      "step": 4260
+    },
+    {
+      "epoch": 0.3944685529132674,
+      "grad_norm": 0.1416015625,
+      "learning_rate": 0.02,
+      "loss": 1.5656,
+      "step": 4261
+    },
+    {
+      "epoch": 0.39456112943354743,
+      "grad_norm": 0.138671875,
+      "learning_rate": 0.02,
+      "loss": 1.5968,
+      "step": 4262
+    },
+    {
+      "epoch": 0.3946537059538275,
+      "grad_norm": 0.1474609375,
+      "learning_rate": 0.02,
+      "loss": 1.6659,
+      "step": 4263
+    },
+    {
+      "epoch": 0.3947462824741075,
+      "grad_norm": 0.14453125,
+      "learning_rate": 0.02,
+      "loss": 1.6445,
+      "step": 4264
+    },
+    {
+      "epoch": 0.39483885899438753,
+      "grad_norm": 0.1455078125,
+      "learning_rate": 0.02,
+      "loss": 1.6074,
+      "step": 4265
+    },
+    {
+      "epoch": 0.3949314355146676,
+      "grad_norm": 0.1484375,
+      "learning_rate": 0.02,
+      "loss": 1.6087,
+      "step": 4266
+    },
+    {
+      "epoch": 0.39502401203494764,
+      "grad_norm": 0.1552734375,
+      "learning_rate": 0.02,
+      "loss": 1.5641,
+      "step": 4267
+    },
+    {
+      "epoch": 0.3951165885552277,
+      "grad_norm": 0.1455078125,
+      "learning_rate": 0.02,
+      "loss": 1.6064,
+      "step": 4268
+    },
+    {
+      "epoch": 0.39520916507550774,
+      "grad_norm": 0.1494140625,
+      "learning_rate": 0.02,
+      "loss": 1.6465,
+      "step": 4269
+    },
+    {
+      "epoch": 0.3953017415957878,
+      "grad_norm": 0.1455078125,
+      "learning_rate": 0.02,
+      "loss": 1.5649,
+      "step": 4270
+    },
+    {
+      "epoch": 0.3953943181160678,
+      "grad_norm": 0.1416015625,
+      "learning_rate": 0.02,
+      "loss": 1.5933,
+      "step": 4271
+    },
+    {
+      "epoch": 0.39548689463634784,
+      "grad_norm": 0.1513671875,
+      "learning_rate": 0.02,
+      "loss": 1.6468,
+      "step": 4272
+    },
+    {
+      "epoch": 0.3955794711566279,
+      "grad_norm": 0.1474609375,
+      "learning_rate": 0.02,
+      "loss": 1.6289,
+      "step": 4273
+    },
+    {
+      "epoch": 0.39567204767690795,
+      "grad_norm": 0.146484375,
+      "learning_rate": 0.02,
+      "loss": 1.6267,
+      "step": 4274
+    },
+    {
+      "epoch": 0.395764624197188,
+      "grad_norm": 0.1416015625,
+      "learning_rate": 0.02,
+      "loss": 1.5651,
+      "step": 4275
+    },
+    {
+      "epoch": 0.39585720071746805,
+      "grad_norm": 0.1474609375,
+      "learning_rate": 0.02,
+      "loss": 1.5928,
+      "step": 4276
+    },
+    {
+      "epoch": 0.3959497772377481,
+      "grad_norm": 0.15625,
+      "learning_rate": 0.02,
+      "loss": 1.533,
+      "step": 4277
+    },
+    {
+      "epoch": 0.3960423537580281,
+      "grad_norm": 0.1513671875,
+      "learning_rate": 0.02,
+      "loss": 1.5808,
+      "step": 4278
+    },
+    {
+      "epoch": 0.39613493027830815,
+      "grad_norm": 0.1474609375,
+      "learning_rate": 0.02,
+      "loss": 1.6344,
+      "step": 4279
+    },
+    {
+      "epoch": 0.3962275067985882,
+      "grad_norm": 0.1552734375,
+      "learning_rate": 0.02,
+      "loss": 1.619,
+      "step": 4280
+    },
+    {
+      "epoch": 0.39632008331886825,
+      "grad_norm": 0.150390625,
+      "learning_rate": 0.02,
+      "loss": 1.538,
+      "step": 4281
+    },
+    {
+      "epoch": 0.3964126598391483,
+      "grad_norm": 0.1513671875,
+      "learning_rate": 0.02,
+      "loss": 1.5859,
+      "step": 4282
+    },
+    {
+      "epoch": 0.39650523635942836,
+      "grad_norm": 0.1572265625,
+      "learning_rate": 0.02,
+      "loss": 1.6614,
+      "step": 4283
+    },
+    {
+      "epoch": 0.3965978128797084,
+      "grad_norm": 0.146484375,
+      "learning_rate": 0.02,
+      "loss": 1.5336,
+      "step": 4284
+    },
+    {
+      "epoch": 0.3966903893999884,
+      "grad_norm": 0.14453125,
+      "learning_rate": 0.02,
+      "loss": 1.6162,
+      "step": 4285
+    },
+    {
+      "epoch": 0.39678296592026846,
+      "grad_norm": 0.14453125,
+      "learning_rate": 0.02,
+      "loss": 1.576,
+      "step": 4286
+    },
+    {
+      "epoch": 0.3968755424405485,
+      "grad_norm": 0.1455078125,
+      "learning_rate": 0.02,
+      "loss": 1.6403,
+      "step": 4287
+    },
+    {
+      "epoch": 0.39696811896082856,
+      "grad_norm": 0.1484375,
+      "learning_rate": 0.02,
+      "loss": 1.6411,
+      "step": 4288
+    },
+    {
+      "epoch": 0.3970606954811086,
+      "grad_norm": 0.1484375,
+      "learning_rate": 0.02,
+      "loss": 1.5927,
+      "step": 4289
+    },
+    {
+      "epoch": 0.39715327200138867,
+      "grad_norm": 0.1455078125,
+      "learning_rate": 0.02,
+      "loss": 1.5979,
+      "step": 4290
+    },
+    {
+      "epoch": 0.39724584852166867,
+      "grad_norm": 0.142578125,
+      "learning_rate": 0.02,
+      "loss": 1.5287,
+      "step": 4291
+    },
+    {
+      "epoch": 0.3973384250419487,
+      "grad_norm": 0.1435546875,
+      "learning_rate": 0.02,
+      "loss": 1.5715,
+      "step": 4292
+    },
+    {
+      "epoch": 0.39743100156222877,
+      "grad_norm": 0.1484375,
+      "learning_rate": 0.02,
+      "loss": 1.6159,
+      "step": 4293
+    },
+    {
+      "epoch": 0.3975235780825088,
+      "grad_norm": 0.1455078125,
+      "learning_rate": 0.02,
+      "loss": 1.655,
+      "step": 4294
+    },
+    {
+      "epoch": 0.3976161546027889,
+      "grad_norm": 0.1572265625,
+      "learning_rate": 0.02,
+      "loss": 1.6396,
+      "step": 4295
+    },
+    {
+      "epoch": 0.3977087311230689,
+      "grad_norm": 0.138671875,
+      "learning_rate": 0.02,
+      "loss": 1.6249,
+      "step": 4296
+    },
+    {
+      "epoch": 0.397801307643349,
+      "grad_norm": 0.1376953125,
+      "learning_rate": 0.02,
+      "loss": 1.5371,
+      "step": 4297
+    },
+    {
+      "epoch": 0.397893884163629,
+      "grad_norm": 0.1533203125,
+      "learning_rate": 0.02,
+      "loss": 1.5571,
+      "step": 4298
+    },
+    {
+      "epoch": 0.397986460683909,
+      "grad_norm": 0.1533203125,
+      "learning_rate": 0.02,
+      "loss": 1.615,
+      "step": 4299
+    },
+    {
+      "epoch": 0.3980790372041891,
+      "grad_norm": 0.154296875,
+      "learning_rate": 0.02,
+      "loss": 1.6152,
+      "step": 4300
+    },
+    {
+      "epoch": 0.39817161372446913,
+      "grad_norm": 0.15234375,
+      "learning_rate": 0.02,
+      "loss": 1.5848,
+      "step": 4301
+    },
+    {
+      "epoch": 0.3982641902447492,
+      "grad_norm": 0.13671875,
+      "learning_rate": 0.02,
+      "loss": 1.5751,
+      "step": 4302
+    },
+    {
+      "epoch": 0.39835676676502924,
+      "grad_norm": 0.146484375,
+      "learning_rate": 0.02,
+      "loss": 1.6026,
+      "step": 4303
+    },
+    {
+      "epoch": 0.3984493432853093,
+      "grad_norm": 0.142578125,
+      "learning_rate": 0.02,
+      "loss": 1.6335,
+      "step": 4304
+    },
+    {
+      "epoch": 0.3985419198055893,
+      "grad_norm": 0.138671875,
+      "learning_rate": 0.02,
+      "loss": 1.5008,
+      "step": 4305
+    },
+    {
+      "epoch": 0.39863449632586934,
+      "grad_norm": 0.1416015625,
+      "learning_rate": 0.02,
+      "loss": 1.6145,
+      "step": 4306
+    },
+    {
+      "epoch": 0.3987270728461494,
+      "grad_norm": 0.1416015625,
+      "learning_rate": 0.02,
+      "loss": 1.5593,
+      "step": 4307
+    },
+    {
+      "epoch": 0.39881964936642944,
+      "grad_norm": 0.138671875,
+      "learning_rate": 0.02,
+      "loss": 1.5442,
+      "step": 4308
+    },
+    {
+      "epoch": 0.3989122258867095,
+      "grad_norm": 0.1416015625,
+      "learning_rate": 0.02,
+      "loss": 1.5972,
+      "step": 4309
+    },
+    {
+      "epoch": 0.39900480240698954,
+      "grad_norm": 0.1484375,
+      "learning_rate": 0.02,
+      "loss": 1.5826,
+      "step": 4310
+    },
+    {
+      "epoch": 0.3990973789272696,
+      "grad_norm": 0.150390625,
+      "learning_rate": 0.02,
+      "loss": 1.6189,
+      "step": 4311
+    },
+    {
+      "epoch": 0.3991899554475496,
+      "grad_norm": 0.1513671875,
+      "learning_rate": 0.02,
+      "loss": 1.5986,
+      "step": 4312
+    },
+    {
+      "epoch": 0.39928253196782965,
+      "grad_norm": 0.1513671875,
+      "learning_rate": 0.02,
+      "loss": 1.5762,
+      "step": 4313
+    },
+    {
+      "epoch": 0.3993751084881097,
+      "grad_norm": 0.16015625,
+      "learning_rate": 0.02,
+      "loss": 1.63,
+      "step": 4314
+    },
+    {
+      "epoch": 0.39946768500838975,
+      "grad_norm": 0.1416015625,
+      "learning_rate": 0.02,
+      "loss": 1.6487,
+      "step": 4315
+    },
+    {
+      "epoch": 0.3995602615286698,
+      "grad_norm": 0.1435546875,
+      "learning_rate": 0.02,
+      "loss": 1.578,
+      "step": 4316
+    },
+    {
+      "epoch": 0.39965283804894985,
+      "grad_norm": 0.1416015625,
+      "learning_rate": 0.02,
+      "loss": 1.6128,
+      "step": 4317
+    },
+    {
+      "epoch": 0.3997454145692299,
+      "grad_norm": 0.1484375,
+      "learning_rate": 0.02,
+      "loss": 1.5804,
+      "step": 4318
+    },
+    {
+      "epoch": 0.3998379910895099,
+      "grad_norm": 0.138671875,
+      "learning_rate": 0.02,
+      "loss": 1.5669,
+      "step": 4319
+    },
+    {
+      "epoch": 0.39993056760978996,
+      "grad_norm": 0.1435546875,
+      "learning_rate": 0.02,
+      "loss": 1.5968,
+      "step": 4320
+    },
+    {
+      "epoch": 0.40002314413007,
+      "grad_norm": 0.14453125,
+      "learning_rate": 0.02,
+      "loss": 1.6179,
+      "step": 4321
+    },
+    {
+      "epoch": 0.40011572065035006,
+      "grad_norm": 0.142578125,
+      "learning_rate": 0.02,
+      "loss": 1.5544,
+      "step": 4322
+    },
+    {
+      "epoch": 0.4002082971706301,
+      "grad_norm": 0.14453125,
+      "learning_rate": 0.02,
+      "loss": 1.5982,
+      "step": 4323
+    },
+    {
+      "epoch": 0.40030087369091016,
+      "grad_norm": 0.14453125,
+      "learning_rate": 0.02,
+      "loss": 1.6672,
+      "step": 4324
+    },
+    {
+      "epoch": 0.40039345021119016,
+      "grad_norm": 0.1474609375,
+      "learning_rate": 0.02,
+      "loss": 1.591,
+      "step": 4325
+    },
+    {
+      "epoch": 0.4004860267314702,
+      "grad_norm": 0.146484375,
+      "learning_rate": 0.02,
+      "loss": 1.6131,
+      "step": 4326
+    },
+    {
+      "epoch": 0.40057860325175026,
+      "grad_norm": 0.15234375,
+      "learning_rate": 0.02,
+      "loss": 1.5896,
+      "step": 4327
+    },
+    {
+      "epoch": 0.4006711797720303,
+      "grad_norm": 0.1513671875,
+      "learning_rate": 0.02,
+      "loss": 1.6251,
+      "step": 4328
+    },
+    {
+      "epoch": 0.40076375629231037,
+      "grad_norm": 0.15625,
+      "learning_rate": 0.02,
+      "loss": 1.6239,
+      "step": 4329
+    },
+    {
+      "epoch": 0.4008563328125904,
+      "grad_norm": 0.1435546875,
+      "learning_rate": 0.02,
+      "loss": 1.5361,
+      "step": 4330
+    },
+    {
+      "epoch": 0.4009489093328705,
+      "grad_norm": 0.142578125,
+      "learning_rate": 0.02,
+      "loss": 1.5812,
+      "step": 4331
+    },
+    {
+      "epoch": 0.40104148585315047,
+      "grad_norm": 0.14453125,
+      "learning_rate": 0.02,
+      "loss": 1.543,
+      "step": 4332
+    },
+    {
+      "epoch": 0.4011340623734305,
+      "grad_norm": 0.1533203125,
+      "learning_rate": 0.02,
+      "loss": 1.5317,
+      "step": 4333
+    },
+    {
+      "epoch": 0.4012266388937106,
+      "grad_norm": 0.1455078125,
+      "learning_rate": 0.02,
+      "loss": 1.5782,
+      "step": 4334
+    },
+    {
+      "epoch": 0.4013192154139906,
+      "grad_norm": 0.142578125,
+      "learning_rate": 0.02,
+      "loss": 1.509,
+      "step": 4335
+    },
+    {
+      "epoch": 0.4014117919342707,
+      "grad_norm": 0.1455078125,
+      "learning_rate": 0.02,
+      "loss": 1.6284,
+      "step": 4336
+    },
+    {
+      "epoch": 0.40150436845455073,
+      "grad_norm": 0.146484375,
+      "learning_rate": 0.02,
+      "loss": 1.5663,
+      "step": 4337
+    },
+    {
+      "epoch": 0.4015969449748308,
+      "grad_norm": 0.1455078125,
+      "learning_rate": 0.02,
+      "loss": 1.5685,
+      "step": 4338
+    },
+    {
+      "epoch": 0.4016895214951108,
+      "grad_norm": 0.1552734375,
+      "learning_rate": 0.02,
+      "loss": 1.6109,
+      "step": 4339
+    },
+    {
+      "epoch": 0.40178209801539083,
+      "grad_norm": 0.1552734375,
+      "learning_rate": 0.02,
+      "loss": 1.5703,
+      "step": 4340
+    },
+    {
+      "epoch": 0.4018746745356709,
+      "grad_norm": 0.1484375,
+      "learning_rate": 0.02,
+      "loss": 1.5928,
+      "step": 4341
+    },
+    {
+      "epoch": 0.40196725105595094,
+      "grad_norm": 0.15234375,
+      "learning_rate": 0.02,
+      "loss": 1.5843,
+      "step": 4342
+    },
+    {
+      "epoch": 0.402059827576231,
+      "grad_norm": 0.1376953125,
+      "learning_rate": 0.02,
+      "loss": 1.5141,
+      "step": 4343
+    },
+    {
+      "epoch": 0.40215240409651104,
+      "grad_norm": 0.146484375,
+      "learning_rate": 0.02,
+      "loss": 1.5929,
+      "step": 4344
+    },
+    {
+      "epoch": 0.4022449806167911,
+      "grad_norm": 0.1396484375,
+      "learning_rate": 0.02,
+      "loss": 1.586,
+      "step": 4345
+    },
+    {
+      "epoch": 0.4023375571370711,
+      "grad_norm": 0.1484375,
+      "learning_rate": 0.02,
+      "loss": 1.5676,
+      "step": 4346
+    },
+    {
+      "epoch": 0.40243013365735114,
+      "grad_norm": 0.1357421875,
+      "learning_rate": 0.02,
+      "loss": 1.5804,
+      "step": 4347
+    },
+    {
+      "epoch": 0.4025227101776312,
+      "grad_norm": 0.1494140625,
+      "learning_rate": 0.02,
+      "loss": 1.6231,
+      "step": 4348
+    },
+    {
+      "epoch": 0.40261528669791125,
+      "grad_norm": 0.134765625,
+      "learning_rate": 0.02,
+      "loss": 1.6123,
+      "step": 4349
+    },
+    {
+      "epoch": 0.4027078632181913,
+      "grad_norm": 0.146484375,
+      "learning_rate": 0.02,
+      "loss": 1.6297,
+      "step": 4350
+    },
+    {
+      "epoch": 0.40280043973847135,
+      "grad_norm": 0.1416015625,
+      "learning_rate": 0.02,
+      "loss": 1.5915,
+      "step": 4351
+    },
+    {
+      "epoch": 0.4028930162587514,
+      "grad_norm": 0.1376953125,
+      "learning_rate": 0.02,
+      "loss": 1.557,
+      "step": 4352
+    },
+    {
+      "epoch": 0.4029855927790314,
+      "grad_norm": 0.1416015625,
+      "learning_rate": 0.02,
+      "loss": 1.5881,
+      "step": 4353
+    },
+    {
+      "epoch": 0.40307816929931145,
+      "grad_norm": 0.142578125,
+      "learning_rate": 0.02,
+      "loss": 1.5967,
+      "step": 4354
+    },
+    {
+      "epoch": 0.4031707458195915,
+      "grad_norm": 0.1552734375,
+      "learning_rate": 0.02,
+      "loss": 1.6133,
+      "step": 4355
+    },
+    {
+      "epoch": 0.40326332233987155,
+      "grad_norm": 0.1416015625,
+      "learning_rate": 0.02,
+      "loss": 1.5454,
+      "step": 4356
+    },
+    {
+      "epoch": 0.4033558988601516,
+      "grad_norm": 0.15234375,
+      "learning_rate": 0.02,
+      "loss": 1.6063,
+      "step": 4357
+    },
+    {
+      "epoch": 0.40344847538043166,
+      "grad_norm": 0.15625,
+      "learning_rate": 0.02,
+      "loss": 1.6164,
+      "step": 4358
+    },
+    {
+      "epoch": 0.40354105190071166,
+      "grad_norm": 0.1337890625,
+      "learning_rate": 0.02,
+      "loss": 1.5722,
+      "step": 4359
+    },
+    {
+      "epoch": 0.4036336284209917,
+      "grad_norm": 0.150390625,
+      "learning_rate": 0.02,
+      "loss": 1.6426,
+      "step": 4360
+    },
+    {
+      "epoch": 0.40372620494127176,
+      "grad_norm": 0.1435546875,
+      "learning_rate": 0.02,
+      "loss": 1.6325,
+      "step": 4361
+    },
+    {
+      "epoch": 0.4038187814615518,
+      "grad_norm": 0.14453125,
+      "learning_rate": 0.02,
+      "loss": 1.6269,
+      "step": 4362
+    },
+    {
+      "epoch": 0.40391135798183186,
+      "grad_norm": 0.154296875,
+      "learning_rate": 0.02,
+      "loss": 1.6531,
+      "step": 4363
+    },
+    {
+      "epoch": 0.4040039345021119,
+      "grad_norm": 0.15625,
+      "learning_rate": 0.02,
+      "loss": 1.5789,
+      "step": 4364
+    },
+    {
+      "epoch": 0.40409651102239197,
+      "grad_norm": 0.1513671875,
+      "learning_rate": 0.02,
+      "loss": 1.5905,
+      "step": 4365
+    },
+    {
+      "epoch": 0.40418908754267197,
+      "grad_norm": 0.15625,
+      "learning_rate": 0.02,
+      "loss": 1.6093,
+      "step": 4366
+    },
+    {
+      "epoch": 0.404281664062952,
+      "grad_norm": 0.1474609375,
+      "learning_rate": 0.02,
+      "loss": 1.5237,
+      "step": 4367
+    },
+    {
+      "epoch": 0.40437424058323207,
+      "grad_norm": 0.1640625,
+      "learning_rate": 0.02,
+      "loss": 1.6448,
+      "step": 4368
+    },
+    {
+      "epoch": 0.4044668171035121,
+      "grad_norm": 0.15234375,
+      "learning_rate": 0.02,
+      "loss": 1.5699,
+      "step": 4369
+    },
+    {
+      "epoch": 0.4045593936237922,
+      "grad_norm": 0.150390625,
+      "learning_rate": 0.02,
+      "loss": 1.5926,
+      "step": 4370
+    },
+    {
+      "epoch": 0.4046519701440722,
+      "grad_norm": 0.1484375,
+      "learning_rate": 0.02,
+      "loss": 1.5787,
+      "step": 4371
+    },
+    {
+      "epoch": 0.4047445466643523,
+      "grad_norm": 0.1435546875,
+      "learning_rate": 0.02,
+      "loss": 1.611,
+      "step": 4372
+    },
+    {
+      "epoch": 0.4048371231846323,
+      "grad_norm": 0.1455078125,
+      "learning_rate": 0.02,
+      "loss": 1.5798,
+      "step": 4373
+    },
+    {
+      "epoch": 0.4049296997049123,
+      "grad_norm": 0.1455078125,
+      "learning_rate": 0.02,
+      "loss": 1.5685,
+      "step": 4374
+    },
+    {
+      "epoch": 0.4050222762251924,
+      "grad_norm": 0.16015625,
+      "learning_rate": 0.02,
+      "loss": 1.5967,
+      "step": 4375
+    },
+    {
+      "epoch": 0.40511485274547243,
+      "grad_norm": 0.146484375,
+      "learning_rate": 0.02,
+      "loss": 1.6312,
+      "step": 4376
+    },
+    {
+      "epoch": 0.4052074292657525,
+      "grad_norm": 0.1552734375,
+      "learning_rate": 0.02,
+      "loss": 1.5789,
+      "step": 4377
+    },
+    {
+      "epoch": 0.40530000578603254,
+      "grad_norm": 0.1474609375,
+      "learning_rate": 0.02,
+      "loss": 1.5515,
+      "step": 4378
+    },
+    {
+      "epoch": 0.4053925823063126,
+      "grad_norm": 0.1474609375,
+      "learning_rate": 0.02,
+      "loss": 1.5905,
+      "step": 4379
+    },
+    {
+      "epoch": 0.4054851588265926,
+      "grad_norm": 0.1455078125,
+      "learning_rate": 0.02,
+      "loss": 1.6472,
+      "step": 4380
+    },
+    {
+      "epoch": 0.40557773534687264,
+      "grad_norm": 0.146484375,
+      "learning_rate": 0.02,
+      "loss": 1.6021,
+      "step": 4381
+    },
+    {
+      "epoch": 0.4056703118671527,
+      "grad_norm": 0.1376953125,
+      "learning_rate": 0.02,
+      "loss": 1.6058,
+      "step": 4382
+    },
+    {
+      "epoch": 0.40576288838743274,
+      "grad_norm": 0.150390625,
+      "learning_rate": 0.02,
+      "loss": 1.6217,
+      "step": 4383
+    },
+    {
+      "epoch": 0.4058554649077128,
+      "grad_norm": 0.138671875,
+      "learning_rate": 0.02,
+      "loss": 1.5953,
+      "step": 4384
+    },
+    {
+      "epoch": 0.40594804142799285,
+      "grad_norm": 0.1455078125,
+      "learning_rate": 0.02,
+      "loss": 1.5417,
+      "step": 4385
+    },
+    {
+      "epoch": 0.40604061794827284,
+      "grad_norm": 0.138671875,
+      "learning_rate": 0.02,
+      "loss": 1.5838,
+      "step": 4386
+    },
+    {
+      "epoch": 0.4061331944685529,
+      "grad_norm": 0.1435546875,
+      "learning_rate": 0.02,
+      "loss": 1.5173,
+      "step": 4387
+    },
+    {
+      "epoch": 0.40622577098883295,
+      "grad_norm": 0.1455078125,
+      "learning_rate": 0.02,
+      "loss": 1.5733,
+      "step": 4388
+    },
+    {
+      "epoch": 0.406318347509113,
+      "grad_norm": 0.1513671875,
+      "learning_rate": 0.02,
+      "loss": 1.6026,
+      "step": 4389
+    },
+    {
+      "epoch": 0.40641092402939305,
+      "grad_norm": 0.1494140625,
+      "learning_rate": 0.02,
+      "loss": 1.5971,
+      "step": 4390
+    },
+    {
+      "epoch": 0.4065035005496731,
+      "grad_norm": 0.15625,
+      "learning_rate": 0.02,
+      "loss": 1.6225,
+      "step": 4391
+    },
+    {
+      "epoch": 0.40659607706995315,
+      "grad_norm": 0.140625,
+      "learning_rate": 0.02,
+      "loss": 1.5724,
+      "step": 4392
+    },
+    {
+      "epoch": 0.40668865359023315,
+      "grad_norm": 0.140625,
+      "learning_rate": 0.02,
+      "loss": 1.6342,
+      "step": 4393
+    },
+    {
+      "epoch": 0.4067812301105132,
+      "grad_norm": 0.1337890625,
+      "learning_rate": 0.02,
+      "loss": 1.591,
+      "step": 4394
+    },
+    {
+      "epoch": 0.40687380663079326,
+      "grad_norm": 0.1494140625,
+      "learning_rate": 0.02,
+      "loss": 1.6436,
+      "step": 4395
+    },
+    {
+      "epoch": 0.4069663831510733,
+      "grad_norm": 0.140625,
+      "learning_rate": 0.02,
+      "loss": 1.5401,
+      "step": 4396
+    },
+    {
+      "epoch": 0.40705895967135336,
+      "grad_norm": 0.1376953125,
+      "learning_rate": 0.02,
+      "loss": 1.5578,
+      "step": 4397
+    },
+    {
+      "epoch": 0.4071515361916334,
+      "grad_norm": 0.142578125,
+      "learning_rate": 0.02,
+      "loss": 1.6371,
+      "step": 4398
+    },
+    {
+      "epoch": 0.40724411271191346,
+      "grad_norm": 0.1337890625,
+      "learning_rate": 0.02,
+      "loss": 1.585,
+      "step": 4399
+    },
+    {
+      "epoch": 0.40733668923219346,
+      "grad_norm": 0.15234375,
+      "learning_rate": 0.02,
+      "loss": 1.5539,
+      "step": 4400
+    },
+    {
+      "epoch": 0.4074292657524735,
+      "grad_norm": 0.150390625,
+      "learning_rate": 0.02,
+      "loss": 1.5983,
+      "step": 4401
+    },
+    {
+      "epoch": 0.40752184227275356,
+      "grad_norm": 0.1591796875,
+      "learning_rate": 0.02,
+      "loss": 1.657,
+      "step": 4402
+    },
+    {
+      "epoch": 0.4076144187930336,
+      "grad_norm": 0.1484375,
+      "learning_rate": 0.02,
+      "loss": 1.5813,
+      "step": 4403
+    },
+    {
+      "epoch": 0.40770699531331367,
+      "grad_norm": 0.1484375,
+      "learning_rate": 0.02,
+      "loss": 1.5999,
+      "step": 4404
+    },
+    {
+      "epoch": 0.4077995718335937,
+      "grad_norm": 0.1494140625,
+      "learning_rate": 0.02,
+      "loss": 1.6276,
+      "step": 4405
+    },
+    {
+      "epoch": 0.4078921483538738,
+      "grad_norm": 0.150390625,
+      "learning_rate": 0.02,
+      "loss": 1.6393,
+      "step": 4406
+    },
+    {
+      "epoch": 0.40798472487415377,
+      "grad_norm": 0.15625,
+      "learning_rate": 0.02,
+      "loss": 1.5902,
+      "step": 4407
+    },
+    {
+      "epoch": 0.4080773013944338,
+      "grad_norm": 0.1572265625,
+      "learning_rate": 0.02,
+      "loss": 1.6684,
+      "step": 4408
+    },
+    {
+      "epoch": 0.4081698779147139,
+      "grad_norm": 0.14453125,
+      "learning_rate": 0.02,
+      "loss": 1.5642,
+      "step": 4409
+    },
+    {
+      "epoch": 0.4082624544349939,
+      "grad_norm": 0.1455078125,
+      "learning_rate": 0.02,
+      "loss": 1.6028,
+      "step": 4410
+    },
+    {
+      "epoch": 0.408355030955274,
+      "grad_norm": 0.1494140625,
+      "learning_rate": 0.02,
+      "loss": 1.6342,
+      "step": 4411
+    },
+    {
+      "epoch": 0.40844760747555403,
+      "grad_norm": 0.154296875,
+      "learning_rate": 0.02,
+      "loss": 1.5859,
+      "step": 4412
+    },
+    {
+      "epoch": 0.4085401839958341,
+      "grad_norm": 0.1533203125,
+      "learning_rate": 0.02,
+      "loss": 1.6234,
+      "step": 4413
+    },
+    {
+      "epoch": 0.4086327605161141,
+      "grad_norm": 0.1357421875,
+      "learning_rate": 0.02,
+      "loss": 1.6188,
+      "step": 4414
+    },
+    {
+      "epoch": 0.40872533703639413,
+      "grad_norm": 0.146484375,
+      "learning_rate": 0.02,
+      "loss": 1.6022,
+      "step": 4415
+    },
+    {
+      "epoch": 0.4088179135566742,
+      "grad_norm": 0.1484375,
+      "learning_rate": 0.02,
+      "loss": 1.5897,
+      "step": 4416
+    },
+    {
+      "epoch": 0.40891049007695424,
+      "grad_norm": 0.15625,
+      "learning_rate": 0.02,
+      "loss": 1.6064,
+      "step": 4417
+    },
+    {
+      "epoch": 0.4090030665972343,
+      "grad_norm": 0.1494140625,
+      "learning_rate": 0.02,
+      "loss": 1.5657,
+      "step": 4418
+    },
+    {
+      "epoch": 0.40909564311751434,
+      "grad_norm": 0.1513671875,
+      "learning_rate": 0.02,
+      "loss": 1.5803,
+      "step": 4419
+    },
+    {
+      "epoch": 0.40918821963779434,
+      "grad_norm": 0.130859375,
+      "learning_rate": 0.02,
+      "loss": 1.5977,
+      "step": 4420
+    },
+    {
+      "epoch": 0.4092807961580744,
+      "grad_norm": 0.1416015625,
+      "learning_rate": 0.02,
+      "loss": 1.6341,
+      "step": 4421
+    },
+    {
+      "epoch": 0.40937337267835444,
+      "grad_norm": 0.1337890625,
+      "learning_rate": 0.02,
+      "loss": 1.5287,
+      "step": 4422
+    },
+    {
+      "epoch": 0.4094659491986345,
+      "grad_norm": 0.1533203125,
+      "learning_rate": 0.02,
+      "loss": 1.6646,
+      "step": 4423
+    },
+    {
+      "epoch": 0.40955852571891455,
+      "grad_norm": 0.1474609375,
+      "learning_rate": 0.02,
+      "loss": 1.5776,
+      "step": 4424
+    },
+    {
+      "epoch": 0.4096511022391946,
+      "grad_norm": 0.1513671875,
+      "learning_rate": 0.02,
+      "loss": 1.609,
+      "step": 4425
+    },
+    {
+      "epoch": 0.40974367875947465,
+      "grad_norm": 0.15625,
+      "learning_rate": 0.02,
+      "loss": 1.5847,
+      "step": 4426
+    },
+    {
+      "epoch": 0.40983625527975465,
+      "grad_norm": 0.158203125,
+      "learning_rate": 0.02,
+      "loss": 1.6153,
+      "step": 4427
+    },
+    {
+      "epoch": 0.4099288318000347,
+      "grad_norm": 0.1484375,
+      "learning_rate": 0.02,
+      "loss": 1.6572,
+      "step": 4428
+    },
+    {
+      "epoch": 0.41002140832031475,
+      "grad_norm": 0.14453125,
+      "learning_rate": 0.02,
+      "loss": 1.5959,
+      "step": 4429
+    },
+    {
+      "epoch": 0.4101139848405948,
+      "grad_norm": 0.1494140625,
+      "learning_rate": 0.02,
+      "loss": 1.629,
+      "step": 4430
+    },
+    {
+      "epoch": 0.41020656136087486,
+      "grad_norm": 0.1396484375,
+      "learning_rate": 0.02,
+      "loss": 1.5934,
+      "step": 4431
+    },
+    {
+      "epoch": 0.4102991378811549,
+      "grad_norm": 0.1474609375,
+      "learning_rate": 0.02,
+      "loss": 1.6031,
+      "step": 4432
+    },
+    {
+      "epoch": 0.41039171440143496,
+      "grad_norm": 0.142578125,
+      "learning_rate": 0.02,
+      "loss": 1.6148,
+      "step": 4433
+    },
+    {
+      "epoch": 0.41048429092171496,
+      "grad_norm": 0.146484375,
+      "learning_rate": 0.02,
+      "loss": 1.5342,
+      "step": 4434
+    },
+    {
+      "epoch": 0.410576867441995,
+      "grad_norm": 0.142578125,
+      "learning_rate": 0.02,
+      "loss": 1.593,
+      "step": 4435
+    },
+    {
+      "epoch": 0.41066944396227506,
+      "grad_norm": 0.1396484375,
+      "learning_rate": 0.02,
+      "loss": 1.5406,
+      "step": 4436
+    },
+    {
+      "epoch": 0.4107620204825551,
+      "grad_norm": 0.14453125,
+      "learning_rate": 0.02,
+      "loss": 1.5595,
+      "step": 4437
+    },
+    {
+      "epoch": 0.41085459700283516,
+      "grad_norm": 0.1513671875,
+      "learning_rate": 0.02,
+      "loss": 1.5677,
+      "step": 4438
+    },
+    {
+      "epoch": 0.4109471735231152,
+      "grad_norm": 0.1474609375,
+      "learning_rate": 0.02,
+      "loss": 1.578,
+      "step": 4439
+    },
+    {
+      "epoch": 0.41103975004339527,
+      "grad_norm": 0.154296875,
+      "learning_rate": 0.02,
+      "loss": 1.5751,
+      "step": 4440
+    },
+    {
+      "epoch": 0.41113232656367527,
+      "grad_norm": 0.1494140625,
+      "learning_rate": 0.02,
+      "loss": 1.5438,
+      "step": 4441
+    },
+    {
+      "epoch": 0.4112249030839553,
+      "grad_norm": 0.15234375,
+      "learning_rate": 0.02,
+      "loss": 1.5558,
+      "step": 4442
+    },
+    {
+      "epoch": 0.41131747960423537,
+      "grad_norm": 0.154296875,
+      "learning_rate": 0.02,
+      "loss": 1.5849,
+      "step": 4443
+    },
+    {
+      "epoch": 0.4114100561245154,
+      "grad_norm": 0.146484375,
+      "learning_rate": 0.02,
+      "loss": 1.6392,
+      "step": 4444
+    },
+    {
+      "epoch": 0.4115026326447955,
+      "grad_norm": 0.1455078125,
+      "learning_rate": 0.02,
+      "loss": 1.5588,
+      "step": 4445
+    },
+    {
+      "epoch": 0.4115952091650755,
+      "grad_norm": 0.15234375,
+      "learning_rate": 0.02,
+      "loss": 1.6405,
+      "step": 4446
+    },
+    {
+      "epoch": 0.4116877856853556,
+      "grad_norm": 0.142578125,
+      "learning_rate": 0.02,
+      "loss": 1.6484,
+      "step": 4447
+    },
+    {
+      "epoch": 0.4117803622056356,
+      "grad_norm": 0.1494140625,
+      "learning_rate": 0.02,
+      "loss": 1.585,
+      "step": 4448
+    },
+    {
+      "epoch": 0.4118729387259156,
+      "grad_norm": 0.1416015625,
+      "learning_rate": 0.02,
+      "loss": 1.5502,
+      "step": 4449
+    },
+    {
+      "epoch": 0.4119655152461957,
+      "grad_norm": 0.158203125,
+      "learning_rate": 0.02,
+      "loss": 1.591,
+      "step": 4450
+    },
+    {
+      "epoch": 0.41205809176647573,
+      "grad_norm": 0.1494140625,
+      "learning_rate": 0.02,
+      "loss": 1.5943,
+      "step": 4451
+    },
+    {
+      "epoch": 0.4121506682867558,
+      "grad_norm": 0.1552734375,
+      "learning_rate": 0.02,
+      "loss": 1.6667,
+      "step": 4452
+    },
+    {
+      "epoch": 0.41224324480703584,
+      "grad_norm": 0.1533203125,
+      "learning_rate": 0.02,
+      "loss": 1.5859,
+      "step": 4453
+    },
+    {
+      "epoch": 0.41233582132731583,
+      "grad_norm": 0.15234375,
+      "learning_rate": 0.02,
+      "loss": 1.5311,
+      "step": 4454
+    },
+    {
+      "epoch": 0.4124283978475959,
+      "grad_norm": 0.1416015625,
+      "learning_rate": 0.02,
+      "loss": 1.5879,
+      "step": 4455
+    },
+    {
+      "epoch": 0.41252097436787594,
+      "grad_norm": 0.14453125,
+      "learning_rate": 0.02,
+      "loss": 1.6047,
+      "step": 4456
+    },
+    {
+      "epoch": 0.412613550888156,
+      "grad_norm": 0.142578125,
+      "learning_rate": 0.02,
+      "loss": 1.607,
+      "step": 4457
+    },
+    {
+      "epoch": 0.41270612740843604,
+      "grad_norm": 0.1533203125,
+      "learning_rate": 0.02,
+      "loss": 1.6122,
+      "step": 4458
+    },
+    {
+      "epoch": 0.4127987039287161,
+      "grad_norm": 0.154296875,
+      "learning_rate": 0.02,
+      "loss": 1.6209,
+      "step": 4459
+    },
+    {
+      "epoch": 0.41289128044899615,
+      "grad_norm": 0.1572265625,
+      "learning_rate": 0.02,
+      "loss": 1.6432,
+      "step": 4460
+    },
+    {
+      "epoch": 0.41298385696927614,
+      "grad_norm": 0.1376953125,
+      "learning_rate": 0.02,
+      "loss": 1.5889,
+      "step": 4461
+    },
+    {
+      "epoch": 0.4130764334895562,
+      "grad_norm": 0.1552734375,
+      "learning_rate": 0.02,
+      "loss": 1.5624,
+      "step": 4462
+    },
+    {
+      "epoch": 0.41316901000983625,
+      "grad_norm": 0.142578125,
+      "learning_rate": 0.02,
+      "loss": 1.5778,
+      "step": 4463
+    },
+    {
+      "epoch": 0.4132615865301163,
+      "grad_norm": 0.14453125,
+      "learning_rate": 0.02,
+      "loss": 1.6137,
+      "step": 4464
+    },
+    {
+      "epoch": 0.41335416305039635,
+      "grad_norm": 0.1396484375,
+      "learning_rate": 0.02,
+      "loss": 1.6035,
+      "step": 4465
+    },
+    {
+      "epoch": 0.4134467395706764,
+      "grad_norm": 0.140625,
+      "learning_rate": 0.02,
+      "loss": 1.5849,
+      "step": 4466
+    },
+    {
+      "epoch": 0.41353931609095645,
+      "grad_norm": 0.1357421875,
+      "learning_rate": 0.02,
+      "loss": 1.5956,
+      "step": 4467
+    },
+    {
+      "epoch": 0.41363189261123645,
+      "grad_norm": 0.1416015625,
+      "learning_rate": 0.02,
+      "loss": 1.6086,
+      "step": 4468
+    },
+    {
+      "epoch": 0.4137244691315165,
+      "grad_norm": 0.1357421875,
+      "learning_rate": 0.02,
+      "loss": 1.5476,
+      "step": 4469
+    },
+    {
+      "epoch": 0.41381704565179656,
+      "grad_norm": 0.1396484375,
+      "learning_rate": 0.02,
+      "loss": 1.6172,
+      "step": 4470
+    },
+    {
+      "epoch": 0.4139096221720766,
+      "grad_norm": 0.15234375,
+      "learning_rate": 0.02,
+      "loss": 1.6261,
+      "step": 4471
+    },
+    {
+      "epoch": 0.41400219869235666,
+      "grad_norm": 0.1455078125,
+      "learning_rate": 0.02,
+      "loss": 1.6049,
+      "step": 4472
+    },
+    {
+      "epoch": 0.4140947752126367,
+      "grad_norm": 0.14453125,
+      "learning_rate": 0.02,
+      "loss": 1.6064,
+      "step": 4473
+    },
+    {
+      "epoch": 0.41418735173291676,
+      "grad_norm": 0.1455078125,
+      "learning_rate": 0.02,
+      "loss": 1.6408,
+      "step": 4474
+    },
+    {
+      "epoch": 0.41427992825319676,
+      "grad_norm": 0.158203125,
+      "learning_rate": 0.02,
+      "loss": 1.5912,
+      "step": 4475
+    },
+    {
+      "epoch": 0.4143725047734768,
+      "grad_norm": 0.146484375,
+      "learning_rate": 0.02,
+      "loss": 1.5729,
+      "step": 4476
+    },
+    {
+      "epoch": 0.41446508129375687,
+      "grad_norm": 0.14453125,
+      "learning_rate": 0.02,
+      "loss": 1.6144,
+      "step": 4477
+    },
+    {
+      "epoch": 0.4145576578140369,
+      "grad_norm": 0.154296875,
+      "learning_rate": 0.02,
+      "loss": 1.5756,
+      "step": 4478
+    },
+    {
+      "epoch": 0.41465023433431697,
+      "grad_norm": 0.1435546875,
+      "learning_rate": 0.02,
+      "loss": 1.5629,
+      "step": 4479
+    },
+    {
+      "epoch": 0.414742810854597,
+      "grad_norm": 0.1484375,
+      "learning_rate": 0.02,
+      "loss": 1.6255,
+      "step": 4480
+    },
+    {
+      "epoch": 0.4148353873748771,
+      "grad_norm": 0.15625,
+      "learning_rate": 0.02,
+      "loss": 1.6016,
+      "step": 4481
+    },
+    {
+      "epoch": 0.41492796389515707,
+      "grad_norm": 0.1533203125,
+      "learning_rate": 0.02,
+      "loss": 1.6495,
+      "step": 4482
+    },
+    {
+      "epoch": 0.4150205404154371,
+      "grad_norm": 0.1484375,
+      "learning_rate": 0.02,
+      "loss": 1.5437,
+      "step": 4483
+    },
+    {
+      "epoch": 0.4151131169357172,
+      "grad_norm": 0.1357421875,
+      "learning_rate": 0.02,
+      "loss": 1.6318,
+      "step": 4484
+    },
+    {
+      "epoch": 0.4152056934559972,
+      "grad_norm": 0.1474609375,
+      "learning_rate": 0.02,
+      "loss": 1.6394,
+      "step": 4485
+    },
+    {
+      "epoch": 0.4152982699762773,
+      "grad_norm": 0.1396484375,
+      "learning_rate": 0.02,
+      "loss": 1.571,
+      "step": 4486
+    },
+    {
+      "epoch": 0.41539084649655733,
+      "grad_norm": 0.15625,
+      "learning_rate": 0.02,
+      "loss": 1.6369,
+      "step": 4487
+    },
+    {
+      "epoch": 0.4154834230168373,
+      "grad_norm": 0.1533203125,
+      "learning_rate": 0.02,
+      "loss": 1.5549,
+      "step": 4488
+    },
+    {
+      "epoch": 0.4155759995371174,
+      "grad_norm": 0.1533203125,
+      "learning_rate": 0.02,
+      "loss": 1.5776,
+      "step": 4489
+    },
+    {
+      "epoch": 0.41566857605739743,
+      "grad_norm": 0.1435546875,
+      "learning_rate": 0.02,
+      "loss": 1.5503,
+      "step": 4490
+    },
+    {
+      "epoch": 0.4157611525776775,
+      "grad_norm": 0.158203125,
+      "learning_rate": 0.02,
+      "loss": 1.6125,
+      "step": 4491
+    },
+    {
+      "epoch": 0.41585372909795754,
+      "grad_norm": 0.1533203125,
+      "learning_rate": 0.02,
+      "loss": 1.5797,
+      "step": 4492
+    },
+    {
+      "epoch": 0.4159463056182376,
+      "grad_norm": 0.1513671875,
+      "learning_rate": 0.02,
+      "loss": 1.5436,
+      "step": 4493
+    },
+    {
+      "epoch": 0.41603888213851764,
+      "grad_norm": 0.1533203125,
+      "learning_rate": 0.02,
+      "loss": 1.5828,
+      "step": 4494
+    },
+    {
+      "epoch": 0.41613145865879764,
+      "grad_norm": 0.15625,
+      "learning_rate": 0.02,
+      "loss": 1.6467,
+      "step": 4495
+    },
+    {
+      "epoch": 0.4162240351790777,
+      "grad_norm": 0.1494140625,
+      "learning_rate": 0.02,
+      "loss": 1.6121,
+      "step": 4496
+    },
+    {
+      "epoch": 0.41631661169935774,
+      "grad_norm": 0.150390625,
+      "learning_rate": 0.02,
+      "loss": 1.5894,
+      "step": 4497
+    },
+    {
+      "epoch": 0.4164091882196378,
+      "grad_norm": 0.150390625,
+      "learning_rate": 0.02,
+      "loss": 1.5417,
+      "step": 4498
+    },
+    {
+      "epoch": 0.41650176473991785,
+      "grad_norm": 0.1494140625,
+      "learning_rate": 0.02,
+      "loss": 1.6291,
+      "step": 4499
+    },
+    {
+      "epoch": 0.4165943412601979,
+      "grad_norm": 0.158203125,
+      "learning_rate": 0.02,
+      "loss": 1.5477,
+      "step": 4500
+    },
+    {
+      "epoch": 0.41668691778047795,
+      "grad_norm": 0.1513671875,
+      "learning_rate": 0.02,
+      "loss": 1.6016,
+      "step": 4501
+    },
+    {
+      "epoch": 0.41677949430075795,
+      "grad_norm": 0.15234375,
+      "learning_rate": 0.02,
+      "loss": 1.562,
+      "step": 4502
+    },
+    {
+      "epoch": 0.416872070821038,
+      "grad_norm": 0.1513671875,
+      "learning_rate": 0.02,
+      "loss": 1.6075,
+      "step": 4503
+    },
+    {
+      "epoch": 0.41696464734131805,
+      "grad_norm": 0.1630859375,
+      "learning_rate": 0.02,
+      "loss": 1.5617,
+      "step": 4504
+    },
+    {
+      "epoch": 0.4170572238615981,
+      "grad_norm": 0.1396484375,
+      "learning_rate": 0.02,
+      "loss": 1.5996,
+      "step": 4505
+    },
+    {
+      "epoch": 0.41714980038187816,
+      "grad_norm": 0.14453125,
+      "learning_rate": 0.02,
+      "loss": 1.5851,
+      "step": 4506
+    },
+    {
+      "epoch": 0.4172423769021582,
+      "grad_norm": 0.1435546875,
+      "learning_rate": 0.02,
+      "loss": 1.5739,
+      "step": 4507
+    },
+    {
+      "epoch": 0.41733495342243826,
+      "grad_norm": 0.1494140625,
+      "learning_rate": 0.02,
+      "loss": 1.5689,
+      "step": 4508
+    },
+    {
+      "epoch": 0.41742752994271826,
+      "grad_norm": 0.1455078125,
+      "learning_rate": 0.02,
+      "loss": 1.5432,
+      "step": 4509
+    },
+    {
+      "epoch": 0.4175201064629983,
+      "grad_norm": 0.15234375,
+      "learning_rate": 0.02,
+      "loss": 1.5707,
+      "step": 4510
+    },
+    {
+      "epoch": 0.41761268298327836,
+      "grad_norm": 0.1455078125,
+      "learning_rate": 0.02,
+      "loss": 1.6407,
+      "step": 4511
+    },
+    {
+      "epoch": 0.4177052595035584,
+      "grad_norm": 0.15234375,
+      "learning_rate": 0.02,
+      "loss": 1.6059,
+      "step": 4512
+    },
+    {
+      "epoch": 0.41779783602383846,
+      "grad_norm": 0.150390625,
+      "learning_rate": 0.02,
+      "loss": 1.5431,
+      "step": 4513
+    },
+    {
+      "epoch": 0.4178904125441185,
+      "grad_norm": 0.1416015625,
+      "learning_rate": 0.02,
+      "loss": 1.6293,
+      "step": 4514
+    },
+    {
+      "epoch": 0.41798298906439857,
+      "grad_norm": 0.1376953125,
+      "learning_rate": 0.02,
+      "loss": 1.5335,
+      "step": 4515
+    },
+    {
+      "epoch": 0.41807556558467857,
+      "grad_norm": 0.142578125,
+      "learning_rate": 0.02,
+      "loss": 1.5229,
+      "step": 4516
+    },
+    {
+      "epoch": 0.4181681421049586,
+      "grad_norm": 0.150390625,
+      "learning_rate": 0.02,
+      "loss": 1.6021,
+      "step": 4517
+    },
+    {
+      "epoch": 0.41826071862523867,
+      "grad_norm": 0.1455078125,
+      "learning_rate": 0.02,
+      "loss": 1.6042,
+      "step": 4518
+    },
+    {
+      "epoch": 0.4183532951455187,
+      "grad_norm": 0.15625,
+      "learning_rate": 0.02,
+      "loss": 1.6203,
+      "step": 4519
+    },
+    {
+      "epoch": 0.4184458716657988,
+      "grad_norm": 0.150390625,
+      "learning_rate": 0.02,
+      "loss": 1.6173,
+      "step": 4520
+    },
+    {
+      "epoch": 0.4185384481860788,
+      "grad_norm": 0.1484375,
+      "learning_rate": 0.02,
+      "loss": 1.5846,
+      "step": 4521
+    },
+    {
+      "epoch": 0.4186310247063588,
+      "grad_norm": 0.1494140625,
+      "learning_rate": 0.02,
+      "loss": 1.6173,
+      "step": 4522
+    },
+    {
+      "epoch": 0.4187236012266389,
+      "grad_norm": 0.1494140625,
+      "learning_rate": 0.02,
+      "loss": 1.5804,
+      "step": 4523
+    },
+    {
+      "epoch": 0.4188161777469189,
+      "grad_norm": 0.1376953125,
+      "learning_rate": 0.02,
+      "loss": 1.5647,
+      "step": 4524
+    },
+    {
+      "epoch": 0.418908754267199,
+      "grad_norm": 0.138671875,
+      "learning_rate": 0.02,
+      "loss": 1.5745,
+      "step": 4525
+    },
+    {
+      "epoch": 0.41900133078747903,
+      "grad_norm": 0.1337890625,
+      "learning_rate": 0.02,
+      "loss": 1.6194,
+      "step": 4526
+    },
+    {
+      "epoch": 0.4190939073077591,
+      "grad_norm": 0.1396484375,
+      "learning_rate": 0.02,
+      "loss": 1.6075,
+      "step": 4527
+    },
+    {
+      "epoch": 0.41918648382803914,
+      "grad_norm": 0.1396484375,
+      "learning_rate": 0.02,
+      "loss": 1.6039,
+      "step": 4528
+    },
+    {
+      "epoch": 0.41927906034831913,
+      "grad_norm": 0.1357421875,
+      "learning_rate": 0.02,
+      "loss": 1.6012,
+      "step": 4529
+    },
+    {
+      "epoch": 0.4193716368685992,
+      "grad_norm": 0.134765625,
+      "learning_rate": 0.02,
+      "loss": 1.5522,
+      "step": 4530
+    },
+    {
+      "epoch": 0.41946421338887924,
+      "grad_norm": 0.13671875,
+      "learning_rate": 0.02,
+      "loss": 1.5983,
+      "step": 4531
+    },
+    {
+      "epoch": 0.4195567899091593,
+      "grad_norm": 0.1474609375,
+      "learning_rate": 0.02,
+      "loss": 1.5856,
+      "step": 4532
+    },
+    {
+      "epoch": 0.41964936642943934,
+      "grad_norm": 0.154296875,
+      "learning_rate": 0.02,
+      "loss": 1.6198,
+      "step": 4533
+    },
+    {
+      "epoch": 0.4197419429497194,
+      "grad_norm": 0.154296875,
+      "learning_rate": 0.02,
+      "loss": 1.6228,
+      "step": 4534
+    },
+    {
+      "epoch": 0.41983451946999945,
+      "grad_norm": 0.1484375,
+      "learning_rate": 0.02,
+      "loss": 1.6646,
+      "step": 4535
+    },
+    {
+      "epoch": 0.41992709599027944,
+      "grad_norm": 0.1494140625,
+      "learning_rate": 0.02,
+      "loss": 1.6177,
+      "step": 4536
+    },
+    {
+      "epoch": 0.4200196725105595,
+      "grad_norm": 0.1435546875,
+      "learning_rate": 0.02,
+      "loss": 1.5565,
+      "step": 4537
+    },
+    {
+      "epoch": 0.42011224903083955,
+      "grad_norm": 0.146484375,
+      "learning_rate": 0.02,
+      "loss": 1.6071,
+      "step": 4538
+    },
+    {
+      "epoch": 0.4202048255511196,
+      "grad_norm": 0.1318359375,
+      "learning_rate": 0.02,
+      "loss": 1.5883,
+      "step": 4539
+    },
+    {
+      "epoch": 0.42029740207139965,
+      "grad_norm": 0.1455078125,
+      "learning_rate": 0.02,
+      "loss": 1.539,
+      "step": 4540
+    },
+    {
+      "epoch": 0.4203899785916797,
+      "grad_norm": 0.1416015625,
+      "learning_rate": 0.02,
+      "loss": 1.596,
+      "step": 4541
+    },
+    {
+      "epoch": 0.42048255511195975,
+      "grad_norm": 0.1474609375,
+      "learning_rate": 0.02,
+      "loss": 1.6224,
+      "step": 4542
+    },
+    {
+      "epoch": 0.42057513163223975,
+      "grad_norm": 0.15625,
+      "learning_rate": 0.02,
+      "loss": 1.6061,
+      "step": 4543
+    },
+    {
+      "epoch": 0.4206677081525198,
+      "grad_norm": 0.15234375,
+      "learning_rate": 0.02,
+      "loss": 1.6681,
+      "step": 4544
+    },
+    {
+      "epoch": 0.42076028467279986,
+      "grad_norm": 0.140625,
+      "learning_rate": 0.02,
+      "loss": 1.5662,
+      "step": 4545
+    },
+    {
+      "epoch": 0.4208528611930799,
+      "grad_norm": 0.1435546875,
+      "learning_rate": 0.02,
+      "loss": 1.6055,
+      "step": 4546
+    },
+    {
+      "epoch": 0.42094543771335996,
+      "grad_norm": 0.1396484375,
+      "learning_rate": 0.02,
+      "loss": 1.6651,
+      "step": 4547
+    },
+    {
+      "epoch": 0.42103801423364,
+      "grad_norm": 0.1474609375,
+      "learning_rate": 0.02,
+      "loss": 1.6726,
+      "step": 4548
+    },
+    {
+      "epoch": 0.42113059075392006,
+      "grad_norm": 0.1552734375,
+      "learning_rate": 0.02,
+      "loss": 1.5954,
+      "step": 4549
+    },
+    {
+      "epoch": 0.42122316727420006,
+      "grad_norm": 0.138671875,
+      "learning_rate": 0.02,
+      "loss": 1.5733,
+      "step": 4550
+    },
+    {
+      "epoch": 0.4213157437944801,
+      "grad_norm": 0.1474609375,
+      "learning_rate": 0.02,
+      "loss": 1.616,
+      "step": 4551
+    },
+    {
+      "epoch": 0.42140832031476017,
+      "grad_norm": 0.134765625,
+      "learning_rate": 0.02,
+      "loss": 1.5566,
+      "step": 4552
+    },
+    {
+      "epoch": 0.4215008968350402,
+      "grad_norm": 0.1435546875,
+      "learning_rate": 0.02,
+      "loss": 1.571,
+      "step": 4553
+    },
+    {
+      "epoch": 0.42159347335532027,
+      "grad_norm": 0.1474609375,
+      "learning_rate": 0.02,
+      "loss": 1.6187,
+      "step": 4554
+    },
+    {
+      "epoch": 0.4216860498756003,
+      "grad_norm": 0.142578125,
+      "learning_rate": 0.02,
+      "loss": 1.5604,
+      "step": 4555
+    },
+    {
+      "epoch": 0.4217786263958803,
+      "grad_norm": 0.140625,
+      "learning_rate": 0.02,
+      "loss": 1.5655,
+      "step": 4556
+    },
+    {
+      "epoch": 0.42187120291616037,
+      "grad_norm": 0.140625,
+      "learning_rate": 0.02,
+      "loss": 1.5874,
+      "step": 4557
+    },
+    {
+      "epoch": 0.4219637794364404,
+      "grad_norm": 0.142578125,
+      "learning_rate": 0.02,
+      "loss": 1.5224,
+      "step": 4558
+    },
+    {
+      "epoch": 0.4220563559567205,
+      "grad_norm": 0.16015625,
+      "learning_rate": 0.02,
+      "loss": 1.6681,
+      "step": 4559
+    },
+    {
+      "epoch": 0.4221489324770005,
+      "grad_norm": 0.1513671875,
+      "learning_rate": 0.02,
+      "loss": 1.5736,
+      "step": 4560
+    },
+    {
+      "epoch": 0.4222415089972806,
+      "grad_norm": 0.1484375,
+      "learning_rate": 0.02,
+      "loss": 1.5621,
+      "step": 4561
+    },
+    {
+      "epoch": 0.42233408551756063,
+      "grad_norm": 0.1513671875,
+      "learning_rate": 0.02,
+      "loss": 1.6073,
+      "step": 4562
+    },
+    {
+      "epoch": 0.42242666203784063,
+      "grad_norm": 0.1630859375,
+      "learning_rate": 0.02,
+      "loss": 1.5996,
+      "step": 4563
+    },
+    {
+      "epoch": 0.4225192385581207,
+      "grad_norm": 0.1533203125,
+      "learning_rate": 0.02,
+      "loss": 1.5678,
+      "step": 4564
+    },
+    {
+      "epoch": 0.42261181507840073,
+      "grad_norm": 0.1591796875,
+      "learning_rate": 0.02,
+      "loss": 1.5855,
+      "step": 4565
+    },
+    {
+      "epoch": 0.4227043915986808,
+      "grad_norm": 0.14453125,
+      "learning_rate": 0.02,
+      "loss": 1.5989,
+      "step": 4566
+    },
+    {
+      "epoch": 0.42279696811896084,
+      "grad_norm": 0.1533203125,
+      "learning_rate": 0.02,
+      "loss": 1.6074,
+      "step": 4567
+    },
+    {
+      "epoch": 0.4228895446392409,
+      "grad_norm": 0.1396484375,
+      "learning_rate": 0.02,
+      "loss": 1.5978,
+      "step": 4568
+    },
+    {
+      "epoch": 0.42298212115952094,
+      "grad_norm": 0.1533203125,
+      "learning_rate": 0.02,
+      "loss": 1.6012,
+      "step": 4569
+    },
+    {
+      "epoch": 0.42307469767980094,
+      "grad_norm": 0.1474609375,
+      "learning_rate": 0.02,
+      "loss": 1.6151,
+      "step": 4570
+    },
+    {
+      "epoch": 0.423167274200081,
+      "grad_norm": 0.150390625,
+      "learning_rate": 0.02,
+      "loss": 1.5633,
+      "step": 4571
+    },
+    {
+      "epoch": 0.42325985072036104,
+      "grad_norm": 0.14453125,
+      "learning_rate": 0.02,
+      "loss": 1.4883,
+      "step": 4572
+    },
+    {
+      "epoch": 0.4233524272406411,
+      "grad_norm": 0.1455078125,
+      "learning_rate": 0.02,
+      "loss": 1.6219,
+      "step": 4573
+    },
+    {
+      "epoch": 0.42344500376092115,
+      "grad_norm": 0.142578125,
+      "learning_rate": 0.02,
+      "loss": 1.6101,
+      "step": 4574
+    },
+    {
+      "epoch": 0.4235375802812012,
+      "grad_norm": 0.1416015625,
+      "learning_rate": 0.02,
+      "loss": 1.598,
+      "step": 4575
+    },
+    {
+      "epoch": 0.42363015680148125,
+      "grad_norm": 0.1455078125,
+      "learning_rate": 0.02,
+      "loss": 1.554,
+      "step": 4576
+    },
+    {
+      "epoch": 0.42372273332176125,
+      "grad_norm": 0.1494140625,
+      "learning_rate": 0.02,
+      "loss": 1.5637,
+      "step": 4577
+    },
+    {
+      "epoch": 0.4238153098420413,
+      "grad_norm": 0.14453125,
+      "learning_rate": 0.02,
+      "loss": 1.6173,
+      "step": 4578
+    },
+    {
+      "epoch": 0.42390788636232135,
+      "grad_norm": 0.1484375,
+      "learning_rate": 0.02,
+      "loss": 1.6129,
+      "step": 4579
+    },
+    {
+      "epoch": 0.4240004628826014,
+      "grad_norm": 0.14453125,
+      "learning_rate": 0.02,
+      "loss": 1.5938,
+      "step": 4580
+    },
+    {
+      "epoch": 0.42409303940288146,
+      "grad_norm": 0.146484375,
+      "learning_rate": 0.02,
+      "loss": 1.5648,
+      "step": 4581
+    },
+    {
+      "epoch": 0.4241856159231615,
+      "grad_norm": 0.1484375,
+      "learning_rate": 0.02,
+      "loss": 1.6303,
+      "step": 4582
+    },
+    {
+      "epoch": 0.42427819244344156,
+      "grad_norm": 0.1533203125,
+      "learning_rate": 0.02,
+      "loss": 1.578,
+      "step": 4583
+    },
+    {
+      "epoch": 0.42437076896372156,
+      "grad_norm": 0.14453125,
+      "learning_rate": 0.02,
+      "loss": 1.5716,
+      "step": 4584
+    },
+    {
+      "epoch": 0.4244633454840016,
+      "grad_norm": 0.1513671875,
+      "learning_rate": 0.02,
+      "loss": 1.5285,
+      "step": 4585
+    },
+    {
+      "epoch": 0.42455592200428166,
+      "grad_norm": 0.1455078125,
+      "learning_rate": 0.02,
+      "loss": 1.6214,
+      "step": 4586
+    },
+    {
+      "epoch": 0.4246484985245617,
+      "grad_norm": 0.146484375,
+      "learning_rate": 0.02,
+      "loss": 1.6134,
+      "step": 4587
+    },
+    {
+      "epoch": 0.42474107504484176,
+      "grad_norm": 0.146484375,
+      "learning_rate": 0.02,
+      "loss": 1.5909,
+      "step": 4588
+    },
+    {
+      "epoch": 0.4248336515651218,
+      "grad_norm": 0.1513671875,
+      "learning_rate": 0.02,
+      "loss": 1.5924,
+      "step": 4589
+    },
+    {
+      "epoch": 0.4249262280854018,
+      "grad_norm": 0.140625,
+      "learning_rate": 0.02,
+      "loss": 1.5629,
+      "step": 4590
+    },
+    {
+      "epoch": 0.42501880460568187,
+      "grad_norm": 0.1376953125,
+      "learning_rate": 0.02,
+      "loss": 1.5216,
+      "step": 4591
+    },
+    {
+      "epoch": 0.4251113811259619,
+      "grad_norm": 0.14453125,
+      "learning_rate": 0.02,
+      "loss": 1.6209,
+      "step": 4592
+    },
+    {
+      "epoch": 0.42520395764624197,
+      "grad_norm": 0.1416015625,
+      "learning_rate": 0.02,
+      "loss": 1.6058,
+      "step": 4593
+    },
+    {
+      "epoch": 0.425296534166522,
+      "grad_norm": 0.1435546875,
+      "learning_rate": 0.02,
+      "loss": 1.5569,
+      "step": 4594
+    },
+    {
+      "epoch": 0.4253891106868021,
+      "grad_norm": 0.1416015625,
+      "learning_rate": 0.02,
+      "loss": 1.538,
+      "step": 4595
+    },
+    {
+      "epoch": 0.4254816872070821,
+      "grad_norm": 0.150390625,
+      "learning_rate": 0.02,
+      "loss": 1.5408,
+      "step": 4596
+    },
+    {
+      "epoch": 0.4255742637273621,
+      "grad_norm": 0.15234375,
+      "learning_rate": 0.02,
+      "loss": 1.6279,
+      "step": 4597
+    },
+    {
+      "epoch": 0.4256668402476422,
+      "grad_norm": 0.15234375,
+      "learning_rate": 0.02,
+      "loss": 1.5667,
+      "step": 4598
+    },
+    {
+      "epoch": 0.4257594167679222,
+      "grad_norm": 0.158203125,
+      "learning_rate": 0.02,
+      "loss": 1.5523,
+      "step": 4599
+    },
+    {
+      "epoch": 0.4258519932882023,
+      "grad_norm": 0.1640625,
+      "learning_rate": 0.02,
+      "loss": 1.6297,
+      "step": 4600
+    },
+    {
+      "epoch": 0.42594456980848233,
+      "grad_norm": 0.158203125,
+      "learning_rate": 0.02,
+      "loss": 1.5951,
+      "step": 4601
+    },
+    {
+      "epoch": 0.4260371463287624,
+      "grad_norm": 0.166015625,
+      "learning_rate": 0.02,
+      "loss": 1.5858,
+      "step": 4602
+    },
+    {
+      "epoch": 0.42612972284904244,
+      "grad_norm": 0.15234375,
+      "learning_rate": 0.02,
+      "loss": 1.6128,
+      "step": 4603
+    },
+    {
+      "epoch": 0.42622229936932243,
+      "grad_norm": 0.154296875,
+      "learning_rate": 0.02,
+      "loss": 1.5569,
+      "step": 4604
+    },
+    {
+      "epoch": 0.4263148758896025,
+      "grad_norm": 0.1513671875,
+      "learning_rate": 0.02,
+      "loss": 1.6101,
+      "step": 4605
+    },
+    {
+      "epoch": 0.42640745240988254,
+      "grad_norm": 0.1533203125,
+      "learning_rate": 0.02,
+      "loss": 1.6742,
+      "step": 4606
+    },
+    {
+      "epoch": 0.4265000289301626,
+      "grad_norm": 0.1513671875,
+      "learning_rate": 0.02,
+      "loss": 1.5509,
+      "step": 4607
+    },
+    {
+      "epoch": 0.42659260545044264,
+      "grad_norm": 0.1513671875,
+      "learning_rate": 0.02,
+      "loss": 1.6307,
+      "step": 4608
+    },
+    {
+      "epoch": 0.4266851819707227,
+      "grad_norm": 0.1455078125,
+      "learning_rate": 0.02,
+      "loss": 1.5433,
+      "step": 4609
+    },
+    {
+      "epoch": 0.42677775849100275,
+      "grad_norm": 0.1484375,
+      "learning_rate": 0.02,
+      "loss": 1.6117,
+      "step": 4610
+    },
+    {
+      "epoch": 0.42687033501128274,
+      "grad_norm": 0.1494140625,
+      "learning_rate": 0.02,
+      "loss": 1.5966,
+      "step": 4611
+    },
+    {
+      "epoch": 0.4269629115315628,
+      "grad_norm": 0.140625,
+      "learning_rate": 0.02,
+      "loss": 1.6227,
+      "step": 4612
+    },
+    {
+      "epoch": 0.42705548805184285,
+      "grad_norm": 0.13671875,
+      "learning_rate": 0.02,
+      "loss": 1.5412,
+      "step": 4613
+    },
+    {
+      "epoch": 0.4271480645721229,
+      "grad_norm": 0.1494140625,
+      "learning_rate": 0.02,
+      "loss": 1.6433,
+      "step": 4614
+    },
+    {
+      "epoch": 0.42724064109240295,
+      "grad_norm": 0.142578125,
+      "learning_rate": 0.02,
+      "loss": 1.5747,
+      "step": 4615
+    },
+    {
+      "epoch": 0.427333217612683,
+      "grad_norm": 0.1494140625,
+      "learning_rate": 0.02,
+      "loss": 1.5988,
+      "step": 4616
+    },
+    {
+      "epoch": 0.427425794132963,
+      "grad_norm": 0.1396484375,
+      "learning_rate": 0.02,
+      "loss": 1.6299,
+      "step": 4617
+    },
+    {
+      "epoch": 0.42751837065324305,
+      "grad_norm": 0.1396484375,
+      "learning_rate": 0.02,
+      "loss": 1.5879,
+      "step": 4618
+    },
+    {
+      "epoch": 0.4276109471735231,
+      "grad_norm": 0.1376953125,
+      "learning_rate": 0.02,
+      "loss": 1.5556,
+      "step": 4619
+    },
+    {
+      "epoch": 0.42770352369380316,
+      "grad_norm": 0.1435546875,
+      "learning_rate": 0.02,
+      "loss": 1.6339,
+      "step": 4620
+    },
+    {
+      "epoch": 0.4277961002140832,
+      "grad_norm": 0.1455078125,
+      "learning_rate": 0.02,
+      "loss": 1.5919,
+      "step": 4621
+    },
+    {
+      "epoch": 0.42788867673436326,
+      "grad_norm": 0.138671875,
+      "learning_rate": 0.02,
+      "loss": 1.6252,
+      "step": 4622
+    },
+    {
+      "epoch": 0.4279812532546433,
+      "grad_norm": 0.1455078125,
+      "learning_rate": 0.02,
+      "loss": 1.5898,
+      "step": 4623
+    },
+    {
+      "epoch": 0.4280738297749233,
+      "grad_norm": 0.1474609375,
+      "learning_rate": 0.02,
+      "loss": 1.589,
+      "step": 4624
+    },
+    {
+      "epoch": 0.42816640629520336,
+      "grad_norm": 0.1552734375,
+      "learning_rate": 0.02,
+      "loss": 1.625,
+      "step": 4625
+    },
+    {
+      "epoch": 0.4282589828154834,
+      "grad_norm": 0.1416015625,
+      "learning_rate": 0.02,
+      "loss": 1.618,
+      "step": 4626
+    },
+    {
+      "epoch": 0.42835155933576347,
+      "grad_norm": 0.15234375,
+      "learning_rate": 0.02,
+      "loss": 1.6095,
+      "step": 4627
+    },
+    {
+      "epoch": 0.4284441358560435,
+      "grad_norm": 0.1396484375,
+      "learning_rate": 0.02,
+      "loss": 1.6287,
+      "step": 4628
+    },
+    {
+      "epoch": 0.42853671237632357,
+      "grad_norm": 0.1494140625,
+      "learning_rate": 0.02,
+      "loss": 1.6039,
+      "step": 4629
+    },
+    {
+      "epoch": 0.4286292888966036,
+      "grad_norm": 0.1376953125,
+      "learning_rate": 0.02,
+      "loss": 1.5732,
+      "step": 4630
+    },
+    {
+      "epoch": 0.4287218654168836,
+      "grad_norm": 0.150390625,
+      "learning_rate": 0.02,
+      "loss": 1.6061,
+      "step": 4631
+    },
+    {
+      "epoch": 0.42881444193716367,
+      "grad_norm": 0.134765625,
+      "learning_rate": 0.02,
+      "loss": 1.5666,
+      "step": 4632
+    },
+    {
+      "epoch": 0.4289070184574437,
+      "grad_norm": 0.1396484375,
+      "learning_rate": 0.02,
+      "loss": 1.5948,
+      "step": 4633
+    },
+    {
+      "epoch": 0.4289995949777238,
+      "grad_norm": 0.1337890625,
+      "learning_rate": 0.02,
+      "loss": 1.5487,
+      "step": 4634
+    },
+    {
+      "epoch": 0.4290921714980038,
+      "grad_norm": 0.1513671875,
+      "learning_rate": 0.02,
+      "loss": 1.6123,
+      "step": 4635
+    },
+    {
+      "epoch": 0.4291847480182839,
+      "grad_norm": 0.1513671875,
+      "learning_rate": 0.02,
+      "loss": 1.6444,
+      "step": 4636
+    },
+    {
+      "epoch": 0.42927732453856393,
+      "grad_norm": 0.146484375,
+      "learning_rate": 0.02,
+      "loss": 1.5958,
+      "step": 4637
+    },
+    {
+      "epoch": 0.42936990105884393,
+      "grad_norm": 0.138671875,
+      "learning_rate": 0.02,
+      "loss": 1.5434,
+      "step": 4638
+    },
+    {
+      "epoch": 0.429462477579124,
+      "grad_norm": 0.1484375,
+      "learning_rate": 0.02,
+      "loss": 1.6169,
+      "step": 4639
+    },
+    {
+      "epoch": 0.42955505409940403,
+      "grad_norm": 0.1435546875,
+      "learning_rate": 0.02,
+      "loss": 1.5716,
+      "step": 4640
+    },
+    {
+      "epoch": 0.4296476306196841,
+      "grad_norm": 0.15234375,
+      "learning_rate": 0.02,
+      "loss": 1.6589,
+      "step": 4641
+    },
+    {
+      "epoch": 0.42974020713996414,
+      "grad_norm": 0.1494140625,
+      "learning_rate": 0.02,
+      "loss": 1.5798,
+      "step": 4642
+    },
+    {
+      "epoch": 0.4298327836602442,
+      "grad_norm": 0.1484375,
+      "learning_rate": 0.02,
+      "loss": 1.5836,
+      "step": 4643
+    },
+    {
+      "epoch": 0.42992536018052424,
+      "grad_norm": 0.158203125,
+      "learning_rate": 0.02,
+      "loss": 1.6254,
+      "step": 4644
+    },
+    {
+      "epoch": 0.43001793670080424,
+      "grad_norm": 0.1416015625,
+      "learning_rate": 0.02,
+      "loss": 1.5951,
+      "step": 4645
+    },
+    {
+      "epoch": 0.4301105132210843,
+      "grad_norm": 0.142578125,
+      "learning_rate": 0.02,
+      "loss": 1.6165,
+      "step": 4646
+    },
+    {
+      "epoch": 0.43020308974136434,
+      "grad_norm": 0.1435546875,
+      "learning_rate": 0.02,
+      "loss": 1.5704,
+      "step": 4647
+    },
+    {
+      "epoch": 0.4302956662616444,
+      "grad_norm": 0.1484375,
+      "learning_rate": 0.02,
+      "loss": 1.5434,
+      "step": 4648
+    },
+    {
+      "epoch": 0.43038824278192445,
+      "grad_norm": 0.1357421875,
+      "learning_rate": 0.02,
+      "loss": 1.5505,
+      "step": 4649
+    },
+    {
+      "epoch": 0.4304808193022045,
+      "grad_norm": 0.142578125,
+      "learning_rate": 0.02,
+      "loss": 1.5706,
+      "step": 4650
+    },
+    {
+      "epoch": 0.4305733958224845,
+      "grad_norm": 0.1357421875,
+      "learning_rate": 0.02,
+      "loss": 1.5844,
+      "step": 4651
+    },
+    {
+      "epoch": 0.43066597234276455,
+      "grad_norm": 0.138671875,
+      "learning_rate": 0.02,
+      "loss": 1.6047,
+      "step": 4652
+    },
+    {
+      "epoch": 0.4307585488630446,
+      "grad_norm": 0.146484375,
+      "learning_rate": 0.02,
+      "loss": 1.5945,
+      "step": 4653
+    },
+    {
+      "epoch": 0.43085112538332465,
+      "grad_norm": 0.146484375,
+      "learning_rate": 0.02,
+      "loss": 1.5545,
+      "step": 4654
+    },
+    {
+      "epoch": 0.4309437019036047,
+      "grad_norm": 0.150390625,
+      "learning_rate": 0.02,
+      "loss": 1.6208,
+      "step": 4655
+    },
+    {
+      "epoch": 0.43103627842388476,
+      "grad_norm": 0.1474609375,
+      "learning_rate": 0.02,
+      "loss": 1.6051,
+      "step": 4656
+    },
+    {
+      "epoch": 0.4311288549441648,
+      "grad_norm": 0.1552734375,
+      "learning_rate": 0.02,
+      "loss": 1.5962,
+      "step": 4657
+    },
+    {
+      "epoch": 0.4312214314644448,
+      "grad_norm": 0.1455078125,
+      "learning_rate": 0.02,
+      "loss": 1.6331,
+      "step": 4658
+    },
+    {
+      "epoch": 0.43131400798472486,
+      "grad_norm": 0.16015625,
+      "learning_rate": 0.02,
+      "loss": 1.6221,
+      "step": 4659
+    },
+    {
+      "epoch": 0.4314065845050049,
+      "grad_norm": 0.14453125,
+      "learning_rate": 0.02,
+      "loss": 1.5742,
+      "step": 4660
+    },
+    {
+      "epoch": 0.43149916102528496,
+      "grad_norm": 0.150390625,
+      "learning_rate": 0.02,
+      "loss": 1.5714,
+      "step": 4661
+    },
+    {
+      "epoch": 0.431591737545565,
+      "grad_norm": 0.1337890625,
+      "learning_rate": 0.02,
+      "loss": 1.5328,
+      "step": 4662
+    },
+    {
+      "epoch": 0.43168431406584507,
+      "grad_norm": 0.1513671875,
+      "learning_rate": 0.02,
+      "loss": 1.5598,
+      "step": 4663
+    },
+    {
+      "epoch": 0.4317768905861251,
+      "grad_norm": 0.1357421875,
+      "learning_rate": 0.02,
+      "loss": 1.5923,
+      "step": 4664
+    },
+    {
+      "epoch": 0.4318694671064051,
+      "grad_norm": 0.1376953125,
+      "learning_rate": 0.02,
+      "loss": 1.5421,
+      "step": 4665
+    },
+    {
+      "epoch": 0.43196204362668517,
+      "grad_norm": 0.1435546875,
+      "learning_rate": 0.02,
+      "loss": 1.6316,
+      "step": 4666
+    },
+    {
+      "epoch": 0.4320546201469652,
+      "grad_norm": 0.14453125,
+      "learning_rate": 0.02,
+      "loss": 1.611,
+      "step": 4667
+    },
+    {
+      "epoch": 0.43214719666724527,
+      "grad_norm": 0.146484375,
+      "learning_rate": 0.02,
+      "loss": 1.5944,
+      "step": 4668
+    },
+    {
+      "epoch": 0.4322397731875253,
+      "grad_norm": 0.1435546875,
+      "learning_rate": 0.02,
+      "loss": 1.572,
+      "step": 4669
+    },
+    {
+      "epoch": 0.4323323497078054,
+      "grad_norm": 0.150390625,
+      "learning_rate": 0.02,
+      "loss": 1.534,
+      "step": 4670
+    },
+    {
+      "epoch": 0.4324249262280854,
+      "grad_norm": 0.13671875,
+      "learning_rate": 0.02,
+      "loss": 1.5814,
+      "step": 4671
+    },
+    {
+      "epoch": 0.4325175027483654,
+      "grad_norm": 0.14453125,
+      "learning_rate": 0.02,
+      "loss": 1.5709,
+      "step": 4672
+    },
+    {
+      "epoch": 0.4326100792686455,
+      "grad_norm": 0.138671875,
+      "learning_rate": 0.02,
+      "loss": 1.5352,
+      "step": 4673
+    },
+    {
+      "epoch": 0.4327026557889255,
+      "grad_norm": 0.1484375,
+      "learning_rate": 0.02,
+      "loss": 1.5398,
+      "step": 4674
+    },
+    {
+      "epoch": 0.4327952323092056,
+      "grad_norm": 0.15625,
+      "learning_rate": 0.02,
+      "loss": 1.6126,
+      "step": 4675
+    },
+    {
+      "epoch": 0.43288780882948563,
+      "grad_norm": 0.1484375,
+      "learning_rate": 0.02,
+      "loss": 1.5753,
+      "step": 4676
+    },
+    {
+      "epoch": 0.4329803853497657,
+      "grad_norm": 0.15234375,
+      "learning_rate": 0.02,
+      "loss": 1.5966,
+      "step": 4677
+    },
+    {
+      "epoch": 0.43307296187004574,
+      "grad_norm": 0.150390625,
+      "learning_rate": 0.02,
+      "loss": 1.5827,
+      "step": 4678
+    },
+    {
+      "epoch": 0.43316553839032573,
+      "grad_norm": 0.1474609375,
+      "learning_rate": 0.02,
+      "loss": 1.537,
+      "step": 4679
+    },
+    {
+      "epoch": 0.4332581149106058,
+      "grad_norm": 0.150390625,
+      "learning_rate": 0.02,
+      "loss": 1.5989,
+      "step": 4680
+    },
+    {
+      "epoch": 0.43335069143088584,
+      "grad_norm": 0.1494140625,
+      "learning_rate": 0.02,
+      "loss": 1.6216,
+      "step": 4681
+    },
+    {
+      "epoch": 0.4334432679511659,
+      "grad_norm": 0.14453125,
+      "learning_rate": 0.02,
+      "loss": 1.523,
+      "step": 4682
+    },
+    {
+      "epoch": 0.43353584447144594,
+      "grad_norm": 0.1494140625,
+      "learning_rate": 0.02,
+      "loss": 1.6209,
+      "step": 4683
+    },
+    {
+      "epoch": 0.433628420991726,
+      "grad_norm": 0.1474609375,
+      "learning_rate": 0.02,
+      "loss": 1.5372,
+      "step": 4684
+    },
+    {
+      "epoch": 0.433720997512006,
+      "grad_norm": 0.146484375,
+      "learning_rate": 0.02,
+      "loss": 1.6042,
+      "step": 4685
+    },
+    {
+      "epoch": 0.43381357403228604,
+      "grad_norm": 0.1494140625,
+      "learning_rate": 0.02,
+      "loss": 1.576,
+      "step": 4686
+    },
+    {
+      "epoch": 0.4339061505525661,
+      "grad_norm": 0.1494140625,
+      "learning_rate": 0.02,
+      "loss": 1.5958,
+      "step": 4687
+    },
+    {
+      "epoch": 0.43399872707284615,
+      "grad_norm": 0.1552734375,
+      "learning_rate": 0.02,
+      "loss": 1.5918,
+      "step": 4688
+    },
+    {
+      "epoch": 0.4340913035931262,
+      "grad_norm": 0.1435546875,
+      "learning_rate": 0.02,
+      "loss": 1.6306,
+      "step": 4689
+    },
+    {
+      "epoch": 0.43418388011340625,
+      "grad_norm": 0.1494140625,
+      "learning_rate": 0.02,
+      "loss": 1.6193,
+      "step": 4690
+    },
+    {
+      "epoch": 0.4342764566336863,
+      "grad_norm": 0.14453125,
+      "learning_rate": 0.02,
+      "loss": 1.6097,
+      "step": 4691
+    },
+    {
+      "epoch": 0.4343690331539663,
+      "grad_norm": 0.1513671875,
+      "learning_rate": 0.02,
+      "loss": 1.5536,
+      "step": 4692
+    },
+    {
+      "epoch": 0.43446160967424635,
+      "grad_norm": 0.1640625,
+      "learning_rate": 0.02,
+      "loss": 1.6389,
+      "step": 4693
+    },
+    {
+      "epoch": 0.4345541861945264,
+      "grad_norm": 0.1533203125,
+      "learning_rate": 0.02,
+      "loss": 1.5533,
+      "step": 4694
+    },
+    {
+      "epoch": 0.43464676271480646,
+      "grad_norm": 0.1572265625,
+      "learning_rate": 0.02,
+      "loss": 1.6027,
+      "step": 4695
+    },
+    {
+      "epoch": 0.4347393392350865,
+      "grad_norm": 0.138671875,
+      "learning_rate": 0.02,
+      "loss": 1.5512,
+      "step": 4696
+    },
+    {
+      "epoch": 0.43483191575536656,
+      "grad_norm": 0.1474609375,
+      "learning_rate": 0.02,
+      "loss": 1.6184,
+      "step": 4697
+    },
+    {
+      "epoch": 0.4349244922756466,
+      "grad_norm": 0.1376953125,
+      "learning_rate": 0.02,
+      "loss": 1.6185,
+      "step": 4698
+    },
+    {
+      "epoch": 0.4350170687959266,
+      "grad_norm": 0.1396484375,
+      "learning_rate": 0.02,
+      "loss": 1.6005,
+      "step": 4699
+    },
+    {
+      "epoch": 0.43510964531620666,
+      "grad_norm": 0.1455078125,
+      "learning_rate": 0.02,
+      "loss": 1.628,
+      "step": 4700
+    },
+    {
+      "epoch": 0.4352022218364867,
+      "grad_norm": 0.150390625,
+      "learning_rate": 0.02,
+      "loss": 1.5702,
+      "step": 4701
+    },
+    {
+      "epoch": 0.43529479835676677,
+      "grad_norm": 0.1396484375,
+      "learning_rate": 0.02,
+      "loss": 1.5505,
+      "step": 4702
+    },
+    {
+      "epoch": 0.4353873748770468,
+      "grad_norm": 0.1494140625,
+      "learning_rate": 0.02,
+      "loss": 1.6205,
+      "step": 4703
+    },
+    {
+      "epoch": 0.43547995139732687,
+      "grad_norm": 0.142578125,
+      "learning_rate": 0.02,
+      "loss": 1.5705,
+      "step": 4704
+    },
+    {
+      "epoch": 0.4355725279176069,
+      "grad_norm": 0.1591796875,
+      "learning_rate": 0.02,
+      "loss": 1.6085,
+      "step": 4705
+    },
+    {
+      "epoch": 0.4356651044378869,
+      "grad_norm": 0.1396484375,
+      "learning_rate": 0.02,
+      "loss": 1.6355,
+      "step": 4706
+    },
+    {
+      "epoch": 0.43575768095816697,
+      "grad_norm": 0.15234375,
+      "learning_rate": 0.02,
+      "loss": 1.6231,
+      "step": 4707
+    },
+    {
+      "epoch": 0.435850257478447,
+      "grad_norm": 0.1396484375,
+      "learning_rate": 0.02,
+      "loss": 1.5542,
+      "step": 4708
+    },
+    {
+      "epoch": 0.4359428339987271,
+      "grad_norm": 0.1552734375,
+      "learning_rate": 0.02,
+      "loss": 1.6085,
+      "step": 4709
+    },
+    {
+      "epoch": 0.4360354105190071,
+      "grad_norm": 0.1533203125,
+      "learning_rate": 0.02,
+      "loss": 1.6192,
+      "step": 4710
+    },
+    {
+      "epoch": 0.4361279870392872,
+      "grad_norm": 0.154296875,
+      "learning_rate": 0.02,
+      "loss": 1.5708,
+      "step": 4711
+    },
+    {
+      "epoch": 0.43622056355956723,
+      "grad_norm": 0.1494140625,
+      "learning_rate": 0.02,
+      "loss": 1.6533,
+      "step": 4712
+    },
+    {
+      "epoch": 0.43631314007984723,
+      "grad_norm": 0.15625,
+      "learning_rate": 0.02,
+      "loss": 1.6195,
+      "step": 4713
+    },
+    {
+      "epoch": 0.4364057166001273,
+      "grad_norm": 0.1474609375,
+      "learning_rate": 0.02,
+      "loss": 1.6226,
+      "step": 4714
+    },
+    {
+      "epoch": 0.43649829312040733,
+      "grad_norm": 0.154296875,
+      "learning_rate": 0.02,
+      "loss": 1.5419,
+      "step": 4715
+    },
+    {
+      "epoch": 0.4365908696406874,
+      "grad_norm": 0.146484375,
+      "learning_rate": 0.02,
+      "loss": 1.5818,
+      "step": 4716
+    },
+    {
+      "epoch": 0.43668344616096744,
+      "grad_norm": 0.1552734375,
+      "learning_rate": 0.02,
+      "loss": 1.5584,
+      "step": 4717
+    },
+    {
+      "epoch": 0.4367760226812475,
+      "grad_norm": 0.1552734375,
+      "learning_rate": 0.02,
+      "loss": 1.5771,
+      "step": 4718
+    },
+    {
+      "epoch": 0.4368685992015275,
+      "grad_norm": 0.16796875,
+      "learning_rate": 0.02,
+      "loss": 1.5385,
+      "step": 4719
+    },
+    {
+      "epoch": 0.43696117572180754,
+      "grad_norm": 0.166015625,
+      "learning_rate": 0.02,
+      "loss": 1.6139,
+      "step": 4720
+    },
+    {
+      "epoch": 0.4370537522420876,
+      "grad_norm": 0.15234375,
+      "learning_rate": 0.02,
+      "loss": 1.5875,
+      "step": 4721
+    },
+    {
+      "epoch": 0.43714632876236764,
+      "grad_norm": 0.1533203125,
+      "learning_rate": 0.02,
+      "loss": 1.5578,
+      "step": 4722
+    },
+    {
+      "epoch": 0.4372389052826477,
+      "grad_norm": 0.1494140625,
+      "learning_rate": 0.02,
+      "loss": 1.5593,
+      "step": 4723
+    },
+    {
+      "epoch": 0.43733148180292775,
+      "grad_norm": 0.162109375,
+      "learning_rate": 0.02,
+      "loss": 1.5683,
+      "step": 4724
+    },
+    {
+      "epoch": 0.4374240583232078,
+      "grad_norm": 0.1376953125,
+      "learning_rate": 0.02,
+      "loss": 1.5324,
+      "step": 4725
+    },
+    {
+      "epoch": 0.4375166348434878,
+      "grad_norm": 0.1591796875,
+      "learning_rate": 0.02,
+      "loss": 1.6243,
+      "step": 4726
+    },
+    {
+      "epoch": 0.43760921136376785,
+      "grad_norm": 0.146484375,
+      "learning_rate": 0.02,
+      "loss": 1.5824,
+      "step": 4727
+    },
+    {
+      "epoch": 0.4377017878840479,
+      "grad_norm": 0.1474609375,
+      "learning_rate": 0.02,
+      "loss": 1.5923,
+      "step": 4728
+    },
+    {
+      "epoch": 0.43779436440432795,
+      "grad_norm": 0.1396484375,
+      "learning_rate": 0.02,
+      "loss": 1.5624,
+      "step": 4729
+    },
+    {
+      "epoch": 0.437886940924608,
+      "grad_norm": 0.1416015625,
+      "learning_rate": 0.02,
+      "loss": 1.5118,
+      "step": 4730
+    },
+    {
+      "epoch": 0.43797951744488806,
+      "grad_norm": 0.138671875,
+      "learning_rate": 0.02,
+      "loss": 1.5556,
+      "step": 4731
+    },
+    {
+      "epoch": 0.4380720939651681,
+      "grad_norm": 0.1484375,
+      "learning_rate": 0.02,
+      "loss": 1.6186,
+      "step": 4732
+    },
+    {
+      "epoch": 0.4381646704854481,
+      "grad_norm": 0.1513671875,
+      "learning_rate": 0.02,
+      "loss": 1.6069,
+      "step": 4733
+    },
+    {
+      "epoch": 0.43825724700572816,
+      "grad_norm": 0.154296875,
+      "learning_rate": 0.02,
+      "loss": 1.5917,
+      "step": 4734
+    },
+    {
+      "epoch": 0.4383498235260082,
+      "grad_norm": 0.14453125,
+      "learning_rate": 0.02,
+      "loss": 1.6041,
+      "step": 4735
+    },
+    {
+      "epoch": 0.43844240004628826,
+      "grad_norm": 0.1455078125,
+      "learning_rate": 0.02,
+      "loss": 1.5238,
+      "step": 4736
+    },
+    {
+      "epoch": 0.4385349765665683,
+      "grad_norm": 0.14453125,
+      "learning_rate": 0.02,
+      "loss": 1.5837,
+      "step": 4737
+    },
+    {
+      "epoch": 0.43862755308684837,
+      "grad_norm": 0.142578125,
+      "learning_rate": 0.02,
+      "loss": 1.5512,
+      "step": 4738
+    },
+    {
+      "epoch": 0.4387201296071284,
+      "grad_norm": 0.1494140625,
+      "learning_rate": 0.02,
+      "loss": 1.6138,
+      "step": 4739
+    },
+    {
+      "epoch": 0.4388127061274084,
+      "grad_norm": 0.1474609375,
+      "learning_rate": 0.02,
+      "loss": 1.5986,
+      "step": 4740
+    },
+    {
+      "epoch": 0.43890528264768847,
+      "grad_norm": 0.1435546875,
+      "learning_rate": 0.02,
+      "loss": 1.5692,
+      "step": 4741
+    },
+    {
+      "epoch": 0.4389978591679685,
+      "grad_norm": 0.1455078125,
+      "learning_rate": 0.02,
+      "loss": 1.5533,
+      "step": 4742
+    },
+    {
+      "epoch": 0.43909043568824857,
+      "grad_norm": 0.1435546875,
+      "learning_rate": 0.02,
+      "loss": 1.6276,
+      "step": 4743
+    },
+    {
+      "epoch": 0.4391830122085286,
+      "grad_norm": 0.138671875,
+      "learning_rate": 0.02,
+      "loss": 1.6314,
+      "step": 4744
+    },
+    {
+      "epoch": 0.4392755887288087,
+      "grad_norm": 0.1474609375,
+      "learning_rate": 0.02,
+      "loss": 1.6274,
+      "step": 4745
+    },
+    {
+      "epoch": 0.4393681652490887,
+      "grad_norm": 0.1455078125,
+      "learning_rate": 0.02,
+      "loss": 1.5821,
+      "step": 4746
+    },
+    {
+      "epoch": 0.4394607417693687,
+      "grad_norm": 0.1416015625,
+      "learning_rate": 0.02,
+      "loss": 1.5417,
+      "step": 4747
+    },
+    {
+      "epoch": 0.4395533182896488,
+      "grad_norm": 0.1376953125,
+      "learning_rate": 0.02,
+      "loss": 1.5658,
+      "step": 4748
+    },
+    {
+      "epoch": 0.4396458948099288,
+      "grad_norm": 0.14453125,
+      "learning_rate": 0.02,
+      "loss": 1.6306,
+      "step": 4749
+    },
+    {
+      "epoch": 0.4397384713302089,
+      "grad_norm": 0.1484375,
+      "learning_rate": 0.02,
+      "loss": 1.5915,
+      "step": 4750
+    },
+    {
+      "epoch": 0.43983104785048893,
+      "grad_norm": 0.150390625,
+      "learning_rate": 0.02,
+      "loss": 1.5479,
+      "step": 4751
+    },
+    {
+      "epoch": 0.439923624370769,
+      "grad_norm": 0.14453125,
+      "learning_rate": 0.02,
+      "loss": 1.6272,
+      "step": 4752
+    },
+    {
+      "epoch": 0.440016200891049,
+      "grad_norm": 0.1513671875,
+      "learning_rate": 0.02,
+      "loss": 1.5506,
+      "step": 4753
+    },
+    {
+      "epoch": 0.44010877741132903,
+      "grad_norm": 0.1455078125,
+      "learning_rate": 0.02,
+      "loss": 1.5466,
+      "step": 4754
+    },
+    {
+      "epoch": 0.4402013539316091,
+      "grad_norm": 0.1484375,
+      "learning_rate": 0.02,
+      "loss": 1.6237,
+      "step": 4755
+    },
+    {
+      "epoch": 0.44029393045188914,
+      "grad_norm": 0.1513671875,
+      "learning_rate": 0.02,
+      "loss": 1.6163,
+      "step": 4756
+    },
+    {
+      "epoch": 0.4403865069721692,
+      "grad_norm": 0.150390625,
+      "learning_rate": 0.02,
+      "loss": 1.5718,
+      "step": 4757
+    },
+    {
+      "epoch": 0.44047908349244924,
+      "grad_norm": 0.13671875,
+      "learning_rate": 0.02,
+      "loss": 1.6571,
+      "step": 4758
+    },
+    {
+      "epoch": 0.4405716600127293,
+      "grad_norm": 0.14453125,
+      "learning_rate": 0.02,
+      "loss": 1.5623,
+      "step": 4759
+    },
+    {
+      "epoch": 0.4406642365330093,
+      "grad_norm": 0.1455078125,
+      "learning_rate": 0.02,
+      "loss": 1.6397,
+      "step": 4760
+    },
+    {
+      "epoch": 0.44075681305328934,
+      "grad_norm": 0.1533203125,
+      "learning_rate": 0.02,
+      "loss": 1.627,
+      "step": 4761
+    },
+    {
+      "epoch": 0.4408493895735694,
+      "grad_norm": 0.1474609375,
+      "learning_rate": 0.02,
+      "loss": 1.6514,
+      "step": 4762
+    },
+    {
+      "epoch": 0.44094196609384945,
+      "grad_norm": 0.1435546875,
+      "learning_rate": 0.02,
+      "loss": 1.5836,
+      "step": 4763
+    },
+    {
+      "epoch": 0.4410345426141295,
+      "grad_norm": 0.1337890625,
+      "learning_rate": 0.02,
+      "loss": 1.5145,
+      "step": 4764
+    },
+    {
+      "epoch": 0.44112711913440955,
+      "grad_norm": 0.142578125,
+      "learning_rate": 0.02,
+      "loss": 1.6188,
+      "step": 4765
+    },
+    {
+      "epoch": 0.4412196956546896,
+      "grad_norm": 0.1435546875,
+      "learning_rate": 0.02,
+      "loss": 1.5971,
+      "step": 4766
+    },
+    {
+      "epoch": 0.4413122721749696,
+      "grad_norm": 0.140625,
+      "learning_rate": 0.02,
+      "loss": 1.67,
+      "step": 4767
+    },
+    {
+      "epoch": 0.44140484869524965,
+      "grad_norm": 0.1494140625,
+      "learning_rate": 0.02,
+      "loss": 1.6274,
+      "step": 4768
+    },
+    {
+      "epoch": 0.4414974252155297,
+      "grad_norm": 0.15234375,
+      "learning_rate": 0.02,
+      "loss": 1.5865,
+      "step": 4769
+    },
+    {
+      "epoch": 0.44159000173580976,
+      "grad_norm": 0.1357421875,
+      "learning_rate": 0.02,
+      "loss": 1.6749,
+      "step": 4770
+    },
+    {
+      "epoch": 0.4416825782560898,
+      "grad_norm": 0.140625,
+      "learning_rate": 0.02,
+      "loss": 1.5823,
+      "step": 4771
+    },
+    {
+      "epoch": 0.44177515477636986,
+      "grad_norm": 0.140625,
+      "learning_rate": 0.02,
+      "loss": 1.5746,
+      "step": 4772
+    },
+    {
+      "epoch": 0.4418677312966499,
+      "grad_norm": 0.5703125,
+      "learning_rate": 0.02,
+      "loss": 1.6225,
+      "step": 4773
+    },
+    {
+      "epoch": 0.4419603078169299,
+      "grad_norm": 0.1337890625,
+      "learning_rate": 0.02,
+      "loss": 1.5649,
+      "step": 4774
+    },
+    {
+      "epoch": 0.44205288433720996,
+      "grad_norm": 0.1484375,
+      "learning_rate": 0.02,
+      "loss": 1.5969,
+      "step": 4775
+    },
+    {
+      "epoch": 0.44214546085749,
+      "grad_norm": 0.1337890625,
+      "learning_rate": 0.02,
+      "loss": 1.5714,
+      "step": 4776
+    },
+    {
+      "epoch": 0.44223803737777007,
+      "grad_norm": 0.1494140625,
+      "learning_rate": 0.02,
+      "loss": 1.5154,
+      "step": 4777
+    },
+    {
+      "epoch": 0.4423306138980501,
+      "grad_norm": 0.1484375,
+      "learning_rate": 0.02,
+      "loss": 1.5613,
+      "step": 4778
+    },
+    {
+      "epoch": 0.44242319041833017,
+      "grad_norm": 0.140625,
+      "learning_rate": 0.02,
+      "loss": 1.5313,
+      "step": 4779
+    },
+    {
+      "epoch": 0.4425157669386102,
+      "grad_norm": 0.140625,
+      "learning_rate": 0.02,
+      "loss": 1.6095,
+      "step": 4780
+    },
+    {
+      "epoch": 0.4426083434588902,
+      "grad_norm": 0.138671875,
+      "learning_rate": 0.02,
+      "loss": 1.5675,
+      "step": 4781
+    },
+    {
+      "epoch": 0.44270091997917027,
+      "grad_norm": 0.1494140625,
+      "learning_rate": 0.02,
+      "loss": 1.5443,
+      "step": 4782
+    },
+    {
+      "epoch": 0.4427934964994503,
+      "grad_norm": 0.1337890625,
+      "learning_rate": 0.02,
+      "loss": 1.5985,
+      "step": 4783
+    },
+    {
+      "epoch": 0.4428860730197304,
+      "grad_norm": 0.14453125,
+      "learning_rate": 0.02,
+      "loss": 1.5337,
+      "step": 4784
+    },
+    {
+      "epoch": 0.4429786495400104,
+      "grad_norm": 0.138671875,
+      "learning_rate": 0.02,
+      "loss": 1.623,
+      "step": 4785
+    },
+    {
+      "epoch": 0.4430712260602905,
+      "grad_norm": 0.1572265625,
+      "learning_rate": 0.02,
+      "loss": 1.6207,
+      "step": 4786
+    },
+    {
+      "epoch": 0.4431638025805705,
+      "grad_norm": 0.1552734375,
+      "learning_rate": 0.02,
+      "loss": 1.6004,
+      "step": 4787
+    },
+    {
+      "epoch": 0.44325637910085053,
+      "grad_norm": 0.1611328125,
+      "learning_rate": 0.02,
+      "loss": 1.5711,
+      "step": 4788
+    },
+    {
+      "epoch": 0.4433489556211306,
+      "grad_norm": 0.1533203125,
+      "learning_rate": 0.02,
+      "loss": 1.5864,
+      "step": 4789
+    },
+    {
+      "epoch": 0.44344153214141063,
+      "grad_norm": 0.150390625,
+      "learning_rate": 0.02,
+      "loss": 1.5904,
+      "step": 4790
+    },
+    {
+      "epoch": 0.4435341086616907,
+      "grad_norm": 0.1572265625,
+      "learning_rate": 0.02,
+      "loss": 1.6189,
+      "step": 4791
+    },
+    {
+      "epoch": 0.44362668518197074,
+      "grad_norm": 0.140625,
+      "learning_rate": 0.02,
+      "loss": 1.5644,
+      "step": 4792
+    },
+    {
+      "epoch": 0.4437192617022508,
+      "grad_norm": 0.15234375,
+      "learning_rate": 0.02,
+      "loss": 1.5978,
+      "step": 4793
+    },
+    {
+      "epoch": 0.4438118382225308,
+      "grad_norm": 0.1494140625,
+      "learning_rate": 0.02,
+      "loss": 1.6133,
+      "step": 4794
+    },
+    {
+      "epoch": 0.44390441474281084,
+      "grad_norm": 0.1591796875,
+      "learning_rate": 0.02,
+      "loss": 1.6147,
+      "step": 4795
+    },
+    {
+      "epoch": 0.4439969912630909,
+      "grad_norm": 0.158203125,
+      "learning_rate": 0.02,
+      "loss": 1.5561,
+      "step": 4796
+    },
+    {
+      "epoch": 0.44408956778337094,
+      "grad_norm": 0.154296875,
+      "learning_rate": 0.02,
+      "loss": 1.5985,
+      "step": 4797
+    },
+    {
+      "epoch": 0.444182144303651,
+      "grad_norm": 0.15625,
+      "learning_rate": 0.02,
+      "loss": 1.5668,
+      "step": 4798
+    },
+    {
+      "epoch": 0.44427472082393105,
+      "grad_norm": 0.154296875,
+      "learning_rate": 0.02,
+      "loss": 1.6194,
+      "step": 4799
+    },
+    {
+      "epoch": 0.4443672973442111,
+      "grad_norm": 0.1494140625,
+      "learning_rate": 0.02,
+      "loss": 1.5684,
+      "step": 4800
+    },
+    {
+      "epoch": 0.4444598738644911,
+      "grad_norm": 0.1484375,
+      "learning_rate": 0.02,
+      "loss": 1.5336,
+      "step": 4801
+    },
+    {
+      "epoch": 0.44455245038477115,
+      "grad_norm": 0.1474609375,
+      "learning_rate": 0.02,
+      "loss": 1.619,
+      "step": 4802
+    },
+    {
+      "epoch": 0.4446450269050512,
+      "grad_norm": 0.154296875,
+      "learning_rate": 0.02,
+      "loss": 1.6381,
+      "step": 4803
+    },
+    {
+      "epoch": 0.44473760342533125,
+      "grad_norm": 0.1474609375,
+      "learning_rate": 0.02,
+      "loss": 1.5861,
+      "step": 4804
+    },
+    {
+      "epoch": 0.4448301799456113,
+      "grad_norm": 0.1396484375,
+      "learning_rate": 0.02,
+      "loss": 1.5925,
+      "step": 4805
+    },
+    {
+      "epoch": 0.44492275646589136,
+      "grad_norm": 0.158203125,
+      "learning_rate": 0.02,
+      "loss": 1.6528,
+      "step": 4806
+    },
+    {
+      "epoch": 0.4450153329861714,
+      "grad_norm": 0.146484375,
+      "learning_rate": 0.02,
+      "loss": 1.5431,
+      "step": 4807
+    },
+    {
+      "epoch": 0.4451079095064514,
+      "grad_norm": 0.142578125,
+      "learning_rate": 0.02,
+      "loss": 1.6005,
+      "step": 4808
+    },
+    {
+      "epoch": 0.44520048602673146,
+      "grad_norm": 0.1435546875,
+      "learning_rate": 0.02,
+      "loss": 1.6329,
+      "step": 4809
+    },
+    {
+      "epoch": 0.4452930625470115,
+      "grad_norm": 0.154296875,
+      "learning_rate": 0.02,
+      "loss": 1.5346,
+      "step": 4810
+    },
+    {
+      "epoch": 0.44538563906729156,
+      "grad_norm": 0.1572265625,
+      "learning_rate": 0.02,
+      "loss": 1.5869,
+      "step": 4811
+    },
+    {
+      "epoch": 0.4454782155875716,
+      "grad_norm": 0.1376953125,
+      "learning_rate": 0.02,
+      "loss": 1.5585,
+      "step": 4812
+    },
+    {
+      "epoch": 0.44557079210785167,
+      "grad_norm": 0.1474609375,
+      "learning_rate": 0.02,
+      "loss": 1.5826,
+      "step": 4813
+    },
+    {
+      "epoch": 0.4456633686281317,
+      "grad_norm": 0.1484375,
+      "learning_rate": 0.02,
+      "loss": 1.512,
+      "step": 4814
+    },
+    {
+      "epoch": 0.4457559451484117,
+      "grad_norm": 0.14453125,
+      "learning_rate": 0.02,
+      "loss": 1.5581,
+      "step": 4815
+    },
+    {
+      "epoch": 0.44584852166869177,
+      "grad_norm": 0.1552734375,
+      "learning_rate": 0.02,
+      "loss": 1.5445,
+      "step": 4816
+    },
+    {
+      "epoch": 0.4459410981889718,
+      "grad_norm": 0.1474609375,
+      "learning_rate": 0.02,
+      "loss": 1.5623,
+      "step": 4817
+    },
+    {
+      "epoch": 0.44603367470925187,
+      "grad_norm": 0.1484375,
+      "learning_rate": 0.02,
+      "loss": 1.6114,
+      "step": 4818
+    },
+    {
+      "epoch": 0.4461262512295319,
+      "grad_norm": 0.1513671875,
+      "learning_rate": 0.02,
+      "loss": 1.5814,
+      "step": 4819
+    },
+    {
+      "epoch": 0.446218827749812,
+      "grad_norm": 0.1455078125,
+      "learning_rate": 0.02,
+      "loss": 1.638,
+      "step": 4820
+    },
+    {
+      "epoch": 0.44631140427009197,
+      "grad_norm": 0.1435546875,
+      "learning_rate": 0.02,
+      "loss": 1.6134,
+      "step": 4821
+    },
+    {
+      "epoch": 0.446403980790372,
+      "grad_norm": 0.14453125,
+      "learning_rate": 0.02,
+      "loss": 1.5706,
+      "step": 4822
+    },
+    {
+      "epoch": 0.4464965573106521,
+      "grad_norm": 0.13671875,
+      "learning_rate": 0.02,
+      "loss": 1.5607,
+      "step": 4823
+    },
+    {
+      "epoch": 0.44658913383093213,
+      "grad_norm": 0.1455078125,
+      "learning_rate": 0.02,
+      "loss": 1.5669,
+      "step": 4824
+    },
+    {
+      "epoch": 0.4466817103512122,
+      "grad_norm": 0.14453125,
+      "learning_rate": 0.02,
+      "loss": 1.5869,
+      "step": 4825
+    },
+    {
+      "epoch": 0.44677428687149223,
+      "grad_norm": 0.1591796875,
+      "learning_rate": 0.02,
+      "loss": 1.6242,
+      "step": 4826
+    },
+    {
+      "epoch": 0.4468668633917723,
+      "grad_norm": 0.1513671875,
+      "learning_rate": 0.02,
+      "loss": 1.5901,
+      "step": 4827
+    },
+    {
+      "epoch": 0.4469594399120523,
+      "grad_norm": 0.154296875,
+      "learning_rate": 0.02,
+      "loss": 1.5699,
+      "step": 4828
+    },
+    {
+      "epoch": 0.44705201643233233,
+      "grad_norm": 0.1484375,
+      "learning_rate": 0.02,
+      "loss": 1.6239,
+      "step": 4829
+    },
+    {
+      "epoch": 0.4471445929526124,
+      "grad_norm": 0.158203125,
+      "learning_rate": 0.02,
+      "loss": 1.6225,
+      "step": 4830
+    },
+    {
+      "epoch": 0.44723716947289244,
+      "grad_norm": 0.1484375,
+      "learning_rate": 0.02,
+      "loss": 1.5669,
+      "step": 4831
+    },
+    {
+      "epoch": 0.4473297459931725,
+      "grad_norm": 0.15234375,
+      "learning_rate": 0.02,
+      "loss": 1.5358,
+      "step": 4832
+    },
+    {
+      "epoch": 0.44742232251345254,
+      "grad_norm": 0.1455078125,
+      "learning_rate": 0.02,
+      "loss": 1.5854,
+      "step": 4833
+    },
+    {
+      "epoch": 0.4475148990337326,
+      "grad_norm": 0.1416015625,
+      "learning_rate": 0.02,
+      "loss": 1.5541,
+      "step": 4834
+    },
+    {
+      "epoch": 0.4476074755540126,
+      "grad_norm": 0.1435546875,
+      "learning_rate": 0.02,
+      "loss": 1.4802,
+      "step": 4835
+    },
+    {
+      "epoch": 0.44770005207429264,
+      "grad_norm": 0.1435546875,
+      "learning_rate": 0.02,
+      "loss": 1.6207,
+      "step": 4836
+    },
+    {
+      "epoch": 0.4477926285945727,
+      "grad_norm": 0.1416015625,
+      "learning_rate": 0.02,
+      "loss": 1.6119,
+      "step": 4837
+    },
+    {
+      "epoch": 0.44788520511485275,
+      "grad_norm": 0.1416015625,
+      "learning_rate": 0.02,
+      "loss": 1.5677,
+      "step": 4838
+    },
+    {
+      "epoch": 0.4479777816351328,
+      "grad_norm": 0.1455078125,
+      "learning_rate": 0.02,
+      "loss": 1.624,
+      "step": 4839
+    },
+    {
+      "epoch": 0.44807035815541285,
+      "grad_norm": 0.1416015625,
+      "learning_rate": 0.02,
+      "loss": 1.6127,
+      "step": 4840
+    },
+    {
+      "epoch": 0.4481629346756929,
+      "grad_norm": 0.150390625,
+      "learning_rate": 0.02,
+      "loss": 1.5657,
+      "step": 4841
+    },
+    {
+      "epoch": 0.4482555111959729,
+      "grad_norm": 0.1435546875,
+      "learning_rate": 0.02,
+      "loss": 1.5734,
+      "step": 4842
+    },
+    {
+      "epoch": 0.44834808771625295,
+      "grad_norm": 0.146484375,
+      "learning_rate": 0.02,
+      "loss": 1.4944,
+      "step": 4843
+    },
+    {
+      "epoch": 0.448440664236533,
+      "grad_norm": 0.1435546875,
+      "learning_rate": 0.02,
+      "loss": 1.5631,
+      "step": 4844
+    },
+    {
+      "epoch": 0.44853324075681306,
+      "grad_norm": 0.1533203125,
+      "learning_rate": 0.02,
+      "loss": 1.6115,
+      "step": 4845
+    },
+    {
+      "epoch": 0.4486258172770931,
+      "grad_norm": 0.1455078125,
+      "learning_rate": 0.02,
+      "loss": 1.5945,
+      "step": 4846
+    },
+    {
+      "epoch": 0.44871839379737316,
+      "grad_norm": 0.1455078125,
+      "learning_rate": 0.02,
+      "loss": 1.5627,
+      "step": 4847
+    },
+    {
+      "epoch": 0.4488109703176532,
+      "grad_norm": 0.1357421875,
+      "learning_rate": 0.02,
+      "loss": 1.6345,
+      "step": 4848
+    },
+    {
+      "epoch": 0.4489035468379332,
+      "grad_norm": 0.15234375,
+      "learning_rate": 0.02,
+      "loss": 1.598,
+      "step": 4849
+    },
+    {
+      "epoch": 0.44899612335821326,
+      "grad_norm": 0.1435546875,
+      "learning_rate": 0.02,
+      "loss": 1.5791,
+      "step": 4850
+    },
+    {
+      "epoch": 0.4490886998784933,
+      "grad_norm": 0.1513671875,
+      "learning_rate": 0.02,
+      "loss": 1.637,
+      "step": 4851
+    },
+    {
+      "epoch": 0.44918127639877337,
+      "grad_norm": 0.1435546875,
+      "learning_rate": 0.02,
+      "loss": 1.5911,
+      "step": 4852
+    },
+    {
+      "epoch": 0.4492738529190534,
+      "grad_norm": 0.1484375,
+      "learning_rate": 0.02,
+      "loss": 1.6176,
+      "step": 4853
+    },
+    {
+      "epoch": 0.44936642943933347,
+      "grad_norm": 0.1337890625,
+      "learning_rate": 0.02,
+      "loss": 1.5884,
+      "step": 4854
+    },
+    {
+      "epoch": 0.44945900595961347,
+      "grad_norm": 0.1474609375,
+      "learning_rate": 0.02,
+      "loss": 1.5821,
+      "step": 4855
+    },
+    {
+      "epoch": 0.4495515824798935,
+      "grad_norm": 0.1435546875,
+      "learning_rate": 0.02,
+      "loss": 1.5949,
+      "step": 4856
+    },
+    {
+      "epoch": 0.44964415900017357,
+      "grad_norm": 0.1484375,
+      "learning_rate": 0.02,
+      "loss": 1.5593,
+      "step": 4857
+    },
+    {
+      "epoch": 0.4497367355204536,
+      "grad_norm": 0.1533203125,
+      "learning_rate": 0.02,
+      "loss": 1.597,
+      "step": 4858
+    },
+    {
+      "epoch": 0.4498293120407337,
+      "grad_norm": 0.16015625,
+      "learning_rate": 0.02,
+      "loss": 1.6091,
+      "step": 4859
+    },
+    {
+      "epoch": 0.4499218885610137,
+      "grad_norm": 0.1494140625,
+      "learning_rate": 0.02,
+      "loss": 1.5986,
+      "step": 4860
+    },
+    {
+      "epoch": 0.4500144650812938,
+      "grad_norm": 0.1474609375,
+      "learning_rate": 0.02,
+      "loss": 1.6156,
+      "step": 4861
+    },
+    {
+      "epoch": 0.4501070416015738,
+      "grad_norm": 0.150390625,
+      "learning_rate": 0.02,
+      "loss": 1.521,
+      "step": 4862
+    },
+    {
+      "epoch": 0.45019961812185383,
+      "grad_norm": 0.1474609375,
+      "learning_rate": 0.02,
+      "loss": 1.5272,
+      "step": 4863
+    },
+    {
+      "epoch": 0.4502921946421339,
+      "grad_norm": 0.1474609375,
+      "learning_rate": 0.02,
+      "loss": 1.5673,
+      "step": 4864
+    },
+    {
+      "epoch": 0.45038477116241393,
+      "grad_norm": 0.14453125,
+      "learning_rate": 0.02,
+      "loss": 1.5691,
+      "step": 4865
+    },
+    {
+      "epoch": 0.450477347682694,
+      "grad_norm": 0.14453125,
+      "learning_rate": 0.02,
+      "loss": 1.5539,
+      "step": 4866
+    },
+    {
+      "epoch": 0.45056992420297404,
+      "grad_norm": 0.134765625,
+      "learning_rate": 0.02,
+      "loss": 1.5198,
+      "step": 4867
+    },
+    {
+      "epoch": 0.4506625007232541,
+      "grad_norm": 0.1416015625,
+      "learning_rate": 0.02,
+      "loss": 1.5394,
+      "step": 4868
+    },
+    {
+      "epoch": 0.4507550772435341,
+      "grad_norm": 0.150390625,
+      "learning_rate": 0.02,
+      "loss": 1.5548,
+      "step": 4869
+    },
+    {
+      "epoch": 0.45084765376381414,
+      "grad_norm": 0.1513671875,
+      "learning_rate": 0.02,
+      "loss": 1.5975,
+      "step": 4870
+    },
+    {
+      "epoch": 0.4509402302840942,
+      "grad_norm": 0.140625,
+      "learning_rate": 0.02,
+      "loss": 1.5192,
+      "step": 4871
+    },
+    {
+      "epoch": 0.45103280680437424,
+      "grad_norm": 0.1455078125,
+      "learning_rate": 0.02,
+      "loss": 1.5534,
+      "step": 4872
+    },
+    {
+      "epoch": 0.4511253833246543,
+      "grad_norm": 0.1416015625,
+      "learning_rate": 0.02,
+      "loss": 1.581,
+      "step": 4873
+    },
+    {
+      "epoch": 0.45121795984493435,
+      "grad_norm": 0.146484375,
+      "learning_rate": 0.02,
+      "loss": 1.5944,
+      "step": 4874
+    },
+    {
+      "epoch": 0.4513105363652144,
+      "grad_norm": 0.140625,
+      "learning_rate": 0.02,
+      "loss": 1.6471,
+      "step": 4875
+    },
+    {
+      "epoch": 0.4514031128854944,
+      "grad_norm": 0.146484375,
+      "learning_rate": 0.02,
+      "loss": 1.5724,
+      "step": 4876
+    },
+    {
+      "epoch": 0.45149568940577445,
+      "grad_norm": 0.1376953125,
+      "learning_rate": 0.02,
+      "loss": 1.5459,
+      "step": 4877
+    },
+    {
+      "epoch": 0.4515882659260545,
+      "grad_norm": 0.140625,
+      "learning_rate": 0.02,
+      "loss": 1.5167,
+      "step": 4878
+    },
+    {
+      "epoch": 0.45168084244633455,
+      "grad_norm": 0.134765625,
+      "learning_rate": 0.02,
+      "loss": 1.6038,
+      "step": 4879
+    },
+    {
+      "epoch": 0.4517734189666146,
+      "grad_norm": 0.14453125,
+      "learning_rate": 0.02,
+      "loss": 1.5275,
+      "step": 4880
+    },
+    {
+      "epoch": 0.45186599548689466,
+      "grad_norm": 0.150390625,
+      "learning_rate": 0.02,
+      "loss": 1.6062,
+      "step": 4881
+    },
+    {
+      "epoch": 0.45195857200717465,
+      "grad_norm": 0.1484375,
+      "learning_rate": 0.02,
+      "loss": 1.6199,
+      "step": 4882
+    },
+    {
+      "epoch": 0.4520511485274547,
+      "grad_norm": 0.15625,
+      "learning_rate": 0.02,
+      "loss": 1.5715,
+      "step": 4883
+    },
+    {
+      "epoch": 0.45214372504773476,
+      "grad_norm": 0.1474609375,
+      "learning_rate": 0.02,
+      "loss": 1.5391,
+      "step": 4884
+    },
+    {
+      "epoch": 0.4522363015680148,
+      "grad_norm": 0.146484375,
+      "learning_rate": 0.02,
+      "loss": 1.678,
+      "step": 4885
+    },
+    {
+      "epoch": 0.45232887808829486,
+      "grad_norm": 0.1474609375,
+      "learning_rate": 0.02,
+      "loss": 1.5174,
+      "step": 4886
+    },
+    {
+      "epoch": 0.4524214546085749,
+      "grad_norm": 0.1513671875,
+      "learning_rate": 0.02,
+      "loss": 1.6053,
+      "step": 4887
+    },
+    {
+      "epoch": 0.45251403112885497,
+      "grad_norm": 0.1494140625,
+      "learning_rate": 0.02,
+      "loss": 1.6344,
+      "step": 4888
+    },
+    {
+      "epoch": 0.45260660764913496,
+      "grad_norm": 0.1416015625,
+      "learning_rate": 0.02,
+      "loss": 1.5213,
+      "step": 4889
+    },
+    {
+      "epoch": 0.452699184169415,
+      "grad_norm": 0.1328125,
+      "learning_rate": 0.02,
+      "loss": 1.6058,
+      "step": 4890
+    },
+    {
+      "epoch": 0.45279176068969507,
+      "grad_norm": 0.14453125,
+      "learning_rate": 0.02,
+      "loss": 1.5874,
+      "step": 4891
+    },
+    {
+      "epoch": 0.4528843372099751,
+      "grad_norm": 0.150390625,
+      "learning_rate": 0.02,
+      "loss": 1.5516,
+      "step": 4892
+    },
+    {
+      "epoch": 0.45297691373025517,
+      "grad_norm": 0.142578125,
+      "learning_rate": 0.02,
+      "loss": 1.5429,
+      "step": 4893
+    },
+    {
+      "epoch": 0.4530694902505352,
+      "grad_norm": 0.150390625,
+      "learning_rate": 0.02,
+      "loss": 1.6065,
+      "step": 4894
+    },
+    {
+      "epoch": 0.4531620667708153,
+      "grad_norm": 0.146484375,
+      "learning_rate": 0.02,
+      "loss": 1.6063,
+      "step": 4895
+    },
+    {
+      "epoch": 0.45325464329109527,
+      "grad_norm": 0.146484375,
+      "learning_rate": 0.02,
+      "loss": 1.6133,
+      "step": 4896
+    },
+    {
+      "epoch": 0.4533472198113753,
+      "grad_norm": 0.1484375,
+      "learning_rate": 0.02,
+      "loss": 1.5856,
+      "step": 4897
+    },
+    {
+      "epoch": 0.4534397963316554,
+      "grad_norm": 0.1484375,
+      "learning_rate": 0.02,
+      "loss": 1.5771,
+      "step": 4898
+    },
+    {
+      "epoch": 0.45353237285193543,
+      "grad_norm": 0.14453125,
+      "learning_rate": 0.02,
+      "loss": 1.5586,
+      "step": 4899
+    },
+    {
+      "epoch": 0.4536249493722155,
+      "grad_norm": 0.1416015625,
+      "learning_rate": 0.02,
+      "loss": 1.5948,
+      "step": 4900
+    },
+    {
+      "epoch": 0.45371752589249553,
+      "grad_norm": 0.15234375,
+      "learning_rate": 0.02,
+      "loss": 1.535,
+      "step": 4901
+    },
+    {
+      "epoch": 0.4538101024127756,
+      "grad_norm": 0.1630859375,
+      "learning_rate": 0.02,
+      "loss": 1.6473,
+      "step": 4902
+    },
+    {
+      "epoch": 0.4539026789330556,
+      "grad_norm": 0.1474609375,
+      "learning_rate": 0.02,
+      "loss": 1.5765,
+      "step": 4903
+    },
+    {
+      "epoch": 0.45399525545333563,
+      "grad_norm": 0.140625,
+      "learning_rate": 0.02,
+      "loss": 1.552,
+      "step": 4904
+    },
+    {
+      "epoch": 0.4540878319736157,
+      "grad_norm": 0.1396484375,
+      "learning_rate": 0.02,
+      "loss": 1.5337,
+      "step": 4905
+    },
+    {
+      "epoch": 0.45418040849389574,
+      "grad_norm": 0.13671875,
+      "learning_rate": 0.02,
+      "loss": 1.5063,
+      "step": 4906
+    },
+    {
+      "epoch": 0.4542729850141758,
+      "grad_norm": 0.1396484375,
+      "learning_rate": 0.02,
+      "loss": 1.5422,
+      "step": 4907
+    },
+    {
+      "epoch": 0.45436556153445584,
+      "grad_norm": 0.134765625,
+      "learning_rate": 0.02,
+      "loss": 1.5495,
+      "step": 4908
+    },
+    {
+      "epoch": 0.4544581380547359,
+      "grad_norm": 0.1435546875,
+      "learning_rate": 0.02,
+      "loss": 1.6485,
+      "step": 4909
+    },
+    {
+      "epoch": 0.4545507145750159,
+      "grad_norm": 0.1416015625,
+      "learning_rate": 0.02,
+      "loss": 1.6041,
+      "step": 4910
+    },
+    {
+      "epoch": 0.45464329109529594,
+      "grad_norm": 0.1318359375,
+      "learning_rate": 0.02,
+      "loss": 1.5321,
+      "step": 4911
+    },
+    {
+      "epoch": 0.454735867615576,
+      "grad_norm": 0.1357421875,
+      "learning_rate": 0.02,
+      "loss": 1.545,
+      "step": 4912
+    },
+    {
+      "epoch": 0.45482844413585605,
+      "grad_norm": 0.142578125,
+      "learning_rate": 0.02,
+      "loss": 1.6113,
+      "step": 4913
+    },
+    {
+      "epoch": 0.4549210206561361,
+      "grad_norm": 0.142578125,
+      "learning_rate": 0.02,
+      "loss": 1.6505,
+      "step": 4914
+    },
+    {
+      "epoch": 0.45501359717641615,
+      "grad_norm": 0.130859375,
+      "learning_rate": 0.02,
+      "loss": 1.589,
+      "step": 4915
+    },
+    {
+      "epoch": 0.45510617369669615,
+      "grad_norm": 0.130859375,
+      "learning_rate": 0.02,
+      "loss": 1.4747,
+      "step": 4916
+    },
+    {
+      "epoch": 0.4551987502169762,
+      "grad_norm": 0.1298828125,
+      "learning_rate": 0.02,
+      "loss": 1.623,
+      "step": 4917
+    },
+    {
+      "epoch": 0.45529132673725625,
+      "grad_norm": 0.1376953125,
+      "learning_rate": 0.02,
+      "loss": 1.5873,
+      "step": 4918
+    },
+    {
+      "epoch": 0.4553839032575363,
+      "grad_norm": 0.1357421875,
+      "learning_rate": 0.02,
+      "loss": 1.605,
+      "step": 4919
+    },
+    {
+      "epoch": 0.45547647977781636,
+      "grad_norm": 0.1416015625,
+      "learning_rate": 0.02,
+      "loss": 1.5821,
+      "step": 4920
+    },
+    {
+      "epoch": 0.4555690562980964,
+      "grad_norm": 0.1416015625,
+      "learning_rate": 0.02,
+      "loss": 1.6358,
+      "step": 4921
+    },
+    {
+      "epoch": 0.45566163281837646,
+      "grad_norm": 0.1396484375,
+      "learning_rate": 0.02,
+      "loss": 1.5615,
+      "step": 4922
+    },
+    {
+      "epoch": 0.45575420933865646,
+      "grad_norm": 0.1455078125,
+      "learning_rate": 0.02,
+      "loss": 1.5608,
+      "step": 4923
+    },
+    {
+      "epoch": 0.4558467858589365,
+      "grad_norm": 0.1416015625,
+      "learning_rate": 0.02,
+      "loss": 1.5909,
+      "step": 4924
+    },
+    {
+      "epoch": 0.45593936237921656,
+      "grad_norm": 0.1494140625,
+      "learning_rate": 0.02,
+      "loss": 1.6073,
+      "step": 4925
+    },
+    {
+      "epoch": 0.4560319388994966,
+      "grad_norm": 0.1435546875,
+      "learning_rate": 0.02,
+      "loss": 1.6187,
+      "step": 4926
+    },
+    {
+      "epoch": 0.45612451541977667,
+      "grad_norm": 0.150390625,
+      "learning_rate": 0.02,
+      "loss": 1.6156,
+      "step": 4927
+    },
+    {
+      "epoch": 0.4562170919400567,
+      "grad_norm": 0.1455078125,
+      "learning_rate": 0.02,
+      "loss": 1.5838,
+      "step": 4928
+    },
+    {
+      "epoch": 0.45630966846033677,
+      "grad_norm": 0.14453125,
+      "learning_rate": 0.02,
+      "loss": 1.5458,
+      "step": 4929
+    },
+    {
+      "epoch": 0.45640224498061677,
+      "grad_norm": 0.1494140625,
+      "learning_rate": 0.02,
+      "loss": 1.6099,
+      "step": 4930
+    },
+    {
+      "epoch": 0.4564948215008968,
+      "grad_norm": 0.1591796875,
+      "learning_rate": 0.02,
+      "loss": 1.5513,
+      "step": 4931
+    },
+    {
+      "epoch": 0.45658739802117687,
+      "grad_norm": 0.150390625,
+      "learning_rate": 0.02,
+      "loss": 1.654,
+      "step": 4932
+    },
+    {
+      "epoch": 0.4566799745414569,
+      "grad_norm": 0.15625,
+      "learning_rate": 0.02,
+      "loss": 1.5393,
+      "step": 4933
+    },
+    {
+      "epoch": 0.456772551061737,
+      "grad_norm": 0.1474609375,
+      "learning_rate": 0.02,
+      "loss": 1.5586,
+      "step": 4934
+    },
+    {
+      "epoch": 0.456865127582017,
+      "grad_norm": 0.1533203125,
+      "learning_rate": 0.02,
+      "loss": 1.604,
+      "step": 4935
+    },
+    {
+      "epoch": 0.4569577041022971,
+      "grad_norm": 0.1611328125,
+      "learning_rate": 0.02,
+      "loss": 1.5997,
+      "step": 4936
+    },
+    {
+      "epoch": 0.4570502806225771,
+      "grad_norm": 0.1494140625,
+      "learning_rate": 0.02,
+      "loss": 1.5936,
+      "step": 4937
+    },
+    {
+      "epoch": 0.45714285714285713,
+      "grad_norm": 0.1484375,
+      "learning_rate": 0.02,
+      "loss": 1.529,
+      "step": 4938
+    },
+    {
+      "epoch": 0.4572354336631372,
+      "grad_norm": 0.1357421875,
+      "learning_rate": 0.02,
+      "loss": 1.6017,
+      "step": 4939
+    },
+    {
+      "epoch": 0.45732801018341723,
+      "grad_norm": 0.1474609375,
+      "learning_rate": 0.02,
+      "loss": 1.5797,
+      "step": 4940
+    },
+    {
+      "epoch": 0.4574205867036973,
+      "grad_norm": 0.14453125,
+      "learning_rate": 0.02,
+      "loss": 1.6091,
+      "step": 4941
+    },
+    {
+      "epoch": 0.45751316322397734,
+      "grad_norm": 0.1484375,
+      "learning_rate": 0.02,
+      "loss": 1.5932,
+      "step": 4942
+    },
+    {
+      "epoch": 0.4576057397442574,
+      "grad_norm": 0.14453125,
+      "learning_rate": 0.02,
+      "loss": 1.5603,
+      "step": 4943
+    },
+    {
+      "epoch": 0.4576983162645374,
+      "grad_norm": 0.1474609375,
+      "learning_rate": 0.02,
+      "loss": 1.5473,
+      "step": 4944
+    },
+    {
+      "epoch": 0.45779089278481744,
+      "grad_norm": 0.134765625,
+      "learning_rate": 0.02,
+      "loss": 1.5749,
+      "step": 4945
+    },
+    {
+      "epoch": 0.4578834693050975,
+      "grad_norm": 0.1513671875,
+      "learning_rate": 0.02,
+      "loss": 1.5967,
+      "step": 4946
+    },
+    {
+      "epoch": 0.45797604582537754,
+      "grad_norm": 0.1435546875,
+      "learning_rate": 0.02,
+      "loss": 1.6556,
+      "step": 4947
+    },
+    {
+      "epoch": 0.4580686223456576,
+      "grad_norm": 0.15625,
+      "learning_rate": 0.02,
+      "loss": 1.5833,
+      "step": 4948
+    },
+    {
+      "epoch": 0.45816119886593765,
+      "grad_norm": 0.1435546875,
+      "learning_rate": 0.02,
+      "loss": 1.5878,
+      "step": 4949
+    },
+    {
+      "epoch": 0.45825377538621764,
+      "grad_norm": 0.15625,
+      "learning_rate": 0.02,
+      "loss": 1.6003,
+      "step": 4950
+    },
+    {
+      "epoch": 0.4583463519064977,
+      "grad_norm": 0.1484375,
+      "learning_rate": 0.02,
+      "loss": 1.5656,
+      "step": 4951
+    },
+    {
+      "epoch": 0.45843892842677775,
+      "grad_norm": 0.166015625,
+      "learning_rate": 0.02,
+      "loss": 1.6149,
+      "step": 4952
+    },
+    {
+      "epoch": 0.4585315049470578,
+      "grad_norm": 0.142578125,
+      "learning_rate": 0.02,
+      "loss": 1.5574,
+      "step": 4953
+    },
+    {
+      "epoch": 0.45862408146733785,
+      "grad_norm": 0.1552734375,
+      "learning_rate": 0.02,
+      "loss": 1.5644,
+      "step": 4954
+    },
+    {
+      "epoch": 0.4587166579876179,
+      "grad_norm": 0.1416015625,
+      "learning_rate": 0.02,
+      "loss": 1.5955,
+      "step": 4955
+    },
+    {
+      "epoch": 0.45880923450789796,
+      "grad_norm": 0.1435546875,
+      "learning_rate": 0.02,
+      "loss": 1.5727,
+      "step": 4956
+    },
+    {
+      "epoch": 0.45890181102817795,
+      "grad_norm": 0.15234375,
+      "learning_rate": 0.02,
+      "loss": 1.561,
+      "step": 4957
+    },
+    {
+      "epoch": 0.458994387548458,
+      "grad_norm": 0.154296875,
+      "learning_rate": 0.02,
+      "loss": 1.5827,
+      "step": 4958
+    },
+    {
+      "epoch": 0.45908696406873806,
+      "grad_norm": 0.14453125,
+      "learning_rate": 0.02,
+      "loss": 1.6292,
+      "step": 4959
+    },
+    {
+      "epoch": 0.4591795405890181,
+      "grad_norm": 0.1396484375,
+      "learning_rate": 0.02,
+      "loss": 1.5529,
+      "step": 4960
+    },
+    {
+      "epoch": 0.45927211710929816,
+      "grad_norm": 0.1416015625,
+      "learning_rate": 0.02,
+      "loss": 1.5898,
+      "step": 4961
+    },
+    {
+      "epoch": 0.4593646936295782,
+      "grad_norm": 0.138671875,
+      "learning_rate": 0.02,
+      "loss": 1.593,
+      "step": 4962
+    },
+    {
+      "epoch": 0.45945727014985827,
+      "grad_norm": 0.138671875,
+      "learning_rate": 0.02,
+      "loss": 1.5675,
+      "step": 4963
+    },
+    {
+      "epoch": 0.45954984667013826,
+      "grad_norm": 0.1435546875,
+      "learning_rate": 0.02,
+      "loss": 1.5739,
+      "step": 4964
+    },
+    {
+      "epoch": 0.4596424231904183,
+      "grad_norm": 0.1513671875,
+      "learning_rate": 0.02,
+      "loss": 1.6362,
+      "step": 4965
+    },
+    {
+      "epoch": 0.45973499971069837,
+      "grad_norm": 0.142578125,
+      "learning_rate": 0.02,
+      "loss": 1.5963,
+      "step": 4966
+    },
+    {
+      "epoch": 0.4598275762309784,
+      "grad_norm": 0.1533203125,
+      "learning_rate": 0.02,
+      "loss": 1.5341,
+      "step": 4967
+    },
+    {
+      "epoch": 0.45992015275125847,
+      "grad_norm": 0.1455078125,
+      "learning_rate": 0.02,
+      "loss": 1.5792,
+      "step": 4968
+    },
+    {
+      "epoch": 0.4600127292715385,
+      "grad_norm": 0.1513671875,
+      "learning_rate": 0.02,
+      "loss": 1.6138,
+      "step": 4969
+    },
+    {
+      "epoch": 0.4601053057918186,
+      "grad_norm": 0.142578125,
+      "learning_rate": 0.02,
+      "loss": 1.6301,
+      "step": 4970
+    },
+    {
+      "epoch": 0.46019788231209857,
+      "grad_norm": 0.158203125,
+      "learning_rate": 0.02,
+      "loss": 1.5923,
+      "step": 4971
+    },
+    {
+      "epoch": 0.4602904588323786,
+      "grad_norm": 0.158203125,
+      "learning_rate": 0.02,
+      "loss": 1.593,
+      "step": 4972
+    },
+    {
+      "epoch": 0.4603830353526587,
+      "grad_norm": 0.14453125,
+      "learning_rate": 0.02,
+      "loss": 1.5555,
+      "step": 4973
+    },
+    {
+      "epoch": 0.46047561187293873,
+      "grad_norm": 0.1474609375,
+      "learning_rate": 0.02,
+      "loss": 1.6212,
+      "step": 4974
+    },
+    {
+      "epoch": 0.4605681883932188,
+      "grad_norm": 0.1455078125,
+      "learning_rate": 0.02,
+      "loss": 1.5977,
+      "step": 4975
+    },
+    {
+      "epoch": 0.46066076491349883,
+      "grad_norm": 0.1572265625,
+      "learning_rate": 0.02,
+      "loss": 1.542,
+      "step": 4976
+    },
+    {
+      "epoch": 0.4607533414337789,
+      "grad_norm": 0.1513671875,
+      "learning_rate": 0.02,
+      "loss": 1.5552,
+      "step": 4977
+    },
+    {
+      "epoch": 0.4608459179540589,
+      "grad_norm": 0.15234375,
+      "learning_rate": 0.02,
+      "loss": 1.5959,
+      "step": 4978
+    },
+    {
+      "epoch": 0.46093849447433893,
+      "grad_norm": 0.1455078125,
+      "learning_rate": 0.02,
+      "loss": 1.6141,
+      "step": 4979
+    },
+    {
+      "epoch": 0.461031070994619,
+      "grad_norm": 0.1640625,
+      "learning_rate": 0.02,
+      "loss": 1.6296,
+      "step": 4980
+    },
+    {
+      "epoch": 0.46112364751489904,
+      "grad_norm": 0.15625,
+      "learning_rate": 0.02,
+      "loss": 1.647,
+      "step": 4981
+    },
+    {
+      "epoch": 0.4612162240351791,
+      "grad_norm": 0.154296875,
+      "learning_rate": 0.02,
+      "loss": 1.5912,
+      "step": 4982
+    },
+    {
+      "epoch": 0.46130880055545914,
+      "grad_norm": 0.14453125,
+      "learning_rate": 0.02,
+      "loss": 1.5685,
+      "step": 4983
+    },
+    {
+      "epoch": 0.46140137707573914,
+      "grad_norm": 0.1494140625,
+      "learning_rate": 0.02,
+      "loss": 1.5693,
+      "step": 4984
+    },
+    {
+      "epoch": 0.4614939535960192,
+      "grad_norm": 0.1455078125,
+      "learning_rate": 0.02,
+      "loss": 1.5598,
+      "step": 4985
+    },
+    {
+      "epoch": 0.46158653011629924,
+      "grad_norm": 0.1591796875,
+      "learning_rate": 0.02,
+      "loss": 1.5667,
+      "step": 4986
+    },
+    {
+      "epoch": 0.4616791066365793,
+      "grad_norm": 0.13671875,
+      "learning_rate": 0.02,
+      "loss": 1.5461,
+      "step": 4987
+    },
+    {
+      "epoch": 0.46177168315685935,
+      "grad_norm": 0.1435546875,
+      "learning_rate": 0.02,
+      "loss": 1.5805,
+      "step": 4988
+    },
+    {
+      "epoch": 0.4618642596771394,
+      "grad_norm": 0.140625,
+      "learning_rate": 0.02,
+      "loss": 1.6048,
+      "step": 4989
+    },
+    {
+      "epoch": 0.46195683619741945,
+      "grad_norm": 0.1494140625,
+      "learning_rate": 0.02,
+      "loss": 1.5919,
+      "step": 4990
+    },
+    {
+      "epoch": 0.46204941271769945,
+      "grad_norm": 0.146484375,
+      "learning_rate": 0.02,
+      "loss": 1.5872,
+      "step": 4991
+    },
+    {
+      "epoch": 0.4621419892379795,
+      "grad_norm": 0.150390625,
+      "learning_rate": 0.02,
+      "loss": 1.5659,
+      "step": 4992
+    },
+    {
+      "epoch": 0.46223456575825955,
+      "grad_norm": 0.1513671875,
+      "learning_rate": 0.02,
+      "loss": 1.594,
+      "step": 4993
+    },
+    {
+      "epoch": 0.4623271422785396,
+      "grad_norm": 0.1484375,
+      "learning_rate": 0.02,
+      "loss": 1.6401,
+      "step": 4994
+    },
+    {
+      "epoch": 0.46241971879881966,
+      "grad_norm": 0.1416015625,
+      "learning_rate": 0.02,
+      "loss": 1.6329,
+      "step": 4995
+    },
+    {
+      "epoch": 0.4625122953190997,
+      "grad_norm": 0.1533203125,
+      "learning_rate": 0.02,
+      "loss": 1.6339,
+      "step": 4996
+    },
+    {
+      "epoch": 0.46260487183937976,
+      "grad_norm": 0.1416015625,
+      "learning_rate": 0.02,
+      "loss": 1.5022,
+      "step": 4997
+    },
+    {
+      "epoch": 0.46269744835965976,
+      "grad_norm": 0.1494140625,
+      "learning_rate": 0.02,
+      "loss": 1.63,
+      "step": 4998
+    },
+    {
+      "epoch": 0.4627900248799398,
+      "grad_norm": 0.142578125,
+      "learning_rate": 0.02,
+      "loss": 1.537,
+      "step": 4999
+    },
+    {
+      "epoch": 0.46288260140021986,
+      "grad_norm": 0.1455078125,
+      "learning_rate": 0.02,
+      "loss": 1.6122,
+      "step": 5000
+    },
+    {
+      "epoch": 0.4629751779204999,
+      "grad_norm": 0.14453125,
+      "learning_rate": 0.02,
+      "loss": 1.6579,
+      "step": 5001
+    },
+    {
+      "epoch": 0.46306775444077997,
+      "grad_norm": 0.1455078125,
+      "learning_rate": 0.02,
+      "loss": 1.5961,
+      "step": 5002
+    },
+    {
+      "epoch": 0.46316033096106,
+      "grad_norm": 0.1474609375,
+      "learning_rate": 0.02,
+      "loss": 1.5256,
+      "step": 5003
+    },
+    {
+      "epoch": 0.46325290748134007,
+      "grad_norm": 0.1513671875,
+      "learning_rate": 0.02,
+      "loss": 1.5616,
+      "step": 5004
+    },
+    {
+      "epoch": 0.46334548400162007,
+      "grad_norm": 0.142578125,
+      "learning_rate": 0.02,
+      "loss": 1.5908,
+      "step": 5005
+    },
+    {
+      "epoch": 0.4634380605219001,
+      "grad_norm": 0.15234375,
+      "learning_rate": 0.02,
+      "loss": 1.6272,
+      "step": 5006
+    },
+    {
+      "epoch": 0.46353063704218017,
+      "grad_norm": 0.146484375,
+      "learning_rate": 0.02,
+      "loss": 1.5789,
+      "step": 5007
+    },
+    {
+      "epoch": 0.4636232135624602,
+      "grad_norm": 0.1455078125,
+      "learning_rate": 0.02,
+      "loss": 1.6178,
+      "step": 5008
+    },
+    {
+      "epoch": 0.4637157900827403,
+      "grad_norm": 0.140625,
+      "learning_rate": 0.02,
+      "loss": 1.6064,
+      "step": 5009
+    },
+    {
+      "epoch": 0.46380836660302033,
+      "grad_norm": 0.140625,
+      "learning_rate": 0.02,
+      "loss": 1.5885,
+      "step": 5010
+    },
+    {
+      "epoch": 0.4639009431233004,
+      "grad_norm": 0.1552734375,
+      "learning_rate": 0.02,
+      "loss": 1.5668,
+      "step": 5011
+    },
+    {
+      "epoch": 0.4639935196435804,
+      "grad_norm": 0.1455078125,
+      "learning_rate": 0.02,
+      "loss": 1.5139,
+      "step": 5012
+    },
+    {
+      "epoch": 0.46408609616386043,
+      "grad_norm": 0.14453125,
+      "learning_rate": 0.02,
+      "loss": 1.603,
+      "step": 5013
+    },
+    {
+      "epoch": 0.4641786726841405,
+      "grad_norm": 0.1513671875,
+      "learning_rate": 0.02,
+      "loss": 1.5632,
+      "step": 5014
+    },
+    {
+      "epoch": 0.46427124920442053,
+      "grad_norm": 0.1455078125,
+      "learning_rate": 0.02,
+      "loss": 1.5896,
+      "step": 5015
+    },
+    {
+      "epoch": 0.4643638257247006,
+      "grad_norm": 0.1376953125,
+      "learning_rate": 0.02,
+      "loss": 1.6015,
+      "step": 5016
+    },
+    {
+      "epoch": 0.46445640224498064,
+      "grad_norm": 0.1416015625,
+      "learning_rate": 0.02,
+      "loss": 1.6165,
+      "step": 5017
+    },
+    {
+      "epoch": 0.46454897876526063,
+      "grad_norm": 0.14453125,
+      "learning_rate": 0.02,
+      "loss": 1.6327,
+      "step": 5018
+    },
+    {
+      "epoch": 0.4646415552855407,
+      "grad_norm": 0.1376953125,
+      "learning_rate": 0.02,
+      "loss": 1.587,
+      "step": 5019
+    },
+    {
+      "epoch": 0.46473413180582074,
+      "grad_norm": 0.142578125,
+      "learning_rate": 0.02,
+      "loss": 1.626,
+      "step": 5020
+    },
+    {
+      "epoch": 0.4648267083261008,
+      "grad_norm": 0.1484375,
+      "learning_rate": 0.02,
+      "loss": 1.6121,
+      "step": 5021
+    },
+    {
+      "epoch": 0.46491928484638084,
+      "grad_norm": 0.142578125,
+      "learning_rate": 0.02,
+      "loss": 1.5399,
+      "step": 5022
+    },
+    {
+      "epoch": 0.4650118613666609,
+      "grad_norm": 0.1474609375,
+      "learning_rate": 0.02,
+      "loss": 1.6237,
+      "step": 5023
+    },
+    {
+      "epoch": 0.46510443788694095,
+      "grad_norm": 0.150390625,
+      "learning_rate": 0.02,
+      "loss": 1.5081,
+      "step": 5024
+    },
+    {
+      "epoch": 0.46519701440722094,
+      "grad_norm": 0.1435546875,
+      "learning_rate": 0.02,
+      "loss": 1.5853,
+      "step": 5025
+    },
+    {
+      "epoch": 0.465289590927501,
+      "grad_norm": 0.1513671875,
+      "learning_rate": 0.02,
+      "loss": 1.6273,
+      "step": 5026
+    },
+    {
+      "epoch": 0.46538216744778105,
+      "grad_norm": 0.154296875,
+      "learning_rate": 0.02,
+      "loss": 1.6433,
+      "step": 5027
+    },
+    {
+      "epoch": 0.4654747439680611,
+      "grad_norm": 0.154296875,
+      "learning_rate": 0.02,
+      "loss": 1.5992,
+      "step": 5028
+    },
+    {
+      "epoch": 0.46556732048834115,
+      "grad_norm": 0.1591796875,
+      "learning_rate": 0.02,
+      "loss": 1.5682,
+      "step": 5029
+    },
+    {
+      "epoch": 0.4656598970086212,
+      "grad_norm": 0.1533203125,
+      "learning_rate": 0.02,
+      "loss": 1.5806,
+      "step": 5030
+    },
+    {
+      "epoch": 0.46575247352890126,
+      "grad_norm": 0.150390625,
+      "learning_rate": 0.02,
+      "loss": 1.5322,
+      "step": 5031
+    },
+    {
+      "epoch": 0.46584505004918125,
+      "grad_norm": 0.1552734375,
+      "learning_rate": 0.02,
+      "loss": 1.5194,
+      "step": 5032
+    },
+    {
+      "epoch": 0.4659376265694613,
+      "grad_norm": 0.1552734375,
+      "learning_rate": 0.02,
+      "loss": 1.6416,
+      "step": 5033
+    },
+    {
+      "epoch": 0.46603020308974136,
+      "grad_norm": 0.15625,
+      "learning_rate": 0.02,
+      "loss": 1.613,
+      "step": 5034
+    },
+    {
+      "epoch": 0.4661227796100214,
+      "grad_norm": 0.1494140625,
+      "learning_rate": 0.02,
+      "loss": 1.5531,
+      "step": 5035
+    },
+    {
+      "epoch": 0.46621535613030146,
+      "grad_norm": 0.146484375,
+      "learning_rate": 0.02,
+      "loss": 1.4898,
+      "step": 5036
+    },
+    {
+      "epoch": 0.4663079326505815,
+      "grad_norm": 0.1484375,
+      "learning_rate": 0.02,
+      "loss": 1.6057,
+      "step": 5037
+    },
+    {
+      "epoch": 0.46640050917086157,
+      "grad_norm": 0.1474609375,
+      "learning_rate": 0.02,
+      "loss": 1.5574,
+      "step": 5038
+    },
+    {
+      "epoch": 0.46649308569114156,
+      "grad_norm": 0.1455078125,
+      "learning_rate": 0.02,
+      "loss": 1.6665,
+      "step": 5039
+    },
+    {
+      "epoch": 0.4665856622114216,
+      "grad_norm": 0.1533203125,
+      "learning_rate": 0.02,
+      "loss": 1.5461,
+      "step": 5040
+    },
+    {
+      "epoch": 0.46667823873170167,
+      "grad_norm": 0.150390625,
+      "learning_rate": 0.02,
+      "loss": 1.614,
+      "step": 5041
+    },
+    {
+      "epoch": 0.4667708152519817,
+      "grad_norm": 0.1494140625,
+      "learning_rate": 0.02,
+      "loss": 1.5602,
+      "step": 5042
+    },
+    {
+      "epoch": 0.46686339177226177,
+      "grad_norm": 0.162109375,
+      "learning_rate": 0.02,
+      "loss": 1.6208,
+      "step": 5043
+    },
+    {
+      "epoch": 0.4669559682925418,
+      "grad_norm": 0.15234375,
+      "learning_rate": 0.02,
+      "loss": 1.6067,
+      "step": 5044
+    },
+    {
+      "epoch": 0.4670485448128219,
+      "grad_norm": 0.16015625,
+      "learning_rate": 0.02,
+      "loss": 1.5679,
+      "step": 5045
+    },
+    {
+      "epoch": 0.46714112133310187,
+      "grad_norm": 0.14453125,
+      "learning_rate": 0.02,
+      "loss": 1.606,
+      "step": 5046
+    },
+    {
+      "epoch": 0.4672336978533819,
+      "grad_norm": 0.1513671875,
+      "learning_rate": 0.02,
+      "loss": 1.607,
+      "step": 5047
+    },
+    {
+      "epoch": 0.467326274373662,
+      "grad_norm": 0.1630859375,
+      "learning_rate": 0.02,
+      "loss": 1.5772,
+      "step": 5048
+    },
+    {
+      "epoch": 0.46741885089394203,
+      "grad_norm": 0.1630859375,
+      "learning_rate": 0.02,
+      "loss": 1.6084,
+      "step": 5049
+    },
+    {
+      "epoch": 0.4675114274142221,
+      "grad_norm": 0.1552734375,
+      "learning_rate": 0.02,
+      "loss": 1.5495,
+      "step": 5050
+    },
+    {
+      "epoch": 0.46760400393450213,
+      "grad_norm": 0.150390625,
+      "learning_rate": 0.02,
+      "loss": 1.5244,
+      "step": 5051
+    },
+    {
+      "epoch": 0.46769658045478213,
+      "grad_norm": 0.14453125,
+      "learning_rate": 0.02,
+      "loss": 1.5103,
+      "step": 5052
+    },
+    {
+      "epoch": 0.4677891569750622,
+      "grad_norm": 0.1552734375,
+      "learning_rate": 0.02,
+      "loss": 1.661,
+      "step": 5053
+    },
+    {
+      "epoch": 0.46788173349534223,
+      "grad_norm": 0.1455078125,
+      "learning_rate": 0.02,
+      "loss": 1.5923,
+      "step": 5054
+    },
+    {
+      "epoch": 0.4679743100156223,
+      "grad_norm": 0.1494140625,
+      "learning_rate": 0.02,
+      "loss": 1.5992,
+      "step": 5055
+    },
+    {
+      "epoch": 0.46806688653590234,
+      "grad_norm": 0.154296875,
+      "learning_rate": 0.02,
+      "loss": 1.5888,
+      "step": 5056
+    },
+    {
+      "epoch": 0.4681594630561824,
+      "grad_norm": 0.154296875,
+      "learning_rate": 0.02,
+      "loss": 1.6233,
+      "step": 5057
+    },
+    {
+      "epoch": 0.46825203957646244,
+      "grad_norm": 0.15234375,
+      "learning_rate": 0.02,
+      "loss": 1.5452,
+      "step": 5058
+    },
+    {
+      "epoch": 0.46834461609674244,
+      "grad_norm": 0.1611328125,
+      "learning_rate": 0.02,
+      "loss": 1.6027,
+      "step": 5059
+    },
+    {
+      "epoch": 0.4684371926170225,
+      "grad_norm": 0.158203125,
+      "learning_rate": 0.02,
+      "loss": 1.6228,
+      "step": 5060
+    },
+    {
+      "epoch": 0.46852976913730254,
+      "grad_norm": 0.142578125,
+      "learning_rate": 0.02,
+      "loss": 1.519,
+      "step": 5061
+    },
+    {
+      "epoch": 0.4686223456575826,
+      "grad_norm": 0.146484375,
+      "learning_rate": 0.02,
+      "loss": 1.5966,
+      "step": 5062
+    },
+    {
+      "epoch": 0.46871492217786265,
+      "grad_norm": 0.1474609375,
+      "learning_rate": 0.02,
+      "loss": 1.5847,
+      "step": 5063
+    },
+    {
+      "epoch": 0.4688074986981427,
+      "grad_norm": 0.1435546875,
+      "learning_rate": 0.02,
+      "loss": 1.6143,
+      "step": 5064
+    },
+    {
+      "epoch": 0.46890007521842275,
+      "grad_norm": 0.142578125,
+      "learning_rate": 0.02,
+      "loss": 1.5503,
+      "step": 5065
+    },
+    {
+      "epoch": 0.46899265173870275,
+      "grad_norm": 0.1494140625,
+      "learning_rate": 0.02,
+      "loss": 1.5431,
+      "step": 5066
+    },
+    {
+      "epoch": 0.4690852282589828,
+      "grad_norm": 0.1435546875,
+      "learning_rate": 0.02,
+      "loss": 1.579,
+      "step": 5067
+    },
+    {
+      "epoch": 0.46917780477926285,
+      "grad_norm": 0.146484375,
+      "learning_rate": 0.02,
+      "loss": 1.5345,
+      "step": 5068
+    },
+    {
+      "epoch": 0.4692703812995429,
+      "grad_norm": 0.14453125,
+      "learning_rate": 0.02,
+      "loss": 1.561,
+      "step": 5069
+    },
+    {
+      "epoch": 0.46936295781982296,
+      "grad_norm": 0.1630859375,
+      "learning_rate": 0.02,
+      "loss": 1.613,
+      "step": 5070
+    },
+    {
+      "epoch": 0.469455534340103,
+      "grad_norm": 0.154296875,
+      "learning_rate": 0.02,
+      "loss": 1.5814,
+      "step": 5071
+    },
+    {
+      "epoch": 0.46954811086038306,
+      "grad_norm": 0.15234375,
+      "learning_rate": 0.02,
+      "loss": 1.5992,
+      "step": 5072
+    },
+    {
+      "epoch": 0.46964068738066306,
+      "grad_norm": 0.150390625,
+      "learning_rate": 0.02,
+      "loss": 1.5806,
+      "step": 5073
+    },
+    {
+      "epoch": 0.4697332639009431,
+      "grad_norm": 0.1455078125,
+      "learning_rate": 0.02,
+      "loss": 1.5307,
+      "step": 5074
+    },
+    {
+      "epoch": 0.46982584042122316,
+      "grad_norm": 0.140625,
+      "learning_rate": 0.02,
+      "loss": 1.597,
+      "step": 5075
+    },
+    {
+      "epoch": 0.4699184169415032,
+      "grad_norm": 0.1474609375,
+      "learning_rate": 0.02,
+      "loss": 1.5508,
+      "step": 5076
+    },
+    {
+      "epoch": 0.47001099346178327,
+      "grad_norm": 0.146484375,
+      "learning_rate": 0.02,
+      "loss": 1.5679,
+      "step": 5077
+    },
+    {
+      "epoch": 0.4701035699820633,
+      "grad_norm": 0.1572265625,
+      "learning_rate": 0.02,
+      "loss": 1.64,
+      "step": 5078
+    },
+    {
+      "epoch": 0.47019614650234337,
+      "grad_norm": 0.140625,
+      "learning_rate": 0.02,
+      "loss": 1.556,
+      "step": 5079
+    },
+    {
+      "epoch": 0.47028872302262337,
+      "grad_norm": 0.154296875,
+      "learning_rate": 0.02,
+      "loss": 1.6323,
+      "step": 5080
+    },
+    {
+      "epoch": 0.4703812995429034,
+      "grad_norm": 0.15625,
+      "learning_rate": 0.02,
+      "loss": 1.6241,
+      "step": 5081
+    },
+    {
+      "epoch": 0.47047387606318347,
+      "grad_norm": 0.1650390625,
+      "learning_rate": 0.02,
+      "loss": 1.6182,
+      "step": 5082
+    },
+    {
+      "epoch": 0.4705664525834635,
+      "grad_norm": 0.150390625,
+      "learning_rate": 0.02,
+      "loss": 1.5931,
+      "step": 5083
+    },
+    {
+      "epoch": 0.4706590291037436,
+      "grad_norm": 0.1552734375,
+      "learning_rate": 0.02,
+      "loss": 1.5805,
+      "step": 5084
+    },
+    {
+      "epoch": 0.47075160562402363,
+      "grad_norm": 0.1474609375,
+      "learning_rate": 0.02,
+      "loss": 1.633,
+      "step": 5085
+    },
+    {
+      "epoch": 0.4708441821443036,
+      "grad_norm": 0.171875,
+      "learning_rate": 0.02,
+      "loss": 1.5754,
+      "step": 5086
+    },
+    {
+      "epoch": 0.4709367586645837,
+      "grad_norm": 0.1474609375,
+      "learning_rate": 0.02,
+      "loss": 1.5537,
+      "step": 5087
+    },
+    {
+      "epoch": 0.47102933518486373,
+      "grad_norm": 0.1552734375,
+      "learning_rate": 0.02,
+      "loss": 1.5757,
+      "step": 5088
+    },
+    {
+      "epoch": 0.4711219117051438,
+      "grad_norm": 0.1513671875,
+      "learning_rate": 0.02,
+      "loss": 1.6048,
+      "step": 5089
+    },
+    {
+      "epoch": 0.47121448822542383,
+      "grad_norm": 0.1396484375,
+      "learning_rate": 0.02,
+      "loss": 1.5495,
+      "step": 5090
+    },
+    {
+      "epoch": 0.4713070647457039,
+      "grad_norm": 0.1494140625,
+      "learning_rate": 0.02,
+      "loss": 1.6546,
+      "step": 5091
+    },
+    {
+      "epoch": 0.47139964126598394,
+      "grad_norm": 0.1376953125,
+      "learning_rate": 0.02,
+      "loss": 1.5512,
+      "step": 5092
+    },
+    {
+      "epoch": 0.47149221778626393,
+      "grad_norm": 0.15234375,
+      "learning_rate": 0.02,
+      "loss": 1.5345,
+      "step": 5093
+    },
+    {
+      "epoch": 0.471584794306544,
+      "grad_norm": 0.1455078125,
+      "learning_rate": 0.02,
+      "loss": 1.586,
+      "step": 5094
+    },
+    {
+      "epoch": 0.47167737082682404,
+      "grad_norm": 0.1435546875,
+      "learning_rate": 0.02,
+      "loss": 1.5092,
+      "step": 5095
+    },
+    {
+      "epoch": 0.4717699473471041,
+      "grad_norm": 0.1298828125,
+      "learning_rate": 0.02,
+      "loss": 1.5619,
+      "step": 5096
+    },
+    {
+      "epoch": 0.47186252386738414,
+      "grad_norm": 0.134765625,
+      "learning_rate": 0.02,
+      "loss": 1.53,
+      "step": 5097
+    },
+    {
+      "epoch": 0.4719551003876642,
+      "grad_norm": 0.1357421875,
+      "learning_rate": 0.02,
+      "loss": 1.5472,
+      "step": 5098
+    },
+    {
+      "epoch": 0.47204767690794425,
+      "grad_norm": 0.1513671875,
+      "learning_rate": 0.02,
+      "loss": 1.5747,
+      "step": 5099
+    },
+    {
+      "epoch": 0.47214025342822424,
+      "grad_norm": 0.1396484375,
+      "learning_rate": 0.02,
+      "loss": 1.5923,
+      "step": 5100
+    },
+    {
+      "epoch": 0.4722328299485043,
+      "grad_norm": 0.1474609375,
+      "learning_rate": 0.02,
+      "loss": 1.5399,
+      "step": 5101
+    },
+    {
+      "epoch": 0.47232540646878435,
+      "grad_norm": 0.1513671875,
+      "learning_rate": 0.02,
+      "loss": 1.506,
+      "step": 5102
+    },
+    {
+      "epoch": 0.4724179829890644,
+      "grad_norm": 0.1533203125,
+      "learning_rate": 0.02,
+      "loss": 1.565,
+      "step": 5103
+    },
+    {
+      "epoch": 0.47251055950934445,
+      "grad_norm": 0.150390625,
+      "learning_rate": 0.02,
+      "loss": 1.5535,
+      "step": 5104
+    },
+    {
+      "epoch": 0.4726031360296245,
+      "grad_norm": 0.158203125,
+      "learning_rate": 0.02,
+      "loss": 1.5861,
+      "step": 5105
+    },
+    {
+      "epoch": 0.47269571254990456,
+      "grad_norm": 0.162109375,
+      "learning_rate": 0.02,
+      "loss": 1.6176,
+      "step": 5106
+    },
+    {
+      "epoch": 0.47278828907018455,
+      "grad_norm": 0.1630859375,
+      "learning_rate": 0.02,
+      "loss": 1.6228,
+      "step": 5107
+    },
+    {
+      "epoch": 0.4728808655904646,
+      "grad_norm": 0.162109375,
+      "learning_rate": 0.02,
+      "loss": 1.5279,
+      "step": 5108
+    },
+    {
+      "epoch": 0.47297344211074466,
+      "grad_norm": 0.1416015625,
+      "learning_rate": 0.02,
+      "loss": 1.5505,
+      "step": 5109
+    },
+    {
+      "epoch": 0.4730660186310247,
+      "grad_norm": 0.1552734375,
+      "learning_rate": 0.02,
+      "loss": 1.5624,
+      "step": 5110
+    },
+    {
+      "epoch": 0.47315859515130476,
+      "grad_norm": 0.1572265625,
+      "learning_rate": 0.02,
+      "loss": 1.6392,
+      "step": 5111
+    },
+    {
+      "epoch": 0.4732511716715848,
+      "grad_norm": 0.1533203125,
+      "learning_rate": 0.02,
+      "loss": 1.5223,
+      "step": 5112
+    },
+    {
+      "epoch": 0.47334374819186487,
+      "grad_norm": 0.1435546875,
+      "learning_rate": 0.02,
+      "loss": 1.5969,
+      "step": 5113
+    },
+    {
+      "epoch": 0.47343632471214486,
+      "grad_norm": 0.1572265625,
+      "learning_rate": 0.02,
+      "loss": 1.5696,
+      "step": 5114
+    },
+    {
+      "epoch": 0.4735289012324249,
+      "grad_norm": 0.1572265625,
+      "learning_rate": 0.02,
+      "loss": 1.5923,
+      "step": 5115
+    },
+    {
+      "epoch": 0.47362147775270497,
+      "grad_norm": 0.15234375,
+      "learning_rate": 0.02,
+      "loss": 1.5744,
+      "step": 5116
+    },
+    {
+      "epoch": 0.473714054272985,
+      "grad_norm": 0.138671875,
+      "learning_rate": 0.02,
+      "loss": 1.5612,
+      "step": 5117
+    },
+    {
+      "epoch": 0.47380663079326507,
+      "grad_norm": 0.15234375,
+      "learning_rate": 0.02,
+      "loss": 1.5646,
+      "step": 5118
+    },
+    {
+      "epoch": 0.4738992073135451,
+      "grad_norm": 0.1455078125,
+      "learning_rate": 0.02,
+      "loss": 1.6037,
+      "step": 5119
+    },
+    {
+      "epoch": 0.4739917838338251,
+      "grad_norm": 0.138671875,
+      "learning_rate": 0.02,
+      "loss": 1.5591,
+      "step": 5120
+    },
+    {
+      "epoch": 0.47408436035410517,
+      "grad_norm": 0.138671875,
+      "learning_rate": 0.02,
+      "loss": 1.5739,
+      "step": 5121
+    },
+    {
+      "epoch": 0.4741769368743852,
+      "grad_norm": 0.1416015625,
+      "learning_rate": 0.02,
+      "loss": 1.6316,
+      "step": 5122
+    },
+    {
+      "epoch": 0.4742695133946653,
+      "grad_norm": 0.12890625,
+      "learning_rate": 0.02,
+      "loss": 1.4958,
+      "step": 5123
+    },
+    {
+      "epoch": 0.47436208991494533,
+      "grad_norm": 0.142578125,
+      "learning_rate": 0.02,
+      "loss": 1.5957,
+      "step": 5124
+    },
+    {
+      "epoch": 0.4744546664352254,
+      "grad_norm": 0.1435546875,
+      "learning_rate": 0.02,
+      "loss": 1.5495,
+      "step": 5125
+    },
+    {
+      "epoch": 0.47454724295550543,
+      "grad_norm": 0.130859375,
+      "learning_rate": 0.02,
+      "loss": 1.5666,
+      "step": 5126
+    },
+    {
+      "epoch": 0.47463981947578543,
+      "grad_norm": 0.14453125,
+      "learning_rate": 0.02,
+      "loss": 1.598,
+      "step": 5127
+    },
+    {
+      "epoch": 0.4747323959960655,
+      "grad_norm": 0.134765625,
+      "learning_rate": 0.02,
+      "loss": 1.6183,
+      "step": 5128
+    },
+    {
+      "epoch": 0.47482497251634553,
+      "grad_norm": 0.1416015625,
+      "learning_rate": 0.02,
+      "loss": 1.6022,
+      "step": 5129
+    },
+    {
+      "epoch": 0.4749175490366256,
+      "grad_norm": 0.1396484375,
+      "learning_rate": 0.02,
+      "loss": 1.5978,
+      "step": 5130
+    },
+    {
+      "epoch": 0.47501012555690564,
+      "grad_norm": 0.1435546875,
+      "learning_rate": 0.02,
+      "loss": 1.5331,
+      "step": 5131
+    },
+    {
+      "epoch": 0.4751027020771857,
+      "grad_norm": 0.1494140625,
+      "learning_rate": 0.02,
+      "loss": 1.6638,
+      "step": 5132
+    },
+    {
+      "epoch": 0.47519527859746574,
+      "grad_norm": 0.1455078125,
+      "learning_rate": 0.02,
+      "loss": 1.5853,
+      "step": 5133
+    },
+    {
+      "epoch": 0.47528785511774574,
+      "grad_norm": 0.1474609375,
+      "learning_rate": 0.02,
+      "loss": 1.5829,
+      "step": 5134
+    },
+    {
+      "epoch": 0.4753804316380258,
+      "grad_norm": 0.1494140625,
+      "learning_rate": 0.02,
+      "loss": 1.6233,
+      "step": 5135
+    },
+    {
+      "epoch": 0.47547300815830584,
+      "grad_norm": 0.16015625,
+      "learning_rate": 0.02,
+      "loss": 1.5859,
+      "step": 5136
+    },
+    {
+      "epoch": 0.4755655846785859,
+      "grad_norm": 0.1572265625,
+      "learning_rate": 0.02,
+      "loss": 1.5846,
+      "step": 5137
+    },
+    {
+      "epoch": 0.47565816119886595,
+      "grad_norm": 0.1552734375,
+      "learning_rate": 0.02,
+      "loss": 1.6065,
+      "step": 5138
+    },
+    {
+      "epoch": 0.475750737719146,
+      "grad_norm": 0.142578125,
+      "learning_rate": 0.02,
+      "loss": 1.6059,
+      "step": 5139
+    },
+    {
+      "epoch": 0.47584331423942605,
+      "grad_norm": 0.1474609375,
+      "learning_rate": 0.02,
+      "loss": 1.6168,
+      "step": 5140
+    },
+    {
+      "epoch": 0.47593589075970605,
+      "grad_norm": 0.1552734375,
+      "learning_rate": 0.02,
+      "loss": 1.5622,
+      "step": 5141
+    },
+    {
+      "epoch": 0.4760284672799861,
+      "grad_norm": 0.14453125,
+      "learning_rate": 0.02,
+      "loss": 1.5803,
+      "step": 5142
+    },
+    {
+      "epoch": 0.47612104380026615,
+      "grad_norm": 0.1572265625,
+      "learning_rate": 0.02,
+      "loss": 1.6031,
+      "step": 5143
+    },
+    {
+      "epoch": 0.4762136203205462,
+      "grad_norm": 0.1494140625,
+      "learning_rate": 0.02,
+      "loss": 1.5645,
+      "step": 5144
+    },
+    {
+      "epoch": 0.47630619684082626,
+      "grad_norm": 0.1533203125,
+      "learning_rate": 0.02,
+      "loss": 1.6759,
+      "step": 5145
+    },
+    {
+      "epoch": 0.4763987733611063,
+      "grad_norm": 0.142578125,
+      "learning_rate": 0.02,
+      "loss": 1.5746,
+      "step": 5146
+    },
+    {
+      "epoch": 0.4764913498813863,
+      "grad_norm": 0.1494140625,
+      "learning_rate": 0.02,
+      "loss": 1.537,
+      "step": 5147
+    },
+    {
+      "epoch": 0.47658392640166636,
+      "grad_norm": 0.1455078125,
+      "learning_rate": 0.02,
+      "loss": 1.5538,
+      "step": 5148
+    },
+    {
+      "epoch": 0.4766765029219464,
+      "grad_norm": 0.158203125,
+      "learning_rate": 0.02,
+      "loss": 1.6009,
+      "step": 5149
+    },
+    {
+      "epoch": 0.47676907944222646,
+      "grad_norm": 0.158203125,
+      "learning_rate": 0.02,
+      "loss": 1.57,
+      "step": 5150
+    },
+    {
+      "epoch": 0.4768616559625065,
+      "grad_norm": 0.16015625,
+      "learning_rate": 0.02,
+      "loss": 1.5735,
+      "step": 5151
+    },
+    {
+      "epoch": 0.47695423248278657,
+      "grad_norm": 0.15234375,
+      "learning_rate": 0.02,
+      "loss": 1.5598,
+      "step": 5152
+    },
+    {
+      "epoch": 0.4770468090030666,
+      "grad_norm": 0.1484375,
+      "learning_rate": 0.02,
+      "loss": 1.598,
+      "step": 5153
+    },
+    {
+      "epoch": 0.4771393855233466,
+      "grad_norm": 0.146484375,
+      "learning_rate": 0.02,
+      "loss": 1.5372,
+      "step": 5154
+    },
+    {
+      "epoch": 0.47723196204362667,
+      "grad_norm": 0.146484375,
+      "learning_rate": 0.02,
+      "loss": 1.5614,
+      "step": 5155
+    },
+    {
+      "epoch": 0.4773245385639067,
+      "grad_norm": 0.1494140625,
+      "learning_rate": 0.02,
+      "loss": 1.5483,
+      "step": 5156
+    },
+    {
+      "epoch": 0.47741711508418677,
+      "grad_norm": 0.1474609375,
+      "learning_rate": 0.02,
+      "loss": 1.5707,
+      "step": 5157
+    },
+    {
+      "epoch": 0.4775096916044668,
+      "grad_norm": 0.142578125,
+      "learning_rate": 0.02,
+      "loss": 1.5155,
+      "step": 5158
+    },
+    {
+      "epoch": 0.4776022681247469,
+      "grad_norm": 0.1416015625,
+      "learning_rate": 0.02,
+      "loss": 1.5481,
+      "step": 5159
+    },
+    {
+      "epoch": 0.47769484464502693,
+      "grad_norm": 0.1474609375,
+      "learning_rate": 0.02,
+      "loss": 1.631,
+      "step": 5160
+    },
+    {
+      "epoch": 0.4777874211653069,
+      "grad_norm": 0.138671875,
+      "learning_rate": 0.02,
+      "loss": 1.5817,
+      "step": 5161
+    },
+    {
+      "epoch": 0.477879997685587,
+      "grad_norm": 0.13671875,
+      "learning_rate": 0.02,
+      "loss": 1.5436,
+      "step": 5162
+    },
+    {
+      "epoch": 0.47797257420586703,
+      "grad_norm": 0.13671875,
+      "learning_rate": 0.02,
+      "loss": 1.5785,
+      "step": 5163
+    },
+    {
+      "epoch": 0.4780651507261471,
+      "grad_norm": 0.142578125,
+      "learning_rate": 0.02,
+      "loss": 1.5564,
+      "step": 5164
+    },
+    {
+      "epoch": 0.47815772724642713,
+      "grad_norm": 0.13671875,
+      "learning_rate": 0.02,
+      "loss": 1.6239,
+      "step": 5165
+    },
+    {
+      "epoch": 0.4782503037667072,
+      "grad_norm": 0.134765625,
+      "learning_rate": 0.02,
+      "loss": 1.5507,
+      "step": 5166
+    },
+    {
+      "epoch": 0.47834288028698724,
+      "grad_norm": 0.138671875,
+      "learning_rate": 0.02,
+      "loss": 1.629,
+      "step": 5167
+    },
+    {
+      "epoch": 0.47843545680726723,
+      "grad_norm": 0.1416015625,
+      "learning_rate": 0.02,
+      "loss": 1.5894,
+      "step": 5168
+    },
+    {
+      "epoch": 0.4785280333275473,
+      "grad_norm": 0.1416015625,
+      "learning_rate": 0.02,
+      "loss": 1.5949,
+      "step": 5169
+    },
+    {
+      "epoch": 0.47862060984782734,
+      "grad_norm": 0.1396484375,
+      "learning_rate": 0.02,
+      "loss": 1.5474,
+      "step": 5170
+    },
+    {
+      "epoch": 0.4787131863681074,
+      "grad_norm": 0.140625,
+      "learning_rate": 0.02,
+      "loss": 1.6076,
+      "step": 5171
+    },
+    {
+      "epoch": 0.47880576288838744,
+      "grad_norm": 0.154296875,
+      "learning_rate": 0.02,
+      "loss": 1.5487,
+      "step": 5172
+    },
+    {
+      "epoch": 0.4788983394086675,
+      "grad_norm": 0.1455078125,
+      "learning_rate": 0.02,
+      "loss": 1.6559,
+      "step": 5173
+    },
+    {
+      "epoch": 0.47899091592894755,
+      "grad_norm": 0.1435546875,
+      "learning_rate": 0.02,
+      "loss": 1.613,
+      "step": 5174
+    },
+    {
+      "epoch": 0.47908349244922754,
+      "grad_norm": 0.154296875,
+      "learning_rate": 0.02,
+      "loss": 1.5819,
+      "step": 5175
+    },
+    {
+      "epoch": 0.4791760689695076,
+      "grad_norm": 0.14453125,
+      "learning_rate": 0.02,
+      "loss": 1.5748,
+      "step": 5176
+    },
+    {
+      "epoch": 0.47926864548978765,
+      "grad_norm": 0.1455078125,
+      "learning_rate": 0.02,
+      "loss": 1.5541,
+      "step": 5177
+    },
+    {
+      "epoch": 0.4793612220100677,
+      "grad_norm": 0.138671875,
+      "learning_rate": 0.02,
+      "loss": 1.588,
+      "step": 5178
+    },
+    {
+      "epoch": 0.47945379853034775,
+      "grad_norm": 0.1513671875,
+      "learning_rate": 0.02,
+      "loss": 1.5873,
+      "step": 5179
+    },
+    {
+      "epoch": 0.4795463750506278,
+      "grad_norm": 0.1494140625,
+      "learning_rate": 0.02,
+      "loss": 1.5254,
+      "step": 5180
+    },
+    {
+      "epoch": 0.4796389515709078,
+      "grad_norm": 0.15234375,
+      "learning_rate": 0.02,
+      "loss": 1.5601,
+      "step": 5181
+    },
+    {
+      "epoch": 0.47973152809118785,
+      "grad_norm": 0.1455078125,
+      "learning_rate": 0.02,
+      "loss": 1.6135,
+      "step": 5182
+    },
+    {
+      "epoch": 0.4798241046114679,
+      "grad_norm": 0.150390625,
+      "learning_rate": 0.02,
+      "loss": 1.5853,
+      "step": 5183
+    },
+    {
+      "epoch": 0.47991668113174796,
+      "grad_norm": 0.150390625,
+      "learning_rate": 0.02,
+      "loss": 1.6079,
+      "step": 5184
+    },
+    {
+      "epoch": 0.480009257652028,
+      "grad_norm": 0.1494140625,
+      "learning_rate": 0.02,
+      "loss": 1.603,
+      "step": 5185
+    },
+    {
+      "epoch": 0.48010183417230806,
+      "grad_norm": 0.146484375,
+      "learning_rate": 0.02,
+      "loss": 1.6292,
+      "step": 5186
+    },
+    {
+      "epoch": 0.4801944106925881,
+      "grad_norm": 0.1396484375,
+      "learning_rate": 0.02,
+      "loss": 1.4767,
+      "step": 5187
+    },
+    {
+      "epoch": 0.4802869872128681,
+      "grad_norm": 0.1357421875,
+      "learning_rate": 0.02,
+      "loss": 1.5956,
+      "step": 5188
+    },
+    {
+      "epoch": 0.48037956373314816,
+      "grad_norm": 0.146484375,
+      "learning_rate": 0.02,
+      "loss": 1.5793,
+      "step": 5189
+    },
+    {
+      "epoch": 0.4804721402534282,
+      "grad_norm": 0.140625,
+      "learning_rate": 0.02,
+      "loss": 1.6129,
+      "step": 5190
+    },
+    {
+      "epoch": 0.48056471677370827,
+      "grad_norm": 0.1474609375,
+      "learning_rate": 0.02,
+      "loss": 1.5421,
+      "step": 5191
+    },
+    {
+      "epoch": 0.4806572932939883,
+      "grad_norm": 0.146484375,
+      "learning_rate": 0.02,
+      "loss": 1.5508,
+      "step": 5192
+    },
+    {
+      "epoch": 0.48074986981426837,
+      "grad_norm": 0.1533203125,
+      "learning_rate": 0.02,
+      "loss": 1.5944,
+      "step": 5193
+    },
+    {
+      "epoch": 0.4808424463345484,
+      "grad_norm": 0.150390625,
+      "learning_rate": 0.02,
+      "loss": 1.5731,
+      "step": 5194
+    },
+    {
+      "epoch": 0.4809350228548284,
+      "grad_norm": 0.1533203125,
+      "learning_rate": 0.02,
+      "loss": 1.5523,
+      "step": 5195
+    },
+    {
+      "epoch": 0.48102759937510847,
+      "grad_norm": 0.1474609375,
+      "learning_rate": 0.02,
+      "loss": 1.5948,
+      "step": 5196
+    },
+    {
+      "epoch": 0.4811201758953885,
+      "grad_norm": 0.1494140625,
+      "learning_rate": 0.02,
+      "loss": 1.5689,
+      "step": 5197
+    },
+    {
+      "epoch": 0.4812127524156686,
+      "grad_norm": 0.1552734375,
+      "learning_rate": 0.02,
+      "loss": 1.5545,
+      "step": 5198
+    },
+    {
+      "epoch": 0.48130532893594863,
+      "grad_norm": 0.1435546875,
+      "learning_rate": 0.02,
+      "loss": 1.6645,
+      "step": 5199
+    },
+    {
+      "epoch": 0.4813979054562287,
+      "grad_norm": 0.146484375,
+      "learning_rate": 0.02,
+      "loss": 1.5953,
+      "step": 5200
+    },
+    {
+      "epoch": 0.48149048197650873,
+      "grad_norm": 0.1396484375,
+      "learning_rate": 0.02,
+      "loss": 1.5307,
+      "step": 5201
+    },
+    {
+      "epoch": 0.48158305849678873,
+      "grad_norm": 0.1484375,
+      "learning_rate": 0.02,
+      "loss": 1.5539,
+      "step": 5202
+    },
+    {
+      "epoch": 0.4816756350170688,
+      "grad_norm": 0.1494140625,
+      "learning_rate": 0.02,
+      "loss": 1.5936,
+      "step": 5203
+    },
+    {
+      "epoch": 0.48176821153734883,
+      "grad_norm": 0.15234375,
+      "learning_rate": 0.02,
+      "loss": 1.5998,
+      "step": 5204
+    },
+    {
+      "epoch": 0.4818607880576289,
+      "grad_norm": 0.14453125,
+      "learning_rate": 0.02,
+      "loss": 1.5581,
+      "step": 5205
+    },
+    {
+      "epoch": 0.48195336457790894,
+      "grad_norm": 0.146484375,
+      "learning_rate": 0.02,
+      "loss": 1.5773,
+      "step": 5206
+    },
+    {
+      "epoch": 0.482045941098189,
+      "grad_norm": 0.142578125,
+      "learning_rate": 0.02,
+      "loss": 1.6036,
+      "step": 5207
+    },
+    {
+      "epoch": 0.48213851761846904,
+      "grad_norm": 0.1552734375,
+      "learning_rate": 0.02,
+      "loss": 1.5793,
+      "step": 5208
+    },
+    {
+      "epoch": 0.48223109413874904,
+      "grad_norm": 0.1455078125,
+      "learning_rate": 0.02,
+      "loss": 1.6449,
+      "step": 5209
+    },
+    {
+      "epoch": 0.4823236706590291,
+      "grad_norm": 0.142578125,
+      "learning_rate": 0.02,
+      "loss": 1.5928,
+      "step": 5210
+    },
+    {
+      "epoch": 0.48241624717930914,
+      "grad_norm": 0.1484375,
+      "learning_rate": 0.02,
+      "loss": 1.5432,
+      "step": 5211
+    },
+    {
+      "epoch": 0.4825088236995892,
+      "grad_norm": 0.142578125,
+      "learning_rate": 0.02,
+      "loss": 1.6019,
+      "step": 5212
+    },
+    {
+      "epoch": 0.48260140021986925,
+      "grad_norm": 0.1533203125,
+      "learning_rate": 0.02,
+      "loss": 1.5829,
+      "step": 5213
+    },
+    {
+      "epoch": 0.4826939767401493,
+      "grad_norm": 0.146484375,
+      "learning_rate": 0.02,
+      "loss": 1.5677,
+      "step": 5214
+    },
+    {
+      "epoch": 0.4827865532604293,
+      "grad_norm": 0.142578125,
+      "learning_rate": 0.02,
+      "loss": 1.6137,
+      "step": 5215
+    },
+    {
+      "epoch": 0.48287912978070935,
+      "grad_norm": 0.1474609375,
+      "learning_rate": 0.02,
+      "loss": 1.5235,
+      "step": 5216
+    },
+    {
+      "epoch": 0.4829717063009894,
+      "grad_norm": 0.1591796875,
+      "learning_rate": 0.02,
+      "loss": 1.5868,
+      "step": 5217
+    },
+    {
+      "epoch": 0.48306428282126945,
+      "grad_norm": 0.146484375,
+      "learning_rate": 0.02,
+      "loss": 1.5665,
+      "step": 5218
+    },
+    {
+      "epoch": 0.4831568593415495,
+      "grad_norm": 0.15234375,
+      "learning_rate": 0.02,
+      "loss": 1.5567,
+      "step": 5219
+    },
+    {
+      "epoch": 0.48324943586182956,
+      "grad_norm": 0.1494140625,
+      "learning_rate": 0.02,
+      "loss": 1.5403,
+      "step": 5220
+    },
+    {
+      "epoch": 0.4833420123821096,
+      "grad_norm": 0.1484375,
+      "learning_rate": 0.02,
+      "loss": 1.6335,
+      "step": 5221
+    },
+    {
+      "epoch": 0.4834345889023896,
+      "grad_norm": 0.154296875,
+      "learning_rate": 0.02,
+      "loss": 1.5777,
+      "step": 5222
+    },
+    {
+      "epoch": 0.48352716542266966,
+      "grad_norm": 0.1533203125,
+      "learning_rate": 0.02,
+      "loss": 1.5708,
+      "step": 5223
+    },
+    {
+      "epoch": 0.4836197419429497,
+      "grad_norm": 0.142578125,
+      "learning_rate": 0.02,
+      "loss": 1.5493,
+      "step": 5224
+    },
+    {
+      "epoch": 0.48371231846322976,
+      "grad_norm": 0.150390625,
+      "learning_rate": 0.02,
+      "loss": 1.6235,
+      "step": 5225
+    },
+    {
+      "epoch": 0.4838048949835098,
+      "grad_norm": 0.1416015625,
+      "learning_rate": 0.02,
+      "loss": 1.5845,
+      "step": 5226
+    },
+    {
+      "epoch": 0.48389747150378987,
+      "grad_norm": 0.1416015625,
+      "learning_rate": 0.02,
+      "loss": 1.5669,
+      "step": 5227
+    },
+    {
+      "epoch": 0.4839900480240699,
+      "grad_norm": 0.15234375,
+      "learning_rate": 0.02,
+      "loss": 1.5014,
+      "step": 5228
+    },
+    {
+      "epoch": 0.4840826245443499,
+      "grad_norm": 0.1396484375,
+      "learning_rate": 0.02,
+      "loss": 1.5104,
+      "step": 5229
+    },
+    {
+      "epoch": 0.48417520106462997,
+      "grad_norm": 0.138671875,
+      "learning_rate": 0.02,
+      "loss": 1.5567,
+      "step": 5230
+    },
+    {
+      "epoch": 0.48426777758491,
+      "grad_norm": 0.1484375,
+      "learning_rate": 0.02,
+      "loss": 1.5419,
+      "step": 5231
+    },
+    {
+      "epoch": 0.48436035410519007,
+      "grad_norm": 0.1416015625,
+      "learning_rate": 0.02,
+      "loss": 1.5595,
+      "step": 5232
+    },
+    {
+      "epoch": 0.4844529306254701,
+      "grad_norm": 0.154296875,
+      "learning_rate": 0.02,
+      "loss": 1.5907,
+      "step": 5233
+    },
+    {
+      "epoch": 0.4845455071457502,
+      "grad_norm": 0.150390625,
+      "learning_rate": 0.02,
+      "loss": 1.6157,
+      "step": 5234
+    },
+    {
+      "epoch": 0.48463808366603023,
+      "grad_norm": 0.1572265625,
+      "learning_rate": 0.02,
+      "loss": 1.6121,
+      "step": 5235
+    },
+    {
+      "epoch": 0.4847306601863102,
+      "grad_norm": 0.1396484375,
+      "learning_rate": 0.02,
+      "loss": 1.5972,
+      "step": 5236
+    },
+    {
+      "epoch": 0.4848232367065903,
+      "grad_norm": 0.15234375,
+      "learning_rate": 0.02,
+      "loss": 1.6024,
+      "step": 5237
+    },
+    {
+      "epoch": 0.48491581322687033,
+      "grad_norm": 0.1337890625,
+      "learning_rate": 0.02,
+      "loss": 1.5984,
+      "step": 5238
+    },
+    {
+      "epoch": 0.4850083897471504,
+      "grad_norm": 0.14453125,
+      "learning_rate": 0.02,
+      "loss": 1.6026,
+      "step": 5239
+    },
+    {
+      "epoch": 0.48510096626743043,
+      "grad_norm": 0.1337890625,
+      "learning_rate": 0.02,
+      "loss": 1.5711,
+      "step": 5240
+    },
+    {
+      "epoch": 0.4851935427877105,
+      "grad_norm": 0.1435546875,
+      "learning_rate": 0.02,
+      "loss": 1.5659,
+      "step": 5241
+    },
+    {
+      "epoch": 0.48528611930799054,
+      "grad_norm": 0.13671875,
+      "learning_rate": 0.02,
+      "loss": 1.5055,
+      "step": 5242
+    },
+    {
+      "epoch": 0.48537869582827053,
+      "grad_norm": 0.1416015625,
+      "learning_rate": 0.02,
+      "loss": 1.5643,
+      "step": 5243
+    },
+    {
+      "epoch": 0.4854712723485506,
+      "grad_norm": 0.13671875,
+      "learning_rate": 0.02,
+      "loss": 1.6087,
+      "step": 5244
+    },
+    {
+      "epoch": 0.48556384886883064,
+      "grad_norm": 0.158203125,
+      "learning_rate": 0.02,
+      "loss": 1.5798,
+      "step": 5245
+    },
+    {
+      "epoch": 0.4856564253891107,
+      "grad_norm": 0.1376953125,
+      "learning_rate": 0.02,
+      "loss": 1.5219,
+      "step": 5246
+    },
+    {
+      "epoch": 0.48574900190939074,
+      "grad_norm": 0.1455078125,
+      "learning_rate": 0.02,
+      "loss": 1.6027,
+      "step": 5247
+    },
+    {
+      "epoch": 0.4858415784296708,
+      "grad_norm": 0.138671875,
+      "learning_rate": 0.02,
+      "loss": 1.5147,
+      "step": 5248
+    },
+    {
+      "epoch": 0.4859341549499508,
+      "grad_norm": 0.150390625,
+      "learning_rate": 0.02,
+      "loss": 1.5634,
+      "step": 5249
+    },
+    {
+      "epoch": 0.48602673147023084,
+      "grad_norm": 0.1396484375,
+      "learning_rate": 0.02,
+      "loss": 1.617,
+      "step": 5250
+    },
+    {
+      "epoch": 0.4861193079905109,
+      "grad_norm": 0.1484375,
+      "learning_rate": 0.02,
+      "loss": 1.5371,
+      "step": 5251
+    },
+    {
+      "epoch": 0.48621188451079095,
+      "grad_norm": 0.142578125,
+      "learning_rate": 0.02,
+      "loss": 1.5322,
+      "step": 5252
+    },
+    {
+      "epoch": 0.486304461031071,
+      "grad_norm": 0.1494140625,
+      "learning_rate": 0.02,
+      "loss": 1.6197,
+      "step": 5253
+    },
+    {
+      "epoch": 0.48639703755135105,
+      "grad_norm": 0.150390625,
+      "learning_rate": 0.02,
+      "loss": 1.6107,
+      "step": 5254
+    },
+    {
+      "epoch": 0.4864896140716311,
+      "grad_norm": 0.154296875,
+      "learning_rate": 0.02,
+      "loss": 1.6126,
+      "step": 5255
+    },
+    {
+      "epoch": 0.4865821905919111,
+      "grad_norm": 0.140625,
+      "learning_rate": 0.02,
+      "loss": 1.487,
+      "step": 5256
+    },
+    {
+      "epoch": 0.48667476711219115,
+      "grad_norm": 0.15234375,
+      "learning_rate": 0.02,
+      "loss": 1.581,
+      "step": 5257
+    },
+    {
+      "epoch": 0.4867673436324712,
+      "grad_norm": 0.1572265625,
+      "learning_rate": 0.02,
+      "loss": 1.5984,
+      "step": 5258
+    },
+    {
+      "epoch": 0.48685992015275126,
+      "grad_norm": 0.14453125,
+      "learning_rate": 0.02,
+      "loss": 1.6304,
+      "step": 5259
+    },
+    {
+      "epoch": 0.4869524966730313,
+      "grad_norm": 0.1533203125,
+      "learning_rate": 0.02,
+      "loss": 1.5557,
+      "step": 5260
+    },
+    {
+      "epoch": 0.48704507319331136,
+      "grad_norm": 0.140625,
+      "learning_rate": 0.02,
+      "loss": 1.5914,
+      "step": 5261
+    },
+    {
+      "epoch": 0.4871376497135914,
+      "grad_norm": 0.1416015625,
+      "learning_rate": 0.02,
+      "loss": 1.5066,
+      "step": 5262
+    },
+    {
+      "epoch": 0.4872302262338714,
+      "grad_norm": 0.14453125,
+      "learning_rate": 0.02,
+      "loss": 1.5189,
+      "step": 5263
+    },
+    {
+      "epoch": 0.48732280275415146,
+      "grad_norm": 0.1484375,
+      "learning_rate": 0.02,
+      "loss": 1.5536,
+      "step": 5264
+    },
+    {
+      "epoch": 0.4874153792744315,
+      "grad_norm": 0.14453125,
+      "learning_rate": 0.02,
+      "loss": 1.6115,
+      "step": 5265
+    },
+    {
+      "epoch": 0.48750795579471157,
+      "grad_norm": 0.13671875,
+      "learning_rate": 0.02,
+      "loss": 1.5907,
+      "step": 5266
+    },
+    {
+      "epoch": 0.4876005323149916,
+      "grad_norm": 0.140625,
+      "learning_rate": 0.02,
+      "loss": 1.5773,
+      "step": 5267
+    },
+    {
+      "epoch": 0.48769310883527167,
+      "grad_norm": 0.1455078125,
+      "learning_rate": 0.02,
+      "loss": 1.6089,
+      "step": 5268
+    },
+    {
+      "epoch": 0.4877856853555517,
+      "grad_norm": 0.1396484375,
+      "learning_rate": 0.02,
+      "loss": 1.5679,
+      "step": 5269
+    },
+    {
+      "epoch": 0.4878782618758317,
+      "grad_norm": 0.1484375,
+      "learning_rate": 0.02,
+      "loss": 1.6113,
+      "step": 5270
+    },
+    {
+      "epoch": 0.4879708383961118,
+      "grad_norm": 0.1455078125,
+      "learning_rate": 0.02,
+      "loss": 1.5508,
+      "step": 5271
+    },
+    {
+      "epoch": 0.4880634149163918,
+      "grad_norm": 0.1494140625,
+      "learning_rate": 0.02,
+      "loss": 1.5656,
+      "step": 5272
+    },
+    {
+      "epoch": 0.4881559914366719,
+      "grad_norm": 0.13671875,
+      "learning_rate": 0.02,
+      "loss": 1.5553,
+      "step": 5273
+    },
+    {
+      "epoch": 0.48824856795695193,
+      "grad_norm": 0.138671875,
+      "learning_rate": 0.02,
+      "loss": 1.5922,
+      "step": 5274
+    },
+    {
+      "epoch": 0.488341144477232,
+      "grad_norm": 0.1484375,
+      "learning_rate": 0.02,
+      "loss": 1.5054,
+      "step": 5275
+    },
+    {
+      "epoch": 0.48843372099751203,
+      "grad_norm": 0.1328125,
+      "learning_rate": 0.02,
+      "loss": 1.5433,
+      "step": 5276
+    },
+    {
+      "epoch": 0.48852629751779203,
+      "grad_norm": 0.1455078125,
+      "learning_rate": 0.02,
+      "loss": 1.5845,
+      "step": 5277
+    },
+    {
+      "epoch": 0.4886188740380721,
+      "grad_norm": 0.146484375,
+      "learning_rate": 0.02,
+      "loss": 1.5008,
+      "step": 5278
+    },
+    {
+      "epoch": 0.48871145055835213,
+      "grad_norm": 0.1455078125,
+      "learning_rate": 0.02,
+      "loss": 1.5799,
+      "step": 5279
+    },
+    {
+      "epoch": 0.4888040270786322,
+      "grad_norm": 0.142578125,
+      "learning_rate": 0.02,
+      "loss": 1.5477,
+      "step": 5280
+    },
+    {
+      "epoch": 0.48889660359891224,
+      "grad_norm": 0.1494140625,
+      "learning_rate": 0.02,
+      "loss": 1.5784,
+      "step": 5281
+    },
+    {
+      "epoch": 0.4889891801191923,
+      "grad_norm": 0.15234375,
+      "learning_rate": 0.02,
+      "loss": 1.5902,
+      "step": 5282
+    },
+    {
+      "epoch": 0.4890817566394723,
+      "grad_norm": 0.1552734375,
+      "learning_rate": 0.02,
+      "loss": 1.5798,
+      "step": 5283
+    },
+    {
+      "epoch": 0.48917433315975234,
+      "grad_norm": 0.1484375,
+      "learning_rate": 0.02,
+      "loss": 1.563,
+      "step": 5284
+    },
+    {
+      "epoch": 0.4892669096800324,
+      "grad_norm": 0.158203125,
+      "learning_rate": 0.02,
+      "loss": 1.5473,
+      "step": 5285
+    },
+    {
+      "epoch": 0.48935948620031244,
+      "grad_norm": 0.1533203125,
+      "learning_rate": 0.02,
+      "loss": 1.5577,
+      "step": 5286
+    },
+    {
+      "epoch": 0.4894520627205925,
+      "grad_norm": 0.1474609375,
+      "learning_rate": 0.02,
+      "loss": 1.5503,
+      "step": 5287
+    },
+    {
+      "epoch": 0.48954463924087255,
+      "grad_norm": 0.1552734375,
+      "learning_rate": 0.02,
+      "loss": 1.5766,
+      "step": 5288
+    },
+    {
+      "epoch": 0.4896372157611526,
+      "grad_norm": 0.1591796875,
+      "learning_rate": 0.02,
+      "loss": 1.5315,
+      "step": 5289
+    },
+    {
+      "epoch": 0.4897297922814326,
+      "grad_norm": 0.14453125,
+      "learning_rate": 0.02,
+      "loss": 1.5873,
+      "step": 5290
+    },
+    {
+      "epoch": 0.48982236880171265,
+      "grad_norm": 0.15625,
+      "learning_rate": 0.02,
+      "loss": 1.6014,
+      "step": 5291
+    },
+    {
+      "epoch": 0.4899149453219927,
+      "grad_norm": 0.146484375,
+      "learning_rate": 0.02,
+      "loss": 1.5605,
+      "step": 5292
+    },
+    {
+      "epoch": 0.49000752184227275,
+      "grad_norm": 0.1533203125,
+      "learning_rate": 0.02,
+      "loss": 1.5215,
+      "step": 5293
+    },
+    {
+      "epoch": 0.4901000983625528,
+      "grad_norm": 0.154296875,
+      "learning_rate": 0.02,
+      "loss": 1.5837,
+      "step": 5294
+    },
+    {
+      "epoch": 0.49019267488283286,
+      "grad_norm": 0.15625,
+      "learning_rate": 0.02,
+      "loss": 1.5939,
+      "step": 5295
+    },
+    {
+      "epoch": 0.4902852514031129,
+      "grad_norm": 0.1484375,
+      "learning_rate": 0.02,
+      "loss": 1.5319,
+      "step": 5296
+    },
+    {
+      "epoch": 0.4903778279233929,
+      "grad_norm": 0.1494140625,
+      "learning_rate": 0.02,
+      "loss": 1.586,
+      "step": 5297
+    },
+    {
+      "epoch": 0.49047040444367296,
+      "grad_norm": 0.146484375,
+      "learning_rate": 0.02,
+      "loss": 1.6067,
+      "step": 5298
+    },
+    {
+      "epoch": 0.490562980963953,
+      "grad_norm": 0.1435546875,
+      "learning_rate": 0.02,
+      "loss": 1.5843,
+      "step": 5299
+    },
+    {
+      "epoch": 0.49065555748423306,
+      "grad_norm": 0.1435546875,
+      "learning_rate": 0.02,
+      "loss": 1.5874,
+      "step": 5300
+    },
+    {
+      "epoch": 0.4907481340045131,
+      "grad_norm": 0.142578125,
+      "learning_rate": 0.02,
+      "loss": 1.5642,
+      "step": 5301
+    },
+    {
+      "epoch": 0.49084071052479317,
+      "grad_norm": 0.1435546875,
+      "learning_rate": 0.02,
+      "loss": 1.592,
+      "step": 5302
+    },
+    {
+      "epoch": 0.4909332870450732,
+      "grad_norm": 0.1494140625,
+      "learning_rate": 0.02,
+      "loss": 1.6081,
+      "step": 5303
+    },
+    {
+      "epoch": 0.4910258635653532,
+      "grad_norm": 0.1533203125,
+      "learning_rate": 0.02,
+      "loss": 1.554,
+      "step": 5304
+    },
+    {
+      "epoch": 0.49111844008563327,
+      "grad_norm": 0.1396484375,
+      "learning_rate": 0.02,
+      "loss": 1.5694,
+      "step": 5305
+    },
+    {
+      "epoch": 0.4912110166059133,
+      "grad_norm": 0.1533203125,
+      "learning_rate": 0.02,
+      "loss": 1.5816,
+      "step": 5306
+    },
+    {
+      "epoch": 0.49130359312619337,
+      "grad_norm": 0.150390625,
+      "learning_rate": 0.02,
+      "loss": 1.589,
+      "step": 5307
+    },
+    {
+      "epoch": 0.4913961696464734,
+      "grad_norm": 0.146484375,
+      "learning_rate": 0.02,
+      "loss": 1.5347,
+      "step": 5308
+    },
+    {
+      "epoch": 0.4914887461667535,
+      "grad_norm": 0.150390625,
+      "learning_rate": 0.02,
+      "loss": 1.486,
+      "step": 5309
+    },
+    {
+      "epoch": 0.49158132268703353,
+      "grad_norm": 0.1474609375,
+      "learning_rate": 0.02,
+      "loss": 1.5794,
+      "step": 5310
+    },
+    {
+      "epoch": 0.4916738992073135,
+      "grad_norm": 0.1435546875,
+      "learning_rate": 0.02,
+      "loss": 1.623,
+      "step": 5311
+    },
+    {
+      "epoch": 0.4917664757275936,
+      "grad_norm": 0.1513671875,
+      "learning_rate": 0.02,
+      "loss": 1.5892,
+      "step": 5312
+    },
+    {
+      "epoch": 0.49185905224787363,
+      "grad_norm": 0.140625,
+      "learning_rate": 0.02,
+      "loss": 1.6062,
+      "step": 5313
+    },
+    {
+      "epoch": 0.4919516287681537,
+      "grad_norm": 0.1474609375,
+      "learning_rate": 0.02,
+      "loss": 1.5238,
+      "step": 5314
+    },
+    {
+      "epoch": 0.49204420528843373,
+      "grad_norm": 0.1396484375,
+      "learning_rate": 0.02,
+      "loss": 1.5722,
+      "step": 5315
+    },
+    {
+      "epoch": 0.4921367818087138,
+      "grad_norm": 0.1435546875,
+      "learning_rate": 0.02,
+      "loss": 1.5742,
+      "step": 5316
+    },
+    {
+      "epoch": 0.4922293583289938,
+      "grad_norm": 0.1416015625,
+      "learning_rate": 0.02,
+      "loss": 1.5703,
+      "step": 5317
+    },
+    {
+      "epoch": 0.49232193484927383,
+      "grad_norm": 0.1376953125,
+      "learning_rate": 0.02,
+      "loss": 1.5686,
+      "step": 5318
+    },
+    {
+      "epoch": 0.4924145113695539,
+      "grad_norm": 0.1435546875,
+      "learning_rate": 0.02,
+      "loss": 1.6056,
+      "step": 5319
+    },
+    {
+      "epoch": 0.49250708788983394,
+      "grad_norm": 0.1494140625,
+      "learning_rate": 0.02,
+      "loss": 1.5212,
+      "step": 5320
+    },
+    {
+      "epoch": 0.492599664410114,
+      "grad_norm": 0.140625,
+      "learning_rate": 0.02,
+      "loss": 1.5276,
+      "step": 5321
+    },
+    {
+      "epoch": 0.49269224093039404,
+      "grad_norm": 0.146484375,
+      "learning_rate": 0.02,
+      "loss": 1.5485,
+      "step": 5322
+    },
+    {
+      "epoch": 0.4927848174506741,
+      "grad_norm": 0.154296875,
+      "learning_rate": 0.02,
+      "loss": 1.6596,
+      "step": 5323
+    },
+    {
+      "epoch": 0.4928773939709541,
+      "grad_norm": 0.1513671875,
+      "learning_rate": 0.02,
+      "loss": 1.6072,
+      "step": 5324
+    },
+    {
+      "epoch": 0.49296997049123414,
+      "grad_norm": 0.1513671875,
+      "learning_rate": 0.02,
+      "loss": 1.5931,
+      "step": 5325
+    },
+    {
+      "epoch": 0.4930625470115142,
+      "grad_norm": 0.1533203125,
+      "learning_rate": 0.02,
+      "loss": 1.6059,
+      "step": 5326
+    },
+    {
+      "epoch": 0.49315512353179425,
+      "grad_norm": 0.1455078125,
+      "learning_rate": 0.02,
+      "loss": 1.5387,
+      "step": 5327
+    },
+    {
+      "epoch": 0.4932477000520743,
+      "grad_norm": 0.1318359375,
+      "learning_rate": 0.02,
+      "loss": 1.5824,
+      "step": 5328
+    },
+    {
+      "epoch": 0.49334027657235435,
+      "grad_norm": 0.1494140625,
+      "learning_rate": 0.02,
+      "loss": 1.5485,
+      "step": 5329
+    },
+    {
+      "epoch": 0.4934328530926344,
+      "grad_norm": 0.146484375,
+      "learning_rate": 0.02,
+      "loss": 1.5829,
+      "step": 5330
+    },
+    {
+      "epoch": 0.4935254296129144,
+      "grad_norm": 0.154296875,
+      "learning_rate": 0.02,
+      "loss": 1.6059,
+      "step": 5331
+    },
+    {
+      "epoch": 0.49361800613319445,
+      "grad_norm": 0.1552734375,
+      "learning_rate": 0.02,
+      "loss": 1.525,
+      "step": 5332
+    },
+    {
+      "epoch": 0.4937105826534745,
+      "grad_norm": 0.1611328125,
+      "learning_rate": 0.02,
+      "loss": 1.5624,
+      "step": 5333
+    },
+    {
+      "epoch": 0.49380315917375456,
+      "grad_norm": 0.1474609375,
+      "learning_rate": 0.02,
+      "loss": 1.5449,
+      "step": 5334
+    },
+    {
+      "epoch": 0.4938957356940346,
+      "grad_norm": 0.15234375,
+      "learning_rate": 0.02,
+      "loss": 1.5593,
+      "step": 5335
+    },
+    {
+      "epoch": 0.49398831221431466,
+      "grad_norm": 0.1494140625,
+      "learning_rate": 0.02,
+      "loss": 1.5092,
+      "step": 5336
+    },
+    {
+      "epoch": 0.4940808887345947,
+      "grad_norm": 0.1474609375,
+      "learning_rate": 0.02,
+      "loss": 1.5592,
+      "step": 5337
+    },
+    {
+      "epoch": 0.4941734652548747,
+      "grad_norm": 0.146484375,
+      "learning_rate": 0.02,
+      "loss": 1.5401,
+      "step": 5338
+    },
+    {
+      "epoch": 0.49426604177515476,
+      "grad_norm": 0.150390625,
+      "learning_rate": 0.02,
+      "loss": 1.6003,
+      "step": 5339
+    },
+    {
+      "epoch": 0.4943586182954348,
+      "grad_norm": 0.1435546875,
+      "learning_rate": 0.02,
+      "loss": 1.5927,
+      "step": 5340
+    },
+    {
+      "epoch": 0.49445119481571487,
+      "grad_norm": 0.14453125,
+      "learning_rate": 0.02,
+      "loss": 1.5788,
+      "step": 5341
+    },
+    {
+      "epoch": 0.4945437713359949,
+      "grad_norm": 0.1484375,
+      "learning_rate": 0.02,
+      "loss": 1.5713,
+      "step": 5342
+    },
+    {
+      "epoch": 0.49463634785627497,
+      "grad_norm": 0.15625,
+      "learning_rate": 0.02,
+      "loss": 1.57,
+      "step": 5343
+    },
+    {
+      "epoch": 0.494728924376555,
+      "grad_norm": 0.1484375,
+      "learning_rate": 0.02,
+      "loss": 1.5222,
+      "step": 5344
+    },
+    {
+      "epoch": 0.494821500896835,
+      "grad_norm": 0.146484375,
+      "learning_rate": 0.02,
+      "loss": 1.5395,
+      "step": 5345
+    },
+    {
+      "epoch": 0.4949140774171151,
+      "grad_norm": 0.14453125,
+      "learning_rate": 0.02,
+      "loss": 1.5909,
+      "step": 5346
+    },
+    {
+      "epoch": 0.4950066539373951,
+      "grad_norm": 0.1494140625,
+      "learning_rate": 0.02,
+      "loss": 1.5382,
+      "step": 5347
+    },
+    {
+      "epoch": 0.4950992304576752,
+      "grad_norm": 0.1416015625,
+      "learning_rate": 0.02,
+      "loss": 1.5869,
+      "step": 5348
+    },
+    {
+      "epoch": 0.49519180697795523,
+      "grad_norm": 0.1298828125,
+      "learning_rate": 0.02,
+      "loss": 1.5435,
+      "step": 5349
+    },
+    {
+      "epoch": 0.4952843834982353,
+      "grad_norm": 0.1337890625,
+      "learning_rate": 0.02,
+      "loss": 1.5902,
+      "step": 5350
+    },
+    {
+      "epoch": 0.4953769600185153,
+      "grad_norm": 0.15234375,
+      "learning_rate": 0.02,
+      "loss": 1.6712,
+      "step": 5351
+    },
+    {
+      "epoch": 0.49546953653879533,
+      "grad_norm": 0.15234375,
+      "learning_rate": 0.02,
+      "loss": 1.6068,
+      "step": 5352
+    },
+    {
+      "epoch": 0.4955621130590754,
+      "grad_norm": 0.158203125,
+      "learning_rate": 0.02,
+      "loss": 1.5549,
+      "step": 5353
+    },
+    {
+      "epoch": 0.49565468957935543,
+      "grad_norm": 0.14453125,
+      "learning_rate": 0.02,
+      "loss": 1.5191,
+      "step": 5354
+    },
+    {
+      "epoch": 0.4957472660996355,
+      "grad_norm": 0.1376953125,
+      "learning_rate": 0.02,
+      "loss": 1.5377,
+      "step": 5355
+    },
+    {
+      "epoch": 0.49583984261991554,
+      "grad_norm": 0.1484375,
+      "learning_rate": 0.02,
+      "loss": 1.5552,
+      "step": 5356
+    },
+    {
+      "epoch": 0.4959324191401956,
+      "grad_norm": 0.1435546875,
+      "learning_rate": 0.02,
+      "loss": 1.5378,
+      "step": 5357
+    },
+    {
+      "epoch": 0.4960249956604756,
+      "grad_norm": 0.138671875,
+      "learning_rate": 0.02,
+      "loss": 1.5607,
+      "step": 5358
+    },
+    {
+      "epoch": 0.49611757218075564,
+      "grad_norm": 0.1533203125,
+      "learning_rate": 0.02,
+      "loss": 1.595,
+      "step": 5359
+    },
+    {
+      "epoch": 0.4962101487010357,
+      "grad_norm": 0.1416015625,
+      "learning_rate": 0.02,
+      "loss": 1.5896,
+      "step": 5360
+    },
+    {
+      "epoch": 0.49630272522131574,
+      "grad_norm": 0.1376953125,
+      "learning_rate": 0.02,
+      "loss": 1.5742,
+      "step": 5361
+    },
+    {
+      "epoch": 0.4963953017415958,
+      "grad_norm": 0.134765625,
+      "learning_rate": 0.02,
+      "loss": 1.5511,
+      "step": 5362
+    },
+    {
+      "epoch": 0.49648787826187585,
+      "grad_norm": 0.1494140625,
+      "learning_rate": 0.02,
+      "loss": 1.6353,
+      "step": 5363
+    },
+    {
+      "epoch": 0.4965804547821559,
+      "grad_norm": 0.1484375,
+      "learning_rate": 0.02,
+      "loss": 1.5846,
+      "step": 5364
+    },
+    {
+      "epoch": 0.4966730313024359,
+      "grad_norm": 0.1474609375,
+      "learning_rate": 0.02,
+      "loss": 1.5004,
+      "step": 5365
+    },
+    {
+      "epoch": 0.49676560782271595,
+      "grad_norm": 0.1494140625,
+      "learning_rate": 0.02,
+      "loss": 1.5388,
+      "step": 5366
+    },
+    {
+      "epoch": 0.496858184342996,
+      "grad_norm": 0.1572265625,
+      "learning_rate": 0.02,
+      "loss": 1.6027,
+      "step": 5367
+    },
+    {
+      "epoch": 0.49695076086327605,
+      "grad_norm": 0.1533203125,
+      "learning_rate": 0.02,
+      "loss": 1.5493,
+      "step": 5368
+    },
+    {
+      "epoch": 0.4970433373835561,
+      "grad_norm": 0.140625,
+      "learning_rate": 0.02,
+      "loss": 1.5243,
+      "step": 5369
+    },
+    {
+      "epoch": 0.49713591390383616,
+      "grad_norm": 0.1416015625,
+      "learning_rate": 0.02,
+      "loss": 1.5527,
+      "step": 5370
+    },
+    {
+      "epoch": 0.4972284904241162,
+      "grad_norm": 0.14453125,
+      "learning_rate": 0.02,
+      "loss": 1.5678,
+      "step": 5371
+    },
+    {
+      "epoch": 0.4973210669443962,
+      "grad_norm": 0.15234375,
+      "learning_rate": 0.02,
+      "loss": 1.5361,
+      "step": 5372
+    },
+    {
+      "epoch": 0.49741364346467626,
+      "grad_norm": 0.14453125,
+      "learning_rate": 0.02,
+      "loss": 1.5033,
+      "step": 5373
+    },
+    {
+      "epoch": 0.4975062199849563,
+      "grad_norm": 0.142578125,
+      "learning_rate": 0.02,
+      "loss": 1.5937,
+      "step": 5374
+    },
+    {
+      "epoch": 0.49759879650523636,
+      "grad_norm": 0.15625,
+      "learning_rate": 0.02,
+      "loss": 1.5392,
+      "step": 5375
+    },
+    {
+      "epoch": 0.4976913730255164,
+      "grad_norm": 0.1591796875,
+      "learning_rate": 0.02,
+      "loss": 1.5877,
+      "step": 5376
+    },
+    {
+      "epoch": 0.49778394954579647,
+      "grad_norm": 0.1494140625,
+      "learning_rate": 0.02,
+      "loss": 1.5377,
+      "step": 5377
+    },
+    {
+      "epoch": 0.4978765260660765,
+      "grad_norm": 0.16015625,
+      "learning_rate": 0.02,
+      "loss": 1.5806,
+      "step": 5378
+    },
+    {
+      "epoch": 0.4979691025863565,
+      "grad_norm": 0.1513671875,
+      "learning_rate": 0.02,
+      "loss": 1.5221,
+      "step": 5379
+    },
+    {
+      "epoch": 0.49806167910663657,
+      "grad_norm": 0.1484375,
+      "learning_rate": 0.02,
+      "loss": 1.5749,
+      "step": 5380
+    },
+    {
+      "epoch": 0.4981542556269166,
+      "grad_norm": 0.1474609375,
+      "learning_rate": 0.02,
+      "loss": 1.5265,
+      "step": 5381
+    },
+    {
+      "epoch": 0.49824683214719667,
+      "grad_norm": 0.150390625,
+      "learning_rate": 0.02,
+      "loss": 1.6026,
+      "step": 5382
+    },
+    {
+      "epoch": 0.4983394086674767,
+      "grad_norm": 0.1435546875,
+      "learning_rate": 0.02,
+      "loss": 1.5605,
+      "step": 5383
+    },
+    {
+      "epoch": 0.4984319851877568,
+      "grad_norm": 0.1494140625,
+      "learning_rate": 0.02,
+      "loss": 1.528,
+      "step": 5384
+    },
+    {
+      "epoch": 0.4985245617080368,
+      "grad_norm": 0.1533203125,
+      "learning_rate": 0.02,
+      "loss": 1.5318,
+      "step": 5385
+    },
+    {
+      "epoch": 0.4986171382283168,
+      "grad_norm": 0.1630859375,
+      "learning_rate": 0.02,
+      "loss": 1.5616,
+      "step": 5386
+    },
+    {
+      "epoch": 0.4987097147485969,
+      "grad_norm": 0.1455078125,
+      "learning_rate": 0.02,
+      "loss": 1.5163,
+      "step": 5387
+    },
+    {
+      "epoch": 0.49880229126887693,
+      "grad_norm": 0.1533203125,
+      "learning_rate": 0.02,
+      "loss": 1.5446,
+      "step": 5388
+    },
+    {
+      "epoch": 0.498894867789157,
+      "grad_norm": 0.1484375,
+      "learning_rate": 0.02,
+      "loss": 1.563,
+      "step": 5389
+    },
+    {
+      "epoch": 0.49898744430943703,
+      "grad_norm": 0.1484375,
+      "learning_rate": 0.02,
+      "loss": 1.5657,
+      "step": 5390
+    },
+    {
+      "epoch": 0.4990800208297171,
+      "grad_norm": 0.1376953125,
+      "learning_rate": 0.02,
+      "loss": 1.5155,
+      "step": 5391
+    },
+    {
+      "epoch": 0.4991725973499971,
+      "grad_norm": 0.1455078125,
+      "learning_rate": 0.02,
+      "loss": 1.5253,
+      "step": 5392
+    },
+    {
+      "epoch": 0.49926517387027713,
+      "grad_norm": 0.1376953125,
+      "learning_rate": 0.02,
+      "loss": 1.559,
+      "step": 5393
+    },
+    {
+      "epoch": 0.4993577503905572,
+      "grad_norm": 0.1552734375,
+      "learning_rate": 0.02,
+      "loss": 1.6396,
+      "step": 5394
+    },
+    {
+      "epoch": 0.49945032691083724,
+      "grad_norm": 0.154296875,
+      "learning_rate": 0.02,
+      "loss": 1.5363,
+      "step": 5395
+    },
+    {
+      "epoch": 0.4995429034311173,
+      "grad_norm": 0.154296875,
+      "learning_rate": 0.02,
+      "loss": 1.5384,
+      "step": 5396
+    },
+    {
+      "epoch": 0.49963547995139734,
+      "grad_norm": 0.1455078125,
+      "learning_rate": 0.02,
+      "loss": 1.4743,
+      "step": 5397
+    },
+    {
+      "epoch": 0.4997280564716774,
+      "grad_norm": 0.1484375,
+      "learning_rate": 0.02,
+      "loss": 1.6183,
+      "step": 5398
+    },
+    {
+      "epoch": 0.4998206329919574,
+      "grad_norm": 0.1513671875,
+      "learning_rate": 0.02,
+      "loss": 1.5957,
+      "step": 5399
+    },
+    {
+      "epoch": 0.49991320951223744,
+      "grad_norm": 0.1591796875,
+      "learning_rate": 0.02,
+      "loss": 1.5754,
+      "step": 5400
+    },
+    {
+      "epoch": 0.5000057860325176,
+      "grad_norm": 0.15625,
+      "learning_rate": 0.02,
+      "loss": 1.5451,
+      "step": 5401
+    },
+    {
+      "epoch": 0.5000983625527975,
+      "grad_norm": 0.15625,
+      "learning_rate": 0.02,
+      "loss": 1.5401,
+      "step": 5402
+    },
+    {
+      "epoch": 0.5001909390730775,
+      "grad_norm": 0.15625,
+      "learning_rate": 0.02,
+      "loss": 1.6026,
+      "step": 5403
+    },
+    {
+      "epoch": 0.5002835155933576,
+      "grad_norm": 0.1533203125,
+      "learning_rate": 0.02,
+      "loss": 1.5788,
+      "step": 5404
+    },
+    {
+      "epoch": 0.5003760921136376,
+      "grad_norm": 0.1435546875,
+      "learning_rate": 0.02,
+      "loss": 1.4953,
+      "step": 5405
+    },
+    {
+      "epoch": 0.5004686686339177,
+      "grad_norm": 0.142578125,
+      "learning_rate": 0.02,
+      "loss": 1.5448,
+      "step": 5406
+    },
+    {
+      "epoch": 0.5005612451541978,
+      "grad_norm": 0.1396484375,
+      "learning_rate": 0.02,
+      "loss": 1.5784,
+      "step": 5407
+    },
+    {
+      "epoch": 0.5006538216744778,
+      "grad_norm": 0.154296875,
+      "learning_rate": 0.02,
+      "loss": 1.5271,
+      "step": 5408
+    },
+    {
+      "epoch": 0.5007463981947579,
+      "grad_norm": 0.1416015625,
+      "learning_rate": 0.02,
+      "loss": 1.5274,
+      "step": 5409
+    },
+    {
+      "epoch": 0.5008389747150379,
+      "grad_norm": 0.158203125,
+      "learning_rate": 0.02,
+      "loss": 1.6237,
+      "step": 5410
+    },
+    {
+      "epoch": 0.500931551235318,
+      "grad_norm": 0.142578125,
+      "learning_rate": 0.02,
+      "loss": 1.5826,
+      "step": 5411
+    },
+    {
+      "epoch": 0.501024127755598,
+      "grad_norm": 0.1513671875,
+      "learning_rate": 0.02,
+      "loss": 1.5492,
+      "step": 5412
+    },
+    {
+      "epoch": 0.5011167042758781,
+      "grad_norm": 0.1416015625,
+      "learning_rate": 0.02,
+      "loss": 1.6137,
+      "step": 5413
+    },
+    {
+      "epoch": 0.5012092807961581,
+      "grad_norm": 0.138671875,
+      "learning_rate": 0.02,
+      "loss": 1.5293,
+      "step": 5414
+    },
+    {
+      "epoch": 0.5013018573164382,
+      "grad_norm": 0.1396484375,
+      "learning_rate": 0.02,
+      "loss": 1.5341,
+      "step": 5415
+    },
+    {
+      "epoch": 0.5013944338367181,
+      "grad_norm": 0.1435546875,
+      "learning_rate": 0.02,
+      "loss": 1.6153,
+      "step": 5416
+    },
+    {
+      "epoch": 0.5014870103569982,
+      "grad_norm": 0.1416015625,
+      "learning_rate": 0.02,
+      "loss": 1.5939,
+      "step": 5417
+    },
+    {
+      "epoch": 0.5015795868772782,
+      "grad_norm": 0.146484375,
+      "learning_rate": 0.02,
+      "loss": 1.5577,
+      "step": 5418
+    },
+    {
+      "epoch": 0.5016721633975583,
+      "grad_norm": 0.142578125,
+      "learning_rate": 0.02,
+      "loss": 1.5556,
+      "step": 5419
+    },
+    {
+      "epoch": 0.5017647399178383,
+      "grad_norm": 0.1474609375,
+      "learning_rate": 0.02,
+      "loss": 1.5994,
+      "step": 5420
+    },
+    {
+      "epoch": 0.5018573164381184,
+      "grad_norm": 0.142578125,
+      "learning_rate": 0.02,
+      "loss": 1.5808,
+      "step": 5421
+    },
+    {
+      "epoch": 0.5019498929583984,
+      "grad_norm": 0.13671875,
+      "learning_rate": 0.02,
+      "loss": 1.5485,
+      "step": 5422
+    },
+    {
+      "epoch": 0.5020424694786785,
+      "grad_norm": 0.14453125,
+      "learning_rate": 0.02,
+      "loss": 1.6195,
+      "step": 5423
+    },
+    {
+      "epoch": 0.5021350459989585,
+      "grad_norm": 0.16015625,
+      "learning_rate": 0.02,
+      "loss": 1.6219,
+      "step": 5424
+    },
+    {
+      "epoch": 0.5022276225192386,
+      "grad_norm": 0.15234375,
+      "learning_rate": 0.02,
+      "loss": 1.6012,
+      "step": 5425
+    },
+    {
+      "epoch": 0.5023201990395186,
+      "grad_norm": 0.1494140625,
+      "learning_rate": 0.02,
+      "loss": 1.6037,
+      "step": 5426
+    },
+    {
+      "epoch": 0.5024127755597987,
+      "grad_norm": 0.15625,
+      "learning_rate": 0.02,
+      "loss": 1.5328,
+      "step": 5427
+    },
+    {
+      "epoch": 0.5025053520800787,
+      "grad_norm": 0.1484375,
+      "learning_rate": 0.02,
+      "loss": 1.5461,
+      "step": 5428
+    },
+    {
+      "epoch": 0.5025979286003587,
+      "grad_norm": 0.142578125,
+      "learning_rate": 0.02,
+      "loss": 1.5378,
+      "step": 5429
+    },
+    {
+      "epoch": 0.5026905051206387,
+      "grad_norm": 0.1474609375,
+      "learning_rate": 0.02,
+      "loss": 1.5835,
+      "step": 5430
+    },
+    {
+      "epoch": 0.5027830816409188,
+      "grad_norm": 0.1591796875,
+      "learning_rate": 0.02,
+      "loss": 1.6232,
+      "step": 5431
+    },
+    {
+      "epoch": 0.5028756581611988,
+      "grad_norm": 0.1484375,
+      "learning_rate": 0.02,
+      "loss": 1.6271,
+      "step": 5432
+    },
+    {
+      "epoch": 0.5029682346814789,
+      "grad_norm": 0.1572265625,
+      "learning_rate": 0.02,
+      "loss": 1.6059,
+      "step": 5433
+    },
+    {
+      "epoch": 0.5030608112017589,
+      "grad_norm": 0.154296875,
+      "learning_rate": 0.02,
+      "loss": 1.5599,
+      "step": 5434
+    },
+    {
+      "epoch": 0.503153387722039,
+      "grad_norm": 0.1513671875,
+      "learning_rate": 0.02,
+      "loss": 1.5803,
+      "step": 5435
+    },
+    {
+      "epoch": 0.503245964242319,
+      "grad_norm": 0.1455078125,
+      "learning_rate": 0.02,
+      "loss": 1.5508,
+      "step": 5436
+    },
+    {
+      "epoch": 0.5033385407625991,
+      "grad_norm": 0.1435546875,
+      "learning_rate": 0.02,
+      "loss": 1.6099,
+      "step": 5437
+    },
+    {
+      "epoch": 0.5034311172828791,
+      "grad_norm": 0.1513671875,
+      "learning_rate": 0.02,
+      "loss": 1.5877,
+      "step": 5438
+    },
+    {
+      "epoch": 0.5035236938031592,
+      "grad_norm": 0.1396484375,
+      "learning_rate": 0.02,
+      "loss": 1.508,
+      "step": 5439
+    },
+    {
+      "epoch": 0.5036162703234393,
+      "grad_norm": 0.1494140625,
+      "learning_rate": 0.02,
+      "loss": 1.5323,
+      "step": 5440
+    },
+    {
+      "epoch": 0.5037088468437193,
+      "grad_norm": 0.14453125,
+      "learning_rate": 0.02,
+      "loss": 1.5173,
+      "step": 5441
+    },
+    {
+      "epoch": 0.5038014233639994,
+      "grad_norm": 0.1494140625,
+      "learning_rate": 0.02,
+      "loss": 1.5442,
+      "step": 5442
+    },
+    {
+      "epoch": 0.5038939998842793,
+      "grad_norm": 0.1455078125,
+      "learning_rate": 0.02,
+      "loss": 1.6403,
+      "step": 5443
+    },
+    {
+      "epoch": 0.5039865764045593,
+      "grad_norm": 0.15625,
+      "learning_rate": 0.02,
+      "loss": 1.6548,
+      "step": 5444
+    },
+    {
+      "epoch": 0.5040791529248394,
+      "grad_norm": 0.1416015625,
+      "learning_rate": 0.02,
+      "loss": 1.5687,
+      "step": 5445
+    },
+    {
+      "epoch": 0.5041717294451195,
+      "grad_norm": 0.142578125,
+      "learning_rate": 0.02,
+      "loss": 1.5597,
+      "step": 5446
+    },
+    {
+      "epoch": 0.5042643059653995,
+      "grad_norm": 0.142578125,
+      "learning_rate": 0.02,
+      "loss": 1.6117,
+      "step": 5447
+    },
+    {
+      "epoch": 0.5043568824856796,
+      "grad_norm": 0.1494140625,
+      "learning_rate": 0.02,
+      "loss": 1.5894,
+      "step": 5448
+    },
+    {
+      "epoch": 0.5044494590059596,
+      "grad_norm": 0.1435546875,
+      "learning_rate": 0.02,
+      "loss": 1.5926,
+      "step": 5449
+    },
+    {
+      "epoch": 0.5045420355262397,
+      "grad_norm": 0.138671875,
+      "learning_rate": 0.02,
+      "loss": 1.5914,
+      "step": 5450
+    },
+    {
+      "epoch": 0.5046346120465197,
+      "grad_norm": 0.1396484375,
+      "learning_rate": 0.02,
+      "loss": 1.53,
+      "step": 5451
+    },
+    {
+      "epoch": 0.5047271885667998,
+      "grad_norm": 0.1494140625,
+      "learning_rate": 0.02,
+      "loss": 1.6101,
+      "step": 5452
+    },
+    {
+      "epoch": 0.5048197650870798,
+      "grad_norm": 0.146484375,
+      "learning_rate": 0.02,
+      "loss": 1.6035,
+      "step": 5453
+    },
+    {
+      "epoch": 0.5049123416073599,
+      "grad_norm": 0.1552734375,
+      "learning_rate": 0.02,
+      "loss": 1.6262,
+      "step": 5454
+    },
+    {
+      "epoch": 0.5050049181276399,
+      "grad_norm": 0.1455078125,
+      "learning_rate": 0.02,
+      "loss": 1.5474,
+      "step": 5455
+    },
+    {
+      "epoch": 0.50509749464792,
+      "grad_norm": 0.1474609375,
+      "learning_rate": 0.02,
+      "loss": 1.5535,
+      "step": 5456
+    },
+    {
+      "epoch": 0.5051900711681999,
+      "grad_norm": 0.142578125,
+      "learning_rate": 0.02,
+      "loss": 1.5687,
+      "step": 5457
+    },
+    {
+      "epoch": 0.50528264768848,
+      "grad_norm": 0.154296875,
+      "learning_rate": 0.02,
+      "loss": 1.5846,
+      "step": 5458
+    },
+    {
+      "epoch": 0.50537522420876,
+      "grad_norm": 0.150390625,
+      "learning_rate": 0.02,
+      "loss": 1.5515,
+      "step": 5459
+    },
+    {
+      "epoch": 0.5054678007290401,
+      "grad_norm": 0.1474609375,
+      "learning_rate": 0.02,
+      "loss": 1.5568,
+      "step": 5460
+    },
+    {
+      "epoch": 0.5055603772493201,
+      "grad_norm": 0.1357421875,
+      "learning_rate": 0.02,
+      "loss": 1.541,
+      "step": 5461
+    },
+    {
+      "epoch": 0.5056529537696002,
+      "grad_norm": 0.150390625,
+      "learning_rate": 0.02,
+      "loss": 1.5909,
+      "step": 5462
+    },
+    {
+      "epoch": 0.5057455302898802,
+      "grad_norm": 0.15234375,
+      "learning_rate": 0.02,
+      "loss": 1.5745,
+      "step": 5463
+    },
+    {
+      "epoch": 0.5058381068101603,
+      "grad_norm": 0.1552734375,
+      "learning_rate": 0.02,
+      "loss": 1.5156,
+      "step": 5464
+    },
+    {
+      "epoch": 0.5059306833304403,
+      "grad_norm": 0.138671875,
+      "learning_rate": 0.02,
+      "loss": 1.5531,
+      "step": 5465
+    },
+    {
+      "epoch": 0.5060232598507204,
+      "grad_norm": 0.1552734375,
+      "learning_rate": 0.02,
+      "loss": 1.6237,
+      "step": 5466
+    },
+    {
+      "epoch": 0.5061158363710004,
+      "grad_norm": 0.1416015625,
+      "learning_rate": 0.02,
+      "loss": 1.5672,
+      "step": 5467
+    },
+    {
+      "epoch": 0.5062084128912805,
+      "grad_norm": 0.154296875,
+      "learning_rate": 0.02,
+      "loss": 1.5292,
+      "step": 5468
+    },
+    {
+      "epoch": 0.5063009894115605,
+      "grad_norm": 0.1630859375,
+      "learning_rate": 0.02,
+      "loss": 1.5306,
+      "step": 5469
+    },
+    {
+      "epoch": 0.5063935659318405,
+      "grad_norm": 0.1630859375,
+      "learning_rate": 0.02,
+      "loss": 1.5699,
+      "step": 5470
+    },
+    {
+      "epoch": 0.5064861424521205,
+      "grad_norm": 0.1513671875,
+      "learning_rate": 0.02,
+      "loss": 1.5358,
+      "step": 5471
+    },
+    {
+      "epoch": 0.5065787189724006,
+      "grad_norm": 0.158203125,
+      "learning_rate": 0.02,
+      "loss": 1.643,
+      "step": 5472
+    },
+    {
+      "epoch": 0.5066712954926806,
+      "grad_norm": 0.150390625,
+      "learning_rate": 0.02,
+      "loss": 1.6485,
+      "step": 5473
+    },
+    {
+      "epoch": 0.5067638720129607,
+      "grad_norm": 0.16015625,
+      "learning_rate": 0.02,
+      "loss": 1.6087,
+      "step": 5474
+    },
+    {
+      "epoch": 0.5068564485332407,
+      "grad_norm": 0.150390625,
+      "learning_rate": 0.02,
+      "loss": 1.6516,
+      "step": 5475
+    },
+    {
+      "epoch": 0.5069490250535208,
+      "grad_norm": 0.15234375,
+      "learning_rate": 0.02,
+      "loss": 1.5645,
+      "step": 5476
+    },
+    {
+      "epoch": 0.5070416015738008,
+      "grad_norm": 0.1494140625,
+      "learning_rate": 0.02,
+      "loss": 1.6008,
+      "step": 5477
+    },
+    {
+      "epoch": 0.5071341780940809,
+      "grad_norm": 0.1455078125,
+      "learning_rate": 0.02,
+      "loss": 1.5659,
+      "step": 5478
+    },
+    {
+      "epoch": 0.507226754614361,
+      "grad_norm": 0.1455078125,
+      "learning_rate": 0.02,
+      "loss": 1.5733,
+      "step": 5479
+    },
+    {
+      "epoch": 0.507319331134641,
+      "grad_norm": 0.1513671875,
+      "learning_rate": 0.02,
+      "loss": 1.5633,
+      "step": 5480
+    },
+    {
+      "epoch": 0.5074119076549211,
+      "grad_norm": 0.1533203125,
+      "learning_rate": 0.02,
+      "loss": 1.6251,
+      "step": 5481
+    },
+    {
+      "epoch": 0.5075044841752011,
+      "grad_norm": 0.14453125,
+      "learning_rate": 0.02,
+      "loss": 1.5369,
+      "step": 5482
+    },
+    {
+      "epoch": 0.5075970606954812,
+      "grad_norm": 0.1435546875,
+      "learning_rate": 0.02,
+      "loss": 1.5577,
+      "step": 5483
+    },
+    {
+      "epoch": 0.5076896372157611,
+      "grad_norm": 0.1484375,
+      "learning_rate": 0.02,
+      "loss": 1.583,
+      "step": 5484
+    },
+    {
+      "epoch": 0.5077822137360412,
+      "grad_norm": 0.1474609375,
+      "learning_rate": 0.02,
+      "loss": 1.5729,
+      "step": 5485
+    },
+    {
+      "epoch": 0.5078747902563212,
+      "grad_norm": 0.1474609375,
+      "learning_rate": 0.02,
+      "loss": 1.619,
+      "step": 5486
+    },
+    {
+      "epoch": 0.5079673667766013,
+      "grad_norm": 0.1435546875,
+      "learning_rate": 0.02,
+      "loss": 1.604,
+      "step": 5487
+    },
+    {
+      "epoch": 0.5080599432968813,
+      "grad_norm": 0.138671875,
+      "learning_rate": 0.02,
+      "loss": 1.5296,
+      "step": 5488
+    },
+    {
+      "epoch": 0.5081525198171614,
+      "grad_norm": 0.1318359375,
+      "learning_rate": 0.02,
+      "loss": 1.5624,
+      "step": 5489
+    },
+    {
+      "epoch": 0.5082450963374414,
+      "grad_norm": 0.1494140625,
+      "learning_rate": 0.02,
+      "loss": 1.6056,
+      "step": 5490
+    },
+    {
+      "epoch": 0.5083376728577215,
+      "grad_norm": 0.14453125,
+      "learning_rate": 0.02,
+      "loss": 1.5433,
+      "step": 5491
+    },
+    {
+      "epoch": 0.5084302493780015,
+      "grad_norm": 0.154296875,
+      "learning_rate": 0.02,
+      "loss": 1.6075,
+      "step": 5492
+    },
+    {
+      "epoch": 0.5085228258982816,
+      "grad_norm": 0.142578125,
+      "learning_rate": 0.02,
+      "loss": 1.5074,
+      "step": 5493
+    },
+    {
+      "epoch": 0.5086154024185616,
+      "grad_norm": 0.15625,
+      "learning_rate": 0.02,
+      "loss": 1.6281,
+      "step": 5494
+    },
+    {
+      "epoch": 0.5087079789388417,
+      "grad_norm": 0.1494140625,
+      "learning_rate": 0.02,
+      "loss": 1.5236,
+      "step": 5495
+    },
+    {
+      "epoch": 0.5088005554591217,
+      "grad_norm": 0.1494140625,
+      "learning_rate": 0.02,
+      "loss": 1.5889,
+      "step": 5496
+    },
+    {
+      "epoch": 0.5088931319794017,
+      "grad_norm": 0.14453125,
+      "learning_rate": 0.02,
+      "loss": 1.5567,
+      "step": 5497
+    },
+    {
+      "epoch": 0.5089857084996817,
+      "grad_norm": 0.1552734375,
+      "learning_rate": 0.02,
+      "loss": 1.446,
+      "step": 5498
+    },
+    {
+      "epoch": 0.5090782850199618,
+      "grad_norm": 0.1484375,
+      "learning_rate": 0.02,
+      "loss": 1.5453,
+      "step": 5499
+    },
+    {
+      "epoch": 0.5091708615402418,
+      "grad_norm": 0.15625,
+      "learning_rate": 0.02,
+      "loss": 1.61,
+      "step": 5500
+    },
+    {
+      "epoch": 0.5092634380605219,
+      "grad_norm": 0.15234375,
+      "learning_rate": 0.02,
+      "loss": 1.6106,
+      "step": 5501
+    },
+    {
+      "epoch": 0.5093560145808019,
+      "grad_norm": 0.14453125,
+      "learning_rate": 0.02,
+      "loss": 1.4881,
+      "step": 5502
+    },
+    {
+      "epoch": 0.509448591101082,
+      "grad_norm": 0.1494140625,
+      "learning_rate": 0.02,
+      "loss": 1.5881,
+      "step": 5503
+    },
+    {
+      "epoch": 0.509541167621362,
+      "grad_norm": 0.146484375,
+      "learning_rate": 0.02,
+      "loss": 1.5341,
+      "step": 5504
+    },
+    {
+      "epoch": 0.5096337441416421,
+      "grad_norm": 0.150390625,
+      "learning_rate": 0.02,
+      "loss": 1.5998,
+      "step": 5505
+    },
+    {
+      "epoch": 0.5097263206619221,
+      "grad_norm": 0.140625,
+      "learning_rate": 0.02,
+      "loss": 1.5046,
+      "step": 5506
+    },
+    {
+      "epoch": 0.5098188971822022,
+      "grad_norm": 0.146484375,
+      "learning_rate": 0.02,
+      "loss": 1.5171,
+      "step": 5507
+    },
+    {
+      "epoch": 0.5099114737024822,
+      "grad_norm": 0.154296875,
+      "learning_rate": 0.02,
+      "loss": 1.5175,
+      "step": 5508
+    },
+    {
+      "epoch": 0.5100040502227623,
+      "grad_norm": 0.1474609375,
+      "learning_rate": 0.02,
+      "loss": 1.5656,
+      "step": 5509
+    },
+    {
+      "epoch": 0.5100966267430423,
+      "grad_norm": 0.154296875,
+      "learning_rate": 0.02,
+      "loss": 1.525,
+      "step": 5510
+    },
+    {
+      "epoch": 0.5101892032633223,
+      "grad_norm": 0.1494140625,
+      "learning_rate": 0.02,
+      "loss": 1.5537,
+      "step": 5511
+    },
+    {
+      "epoch": 0.5102817797836023,
+      "grad_norm": 0.1455078125,
+      "learning_rate": 0.02,
+      "loss": 1.5825,
+      "step": 5512
+    },
+    {
+      "epoch": 0.5103743563038824,
+      "grad_norm": 0.1416015625,
+      "learning_rate": 0.02,
+      "loss": 1.5682,
+      "step": 5513
+    },
+    {
+      "epoch": 0.5104669328241624,
+      "grad_norm": 0.1474609375,
+      "learning_rate": 0.02,
+      "loss": 1.6305,
+      "step": 5514
+    },
+    {
+      "epoch": 0.5105595093444425,
+      "grad_norm": 0.15625,
+      "learning_rate": 0.02,
+      "loss": 1.6275,
+      "step": 5515
+    },
+    {
+      "epoch": 0.5106520858647225,
+      "grad_norm": 0.14453125,
+      "learning_rate": 0.02,
+      "loss": 1.5814,
+      "step": 5516
+    },
+    {
+      "epoch": 0.5107446623850026,
+      "grad_norm": 0.150390625,
+      "learning_rate": 0.02,
+      "loss": 1.5702,
+      "step": 5517
+    },
+    {
+      "epoch": 0.5108372389052827,
+      "grad_norm": 0.146484375,
+      "learning_rate": 0.02,
+      "loss": 1.6146,
+      "step": 5518
+    },
+    {
+      "epoch": 0.5109298154255627,
+      "grad_norm": 0.1513671875,
+      "learning_rate": 0.02,
+      "loss": 1.5855,
+      "step": 5519
+    },
+    {
+      "epoch": 0.5110223919458428,
+      "grad_norm": 0.15234375,
+      "learning_rate": 0.02,
+      "loss": 1.6187,
+      "step": 5520
+    },
+    {
+      "epoch": 0.5111149684661228,
+      "grad_norm": 0.150390625,
+      "learning_rate": 0.02,
+      "loss": 1.5857,
+      "step": 5521
+    },
+    {
+      "epoch": 0.5112075449864029,
+      "grad_norm": 0.15625,
+      "learning_rate": 0.02,
+      "loss": 1.6048,
+      "step": 5522
+    },
+    {
+      "epoch": 0.5113001215066829,
+      "grad_norm": 0.1513671875,
+      "learning_rate": 0.02,
+      "loss": 1.5783,
+      "step": 5523
+    },
+    {
+      "epoch": 0.511392698026963,
+      "grad_norm": 0.15234375,
+      "learning_rate": 0.02,
+      "loss": 1.5406,
+      "step": 5524
+    },
+    {
+      "epoch": 0.5114852745472429,
+      "grad_norm": 0.142578125,
+      "learning_rate": 0.02,
+      "loss": 1.5278,
+      "step": 5525
+    },
+    {
+      "epoch": 0.511577851067523,
+      "grad_norm": 0.1533203125,
+      "learning_rate": 0.02,
+      "loss": 1.5315,
+      "step": 5526
+    },
+    {
+      "epoch": 0.511670427587803,
+      "grad_norm": 0.1435546875,
+      "learning_rate": 0.02,
+      "loss": 1.4632,
+      "step": 5527
+    },
+    {
+      "epoch": 0.5117630041080831,
+      "grad_norm": 0.1494140625,
+      "learning_rate": 0.02,
+      "loss": 1.5531,
+      "step": 5528
+    },
+    {
+      "epoch": 0.5118555806283631,
+      "grad_norm": 0.146484375,
+      "learning_rate": 0.02,
+      "loss": 1.5129,
+      "step": 5529
+    },
+    {
+      "epoch": 0.5119481571486432,
+      "grad_norm": 0.142578125,
+      "learning_rate": 0.02,
+      "loss": 1.4993,
+      "step": 5530
+    },
+    {
+      "epoch": 0.5120407336689232,
+      "grad_norm": 0.1328125,
+      "learning_rate": 0.02,
+      "loss": 1.6121,
+      "step": 5531
+    },
+    {
+      "epoch": 0.5121333101892033,
+      "grad_norm": 0.1533203125,
+      "learning_rate": 0.02,
+      "loss": 1.5603,
+      "step": 5532
+    },
+    {
+      "epoch": 0.5122258867094833,
+      "grad_norm": 0.15625,
+      "learning_rate": 0.02,
+      "loss": 1.582,
+      "step": 5533
+    },
+    {
+      "epoch": 0.5123184632297634,
+      "grad_norm": 0.1513671875,
+      "learning_rate": 0.02,
+      "loss": 1.5415,
+      "step": 5534
+    },
+    {
+      "epoch": 0.5124110397500434,
+      "grad_norm": 0.1396484375,
+      "learning_rate": 0.02,
+      "loss": 1.6253,
+      "step": 5535
+    },
+    {
+      "epoch": 0.5125036162703235,
+      "grad_norm": 0.146484375,
+      "learning_rate": 0.02,
+      "loss": 1.6506,
+      "step": 5536
+    },
+    {
+      "epoch": 0.5125961927906035,
+      "grad_norm": 0.15234375,
+      "learning_rate": 0.02,
+      "loss": 1.5405,
+      "step": 5537
+    },
+    {
+      "epoch": 0.5126887693108835,
+      "grad_norm": 0.1484375,
+      "learning_rate": 0.02,
+      "loss": 1.5412,
+      "step": 5538
+    },
+    {
+      "epoch": 0.5127813458311635,
+      "grad_norm": 0.1494140625,
+      "learning_rate": 0.02,
+      "loss": 1.5546,
+      "step": 5539
+    },
+    {
+      "epoch": 0.5128739223514436,
+      "grad_norm": 0.154296875,
+      "learning_rate": 0.02,
+      "loss": 1.535,
+      "step": 5540
+    },
+    {
+      "epoch": 0.5129664988717236,
+      "grad_norm": 0.14453125,
+      "learning_rate": 0.02,
+      "loss": 1.589,
+      "step": 5541
+    },
+    {
+      "epoch": 0.5130590753920037,
+      "grad_norm": 0.14453125,
+      "learning_rate": 0.02,
+      "loss": 1.5297,
+      "step": 5542
+    },
+    {
+      "epoch": 0.5131516519122837,
+      "grad_norm": 0.142578125,
+      "learning_rate": 0.02,
+      "loss": 1.5781,
+      "step": 5543
+    },
+    {
+      "epoch": 0.5132442284325638,
+      "grad_norm": 0.146484375,
+      "learning_rate": 0.02,
+      "loss": 1.5871,
+      "step": 5544
+    },
+    {
+      "epoch": 0.5133368049528438,
+      "grad_norm": 0.1357421875,
+      "learning_rate": 0.02,
+      "loss": 1.5262,
+      "step": 5545
+    },
+    {
+      "epoch": 0.5134293814731239,
+      "grad_norm": 0.1494140625,
+      "learning_rate": 0.02,
+      "loss": 1.5194,
+      "step": 5546
+    },
+    {
+      "epoch": 0.5135219579934039,
+      "grad_norm": 0.12890625,
+      "learning_rate": 0.02,
+      "loss": 1.5894,
+      "step": 5547
+    },
+    {
+      "epoch": 0.513614534513684,
+      "grad_norm": 0.1494140625,
+      "learning_rate": 0.02,
+      "loss": 1.5414,
+      "step": 5548
+    },
+    {
+      "epoch": 0.513707111033964,
+      "grad_norm": 0.1494140625,
+      "learning_rate": 0.02,
+      "loss": 1.5235,
+      "step": 5549
+    },
+    {
+      "epoch": 0.5137996875542441,
+      "grad_norm": 0.150390625,
+      "learning_rate": 0.02,
+      "loss": 1.5574,
+      "step": 5550
+    },
+    {
+      "epoch": 0.5138922640745242,
+      "grad_norm": 0.1376953125,
+      "learning_rate": 0.02,
+      "loss": 1.4927,
+      "step": 5551
+    },
+    {
+      "epoch": 0.5139848405948041,
+      "grad_norm": 0.138671875,
+      "learning_rate": 0.02,
+      "loss": 1.5426,
+      "step": 5552
+    },
+    {
+      "epoch": 0.5140774171150841,
+      "grad_norm": 0.1376953125,
+      "learning_rate": 0.02,
+      "loss": 1.5162,
+      "step": 5553
+    },
+    {
+      "epoch": 0.5141699936353642,
+      "grad_norm": 0.15625,
+      "learning_rate": 0.02,
+      "loss": 1.5243,
+      "step": 5554
+    },
+    {
+      "epoch": 0.5142625701556442,
+      "grad_norm": 0.1328125,
+      "learning_rate": 0.02,
+      "loss": 1.5388,
+      "step": 5555
+    },
+    {
+      "epoch": 0.5143551466759243,
+      "grad_norm": 0.142578125,
+      "learning_rate": 0.02,
+      "loss": 1.5877,
+      "step": 5556
+    },
+    {
+      "epoch": 0.5144477231962044,
+      "grad_norm": 0.1376953125,
+      "learning_rate": 0.02,
+      "loss": 1.5154,
+      "step": 5557
+    },
+    {
+      "epoch": 0.5145402997164844,
+      "grad_norm": 0.146484375,
+      "learning_rate": 0.02,
+      "loss": 1.5702,
+      "step": 5558
+    },
+    {
+      "epoch": 0.5146328762367645,
+      "grad_norm": 0.1357421875,
+      "learning_rate": 0.02,
+      "loss": 1.5694,
+      "step": 5559
+    },
+    {
+      "epoch": 0.5147254527570445,
+      "grad_norm": 0.150390625,
+      "learning_rate": 0.02,
+      "loss": 1.6226,
+      "step": 5560
+    },
+    {
+      "epoch": 0.5148180292773246,
+      "grad_norm": 0.1455078125,
+      "learning_rate": 0.02,
+      "loss": 1.5816,
+      "step": 5561
+    },
+    {
+      "epoch": 0.5149106057976046,
+      "grad_norm": 0.154296875,
+      "learning_rate": 0.02,
+      "loss": 1.5232,
+      "step": 5562
+    },
+    {
+      "epoch": 0.5150031823178847,
+      "grad_norm": 0.1416015625,
+      "learning_rate": 0.02,
+      "loss": 1.6383,
+      "step": 5563
+    },
+    {
+      "epoch": 0.5150957588381647,
+      "grad_norm": 0.1474609375,
+      "learning_rate": 0.02,
+      "loss": 1.5337,
+      "step": 5564
+    },
+    {
+      "epoch": 0.5151883353584447,
+      "grad_norm": 0.1455078125,
+      "learning_rate": 0.02,
+      "loss": 1.6234,
+      "step": 5565
+    },
+    {
+      "epoch": 0.5152809118787247,
+      "grad_norm": 0.1455078125,
+      "learning_rate": 0.02,
+      "loss": 1.5013,
+      "step": 5566
+    },
+    {
+      "epoch": 0.5153734883990048,
+      "grad_norm": 0.1416015625,
+      "learning_rate": 0.02,
+      "loss": 1.5613,
+      "step": 5567
+    },
+    {
+      "epoch": 0.5154660649192848,
+      "grad_norm": 0.1435546875,
+      "learning_rate": 0.02,
+      "loss": 1.5847,
+      "step": 5568
+    },
+    {
+      "epoch": 0.5155586414395649,
+      "grad_norm": 0.1455078125,
+      "learning_rate": 0.02,
+      "loss": 1.5715,
+      "step": 5569
+    },
+    {
+      "epoch": 0.5156512179598449,
+      "grad_norm": 0.14453125,
+      "learning_rate": 0.02,
+      "loss": 1.6177,
+      "step": 5570
+    },
+    {
+      "epoch": 0.515743794480125,
+      "grad_norm": 0.150390625,
+      "learning_rate": 0.02,
+      "loss": 1.539,
+      "step": 5571
+    },
+    {
+      "epoch": 0.515836371000405,
+      "grad_norm": 0.1455078125,
+      "learning_rate": 0.02,
+      "loss": 1.5639,
+      "step": 5572
+    },
+    {
+      "epoch": 0.5159289475206851,
+      "grad_norm": 0.150390625,
+      "learning_rate": 0.02,
+      "loss": 1.5264,
+      "step": 5573
+    },
+    {
+      "epoch": 0.5160215240409651,
+      "grad_norm": 0.1533203125,
+      "learning_rate": 0.02,
+      "loss": 1.564,
+      "step": 5574
+    },
+    {
+      "epoch": 0.5161141005612452,
+      "grad_norm": 0.1435546875,
+      "learning_rate": 0.02,
+      "loss": 1.5835,
+      "step": 5575
+    },
+    {
+      "epoch": 0.5162066770815252,
+      "grad_norm": 0.1416015625,
+      "learning_rate": 0.02,
+      "loss": 1.5913,
+      "step": 5576
+    },
+    {
+      "epoch": 0.5162992536018053,
+      "grad_norm": 0.1494140625,
+      "learning_rate": 0.02,
+      "loss": 1.5821,
+      "step": 5577
+    },
+    {
+      "epoch": 0.5163918301220853,
+      "grad_norm": 0.150390625,
+      "learning_rate": 0.02,
+      "loss": 1.5878,
+      "step": 5578
+    },
+    {
+      "epoch": 0.5164844066423653,
+      "grad_norm": 0.1572265625,
+      "learning_rate": 0.02,
+      "loss": 1.5942,
+      "step": 5579
+    },
+    {
+      "epoch": 0.5165769831626453,
+      "grad_norm": 0.1474609375,
+      "learning_rate": 0.02,
+      "loss": 1.5124,
+      "step": 5580
+    },
+    {
+      "epoch": 0.5166695596829254,
+      "grad_norm": 0.1533203125,
+      "learning_rate": 0.02,
+      "loss": 1.6171,
+      "step": 5581
+    },
+    {
+      "epoch": 0.5167621362032054,
+      "grad_norm": 0.1474609375,
+      "learning_rate": 0.02,
+      "loss": 1.5279,
+      "step": 5582
+    },
+    {
+      "epoch": 0.5168547127234855,
+      "grad_norm": 0.146484375,
+      "learning_rate": 0.02,
+      "loss": 1.4688,
+      "step": 5583
+    },
+    {
+      "epoch": 0.5169472892437655,
+      "grad_norm": 0.1474609375,
+      "learning_rate": 0.02,
+      "loss": 1.5892,
+      "step": 5584
+    },
+    {
+      "epoch": 0.5170398657640456,
+      "grad_norm": 0.1591796875,
+      "learning_rate": 0.02,
+      "loss": 1.5418,
+      "step": 5585
+    },
+    {
+      "epoch": 0.5171324422843256,
+      "grad_norm": 0.142578125,
+      "learning_rate": 0.02,
+      "loss": 1.535,
+      "step": 5586
+    },
+    {
+      "epoch": 0.5172250188046057,
+      "grad_norm": 0.1474609375,
+      "learning_rate": 0.02,
+      "loss": 1.5462,
+      "step": 5587
+    },
+    {
+      "epoch": 0.5173175953248857,
+      "grad_norm": 0.142578125,
+      "learning_rate": 0.02,
+      "loss": 1.5502,
+      "step": 5588
+    },
+    {
+      "epoch": 0.5174101718451658,
+      "grad_norm": 0.1474609375,
+      "learning_rate": 0.02,
+      "loss": 1.5061,
+      "step": 5589
+    },
+    {
+      "epoch": 0.5175027483654459,
+      "grad_norm": 0.146484375,
+      "learning_rate": 0.02,
+      "loss": 1.5245,
+      "step": 5590
+    },
+    {
+      "epoch": 0.5175953248857259,
+      "grad_norm": 0.150390625,
+      "learning_rate": 0.02,
+      "loss": 1.5777,
+      "step": 5591
+    },
+    {
+      "epoch": 0.517687901406006,
+      "grad_norm": 0.1572265625,
+      "learning_rate": 0.02,
+      "loss": 1.5995,
+      "step": 5592
+    },
+    {
+      "epoch": 0.5177804779262859,
+      "grad_norm": 0.15625,
+      "learning_rate": 0.02,
+      "loss": 1.542,
+      "step": 5593
+    },
+    {
+      "epoch": 0.517873054446566,
+      "grad_norm": 0.1513671875,
+      "learning_rate": 0.02,
+      "loss": 1.5977,
+      "step": 5594
+    },
+    {
+      "epoch": 0.517965630966846,
+      "grad_norm": 0.154296875,
+      "learning_rate": 0.02,
+      "loss": 1.5656,
+      "step": 5595
+    },
+    {
+      "epoch": 0.518058207487126,
+      "grad_norm": 0.150390625,
+      "learning_rate": 0.02,
+      "loss": 1.5765,
+      "step": 5596
+    },
+    {
+      "epoch": 0.5181507840074061,
+      "grad_norm": 0.162109375,
+      "learning_rate": 0.02,
+      "loss": 1.6417,
+      "step": 5597
+    },
+    {
+      "epoch": 0.5182433605276862,
+      "grad_norm": 0.140625,
+      "learning_rate": 0.02,
+      "loss": 1.6526,
+      "step": 5598
+    },
+    {
+      "epoch": 0.5183359370479662,
+      "grad_norm": 0.1484375,
+      "learning_rate": 0.02,
+      "loss": 1.501,
+      "step": 5599
+    },
+    {
+      "epoch": 0.5184285135682463,
+      "grad_norm": 0.1455078125,
+      "learning_rate": 0.02,
+      "loss": 1.5727,
+      "step": 5600
+    },
+    {
+      "epoch": 0.5185210900885263,
+      "grad_norm": 0.15234375,
+      "learning_rate": 0.02,
+      "loss": 1.5447,
+      "step": 5601
+    },
+    {
+      "epoch": 0.5186136666088064,
+      "grad_norm": 0.142578125,
+      "learning_rate": 0.02,
+      "loss": 1.5363,
+      "step": 5602
+    },
+    {
+      "epoch": 0.5187062431290864,
+      "grad_norm": 0.1591796875,
+      "learning_rate": 0.02,
+      "loss": 1.6159,
+      "step": 5603
+    },
+    {
+      "epoch": 0.5187988196493665,
+      "grad_norm": 0.1494140625,
+      "learning_rate": 0.02,
+      "loss": 1.5919,
+      "step": 5604
+    },
+    {
+      "epoch": 0.5188913961696465,
+      "grad_norm": 0.1494140625,
+      "learning_rate": 0.02,
+      "loss": 1.5805,
+      "step": 5605
+    },
+    {
+      "epoch": 0.5189839726899265,
+      "grad_norm": 0.138671875,
+      "learning_rate": 0.02,
+      "loss": 1.6058,
+      "step": 5606
+    },
+    {
+      "epoch": 0.5190765492102065,
+      "grad_norm": 0.1435546875,
+      "learning_rate": 0.02,
+      "loss": 1.5873,
+      "step": 5607
+    },
+    {
+      "epoch": 0.5191691257304866,
+      "grad_norm": 0.1474609375,
+      "learning_rate": 0.02,
+      "loss": 1.5949,
+      "step": 5608
+    },
+    {
+      "epoch": 0.5192617022507666,
+      "grad_norm": 0.1533203125,
+      "learning_rate": 0.02,
+      "loss": 1.5869,
+      "step": 5609
+    },
+    {
+      "epoch": 0.5193542787710467,
+      "grad_norm": 0.150390625,
+      "learning_rate": 0.02,
+      "loss": 1.621,
+      "step": 5610
+    },
+    {
+      "epoch": 0.5194468552913267,
+      "grad_norm": 0.158203125,
+      "learning_rate": 0.02,
+      "loss": 1.6007,
+      "step": 5611
+    },
+    {
+      "epoch": 0.5195394318116068,
+      "grad_norm": 0.150390625,
+      "learning_rate": 0.02,
+      "loss": 1.5925,
+      "step": 5612
+    },
+    {
+      "epoch": 0.5196320083318868,
+      "grad_norm": 0.1396484375,
+      "learning_rate": 0.02,
+      "loss": 1.5582,
+      "step": 5613
+    },
+    {
+      "epoch": 0.5197245848521669,
+      "grad_norm": 0.15625,
+      "learning_rate": 0.02,
+      "loss": 1.5882,
+      "step": 5614
+    },
+    {
+      "epoch": 0.5198171613724469,
+      "grad_norm": 0.1572265625,
+      "learning_rate": 0.02,
+      "loss": 1.515,
+      "step": 5615
+    },
+    {
+      "epoch": 0.519909737892727,
+      "grad_norm": 0.15625,
+      "learning_rate": 0.02,
+      "loss": 1.5285,
+      "step": 5616
+    },
+    {
+      "epoch": 0.520002314413007,
+      "grad_norm": 0.1357421875,
+      "learning_rate": 0.02,
+      "loss": 1.5238,
+      "step": 5617
+    },
+    {
+      "epoch": 0.5200948909332871,
+      "grad_norm": 0.15234375,
+      "learning_rate": 0.02,
+      "loss": 1.5652,
+      "step": 5618
+    },
+    {
+      "epoch": 0.5201874674535671,
+      "grad_norm": 0.1416015625,
+      "learning_rate": 0.02,
+      "loss": 1.5393,
+      "step": 5619
+    },
+    {
+      "epoch": 0.5202800439738471,
+      "grad_norm": 0.15625,
+      "learning_rate": 0.02,
+      "loss": 1.5613,
+      "step": 5620
+    },
+    {
+      "epoch": 0.5203726204941271,
+      "grad_norm": 0.140625,
+      "learning_rate": 0.02,
+      "loss": 1.5867,
+      "step": 5621
+    },
+    {
+      "epoch": 0.5204651970144072,
+      "grad_norm": 0.1533203125,
+      "learning_rate": 0.02,
+      "loss": 1.5207,
+      "step": 5622
+    },
+    {
+      "epoch": 0.5205577735346872,
+      "grad_norm": 0.142578125,
+      "learning_rate": 0.02,
+      "loss": 1.5611,
+      "step": 5623
+    },
+    {
+      "epoch": 0.5206503500549673,
+      "grad_norm": 0.1533203125,
+      "learning_rate": 0.02,
+      "loss": 1.5276,
+      "step": 5624
+    },
+    {
+      "epoch": 0.5207429265752473,
+      "grad_norm": 0.14453125,
+      "learning_rate": 0.02,
+      "loss": 1.5444,
+      "step": 5625
+    },
+    {
+      "epoch": 0.5208355030955274,
+      "grad_norm": 0.1455078125,
+      "learning_rate": 0.02,
+      "loss": 1.6065,
+      "step": 5626
+    },
+    {
+      "epoch": 0.5209280796158074,
+      "grad_norm": 0.146484375,
+      "learning_rate": 0.02,
+      "loss": 1.5021,
+      "step": 5627
+    },
+    {
+      "epoch": 0.5210206561360875,
+      "grad_norm": 0.154296875,
+      "learning_rate": 0.02,
+      "loss": 1.5988,
+      "step": 5628
+    },
+    {
+      "epoch": 0.5211132326563676,
+      "grad_norm": 0.1494140625,
+      "learning_rate": 0.02,
+      "loss": 1.5045,
+      "step": 5629
+    },
+    {
+      "epoch": 0.5212058091766476,
+      "grad_norm": 0.1640625,
+      "learning_rate": 0.02,
+      "loss": 1.63,
+      "step": 5630
+    },
+    {
+      "epoch": 0.5212983856969277,
+      "grad_norm": 0.134765625,
+      "learning_rate": 0.02,
+      "loss": 1.4858,
+      "step": 5631
+    },
+    {
+      "epoch": 0.5213909622172077,
+      "grad_norm": 0.1494140625,
+      "learning_rate": 0.02,
+      "loss": 1.6249,
+      "step": 5632
+    },
+    {
+      "epoch": 0.5214835387374877,
+      "grad_norm": 0.1572265625,
+      "learning_rate": 0.02,
+      "loss": 1.5701,
+      "step": 5633
+    },
+    {
+      "epoch": 0.5215761152577677,
+      "grad_norm": 0.154296875,
+      "learning_rate": 0.02,
+      "loss": 1.5857,
+      "step": 5634
+    },
+    {
+      "epoch": 0.5216686917780478,
+      "grad_norm": 0.1494140625,
+      "learning_rate": 0.02,
+      "loss": 1.5783,
+      "step": 5635
+    },
+    {
+      "epoch": 0.5217612682983278,
+      "grad_norm": 0.150390625,
+      "learning_rate": 0.02,
+      "loss": 1.587,
+      "step": 5636
+    },
+    {
+      "epoch": 0.5218538448186079,
+      "grad_norm": 0.1416015625,
+      "learning_rate": 0.02,
+      "loss": 1.5809,
+      "step": 5637
+    },
+    {
+      "epoch": 0.5219464213388879,
+      "grad_norm": 0.1435546875,
+      "learning_rate": 0.02,
+      "loss": 1.5329,
+      "step": 5638
+    },
+    {
+      "epoch": 0.522038997859168,
+      "grad_norm": 0.1357421875,
+      "learning_rate": 0.02,
+      "loss": 1.576,
+      "step": 5639
+    },
+    {
+      "epoch": 0.522131574379448,
+      "grad_norm": 0.1435546875,
+      "learning_rate": 0.02,
+      "loss": 1.5481,
+      "step": 5640
+    },
+    {
+      "epoch": 0.5222241508997281,
+      "grad_norm": 0.142578125,
+      "learning_rate": 0.02,
+      "loss": 1.5401,
+      "step": 5641
+    },
+    {
+      "epoch": 0.5223167274200081,
+      "grad_norm": 0.1455078125,
+      "learning_rate": 0.02,
+      "loss": 1.5676,
+      "step": 5642
+    },
+    {
+      "epoch": 0.5224093039402882,
+      "grad_norm": 0.140625,
+      "learning_rate": 0.02,
+      "loss": 1.5298,
+      "step": 5643
+    },
+    {
+      "epoch": 0.5225018804605682,
+      "grad_norm": 0.146484375,
+      "learning_rate": 0.02,
+      "loss": 1.5904,
+      "step": 5644
+    },
+    {
+      "epoch": 0.5225944569808483,
+      "grad_norm": 0.1484375,
+      "learning_rate": 0.02,
+      "loss": 1.5045,
+      "step": 5645
+    },
+    {
+      "epoch": 0.5226870335011283,
+      "grad_norm": 0.142578125,
+      "learning_rate": 0.02,
+      "loss": 1.5447,
+      "step": 5646
+    },
+    {
+      "epoch": 0.5227796100214083,
+      "grad_norm": 0.1455078125,
+      "learning_rate": 0.02,
+      "loss": 1.5638,
+      "step": 5647
+    },
+    {
+      "epoch": 0.5228721865416883,
+      "grad_norm": 0.146484375,
+      "learning_rate": 0.02,
+      "loss": 1.6583,
+      "step": 5648
+    },
+    {
+      "epoch": 0.5229647630619684,
+      "grad_norm": 0.1376953125,
+      "learning_rate": 0.02,
+      "loss": 1.5409,
+      "step": 5649
+    },
+    {
+      "epoch": 0.5230573395822484,
+      "grad_norm": 0.140625,
+      "learning_rate": 0.02,
+      "loss": 1.5308,
+      "step": 5650
+    },
+    {
+      "epoch": 0.5231499161025285,
+      "grad_norm": 0.146484375,
+      "learning_rate": 0.02,
+      "loss": 1.5416,
+      "step": 5651
+    },
+    {
+      "epoch": 0.5232424926228085,
+      "grad_norm": 0.142578125,
+      "learning_rate": 0.02,
+      "loss": 1.54,
+      "step": 5652
+    },
+    {
+      "epoch": 0.5233350691430886,
+      "grad_norm": 0.140625,
+      "learning_rate": 0.02,
+      "loss": 1.5503,
+      "step": 5653
+    },
+    {
+      "epoch": 0.5234276456633686,
+      "grad_norm": 0.150390625,
+      "learning_rate": 0.02,
+      "loss": 1.5672,
+      "step": 5654
+    },
+    {
+      "epoch": 0.5235202221836487,
+      "grad_norm": 0.1435546875,
+      "learning_rate": 0.02,
+      "loss": 1.5595,
+      "step": 5655
+    },
+    {
+      "epoch": 0.5236127987039287,
+      "grad_norm": 0.1455078125,
+      "learning_rate": 0.02,
+      "loss": 1.5893,
+      "step": 5656
+    },
+    {
+      "epoch": 0.5237053752242088,
+      "grad_norm": 0.142578125,
+      "learning_rate": 0.02,
+      "loss": 1.6014,
+      "step": 5657
+    },
+    {
+      "epoch": 0.5237979517444888,
+      "grad_norm": 0.158203125,
+      "learning_rate": 0.02,
+      "loss": 1.6032,
+      "step": 5658
+    },
+    {
+      "epoch": 0.5238905282647689,
+      "grad_norm": 0.154296875,
+      "learning_rate": 0.02,
+      "loss": 1.5902,
+      "step": 5659
+    },
+    {
+      "epoch": 0.5239831047850488,
+      "grad_norm": 0.1513671875,
+      "learning_rate": 0.02,
+      "loss": 1.4611,
+      "step": 5660
+    },
+    {
+      "epoch": 0.5240756813053289,
+      "grad_norm": 0.1416015625,
+      "learning_rate": 0.02,
+      "loss": 1.6066,
+      "step": 5661
+    },
+    {
+      "epoch": 0.5241682578256089,
+      "grad_norm": 0.154296875,
+      "learning_rate": 0.02,
+      "loss": 1.5992,
+      "step": 5662
+    },
+    {
+      "epoch": 0.524260834345889,
+      "grad_norm": 0.162109375,
+      "learning_rate": 0.02,
+      "loss": 1.5699,
+      "step": 5663
+    },
+    {
+      "epoch": 0.524353410866169,
+      "grad_norm": 0.1591796875,
+      "learning_rate": 0.02,
+      "loss": 1.5554,
+      "step": 5664
+    },
+    {
+      "epoch": 0.5244459873864491,
+      "grad_norm": 0.154296875,
+      "learning_rate": 0.02,
+      "loss": 1.6391,
+      "step": 5665
+    },
+    {
+      "epoch": 0.5245385639067291,
+      "grad_norm": 0.1591796875,
+      "learning_rate": 0.02,
+      "loss": 1.5671,
+      "step": 5666
+    },
+    {
+      "epoch": 0.5246311404270092,
+      "grad_norm": 0.1494140625,
+      "learning_rate": 0.02,
+      "loss": 1.5848,
+      "step": 5667
+    },
+    {
+      "epoch": 0.5247237169472893,
+      "grad_norm": 0.1435546875,
+      "learning_rate": 0.02,
+      "loss": 1.5572,
+      "step": 5668
+    },
+    {
+      "epoch": 0.5248162934675693,
+      "grad_norm": 0.150390625,
+      "learning_rate": 0.02,
+      "loss": 1.5508,
+      "step": 5669
+    },
+    {
+      "epoch": 0.5249088699878494,
+      "grad_norm": 0.142578125,
+      "learning_rate": 0.02,
+      "loss": 1.6037,
+      "step": 5670
+    },
+    {
+      "epoch": 0.5250014465081294,
+      "grad_norm": 0.146484375,
+      "learning_rate": 0.02,
+      "loss": 1.5491,
+      "step": 5671
+    },
+    {
+      "epoch": 0.5250940230284095,
+      "grad_norm": 0.1455078125,
+      "learning_rate": 0.02,
+      "loss": 1.5112,
+      "step": 5672
+    },
+    {
+      "epoch": 0.5251865995486895,
+      "grad_norm": 0.1474609375,
+      "learning_rate": 0.02,
+      "loss": 1.5568,
+      "step": 5673
+    },
+    {
+      "epoch": 0.5252791760689695,
+      "grad_norm": 0.1533203125,
+      "learning_rate": 0.02,
+      "loss": 1.5084,
+      "step": 5674
+    },
+    {
+      "epoch": 0.5253717525892495,
+      "grad_norm": 0.158203125,
+      "learning_rate": 0.02,
+      "loss": 1.533,
+      "step": 5675
+    },
+    {
+      "epoch": 0.5254643291095296,
+      "grad_norm": 0.1513671875,
+      "learning_rate": 0.02,
+      "loss": 1.6157,
+      "step": 5676
+    },
+    {
+      "epoch": 0.5255569056298096,
+      "grad_norm": 0.154296875,
+      "learning_rate": 0.02,
+      "loss": 1.5642,
+      "step": 5677
+    },
+    {
+      "epoch": 0.5256494821500897,
+      "grad_norm": 0.1533203125,
+      "learning_rate": 0.02,
+      "loss": 1.6299,
+      "step": 5678
+    },
+    {
+      "epoch": 0.5257420586703697,
+      "grad_norm": 0.1494140625,
+      "learning_rate": 0.02,
+      "loss": 1.6029,
+      "step": 5679
+    },
+    {
+      "epoch": 0.5258346351906498,
+      "grad_norm": 0.1591796875,
+      "learning_rate": 0.02,
+      "loss": 1.5389,
+      "step": 5680
+    },
+    {
+      "epoch": 0.5259272117109298,
+      "grad_norm": 0.15625,
+      "learning_rate": 0.02,
+      "loss": 1.6055,
+      "step": 5681
+    },
+    {
+      "epoch": 0.5260197882312099,
+      "grad_norm": 0.1533203125,
+      "learning_rate": 0.02,
+      "loss": 1.6091,
+      "step": 5682
+    },
+    {
+      "epoch": 0.5261123647514899,
+      "grad_norm": 0.1494140625,
+      "learning_rate": 0.02,
+      "loss": 1.5225,
+      "step": 5683
+    },
+    {
+      "epoch": 0.52620494127177,
+      "grad_norm": 0.14453125,
+      "learning_rate": 0.02,
+      "loss": 1.5242,
+      "step": 5684
+    },
+    {
+      "epoch": 0.52629751779205,
+      "grad_norm": 0.1396484375,
+      "learning_rate": 0.02,
+      "loss": 1.5286,
+      "step": 5685
+    },
+    {
+      "epoch": 0.5263900943123301,
+      "grad_norm": 0.1357421875,
+      "learning_rate": 0.02,
+      "loss": 1.4955,
+      "step": 5686
+    },
+    {
+      "epoch": 0.5264826708326101,
+      "grad_norm": 0.1435546875,
+      "learning_rate": 0.02,
+      "loss": 1.525,
+      "step": 5687
+    },
+    {
+      "epoch": 0.5265752473528901,
+      "grad_norm": 0.1357421875,
+      "learning_rate": 0.02,
+      "loss": 1.568,
+      "step": 5688
+    },
+    {
+      "epoch": 0.5266678238731701,
+      "grad_norm": 0.146484375,
+      "learning_rate": 0.02,
+      "loss": 1.6592,
+      "step": 5689
+    },
+    {
+      "epoch": 0.5267604003934502,
+      "grad_norm": 0.130859375,
+      "learning_rate": 0.02,
+      "loss": 1.511,
+      "step": 5690
+    },
+    {
+      "epoch": 0.5268529769137302,
+      "grad_norm": 0.134765625,
+      "learning_rate": 0.02,
+      "loss": 1.5322,
+      "step": 5691
+    },
+    {
+      "epoch": 0.5269455534340103,
+      "grad_norm": 0.1376953125,
+      "learning_rate": 0.02,
+      "loss": 1.6176,
+      "step": 5692
+    },
+    {
+      "epoch": 0.5270381299542903,
+      "grad_norm": 0.14453125,
+      "learning_rate": 0.02,
+      "loss": 1.5095,
+      "step": 5693
+    },
+    {
+      "epoch": 0.5271307064745704,
+      "grad_norm": 0.1416015625,
+      "learning_rate": 0.02,
+      "loss": 1.5558,
+      "step": 5694
+    },
+    {
+      "epoch": 0.5272232829948504,
+      "grad_norm": 0.150390625,
+      "learning_rate": 0.02,
+      "loss": 1.5321,
+      "step": 5695
+    },
+    {
+      "epoch": 0.5273158595151305,
+      "grad_norm": 0.146484375,
+      "learning_rate": 0.02,
+      "loss": 1.5475,
+      "step": 5696
+    },
+    {
+      "epoch": 0.5274084360354105,
+      "grad_norm": 0.1513671875,
+      "learning_rate": 0.02,
+      "loss": 1.525,
+      "step": 5697
+    },
+    {
+      "epoch": 0.5275010125556906,
+      "grad_norm": 0.1396484375,
+      "learning_rate": 0.02,
+      "loss": 1.5447,
+      "step": 5698
+    },
+    {
+      "epoch": 0.5275935890759706,
+      "grad_norm": 0.1513671875,
+      "learning_rate": 0.02,
+      "loss": 1.5739,
+      "step": 5699
+    },
+    {
+      "epoch": 0.5276861655962507,
+      "grad_norm": 0.1455078125,
+      "learning_rate": 0.02,
+      "loss": 1.5947,
+      "step": 5700
+    },
+    {
+      "epoch": 0.5277787421165306,
+      "grad_norm": 0.158203125,
+      "learning_rate": 0.02,
+      "loss": 1.5814,
+      "step": 5701
+    },
+    {
+      "epoch": 0.5278713186368107,
+      "grad_norm": 0.146484375,
+      "learning_rate": 0.02,
+      "loss": 1.5637,
+      "step": 5702
+    },
+    {
+      "epoch": 0.5279638951570907,
+      "grad_norm": 0.16015625,
+      "learning_rate": 0.02,
+      "loss": 1.542,
+      "step": 5703
+    },
+    {
+      "epoch": 0.5280564716773708,
+      "grad_norm": 0.14453125,
+      "learning_rate": 0.02,
+      "loss": 1.6247,
+      "step": 5704
+    },
+    {
+      "epoch": 0.5281490481976508,
+      "grad_norm": 0.1513671875,
+      "learning_rate": 0.02,
+      "loss": 1.5731,
+      "step": 5705
+    },
+    {
+      "epoch": 0.5282416247179309,
+      "grad_norm": 0.146484375,
+      "learning_rate": 0.02,
+      "loss": 1.5614,
+      "step": 5706
+    },
+    {
+      "epoch": 0.528334201238211,
+      "grad_norm": 0.1455078125,
+      "learning_rate": 0.02,
+      "loss": 1.5748,
+      "step": 5707
+    },
+    {
+      "epoch": 0.528426777758491,
+      "grad_norm": 0.154296875,
+      "learning_rate": 0.02,
+      "loss": 1.6907,
+      "step": 5708
+    },
+    {
+      "epoch": 0.5285193542787711,
+      "grad_norm": 0.1494140625,
+      "learning_rate": 0.02,
+      "loss": 1.575,
+      "step": 5709
+    },
+    {
+      "epoch": 0.5286119307990511,
+      "grad_norm": 0.1474609375,
+      "learning_rate": 0.02,
+      "loss": 1.5058,
+      "step": 5710
+    },
+    {
+      "epoch": 0.5287045073193312,
+      "grad_norm": 0.1455078125,
+      "learning_rate": 0.02,
+      "loss": 1.5229,
+      "step": 5711
+    },
+    {
+      "epoch": 0.5287970838396112,
+      "grad_norm": 0.1494140625,
+      "learning_rate": 0.02,
+      "loss": 1.5645,
+      "step": 5712
+    },
+    {
+      "epoch": 0.5288896603598913,
+      "grad_norm": 0.1337890625,
+      "learning_rate": 0.02,
+      "loss": 1.5277,
+      "step": 5713
+    },
+    {
+      "epoch": 0.5289822368801713,
+      "grad_norm": 0.1455078125,
+      "learning_rate": 0.02,
+      "loss": 1.6384,
+      "step": 5714
+    },
+    {
+      "epoch": 0.5290748134004513,
+      "grad_norm": 0.1455078125,
+      "learning_rate": 0.02,
+      "loss": 1.5712,
+      "step": 5715
+    },
+    {
+      "epoch": 0.5291673899207313,
+      "grad_norm": 0.1572265625,
+      "learning_rate": 0.02,
+      "loss": 1.5863,
+      "step": 5716
+    },
+    {
+      "epoch": 0.5292599664410114,
+      "grad_norm": 0.1494140625,
+      "learning_rate": 0.02,
+      "loss": 1.4992,
+      "step": 5717
+    },
+    {
+      "epoch": 0.5293525429612914,
+      "grad_norm": 0.158203125,
+      "learning_rate": 0.02,
+      "loss": 1.589,
+      "step": 5718
+    },
+    {
+      "epoch": 0.5294451194815715,
+      "grad_norm": 0.13671875,
+      "learning_rate": 0.02,
+      "loss": 1.5657,
+      "step": 5719
+    },
+    {
+      "epoch": 0.5295376960018515,
+      "grad_norm": 0.154296875,
+      "learning_rate": 0.02,
+      "loss": 1.6115,
+      "step": 5720
+    },
+    {
+      "epoch": 0.5296302725221316,
+      "grad_norm": 0.146484375,
+      "learning_rate": 0.02,
+      "loss": 1.5699,
+      "step": 5721
+    },
+    {
+      "epoch": 0.5297228490424116,
+      "grad_norm": 0.14453125,
+      "learning_rate": 0.02,
+      "loss": 1.5456,
+      "step": 5722
+    },
+    {
+      "epoch": 0.5298154255626917,
+      "grad_norm": 0.1435546875,
+      "learning_rate": 0.02,
+      "loss": 1.5527,
+      "step": 5723
+    },
+    {
+      "epoch": 0.5299080020829717,
+      "grad_norm": 0.1513671875,
+      "learning_rate": 0.02,
+      "loss": 1.5857,
+      "step": 5724
+    },
+    {
+      "epoch": 0.5300005786032518,
+      "grad_norm": 0.146484375,
+      "learning_rate": 0.02,
+      "loss": 1.5903,
+      "step": 5725
+    },
+    {
+      "epoch": 0.5300931551235318,
+      "grad_norm": 0.15625,
+      "learning_rate": 0.02,
+      "loss": 1.5664,
+      "step": 5726
+    },
+    {
+      "epoch": 0.5301857316438119,
+      "grad_norm": 0.1533203125,
+      "learning_rate": 0.02,
+      "loss": 1.542,
+      "step": 5727
+    },
+    {
+      "epoch": 0.5302783081640918,
+      "grad_norm": 0.14453125,
+      "learning_rate": 0.02,
+      "loss": 1.5147,
+      "step": 5728
+    },
+    {
+      "epoch": 0.5303708846843719,
+      "grad_norm": 0.1513671875,
+      "learning_rate": 0.02,
+      "loss": 1.4981,
+      "step": 5729
+    },
+    {
+      "epoch": 0.5304634612046519,
+      "grad_norm": 0.15625,
+      "learning_rate": 0.02,
+      "loss": 1.5373,
+      "step": 5730
+    },
+    {
+      "epoch": 0.530556037724932,
+      "grad_norm": 0.1591796875,
+      "learning_rate": 0.02,
+      "loss": 1.5503,
+      "step": 5731
+    },
+    {
+      "epoch": 0.530648614245212,
+      "grad_norm": 0.1484375,
+      "learning_rate": 0.02,
+      "loss": 1.552,
+      "step": 5732
+    },
+    {
+      "epoch": 0.5307411907654921,
+      "grad_norm": 0.1552734375,
+      "learning_rate": 0.02,
+      "loss": 1.5609,
+      "step": 5733
+    },
+    {
+      "epoch": 0.5308337672857721,
+      "grad_norm": 0.1376953125,
+      "learning_rate": 0.02,
+      "loss": 1.5659,
+      "step": 5734
+    },
+    {
+      "epoch": 0.5309263438060522,
+      "grad_norm": 0.1513671875,
+      "learning_rate": 0.02,
+      "loss": 1.5689,
+      "step": 5735
+    },
+    {
+      "epoch": 0.5310189203263322,
+      "grad_norm": 0.1455078125,
+      "learning_rate": 0.02,
+      "loss": 1.5796,
+      "step": 5736
+    },
+    {
+      "epoch": 0.5311114968466123,
+      "grad_norm": 0.138671875,
+      "learning_rate": 0.02,
+      "loss": 1.4686,
+      "step": 5737
+    },
+    {
+      "epoch": 0.5312040733668923,
+      "grad_norm": 0.1494140625,
+      "learning_rate": 0.02,
+      "loss": 1.5829,
+      "step": 5738
+    },
+    {
+      "epoch": 0.5312966498871724,
+      "grad_norm": 0.14453125,
+      "learning_rate": 0.02,
+      "loss": 1.5861,
+      "step": 5739
+    },
+    {
+      "epoch": 0.5313892264074525,
+      "grad_norm": 0.150390625,
+      "learning_rate": 0.02,
+      "loss": 1.5235,
+      "step": 5740
+    },
+    {
+      "epoch": 0.5314818029277325,
+      "grad_norm": 0.1572265625,
+      "learning_rate": 0.02,
+      "loss": 1.5315,
+      "step": 5741
+    },
+    {
+      "epoch": 0.5315743794480124,
+      "grad_norm": 0.158203125,
+      "learning_rate": 0.02,
+      "loss": 1.5576,
+      "step": 5742
+    },
+    {
+      "epoch": 0.5316669559682925,
+      "grad_norm": 0.16015625,
+      "learning_rate": 0.02,
+      "loss": 1.5559,
+      "step": 5743
+    },
+    {
+      "epoch": 0.5317595324885726,
+      "grad_norm": 0.1552734375,
+      "learning_rate": 0.02,
+      "loss": 1.5712,
+      "step": 5744
+    },
+    {
+      "epoch": 0.5318521090088526,
+      "grad_norm": 0.1572265625,
+      "learning_rate": 0.02,
+      "loss": 1.5565,
+      "step": 5745
+    },
+    {
+      "epoch": 0.5319446855291327,
+      "grad_norm": 0.1494140625,
+      "learning_rate": 0.02,
+      "loss": 1.5852,
+      "step": 5746
+    },
+    {
+      "epoch": 0.5320372620494127,
+      "grad_norm": 0.154296875,
+      "learning_rate": 0.02,
+      "loss": 1.5734,
+      "step": 5747
+    },
+    {
+      "epoch": 0.5321298385696928,
+      "grad_norm": 0.1630859375,
+      "learning_rate": 0.02,
+      "loss": 1.5662,
+      "step": 5748
+    },
+    {
+      "epoch": 0.5322224150899728,
+      "grad_norm": 0.1591796875,
+      "learning_rate": 0.02,
+      "loss": 1.6231,
+      "step": 5749
+    },
+    {
+      "epoch": 0.5323149916102529,
+      "grad_norm": 0.140625,
+      "learning_rate": 0.02,
+      "loss": 1.6235,
+      "step": 5750
+    },
+    {
+      "epoch": 0.5324075681305329,
+      "grad_norm": 0.1533203125,
+      "learning_rate": 0.02,
+      "loss": 1.5688,
+      "step": 5751
+    },
+    {
+      "epoch": 0.532500144650813,
+      "grad_norm": 0.1513671875,
+      "learning_rate": 0.02,
+      "loss": 1.6294,
+      "step": 5752
+    },
+    {
+      "epoch": 0.532592721171093,
+      "grad_norm": 0.14453125,
+      "learning_rate": 0.02,
+      "loss": 1.6234,
+      "step": 5753
+    },
+    {
+      "epoch": 0.5326852976913731,
+      "grad_norm": 0.14453125,
+      "learning_rate": 0.02,
+      "loss": 1.5155,
+      "step": 5754
+    },
+    {
+      "epoch": 0.5327778742116531,
+      "grad_norm": 0.134765625,
+      "learning_rate": 0.02,
+      "loss": 1.5926,
+      "step": 5755
+    },
+    {
+      "epoch": 0.5328704507319331,
+      "grad_norm": 0.1416015625,
+      "learning_rate": 0.02,
+      "loss": 1.4946,
+      "step": 5756
+    },
+    {
+      "epoch": 0.5329630272522131,
+      "grad_norm": 0.1455078125,
+      "learning_rate": 0.02,
+      "loss": 1.5543,
+      "step": 5757
+    },
+    {
+      "epoch": 0.5330556037724932,
+      "grad_norm": 0.1435546875,
+      "learning_rate": 0.02,
+      "loss": 1.5824,
+      "step": 5758
+    },
+    {
+      "epoch": 0.5331481802927732,
+      "grad_norm": 0.134765625,
+      "learning_rate": 0.02,
+      "loss": 1.5393,
+      "step": 5759
+    },
+    {
+      "epoch": 0.5332407568130533,
+      "grad_norm": 0.1572265625,
+      "learning_rate": 0.02,
+      "loss": 1.5693,
+      "step": 5760
+    },
+    {
+      "epoch": 0.5333333333333333,
+      "grad_norm": 0.15234375,
+      "learning_rate": 0.02,
+      "loss": 1.5481,
+      "step": 5761
+    },
+    {
+      "epoch": 0.5334259098536134,
+      "grad_norm": 0.1484375,
+      "learning_rate": 0.02,
+      "loss": 1.5586,
+      "step": 5762
+    },
+    {
+      "epoch": 0.5335184863738934,
+      "grad_norm": 0.142578125,
+      "learning_rate": 0.02,
+      "loss": 1.5419,
+      "step": 5763
+    },
+    {
+      "epoch": 0.5336110628941735,
+      "grad_norm": 0.146484375,
+      "learning_rate": 0.02,
+      "loss": 1.5489,
+      "step": 5764
+    },
+    {
+      "epoch": 0.5337036394144535,
+      "grad_norm": 0.142578125,
+      "learning_rate": 0.02,
+      "loss": 1.5127,
+      "step": 5765
+    },
+    {
+      "epoch": 0.5337962159347336,
+      "grad_norm": 0.140625,
+      "learning_rate": 0.02,
+      "loss": 1.5612,
+      "step": 5766
+    },
+    {
+      "epoch": 0.5338887924550136,
+      "grad_norm": 0.1474609375,
+      "learning_rate": 0.02,
+      "loss": 1.4899,
+      "step": 5767
+    },
+    {
+      "epoch": 0.5339813689752937,
+      "grad_norm": 0.1455078125,
+      "learning_rate": 0.02,
+      "loss": 1.5442,
+      "step": 5768
+    },
+    {
+      "epoch": 0.5340739454955736,
+      "grad_norm": 0.1337890625,
+      "learning_rate": 0.02,
+      "loss": 1.5307,
+      "step": 5769
+    },
+    {
+      "epoch": 0.5341665220158537,
+      "grad_norm": 0.1474609375,
+      "learning_rate": 0.02,
+      "loss": 1.5282,
+      "step": 5770
+    },
+    {
+      "epoch": 0.5342590985361337,
+      "grad_norm": 0.15234375,
+      "learning_rate": 0.02,
+      "loss": 1.6234,
+      "step": 5771
+    },
+    {
+      "epoch": 0.5343516750564138,
+      "grad_norm": 0.154296875,
+      "learning_rate": 0.02,
+      "loss": 1.5161,
+      "step": 5772
+    },
+    {
+      "epoch": 0.5344442515766938,
+      "grad_norm": 0.1455078125,
+      "learning_rate": 0.02,
+      "loss": 1.5619,
+      "step": 5773
+    },
+    {
+      "epoch": 0.5345368280969739,
+      "grad_norm": 0.15234375,
+      "learning_rate": 0.02,
+      "loss": 1.4989,
+      "step": 5774
+    },
+    {
+      "epoch": 0.534629404617254,
+      "grad_norm": 0.1533203125,
+      "learning_rate": 0.02,
+      "loss": 1.5622,
+      "step": 5775
+    },
+    {
+      "epoch": 0.534721981137534,
+      "grad_norm": 0.154296875,
+      "learning_rate": 0.02,
+      "loss": 1.5197,
+      "step": 5776
+    },
+    {
+      "epoch": 0.534814557657814,
+      "grad_norm": 0.142578125,
+      "learning_rate": 0.02,
+      "loss": 1.5258,
+      "step": 5777
+    },
+    {
+      "epoch": 0.5349071341780941,
+      "grad_norm": 0.1484375,
+      "learning_rate": 0.02,
+      "loss": 1.5959,
+      "step": 5778
+    },
+    {
+      "epoch": 0.5349997106983742,
+      "grad_norm": 0.1474609375,
+      "learning_rate": 0.02,
+      "loss": 1.5726,
+      "step": 5779
+    },
+    {
+      "epoch": 0.5350922872186542,
+      "grad_norm": 0.1435546875,
+      "learning_rate": 0.02,
+      "loss": 1.5748,
+      "step": 5780
+    },
+    {
+      "epoch": 0.5351848637389343,
+      "grad_norm": 0.1474609375,
+      "learning_rate": 0.02,
+      "loss": 1.5471,
+      "step": 5781
+    },
+    {
+      "epoch": 0.5352774402592143,
+      "grad_norm": 0.1337890625,
+      "learning_rate": 0.02,
+      "loss": 1.5755,
+      "step": 5782
+    },
+    {
+      "epoch": 0.5353700167794943,
+      "grad_norm": 0.1376953125,
+      "learning_rate": 0.02,
+      "loss": 1.5694,
+      "step": 5783
+    },
+    {
+      "epoch": 0.5354625932997743,
+      "grad_norm": 0.142578125,
+      "learning_rate": 0.02,
+      "loss": 1.5967,
+      "step": 5784
+    },
+    {
+      "epoch": 0.5355551698200544,
+      "grad_norm": 0.146484375,
+      "learning_rate": 0.02,
+      "loss": 1.618,
+      "step": 5785
+    },
+    {
+      "epoch": 0.5356477463403344,
+      "grad_norm": 0.1484375,
+      "learning_rate": 0.02,
+      "loss": 1.622,
+      "step": 5786
+    },
+    {
+      "epoch": 0.5357403228606145,
+      "grad_norm": 0.1357421875,
+      "learning_rate": 0.02,
+      "loss": 1.5226,
+      "step": 5787
+    },
+    {
+      "epoch": 0.5358328993808945,
+      "grad_norm": 0.1396484375,
+      "learning_rate": 0.02,
+      "loss": 1.5578,
+      "step": 5788
+    },
+    {
+      "epoch": 0.5359254759011746,
+      "grad_norm": 0.1533203125,
+      "learning_rate": 0.02,
+      "loss": 1.5921,
+      "step": 5789
+    },
+    {
+      "epoch": 0.5360180524214546,
+      "grad_norm": 0.1552734375,
+      "learning_rate": 0.02,
+      "loss": 1.5601,
+      "step": 5790
+    },
+    {
+      "epoch": 0.5361106289417347,
+      "grad_norm": 0.146484375,
+      "learning_rate": 0.02,
+      "loss": 1.5137,
+      "step": 5791
+    },
+    {
+      "epoch": 0.5362032054620147,
+      "grad_norm": 0.158203125,
+      "learning_rate": 0.02,
+      "loss": 1.5625,
+      "step": 5792
+    },
+    {
+      "epoch": 0.5362957819822948,
+      "grad_norm": 0.14453125,
+      "learning_rate": 0.02,
+      "loss": 1.5131,
+      "step": 5793
+    },
+    {
+      "epoch": 0.5363883585025748,
+      "grad_norm": 0.1455078125,
+      "learning_rate": 0.02,
+      "loss": 1.5666,
+      "step": 5794
+    },
+    {
+      "epoch": 0.5364809350228549,
+      "grad_norm": 0.1337890625,
+      "learning_rate": 0.02,
+      "loss": 1.5945,
+      "step": 5795
+    },
+    {
+      "epoch": 0.5365735115431348,
+      "grad_norm": 0.1455078125,
+      "learning_rate": 0.02,
+      "loss": 1.5771,
+      "step": 5796
+    },
+    {
+      "epoch": 0.5366660880634149,
+      "grad_norm": 0.14453125,
+      "learning_rate": 0.02,
+      "loss": 1.5952,
+      "step": 5797
+    },
+    {
+      "epoch": 0.5367586645836949,
+      "grad_norm": 0.146484375,
+      "learning_rate": 0.02,
+      "loss": 1.5912,
+      "step": 5798
+    },
+    {
+      "epoch": 0.536851241103975,
+      "grad_norm": 0.1416015625,
+      "learning_rate": 0.02,
+      "loss": 1.5397,
+      "step": 5799
+    },
+    {
+      "epoch": 0.536943817624255,
+      "grad_norm": 0.138671875,
+      "learning_rate": 0.02,
+      "loss": 1.5784,
+      "step": 5800
+    },
+    {
+      "epoch": 0.5370363941445351,
+      "grad_norm": 0.1376953125,
+      "learning_rate": 0.02,
+      "loss": 1.5275,
+      "step": 5801
+    },
+    {
+      "epoch": 0.5371289706648151,
+      "grad_norm": 0.142578125,
+      "learning_rate": 0.02,
+      "loss": 1.5367,
+      "step": 5802
+    },
+    {
+      "epoch": 0.5372215471850952,
+      "grad_norm": 0.1533203125,
+      "learning_rate": 0.02,
+      "loss": 1.6292,
+      "step": 5803
+    },
+    {
+      "epoch": 0.5373141237053752,
+      "grad_norm": 0.138671875,
+      "learning_rate": 0.02,
+      "loss": 1.4849,
+      "step": 5804
+    },
+    {
+      "epoch": 0.5374067002256553,
+      "grad_norm": 0.1455078125,
+      "learning_rate": 0.02,
+      "loss": 1.5716,
+      "step": 5805
+    },
+    {
+      "epoch": 0.5374992767459353,
+      "grad_norm": 0.1494140625,
+      "learning_rate": 0.02,
+      "loss": 1.6047,
+      "step": 5806
+    },
+    {
+      "epoch": 0.5375918532662154,
+      "grad_norm": 0.1533203125,
+      "learning_rate": 0.02,
+      "loss": 1.5968,
+      "step": 5807
+    },
+    {
+      "epoch": 0.5376844297864954,
+      "grad_norm": 0.1376953125,
+      "learning_rate": 0.02,
+      "loss": 1.5546,
+      "step": 5808
+    },
+    {
+      "epoch": 0.5377770063067755,
+      "grad_norm": 0.1455078125,
+      "learning_rate": 0.02,
+      "loss": 1.5526,
+      "step": 5809
+    },
+    {
+      "epoch": 0.5378695828270554,
+      "grad_norm": 0.1357421875,
+      "learning_rate": 0.02,
+      "loss": 1.5768,
+      "step": 5810
+    },
+    {
+      "epoch": 0.5379621593473355,
+      "grad_norm": 0.150390625,
+      "learning_rate": 0.02,
+      "loss": 1.6045,
+      "step": 5811
+    },
+    {
+      "epoch": 0.5380547358676155,
+      "grad_norm": 0.1533203125,
+      "learning_rate": 0.02,
+      "loss": 1.5763,
+      "step": 5812
+    },
+    {
+      "epoch": 0.5381473123878956,
+      "grad_norm": 0.1650390625,
+      "learning_rate": 0.02,
+      "loss": 1.5734,
+      "step": 5813
+    },
+    {
+      "epoch": 0.5382398889081756,
+      "grad_norm": 0.142578125,
+      "learning_rate": 0.02,
+      "loss": 1.4989,
+      "step": 5814
+    },
+    {
+      "epoch": 0.5383324654284557,
+      "grad_norm": 0.1572265625,
+      "learning_rate": 0.02,
+      "loss": 1.5008,
+      "step": 5815
+    },
+    {
+      "epoch": 0.5384250419487357,
+      "grad_norm": 0.1572265625,
+      "learning_rate": 0.02,
+      "loss": 1.5593,
+      "step": 5816
+    },
+    {
+      "epoch": 0.5385176184690158,
+      "grad_norm": 0.166015625,
+      "learning_rate": 0.02,
+      "loss": 1.5773,
+      "step": 5817
+    },
+    {
+      "epoch": 0.5386101949892959,
+      "grad_norm": 0.142578125,
+      "learning_rate": 0.02,
+      "loss": 1.551,
+      "step": 5818
+    },
+    {
+      "epoch": 0.5387027715095759,
+      "grad_norm": 0.1611328125,
+      "learning_rate": 0.02,
+      "loss": 1.6043,
+      "step": 5819
+    },
+    {
+      "epoch": 0.538795348029856,
+      "grad_norm": 0.1455078125,
+      "learning_rate": 0.02,
+      "loss": 1.5852,
+      "step": 5820
+    },
+    {
+      "epoch": 0.538887924550136,
+      "grad_norm": 0.1708984375,
+      "learning_rate": 0.02,
+      "loss": 1.609,
+      "step": 5821
+    },
+    {
+      "epoch": 0.5389805010704161,
+      "grad_norm": 0.146484375,
+      "learning_rate": 0.02,
+      "loss": 1.5678,
+      "step": 5822
+    },
+    {
+      "epoch": 0.5390730775906961,
+      "grad_norm": 0.1552734375,
+      "learning_rate": 0.02,
+      "loss": 1.5503,
+      "step": 5823
+    },
+    {
+      "epoch": 0.5391656541109761,
+      "grad_norm": 0.1494140625,
+      "learning_rate": 0.02,
+      "loss": 1.539,
+      "step": 5824
+    },
+    {
+      "epoch": 0.5392582306312561,
+      "grad_norm": 0.1484375,
+      "learning_rate": 0.02,
+      "loss": 1.533,
+      "step": 5825
+    },
+    {
+      "epoch": 0.5393508071515362,
+      "grad_norm": 0.14453125,
+      "learning_rate": 0.02,
+      "loss": 1.5477,
+      "step": 5826
+    },
+    {
+      "epoch": 0.5394433836718162,
+      "grad_norm": 0.16015625,
+      "learning_rate": 0.02,
+      "loss": 1.5122,
+      "step": 5827
+    },
+    {
+      "epoch": 0.5395359601920963,
+      "grad_norm": 0.1494140625,
+      "learning_rate": 0.02,
+      "loss": 1.5745,
+      "step": 5828
+    },
+    {
+      "epoch": 0.5396285367123763,
+      "grad_norm": 0.154296875,
+      "learning_rate": 0.02,
+      "loss": 1.4951,
+      "step": 5829
+    },
+    {
+      "epoch": 0.5397211132326564,
+      "grad_norm": 0.1435546875,
+      "learning_rate": 0.02,
+      "loss": 1.5755,
+      "step": 5830
+    },
+    {
+      "epoch": 0.5398136897529364,
+      "grad_norm": 0.158203125,
+      "learning_rate": 0.02,
+      "loss": 1.4915,
+      "step": 5831
+    },
+    {
+      "epoch": 0.5399062662732165,
+      "grad_norm": 0.15234375,
+      "learning_rate": 0.02,
+      "loss": 1.6049,
+      "step": 5832
+    },
+    {
+      "epoch": 0.5399988427934965,
+      "grad_norm": 0.15625,
+      "learning_rate": 0.02,
+      "loss": 1.5282,
+      "step": 5833
+    },
+    {
+      "epoch": 0.5400914193137766,
+      "grad_norm": 0.1484375,
+      "learning_rate": 0.02,
+      "loss": 1.5428,
+      "step": 5834
+    },
+    {
+      "epoch": 0.5401839958340566,
+      "grad_norm": 0.1552734375,
+      "learning_rate": 0.02,
+      "loss": 1.6081,
+      "step": 5835
+    },
+    {
+      "epoch": 0.5402765723543367,
+      "grad_norm": 0.15625,
+      "learning_rate": 0.02,
+      "loss": 1.5754,
+      "step": 5836
+    },
+    {
+      "epoch": 0.5403691488746166,
+      "grad_norm": 0.15234375,
+      "learning_rate": 0.02,
+      "loss": 1.5854,
+      "step": 5837
+    },
+    {
+      "epoch": 0.5404617253948967,
+      "grad_norm": 0.1533203125,
+      "learning_rate": 0.02,
+      "loss": 1.5622,
+      "step": 5838
+    },
+    {
+      "epoch": 0.5405543019151767,
+      "grad_norm": 0.1455078125,
+      "learning_rate": 0.02,
+      "loss": 1.537,
+      "step": 5839
+    },
+    {
+      "epoch": 0.5406468784354568,
+      "grad_norm": 0.14453125,
+      "learning_rate": 0.02,
+      "loss": 1.5068,
+      "step": 5840
+    },
+    {
+      "epoch": 0.5407394549557368,
+      "grad_norm": 0.138671875,
+      "learning_rate": 0.02,
+      "loss": 1.4914,
+      "step": 5841
+    },
+    {
+      "epoch": 0.5408320314760169,
+      "grad_norm": 0.1484375,
+      "learning_rate": 0.02,
+      "loss": 1.5127,
+      "step": 5842
+    },
+    {
+      "epoch": 0.5409246079962969,
+      "grad_norm": 0.1455078125,
+      "learning_rate": 0.02,
+      "loss": 1.5356,
+      "step": 5843
+    },
+    {
+      "epoch": 0.541017184516577,
+      "grad_norm": 0.140625,
+      "learning_rate": 0.02,
+      "loss": 1.5117,
+      "step": 5844
+    },
+    {
+      "epoch": 0.541109761036857,
+      "grad_norm": 0.1357421875,
+      "learning_rate": 0.02,
+      "loss": 1.5753,
+      "step": 5845
+    },
+    {
+      "epoch": 0.5412023375571371,
+      "grad_norm": 0.150390625,
+      "learning_rate": 0.02,
+      "loss": 1.5767,
+      "step": 5846
+    },
+    {
+      "epoch": 0.5412949140774171,
+      "grad_norm": 0.1513671875,
+      "learning_rate": 0.02,
+      "loss": 1.5817,
+      "step": 5847
+    },
+    {
+      "epoch": 0.5413874905976972,
+      "grad_norm": 0.154296875,
+      "learning_rate": 0.02,
+      "loss": 1.5504,
+      "step": 5848
+    },
+    {
+      "epoch": 0.5414800671179772,
+      "grad_norm": 0.1435546875,
+      "learning_rate": 0.02,
+      "loss": 1.5873,
+      "step": 5849
+    },
+    {
+      "epoch": 0.5415726436382573,
+      "grad_norm": 0.150390625,
+      "learning_rate": 0.02,
+      "loss": 1.609,
+      "step": 5850
+    },
+    {
+      "epoch": 0.5416652201585372,
+      "grad_norm": 0.1435546875,
+      "learning_rate": 0.02,
+      "loss": 1.6099,
+      "step": 5851
+    },
+    {
+      "epoch": 0.5417577966788173,
+      "grad_norm": 0.150390625,
+      "learning_rate": 0.02,
+      "loss": 1.5399,
+      "step": 5852
+    },
+    {
+      "epoch": 0.5418503731990973,
+      "grad_norm": 0.146484375,
+      "learning_rate": 0.02,
+      "loss": 1.4805,
+      "step": 5853
+    },
+    {
+      "epoch": 0.5419429497193774,
+      "grad_norm": 0.14453125,
+      "learning_rate": 0.02,
+      "loss": 1.5661,
+      "step": 5854
+    },
+    {
+      "epoch": 0.5420355262396575,
+      "grad_norm": 0.1435546875,
+      "learning_rate": 0.02,
+      "loss": 1.5368,
+      "step": 5855
+    },
+    {
+      "epoch": 0.5421281027599375,
+      "grad_norm": 0.1474609375,
+      "learning_rate": 0.02,
+      "loss": 1.5484,
+      "step": 5856
+    },
+    {
+      "epoch": 0.5422206792802176,
+      "grad_norm": 0.150390625,
+      "learning_rate": 0.02,
+      "loss": 1.6101,
+      "step": 5857
+    },
+    {
+      "epoch": 0.5423132558004976,
+      "grad_norm": 0.1484375,
+      "learning_rate": 0.02,
+      "loss": 1.5662,
+      "step": 5858
+    },
+    {
+      "epoch": 0.5424058323207777,
+      "grad_norm": 0.140625,
+      "learning_rate": 0.02,
+      "loss": 1.5391,
+      "step": 5859
+    },
+    {
+      "epoch": 0.5424984088410577,
+      "grad_norm": 0.142578125,
+      "learning_rate": 0.02,
+      "loss": 1.5583,
+      "step": 5860
+    },
+    {
+      "epoch": 0.5425909853613378,
+      "grad_norm": 0.1337890625,
+      "learning_rate": 0.02,
+      "loss": 1.5513,
+      "step": 5861
+    },
+    {
+      "epoch": 0.5426835618816178,
+      "grad_norm": 0.1396484375,
+      "learning_rate": 0.02,
+      "loss": 1.5707,
+      "step": 5862
+    },
+    {
+      "epoch": 0.5427761384018979,
+      "grad_norm": 0.1533203125,
+      "learning_rate": 0.02,
+      "loss": 1.5909,
+      "step": 5863
+    },
+    {
+      "epoch": 0.5428687149221778,
+      "grad_norm": 0.154296875,
+      "learning_rate": 0.02,
+      "loss": 1.5945,
+      "step": 5864
+    },
+    {
+      "epoch": 0.5429612914424579,
+      "grad_norm": 0.1669921875,
+      "learning_rate": 0.02,
+      "loss": 1.5323,
+      "step": 5865
+    },
+    {
+      "epoch": 0.5430538679627379,
+      "grad_norm": 0.1533203125,
+      "learning_rate": 0.02,
+      "loss": 1.5626,
+      "step": 5866
+    },
+    {
+      "epoch": 0.543146444483018,
+      "grad_norm": 0.1650390625,
+      "learning_rate": 0.02,
+      "loss": 1.5511,
+      "step": 5867
+    },
+    {
+      "epoch": 0.543239021003298,
+      "grad_norm": 0.142578125,
+      "learning_rate": 0.02,
+      "loss": 1.5997,
+      "step": 5868
+    },
+    {
+      "epoch": 0.5433315975235781,
+      "grad_norm": 0.1591796875,
+      "learning_rate": 0.02,
+      "loss": 1.6159,
+      "step": 5869
+    },
+    {
+      "epoch": 0.5434241740438581,
+      "grad_norm": 0.1572265625,
+      "learning_rate": 0.02,
+      "loss": 1.57,
+      "step": 5870
+    },
+    {
+      "epoch": 0.5435167505641382,
+      "grad_norm": 0.1552734375,
+      "learning_rate": 0.02,
+      "loss": 1.4412,
+      "step": 5871
+    },
+    {
+      "epoch": 0.5436093270844182,
+      "grad_norm": 0.1435546875,
+      "learning_rate": 0.02,
+      "loss": 1.5522,
+      "step": 5872
+    },
+    {
+      "epoch": 0.5437019036046983,
+      "grad_norm": 0.1455078125,
+      "learning_rate": 0.02,
+      "loss": 1.4979,
+      "step": 5873
+    },
+    {
+      "epoch": 0.5437944801249783,
+      "grad_norm": 0.1513671875,
+      "learning_rate": 0.02,
+      "loss": 1.5915,
+      "step": 5874
+    },
+    {
+      "epoch": 0.5438870566452584,
+      "grad_norm": 0.140625,
+      "learning_rate": 0.02,
+      "loss": 1.6201,
+      "step": 5875
+    },
+    {
+      "epoch": 0.5439796331655384,
+      "grad_norm": 0.1435546875,
+      "learning_rate": 0.02,
+      "loss": 1.5164,
+      "step": 5876
+    },
+    {
+      "epoch": 0.5440722096858185,
+      "grad_norm": 0.138671875,
+      "learning_rate": 0.02,
+      "loss": 1.572,
+      "step": 5877
+    },
+    {
+      "epoch": 0.5441647862060984,
+      "grad_norm": 0.1455078125,
+      "learning_rate": 0.02,
+      "loss": 1.6121,
+      "step": 5878
+    },
+    {
+      "epoch": 0.5442573627263785,
+      "grad_norm": 0.150390625,
+      "learning_rate": 0.02,
+      "loss": 1.5557,
+      "step": 5879
+    },
+    {
+      "epoch": 0.5443499392466585,
+      "grad_norm": 0.142578125,
+      "learning_rate": 0.02,
+      "loss": 1.4983,
+      "step": 5880
+    },
+    {
+      "epoch": 0.5444425157669386,
+      "grad_norm": 0.1513671875,
+      "learning_rate": 0.02,
+      "loss": 1.5493,
+      "step": 5881
+    },
+    {
+      "epoch": 0.5445350922872186,
+      "grad_norm": 0.15625,
+      "learning_rate": 0.02,
+      "loss": 1.5164,
+      "step": 5882
+    },
+    {
+      "epoch": 0.5446276688074987,
+      "grad_norm": 0.14453125,
+      "learning_rate": 0.02,
+      "loss": 1.581,
+      "step": 5883
+    },
+    {
+      "epoch": 0.5447202453277787,
+      "grad_norm": 0.1484375,
+      "learning_rate": 0.02,
+      "loss": 1.5421,
+      "step": 5884
+    },
+    {
+      "epoch": 0.5448128218480588,
+      "grad_norm": 0.1474609375,
+      "learning_rate": 0.02,
+      "loss": 1.5711,
+      "step": 5885
+    },
+    {
+      "epoch": 0.5449053983683388,
+      "grad_norm": 0.1533203125,
+      "learning_rate": 0.02,
+      "loss": 1.5623,
+      "step": 5886
+    },
+    {
+      "epoch": 0.5449979748886189,
+      "grad_norm": 0.140625,
+      "learning_rate": 0.02,
+      "loss": 1.4868,
+      "step": 5887
+    },
+    {
+      "epoch": 0.545090551408899,
+      "grad_norm": 0.1572265625,
+      "learning_rate": 0.02,
+      "loss": 1.6125,
+      "step": 5888
+    },
+    {
+      "epoch": 0.545183127929179,
+      "grad_norm": 0.142578125,
+      "learning_rate": 0.02,
+      "loss": 1.5816,
+      "step": 5889
+    },
+    {
+      "epoch": 0.545275704449459,
+      "grad_norm": 0.1376953125,
+      "learning_rate": 0.02,
+      "loss": 1.5506,
+      "step": 5890
+    },
+    {
+      "epoch": 0.5453682809697391,
+      "grad_norm": 0.1376953125,
+      "learning_rate": 0.02,
+      "loss": 1.5072,
+      "step": 5891
+    },
+    {
+      "epoch": 0.545460857490019,
+      "grad_norm": 0.14453125,
+      "learning_rate": 0.02,
+      "loss": 1.5341,
+      "step": 5892
+    },
+    {
+      "epoch": 0.5455534340102991,
+      "grad_norm": 0.142578125,
+      "learning_rate": 0.02,
+      "loss": 1.5754,
+      "step": 5893
+    },
+    {
+      "epoch": 0.5456460105305792,
+      "grad_norm": 0.140625,
+      "learning_rate": 0.02,
+      "loss": 1.4926,
+      "step": 5894
+    },
+    {
+      "epoch": 0.5457385870508592,
+      "grad_norm": 0.1376953125,
+      "learning_rate": 0.02,
+      "loss": 1.562,
+      "step": 5895
+    },
+    {
+      "epoch": 0.5458311635711393,
+      "grad_norm": 0.1474609375,
+      "learning_rate": 0.02,
+      "loss": 1.4652,
+      "step": 5896
+    },
+    {
+      "epoch": 0.5459237400914193,
+      "grad_norm": 0.1396484375,
+      "learning_rate": 0.02,
+      "loss": 1.5794,
+      "step": 5897
+    },
+    {
+      "epoch": 0.5460163166116994,
+      "grad_norm": 0.1474609375,
+      "learning_rate": 0.02,
+      "loss": 1.5151,
+      "step": 5898
+    },
+    {
+      "epoch": 0.5461088931319794,
+      "grad_norm": 0.1435546875,
+      "learning_rate": 0.02,
+      "loss": 1.5507,
+      "step": 5899
+    },
+    {
+      "epoch": 0.5462014696522595,
+      "grad_norm": 0.1474609375,
+      "learning_rate": 0.02,
+      "loss": 1.5732,
+      "step": 5900
+    },
+    {
+      "epoch": 0.5462940461725395,
+      "grad_norm": 0.140625,
+      "learning_rate": 0.02,
+      "loss": 1.6071,
+      "step": 5901
+    },
+    {
+      "epoch": 0.5463866226928196,
+      "grad_norm": 0.14453125,
+      "learning_rate": 0.02,
+      "loss": 1.5658,
+      "step": 5902
+    },
+    {
+      "epoch": 0.5463866226928196,
+      "eval_loss": 1.5515261888504028,
+      "eval_runtime": 59.9872,
+      "eval_samples_per_second": 25.405,
+      "eval_steps_per_second": 6.351,
+      "step": 5902
+    },
+    {
+      "epoch": 0.5464791992130996,
+      "grad_norm": 0.142578125,
+      "learning_rate": 0.02,
+      "loss": 1.6059,
+      "step": 5903
+    },
+    {
+      "epoch": 0.5465717757333797,
+      "grad_norm": 0.140625,
+      "learning_rate": 0.02,
+      "loss": 1.565,
+      "step": 5904
+    },
+    {
+      "epoch": 0.5466643522536596,
+      "grad_norm": 0.138671875,
+      "learning_rate": 0.02,
+      "loss": 1.5874,
+      "step": 5905
+    },
+    {
+      "epoch": 0.5467569287739397,
+      "grad_norm": 0.15234375,
+      "learning_rate": 0.02,
+      "loss": 1.595,
+      "step": 5906
+    },
+    {
+      "epoch": 0.5468495052942197,
+      "grad_norm": 0.150390625,
+      "learning_rate": 0.02,
+      "loss": 1.5294,
+      "step": 5907
+    },
+    {
+      "epoch": 0.5469420818144998,
+      "grad_norm": 0.146484375,
+      "learning_rate": 0.02,
+      "loss": 1.5086,
+      "step": 5908
+    },
+    {
+      "epoch": 0.5470346583347798,
+      "grad_norm": 0.150390625,
+      "learning_rate": 0.02,
+      "loss": 1.5447,
+      "step": 5909
+    },
+    {
+      "epoch": 0.5471272348550599,
+      "grad_norm": 0.1572265625,
+      "learning_rate": 0.02,
+      "loss": 1.5581,
+      "step": 5910
+    },
+    {
+      "epoch": 0.5472198113753399,
+      "grad_norm": 0.15625,
+      "learning_rate": 0.02,
+      "loss": 1.5498,
+      "step": 5911
+    },
+    {
+      "epoch": 0.54731238789562,
+      "grad_norm": 0.158203125,
+      "learning_rate": 0.02,
+      "loss": 1.6479,
+      "step": 5912
+    },
+    {
+      "epoch": 0.5474049644159,
+      "grad_norm": 0.158203125,
+      "learning_rate": 0.02,
+      "loss": 1.581,
+      "step": 5913
+    },
+    {
+      "epoch": 0.5474975409361801,
+      "grad_norm": 0.1533203125,
+      "learning_rate": 0.02,
+      "loss": 1.569,
+      "step": 5914
+    },
+    {
+      "epoch": 0.5475901174564601,
+      "grad_norm": 0.1572265625,
+      "learning_rate": 0.02,
+      "loss": 1.5129,
+      "step": 5915
+    },
+    {
+      "epoch": 0.5476826939767402,
+      "grad_norm": 0.1376953125,
+      "learning_rate": 0.02,
+      "loss": 1.5106,
+      "step": 5916
+    },
+    {
+      "epoch": 0.5477752704970202,
+      "grad_norm": 0.154296875,
+      "learning_rate": 0.02,
+      "loss": 1.543,
+      "step": 5917
+    },
+    {
+      "epoch": 0.5478678470173003,
+      "grad_norm": 0.146484375,
+      "learning_rate": 0.02,
+      "loss": 1.6299,
+      "step": 5918
+    },
+    {
+      "epoch": 0.5479604235375802,
+      "grad_norm": 0.1474609375,
+      "learning_rate": 0.02,
+      "loss": 1.5737,
+      "step": 5919
+    },
+    {
+      "epoch": 0.5480530000578603,
+      "grad_norm": 0.138671875,
+      "learning_rate": 0.02,
+      "loss": 1.5662,
+      "step": 5920
+    },
+    {
+      "epoch": 0.5481455765781403,
+      "grad_norm": 0.150390625,
+      "learning_rate": 0.02,
+      "loss": 1.5858,
+      "step": 5921
+    },
+    {
+      "epoch": 0.5482381530984204,
+      "grad_norm": 0.142578125,
+      "learning_rate": 0.02,
+      "loss": 1.5562,
+      "step": 5922
+    },
+    {
+      "epoch": 0.5483307296187004,
+      "grad_norm": 0.1416015625,
+      "learning_rate": 0.02,
+      "loss": 1.566,
+      "step": 5923
+    },
+    {
+      "epoch": 0.5484233061389805,
+      "grad_norm": 0.1396484375,
+      "learning_rate": 0.02,
+      "loss": 1.5291,
+      "step": 5924
+    },
+    {
+      "epoch": 0.5485158826592605,
+      "grad_norm": 0.1474609375,
+      "learning_rate": 0.02,
+      "loss": 1.5176,
+      "step": 5925
+    },
+    {
+      "epoch": 0.5486084591795406,
+      "grad_norm": 0.1552734375,
+      "learning_rate": 0.02,
+      "loss": 1.5948,
+      "step": 5926
+    },
+    {
+      "epoch": 0.5487010356998206,
+      "grad_norm": 0.1513671875,
+      "learning_rate": 0.02,
+      "loss": 1.562,
+      "step": 5927
+    },
+    {
+      "epoch": 0.5487936122201007,
+      "grad_norm": 0.154296875,
+      "learning_rate": 0.02,
+      "loss": 1.5255,
+      "step": 5928
+    },
+    {
+      "epoch": 0.5488861887403808,
+      "grad_norm": 0.1513671875,
+      "learning_rate": 0.02,
+      "loss": 1.4922,
+      "step": 5929
+    },
+    {
+      "epoch": 0.5489787652606608,
+      "grad_norm": 0.1474609375,
+      "learning_rate": 0.02,
+      "loss": 1.5217,
+      "step": 5930
+    },
+    {
+      "epoch": 0.5490713417809409,
+      "grad_norm": 0.158203125,
+      "learning_rate": 0.02,
+      "loss": 1.5319,
+      "step": 5931
+    },
+    {
+      "epoch": 0.5491639183012208,
+      "grad_norm": 0.15625,
+      "learning_rate": 0.02,
+      "loss": 1.5527,
+      "step": 5932
+    },
+    {
+      "epoch": 0.5492564948215009,
+      "grad_norm": 0.154296875,
+      "learning_rate": 0.02,
+      "loss": 1.5471,
+      "step": 5933
+    },
+    {
+      "epoch": 0.5493490713417809,
+      "grad_norm": 0.1572265625,
+      "learning_rate": 0.02,
+      "loss": 1.545,
+      "step": 5934
+    },
+    {
+      "epoch": 0.549441647862061,
+      "grad_norm": 0.1513671875,
+      "learning_rate": 0.02,
+      "loss": 1.5043,
+      "step": 5935
+    },
+    {
+      "epoch": 0.549534224382341,
+      "grad_norm": 0.1572265625,
+      "learning_rate": 0.02,
+      "loss": 1.5489,
+      "step": 5936
+    },
+    {
+      "epoch": 0.5496268009026211,
+      "grad_norm": 0.146484375,
+      "learning_rate": 0.02,
+      "loss": 1.5497,
+      "step": 5937
+    },
+    {
+      "epoch": 0.5497193774229011,
+      "grad_norm": 0.15234375,
+      "learning_rate": 0.02,
+      "loss": 1.5397,
+      "step": 5938
+    },
+    {
+      "epoch": 0.5498119539431812,
+      "grad_norm": 0.158203125,
+      "learning_rate": 0.02,
+      "loss": 1.519,
+      "step": 5939
+    },
+    {
+      "epoch": 0.5499045304634612,
+      "grad_norm": 0.15625,
+      "learning_rate": 0.02,
+      "loss": 1.593,
+      "step": 5940
+    },
+    {
+      "epoch": 0.5499971069837413,
+      "grad_norm": 0.1513671875,
+      "learning_rate": 0.02,
+      "loss": 1.5034,
+      "step": 5941
+    },
+    {
+      "epoch": 0.5500896835040213,
+      "grad_norm": 0.14453125,
+      "learning_rate": 0.02,
+      "loss": 1.5427,
+      "step": 5942
+    },
+    {
+      "epoch": 0.5501822600243014,
+      "grad_norm": 0.158203125,
+      "learning_rate": 0.02,
+      "loss": 1.5985,
+      "step": 5943
+    },
+    {
+      "epoch": 0.5502748365445814,
+      "grad_norm": 0.1435546875,
+      "learning_rate": 0.02,
+      "loss": 1.5973,
+      "step": 5944
+    },
+    {
+      "epoch": 0.5503674130648615,
+      "grad_norm": 0.15234375,
+      "learning_rate": 0.02,
+      "loss": 1.5879,
+      "step": 5945
+    },
+    {
+      "epoch": 0.5504599895851414,
+      "grad_norm": 0.1552734375,
+      "learning_rate": 0.02,
+      "loss": 1.5132,
+      "step": 5946
+    },
+    {
+      "epoch": 0.5505525661054215,
+      "grad_norm": 0.154296875,
+      "learning_rate": 0.02,
+      "loss": 1.5282,
+      "step": 5947
+    },
+    {
+      "epoch": 0.5506451426257015,
+      "grad_norm": 0.1572265625,
+      "learning_rate": 0.02,
+      "loss": 1.5109,
+      "step": 5948
+    },
+    {
+      "epoch": 0.5507377191459816,
+      "grad_norm": 0.15234375,
+      "learning_rate": 0.02,
+      "loss": 1.5365,
+      "step": 5949
+    },
+    {
+      "epoch": 0.5508302956662616,
+      "grad_norm": 0.1640625,
+      "learning_rate": 0.02,
+      "loss": 1.5245,
+      "step": 5950
+    },
+    {
+      "epoch": 0.5509228721865417,
+      "grad_norm": 0.1591796875,
+      "learning_rate": 0.02,
+      "loss": 1.5004,
+      "step": 5951
+    },
+    {
+      "epoch": 0.5510154487068217,
+      "grad_norm": 0.158203125,
+      "learning_rate": 0.02,
+      "loss": 1.5682,
+      "step": 5952
+    },
+    {
+      "epoch": 0.5511080252271018,
+      "grad_norm": 0.16015625,
+      "learning_rate": 0.02,
+      "loss": 1.5883,
+      "step": 5953
+    },
+    {
+      "epoch": 0.5512006017473818,
+      "grad_norm": 0.1611328125,
+      "learning_rate": 0.02,
+      "loss": 1.5506,
+      "step": 5954
+    },
+    {
+      "epoch": 0.5512931782676619,
+      "grad_norm": 0.1591796875,
+      "learning_rate": 0.02,
+      "loss": 1.5621,
+      "step": 5955
+    },
+    {
+      "epoch": 0.5513857547879419,
+      "grad_norm": 0.1552734375,
+      "learning_rate": 0.02,
+      "loss": 1.5737,
+      "step": 5956
+    },
+    {
+      "epoch": 0.551478331308222,
+      "grad_norm": 0.1396484375,
+      "learning_rate": 0.02,
+      "loss": 1.5726,
+      "step": 5957
+    },
+    {
+      "epoch": 0.551570907828502,
+      "grad_norm": 0.1455078125,
+      "learning_rate": 0.02,
+      "loss": 1.5612,
+      "step": 5958
+    },
+    {
+      "epoch": 0.551663484348782,
+      "grad_norm": 0.15234375,
+      "learning_rate": 0.02,
+      "loss": 1.448,
+      "step": 5959
+    },
+    {
+      "epoch": 0.551756060869062,
+      "grad_norm": 0.1640625,
+      "learning_rate": 0.02,
+      "loss": 1.5272,
+      "step": 5960
+    },
+    {
+      "epoch": 0.5518486373893421,
+      "grad_norm": 0.1552734375,
+      "learning_rate": 0.02,
+      "loss": 1.5679,
+      "step": 5961
+    },
+    {
+      "epoch": 0.5519412139096221,
+      "grad_norm": 0.1474609375,
+      "learning_rate": 0.02,
+      "loss": 1.4825,
+      "step": 5962
+    },
+    {
+      "epoch": 0.5520337904299022,
+      "grad_norm": 0.1416015625,
+      "learning_rate": 0.02,
+      "loss": 1.5138,
+      "step": 5963
+    },
+    {
+      "epoch": 0.5521263669501822,
+      "grad_norm": 0.146484375,
+      "learning_rate": 0.02,
+      "loss": 1.5615,
+      "step": 5964
+    },
+    {
+      "epoch": 0.5522189434704623,
+      "grad_norm": 0.142578125,
+      "learning_rate": 0.02,
+      "loss": 1.5261,
+      "step": 5965
+    },
+    {
+      "epoch": 0.5523115199907424,
+      "grad_norm": 0.1435546875,
+      "learning_rate": 0.02,
+      "loss": 1.5633,
+      "step": 5966
+    },
+    {
+      "epoch": 0.5524040965110224,
+      "grad_norm": 0.1416015625,
+      "learning_rate": 0.02,
+      "loss": 1.5021,
+      "step": 5967
+    },
+    {
+      "epoch": 0.5524966730313025,
+      "grad_norm": 0.142578125,
+      "learning_rate": 0.02,
+      "loss": 1.6194,
+      "step": 5968
+    },
+    {
+      "epoch": 0.5525892495515825,
+      "grad_norm": 0.1474609375,
+      "learning_rate": 0.02,
+      "loss": 1.5489,
+      "step": 5969
+    },
+    {
+      "epoch": 0.5526818260718626,
+      "grad_norm": 0.150390625,
+      "learning_rate": 0.02,
+      "loss": 1.5773,
+      "step": 5970
+    },
+    {
+      "epoch": 0.5527744025921426,
+      "grad_norm": 0.1552734375,
+      "learning_rate": 0.02,
+      "loss": 1.6275,
+      "step": 5971
+    },
+    {
+      "epoch": 0.5528669791124227,
+      "grad_norm": 0.1533203125,
+      "learning_rate": 0.02,
+      "loss": 1.5864,
+      "step": 5972
+    },
+    {
+      "epoch": 0.5529595556327026,
+      "grad_norm": 0.154296875,
+      "learning_rate": 0.02,
+      "loss": 1.5235,
+      "step": 5973
+    },
+    {
+      "epoch": 0.5530521321529827,
+      "grad_norm": 0.150390625,
+      "learning_rate": 0.02,
+      "loss": 1.574,
+      "step": 5974
+    },
+    {
+      "epoch": 0.5531447086732627,
+      "grad_norm": 0.1474609375,
+      "learning_rate": 0.02,
+      "loss": 1.5496,
+      "step": 5975
+    },
+    {
+      "epoch": 0.5532372851935428,
+      "grad_norm": 0.158203125,
+      "learning_rate": 0.02,
+      "loss": 1.5547,
+      "step": 5976
+    },
+    {
+      "epoch": 0.5533298617138228,
+      "grad_norm": 0.1455078125,
+      "learning_rate": 0.02,
+      "loss": 1.5179,
+      "step": 5977
+    },
+    {
+      "epoch": 0.5534224382341029,
+      "grad_norm": 0.154296875,
+      "learning_rate": 0.02,
+      "loss": 1.6324,
+      "step": 5978
+    },
+    {
+      "epoch": 0.5535150147543829,
+      "grad_norm": 0.1474609375,
+      "learning_rate": 0.02,
+      "loss": 1.6043,
+      "step": 5979
+    },
+    {
+      "epoch": 0.553607591274663,
+      "grad_norm": 0.15234375,
+      "learning_rate": 0.02,
+      "loss": 1.5972,
+      "step": 5980
+    },
+    {
+      "epoch": 0.553700167794943,
+      "grad_norm": 0.1513671875,
+      "learning_rate": 0.02,
+      "loss": 1.5851,
+      "step": 5981
+    },
+    {
+      "epoch": 0.5537927443152231,
+      "grad_norm": 0.1494140625,
+      "learning_rate": 0.02,
+      "loss": 1.5137,
+      "step": 5982
+    },
+    {
+      "epoch": 0.5538853208355031,
+      "grad_norm": 0.14453125,
+      "learning_rate": 0.02,
+      "loss": 1.5073,
+      "step": 5983
+    },
+    {
+      "epoch": 0.5539778973557832,
+      "grad_norm": 0.158203125,
+      "learning_rate": 0.02,
+      "loss": 1.5631,
+      "step": 5984
+    },
+    {
+      "epoch": 0.5540704738760632,
+      "grad_norm": 0.14453125,
+      "learning_rate": 0.02,
+      "loss": 1.5708,
+      "step": 5985
+    },
+    {
+      "epoch": 0.5541630503963433,
+      "grad_norm": 0.150390625,
+      "learning_rate": 0.02,
+      "loss": 1.505,
+      "step": 5986
+    },
+    {
+      "epoch": 0.5542556269166232,
+      "grad_norm": 0.1640625,
+      "learning_rate": 0.02,
+      "loss": 1.513,
+      "step": 5987
+    },
+    {
+      "epoch": 0.5543482034369033,
+      "grad_norm": 0.154296875,
+      "learning_rate": 0.02,
+      "loss": 1.5705,
+      "step": 5988
+    },
+    {
+      "epoch": 0.5544407799571833,
+      "grad_norm": 0.154296875,
+      "learning_rate": 0.02,
+      "loss": 1.6026,
+      "step": 5989
+    },
+    {
+      "epoch": 0.5545333564774634,
+      "grad_norm": 0.1484375,
+      "learning_rate": 0.02,
+      "loss": 1.505,
+      "step": 5990
+    },
+    {
+      "epoch": 0.5546259329977434,
+      "grad_norm": 0.14453125,
+      "learning_rate": 0.02,
+      "loss": 1.5464,
+      "step": 5991
+    },
+    {
+      "epoch": 0.5547185095180235,
+      "grad_norm": 0.154296875,
+      "learning_rate": 0.02,
+      "loss": 1.5789,
+      "step": 5992
+    },
+    {
+      "epoch": 0.5548110860383035,
+      "grad_norm": 0.1533203125,
+      "learning_rate": 0.02,
+      "loss": 1.6512,
+      "step": 5993
+    },
+    {
+      "epoch": 0.5549036625585836,
+      "grad_norm": 0.1533203125,
+      "learning_rate": 0.02,
+      "loss": 1.6024,
+      "step": 5994
+    },
+    {
+      "epoch": 0.5549962390788636,
+      "grad_norm": 0.142578125,
+      "learning_rate": 0.02,
+      "loss": 1.5744,
+      "step": 5995
+    },
+    {
+      "epoch": 0.5550888155991437,
+      "grad_norm": 0.1435546875,
+      "learning_rate": 0.02,
+      "loss": 1.5413,
+      "step": 5996
+    },
+    {
+      "epoch": 0.5551813921194237,
+      "grad_norm": 0.1513671875,
+      "learning_rate": 0.02,
+      "loss": 1.5813,
+      "step": 5997
+    },
+    {
+      "epoch": 0.5552739686397038,
+      "grad_norm": 0.1455078125,
+      "learning_rate": 0.02,
+      "loss": 1.55,
+      "step": 5998
+    },
+    {
+      "epoch": 0.5553665451599838,
+      "grad_norm": 0.14453125,
+      "learning_rate": 0.02,
+      "loss": 1.6071,
+      "step": 5999
+    },
+    {
+      "epoch": 0.5554591216802638,
+      "grad_norm": 0.1396484375,
+      "learning_rate": 0.02,
+      "loss": 1.6078,
+      "step": 6000
+    },
+    {
+      "epoch": 0.5555516982005438,
+      "grad_norm": 0.1494140625,
+      "learning_rate": 0.02,
+      "loss": 1.5247,
+      "step": 6001
+    },
+    {
+      "epoch": 0.5556442747208239,
+      "grad_norm": 0.1416015625,
+      "learning_rate": 0.02,
+      "loss": 1.6026,
+      "step": 6002
+    },
+    {
+      "epoch": 0.555736851241104,
+      "grad_norm": 0.1455078125,
+      "learning_rate": 0.02,
+      "loss": 1.5895,
+      "step": 6003
+    },
+    {
+      "epoch": 0.555829427761384,
+      "grad_norm": 0.142578125,
+      "learning_rate": 0.02,
+      "loss": 1.4875,
+      "step": 6004
+    },
+    {
+      "epoch": 0.555922004281664,
+      "grad_norm": 0.1455078125,
+      "learning_rate": 0.02,
+      "loss": 1.5848,
+      "step": 6005
+    },
+    {
+      "epoch": 0.5560145808019441,
+      "grad_norm": 0.14453125,
+      "learning_rate": 0.02,
+      "loss": 1.523,
+      "step": 6006
+    },
+    {
+      "epoch": 0.5561071573222242,
+      "grad_norm": 0.1435546875,
+      "learning_rate": 0.02,
+      "loss": 1.6117,
+      "step": 6007
+    },
+    {
+      "epoch": 0.5561997338425042,
+      "grad_norm": 0.1435546875,
+      "learning_rate": 0.02,
+      "loss": 1.5251,
+      "step": 6008
+    },
+    {
+      "epoch": 0.5562923103627843,
+      "grad_norm": 0.14453125,
+      "learning_rate": 0.02,
+      "loss": 1.5458,
+      "step": 6009
+    },
+    {
+      "epoch": 0.5563848868830643,
+      "grad_norm": 0.1416015625,
+      "learning_rate": 0.02,
+      "loss": 1.6022,
+      "step": 6010
+    },
+    {
+      "epoch": 0.5564774634033444,
+      "grad_norm": 0.1474609375,
+      "learning_rate": 0.02,
+      "loss": 1.5169,
+      "step": 6011
+    },
+    {
+      "epoch": 0.5565700399236244,
+      "grad_norm": 0.146484375,
+      "learning_rate": 0.02,
+      "loss": 1.5442,
+      "step": 6012
+    },
+    {
+      "epoch": 0.5566626164439045,
+      "grad_norm": 0.158203125,
+      "learning_rate": 0.02,
+      "loss": 1.5905,
+      "step": 6013
+    },
+    {
+      "epoch": 0.5567551929641844,
+      "grad_norm": 0.1435546875,
+      "learning_rate": 0.02,
+      "loss": 1.5151,
+      "step": 6014
+    },
+    {
+      "epoch": 0.5568477694844645,
+      "grad_norm": 0.150390625,
+      "learning_rate": 0.02,
+      "loss": 1.6318,
+      "step": 6015
+    },
+    {
+      "epoch": 0.5569403460047445,
+      "grad_norm": 0.1494140625,
+      "learning_rate": 0.02,
+      "loss": 1.507,
+      "step": 6016
+    },
+    {
+      "epoch": 0.5570329225250246,
+      "grad_norm": 0.1533203125,
+      "learning_rate": 0.02,
+      "loss": 1.5276,
+      "step": 6017
+    },
+    {
+      "epoch": 0.5571254990453046,
+      "grad_norm": 0.1484375,
+      "learning_rate": 0.02,
+      "loss": 1.552,
+      "step": 6018
+    },
+    {
+      "epoch": 0.5572180755655847,
+      "grad_norm": 0.14453125,
+      "learning_rate": 0.02,
+      "loss": 1.5458,
+      "step": 6019
+    },
+    {
+      "epoch": 0.5573106520858647,
+      "grad_norm": 0.1591796875,
+      "learning_rate": 0.02,
+      "loss": 1.6208,
+      "step": 6020
+    },
+    {
+      "epoch": 0.5574032286061448,
+      "grad_norm": 0.1513671875,
+      "learning_rate": 0.02,
+      "loss": 1.5786,
+      "step": 6021
+    },
+    {
+      "epoch": 0.5574958051264248,
+      "grad_norm": 0.1552734375,
+      "learning_rate": 0.02,
+      "loss": 1.559,
+      "step": 6022
+    },
+    {
+      "epoch": 0.5575883816467049,
+      "grad_norm": 0.1416015625,
+      "learning_rate": 0.02,
+      "loss": 1.5314,
+      "step": 6023
+    },
+    {
+      "epoch": 0.5576809581669849,
+      "grad_norm": 0.1484375,
+      "learning_rate": 0.02,
+      "loss": 1.526,
+      "step": 6024
+    },
+    {
+      "epoch": 0.557773534687265,
+      "grad_norm": 0.15625,
+      "learning_rate": 0.02,
+      "loss": 1.6247,
+      "step": 6025
+    },
+    {
+      "epoch": 0.557866111207545,
+      "grad_norm": 0.1552734375,
+      "learning_rate": 0.02,
+      "loss": 1.6031,
+      "step": 6026
+    },
+    {
+      "epoch": 0.557958687727825,
+      "grad_norm": 0.1484375,
+      "learning_rate": 0.02,
+      "loss": 1.521,
+      "step": 6027
+    },
+    {
+      "epoch": 0.558051264248105,
+      "grad_norm": 0.1494140625,
+      "learning_rate": 0.02,
+      "loss": 1.5293,
+      "step": 6028
+    },
+    {
+      "epoch": 0.5581438407683851,
+      "grad_norm": 0.134765625,
+      "learning_rate": 0.02,
+      "loss": 1.5642,
+      "step": 6029
+    },
+    {
+      "epoch": 0.5582364172886651,
+      "grad_norm": 0.1474609375,
+      "learning_rate": 0.02,
+      "loss": 1.6067,
+      "step": 6030
+    },
+    {
+      "epoch": 0.5583289938089452,
+      "grad_norm": 0.1416015625,
+      "learning_rate": 0.02,
+      "loss": 1.6204,
+      "step": 6031
+    },
+    {
+      "epoch": 0.5584215703292252,
+      "grad_norm": 0.1572265625,
+      "learning_rate": 0.02,
+      "loss": 1.5232,
+      "step": 6032
+    },
+    {
+      "epoch": 0.5585141468495053,
+      "grad_norm": 0.1328125,
+      "learning_rate": 0.02,
+      "loss": 1.6223,
+      "step": 6033
+    },
+    {
+      "epoch": 0.5586067233697853,
+      "grad_norm": 0.1396484375,
+      "learning_rate": 0.02,
+      "loss": 1.5342,
+      "step": 6034
+    },
+    {
+      "epoch": 0.5586992998900654,
+      "grad_norm": 0.1474609375,
+      "learning_rate": 0.02,
+      "loss": 1.5241,
+      "step": 6035
+    },
+    {
+      "epoch": 0.5587918764103454,
+      "grad_norm": 0.154296875,
+      "learning_rate": 0.02,
+      "loss": 1.5571,
+      "step": 6036
+    },
+    {
+      "epoch": 0.5588844529306255,
+      "grad_norm": 0.15234375,
+      "learning_rate": 0.02,
+      "loss": 1.5282,
+      "step": 6037
+    },
+    {
+      "epoch": 0.5589770294509055,
+      "grad_norm": 0.1494140625,
+      "learning_rate": 0.02,
+      "loss": 1.5159,
+      "step": 6038
+    },
+    {
+      "epoch": 0.5590696059711856,
+      "grad_norm": 0.1474609375,
+      "learning_rate": 0.02,
+      "loss": 1.5565,
+      "step": 6039
+    },
+    {
+      "epoch": 0.5591621824914657,
+      "grad_norm": 0.16015625,
+      "learning_rate": 0.02,
+      "loss": 1.5704,
+      "step": 6040
+    },
+    {
+      "epoch": 0.5592547590117456,
+      "grad_norm": 0.1591796875,
+      "learning_rate": 0.02,
+      "loss": 1.5175,
+      "step": 6041
+    },
+    {
+      "epoch": 0.5593473355320256,
+      "grad_norm": 0.150390625,
+      "learning_rate": 0.02,
+      "loss": 1.4938,
+      "step": 6042
+    },
+    {
+      "epoch": 0.5594399120523057,
+      "grad_norm": 0.1416015625,
+      "learning_rate": 0.02,
+      "loss": 1.5651,
+      "step": 6043
+    },
+    {
+      "epoch": 0.5595324885725858,
+      "grad_norm": 0.15625,
+      "learning_rate": 0.02,
+      "loss": 1.5619,
+      "step": 6044
+    },
+    {
+      "epoch": 0.5596250650928658,
+      "grad_norm": 0.15234375,
+      "learning_rate": 0.02,
+      "loss": 1.5835,
+      "step": 6045
+    },
+    {
+      "epoch": 0.5597176416131459,
+      "grad_norm": 0.1533203125,
+      "learning_rate": 0.02,
+      "loss": 1.6105,
+      "step": 6046
+    },
+    {
+      "epoch": 0.5598102181334259,
+      "grad_norm": 0.1484375,
+      "learning_rate": 0.02,
+      "loss": 1.5772,
+      "step": 6047
+    },
+    {
+      "epoch": 0.559902794653706,
+      "grad_norm": 0.150390625,
+      "learning_rate": 0.02,
+      "loss": 1.5302,
+      "step": 6048
+    },
+    {
+      "epoch": 0.559995371173986,
+      "grad_norm": 0.1533203125,
+      "learning_rate": 0.02,
+      "loss": 1.5118,
+      "step": 6049
+    },
+    {
+      "epoch": 0.5600879476942661,
+      "grad_norm": 0.1396484375,
+      "learning_rate": 0.02,
+      "loss": 1.5045,
+      "step": 6050
+    },
+    {
+      "epoch": 0.5601805242145461,
+      "grad_norm": 0.1455078125,
+      "learning_rate": 0.02,
+      "loss": 1.5051,
+      "step": 6051
+    },
+    {
+      "epoch": 0.5602731007348262,
+      "grad_norm": 0.146484375,
+      "learning_rate": 0.02,
+      "loss": 1.6011,
+      "step": 6052
+    },
+    {
+      "epoch": 0.5603656772551062,
+      "grad_norm": 0.138671875,
+      "learning_rate": 0.02,
+      "loss": 1.5594,
+      "step": 6053
+    },
+    {
+      "epoch": 0.5604582537753863,
+      "grad_norm": 0.1416015625,
+      "learning_rate": 0.02,
+      "loss": 1.5485,
+      "step": 6054
+    },
+    {
+      "epoch": 0.5605508302956662,
+      "grad_norm": 0.1484375,
+      "learning_rate": 0.02,
+      "loss": 1.5687,
+      "step": 6055
+    },
+    {
+      "epoch": 0.5606434068159463,
+      "grad_norm": 0.1435546875,
+      "learning_rate": 0.02,
+      "loss": 1.5773,
+      "step": 6056
+    },
+    {
+      "epoch": 0.5607359833362263,
+      "grad_norm": 0.14453125,
+      "learning_rate": 0.02,
+      "loss": 1.5983,
+      "step": 6057
+    },
+    {
+      "epoch": 0.5608285598565064,
+      "grad_norm": 0.1533203125,
+      "learning_rate": 0.02,
+      "loss": 1.5267,
+      "step": 6058
+    },
+    {
+      "epoch": 0.5609211363767864,
+      "grad_norm": 0.146484375,
+      "learning_rate": 0.02,
+      "loss": 1.6068,
+      "step": 6059
+    },
+    {
+      "epoch": 0.5610137128970665,
+      "grad_norm": 0.15234375,
+      "learning_rate": 0.02,
+      "loss": 1.4765,
+      "step": 6060
+    },
+    {
+      "epoch": 0.5611062894173465,
+      "grad_norm": 0.150390625,
+      "learning_rate": 0.02,
+      "loss": 1.5653,
+      "step": 6061
+    },
+    {
+      "epoch": 0.5611988659376266,
+      "grad_norm": 0.16015625,
+      "learning_rate": 0.02,
+      "loss": 1.5231,
+      "step": 6062
+    },
+    {
+      "epoch": 0.5612914424579066,
+      "grad_norm": 0.142578125,
+      "learning_rate": 0.02,
+      "loss": 1.5525,
+      "step": 6063
+    },
+    {
+      "epoch": 0.5613840189781867,
+      "grad_norm": 0.1494140625,
+      "learning_rate": 0.02,
+      "loss": 1.5029,
+      "step": 6064
+    },
+    {
+      "epoch": 0.5614765954984667,
+      "grad_norm": 0.14453125,
+      "learning_rate": 0.02,
+      "loss": 1.5684,
+      "step": 6065
+    },
+    {
+      "epoch": 0.5615691720187468,
+      "grad_norm": 0.1689453125,
+      "learning_rate": 0.02,
+      "loss": 1.6028,
+      "step": 6066
+    },
+    {
+      "epoch": 0.5616617485390268,
+      "grad_norm": 0.1455078125,
+      "learning_rate": 0.02,
+      "loss": 1.5625,
+      "step": 6067
+    },
+    {
+      "epoch": 0.5617543250593068,
+      "grad_norm": 0.1484375,
+      "learning_rate": 0.02,
+      "loss": 1.5099,
+      "step": 6068
+    },
+    {
+      "epoch": 0.5618469015795868,
+      "grad_norm": 0.150390625,
+      "learning_rate": 0.02,
+      "loss": 1.6117,
+      "step": 6069
+    },
+    {
+      "epoch": 0.5619394780998669,
+      "grad_norm": 0.15234375,
+      "learning_rate": 0.02,
+      "loss": 1.558,
+      "step": 6070
+    },
+    {
+      "epoch": 0.5620320546201469,
+      "grad_norm": 0.1337890625,
+      "learning_rate": 0.02,
+      "loss": 1.5471,
+      "step": 6071
+    },
+    {
+      "epoch": 0.562124631140427,
+      "grad_norm": 0.1494140625,
+      "learning_rate": 0.02,
+      "loss": 1.5803,
+      "step": 6072
+    },
+    {
+      "epoch": 0.562217207660707,
+      "grad_norm": 0.140625,
+      "learning_rate": 0.02,
+      "loss": 1.5761,
+      "step": 6073
+    },
+    {
+      "epoch": 0.5623097841809871,
+      "grad_norm": 0.1455078125,
+      "learning_rate": 0.02,
+      "loss": 1.6077,
+      "step": 6074
+    },
+    {
+      "epoch": 0.5624023607012671,
+      "grad_norm": 0.146484375,
+      "learning_rate": 0.02,
+      "loss": 1.5395,
+      "step": 6075
+    },
+    {
+      "epoch": 0.5624949372215472,
+      "grad_norm": 0.150390625,
+      "learning_rate": 0.02,
+      "loss": 1.5342,
+      "step": 6076
+    },
+    {
+      "epoch": 0.5625875137418272,
+      "grad_norm": 0.1455078125,
+      "learning_rate": 0.02,
+      "loss": 1.5355,
+      "step": 6077
+    },
+    {
+      "epoch": 0.5626800902621073,
+      "grad_norm": 0.138671875,
+      "learning_rate": 0.02,
+      "loss": 1.4983,
+      "step": 6078
+    },
+    {
+      "epoch": 0.5627726667823874,
+      "grad_norm": 0.146484375,
+      "learning_rate": 0.02,
+      "loss": 1.5737,
+      "step": 6079
+    },
+    {
+      "epoch": 0.5628652433026674,
+      "grad_norm": 0.150390625,
+      "learning_rate": 0.02,
+      "loss": 1.5429,
+      "step": 6080
+    },
+    {
+      "epoch": 0.5629578198229475,
+      "grad_norm": 0.1533203125,
+      "learning_rate": 0.02,
+      "loss": 1.528,
+      "step": 6081
+    },
+    {
+      "epoch": 0.5630503963432274,
+      "grad_norm": 0.1484375,
+      "learning_rate": 0.02,
+      "loss": 1.562,
+      "step": 6082
+    },
+    {
+      "epoch": 0.5631429728635075,
+      "grad_norm": 0.146484375,
+      "learning_rate": 0.02,
+      "loss": 1.5798,
+      "step": 6083
+    },
+    {
+      "epoch": 0.5632355493837875,
+      "grad_norm": 0.14453125,
+      "learning_rate": 0.02,
+      "loss": 1.5786,
+      "step": 6084
+    },
+    {
+      "epoch": 0.5633281259040676,
+      "grad_norm": 0.146484375,
+      "learning_rate": 0.02,
+      "loss": 1.5148,
+      "step": 6085
+    },
+    {
+      "epoch": 0.5634207024243476,
+      "grad_norm": 0.1484375,
+      "learning_rate": 0.02,
+      "loss": 1.5807,
+      "step": 6086
+    },
+    {
+      "epoch": 0.5635132789446277,
+      "grad_norm": 0.14453125,
+      "learning_rate": 0.02,
+      "loss": 1.5962,
+      "step": 6087
+    },
+    {
+      "epoch": 0.5636058554649077,
+      "grad_norm": 0.146484375,
+      "learning_rate": 0.02,
+      "loss": 1.4903,
+      "step": 6088
+    },
+    {
+      "epoch": 0.5636984319851878,
+      "grad_norm": 0.1474609375,
+      "learning_rate": 0.02,
+      "loss": 1.5515,
+      "step": 6089
+    },
+    {
+      "epoch": 0.5637910085054678,
+      "grad_norm": 0.1484375,
+      "learning_rate": 0.02,
+      "loss": 1.5646,
+      "step": 6090
+    },
+    {
+      "epoch": 0.5638835850257479,
+      "grad_norm": 0.138671875,
+      "learning_rate": 0.02,
+      "loss": 1.4961,
+      "step": 6091
+    },
+    {
+      "epoch": 0.5639761615460279,
+      "grad_norm": 0.146484375,
+      "learning_rate": 0.02,
+      "loss": 1.5592,
+      "step": 6092
+    },
+    {
+      "epoch": 0.564068738066308,
+      "grad_norm": 0.1513671875,
+      "learning_rate": 0.02,
+      "loss": 1.5387,
+      "step": 6093
+    },
+    {
+      "epoch": 0.564161314586588,
+      "grad_norm": 0.1494140625,
+      "learning_rate": 0.02,
+      "loss": 1.5201,
+      "step": 6094
+    },
+    {
+      "epoch": 0.564253891106868,
+      "grad_norm": 0.1474609375,
+      "learning_rate": 0.02,
+      "loss": 1.5454,
+      "step": 6095
+    },
+    {
+      "epoch": 0.564346467627148,
+      "grad_norm": 0.1435546875,
+      "learning_rate": 0.02,
+      "loss": 1.6085,
+      "step": 6096
+    },
+    {
+      "epoch": 0.5644390441474281,
+      "grad_norm": 0.1533203125,
+      "learning_rate": 0.02,
+      "loss": 1.5805,
+      "step": 6097
+    },
+    {
+      "epoch": 0.5645316206677081,
+      "grad_norm": 0.142578125,
+      "learning_rate": 0.02,
+      "loss": 1.5672,
+      "step": 6098
+    },
+    {
+      "epoch": 0.5646241971879882,
+      "grad_norm": 0.146484375,
+      "learning_rate": 0.02,
+      "loss": 1.5656,
+      "step": 6099
+    },
+    {
+      "epoch": 0.5647167737082682,
+      "grad_norm": 0.146484375,
+      "learning_rate": 0.02,
+      "loss": 1.5225,
+      "step": 6100
+    },
+    {
+      "epoch": 0.5648093502285483,
+      "grad_norm": 0.15234375,
+      "learning_rate": 0.02,
+      "loss": 1.5923,
+      "step": 6101
+    },
+    {
+      "epoch": 0.5649019267488283,
+      "grad_norm": 0.1455078125,
+      "learning_rate": 0.02,
+      "loss": 1.5151,
+      "step": 6102
+    },
+    {
+      "epoch": 0.5649945032691084,
+      "grad_norm": 0.15625,
+      "learning_rate": 0.02,
+      "loss": 1.6001,
+      "step": 6103
+    },
+    {
+      "epoch": 0.5650870797893884,
+      "grad_norm": 0.146484375,
+      "learning_rate": 0.02,
+      "loss": 1.6102,
+      "step": 6104
+    },
+    {
+      "epoch": 0.5651796563096685,
+      "grad_norm": 0.1533203125,
+      "learning_rate": 0.02,
+      "loss": 1.5519,
+      "step": 6105
+    },
+    {
+      "epoch": 0.5652722328299485,
+      "grad_norm": 0.12890625,
+      "learning_rate": 0.02,
+      "loss": 1.5612,
+      "step": 6106
+    },
+    {
+      "epoch": 0.5653648093502286,
+      "grad_norm": 0.146484375,
+      "learning_rate": 0.02,
+      "loss": 1.5942,
+      "step": 6107
+    },
+    {
+      "epoch": 0.5654573858705086,
+      "grad_norm": 0.1396484375,
+      "learning_rate": 0.02,
+      "loss": 1.574,
+      "step": 6108
+    },
+    {
+      "epoch": 0.5655499623907886,
+      "grad_norm": 0.1513671875,
+      "learning_rate": 0.02,
+      "loss": 1.5701,
+      "step": 6109
+    },
+    {
+      "epoch": 0.5656425389110686,
+      "grad_norm": 0.1435546875,
+      "learning_rate": 0.02,
+      "loss": 1.5858,
+      "step": 6110
+    },
+    {
+      "epoch": 0.5657351154313487,
+      "grad_norm": 0.16015625,
+      "learning_rate": 0.02,
+      "loss": 1.5318,
+      "step": 6111
+    },
+    {
+      "epoch": 0.5658276919516287,
+      "grad_norm": 0.14453125,
+      "learning_rate": 0.02,
+      "loss": 1.5587,
+      "step": 6112
+    },
+    {
+      "epoch": 0.5659202684719088,
+      "grad_norm": 0.1484375,
+      "learning_rate": 0.02,
+      "loss": 1.571,
+      "step": 6113
+    },
+    {
+      "epoch": 0.5660128449921888,
+      "grad_norm": 0.1474609375,
+      "learning_rate": 0.02,
+      "loss": 1.5667,
+      "step": 6114
+    },
+    {
+      "epoch": 0.5661054215124689,
+      "grad_norm": 0.1552734375,
+      "learning_rate": 0.02,
+      "loss": 1.5175,
+      "step": 6115
+    },
+    {
+      "epoch": 0.566197998032749,
+      "grad_norm": 0.1435546875,
+      "learning_rate": 0.02,
+      "loss": 1.4651,
+      "step": 6116
+    },
+    {
+      "epoch": 0.566290574553029,
+      "grad_norm": 0.15234375,
+      "learning_rate": 0.02,
+      "loss": 1.5533,
+      "step": 6117
+    },
+    {
+      "epoch": 0.566383151073309,
+      "grad_norm": 0.1328125,
+      "learning_rate": 0.02,
+      "loss": 1.6063,
+      "step": 6118
+    },
+    {
+      "epoch": 0.5664757275935891,
+      "grad_norm": 0.1484375,
+      "learning_rate": 0.02,
+      "loss": 1.5644,
+      "step": 6119
+    },
+    {
+      "epoch": 0.5665683041138692,
+      "grad_norm": 0.14453125,
+      "learning_rate": 0.02,
+      "loss": 1.5103,
+      "step": 6120
+    },
+    {
+      "epoch": 0.5666608806341492,
+      "grad_norm": 0.158203125,
+      "learning_rate": 0.02,
+      "loss": 1.5295,
+      "step": 6121
+    },
+    {
+      "epoch": 0.5667534571544293,
+      "grad_norm": 0.1455078125,
+      "learning_rate": 0.02,
+      "loss": 1.4538,
+      "step": 6122
+    },
+    {
+      "epoch": 0.5668460336747092,
+      "grad_norm": 0.1591796875,
+      "learning_rate": 0.02,
+      "loss": 1.5419,
+      "step": 6123
+    },
+    {
+      "epoch": 0.5669386101949893,
+      "grad_norm": 0.1318359375,
+      "learning_rate": 0.02,
+      "loss": 1.5419,
+      "step": 6124
+    },
+    {
+      "epoch": 0.5670311867152693,
+      "grad_norm": 0.1416015625,
+      "learning_rate": 0.02,
+      "loss": 1.6051,
+      "step": 6125
+    },
+    {
+      "epoch": 0.5671237632355494,
+      "grad_norm": 0.1376953125,
+      "learning_rate": 0.02,
+      "loss": 1.5779,
+      "step": 6126
+    },
+    {
+      "epoch": 0.5672163397558294,
+      "grad_norm": 0.1484375,
+      "learning_rate": 0.02,
+      "loss": 1.5425,
+      "step": 6127
+    },
+    {
+      "epoch": 0.5673089162761095,
+      "grad_norm": 0.146484375,
+      "learning_rate": 0.02,
+      "loss": 1.6006,
+      "step": 6128
+    },
+    {
+      "epoch": 0.5674014927963895,
+      "grad_norm": 0.1474609375,
+      "learning_rate": 0.02,
+      "loss": 1.5269,
+      "step": 6129
+    },
+    {
+      "epoch": 0.5674940693166696,
+      "grad_norm": 0.1376953125,
+      "learning_rate": 0.02,
+      "loss": 1.6145,
+      "step": 6130
+    },
+    {
+      "epoch": 0.5675866458369496,
+      "grad_norm": 0.1474609375,
+      "learning_rate": 0.02,
+      "loss": 1.552,
+      "step": 6131
+    },
+    {
+      "epoch": 0.5676792223572297,
+      "grad_norm": 0.1416015625,
+      "learning_rate": 0.02,
+      "loss": 1.4956,
+      "step": 6132
+    },
+    {
+      "epoch": 0.5677717988775097,
+      "grad_norm": 0.14453125,
+      "learning_rate": 0.02,
+      "loss": 1.5368,
+      "step": 6133
+    },
+    {
+      "epoch": 0.5678643753977898,
+      "grad_norm": 0.134765625,
+      "learning_rate": 0.02,
+      "loss": 1.552,
+      "step": 6134
+    },
+    {
+      "epoch": 0.5679569519180698,
+      "grad_norm": 0.154296875,
+      "learning_rate": 0.02,
+      "loss": 1.6023,
+      "step": 6135
+    },
+    {
+      "epoch": 0.5680495284383498,
+      "grad_norm": 0.154296875,
+      "learning_rate": 0.02,
+      "loss": 1.5595,
+      "step": 6136
+    },
+    {
+      "epoch": 0.5681421049586298,
+      "grad_norm": 0.1533203125,
+      "learning_rate": 0.02,
+      "loss": 1.621,
+      "step": 6137
+    },
+    {
+      "epoch": 0.5682346814789099,
+      "grad_norm": 0.1455078125,
+      "learning_rate": 0.02,
+      "loss": 1.6304,
+      "step": 6138
+    },
+    {
+      "epoch": 0.5683272579991899,
+      "grad_norm": 0.1552734375,
+      "learning_rate": 0.02,
+      "loss": 1.5273,
+      "step": 6139
+    },
+    {
+      "epoch": 0.56841983451947,
+      "grad_norm": 0.1494140625,
+      "learning_rate": 0.02,
+      "loss": 1.459,
+      "step": 6140
+    },
+    {
+      "epoch": 0.56851241103975,
+      "grad_norm": 0.1513671875,
+      "learning_rate": 0.02,
+      "loss": 1.5311,
+      "step": 6141
+    },
+    {
+      "epoch": 0.5686049875600301,
+      "grad_norm": 0.1376953125,
+      "learning_rate": 0.02,
+      "loss": 1.5889,
+      "step": 6142
+    },
+    {
+      "epoch": 0.5686975640803101,
+      "grad_norm": 0.1435546875,
+      "learning_rate": 0.02,
+      "loss": 1.5465,
+      "step": 6143
+    },
+    {
+      "epoch": 0.5687901406005902,
+      "grad_norm": 0.140625,
+      "learning_rate": 0.02,
+      "loss": 1.5221,
+      "step": 6144
+    },
+    {
+      "epoch": 0.5688827171208702,
+      "grad_norm": 0.154296875,
+      "learning_rate": 0.02,
+      "loss": 1.6692,
+      "step": 6145
+    },
+    {
+      "epoch": 0.5689752936411503,
+      "grad_norm": 0.1376953125,
+      "learning_rate": 0.02,
+      "loss": 1.5939,
+      "step": 6146
+    },
+    {
+      "epoch": 0.5690678701614303,
+      "grad_norm": 0.1474609375,
+      "learning_rate": 0.02,
+      "loss": 1.6172,
+      "step": 6147
+    },
+    {
+      "epoch": 0.5691604466817104,
+      "grad_norm": 0.154296875,
+      "learning_rate": 0.02,
+      "loss": 1.5864,
+      "step": 6148
+    },
+    {
+      "epoch": 0.5692530232019904,
+      "grad_norm": 0.1318359375,
+      "learning_rate": 0.02,
+      "loss": 1.4831,
+      "step": 6149
+    },
+    {
+      "epoch": 0.5693455997222704,
+      "grad_norm": 0.1484375,
+      "learning_rate": 0.02,
+      "loss": 1.5976,
+      "step": 6150
+    },
+    {
+      "epoch": 0.5694381762425504,
+      "grad_norm": 0.140625,
+      "learning_rate": 0.02,
+      "loss": 1.5434,
+      "step": 6151
+    },
+    {
+      "epoch": 0.5695307527628305,
+      "grad_norm": 0.1396484375,
+      "learning_rate": 0.02,
+      "loss": 1.4685,
+      "step": 6152
+    },
+    {
+      "epoch": 0.5696233292831105,
+      "grad_norm": 0.15234375,
+      "learning_rate": 0.02,
+      "loss": 1.5528,
+      "step": 6153
+    },
+    {
+      "epoch": 0.5697159058033906,
+      "grad_norm": 0.1513671875,
+      "learning_rate": 0.02,
+      "loss": 1.5527,
+      "step": 6154
+    },
+    {
+      "epoch": 0.5698084823236707,
+      "grad_norm": 0.1533203125,
+      "learning_rate": 0.02,
+      "loss": 1.656,
+      "step": 6155
+    },
+    {
+      "epoch": 0.5699010588439507,
+      "grad_norm": 0.1484375,
+      "learning_rate": 0.02,
+      "loss": 1.5508,
+      "step": 6156
+    },
+    {
+      "epoch": 0.5699936353642308,
+      "grad_norm": 0.1513671875,
+      "learning_rate": 0.02,
+      "loss": 1.5532,
+      "step": 6157
+    },
+    {
+      "epoch": 0.5700862118845108,
+      "grad_norm": 0.142578125,
+      "learning_rate": 0.02,
+      "loss": 1.5807,
+      "step": 6158
+    },
+    {
+      "epoch": 0.5701787884047909,
+      "grad_norm": 0.1484375,
+      "learning_rate": 0.02,
+      "loss": 1.5246,
+      "step": 6159
+    },
+    {
+      "epoch": 0.5702713649250709,
+      "grad_norm": 0.142578125,
+      "learning_rate": 0.02,
+      "loss": 1.5369,
+      "step": 6160
+    },
+    {
+      "epoch": 0.570363941445351,
+      "grad_norm": 0.1533203125,
+      "learning_rate": 0.02,
+      "loss": 1.5218,
+      "step": 6161
+    },
+    {
+      "epoch": 0.570456517965631,
+      "grad_norm": 0.1396484375,
+      "learning_rate": 0.02,
+      "loss": 1.535,
+      "step": 6162
+    },
+    {
+      "epoch": 0.570549094485911,
+      "grad_norm": 0.15234375,
+      "learning_rate": 0.02,
+      "loss": 1.5569,
+      "step": 6163
+    },
+    {
+      "epoch": 0.570641671006191,
+      "grad_norm": 0.142578125,
+      "learning_rate": 0.02,
+      "loss": 1.5727,
+      "step": 6164
+    },
+    {
+      "epoch": 0.5707342475264711,
+      "grad_norm": 0.150390625,
+      "learning_rate": 0.02,
+      "loss": 1.5428,
+      "step": 6165
+    },
+    {
+      "epoch": 0.5708268240467511,
+      "grad_norm": 0.1455078125,
+      "learning_rate": 0.02,
+      "loss": 1.5897,
+      "step": 6166
+    },
+    {
+      "epoch": 0.5709194005670312,
+      "grad_norm": 0.14453125,
+      "learning_rate": 0.02,
+      "loss": 1.5386,
+      "step": 6167
+    },
+    {
+      "epoch": 0.5710119770873112,
+      "grad_norm": 0.1435546875,
+      "learning_rate": 0.02,
+      "loss": 1.5456,
+      "step": 6168
+    },
+    {
+      "epoch": 0.5711045536075913,
+      "grad_norm": 0.142578125,
+      "learning_rate": 0.02,
+      "loss": 1.54,
+      "step": 6169
+    },
+    {
+      "epoch": 0.5711971301278713,
+      "grad_norm": 0.140625,
+      "learning_rate": 0.02,
+      "loss": 1.4929,
+      "step": 6170
+    },
+    {
+      "epoch": 0.5712897066481514,
+      "grad_norm": 0.1376953125,
+      "learning_rate": 0.02,
+      "loss": 1.5632,
+      "step": 6171
+    },
+    {
+      "epoch": 0.5713822831684314,
+      "grad_norm": 0.1513671875,
+      "learning_rate": 0.02,
+      "loss": 1.6497,
+      "step": 6172
+    },
+    {
+      "epoch": 0.5714748596887115,
+      "grad_norm": 0.1357421875,
+      "learning_rate": 0.02,
+      "loss": 1.5095,
+      "step": 6173
+    },
+    {
+      "epoch": 0.5715674362089915,
+      "grad_norm": 0.1435546875,
+      "learning_rate": 0.02,
+      "loss": 1.5955,
+      "step": 6174
+    },
+    {
+      "epoch": 0.5716600127292716,
+      "grad_norm": 0.142578125,
+      "learning_rate": 0.02,
+      "loss": 1.577,
+      "step": 6175
+    },
+    {
+      "epoch": 0.5717525892495516,
+      "grad_norm": 0.1484375,
+      "learning_rate": 0.02,
+      "loss": 1.5497,
+      "step": 6176
+    },
+    {
+      "epoch": 0.5718451657698316,
+      "grad_norm": 0.14453125,
+      "learning_rate": 0.02,
+      "loss": 1.5829,
+      "step": 6177
+    },
+    {
+      "epoch": 0.5719377422901116,
+      "grad_norm": 0.1484375,
+      "learning_rate": 0.02,
+      "loss": 1.6288,
+      "step": 6178
+    },
+    {
+      "epoch": 0.5720303188103917,
+      "grad_norm": 0.1484375,
+      "learning_rate": 0.02,
+      "loss": 1.527,
+      "step": 6179
+    },
+    {
+      "epoch": 0.5721228953306717,
+      "grad_norm": 0.154296875,
+      "learning_rate": 0.02,
+      "loss": 1.5059,
+      "step": 6180
+    },
+    {
+      "epoch": 0.5722154718509518,
+      "grad_norm": 0.142578125,
+      "learning_rate": 0.02,
+      "loss": 1.5368,
+      "step": 6181
+    },
+    {
+      "epoch": 0.5723080483712318,
+      "grad_norm": 0.16015625,
+      "learning_rate": 0.02,
+      "loss": 1.6413,
+      "step": 6182
+    },
+    {
+      "epoch": 0.5724006248915119,
+      "grad_norm": 0.15234375,
+      "learning_rate": 0.02,
+      "loss": 1.5558,
+      "step": 6183
+    },
+    {
+      "epoch": 0.5724932014117919,
+      "grad_norm": 0.146484375,
+      "learning_rate": 0.02,
+      "loss": 1.4733,
+      "step": 6184
+    },
+    {
+      "epoch": 0.572585777932072,
+      "grad_norm": 0.14453125,
+      "learning_rate": 0.02,
+      "loss": 1.5474,
+      "step": 6185
+    },
+    {
+      "epoch": 0.572678354452352,
+      "grad_norm": 0.1474609375,
+      "learning_rate": 0.02,
+      "loss": 1.5813,
+      "step": 6186
+    },
+    {
+      "epoch": 0.5727709309726321,
+      "grad_norm": 0.162109375,
+      "learning_rate": 0.02,
+      "loss": 1.6214,
+      "step": 6187
+    },
+    {
+      "epoch": 0.5728635074929121,
+      "grad_norm": 0.158203125,
+      "learning_rate": 0.02,
+      "loss": 1.5323,
+      "step": 6188
+    },
+    {
+      "epoch": 0.5729560840131922,
+      "grad_norm": 0.1484375,
+      "learning_rate": 0.02,
+      "loss": 1.5501,
+      "step": 6189
+    },
+    {
+      "epoch": 0.5730486605334721,
+      "grad_norm": 0.16015625,
+      "learning_rate": 0.02,
+      "loss": 1.5082,
+      "step": 6190
+    },
+    {
+      "epoch": 0.5731412370537522,
+      "grad_norm": 0.1494140625,
+      "learning_rate": 0.02,
+      "loss": 1.5008,
+      "step": 6191
+    },
+    {
+      "epoch": 0.5732338135740322,
+      "grad_norm": 0.1572265625,
+      "learning_rate": 0.02,
+      "loss": 1.5606,
+      "step": 6192
+    },
+    {
+      "epoch": 0.5733263900943123,
+      "grad_norm": 0.1494140625,
+      "learning_rate": 0.02,
+      "loss": 1.5407,
+      "step": 6193
+    },
+    {
+      "epoch": 0.5734189666145924,
+      "grad_norm": 0.146484375,
+      "learning_rate": 0.02,
+      "loss": 1.5331,
+      "step": 6194
+    },
+    {
+      "epoch": 0.5735115431348724,
+      "grad_norm": 0.1474609375,
+      "learning_rate": 0.02,
+      "loss": 1.4957,
+      "step": 6195
+    },
+    {
+      "epoch": 0.5736041196551525,
+      "grad_norm": 0.15234375,
+      "learning_rate": 0.02,
+      "loss": 1.5677,
+      "step": 6196
+    },
+    {
+      "epoch": 0.5736966961754325,
+      "grad_norm": 0.154296875,
+      "learning_rate": 0.02,
+      "loss": 1.6088,
+      "step": 6197
+    },
+    {
+      "epoch": 0.5737892726957126,
+      "grad_norm": 0.15234375,
+      "learning_rate": 0.02,
+      "loss": 1.5389,
+      "step": 6198
+    },
+    {
+      "epoch": 0.5738818492159926,
+      "grad_norm": 0.146484375,
+      "learning_rate": 0.02,
+      "loss": 1.5088,
+      "step": 6199
+    },
+    {
+      "epoch": 0.5739744257362727,
+      "grad_norm": 0.142578125,
+      "learning_rate": 0.02,
+      "loss": 1.5406,
+      "step": 6200
+    },
+    {
+      "epoch": 0.5740670022565527,
+      "grad_norm": 0.1484375,
+      "learning_rate": 0.02,
+      "loss": 1.4772,
+      "step": 6201
+    },
+    {
+      "epoch": 0.5741595787768328,
+      "grad_norm": 0.1494140625,
+      "learning_rate": 0.02,
+      "loss": 1.5772,
+      "step": 6202
+    },
+    {
+      "epoch": 0.5742521552971128,
+      "grad_norm": 0.1513671875,
+      "learning_rate": 0.02,
+      "loss": 1.558,
+      "step": 6203
+    },
+    {
+      "epoch": 0.5743447318173928,
+      "grad_norm": 0.1591796875,
+      "learning_rate": 0.02,
+      "loss": 1.5736,
+      "step": 6204
+    },
+    {
+      "epoch": 0.5744373083376728,
+      "grad_norm": 0.142578125,
+      "learning_rate": 0.02,
+      "loss": 1.4973,
+      "step": 6205
+    },
+    {
+      "epoch": 0.5745298848579529,
+      "grad_norm": 0.1455078125,
+      "learning_rate": 0.02,
+      "loss": 1.5749,
+      "step": 6206
+    },
+    {
+      "epoch": 0.5746224613782329,
+      "grad_norm": 0.15234375,
+      "learning_rate": 0.02,
+      "loss": 1.6057,
+      "step": 6207
+    },
+    {
+      "epoch": 0.574715037898513,
+      "grad_norm": 0.154296875,
+      "learning_rate": 0.02,
+      "loss": 1.6144,
+      "step": 6208
+    },
+    {
+      "epoch": 0.574807614418793,
+      "grad_norm": 0.1513671875,
+      "learning_rate": 0.02,
+      "loss": 1.5222,
+      "step": 6209
+    },
+    {
+      "epoch": 0.5749001909390731,
+      "grad_norm": 0.150390625,
+      "learning_rate": 0.02,
+      "loss": 1.5857,
+      "step": 6210
+    },
+    {
+      "epoch": 0.5749927674593531,
+      "grad_norm": 0.1572265625,
+      "learning_rate": 0.02,
+      "loss": 1.6274,
+      "step": 6211
+    },
+    {
+      "epoch": 0.5750853439796332,
+      "grad_norm": 0.1572265625,
+      "learning_rate": 0.02,
+      "loss": 1.5724,
+      "step": 6212
+    },
+    {
+      "epoch": 0.5751779204999132,
+      "grad_norm": 0.154296875,
+      "learning_rate": 0.02,
+      "loss": 1.5634,
+      "step": 6213
+    },
+    {
+      "epoch": 0.5752704970201933,
+      "grad_norm": 0.1474609375,
+      "learning_rate": 0.02,
+      "loss": 1.5589,
+      "step": 6214
+    },
+    {
+      "epoch": 0.5753630735404733,
+      "grad_norm": 0.142578125,
+      "learning_rate": 0.02,
+      "loss": 1.5267,
+      "step": 6215
+    },
+    {
+      "epoch": 0.5754556500607534,
+      "grad_norm": 0.150390625,
+      "learning_rate": 0.02,
+      "loss": 1.58,
+      "step": 6216
+    },
+    {
+      "epoch": 0.5755482265810334,
+      "grad_norm": 0.1455078125,
+      "learning_rate": 0.02,
+      "loss": 1.5863,
+      "step": 6217
+    },
+    {
+      "epoch": 0.5756408031013134,
+      "grad_norm": 0.15234375,
+      "learning_rate": 0.02,
+      "loss": 1.5335,
+      "step": 6218
+    },
+    {
+      "epoch": 0.5757333796215934,
+      "grad_norm": 0.146484375,
+      "learning_rate": 0.02,
+      "loss": 1.5579,
+      "step": 6219
+    },
+    {
+      "epoch": 0.5758259561418735,
+      "grad_norm": 0.16015625,
+      "learning_rate": 0.02,
+      "loss": 1.5358,
+      "step": 6220
+    },
+    {
+      "epoch": 0.5759185326621535,
+      "grad_norm": 0.1484375,
+      "learning_rate": 0.02,
+      "loss": 1.5837,
+      "step": 6221
+    },
+    {
+      "epoch": 0.5760111091824336,
+      "grad_norm": 0.1533203125,
+      "learning_rate": 0.02,
+      "loss": 1.5617,
+      "step": 6222
+    },
+    {
+      "epoch": 0.5761036857027136,
+      "grad_norm": 0.1435546875,
+      "learning_rate": 0.02,
+      "loss": 1.6017,
+      "step": 6223
+    },
+    {
+      "epoch": 0.5761962622229937,
+      "grad_norm": 0.1591796875,
+      "learning_rate": 0.02,
+      "loss": 1.5398,
+      "step": 6224
+    },
+    {
+      "epoch": 0.5762888387432737,
+      "grad_norm": 0.1494140625,
+      "learning_rate": 0.02,
+      "loss": 1.5637,
+      "step": 6225
+    },
+    {
+      "epoch": 0.5763814152635538,
+      "grad_norm": 0.1484375,
+      "learning_rate": 0.02,
+      "loss": 1.5417,
+      "step": 6226
+    },
+    {
+      "epoch": 0.5764739917838339,
+      "grad_norm": 0.1494140625,
+      "learning_rate": 0.02,
+      "loss": 1.5244,
+      "step": 6227
+    },
+    {
+      "epoch": 0.5765665683041139,
+      "grad_norm": 0.1513671875,
+      "learning_rate": 0.02,
+      "loss": 1.634,
+      "step": 6228
+    },
+    {
+      "epoch": 0.576659144824394,
+      "grad_norm": 0.1435546875,
+      "learning_rate": 0.02,
+      "loss": 1.5735,
+      "step": 6229
+    },
+    {
+      "epoch": 0.576751721344674,
+      "grad_norm": 0.140625,
+      "learning_rate": 0.02,
+      "loss": 1.5315,
+      "step": 6230
+    },
+    {
+      "epoch": 0.576844297864954,
+      "grad_norm": 0.1494140625,
+      "learning_rate": 0.02,
+      "loss": 1.5709,
+      "step": 6231
+    },
+    {
+      "epoch": 0.576936874385234,
+      "grad_norm": 0.142578125,
+      "learning_rate": 0.02,
+      "loss": 1.627,
+      "step": 6232
+    },
+    {
+      "epoch": 0.577029450905514,
+      "grad_norm": 0.1435546875,
+      "learning_rate": 0.02,
+      "loss": 1.5224,
+      "step": 6233
+    },
+    {
+      "epoch": 0.5771220274257941,
+      "grad_norm": 0.1376953125,
+      "learning_rate": 0.02,
+      "loss": 1.527,
+      "step": 6234
+    },
+    {
+      "epoch": 0.5772146039460742,
+      "grad_norm": 0.1474609375,
+      "learning_rate": 0.02,
+      "loss": 1.593,
+      "step": 6235
+    },
+    {
+      "epoch": 0.5773071804663542,
+      "grad_norm": 0.1416015625,
+      "learning_rate": 0.02,
+      "loss": 1.5215,
+      "step": 6236
+    },
+    {
+      "epoch": 0.5773997569866343,
+      "grad_norm": 0.1455078125,
+      "learning_rate": 0.02,
+      "loss": 1.489,
+      "step": 6237
+    },
+    {
+      "epoch": 0.5774923335069143,
+      "grad_norm": 0.1435546875,
+      "learning_rate": 0.02,
+      "loss": 1.5513,
+      "step": 6238
+    },
+    {
+      "epoch": 0.5775849100271944,
+      "grad_norm": 0.1416015625,
+      "learning_rate": 0.02,
+      "loss": 1.5078,
+      "step": 6239
+    },
+    {
+      "epoch": 0.5776774865474744,
+      "grad_norm": 0.1416015625,
+      "learning_rate": 0.02,
+      "loss": 1.4865,
+      "step": 6240
+    },
+    {
+      "epoch": 0.5777700630677545,
+      "grad_norm": 0.15625,
+      "learning_rate": 0.02,
+      "loss": 1.529,
+      "step": 6241
+    },
+    {
+      "epoch": 0.5778626395880345,
+      "grad_norm": 0.140625,
+      "learning_rate": 0.02,
+      "loss": 1.5794,
+      "step": 6242
+    },
+    {
+      "epoch": 0.5779552161083146,
+      "grad_norm": 0.15234375,
+      "learning_rate": 0.02,
+      "loss": 1.5606,
+      "step": 6243
+    },
+    {
+      "epoch": 0.5780477926285946,
+      "grad_norm": 0.1416015625,
+      "learning_rate": 0.02,
+      "loss": 1.5319,
+      "step": 6244
+    },
+    {
+      "epoch": 0.5781403691488746,
+      "grad_norm": 0.16015625,
+      "learning_rate": 0.02,
+      "loss": 1.5463,
+      "step": 6245
+    },
+    {
+      "epoch": 0.5782329456691546,
+      "grad_norm": 0.1533203125,
+      "learning_rate": 0.02,
+      "loss": 1.567,
+      "step": 6246
+    },
+    {
+      "epoch": 0.5783255221894347,
+      "grad_norm": 0.154296875,
+      "learning_rate": 0.02,
+      "loss": 1.5152,
+      "step": 6247
+    },
+    {
+      "epoch": 0.5784180987097147,
+      "grad_norm": 0.1484375,
+      "learning_rate": 0.02,
+      "loss": 1.5863,
+      "step": 6248
+    },
+    {
+      "epoch": 0.5785106752299948,
+      "grad_norm": 0.1455078125,
+      "learning_rate": 0.02,
+      "loss": 1.5728,
+      "step": 6249
+    },
+    {
+      "epoch": 0.5786032517502748,
+      "grad_norm": 0.1416015625,
+      "learning_rate": 0.02,
+      "loss": 1.5268,
+      "step": 6250
+    },
+    {
+      "epoch": 0.5786958282705549,
+      "grad_norm": 0.1416015625,
+      "learning_rate": 0.02,
+      "loss": 1.5605,
+      "step": 6251
+    },
+    {
+      "epoch": 0.5787884047908349,
+      "grad_norm": 0.1416015625,
+      "learning_rate": 0.02,
+      "loss": 1.5085,
+      "step": 6252
+    },
+    {
+      "epoch": 0.578880981311115,
+      "grad_norm": 0.1435546875,
+      "learning_rate": 0.02,
+      "loss": 1.5225,
+      "step": 6253
+    },
+    {
+      "epoch": 0.578973557831395,
+      "grad_norm": 0.140625,
+      "learning_rate": 0.02,
+      "loss": 1.5207,
+      "step": 6254
+    },
+    {
+      "epoch": 0.5790661343516751,
+      "grad_norm": 0.1396484375,
+      "learning_rate": 0.02,
+      "loss": 1.5876,
+      "step": 6255
+    },
+    {
+      "epoch": 0.5791587108719551,
+      "grad_norm": 0.1455078125,
+      "learning_rate": 0.02,
+      "loss": 1.6537,
+      "step": 6256
+    },
+    {
+      "epoch": 0.5792512873922352,
+      "grad_norm": 0.1416015625,
+      "learning_rate": 0.02,
+      "loss": 1.5429,
+      "step": 6257
+    },
+    {
+      "epoch": 0.5793438639125151,
+      "grad_norm": 0.1484375,
+      "learning_rate": 0.02,
+      "loss": 1.5922,
+      "step": 6258
+    },
+    {
+      "epoch": 0.5794364404327952,
+      "grad_norm": 0.14453125,
+      "learning_rate": 0.02,
+      "loss": 1.5318,
+      "step": 6259
+    },
+    {
+      "epoch": 0.5795290169530752,
+      "grad_norm": 0.1455078125,
+      "learning_rate": 0.02,
+      "loss": 1.5117,
+      "step": 6260
+    },
+    {
+      "epoch": 0.5796215934733553,
+      "grad_norm": 0.1640625,
+      "learning_rate": 0.02,
+      "loss": 1.5841,
+      "step": 6261
+    },
+    {
+      "epoch": 0.5797141699936353,
+      "grad_norm": 0.138671875,
+      "learning_rate": 0.02,
+      "loss": 1.541,
+      "step": 6262
+    },
+    {
+      "epoch": 0.5798067465139154,
+      "grad_norm": 0.140625,
+      "learning_rate": 0.02,
+      "loss": 1.5798,
+      "step": 6263
+    },
+    {
+      "epoch": 0.5798993230341954,
+      "grad_norm": 0.1396484375,
+      "learning_rate": 0.02,
+      "loss": 1.551,
+      "step": 6264
+    },
+    {
+      "epoch": 0.5799918995544755,
+      "grad_norm": 0.15234375,
+      "learning_rate": 0.02,
+      "loss": 1.537,
+      "step": 6265
+    },
+    {
+      "epoch": 0.5800844760747556,
+      "grad_norm": 0.1474609375,
+      "learning_rate": 0.02,
+      "loss": 1.5439,
+      "step": 6266
+    },
+    {
+      "epoch": 0.5801770525950356,
+      "grad_norm": 0.1484375,
+      "learning_rate": 0.02,
+      "loss": 1.4923,
+      "step": 6267
+    },
+    {
+      "epoch": 0.5802696291153157,
+      "grad_norm": 0.1396484375,
+      "learning_rate": 0.02,
+      "loss": 1.4585,
+      "step": 6268
+    },
+    {
+      "epoch": 0.5803622056355957,
+      "grad_norm": 0.1533203125,
+      "learning_rate": 0.02,
+      "loss": 1.5752,
+      "step": 6269
+    },
+    {
+      "epoch": 0.5804547821558758,
+      "grad_norm": 0.1572265625,
+      "learning_rate": 0.02,
+      "loss": 1.5792,
+      "step": 6270
+    },
+    {
+      "epoch": 0.5805473586761558,
+      "grad_norm": 0.15625,
+      "learning_rate": 0.02,
+      "loss": 1.5695,
+      "step": 6271
+    },
+    {
+      "epoch": 0.5806399351964358,
+      "grad_norm": 0.13671875,
+      "learning_rate": 0.02,
+      "loss": 1.538,
+      "step": 6272
+    },
+    {
+      "epoch": 0.5807325117167158,
+      "grad_norm": 0.1474609375,
+      "learning_rate": 0.02,
+      "loss": 1.5323,
+      "step": 6273
+    },
+    {
+      "epoch": 0.5808250882369959,
+      "grad_norm": 0.13671875,
+      "learning_rate": 0.02,
+      "loss": 1.5639,
+      "step": 6274
+    },
+    {
+      "epoch": 0.5809176647572759,
+      "grad_norm": 0.1484375,
+      "learning_rate": 0.02,
+      "loss": 1.5445,
+      "step": 6275
+    },
+    {
+      "epoch": 0.581010241277556,
+      "grad_norm": 0.1474609375,
+      "learning_rate": 0.02,
+      "loss": 1.5044,
+      "step": 6276
+    },
+    {
+      "epoch": 0.581102817797836,
+      "grad_norm": 0.1455078125,
+      "learning_rate": 0.02,
+      "loss": 1.4582,
+      "step": 6277
+    },
+    {
+      "epoch": 0.5811953943181161,
+      "grad_norm": 0.14453125,
+      "learning_rate": 0.02,
+      "loss": 1.5625,
+      "step": 6278
+    },
+    {
+      "epoch": 0.5812879708383961,
+      "grad_norm": 0.140625,
+      "learning_rate": 0.02,
+      "loss": 1.553,
+      "step": 6279
+    },
+    {
+      "epoch": 0.5813805473586762,
+      "grad_norm": 0.15234375,
+      "learning_rate": 0.02,
+      "loss": 1.5976,
+      "step": 6280
+    },
+    {
+      "epoch": 0.5814731238789562,
+      "grad_norm": 0.154296875,
+      "learning_rate": 0.02,
+      "loss": 1.5379,
+      "step": 6281
+    },
+    {
+      "epoch": 0.5815657003992363,
+      "grad_norm": 0.142578125,
+      "learning_rate": 0.02,
+      "loss": 1.5477,
+      "step": 6282
+    },
+    {
+      "epoch": 0.5816582769195163,
+      "grad_norm": 0.1572265625,
+      "learning_rate": 0.02,
+      "loss": 1.5659,
+      "step": 6283
+    },
+    {
+      "epoch": 0.5817508534397964,
+      "grad_norm": 0.1484375,
+      "learning_rate": 0.02,
+      "loss": 1.4831,
+      "step": 6284
+    },
+    {
+      "epoch": 0.5818434299600764,
+      "grad_norm": 0.1494140625,
+      "learning_rate": 0.02,
+      "loss": 1.5283,
+      "step": 6285
+    },
+    {
+      "epoch": 0.5819360064803564,
+      "grad_norm": 0.1513671875,
+      "learning_rate": 0.02,
+      "loss": 1.518,
+      "step": 6286
+    },
+    {
+      "epoch": 0.5820285830006364,
+      "grad_norm": 0.1513671875,
+      "learning_rate": 0.02,
+      "loss": 1.5323,
+      "step": 6287
+    },
+    {
+      "epoch": 0.5821211595209165,
+      "grad_norm": 0.1435546875,
+      "learning_rate": 0.02,
+      "loss": 1.5101,
+      "step": 6288
+    },
+    {
+      "epoch": 0.5822137360411965,
+      "grad_norm": 0.1484375,
+      "learning_rate": 0.02,
+      "loss": 1.5626,
+      "step": 6289
+    },
+    {
+      "epoch": 0.5823063125614766,
+      "grad_norm": 0.14453125,
+      "learning_rate": 0.02,
+      "loss": 1.548,
+      "step": 6290
+    },
+    {
+      "epoch": 0.5823988890817566,
+      "grad_norm": 0.1435546875,
+      "learning_rate": 0.02,
+      "loss": 1.4608,
+      "step": 6291
+    },
+    {
+      "epoch": 0.5824914656020367,
+      "grad_norm": 0.146484375,
+      "learning_rate": 0.02,
+      "loss": 1.5382,
+      "step": 6292
+    },
+    {
+      "epoch": 0.5825840421223167,
+      "grad_norm": 0.1513671875,
+      "learning_rate": 0.02,
+      "loss": 1.5461,
+      "step": 6293
+    },
+    {
+      "epoch": 0.5826766186425968,
+      "grad_norm": 0.14453125,
+      "learning_rate": 0.02,
+      "loss": 1.5345,
+      "step": 6294
+    },
+    {
+      "epoch": 0.5827691951628768,
+      "grad_norm": 0.1474609375,
+      "learning_rate": 0.02,
+      "loss": 1.5537,
+      "step": 6295
+    },
+    {
+      "epoch": 0.5828617716831569,
+      "grad_norm": 0.15234375,
+      "learning_rate": 0.02,
+      "loss": 1.5838,
+      "step": 6296
+    },
+    {
+      "epoch": 0.582954348203437,
+      "grad_norm": 0.16015625,
+      "learning_rate": 0.02,
+      "loss": 1.5383,
+      "step": 6297
+    },
+    {
+      "epoch": 0.583046924723717,
+      "grad_norm": 0.1484375,
+      "learning_rate": 0.02,
+      "loss": 1.5583,
+      "step": 6298
+    },
+    {
+      "epoch": 0.5831395012439969,
+      "grad_norm": 0.15234375,
+      "learning_rate": 0.02,
+      "loss": 1.5919,
+      "step": 6299
+    },
+    {
+      "epoch": 0.583232077764277,
+      "grad_norm": 0.1416015625,
+      "learning_rate": 0.02,
+      "loss": 1.5756,
+      "step": 6300
+    },
+    {
+      "epoch": 0.583324654284557,
+      "grad_norm": 0.14453125,
+      "learning_rate": 0.02,
+      "loss": 1.591,
+      "step": 6301
+    },
+    {
+      "epoch": 0.5834172308048371,
+      "grad_norm": 0.1376953125,
+      "learning_rate": 0.02,
+      "loss": 1.4678,
+      "step": 6302
+    },
+    {
+      "epoch": 0.5835098073251171,
+      "grad_norm": 0.1484375,
+      "learning_rate": 0.02,
+      "loss": 1.5689,
+      "step": 6303
+    },
+    {
+      "epoch": 0.5836023838453972,
+      "grad_norm": 0.146484375,
+      "learning_rate": 0.02,
+      "loss": 1.5426,
+      "step": 6304
+    },
+    {
+      "epoch": 0.5836949603656773,
+      "grad_norm": 0.15234375,
+      "learning_rate": 0.02,
+      "loss": 1.5045,
+      "step": 6305
+    },
+    {
+      "epoch": 0.5837875368859573,
+      "grad_norm": 0.140625,
+      "learning_rate": 0.02,
+      "loss": 1.5852,
+      "step": 6306
+    },
+    {
+      "epoch": 0.5838801134062374,
+      "grad_norm": 0.1591796875,
+      "learning_rate": 0.02,
+      "loss": 1.5393,
+      "step": 6307
+    },
+    {
+      "epoch": 0.5839726899265174,
+      "grad_norm": 0.1474609375,
+      "learning_rate": 0.02,
+      "loss": 1.5541,
+      "step": 6308
+    },
+    {
+      "epoch": 0.5840652664467975,
+      "grad_norm": 0.1533203125,
+      "learning_rate": 0.02,
+      "loss": 1.5501,
+      "step": 6309
+    },
+    {
+      "epoch": 0.5841578429670775,
+      "grad_norm": 0.1455078125,
+      "learning_rate": 0.02,
+      "loss": 1.5962,
+      "step": 6310
+    },
+    {
+      "epoch": 0.5842504194873576,
+      "grad_norm": 0.1552734375,
+      "learning_rate": 0.02,
+      "loss": 1.5288,
+      "step": 6311
+    },
+    {
+      "epoch": 0.5843429960076376,
+      "grad_norm": 0.1484375,
+      "learning_rate": 0.02,
+      "loss": 1.4677,
+      "step": 6312
+    },
+    {
+      "epoch": 0.5844355725279176,
+      "grad_norm": 0.142578125,
+      "learning_rate": 0.02,
+      "loss": 1.5353,
+      "step": 6313
+    },
+    {
+      "epoch": 0.5845281490481976,
+      "grad_norm": 0.142578125,
+      "learning_rate": 0.02,
+      "loss": 1.4701,
+      "step": 6314
+    },
+    {
+      "epoch": 0.5846207255684777,
+      "grad_norm": 0.158203125,
+      "learning_rate": 0.02,
+      "loss": 1.5017,
+      "step": 6315
+    },
+    {
+      "epoch": 0.5847133020887577,
+      "grad_norm": 0.1513671875,
+      "learning_rate": 0.02,
+      "loss": 1.4944,
+      "step": 6316
+    },
+    {
+      "epoch": 0.5848058786090378,
+      "grad_norm": 0.15234375,
+      "learning_rate": 0.02,
+      "loss": 1.5353,
+      "step": 6317
+    },
+    {
+      "epoch": 0.5848984551293178,
+      "grad_norm": 0.1611328125,
+      "learning_rate": 0.02,
+      "loss": 1.6036,
+      "step": 6318
+    },
+    {
+      "epoch": 0.5849910316495979,
+      "grad_norm": 0.146484375,
+      "learning_rate": 0.02,
+      "loss": 1.5741,
+      "step": 6319
+    },
+    {
+      "epoch": 0.5850836081698779,
+      "grad_norm": 0.154296875,
+      "learning_rate": 0.02,
+      "loss": 1.5497,
+      "step": 6320
+    },
+    {
+      "epoch": 0.585176184690158,
+      "grad_norm": 0.1455078125,
+      "learning_rate": 0.02,
+      "loss": 1.5268,
+      "step": 6321
+    },
+    {
+      "epoch": 0.585268761210438,
+      "grad_norm": 0.1494140625,
+      "learning_rate": 0.02,
+      "loss": 1.5545,
+      "step": 6322
+    },
+    {
+      "epoch": 0.5853613377307181,
+      "grad_norm": 0.1455078125,
+      "learning_rate": 0.02,
+      "loss": 1.6326,
+      "step": 6323
+    },
+    {
+      "epoch": 0.5854539142509981,
+      "grad_norm": 0.1552734375,
+      "learning_rate": 0.02,
+      "loss": 1.572,
+      "step": 6324
+    },
+    {
+      "epoch": 0.5855464907712782,
+      "grad_norm": 0.150390625,
+      "learning_rate": 0.02,
+      "loss": 1.553,
+      "step": 6325
+    },
+    {
+      "epoch": 0.5856390672915581,
+      "grad_norm": 0.15625,
+      "learning_rate": 0.02,
+      "loss": 1.5608,
+      "step": 6326
+    },
+    {
+      "epoch": 0.5857316438118382,
+      "grad_norm": 0.162109375,
+      "learning_rate": 0.02,
+      "loss": 1.528,
+      "step": 6327
+    },
+    {
+      "epoch": 0.5858242203321182,
+      "grad_norm": 0.169921875,
+      "learning_rate": 0.02,
+      "loss": 1.5681,
+      "step": 6328
+    },
+    {
+      "epoch": 0.5859167968523983,
+      "grad_norm": 0.1552734375,
+      "learning_rate": 0.02,
+      "loss": 1.5855,
+      "step": 6329
+    },
+    {
+      "epoch": 0.5860093733726783,
+      "grad_norm": 0.15625,
+      "learning_rate": 0.02,
+      "loss": 1.5622,
+      "step": 6330
+    },
+    {
+      "epoch": 0.5861019498929584,
+      "grad_norm": 0.16015625,
+      "learning_rate": 0.02,
+      "loss": 1.5444,
+      "step": 6331
+    },
+    {
+      "epoch": 0.5861945264132384,
+      "grad_norm": 0.146484375,
+      "learning_rate": 0.02,
+      "loss": 1.5229,
+      "step": 6332
+    },
+    {
+      "epoch": 0.5862871029335185,
+      "grad_norm": 0.1611328125,
+      "learning_rate": 0.02,
+      "loss": 1.5173,
+      "step": 6333
+    },
+    {
+      "epoch": 0.5863796794537985,
+      "grad_norm": 0.1455078125,
+      "learning_rate": 0.02,
+      "loss": 1.5245,
+      "step": 6334
+    },
+    {
+      "epoch": 0.5864722559740786,
+      "grad_norm": 0.1533203125,
+      "learning_rate": 0.02,
+      "loss": 1.4832,
+      "step": 6335
+    },
+    {
+      "epoch": 0.5865648324943586,
+      "grad_norm": 0.1416015625,
+      "learning_rate": 0.02,
+      "loss": 1.543,
+      "step": 6336
+    },
+    {
+      "epoch": 0.5866574090146387,
+      "grad_norm": 0.15625,
+      "learning_rate": 0.02,
+      "loss": 1.5048,
+      "step": 6337
+    },
+    {
+      "epoch": 0.5867499855349187,
+      "grad_norm": 0.1376953125,
+      "learning_rate": 0.02,
+      "loss": 1.5277,
+      "step": 6338
+    },
+    {
+      "epoch": 0.5868425620551988,
+      "grad_norm": 0.1513671875,
+      "learning_rate": 0.02,
+      "loss": 1.5372,
+      "step": 6339
+    },
+    {
+      "epoch": 0.5869351385754787,
+      "grad_norm": 0.1513671875,
+      "learning_rate": 0.02,
+      "loss": 1.5939,
+      "step": 6340
+    },
+    {
+      "epoch": 0.5870277150957588,
+      "grad_norm": 0.154296875,
+      "learning_rate": 0.02,
+      "loss": 1.5363,
+      "step": 6341
+    },
+    {
+      "epoch": 0.5871202916160388,
+      "grad_norm": 0.150390625,
+      "learning_rate": 0.02,
+      "loss": 1.5968,
+      "step": 6342
+    },
+    {
+      "epoch": 0.5872128681363189,
+      "grad_norm": 0.1572265625,
+      "learning_rate": 0.02,
+      "loss": 1.5293,
+      "step": 6343
+    },
+    {
+      "epoch": 0.587305444656599,
+      "grad_norm": 0.146484375,
+      "learning_rate": 0.02,
+      "loss": 1.5396,
+      "step": 6344
+    },
+    {
+      "epoch": 0.587398021176879,
+      "grad_norm": 0.16015625,
+      "learning_rate": 0.02,
+      "loss": 1.5327,
+      "step": 6345
+    },
+    {
+      "epoch": 0.5874905976971591,
+      "grad_norm": 0.150390625,
+      "learning_rate": 0.02,
+      "loss": 1.5035,
+      "step": 6346
+    },
+    {
+      "epoch": 0.5875831742174391,
+      "grad_norm": 0.1533203125,
+      "learning_rate": 0.02,
+      "loss": 1.5403,
+      "step": 6347
+    },
+    {
+      "epoch": 0.5876757507377192,
+      "grad_norm": 0.150390625,
+      "learning_rate": 0.02,
+      "loss": 1.5447,
+      "step": 6348
+    },
+    {
+      "epoch": 0.5877683272579992,
+      "grad_norm": 0.150390625,
+      "learning_rate": 0.02,
+      "loss": 1.5663,
+      "step": 6349
+    },
+    {
+      "epoch": 0.5878609037782793,
+      "grad_norm": 0.150390625,
+      "learning_rate": 0.02,
+      "loss": 1.5374,
+      "step": 6350
+    },
+    {
+      "epoch": 0.5879534802985593,
+      "grad_norm": 0.1552734375,
+      "learning_rate": 0.02,
+      "loss": 1.587,
+      "step": 6351
+    },
+    {
+      "epoch": 0.5880460568188394,
+      "grad_norm": 0.150390625,
+      "learning_rate": 0.02,
+      "loss": 1.5679,
+      "step": 6352
+    },
+    {
+      "epoch": 0.5881386333391194,
+      "grad_norm": 0.140625,
+      "learning_rate": 0.02,
+      "loss": 1.5836,
+      "step": 6353
+    },
+    {
+      "epoch": 0.5882312098593994,
+      "grad_norm": 0.146484375,
+      "learning_rate": 0.02,
+      "loss": 1.6134,
+      "step": 6354
+    },
+    {
+      "epoch": 0.5883237863796794,
+      "grad_norm": 0.1435546875,
+      "learning_rate": 0.02,
+      "loss": 1.4913,
+      "step": 6355
+    },
+    {
+      "epoch": 0.5884163628999595,
+      "grad_norm": 0.150390625,
+      "learning_rate": 0.02,
+      "loss": 1.5116,
+      "step": 6356
+    },
+    {
+      "epoch": 0.5885089394202395,
+      "grad_norm": 0.14453125,
+      "learning_rate": 0.02,
+      "loss": 1.4725,
+      "step": 6357
+    },
+    {
+      "epoch": 0.5886015159405196,
+      "grad_norm": 0.142578125,
+      "learning_rate": 0.02,
+      "loss": 1.5865,
+      "step": 6358
+    },
+    {
+      "epoch": 0.5886940924607996,
+      "grad_norm": 0.1484375,
+      "learning_rate": 0.02,
+      "loss": 1.4726,
+      "step": 6359
+    },
+    {
+      "epoch": 0.5887866689810797,
+      "grad_norm": 0.1552734375,
+      "learning_rate": 0.02,
+      "loss": 1.6126,
+      "step": 6360
+    },
+    {
+      "epoch": 0.5888792455013597,
+      "grad_norm": 0.142578125,
+      "learning_rate": 0.02,
+      "loss": 1.5546,
+      "step": 6361
+    },
+    {
+      "epoch": 0.5889718220216398,
+      "grad_norm": 0.1455078125,
+      "learning_rate": 0.02,
+      "loss": 1.5451,
+      "step": 6362
+    },
+    {
+      "epoch": 0.5890643985419198,
+      "grad_norm": 0.1474609375,
+      "learning_rate": 0.02,
+      "loss": 1.5333,
+      "step": 6363
+    },
+    {
+      "epoch": 0.5891569750621999,
+      "grad_norm": 0.15234375,
+      "learning_rate": 0.02,
+      "loss": 1.5465,
+      "step": 6364
+    },
+    {
+      "epoch": 0.5892495515824799,
+      "grad_norm": 0.14453125,
+      "learning_rate": 0.02,
+      "loss": 1.5933,
+      "step": 6365
+    },
+    {
+      "epoch": 0.58934212810276,
+      "grad_norm": 0.1494140625,
+      "learning_rate": 0.02,
+      "loss": 1.4986,
+      "step": 6366
+    },
+    {
+      "epoch": 0.5894347046230399,
+      "grad_norm": 0.140625,
+      "learning_rate": 0.02,
+      "loss": 1.5465,
+      "step": 6367
+    },
+    {
+      "epoch": 0.58952728114332,
+      "grad_norm": 0.1455078125,
+      "learning_rate": 0.02,
+      "loss": 1.5526,
+      "step": 6368
+    },
+    {
+      "epoch": 0.5896198576636,
+      "grad_norm": 0.1533203125,
+      "learning_rate": 0.02,
+      "loss": 1.5368,
+      "step": 6369
+    },
+    {
+      "epoch": 0.5897124341838801,
+      "grad_norm": 0.1494140625,
+      "learning_rate": 0.02,
+      "loss": 1.5293,
+      "step": 6370
+    },
+    {
+      "epoch": 0.5898050107041601,
+      "grad_norm": 0.150390625,
+      "learning_rate": 0.02,
+      "loss": 1.5985,
+      "step": 6371
+    },
+    {
+      "epoch": 0.5898975872244402,
+      "grad_norm": 0.154296875,
+      "learning_rate": 0.02,
+      "loss": 1.5128,
+      "step": 6372
+    },
+    {
+      "epoch": 0.5899901637447202,
+      "grad_norm": 0.1484375,
+      "learning_rate": 0.02,
+      "loss": 1.5541,
+      "step": 6373
+    },
+    {
+      "epoch": 0.5900827402650003,
+      "grad_norm": 0.158203125,
+      "learning_rate": 0.02,
+      "loss": 1.5625,
+      "step": 6374
+    },
+    {
+      "epoch": 0.5901753167852803,
+      "grad_norm": 0.1455078125,
+      "learning_rate": 0.02,
+      "loss": 1.5565,
+      "step": 6375
+    },
+    {
+      "epoch": 0.5902678933055604,
+      "grad_norm": 0.15234375,
+      "learning_rate": 0.02,
+      "loss": 1.561,
+      "step": 6376
+    },
+    {
+      "epoch": 0.5903604698258405,
+      "grad_norm": 0.14453125,
+      "learning_rate": 0.02,
+      "loss": 1.5212,
+      "step": 6377
+    },
+    {
+      "epoch": 0.5904530463461205,
+      "grad_norm": 0.140625,
+      "learning_rate": 0.02,
+      "loss": 1.5104,
+      "step": 6378
+    },
+    {
+      "epoch": 0.5905456228664006,
+      "grad_norm": 0.1484375,
+      "learning_rate": 0.02,
+      "loss": 1.5835,
+      "step": 6379
+    },
+    {
+      "epoch": 0.5906381993866806,
+      "grad_norm": 0.1484375,
+      "learning_rate": 0.02,
+      "loss": 1.5764,
+      "step": 6380
+    },
+    {
+      "epoch": 0.5907307759069605,
+      "grad_norm": 0.13671875,
+      "learning_rate": 0.02,
+      "loss": 1.5509,
+      "step": 6381
+    },
+    {
+      "epoch": 0.5908233524272406,
+      "grad_norm": 0.150390625,
+      "learning_rate": 0.02,
+      "loss": 1.5665,
+      "step": 6382
+    },
+    {
+      "epoch": 0.5909159289475207,
+      "grad_norm": 0.146484375,
+      "learning_rate": 0.02,
+      "loss": 1.4999,
+      "step": 6383
+    },
+    {
+      "epoch": 0.5910085054678007,
+      "grad_norm": 0.1455078125,
+      "learning_rate": 0.02,
+      "loss": 1.5412,
+      "step": 6384
+    },
+    {
+      "epoch": 0.5911010819880808,
+      "grad_norm": 0.1484375,
+      "learning_rate": 0.02,
+      "loss": 1.5162,
+      "step": 6385
+    },
+    {
+      "epoch": 0.5911936585083608,
+      "grad_norm": 0.15234375,
+      "learning_rate": 0.02,
+      "loss": 1.4715,
+      "step": 6386
+    },
+    {
+      "epoch": 0.5912862350286409,
+      "grad_norm": 0.15625,
+      "learning_rate": 0.02,
+      "loss": 1.5593,
+      "step": 6387
+    },
+    {
+      "epoch": 0.5913788115489209,
+      "grad_norm": 0.15625,
+      "learning_rate": 0.02,
+      "loss": 1.5979,
+      "step": 6388
+    },
+    {
+      "epoch": 0.591471388069201,
+      "grad_norm": 0.1513671875,
+      "learning_rate": 0.02,
+      "loss": 1.5641,
+      "step": 6389
+    },
+    {
+      "epoch": 0.591563964589481,
+      "grad_norm": 0.15625,
+      "learning_rate": 0.02,
+      "loss": 1.5768,
+      "step": 6390
+    },
+    {
+      "epoch": 0.5916565411097611,
+      "grad_norm": 0.15234375,
+      "learning_rate": 0.02,
+      "loss": 1.5787,
+      "step": 6391
+    },
+    {
+      "epoch": 0.5917491176300411,
+      "grad_norm": 0.1513671875,
+      "learning_rate": 0.02,
+      "loss": 1.543,
+      "step": 6392
+    },
+    {
+      "epoch": 0.5918416941503212,
+      "grad_norm": 0.13671875,
+      "learning_rate": 0.02,
+      "loss": 1.5481,
+      "step": 6393
+    },
+    {
+      "epoch": 0.5919342706706011,
+      "grad_norm": 0.1455078125,
+      "learning_rate": 0.02,
+      "loss": 1.5603,
+      "step": 6394
+    },
+    {
+      "epoch": 0.5920268471908812,
+      "grad_norm": 0.1328125,
+      "learning_rate": 0.02,
+      "loss": 1.5628,
+      "step": 6395
+    },
+    {
+      "epoch": 0.5921194237111612,
+      "grad_norm": 0.1416015625,
+      "learning_rate": 0.02,
+      "loss": 1.5432,
+      "step": 6396
+    },
+    {
+      "epoch": 0.5922120002314413,
+      "grad_norm": 0.1455078125,
+      "learning_rate": 0.02,
+      "loss": 1.5468,
+      "step": 6397
+    },
+    {
+      "epoch": 0.5923045767517213,
+      "grad_norm": 0.14453125,
+      "learning_rate": 0.02,
+      "loss": 1.5261,
+      "step": 6398
+    },
+    {
+      "epoch": 0.5923971532720014,
+      "grad_norm": 0.146484375,
+      "learning_rate": 0.02,
+      "loss": 1.5705,
+      "step": 6399
+    },
+    {
+      "epoch": 0.5924897297922814,
+      "grad_norm": 0.1455078125,
+      "learning_rate": 0.02,
+      "loss": 1.5513,
+      "step": 6400
+    },
+    {
+      "epoch": 0.5925823063125615,
+      "grad_norm": 0.1494140625,
+      "learning_rate": 0.02,
+      "loss": 1.5644,
+      "step": 6401
+    },
+    {
+      "epoch": 0.5926748828328415,
+      "grad_norm": 0.1533203125,
+      "learning_rate": 0.02,
+      "loss": 1.5196,
+      "step": 6402
+    },
+    {
+      "epoch": 0.5927674593531216,
+      "grad_norm": 0.154296875,
+      "learning_rate": 0.02,
+      "loss": 1.5728,
+      "step": 6403
+    },
+    {
+      "epoch": 0.5928600358734016,
+      "grad_norm": 0.1416015625,
+      "learning_rate": 0.02,
+      "loss": 1.5213,
+      "step": 6404
+    },
+    {
+      "epoch": 0.5929526123936817,
+      "grad_norm": 0.146484375,
+      "learning_rate": 0.02,
+      "loss": 1.5383,
+      "step": 6405
+    },
+    {
+      "epoch": 0.5930451889139617,
+      "grad_norm": 0.1376953125,
+      "learning_rate": 0.02,
+      "loss": 1.5566,
+      "step": 6406
+    },
+    {
+      "epoch": 0.5931377654342418,
+      "grad_norm": 0.140625,
+      "learning_rate": 0.02,
+      "loss": 1.5459,
+      "step": 6407
+    },
+    {
+      "epoch": 0.5932303419545217,
+      "grad_norm": 0.1552734375,
+      "learning_rate": 0.02,
+      "loss": 1.4937,
+      "step": 6408
+    },
+    {
+      "epoch": 0.5933229184748018,
+      "grad_norm": 0.1591796875,
+      "learning_rate": 0.02,
+      "loss": 1.6202,
+      "step": 6409
+    },
+    {
+      "epoch": 0.5934154949950818,
+      "grad_norm": 0.1484375,
+      "learning_rate": 0.02,
+      "loss": 1.5716,
+      "step": 6410
+    },
+    {
+      "epoch": 0.5935080715153619,
+      "grad_norm": 0.1533203125,
+      "learning_rate": 0.02,
+      "loss": 1.5845,
+      "step": 6411
+    },
+    {
+      "epoch": 0.5936006480356419,
+      "grad_norm": 0.14453125,
+      "learning_rate": 0.02,
+      "loss": 1.5689,
+      "step": 6412
+    },
+    {
+      "epoch": 0.593693224555922,
+      "grad_norm": 0.1611328125,
+      "learning_rate": 0.02,
+      "loss": 1.5358,
+      "step": 6413
+    },
+    {
+      "epoch": 0.593785801076202,
+      "grad_norm": 0.1357421875,
+      "learning_rate": 0.02,
+      "loss": 1.6157,
+      "step": 6414
+    },
+    {
+      "epoch": 0.5938783775964821,
+      "grad_norm": 0.14453125,
+      "learning_rate": 0.02,
+      "loss": 1.5641,
+      "step": 6415
+    },
+    {
+      "epoch": 0.5939709541167622,
+      "grad_norm": 0.14453125,
+      "learning_rate": 0.02,
+      "loss": 1.6275,
+      "step": 6416
+    },
+    {
+      "epoch": 0.5940635306370422,
+      "grad_norm": 0.16015625,
+      "learning_rate": 0.02,
+      "loss": 1.5707,
+      "step": 6417
+    },
+    {
+      "epoch": 0.5941561071573223,
+      "grad_norm": 0.150390625,
+      "learning_rate": 0.02,
+      "loss": 1.5931,
+      "step": 6418
+    },
+    {
+      "epoch": 0.5942486836776023,
+      "grad_norm": 0.1533203125,
+      "learning_rate": 0.02,
+      "loss": 1.5819,
+      "step": 6419
+    },
+    {
+      "epoch": 0.5943412601978824,
+      "grad_norm": 0.1611328125,
+      "learning_rate": 0.02,
+      "loss": 1.5657,
+      "step": 6420
+    },
+    {
+      "epoch": 0.5944338367181623,
+      "grad_norm": 0.1513671875,
+      "learning_rate": 0.02,
+      "loss": 1.5176,
+      "step": 6421
+    },
+    {
+      "epoch": 0.5945264132384424,
+      "grad_norm": 0.1484375,
+      "learning_rate": 0.02,
+      "loss": 1.5463,
+      "step": 6422
+    },
+    {
+      "epoch": 0.5946189897587224,
+      "grad_norm": 0.1494140625,
+      "learning_rate": 0.02,
+      "loss": 1.5355,
+      "step": 6423
+    },
+    {
+      "epoch": 0.5947115662790025,
+      "grad_norm": 0.1494140625,
+      "learning_rate": 0.02,
+      "loss": 1.6074,
+      "step": 6424
+    },
+    {
+      "epoch": 0.5948041427992825,
+      "grad_norm": 0.1474609375,
+      "learning_rate": 0.02,
+      "loss": 1.5958,
+      "step": 6425
+    },
+    {
+      "epoch": 0.5948967193195626,
+      "grad_norm": 0.1494140625,
+      "learning_rate": 0.02,
+      "loss": 1.5317,
+      "step": 6426
+    },
+    {
+      "epoch": 0.5949892958398426,
+      "grad_norm": 0.1630859375,
+      "learning_rate": 0.02,
+      "loss": 1.5086,
+      "step": 6427
+    },
+    {
+      "epoch": 0.5950818723601227,
+      "grad_norm": 0.158203125,
+      "learning_rate": 0.02,
+      "loss": 1.5121,
+      "step": 6428
+    },
+    {
+      "epoch": 0.5951744488804027,
+      "grad_norm": 0.15234375,
+      "learning_rate": 0.02,
+      "loss": 1.5949,
+      "step": 6429
+    },
+    {
+      "epoch": 0.5952670254006828,
+      "grad_norm": 0.15234375,
+      "learning_rate": 0.02,
+      "loss": 1.5292,
+      "step": 6430
+    },
+    {
+      "epoch": 0.5953596019209628,
+      "grad_norm": 0.158203125,
+      "learning_rate": 0.02,
+      "loss": 1.6059,
+      "step": 6431
+    },
+    {
+      "epoch": 0.5954521784412429,
+      "grad_norm": 0.1611328125,
+      "learning_rate": 0.02,
+      "loss": 1.5515,
+      "step": 6432
+    },
+    {
+      "epoch": 0.5955447549615229,
+      "grad_norm": 0.1455078125,
+      "learning_rate": 0.02,
+      "loss": 1.502,
+      "step": 6433
+    },
+    {
+      "epoch": 0.595637331481803,
+      "grad_norm": 0.15625,
+      "learning_rate": 0.02,
+      "loss": 1.5245,
+      "step": 6434
+    },
+    {
+      "epoch": 0.5957299080020829,
+      "grad_norm": 0.1494140625,
+      "learning_rate": 0.02,
+      "loss": 1.6093,
+      "step": 6435
+    },
+    {
+      "epoch": 0.595822484522363,
+      "grad_norm": 0.150390625,
+      "learning_rate": 0.02,
+      "loss": 1.5204,
+      "step": 6436
+    },
+    {
+      "epoch": 0.595915061042643,
+      "grad_norm": 0.1376953125,
+      "learning_rate": 0.02,
+      "loss": 1.5549,
+      "step": 6437
+    },
+    {
+      "epoch": 0.5960076375629231,
+      "grad_norm": 0.154296875,
+      "learning_rate": 0.02,
+      "loss": 1.59,
+      "step": 6438
+    },
+    {
+      "epoch": 0.5961002140832031,
+      "grad_norm": 0.1484375,
+      "learning_rate": 0.02,
+      "loss": 1.5688,
+      "step": 6439
+    },
+    {
+      "epoch": 0.5961927906034832,
+      "grad_norm": 0.1640625,
+      "learning_rate": 0.02,
+      "loss": 1.5686,
+      "step": 6440
+    },
+    {
+      "epoch": 0.5962853671237632,
+      "grad_norm": 0.166015625,
+      "learning_rate": 0.02,
+      "loss": 1.5391,
+      "step": 6441
+    },
+    {
+      "epoch": 0.5963779436440433,
+      "grad_norm": 0.1630859375,
+      "learning_rate": 0.02,
+      "loss": 1.5544,
+      "step": 6442
+    },
+    {
+      "epoch": 0.5964705201643233,
+      "grad_norm": 0.1611328125,
+      "learning_rate": 0.02,
+      "loss": 1.5486,
+      "step": 6443
+    },
+    {
+      "epoch": 0.5965630966846034,
+      "grad_norm": 0.150390625,
+      "learning_rate": 0.02,
+      "loss": 1.5564,
+      "step": 6444
+    },
+    {
+      "epoch": 0.5966556732048834,
+      "grad_norm": 0.1455078125,
+      "learning_rate": 0.02,
+      "loss": 1.5502,
+      "step": 6445
+    },
+    {
+      "epoch": 0.5967482497251635,
+      "grad_norm": 0.142578125,
+      "learning_rate": 0.02,
+      "loss": 1.5162,
+      "step": 6446
+    },
+    {
+      "epoch": 0.5968408262454435,
+      "grad_norm": 0.1484375,
+      "learning_rate": 0.02,
+      "loss": 1.5785,
+      "step": 6447
+    },
+    {
+      "epoch": 0.5969334027657236,
+      "grad_norm": 0.1455078125,
+      "learning_rate": 0.02,
+      "loss": 1.5395,
+      "step": 6448
+    },
+    {
+      "epoch": 0.5970259792860035,
+      "grad_norm": 0.1494140625,
+      "learning_rate": 0.02,
+      "loss": 1.5465,
+      "step": 6449
+    },
+    {
+      "epoch": 0.5971185558062836,
+      "grad_norm": 0.1513671875,
+      "learning_rate": 0.02,
+      "loss": 1.5495,
+      "step": 6450
+    },
+    {
+      "epoch": 0.5972111323265636,
+      "grad_norm": 0.1533203125,
+      "learning_rate": 0.02,
+      "loss": 1.5932,
+      "step": 6451
+    },
+    {
+      "epoch": 0.5973037088468437,
+      "grad_norm": 0.15234375,
+      "learning_rate": 0.02,
+      "loss": 1.5477,
+      "step": 6452
+    },
+    {
+      "epoch": 0.5973962853671237,
+      "grad_norm": 0.1552734375,
+      "learning_rate": 0.02,
+      "loss": 1.5092,
+      "step": 6453
+    },
+    {
+      "epoch": 0.5974888618874038,
+      "grad_norm": 0.1572265625,
+      "learning_rate": 0.02,
+      "loss": 1.5084,
+      "step": 6454
+    },
+    {
+      "epoch": 0.5975814384076839,
+      "grad_norm": 0.146484375,
+      "learning_rate": 0.02,
+      "loss": 1.5462,
+      "step": 6455
+    },
+    {
+      "epoch": 0.5976740149279639,
+      "grad_norm": 0.1494140625,
+      "learning_rate": 0.02,
+      "loss": 1.54,
+      "step": 6456
+    },
+    {
+      "epoch": 0.597766591448244,
+      "grad_norm": 0.1513671875,
+      "learning_rate": 0.02,
+      "loss": 1.5101,
+      "step": 6457
+    },
+    {
+      "epoch": 0.597859167968524,
+      "grad_norm": 0.150390625,
+      "learning_rate": 0.02,
+      "loss": 1.5885,
+      "step": 6458
+    },
+    {
+      "epoch": 0.5979517444888041,
+      "grad_norm": 0.1494140625,
+      "learning_rate": 0.02,
+      "loss": 1.5022,
+      "step": 6459
+    },
+    {
+      "epoch": 0.5980443210090841,
+      "grad_norm": 0.1357421875,
+      "learning_rate": 0.02,
+      "loss": 1.6019,
+      "step": 6460
+    },
+    {
+      "epoch": 0.5981368975293642,
+      "grad_norm": 0.1484375,
+      "learning_rate": 0.02,
+      "loss": 1.6054,
+      "step": 6461
+    },
+    {
+      "epoch": 0.5982294740496441,
+      "grad_norm": 0.1455078125,
+      "learning_rate": 0.02,
+      "loss": 1.5466,
+      "step": 6462
+    },
+    {
+      "epoch": 0.5983220505699242,
+      "grad_norm": 0.142578125,
+      "learning_rate": 0.02,
+      "loss": 1.4979,
+      "step": 6463
+    },
+    {
+      "epoch": 0.5984146270902042,
+      "grad_norm": 0.1513671875,
+      "learning_rate": 0.02,
+      "loss": 1.4881,
+      "step": 6464
+    },
+    {
+      "epoch": 0.5985072036104843,
+      "grad_norm": 0.15625,
+      "learning_rate": 0.02,
+      "loss": 1.5608,
+      "step": 6465
+    },
+    {
+      "epoch": 0.5985997801307643,
+      "grad_norm": 0.14453125,
+      "learning_rate": 0.02,
+      "loss": 1.5846,
+      "step": 6466
+    },
+    {
+      "epoch": 0.5986923566510444,
+      "grad_norm": 0.16015625,
+      "learning_rate": 0.02,
+      "loss": 1.535,
+      "step": 6467
+    },
+    {
+      "epoch": 0.5987849331713244,
+      "grad_norm": 0.14453125,
+      "learning_rate": 0.02,
+      "loss": 1.4723,
+      "step": 6468
+    },
+    {
+      "epoch": 0.5988775096916045,
+      "grad_norm": 0.1552734375,
+      "learning_rate": 0.02,
+      "loss": 1.6076,
+      "step": 6469
+    },
+    {
+      "epoch": 0.5989700862118845,
+      "grad_norm": 0.1474609375,
+      "learning_rate": 0.02,
+      "loss": 1.5674,
+      "step": 6470
+    },
+    {
+      "epoch": 0.5990626627321646,
+      "grad_norm": 0.154296875,
+      "learning_rate": 0.02,
+      "loss": 1.535,
+      "step": 6471
+    },
+    {
+      "epoch": 0.5991552392524446,
+      "grad_norm": 0.150390625,
+      "learning_rate": 0.02,
+      "loss": 1.6081,
+      "step": 6472
+    },
+    {
+      "epoch": 0.5992478157727247,
+      "grad_norm": 0.1494140625,
+      "learning_rate": 0.02,
+      "loss": 1.4759,
+      "step": 6473
+    },
+    {
+      "epoch": 0.5993403922930047,
+      "grad_norm": 0.150390625,
+      "learning_rate": 0.02,
+      "loss": 1.5287,
+      "step": 6474
+    },
+    {
+      "epoch": 0.5994329688132848,
+      "grad_norm": 0.14453125,
+      "learning_rate": 0.02,
+      "loss": 1.471,
+      "step": 6475
+    },
+    {
+      "epoch": 0.5995255453335647,
+      "grad_norm": 0.1494140625,
+      "learning_rate": 0.02,
+      "loss": 1.5687,
+      "step": 6476
+    },
+    {
+      "epoch": 0.5996181218538448,
+      "grad_norm": 0.142578125,
+      "learning_rate": 0.02,
+      "loss": 1.4771,
+      "step": 6477
+    },
+    {
+      "epoch": 0.5997106983741248,
+      "grad_norm": 0.1474609375,
+      "learning_rate": 0.02,
+      "loss": 1.5557,
+      "step": 6478
+    },
+    {
+      "epoch": 0.5998032748944049,
+      "grad_norm": 0.15234375,
+      "learning_rate": 0.02,
+      "loss": 1.5703,
+      "step": 6479
+    },
+    {
+      "epoch": 0.5998958514146849,
+      "grad_norm": 0.142578125,
+      "learning_rate": 0.02,
+      "loss": 1.5532,
+      "step": 6480
+    },
+    {
+      "epoch": 0.599988427934965,
+      "grad_norm": 0.1533203125,
+      "learning_rate": 0.02,
+      "loss": 1.552,
+      "step": 6481
+    },
+    {
+      "epoch": 0.600081004455245,
+      "grad_norm": 0.1484375,
+      "learning_rate": 0.02,
+      "loss": 1.5465,
+      "step": 6482
+    },
+    {
+      "epoch": 0.6001735809755251,
+      "grad_norm": 0.150390625,
+      "learning_rate": 0.02,
+      "loss": 1.5658,
+      "step": 6483
+    },
+    {
+      "epoch": 0.6002661574958051,
+      "grad_norm": 0.15625,
+      "learning_rate": 0.02,
+      "loss": 1.5777,
+      "step": 6484
+    },
+    {
+      "epoch": 0.6003587340160852,
+      "grad_norm": 0.1669921875,
+      "learning_rate": 0.02,
+      "loss": 1.5446,
+      "step": 6485
+    },
+    {
+      "epoch": 0.6004513105363652,
+      "grad_norm": 0.15625,
+      "learning_rate": 0.02,
+      "loss": 1.5848,
+      "step": 6486
+    },
+    {
+      "epoch": 0.6005438870566453,
+      "grad_norm": 0.154296875,
+      "learning_rate": 0.02,
+      "loss": 1.5498,
+      "step": 6487
+    },
+    {
+      "epoch": 0.6006364635769254,
+      "grad_norm": 0.1494140625,
+      "learning_rate": 0.02,
+      "loss": 1.5528,
+      "step": 6488
+    },
+    {
+      "epoch": 0.6007290400972053,
+      "grad_norm": 0.1552734375,
+      "learning_rate": 0.02,
+      "loss": 1.5257,
+      "step": 6489
+    },
+    {
+      "epoch": 0.6008216166174853,
+      "grad_norm": 0.14453125,
+      "learning_rate": 0.02,
+      "loss": 1.5368,
+      "step": 6490
+    },
+    {
+      "epoch": 0.6009141931377654,
+      "grad_norm": 0.150390625,
+      "learning_rate": 0.02,
+      "loss": 1.4677,
+      "step": 6491
+    },
+    {
+      "epoch": 0.6010067696580454,
+      "grad_norm": 0.14453125,
+      "learning_rate": 0.02,
+      "loss": 1.5882,
+      "step": 6492
+    },
+    {
+      "epoch": 0.6010993461783255,
+      "grad_norm": 0.15625,
+      "learning_rate": 0.02,
+      "loss": 1.5558,
+      "step": 6493
+    },
+    {
+      "epoch": 0.6011919226986056,
+      "grad_norm": 0.158203125,
+      "learning_rate": 0.02,
+      "loss": 1.4939,
+      "step": 6494
+    },
+    {
+      "epoch": 0.6012844992188856,
+      "grad_norm": 0.1572265625,
+      "learning_rate": 0.02,
+      "loss": 1.5687,
+      "step": 6495
+    },
+    {
+      "epoch": 0.6013770757391657,
+      "grad_norm": 0.158203125,
+      "learning_rate": 0.02,
+      "loss": 1.5171,
+      "step": 6496
+    },
+    {
+      "epoch": 0.6014696522594457,
+      "grad_norm": 0.150390625,
+      "learning_rate": 0.02,
+      "loss": 1.4823,
+      "step": 6497
+    },
+    {
+      "epoch": 0.6015622287797258,
+      "grad_norm": 0.16015625,
+      "learning_rate": 0.02,
+      "loss": 1.6103,
+      "step": 6498
+    },
+    {
+      "epoch": 0.6016548053000058,
+      "grad_norm": 0.154296875,
+      "learning_rate": 0.02,
+      "loss": 1.5202,
+      "step": 6499
+    },
+    {
+      "epoch": 0.6017473818202859,
+      "grad_norm": 0.162109375,
+      "learning_rate": 0.02,
+      "loss": 1.519,
+      "step": 6500
+    },
+    {
+      "epoch": 0.6018399583405659,
+      "grad_norm": 0.140625,
+      "learning_rate": 0.02,
+      "loss": 1.5342,
+      "step": 6501
+    },
+    {
+      "epoch": 0.601932534860846,
+      "grad_norm": 0.1552734375,
+      "learning_rate": 0.02,
+      "loss": 1.515,
+      "step": 6502
+    },
+    {
+      "epoch": 0.6020251113811259,
+      "grad_norm": 0.1455078125,
+      "learning_rate": 0.02,
+      "loss": 1.5204,
+      "step": 6503
+    },
+    {
+      "epoch": 0.602117687901406,
+      "grad_norm": 0.1494140625,
+      "learning_rate": 0.02,
+      "loss": 1.5829,
+      "step": 6504
+    },
+    {
+      "epoch": 0.602210264421686,
+      "grad_norm": 0.1435546875,
+      "learning_rate": 0.02,
+      "loss": 1.582,
+      "step": 6505
+    },
+    {
+      "epoch": 0.6023028409419661,
+      "grad_norm": 0.154296875,
+      "learning_rate": 0.02,
+      "loss": 1.5545,
+      "step": 6506
+    },
+    {
+      "epoch": 0.6023954174622461,
+      "grad_norm": 0.1484375,
+      "learning_rate": 0.02,
+      "loss": 1.6172,
+      "step": 6507
+    },
+    {
+      "epoch": 0.6024879939825262,
+      "grad_norm": 0.150390625,
+      "learning_rate": 0.02,
+      "loss": 1.5134,
+      "step": 6508
+    },
+    {
+      "epoch": 0.6025805705028062,
+      "grad_norm": 0.1455078125,
+      "learning_rate": 0.02,
+      "loss": 1.5258,
+      "step": 6509
+    },
+    {
+      "epoch": 0.6026731470230863,
+      "grad_norm": 0.1455078125,
+      "learning_rate": 0.02,
+      "loss": 1.5799,
+      "step": 6510
+    },
+    {
+      "epoch": 0.6027657235433663,
+      "grad_norm": 0.14453125,
+      "learning_rate": 0.02,
+      "loss": 1.5681,
+      "step": 6511
+    },
+    {
+      "epoch": 0.6028583000636464,
+      "grad_norm": 0.13671875,
+      "learning_rate": 0.02,
+      "loss": 1.5795,
+      "step": 6512
+    },
+    {
+      "epoch": 0.6029508765839264,
+      "grad_norm": 0.1455078125,
+      "learning_rate": 0.02,
+      "loss": 1.5378,
+      "step": 6513
+    },
+    {
+      "epoch": 0.6030434531042065,
+      "grad_norm": 0.1357421875,
+      "learning_rate": 0.02,
+      "loss": 1.5804,
+      "step": 6514
+    },
+    {
+      "epoch": 0.6031360296244865,
+      "grad_norm": 0.140625,
+      "learning_rate": 0.02,
+      "loss": 1.4941,
+      "step": 6515
+    },
+    {
+      "epoch": 0.6032286061447666,
+      "grad_norm": 0.146484375,
+      "learning_rate": 0.02,
+      "loss": 1.5601,
+      "step": 6516
+    },
+    {
+      "epoch": 0.6033211826650465,
+      "grad_norm": 0.15234375,
+      "learning_rate": 0.02,
+      "loss": 1.5465,
+      "step": 6517
+    },
+    {
+      "epoch": 0.6034137591853266,
+      "grad_norm": 0.1416015625,
+      "learning_rate": 0.02,
+      "loss": 1.5613,
+      "step": 6518
+    },
+    {
+      "epoch": 0.6035063357056066,
+      "grad_norm": 0.1494140625,
+      "learning_rate": 0.02,
+      "loss": 1.5093,
+      "step": 6519
+    },
+    {
+      "epoch": 0.6035989122258867,
+      "grad_norm": 0.1513671875,
+      "learning_rate": 0.02,
+      "loss": 1.5122,
+      "step": 6520
+    },
+    {
+      "epoch": 0.6036914887461667,
+      "grad_norm": 0.15625,
+      "learning_rate": 0.02,
+      "loss": 1.6274,
+      "step": 6521
+    },
+    {
+      "epoch": 0.6037840652664468,
+      "grad_norm": 0.1533203125,
+      "learning_rate": 0.02,
+      "loss": 1.555,
+      "step": 6522
+    },
+    {
+      "epoch": 0.6038766417867268,
+      "grad_norm": 0.1513671875,
+      "learning_rate": 0.02,
+      "loss": 1.5507,
+      "step": 6523
+    },
+    {
+      "epoch": 0.6039692183070069,
+      "grad_norm": 0.1494140625,
+      "learning_rate": 0.02,
+      "loss": 1.5681,
+      "step": 6524
+    },
+    {
+      "epoch": 0.604061794827287,
+      "grad_norm": 0.1552734375,
+      "learning_rate": 0.02,
+      "loss": 1.5563,
+      "step": 6525
+    },
+    {
+      "epoch": 0.604154371347567,
+      "grad_norm": 0.1689453125,
+      "learning_rate": 0.02,
+      "loss": 1.5846,
+      "step": 6526
+    },
+    {
+      "epoch": 0.604246947867847,
+      "grad_norm": 0.1552734375,
+      "learning_rate": 0.02,
+      "loss": 1.5819,
+      "step": 6527
+    },
+    {
+      "epoch": 0.6043395243881271,
+      "grad_norm": 0.1533203125,
+      "learning_rate": 0.02,
+      "loss": 1.5792,
+      "step": 6528
+    },
+    {
+      "epoch": 0.6044321009084072,
+      "grad_norm": 0.154296875,
+      "learning_rate": 0.02,
+      "loss": 1.5261,
+      "step": 6529
+    },
+    {
+      "epoch": 0.6045246774286871,
+      "grad_norm": 0.15625,
+      "learning_rate": 0.02,
+      "loss": 1.5948,
+      "step": 6530
+    },
+    {
+      "epoch": 0.6046172539489671,
+      "grad_norm": 0.1572265625,
+      "learning_rate": 0.02,
+      "loss": 1.5082,
+      "step": 6531
+    },
+    {
+      "epoch": 0.6047098304692472,
+      "grad_norm": 0.150390625,
+      "learning_rate": 0.02,
+      "loss": 1.5443,
+      "step": 6532
+    },
+    {
+      "epoch": 0.6048024069895273,
+      "grad_norm": 0.1416015625,
+      "learning_rate": 0.02,
+      "loss": 1.4872,
+      "step": 6533
+    },
+    {
+      "epoch": 0.6048949835098073,
+      "grad_norm": 0.146484375,
+      "learning_rate": 0.02,
+      "loss": 1.6424,
+      "step": 6534
+    },
+    {
+      "epoch": 0.6049875600300874,
+      "grad_norm": 0.150390625,
+      "learning_rate": 0.02,
+      "loss": 1.496,
+      "step": 6535
+    },
+    {
+      "epoch": 0.6050801365503674,
+      "grad_norm": 0.15625,
+      "learning_rate": 0.02,
+      "loss": 1.4944,
+      "step": 6536
+    },
+    {
+      "epoch": 0.6051727130706475,
+      "grad_norm": 0.146484375,
+      "learning_rate": 0.02,
+      "loss": 1.507,
+      "step": 6537
+    },
+    {
+      "epoch": 0.6052652895909275,
+      "grad_norm": 0.1484375,
+      "learning_rate": 0.02,
+      "loss": 1.5267,
+      "step": 6538
+    },
+    {
+      "epoch": 0.6053578661112076,
+      "grad_norm": 0.16015625,
+      "learning_rate": 0.02,
+      "loss": 1.5416,
+      "step": 6539
+    },
+    {
+      "epoch": 0.6054504426314876,
+      "grad_norm": 0.1533203125,
+      "learning_rate": 0.02,
+      "loss": 1.5959,
+      "step": 6540
+    },
+    {
+      "epoch": 0.6055430191517677,
+      "grad_norm": 0.154296875,
+      "learning_rate": 0.02,
+      "loss": 1.5292,
+      "step": 6541
+    },
+    {
+      "epoch": 0.6056355956720477,
+      "grad_norm": 0.1552734375,
+      "learning_rate": 0.02,
+      "loss": 1.5418,
+      "step": 6542
+    },
+    {
+      "epoch": 0.6057281721923278,
+      "grad_norm": 0.15625,
+      "learning_rate": 0.02,
+      "loss": 1.5231,
+      "step": 6543
+    },
+    {
+      "epoch": 0.6058207487126077,
+      "grad_norm": 0.158203125,
+      "learning_rate": 0.02,
+      "loss": 1.5853,
+      "step": 6544
+    },
+    {
+      "epoch": 0.6059133252328878,
+      "grad_norm": 0.1435546875,
+      "learning_rate": 0.02,
+      "loss": 1.5003,
+      "step": 6545
+    },
+    {
+      "epoch": 0.6060059017531678,
+      "grad_norm": 0.15234375,
+      "learning_rate": 0.02,
+      "loss": 1.5482,
+      "step": 6546
+    },
+    {
+      "epoch": 0.6060984782734479,
+      "grad_norm": 0.146484375,
+      "learning_rate": 0.02,
+      "loss": 1.6247,
+      "step": 6547
+    },
+    {
+      "epoch": 0.6061910547937279,
+      "grad_norm": 0.1533203125,
+      "learning_rate": 0.02,
+      "loss": 1.4844,
+      "step": 6548
+    },
+    {
+      "epoch": 0.606283631314008,
+      "grad_norm": 0.1455078125,
+      "learning_rate": 0.02,
+      "loss": 1.5107,
+      "step": 6549
+    },
+    {
+      "epoch": 0.606376207834288,
+      "grad_norm": 0.150390625,
+      "learning_rate": 0.02,
+      "loss": 1.5626,
+      "step": 6550
+    },
+    {
+      "epoch": 0.6064687843545681,
+      "grad_norm": 0.142578125,
+      "learning_rate": 0.02,
+      "loss": 1.5299,
+      "step": 6551
+    },
+    {
+      "epoch": 0.6065613608748481,
+      "grad_norm": 0.1572265625,
+      "learning_rate": 0.02,
+      "loss": 1.5385,
+      "step": 6552
+    },
+    {
+      "epoch": 0.6066539373951282,
+      "grad_norm": 0.1435546875,
+      "learning_rate": 0.02,
+      "loss": 1.4735,
+      "step": 6553
+    },
+    {
+      "epoch": 0.6067465139154082,
+      "grad_norm": 0.1435546875,
+      "learning_rate": 0.02,
+      "loss": 1.5624,
+      "step": 6554
+    },
+    {
+      "epoch": 0.6068390904356883,
+      "grad_norm": 0.154296875,
+      "learning_rate": 0.02,
+      "loss": 1.5498,
+      "step": 6555
+    },
+    {
+      "epoch": 0.6069316669559683,
+      "grad_norm": 0.1435546875,
+      "learning_rate": 0.02,
+      "loss": 1.4612,
+      "step": 6556
+    },
+    {
+      "epoch": 0.6070242434762483,
+      "grad_norm": 0.1455078125,
+      "learning_rate": 0.02,
+      "loss": 1.5409,
+      "step": 6557
+    },
+    {
+      "epoch": 0.6071168199965283,
+      "grad_norm": 0.14453125,
+      "learning_rate": 0.02,
+      "loss": 1.5862,
+      "step": 6558
+    },
+    {
+      "epoch": 0.6072093965168084,
+      "grad_norm": 0.146484375,
+      "learning_rate": 0.02,
+      "loss": 1.5758,
+      "step": 6559
+    },
+    {
+      "epoch": 0.6073019730370884,
+      "grad_norm": 0.1484375,
+      "learning_rate": 0.02,
+      "loss": 1.557,
+      "step": 6560
+    },
+    {
+      "epoch": 0.6073945495573685,
+      "grad_norm": 0.146484375,
+      "learning_rate": 0.02,
+      "loss": 1.5748,
+      "step": 6561
+    },
+    {
+      "epoch": 0.6074871260776485,
+      "grad_norm": 0.140625,
+      "learning_rate": 0.02,
+      "loss": 1.565,
+      "step": 6562
+    },
+    {
+      "epoch": 0.6075797025979286,
+      "grad_norm": 0.134765625,
+      "learning_rate": 0.02,
+      "loss": 1.5265,
+      "step": 6563
+    },
+    {
+      "epoch": 0.6076722791182086,
+      "grad_norm": 0.1328125,
+      "learning_rate": 0.02,
+      "loss": 1.5407,
+      "step": 6564
+    },
+    {
+      "epoch": 0.6077648556384887,
+      "grad_norm": 0.1396484375,
+      "learning_rate": 0.02,
+      "loss": 1.5846,
+      "step": 6565
+    },
+    {
+      "epoch": 0.6078574321587688,
+      "grad_norm": 0.1484375,
+      "learning_rate": 0.02,
+      "loss": 1.5542,
+      "step": 6566
+    },
+    {
+      "epoch": 0.6079500086790488,
+      "grad_norm": 0.13671875,
+      "learning_rate": 0.02,
+      "loss": 1.5203,
+      "step": 6567
+    },
+    {
+      "epoch": 0.6080425851993289,
+      "grad_norm": 0.1376953125,
+      "learning_rate": 0.02,
+      "loss": 1.5503,
+      "step": 6568
+    },
+    {
+      "epoch": 0.6081351617196089,
+      "grad_norm": 0.13671875,
+      "learning_rate": 0.02,
+      "loss": 1.5956,
+      "step": 6569
+    },
+    {
+      "epoch": 0.608227738239889,
+      "grad_norm": 0.138671875,
+      "learning_rate": 0.02,
+      "loss": 1.6191,
+      "step": 6570
+    },
+    {
+      "epoch": 0.6083203147601689,
+      "grad_norm": 0.1435546875,
+      "learning_rate": 0.02,
+      "loss": 1.5918,
+      "step": 6571
+    },
+    {
+      "epoch": 0.608412891280449,
+      "grad_norm": 0.14453125,
+      "learning_rate": 0.02,
+      "loss": 1.5376,
+      "step": 6572
+    },
+    {
+      "epoch": 0.608505467800729,
+      "grad_norm": 0.154296875,
+      "learning_rate": 0.02,
+      "loss": 1.5655,
+      "step": 6573
+    },
+    {
+      "epoch": 0.6085980443210091,
+      "grad_norm": 0.1455078125,
+      "learning_rate": 0.02,
+      "loss": 1.5514,
+      "step": 6574
+    },
+    {
+      "epoch": 0.6086906208412891,
+      "grad_norm": 0.1484375,
+      "learning_rate": 0.02,
+      "loss": 1.4902,
+      "step": 6575
+    },
+    {
+      "epoch": 0.6087831973615692,
+      "grad_norm": 0.142578125,
+      "learning_rate": 0.02,
+      "loss": 1.4581,
+      "step": 6576
+    },
+    {
+      "epoch": 0.6088757738818492,
+      "grad_norm": 0.15234375,
+      "learning_rate": 0.02,
+      "loss": 1.5505,
+      "step": 6577
+    },
+    {
+      "epoch": 0.6089683504021293,
+      "grad_norm": 0.154296875,
+      "learning_rate": 0.02,
+      "loss": 1.4879,
+      "step": 6578
+    },
+    {
+      "epoch": 0.6090609269224093,
+      "grad_norm": 0.1474609375,
+      "learning_rate": 0.02,
+      "loss": 1.5769,
+      "step": 6579
+    },
+    {
+      "epoch": 0.6091535034426894,
+      "grad_norm": 0.1494140625,
+      "learning_rate": 0.02,
+      "loss": 1.5708,
+      "step": 6580
+    },
+    {
+      "epoch": 0.6092460799629694,
+      "grad_norm": 0.142578125,
+      "learning_rate": 0.02,
+      "loss": 1.4736,
+      "step": 6581
+    },
+    {
+      "epoch": 0.6093386564832495,
+      "grad_norm": 0.150390625,
+      "learning_rate": 0.02,
+      "loss": 1.5186,
+      "step": 6582
+    },
+    {
+      "epoch": 0.6094312330035295,
+      "grad_norm": 0.150390625,
+      "learning_rate": 0.02,
+      "loss": 1.5759,
+      "step": 6583
+    },
+    {
+      "epoch": 0.6095238095238096,
+      "grad_norm": 0.1533203125,
+      "learning_rate": 0.02,
+      "loss": 1.5327,
+      "step": 6584
+    },
+    {
+      "epoch": 0.6096163860440895,
+      "grad_norm": 0.154296875,
+      "learning_rate": 0.02,
+      "loss": 1.5683,
+      "step": 6585
+    },
+    {
+      "epoch": 0.6097089625643696,
+      "grad_norm": 0.1533203125,
+      "learning_rate": 0.02,
+      "loss": 1.5488,
+      "step": 6586
+    },
+    {
+      "epoch": 0.6098015390846496,
+      "grad_norm": 0.14453125,
+      "learning_rate": 0.02,
+      "loss": 1.5267,
+      "step": 6587
+    },
+    {
+      "epoch": 0.6098941156049297,
+      "grad_norm": 0.1494140625,
+      "learning_rate": 0.02,
+      "loss": 1.5775,
+      "step": 6588
+    },
+    {
+      "epoch": 0.6099866921252097,
+      "grad_norm": 0.15234375,
+      "learning_rate": 0.02,
+      "loss": 1.5455,
+      "step": 6589
+    },
+    {
+      "epoch": 0.6100792686454898,
+      "grad_norm": 0.1484375,
+      "learning_rate": 0.02,
+      "loss": 1.5672,
+      "step": 6590
+    },
+    {
+      "epoch": 0.6101718451657698,
+      "grad_norm": 0.1416015625,
+      "learning_rate": 0.02,
+      "loss": 1.5174,
+      "step": 6591
+    },
+    {
+      "epoch": 0.6102644216860499,
+      "grad_norm": 0.1484375,
+      "learning_rate": 0.02,
+      "loss": 1.5988,
+      "step": 6592
+    },
+    {
+      "epoch": 0.6103569982063299,
+      "grad_norm": 0.1484375,
+      "learning_rate": 0.02,
+      "loss": 1.4693,
+      "step": 6593
+    },
+    {
+      "epoch": 0.61044957472661,
+      "grad_norm": 0.1435546875,
+      "learning_rate": 0.02,
+      "loss": 1.5315,
+      "step": 6594
+    },
+    {
+      "epoch": 0.61054215124689,
+      "grad_norm": 0.134765625,
+      "learning_rate": 0.02,
+      "loss": 1.5557,
+      "step": 6595
+    },
+    {
+      "epoch": 0.6106347277671701,
+      "grad_norm": 0.14453125,
+      "learning_rate": 0.02,
+      "loss": 1.5256,
+      "step": 6596
+    },
+    {
+      "epoch": 0.6107273042874501,
+      "grad_norm": 0.1455078125,
+      "learning_rate": 0.02,
+      "loss": 1.5602,
+      "step": 6597
+    },
+    {
+      "epoch": 0.6108198808077301,
+      "grad_norm": 0.1484375,
+      "learning_rate": 0.02,
+      "loss": 1.5272,
+      "step": 6598
+    },
+    {
+      "epoch": 0.6109124573280101,
+      "grad_norm": 0.1474609375,
+      "learning_rate": 0.02,
+      "loss": 1.5499,
+      "step": 6599
+    },
+    {
+      "epoch": 0.6110050338482902,
+      "grad_norm": 0.1435546875,
+      "learning_rate": 0.02,
+      "loss": 1.5372,
+      "step": 6600
+    },
+    {
+      "epoch": 0.6110976103685702,
+      "grad_norm": 0.146484375,
+      "learning_rate": 0.02,
+      "loss": 1.5684,
+      "step": 6601
+    },
+    {
+      "epoch": 0.6111901868888503,
+      "grad_norm": 0.150390625,
+      "learning_rate": 0.02,
+      "loss": 1.4804,
+      "step": 6602
+    },
+    {
+      "epoch": 0.6112827634091303,
+      "grad_norm": 0.1416015625,
+      "learning_rate": 0.02,
+      "loss": 1.534,
+      "step": 6603
+    },
+    {
+      "epoch": 0.6113753399294104,
+      "grad_norm": 0.142578125,
+      "learning_rate": 0.02,
+      "loss": 1.5451,
+      "step": 6604
+    },
+    {
+      "epoch": 0.6114679164496905,
+      "grad_norm": 0.142578125,
+      "learning_rate": 0.02,
+      "loss": 1.6211,
+      "step": 6605
+    },
+    {
+      "epoch": 0.6115604929699705,
+      "grad_norm": 0.154296875,
+      "learning_rate": 0.02,
+      "loss": 1.5822,
+      "step": 6606
+    },
+    {
+      "epoch": 0.6116530694902506,
+      "grad_norm": 0.15625,
+      "learning_rate": 0.02,
+      "loss": 1.547,
+      "step": 6607
+    },
+    {
+      "epoch": 0.6117456460105306,
+      "grad_norm": 0.158203125,
+      "learning_rate": 0.02,
+      "loss": 1.5704,
+      "step": 6608
+    },
+    {
+      "epoch": 0.6118382225308107,
+      "grad_norm": 0.1513671875,
+      "learning_rate": 0.02,
+      "loss": 1.5664,
+      "step": 6609
+    },
+    {
+      "epoch": 0.6119307990510907,
+      "grad_norm": 0.1591796875,
+      "learning_rate": 0.02,
+      "loss": 1.5201,
+      "step": 6610
+    },
+    {
+      "epoch": 0.6120233755713708,
+      "grad_norm": 0.1396484375,
+      "learning_rate": 0.02,
+      "loss": 1.4944,
+      "step": 6611
+    },
+    {
+      "epoch": 0.6121159520916507,
+      "grad_norm": 0.15234375,
+      "learning_rate": 0.02,
+      "loss": 1.5013,
+      "step": 6612
+    },
+    {
+      "epoch": 0.6122085286119308,
+      "grad_norm": 0.1416015625,
+      "learning_rate": 0.02,
+      "loss": 1.4498,
+      "step": 6613
+    },
+    {
+      "epoch": 0.6123011051322108,
+      "grad_norm": 0.158203125,
+      "learning_rate": 0.02,
+      "loss": 1.5259,
+      "step": 6614
+    },
+    {
+      "epoch": 0.6123936816524909,
+      "grad_norm": 0.140625,
+      "learning_rate": 0.02,
+      "loss": 1.5289,
+      "step": 6615
+    },
+    {
+      "epoch": 0.6124862581727709,
+      "grad_norm": 0.138671875,
+      "learning_rate": 0.02,
+      "loss": 1.5128,
+      "step": 6616
+    },
+    {
+      "epoch": 0.612578834693051,
+      "grad_norm": 0.1416015625,
+      "learning_rate": 0.02,
+      "loss": 1.5374,
+      "step": 6617
+    },
+    {
+      "epoch": 0.612671411213331,
+      "grad_norm": 0.15625,
+      "learning_rate": 0.02,
+      "loss": 1.5595,
+      "step": 6618
+    },
+    {
+      "epoch": 0.6127639877336111,
+      "grad_norm": 0.138671875,
+      "learning_rate": 0.02,
+      "loss": 1.4861,
+      "step": 6619
+    },
+    {
+      "epoch": 0.6128565642538911,
+      "grad_norm": 0.1572265625,
+      "learning_rate": 0.02,
+      "loss": 1.482,
+      "step": 6620
+    },
+    {
+      "epoch": 0.6129491407741712,
+      "grad_norm": 0.15234375,
+      "learning_rate": 0.02,
+      "loss": 1.5353,
+      "step": 6621
+    },
+    {
+      "epoch": 0.6130417172944512,
+      "grad_norm": 0.16015625,
+      "learning_rate": 0.02,
+      "loss": 1.5561,
+      "step": 6622
+    },
+    {
+      "epoch": 0.6131342938147313,
+      "grad_norm": 0.1552734375,
+      "learning_rate": 0.02,
+      "loss": 1.6124,
+      "step": 6623
+    },
+    {
+      "epoch": 0.6132268703350113,
+      "grad_norm": 0.162109375,
+      "learning_rate": 0.02,
+      "loss": 1.5961,
+      "step": 6624
+    },
+    {
+      "epoch": 0.6133194468552913,
+      "grad_norm": 0.1640625,
+      "learning_rate": 0.02,
+      "loss": 1.5311,
+      "step": 6625
+    },
+    {
+      "epoch": 0.6134120233755713,
+      "grad_norm": 0.1474609375,
+      "learning_rate": 0.02,
+      "loss": 1.5189,
+      "step": 6626
+    },
+    {
+      "epoch": 0.6135045998958514,
+      "grad_norm": 0.1494140625,
+      "learning_rate": 0.02,
+      "loss": 1.5266,
+      "step": 6627
+    },
+    {
+      "epoch": 0.6135971764161314,
+      "grad_norm": 0.150390625,
+      "learning_rate": 0.02,
+      "loss": 1.5757,
+      "step": 6628
+    },
+    {
+      "epoch": 0.6136897529364115,
+      "grad_norm": 0.15625,
+      "learning_rate": 0.02,
+      "loss": 1.6004,
+      "step": 6629
+    },
+    {
+      "epoch": 0.6137823294566915,
+      "grad_norm": 0.14453125,
+      "learning_rate": 0.02,
+      "loss": 1.5896,
+      "step": 6630
+    },
+    {
+      "epoch": 0.6138749059769716,
+      "grad_norm": 0.1455078125,
+      "learning_rate": 0.02,
+      "loss": 1.5633,
+      "step": 6631
+    },
+    {
+      "epoch": 0.6139674824972516,
+      "grad_norm": 0.1416015625,
+      "learning_rate": 0.02,
+      "loss": 1.4798,
+      "step": 6632
+    },
+    {
+      "epoch": 0.6140600590175317,
+      "grad_norm": 0.16015625,
+      "learning_rate": 0.02,
+      "loss": 1.5741,
+      "step": 6633
+    },
+    {
+      "epoch": 0.6141526355378117,
+      "grad_norm": 0.1513671875,
+      "learning_rate": 0.02,
+      "loss": 1.5684,
+      "step": 6634
+    },
+    {
+      "epoch": 0.6142452120580918,
+      "grad_norm": 0.1494140625,
+      "learning_rate": 0.02,
+      "loss": 1.5334,
+      "step": 6635
+    },
+    {
+      "epoch": 0.6143377885783718,
+      "grad_norm": 0.140625,
+      "learning_rate": 0.02,
+      "loss": 1.4836,
+      "step": 6636
+    },
+    {
+      "epoch": 0.6144303650986519,
+      "grad_norm": 0.15234375,
+      "learning_rate": 0.02,
+      "loss": 1.57,
+      "step": 6637
+    },
+    {
+      "epoch": 0.614522941618932,
+      "grad_norm": 0.13671875,
+      "learning_rate": 0.02,
+      "loss": 1.5456,
+      "step": 6638
+    },
+    {
+      "epoch": 0.6146155181392119,
+      "grad_norm": 0.1484375,
+      "learning_rate": 0.02,
+      "loss": 1.528,
+      "step": 6639
+    },
+    {
+      "epoch": 0.6147080946594919,
+      "grad_norm": 0.142578125,
+      "learning_rate": 0.02,
+      "loss": 1.5032,
+      "step": 6640
+    },
+    {
+      "epoch": 0.614800671179772,
+      "grad_norm": 0.13671875,
+      "learning_rate": 0.02,
+      "loss": 1.5767,
+      "step": 6641
+    },
+    {
+      "epoch": 0.614893247700052,
+      "grad_norm": 0.142578125,
+      "learning_rate": 0.02,
+      "loss": 1.5037,
+      "step": 6642
+    },
+    {
+      "epoch": 0.6149858242203321,
+      "grad_norm": 0.1533203125,
+      "learning_rate": 0.02,
+      "loss": 1.5108,
+      "step": 6643
+    },
+    {
+      "epoch": 0.6150784007406122,
+      "grad_norm": 0.1494140625,
+      "learning_rate": 0.02,
+      "loss": 1.5914,
+      "step": 6644
+    },
+    {
+      "epoch": 0.6151709772608922,
+      "grad_norm": 0.1533203125,
+      "learning_rate": 0.02,
+      "loss": 1.5933,
+      "step": 6645
+    },
+    {
+      "epoch": 0.6152635537811723,
+      "grad_norm": 0.1474609375,
+      "learning_rate": 0.02,
+      "loss": 1.5199,
+      "step": 6646
+    },
+    {
+      "epoch": 0.6153561303014523,
+      "grad_norm": 0.15234375,
+      "learning_rate": 0.02,
+      "loss": 1.5621,
+      "step": 6647
+    },
+    {
+      "epoch": 0.6154487068217324,
+      "grad_norm": 0.1552734375,
+      "learning_rate": 0.02,
+      "loss": 1.5604,
+      "step": 6648
+    },
+    {
+      "epoch": 0.6155412833420124,
+      "grad_norm": 0.146484375,
+      "learning_rate": 0.02,
+      "loss": 1.4952,
+      "step": 6649
+    },
+    {
+      "epoch": 0.6156338598622925,
+      "grad_norm": 0.14453125,
+      "learning_rate": 0.02,
+      "loss": 1.5301,
+      "step": 6650
+    },
+    {
+      "epoch": 0.6157264363825725,
+      "grad_norm": 0.1513671875,
+      "learning_rate": 0.02,
+      "loss": 1.5563,
+      "step": 6651
+    },
+    {
+      "epoch": 0.6158190129028526,
+      "grad_norm": 0.146484375,
+      "learning_rate": 0.02,
+      "loss": 1.5388,
+      "step": 6652
+    },
+    {
+      "epoch": 0.6159115894231325,
+      "grad_norm": 0.1513671875,
+      "learning_rate": 0.02,
+      "loss": 1.6154,
+      "step": 6653
+    },
+    {
+      "epoch": 0.6160041659434126,
+      "grad_norm": 0.15625,
+      "learning_rate": 0.02,
+      "loss": 1.5262,
+      "step": 6654
+    },
+    {
+      "epoch": 0.6160967424636926,
+      "grad_norm": 0.1689453125,
+      "learning_rate": 0.02,
+      "loss": 1.5928,
+      "step": 6655
+    },
+    {
+      "epoch": 0.6161893189839727,
+      "grad_norm": 0.1572265625,
+      "learning_rate": 0.02,
+      "loss": 1.5204,
+      "step": 6656
+    },
+    {
+      "epoch": 0.6162818955042527,
+      "grad_norm": 0.14453125,
+      "learning_rate": 0.02,
+      "loss": 1.5998,
+      "step": 6657
+    },
+    {
+      "epoch": 0.6163744720245328,
+      "grad_norm": 0.1533203125,
+      "learning_rate": 0.02,
+      "loss": 1.5215,
+      "step": 6658
+    },
+    {
+      "epoch": 0.6164670485448128,
+      "grad_norm": 0.1474609375,
+      "learning_rate": 0.02,
+      "loss": 1.4974,
+      "step": 6659
+    },
+    {
+      "epoch": 0.6165596250650929,
+      "grad_norm": 0.16015625,
+      "learning_rate": 0.02,
+      "loss": 1.5237,
+      "step": 6660
+    },
+    {
+      "epoch": 0.6166522015853729,
+      "grad_norm": 0.1533203125,
+      "learning_rate": 0.02,
+      "loss": 1.5332,
+      "step": 6661
+    },
+    {
+      "epoch": 0.616744778105653,
+      "grad_norm": 0.162109375,
+      "learning_rate": 0.02,
+      "loss": 1.582,
+      "step": 6662
+    },
+    {
+      "epoch": 0.616837354625933,
+      "grad_norm": 0.1435546875,
+      "learning_rate": 0.02,
+      "loss": 1.5228,
+      "step": 6663
+    },
+    {
+      "epoch": 0.6169299311462131,
+      "grad_norm": 0.150390625,
+      "learning_rate": 0.02,
+      "loss": 1.5022,
+      "step": 6664
+    },
+    {
+      "epoch": 0.6170225076664931,
+      "grad_norm": 0.14453125,
+      "learning_rate": 0.02,
+      "loss": 1.5566,
+      "step": 6665
+    },
+    {
+      "epoch": 0.6171150841867731,
+      "grad_norm": 0.1474609375,
+      "learning_rate": 0.02,
+      "loss": 1.5247,
+      "step": 6666
+    },
+    {
+      "epoch": 0.6172076607070531,
+      "grad_norm": 0.14453125,
+      "learning_rate": 0.02,
+      "loss": 1.546,
+      "step": 6667
+    },
+    {
+      "epoch": 0.6173002372273332,
+      "grad_norm": 0.154296875,
+      "learning_rate": 0.02,
+      "loss": 1.5824,
+      "step": 6668
+    },
+    {
+      "epoch": 0.6173928137476132,
+      "grad_norm": 0.158203125,
+      "learning_rate": 0.02,
+      "loss": 1.5199,
+      "step": 6669
+    },
+    {
+      "epoch": 0.6174853902678933,
+      "grad_norm": 0.1591796875,
+      "learning_rate": 0.02,
+      "loss": 1.6074,
+      "step": 6670
+    },
+    {
+      "epoch": 0.6175779667881733,
+      "grad_norm": 0.1552734375,
+      "learning_rate": 0.02,
+      "loss": 1.5608,
+      "step": 6671
+    },
+    {
+      "epoch": 0.6176705433084534,
+      "grad_norm": 0.15625,
+      "learning_rate": 0.02,
+      "loss": 1.5506,
+      "step": 6672
+    },
+    {
+      "epoch": 0.6177631198287334,
+      "grad_norm": 0.142578125,
+      "learning_rate": 0.02,
+      "loss": 1.4967,
+      "step": 6673
+    },
+    {
+      "epoch": 0.6178556963490135,
+      "grad_norm": 0.1484375,
+      "learning_rate": 0.02,
+      "loss": 1.5419,
+      "step": 6674
+    },
+    {
+      "epoch": 0.6179482728692935,
+      "grad_norm": 0.1416015625,
+      "learning_rate": 0.02,
+      "loss": 1.5208,
+      "step": 6675
+    },
+    {
+      "epoch": 0.6180408493895736,
+      "grad_norm": 0.1484375,
+      "learning_rate": 0.02,
+      "loss": 1.4989,
+      "step": 6676
+    },
+    {
+      "epoch": 0.6181334259098537,
+      "grad_norm": 0.15234375,
+      "learning_rate": 0.02,
+      "loss": 1.5445,
+      "step": 6677
+    },
+    {
+      "epoch": 0.6182260024301337,
+      "grad_norm": 0.158203125,
+      "learning_rate": 0.02,
+      "loss": 1.5214,
+      "step": 6678
+    },
+    {
+      "epoch": 0.6183185789504138,
+      "grad_norm": 0.150390625,
+      "learning_rate": 0.02,
+      "loss": 1.5326,
+      "step": 6679
+    },
+    {
+      "epoch": 0.6184111554706937,
+      "grad_norm": 0.15234375,
+      "learning_rate": 0.02,
+      "loss": 1.5612,
+      "step": 6680
+    },
+    {
+      "epoch": 0.6185037319909737,
+      "grad_norm": 0.15625,
+      "learning_rate": 0.02,
+      "loss": 1.5966,
+      "step": 6681
+    },
+    {
+      "epoch": 0.6185963085112538,
+      "grad_norm": 0.15625,
+      "learning_rate": 0.02,
+      "loss": 1.5649,
+      "step": 6682
+    },
+    {
+      "epoch": 0.6186888850315339,
+      "grad_norm": 0.15234375,
+      "learning_rate": 0.02,
+      "loss": 1.5808,
+      "step": 6683
+    },
+    {
+      "epoch": 0.6187814615518139,
+      "grad_norm": 0.1494140625,
+      "learning_rate": 0.02,
+      "loss": 1.5585,
+      "step": 6684
+    },
+    {
+      "epoch": 0.618874038072094,
+      "grad_norm": 0.16015625,
+      "learning_rate": 0.02,
+      "loss": 1.5065,
+      "step": 6685
+    },
+    {
+      "epoch": 0.618966614592374,
+      "grad_norm": 0.146484375,
+      "learning_rate": 0.02,
+      "loss": 1.503,
+      "step": 6686
+    },
+    {
+      "epoch": 0.6190591911126541,
+      "grad_norm": 0.1708984375,
+      "learning_rate": 0.02,
+      "loss": 1.4825,
+      "step": 6687
+    },
+    {
+      "epoch": 0.6191517676329341,
+      "grad_norm": 0.1396484375,
+      "learning_rate": 0.02,
+      "loss": 1.5768,
+      "step": 6688
+    },
+    {
+      "epoch": 0.6192443441532142,
+      "grad_norm": 0.1650390625,
+      "learning_rate": 0.02,
+      "loss": 1.5553,
+      "step": 6689
+    },
+    {
+      "epoch": 0.6193369206734942,
+      "grad_norm": 0.138671875,
+      "learning_rate": 0.02,
+      "loss": 1.6236,
+      "step": 6690
+    },
+    {
+      "epoch": 0.6194294971937743,
+      "grad_norm": 0.1474609375,
+      "learning_rate": 0.02,
+      "loss": 1.4929,
+      "step": 6691
+    },
+    {
+      "epoch": 0.6195220737140543,
+      "grad_norm": 0.1435546875,
+      "learning_rate": 0.02,
+      "loss": 1.4727,
+      "step": 6692
+    },
+    {
+      "epoch": 0.6196146502343343,
+      "grad_norm": 0.1591796875,
+      "learning_rate": 0.02,
+      "loss": 1.4949,
+      "step": 6693
+    },
+    {
+      "epoch": 0.6197072267546143,
+      "grad_norm": 0.140625,
+      "learning_rate": 0.02,
+      "loss": 1.5192,
+      "step": 6694
+    },
+    {
+      "epoch": 0.6197998032748944,
+      "grad_norm": 0.1572265625,
+      "learning_rate": 0.02,
+      "loss": 1.5407,
+      "step": 6695
+    },
+    {
+      "epoch": 0.6198923797951744,
+      "grad_norm": 0.1474609375,
+      "learning_rate": 0.02,
+      "loss": 1.4978,
+      "step": 6696
+    },
+    {
+      "epoch": 0.6199849563154545,
+      "grad_norm": 0.15625,
+      "learning_rate": 0.02,
+      "loss": 1.5381,
+      "step": 6697
+    },
+    {
+      "epoch": 0.6200775328357345,
+      "grad_norm": 0.1396484375,
+      "learning_rate": 0.02,
+      "loss": 1.5135,
+      "step": 6698
+    },
+    {
+      "epoch": 0.6201701093560146,
+      "grad_norm": 0.1611328125,
+      "learning_rate": 0.02,
+      "loss": 1.5321,
+      "step": 6699
+    },
+    {
+      "epoch": 0.6202626858762946,
+      "grad_norm": 0.1591796875,
+      "learning_rate": 0.02,
+      "loss": 1.5232,
+      "step": 6700
+    },
+    {
+      "epoch": 0.6203552623965747,
+      "grad_norm": 0.1611328125,
+      "learning_rate": 0.02,
+      "loss": 1.495,
+      "step": 6701
+    },
+    {
+      "epoch": 0.6204478389168547,
+      "grad_norm": 0.14453125,
+      "learning_rate": 0.02,
+      "loss": 1.5319,
+      "step": 6702
+    },
+    {
+      "epoch": 0.6205404154371348,
+      "grad_norm": 0.1611328125,
+      "learning_rate": 0.02,
+      "loss": 1.5022,
+      "step": 6703
+    },
+    {
+      "epoch": 0.6206329919574148,
+      "grad_norm": 0.1513671875,
+      "learning_rate": 0.02,
+      "loss": 1.5856,
+      "step": 6704
+    },
+    {
+      "epoch": 0.6207255684776949,
+      "grad_norm": 0.1630859375,
+      "learning_rate": 0.02,
+      "loss": 1.5209,
+      "step": 6705
+    },
+    {
+      "epoch": 0.6208181449979749,
+      "grad_norm": 0.1611328125,
+      "learning_rate": 0.02,
+      "loss": 1.5802,
+      "step": 6706
+    },
+    {
+      "epoch": 0.6209107215182549,
+      "grad_norm": 0.14453125,
+      "learning_rate": 0.02,
+      "loss": 1.538,
+      "step": 6707
+    },
+    {
+      "epoch": 0.6210032980385349,
+      "grad_norm": 0.150390625,
+      "learning_rate": 0.02,
+      "loss": 1.5421,
+      "step": 6708
+    },
+    {
+      "epoch": 0.621095874558815,
+      "grad_norm": 0.15234375,
+      "learning_rate": 0.02,
+      "loss": 1.482,
+      "step": 6709
+    },
+    {
+      "epoch": 0.621188451079095,
+      "grad_norm": 0.1474609375,
+      "learning_rate": 0.02,
+      "loss": 1.4862,
+      "step": 6710
+    },
+    {
+      "epoch": 0.6212810275993751,
+      "grad_norm": 0.15234375,
+      "learning_rate": 0.02,
+      "loss": 1.5777,
+      "step": 6711
+    },
+    {
+      "epoch": 0.6213736041196551,
+      "grad_norm": 0.140625,
+      "learning_rate": 0.02,
+      "loss": 1.4995,
+      "step": 6712
+    },
+    {
+      "epoch": 0.6214661806399352,
+      "grad_norm": 0.1552734375,
+      "learning_rate": 0.02,
+      "loss": 1.6032,
+      "step": 6713
+    },
+    {
+      "epoch": 0.6215587571602152,
+      "grad_norm": 0.1396484375,
+      "learning_rate": 0.02,
+      "loss": 1.5751,
+      "step": 6714
+    },
+    {
+      "epoch": 0.6216513336804953,
+      "grad_norm": 0.14453125,
+      "learning_rate": 0.02,
+      "loss": 1.5646,
+      "step": 6715
+    },
+    {
+      "epoch": 0.6217439102007754,
+      "grad_norm": 0.1533203125,
+      "learning_rate": 0.02,
+      "loss": 1.5763,
+      "step": 6716
+    },
+    {
+      "epoch": 0.6218364867210554,
+      "grad_norm": 0.1474609375,
+      "learning_rate": 0.02,
+      "loss": 1.5474,
+      "step": 6717
+    },
+    {
+      "epoch": 0.6219290632413355,
+      "grad_norm": 0.1455078125,
+      "learning_rate": 0.02,
+      "loss": 1.6049,
+      "step": 6718
+    },
+    {
+      "epoch": 0.6220216397616155,
+      "grad_norm": 0.15234375,
+      "learning_rate": 0.02,
+      "loss": 1.5419,
+      "step": 6719
+    },
+    {
+      "epoch": 0.6221142162818954,
+      "grad_norm": 0.1376953125,
+      "learning_rate": 0.02,
+      "loss": 1.5154,
+      "step": 6720
+    },
+    {
+      "epoch": 0.6222067928021755,
+      "grad_norm": 0.1484375,
+      "learning_rate": 0.02,
+      "loss": 1.5755,
+      "step": 6721
+    },
+    {
+      "epoch": 0.6222993693224556,
+      "grad_norm": 0.1484375,
+      "learning_rate": 0.02,
+      "loss": 1.5325,
+      "step": 6722
+    },
+    {
+      "epoch": 0.6223919458427356,
+      "grad_norm": 0.1494140625,
+      "learning_rate": 0.02,
+      "loss": 1.6207,
+      "step": 6723
+    },
+    {
+      "epoch": 0.6224845223630157,
+      "grad_norm": 0.1513671875,
+      "learning_rate": 0.02,
+      "loss": 1.5891,
+      "step": 6724
+    },
+    {
+      "epoch": 0.6225770988832957,
+      "grad_norm": 0.146484375,
+      "learning_rate": 0.02,
+      "loss": 1.5534,
+      "step": 6725
+    },
+    {
+      "epoch": 0.6226696754035758,
+      "grad_norm": 0.1611328125,
+      "learning_rate": 0.02,
+      "loss": 1.5288,
+      "step": 6726
+    },
+    {
+      "epoch": 0.6227622519238558,
+      "grad_norm": 0.1484375,
+      "learning_rate": 0.02,
+      "loss": 1.4771,
+      "step": 6727
+    },
+    {
+      "epoch": 0.6228548284441359,
+      "grad_norm": 0.162109375,
+      "learning_rate": 0.02,
+      "loss": 1.5371,
+      "step": 6728
+    },
+    {
+      "epoch": 0.6229474049644159,
+      "grad_norm": 0.146484375,
+      "learning_rate": 0.02,
+      "loss": 1.4744,
+      "step": 6729
+    },
+    {
+      "epoch": 0.623039981484696,
+      "grad_norm": 0.1435546875,
+      "learning_rate": 0.02,
+      "loss": 1.4472,
+      "step": 6730
+    },
+    {
+      "epoch": 0.623132558004976,
+      "grad_norm": 0.13671875,
+      "learning_rate": 0.02,
+      "loss": 1.5206,
+      "step": 6731
+    },
+    {
+      "epoch": 0.6232251345252561,
+      "grad_norm": 0.16015625,
+      "learning_rate": 0.02,
+      "loss": 1.4971,
+      "step": 6732
+    },
+    {
+      "epoch": 0.6233177110455361,
+      "grad_norm": 0.1474609375,
+      "learning_rate": 0.02,
+      "loss": 1.5427,
+      "step": 6733
+    },
+    {
+      "epoch": 0.6234102875658161,
+      "grad_norm": 0.1533203125,
+      "learning_rate": 0.02,
+      "loss": 1.5065,
+      "step": 6734
+    },
+    {
+      "epoch": 0.6235028640860961,
+      "grad_norm": 0.1376953125,
+      "learning_rate": 0.02,
+      "loss": 1.5552,
+      "step": 6735
+    },
+    {
+      "epoch": 0.6235954406063762,
+      "grad_norm": 0.1552734375,
+      "learning_rate": 0.02,
+      "loss": 1.4611,
+      "step": 6736
+    },
+    {
+      "epoch": 0.6236880171266562,
+      "grad_norm": 0.14453125,
+      "learning_rate": 0.02,
+      "loss": 1.4739,
+      "step": 6737
+    },
+    {
+      "epoch": 0.6237805936469363,
+      "grad_norm": 0.15234375,
+      "learning_rate": 0.02,
+      "loss": 1.5504,
+      "step": 6738
+    },
+    {
+      "epoch": 0.6238731701672163,
+      "grad_norm": 0.1416015625,
+      "learning_rate": 0.02,
+      "loss": 1.511,
+      "step": 6739
+    },
+    {
+      "epoch": 0.6239657466874964,
+      "grad_norm": 0.14453125,
+      "learning_rate": 0.02,
+      "loss": 1.4979,
+      "step": 6740
+    },
+    {
+      "epoch": 0.6240583232077764,
+      "grad_norm": 0.138671875,
+      "learning_rate": 0.02,
+      "loss": 1.5451,
+      "step": 6741
+    },
+    {
+      "epoch": 0.6241508997280565,
+      "grad_norm": 0.1416015625,
+      "learning_rate": 0.02,
+      "loss": 1.5801,
+      "step": 6742
+    },
+    {
+      "epoch": 0.6242434762483365,
+      "grad_norm": 0.150390625,
+      "learning_rate": 0.02,
+      "loss": 1.5192,
+      "step": 6743
+    },
+    {
+      "epoch": 0.6243360527686166,
+      "grad_norm": 0.1328125,
+      "learning_rate": 0.02,
+      "loss": 1.5191,
+      "step": 6744
+    },
+    {
+      "epoch": 0.6244286292888966,
+      "grad_norm": 0.146484375,
+      "learning_rate": 0.02,
+      "loss": 1.5168,
+      "step": 6745
+    },
+    {
+      "epoch": 0.6245212058091767,
+      "grad_norm": 0.138671875,
+      "learning_rate": 0.02,
+      "loss": 1.531,
+      "step": 6746
+    },
+    {
+      "epoch": 0.6246137823294567,
+      "grad_norm": 0.146484375,
+      "learning_rate": 0.02,
+      "loss": 1.5246,
+      "step": 6747
+    },
+    {
+      "epoch": 0.6247063588497367,
+      "grad_norm": 0.142578125,
+      "learning_rate": 0.02,
+      "loss": 1.6065,
+      "step": 6748
+    },
+    {
+      "epoch": 0.6247989353700167,
+      "grad_norm": 0.15234375,
+      "learning_rate": 0.02,
+      "loss": 1.4978,
+      "step": 6749
+    },
+    {
+      "epoch": 0.6248915118902968,
+      "grad_norm": 0.13671875,
+      "learning_rate": 0.02,
+      "loss": 1.5058,
+      "step": 6750
+    },
+    {
+      "epoch": 0.6249840884105768,
+      "grad_norm": 0.15625,
+      "learning_rate": 0.02,
+      "loss": 1.4946,
+      "step": 6751
+    },
+    {
+      "epoch": 0.6250766649308569,
+      "grad_norm": 0.1484375,
+      "learning_rate": 0.02,
+      "loss": 1.6439,
+      "step": 6752
+    },
+    {
+      "epoch": 0.625169241451137,
+      "grad_norm": 0.154296875,
+      "learning_rate": 0.02,
+      "loss": 1.5023,
+      "step": 6753
+    },
+    {
+      "epoch": 0.625261817971417,
+      "grad_norm": 0.134765625,
+      "learning_rate": 0.02,
+      "loss": 1.5575,
+      "step": 6754
+    },
+    {
+      "epoch": 0.625354394491697,
+      "grad_norm": 0.1474609375,
+      "learning_rate": 0.02,
+      "loss": 1.5247,
+      "step": 6755
+    },
+    {
+      "epoch": 0.6254469710119771,
+      "grad_norm": 0.15234375,
+      "learning_rate": 0.02,
+      "loss": 1.5078,
+      "step": 6756
+    },
+    {
+      "epoch": 0.6255395475322572,
+      "grad_norm": 0.15234375,
+      "learning_rate": 0.02,
+      "loss": 1.5392,
+      "step": 6757
+    },
+    {
+      "epoch": 0.6256321240525372,
+      "grad_norm": 0.15234375,
+      "learning_rate": 0.02,
+      "loss": 1.5691,
+      "step": 6758
+    },
+    {
+      "epoch": 0.6257247005728173,
+      "grad_norm": 0.15625,
+      "learning_rate": 0.02,
+      "loss": 1.4953,
+      "step": 6759
+    },
+    {
+      "epoch": 0.6258172770930973,
+      "grad_norm": 0.142578125,
+      "learning_rate": 0.02,
+      "loss": 1.5074,
+      "step": 6760
+    },
+    {
+      "epoch": 0.6259098536133773,
+      "grad_norm": 0.1533203125,
+      "learning_rate": 0.02,
+      "loss": 1.5939,
+      "step": 6761
+    },
+    {
+      "epoch": 0.6260024301336573,
+      "grad_norm": 0.1533203125,
+      "learning_rate": 0.02,
+      "loss": 1.557,
+      "step": 6762
+    },
+    {
+      "epoch": 0.6260950066539374,
+      "grad_norm": 0.1591796875,
+      "learning_rate": 0.02,
+      "loss": 1.6182,
+      "step": 6763
+    },
+    {
+      "epoch": 0.6261875831742174,
+      "grad_norm": 0.1552734375,
+      "learning_rate": 0.02,
+      "loss": 1.6222,
+      "step": 6764
+    },
+    {
+      "epoch": 0.6262801596944975,
+      "grad_norm": 0.158203125,
+      "learning_rate": 0.02,
+      "loss": 1.5027,
+      "step": 6765
+    },
+    {
+      "epoch": 0.6263727362147775,
+      "grad_norm": 0.1484375,
+      "learning_rate": 0.02,
+      "loss": 1.5019,
+      "step": 6766
+    },
+    {
+      "epoch": 0.6264653127350576,
+      "grad_norm": 0.15234375,
+      "learning_rate": 0.02,
+      "loss": 1.5826,
+      "step": 6767
+    },
+    {
+      "epoch": 0.6265578892553376,
+      "grad_norm": 0.14453125,
+      "learning_rate": 0.02,
+      "loss": 1.5582,
+      "step": 6768
+    },
+    {
+      "epoch": 0.6266504657756177,
+      "grad_norm": 0.15234375,
+      "learning_rate": 0.02,
+      "loss": 1.5937,
+      "step": 6769
+    },
+    {
+      "epoch": 0.6267430422958977,
+      "grad_norm": 0.15234375,
+      "learning_rate": 0.02,
+      "loss": 1.6126,
+      "step": 6770
+    },
+    {
+      "epoch": 0.6268356188161778,
+      "grad_norm": 0.1455078125,
+      "learning_rate": 0.02,
+      "loss": 1.527,
+      "step": 6771
+    },
+    {
+      "epoch": 0.6269281953364578,
+      "grad_norm": 0.154296875,
+      "learning_rate": 0.02,
+      "loss": 1.5487,
+      "step": 6772
+    },
+    {
+      "epoch": 0.6270207718567379,
+      "grad_norm": 0.146484375,
+      "learning_rate": 0.02,
+      "loss": 1.5721,
+      "step": 6773
+    },
+    {
+      "epoch": 0.6271133483770179,
+      "grad_norm": 0.150390625,
+      "learning_rate": 0.02,
+      "loss": 1.5953,
+      "step": 6774
+    },
+    {
+      "epoch": 0.6272059248972979,
+      "grad_norm": 0.1474609375,
+      "learning_rate": 0.02,
+      "loss": 1.494,
+      "step": 6775
+    },
+    {
+      "epoch": 0.6272985014175779,
+      "grad_norm": 0.138671875,
+      "learning_rate": 0.02,
+      "loss": 1.5298,
+      "step": 6776
+    },
+    {
+      "epoch": 0.627391077937858,
+      "grad_norm": 0.1494140625,
+      "learning_rate": 0.02,
+      "loss": 1.5538,
+      "step": 6777
+    },
+    {
+      "epoch": 0.627483654458138,
+      "grad_norm": 0.15234375,
+      "learning_rate": 0.02,
+      "loss": 1.5399,
+      "step": 6778
+    },
+    {
+      "epoch": 0.6275762309784181,
+      "grad_norm": 0.1513671875,
+      "learning_rate": 0.02,
+      "loss": 1.5315,
+      "step": 6779
+    },
+    {
+      "epoch": 0.6276688074986981,
+      "grad_norm": 0.1474609375,
+      "learning_rate": 0.02,
+      "loss": 1.5445,
+      "step": 6780
+    },
+    {
+      "epoch": 0.6277613840189782,
+      "grad_norm": 0.1474609375,
+      "learning_rate": 0.02,
+      "loss": 1.5176,
+      "step": 6781
+    },
+    {
+      "epoch": 0.6278539605392582,
+      "grad_norm": 0.146484375,
+      "learning_rate": 0.02,
+      "loss": 1.5561,
+      "step": 6782
+    },
+    {
+      "epoch": 0.6279465370595383,
+      "grad_norm": 0.146484375,
+      "learning_rate": 0.02,
+      "loss": 1.6013,
+      "step": 6783
+    },
+    {
+      "epoch": 0.6280391135798183,
+      "grad_norm": 0.1494140625,
+      "learning_rate": 0.02,
+      "loss": 1.5188,
+      "step": 6784
+    },
+    {
+      "epoch": 0.6281316901000984,
+      "grad_norm": 0.1533203125,
+      "learning_rate": 0.02,
+      "loss": 1.5482,
+      "step": 6785
+    },
+    {
+      "epoch": 0.6282242666203784,
+      "grad_norm": 0.1533203125,
+      "learning_rate": 0.02,
+      "loss": 1.5199,
+      "step": 6786
+    },
+    {
+      "epoch": 0.6283168431406585,
+      "grad_norm": 0.1630859375,
+      "learning_rate": 0.02,
+      "loss": 1.5347,
+      "step": 6787
+    },
+    {
+      "epoch": 0.6284094196609384,
+      "grad_norm": 0.15234375,
+      "learning_rate": 0.02,
+      "loss": 1.6074,
+      "step": 6788
+    },
+    {
+      "epoch": 0.6285019961812185,
+      "grad_norm": 0.1572265625,
+      "learning_rate": 0.02,
+      "loss": 1.5224,
+      "step": 6789
+    },
+    {
+      "epoch": 0.6285945727014985,
+      "grad_norm": 0.1552734375,
+      "learning_rate": 0.02,
+      "loss": 1.566,
+      "step": 6790
+    },
+    {
+      "epoch": 0.6286871492217786,
+      "grad_norm": 0.154296875,
+      "learning_rate": 0.02,
+      "loss": 1.4844,
+      "step": 6791
+    },
+    {
+      "epoch": 0.6287797257420586,
+      "grad_norm": 0.142578125,
+      "learning_rate": 0.02,
+      "loss": 1.5236,
+      "step": 6792
+    },
+    {
+      "epoch": 0.6288723022623387,
+      "grad_norm": 0.154296875,
+      "learning_rate": 0.02,
+      "loss": 1.5125,
+      "step": 6793
+    },
+    {
+      "epoch": 0.6289648787826188,
+      "grad_norm": 0.138671875,
+      "learning_rate": 0.02,
+      "loss": 1.5127,
+      "step": 6794
+    },
+    {
+      "epoch": 0.6290574553028988,
+      "grad_norm": 0.150390625,
+      "learning_rate": 0.02,
+      "loss": 1.5138,
+      "step": 6795
+    },
+    {
+      "epoch": 0.6291500318231789,
+      "grad_norm": 0.13671875,
+      "learning_rate": 0.02,
+      "loss": 1.4401,
+      "step": 6796
+    },
+    {
+      "epoch": 0.6292426083434589,
+      "grad_norm": 0.1533203125,
+      "learning_rate": 0.02,
+      "loss": 1.5157,
+      "step": 6797
+    },
+    {
+      "epoch": 0.629335184863739,
+      "grad_norm": 0.140625,
+      "learning_rate": 0.02,
+      "loss": 1.5402,
+      "step": 6798
+    },
+    {
+      "epoch": 0.629427761384019,
+      "grad_norm": 0.146484375,
+      "learning_rate": 0.02,
+      "loss": 1.5397,
+      "step": 6799
+    },
+    {
+      "epoch": 0.6295203379042991,
+      "grad_norm": 0.1513671875,
+      "learning_rate": 0.02,
+      "loss": 1.634,
+      "step": 6800
+    },
+    {
+      "epoch": 0.6296129144245791,
+      "grad_norm": 0.150390625,
+      "learning_rate": 0.02,
+      "loss": 1.5347,
+      "step": 6801
+    },
+    {
+      "epoch": 0.6297054909448591,
+      "grad_norm": 0.1494140625,
+      "learning_rate": 0.02,
+      "loss": 1.5449,
+      "step": 6802
+    },
+    {
+      "epoch": 0.6297980674651391,
+      "grad_norm": 0.1591796875,
+      "learning_rate": 0.02,
+      "loss": 1.5431,
+      "step": 6803
+    },
+    {
+      "epoch": 0.6298906439854192,
+      "grad_norm": 0.154296875,
+      "learning_rate": 0.02,
+      "loss": 1.5944,
+      "step": 6804
+    },
+    {
+      "epoch": 0.6299832205056992,
+      "grad_norm": 0.1533203125,
+      "learning_rate": 0.02,
+      "loss": 1.5464,
+      "step": 6805
+    },
+    {
+      "epoch": 0.6300757970259793,
+      "grad_norm": 0.1552734375,
+      "learning_rate": 0.02,
+      "loss": 1.5359,
+      "step": 6806
+    },
+    {
+      "epoch": 0.6301683735462593,
+      "grad_norm": 0.1533203125,
+      "learning_rate": 0.02,
+      "loss": 1.5277,
+      "step": 6807
+    },
+    {
+      "epoch": 0.6302609500665394,
+      "grad_norm": 0.1611328125,
+      "learning_rate": 0.02,
+      "loss": 1.5286,
+      "step": 6808
+    },
+    {
+      "epoch": 0.6303535265868194,
+      "grad_norm": 0.138671875,
+      "learning_rate": 0.02,
+      "loss": 1.5568,
+      "step": 6809
+    },
+    {
+      "epoch": 0.6304461031070995,
+      "grad_norm": 0.15234375,
+      "learning_rate": 0.02,
+      "loss": 1.4955,
+      "step": 6810
+    },
+    {
+      "epoch": 0.6305386796273795,
+      "grad_norm": 0.14453125,
+      "learning_rate": 0.02,
+      "loss": 1.6033,
+      "step": 6811
+    },
+    {
+      "epoch": 0.6306312561476596,
+      "grad_norm": 0.150390625,
+      "learning_rate": 0.02,
+      "loss": 1.5297,
+      "step": 6812
+    },
+    {
+      "epoch": 0.6307238326679396,
+      "grad_norm": 0.1455078125,
+      "learning_rate": 0.02,
+      "loss": 1.5824,
+      "step": 6813
+    },
+    {
+      "epoch": 0.6308164091882197,
+      "grad_norm": 0.1474609375,
+      "learning_rate": 0.02,
+      "loss": 1.4772,
+      "step": 6814
+    },
+    {
+      "epoch": 0.6309089857084997,
+      "grad_norm": 0.134765625,
+      "learning_rate": 0.02,
+      "loss": 1.4861,
+      "step": 6815
+    },
+    {
+      "epoch": 0.6310015622287797,
+      "grad_norm": 0.1455078125,
+      "learning_rate": 0.02,
+      "loss": 1.5484,
+      "step": 6816
+    },
+    {
+      "epoch": 0.6310941387490597,
+      "grad_norm": 0.1513671875,
+      "learning_rate": 0.02,
+      "loss": 1.5725,
+      "step": 6817
+    },
+    {
+      "epoch": 0.6311867152693398,
+      "grad_norm": 0.1591796875,
+      "learning_rate": 0.02,
+      "loss": 1.5253,
+      "step": 6818
+    },
+    {
+      "epoch": 0.6312792917896198,
+      "grad_norm": 0.140625,
+      "learning_rate": 0.02,
+      "loss": 1.5673,
+      "step": 6819
+    },
+    {
+      "epoch": 0.6313718683098999,
+      "grad_norm": 0.146484375,
+      "learning_rate": 0.02,
+      "loss": 1.518,
+      "step": 6820
+    },
+    {
+      "epoch": 0.6314644448301799,
+      "grad_norm": 0.14453125,
+      "learning_rate": 0.02,
+      "loss": 1.5634,
+      "step": 6821
+    },
+    {
+      "epoch": 0.63155702135046,
+      "grad_norm": 0.1435546875,
+      "learning_rate": 0.02,
+      "loss": 1.5053,
+      "step": 6822
+    },
+    {
+      "epoch": 0.63164959787074,
+      "grad_norm": 0.1435546875,
+      "learning_rate": 0.02,
+      "loss": 1.6209,
+      "step": 6823
+    },
+    {
+      "epoch": 0.6317421743910201,
+      "grad_norm": 0.1494140625,
+      "learning_rate": 0.02,
+      "loss": 1.5405,
+      "step": 6824
+    },
+    {
+      "epoch": 0.6318347509113001,
+      "grad_norm": 0.1611328125,
+      "learning_rate": 0.02,
+      "loss": 1.5051,
+      "step": 6825
+    },
+    {
+      "epoch": 0.6319273274315802,
+      "grad_norm": 0.1494140625,
+      "learning_rate": 0.02,
+      "loss": 1.6131,
+      "step": 6826
+    },
+    {
+      "epoch": 0.6320199039518603,
+      "grad_norm": 0.1484375,
+      "learning_rate": 0.02,
+      "loss": 1.5352,
+      "step": 6827
+    },
+    {
+      "epoch": 0.6321124804721403,
+      "grad_norm": 0.146484375,
+      "learning_rate": 0.02,
+      "loss": 1.5225,
+      "step": 6828
+    },
+    {
+      "epoch": 0.6322050569924202,
+      "grad_norm": 0.158203125,
+      "learning_rate": 0.02,
+      "loss": 1.5132,
+      "step": 6829
+    },
+    {
+      "epoch": 0.6322976335127003,
+      "grad_norm": 0.1396484375,
+      "learning_rate": 0.02,
+      "loss": 1.5845,
+      "step": 6830
+    },
+    {
+      "epoch": 0.6323902100329803,
+      "grad_norm": 0.154296875,
+      "learning_rate": 0.02,
+      "loss": 1.5778,
+      "step": 6831
+    },
+    {
+      "epoch": 0.6324827865532604,
+      "grad_norm": 0.14453125,
+      "learning_rate": 0.02,
+      "loss": 1.5354,
+      "step": 6832
+    },
+    {
+      "epoch": 0.6325753630735405,
+      "grad_norm": 0.1552734375,
+      "learning_rate": 0.02,
+      "loss": 1.5108,
+      "step": 6833
+    },
+    {
+      "epoch": 0.6326679395938205,
+      "grad_norm": 0.15234375,
+      "learning_rate": 0.02,
+      "loss": 1.5193,
+      "step": 6834
+    },
+    {
+      "epoch": 0.6327605161141006,
+      "grad_norm": 0.1591796875,
+      "learning_rate": 0.02,
+      "loss": 1.5823,
+      "step": 6835
+    },
+    {
+      "epoch": 0.6328530926343806,
+      "grad_norm": 0.1474609375,
+      "learning_rate": 0.02,
+      "loss": 1.4716,
+      "step": 6836
+    },
+    {
+      "epoch": 0.6329456691546607,
+      "grad_norm": 0.15234375,
+      "learning_rate": 0.02,
+      "loss": 1.531,
+      "step": 6837
+    },
+    {
+      "epoch": 0.6330382456749407,
+      "grad_norm": 0.13671875,
+      "learning_rate": 0.02,
+      "loss": 1.5429,
+      "step": 6838
+    },
+    {
+      "epoch": 0.6331308221952208,
+      "grad_norm": 0.15234375,
+      "learning_rate": 0.02,
+      "loss": 1.5171,
+      "step": 6839
+    },
+    {
+      "epoch": 0.6332233987155008,
+      "grad_norm": 0.158203125,
+      "learning_rate": 0.02,
+      "loss": 1.4949,
+      "step": 6840
+    },
+    {
+      "epoch": 0.6333159752357809,
+      "grad_norm": 0.14453125,
+      "learning_rate": 0.02,
+      "loss": 1.5696,
+      "step": 6841
+    },
+    {
+      "epoch": 0.6334085517560609,
+      "grad_norm": 0.1474609375,
+      "learning_rate": 0.02,
+      "loss": 1.5396,
+      "step": 6842
+    },
+    {
+      "epoch": 0.6335011282763409,
+      "grad_norm": 0.14453125,
+      "learning_rate": 0.02,
+      "loss": 1.5314,
+      "step": 6843
+    },
+    {
+      "epoch": 0.6335937047966209,
+      "grad_norm": 0.16015625,
+      "learning_rate": 0.02,
+      "loss": 1.6416,
+      "step": 6844
+    },
+    {
+      "epoch": 0.633686281316901,
+      "grad_norm": 0.1572265625,
+      "learning_rate": 0.02,
+      "loss": 1.5146,
+      "step": 6845
+    },
+    {
+      "epoch": 0.633778857837181,
+      "grad_norm": 0.1572265625,
+      "learning_rate": 0.02,
+      "loss": 1.5109,
+      "step": 6846
+    },
+    {
+      "epoch": 0.6338714343574611,
+      "grad_norm": 0.1435546875,
+      "learning_rate": 0.02,
+      "loss": 1.5465,
+      "step": 6847
+    },
+    {
+      "epoch": 0.6339640108777411,
+      "grad_norm": 0.1494140625,
+      "learning_rate": 0.02,
+      "loss": 1.4715,
+      "step": 6848
+    },
+    {
+      "epoch": 0.6340565873980212,
+      "grad_norm": 0.146484375,
+      "learning_rate": 0.02,
+      "loss": 1.5156,
+      "step": 6849
+    },
+    {
+      "epoch": 0.6341491639183012,
+      "grad_norm": 0.1552734375,
+      "learning_rate": 0.02,
+      "loss": 1.5612,
+      "step": 6850
+    },
+    {
+      "epoch": 0.6342417404385813,
+      "grad_norm": 0.154296875,
+      "learning_rate": 0.02,
+      "loss": 1.5648,
+      "step": 6851
+    },
+    {
+      "epoch": 0.6343343169588613,
+      "grad_norm": 0.1455078125,
+      "learning_rate": 0.02,
+      "loss": 1.5548,
+      "step": 6852
+    },
+    {
+      "epoch": 0.6344268934791414,
+      "grad_norm": 0.14453125,
+      "learning_rate": 0.02,
+      "loss": 1.5017,
+      "step": 6853
+    },
+    {
+      "epoch": 0.6345194699994214,
+      "grad_norm": 0.1455078125,
+      "learning_rate": 0.02,
+      "loss": 1.5028,
+      "step": 6854
+    },
+    {
+      "epoch": 0.6346120465197015,
+      "grad_norm": 0.1513671875,
+      "learning_rate": 0.02,
+      "loss": 1.5272,
+      "step": 6855
+    },
+    {
+      "epoch": 0.6347046230399814,
+      "grad_norm": 0.158203125,
+      "learning_rate": 0.02,
+      "loss": 1.5115,
+      "step": 6856
+    },
+    {
+      "epoch": 0.6347971995602615,
+      "grad_norm": 0.1533203125,
+      "learning_rate": 0.02,
+      "loss": 1.5362,
+      "step": 6857
+    },
+    {
+      "epoch": 0.6348897760805415,
+      "grad_norm": 0.1533203125,
+      "learning_rate": 0.02,
+      "loss": 1.5531,
+      "step": 6858
+    },
+    {
+      "epoch": 0.6349823526008216,
+      "grad_norm": 0.140625,
+      "learning_rate": 0.02,
+      "loss": 1.5135,
+      "step": 6859
+    },
+    {
+      "epoch": 0.6350749291211016,
+      "grad_norm": 0.1455078125,
+      "learning_rate": 0.02,
+      "loss": 1.5148,
+      "step": 6860
+    },
+    {
+      "epoch": 0.6351675056413817,
+      "grad_norm": 0.142578125,
+      "learning_rate": 0.02,
+      "loss": 1.5144,
+      "step": 6861
+    },
+    {
+      "epoch": 0.6352600821616617,
+      "grad_norm": 0.1474609375,
+      "learning_rate": 0.02,
+      "loss": 1.5273,
+      "step": 6862
+    },
+    {
+      "epoch": 0.6353526586819418,
+      "grad_norm": 0.142578125,
+      "learning_rate": 0.02,
+      "loss": 1.5307,
+      "step": 6863
+    },
+    {
+      "epoch": 0.6354452352022218,
+      "grad_norm": 0.150390625,
+      "learning_rate": 0.02,
+      "loss": 1.5497,
+      "step": 6864
+    },
+    {
+      "epoch": 0.6355378117225019,
+      "grad_norm": 0.140625,
+      "learning_rate": 0.02,
+      "loss": 1.5532,
+      "step": 6865
+    },
+    {
+      "epoch": 0.635630388242782,
+      "grad_norm": 0.15234375,
+      "learning_rate": 0.02,
+      "loss": 1.5032,
+      "step": 6866
+    },
+    {
+      "epoch": 0.635722964763062,
+      "grad_norm": 0.1435546875,
+      "learning_rate": 0.02,
+      "loss": 1.484,
+      "step": 6867
+    },
+    {
+      "epoch": 0.6358155412833421,
+      "grad_norm": 0.1513671875,
+      "learning_rate": 0.02,
+      "loss": 1.5252,
+      "step": 6868
+    },
+    {
+      "epoch": 0.6359081178036221,
+      "grad_norm": 0.140625,
+      "learning_rate": 0.02,
+      "loss": 1.6132,
+      "step": 6869
+    },
+    {
+      "epoch": 0.636000694323902,
+      "grad_norm": 0.1611328125,
+      "learning_rate": 0.02,
+      "loss": 1.5779,
+      "step": 6870
+    },
+    {
+      "epoch": 0.6360932708441821,
+      "grad_norm": 0.150390625,
+      "learning_rate": 0.02,
+      "loss": 1.5395,
+      "step": 6871
+    },
+    {
+      "epoch": 0.6361858473644622,
+      "grad_norm": 0.1611328125,
+      "learning_rate": 0.02,
+      "loss": 1.4735,
+      "step": 6872
+    },
+    {
+      "epoch": 0.6362784238847422,
+      "grad_norm": 0.1494140625,
+      "learning_rate": 0.02,
+      "loss": 1.6134,
+      "step": 6873
+    },
+    {
+      "epoch": 0.6363710004050223,
+      "grad_norm": 0.1669921875,
+      "learning_rate": 0.02,
+      "loss": 1.5436,
+      "step": 6874
+    },
+    {
+      "epoch": 0.6364635769253023,
+      "grad_norm": 0.1455078125,
+      "learning_rate": 0.02,
+      "loss": 1.4479,
+      "step": 6875
+    },
+    {
+      "epoch": 0.6365561534455824,
+      "grad_norm": 0.154296875,
+      "learning_rate": 0.02,
+      "loss": 1.5501,
+      "step": 6876
+    },
+    {
+      "epoch": 0.6366487299658624,
+      "grad_norm": 0.1533203125,
+      "learning_rate": 0.02,
+      "loss": 1.5602,
+      "step": 6877
+    },
+    {
+      "epoch": 0.6367413064861425,
+      "grad_norm": 0.16015625,
+      "learning_rate": 0.02,
+      "loss": 1.5308,
+      "step": 6878
+    },
+    {
+      "epoch": 0.6368338830064225,
+      "grad_norm": 0.1513671875,
+      "learning_rate": 0.02,
+      "loss": 1.5616,
+      "step": 6879
+    },
+    {
+      "epoch": 0.6369264595267026,
+      "grad_norm": 0.1591796875,
+      "learning_rate": 0.02,
+      "loss": 1.4953,
+      "step": 6880
+    },
+    {
+      "epoch": 0.6370190360469826,
+      "grad_norm": 0.1494140625,
+      "learning_rate": 0.02,
+      "loss": 1.5021,
+      "step": 6881
+    },
+    {
+      "epoch": 0.6371116125672627,
+      "grad_norm": 0.1552734375,
+      "learning_rate": 0.02,
+      "loss": 1.5274,
+      "step": 6882
+    },
+    {
+      "epoch": 0.6372041890875427,
+      "grad_norm": 0.1416015625,
+      "learning_rate": 0.02,
+      "loss": 1.5295,
+      "step": 6883
+    },
+    {
+      "epoch": 0.6372967656078227,
+      "grad_norm": 0.1494140625,
+      "learning_rate": 0.02,
+      "loss": 1.5061,
+      "step": 6884
+    },
+    {
+      "epoch": 0.6373893421281027,
+      "grad_norm": 0.1484375,
+      "learning_rate": 0.02,
+      "loss": 1.5102,
+      "step": 6885
+    },
+    {
+      "epoch": 0.6374819186483828,
+      "grad_norm": 0.1513671875,
+      "learning_rate": 0.02,
+      "loss": 1.5372,
+      "step": 6886
+    },
+    {
+      "epoch": 0.6375744951686628,
+      "grad_norm": 0.15234375,
+      "learning_rate": 0.02,
+      "loss": 1.5279,
+      "step": 6887
+    },
+    {
+      "epoch": 0.6376670716889429,
+      "grad_norm": 0.1455078125,
+      "learning_rate": 0.02,
+      "loss": 1.5596,
+      "step": 6888
+    },
+    {
+      "epoch": 0.6377596482092229,
+      "grad_norm": 0.1474609375,
+      "learning_rate": 0.02,
+      "loss": 1.5539,
+      "step": 6889
+    },
+    {
+      "epoch": 0.637852224729503,
+      "grad_norm": 0.1669921875,
+      "learning_rate": 0.02,
+      "loss": 1.5243,
+      "step": 6890
+    },
+    {
+      "epoch": 0.637944801249783,
+      "grad_norm": 0.1611328125,
+      "learning_rate": 0.02,
+      "loss": 1.559,
+      "step": 6891
+    },
+    {
+      "epoch": 0.6380373777700631,
+      "grad_norm": 0.1474609375,
+      "learning_rate": 0.02,
+      "loss": 1.4824,
+      "step": 6892
+    },
+    {
+      "epoch": 0.6381299542903431,
+      "grad_norm": 0.1533203125,
+      "learning_rate": 0.02,
+      "loss": 1.5057,
+      "step": 6893
+    },
+    {
+      "epoch": 0.6382225308106232,
+      "grad_norm": 0.1435546875,
+      "learning_rate": 0.02,
+      "loss": 1.5187,
+      "step": 6894
+    },
+    {
+      "epoch": 0.6383151073309032,
+      "grad_norm": 0.1494140625,
+      "learning_rate": 0.02,
+      "loss": 1.5176,
+      "step": 6895
+    },
+    {
+      "epoch": 0.6384076838511833,
+      "grad_norm": 0.150390625,
+      "learning_rate": 0.02,
+      "loss": 1.5905,
+      "step": 6896
+    },
+    {
+      "epoch": 0.6385002603714632,
+      "grad_norm": 0.1455078125,
+      "learning_rate": 0.02,
+      "loss": 1.5079,
+      "step": 6897
+    },
+    {
+      "epoch": 0.6385928368917433,
+      "grad_norm": 0.1484375,
+      "learning_rate": 0.02,
+      "loss": 1.5887,
+      "step": 6898
+    },
+    {
+      "epoch": 0.6386854134120233,
+      "grad_norm": 0.1533203125,
+      "learning_rate": 0.02,
+      "loss": 1.5667,
+      "step": 6899
+    },
+    {
+      "epoch": 0.6387779899323034,
+      "grad_norm": 0.1533203125,
+      "learning_rate": 0.02,
+      "loss": 1.4797,
+      "step": 6900
+    },
+    {
+      "epoch": 0.6388705664525834,
+      "grad_norm": 0.1376953125,
+      "learning_rate": 0.02,
+      "loss": 1.5768,
+      "step": 6901
+    },
+    {
+      "epoch": 0.6389631429728635,
+      "grad_norm": 0.1513671875,
+      "learning_rate": 0.02,
+      "loss": 1.5428,
+      "step": 6902
+    },
+    {
+      "epoch": 0.6390557194931435,
+      "grad_norm": 0.1533203125,
+      "learning_rate": 0.02,
+      "loss": 1.5882,
+      "step": 6903
+    },
+    {
+      "epoch": 0.6391482960134236,
+      "grad_norm": 0.150390625,
+      "learning_rate": 0.02,
+      "loss": 1.5279,
+      "step": 6904
+    },
+    {
+      "epoch": 0.6392408725337037,
+      "grad_norm": 0.146484375,
+      "learning_rate": 0.02,
+      "loss": 1.5541,
+      "step": 6905
+    },
+    {
+      "epoch": 0.6393334490539837,
+      "grad_norm": 0.158203125,
+      "learning_rate": 0.02,
+      "loss": 1.5721,
+      "step": 6906
+    },
+    {
+      "epoch": 0.6394260255742638,
+      "grad_norm": 0.1552734375,
+      "learning_rate": 0.02,
+      "loss": 1.4988,
+      "step": 6907
+    },
+    {
+      "epoch": 0.6395186020945438,
+      "grad_norm": 0.1640625,
+      "learning_rate": 0.02,
+      "loss": 1.5483,
+      "step": 6908
+    },
+    {
+      "epoch": 0.6396111786148239,
+      "grad_norm": 0.15234375,
+      "learning_rate": 0.02,
+      "loss": 1.5715,
+      "step": 6909
+    },
+    {
+      "epoch": 0.6397037551351039,
+      "grad_norm": 0.15234375,
+      "learning_rate": 0.02,
+      "loss": 1.6095,
+      "step": 6910
+    },
+    {
+      "epoch": 0.6397963316553839,
+      "grad_norm": 0.146484375,
+      "learning_rate": 0.02,
+      "loss": 1.5479,
+      "step": 6911
+    },
+    {
+      "epoch": 0.6398889081756639,
+      "grad_norm": 0.1552734375,
+      "learning_rate": 0.02,
+      "loss": 1.5332,
+      "step": 6912
+    },
+    {
+      "epoch": 0.639981484695944,
+      "grad_norm": 0.1484375,
+      "learning_rate": 0.02,
+      "loss": 1.5467,
+      "step": 6913
+    },
+    {
+      "epoch": 0.640074061216224,
+      "grad_norm": 0.1484375,
+      "learning_rate": 0.02,
+      "loss": 1.5142,
+      "step": 6914
+    },
+    {
+      "epoch": 0.6401666377365041,
+      "grad_norm": 0.158203125,
+      "learning_rate": 0.02,
+      "loss": 1.4879,
+      "step": 6915
+    },
+    {
+      "epoch": 0.6402592142567841,
+      "grad_norm": 0.154296875,
+      "learning_rate": 0.02,
+      "loss": 1.517,
+      "step": 6916
+    },
+    {
+      "epoch": 0.6403517907770642,
+      "grad_norm": 0.1474609375,
+      "learning_rate": 0.02,
+      "loss": 1.5169,
+      "step": 6917
+    },
+    {
+      "epoch": 0.6404443672973442,
+      "grad_norm": 0.1611328125,
+      "learning_rate": 0.02,
+      "loss": 1.5689,
+      "step": 6918
+    },
+    {
+      "epoch": 0.6405369438176243,
+      "grad_norm": 0.1513671875,
+      "learning_rate": 0.02,
+      "loss": 1.5503,
+      "step": 6919
+    },
+    {
+      "epoch": 0.6406295203379043,
+      "grad_norm": 0.1591796875,
+      "learning_rate": 0.02,
+      "loss": 1.5423,
+      "step": 6920
+    },
+    {
+      "epoch": 0.6407220968581844,
+      "grad_norm": 0.134765625,
+      "learning_rate": 0.02,
+      "loss": 1.5662,
+      "step": 6921
+    },
+    {
+      "epoch": 0.6408146733784644,
+      "grad_norm": 0.15234375,
+      "learning_rate": 0.02,
+      "loss": 1.5408,
+      "step": 6922
+    },
+    {
+      "epoch": 0.6409072498987445,
+      "grad_norm": 0.1474609375,
+      "learning_rate": 0.02,
+      "loss": 1.5496,
+      "step": 6923
+    },
+    {
+      "epoch": 0.6409998264190244,
+      "grad_norm": 0.15234375,
+      "learning_rate": 0.02,
+      "loss": 1.4823,
+      "step": 6924
+    },
+    {
+      "epoch": 0.6410924029393045,
+      "grad_norm": 0.1416015625,
+      "learning_rate": 0.02,
+      "loss": 1.4633,
+      "step": 6925
+    },
+    {
+      "epoch": 0.6411849794595845,
+      "grad_norm": 0.146484375,
+      "learning_rate": 0.02,
+      "loss": 1.5403,
+      "step": 6926
+    },
+    {
+      "epoch": 0.6412775559798646,
+      "grad_norm": 0.1474609375,
+      "learning_rate": 0.02,
+      "loss": 1.5125,
+      "step": 6927
+    },
+    {
+      "epoch": 0.6413701325001446,
+      "grad_norm": 0.1494140625,
+      "learning_rate": 0.02,
+      "loss": 1.5201,
+      "step": 6928
+    },
+    {
+      "epoch": 0.6414627090204247,
+      "grad_norm": 0.12890625,
+      "learning_rate": 0.02,
+      "loss": 1.4283,
+      "step": 6929
+    },
+    {
+      "epoch": 0.6415552855407047,
+      "grad_norm": 0.1435546875,
+      "learning_rate": 0.02,
+      "loss": 1.5329,
+      "step": 6930
+    },
+    {
+      "epoch": 0.6416478620609848,
+      "grad_norm": 0.142578125,
+      "learning_rate": 0.02,
+      "loss": 1.5394,
+      "step": 6931
+    },
+    {
+      "epoch": 0.6417404385812648,
+      "grad_norm": 0.150390625,
+      "learning_rate": 0.02,
+      "loss": 1.5008,
+      "step": 6932
+    },
+    {
+      "epoch": 0.6418330151015449,
+      "grad_norm": 0.1416015625,
+      "learning_rate": 0.02,
+      "loss": 1.5371,
+      "step": 6933
+    },
+    {
+      "epoch": 0.6419255916218249,
+      "grad_norm": 0.1435546875,
+      "learning_rate": 0.02,
+      "loss": 1.565,
+      "step": 6934
+    },
+    {
+      "epoch": 0.642018168142105,
+      "grad_norm": 0.140625,
+      "learning_rate": 0.02,
+      "loss": 1.5065,
+      "step": 6935
+    },
+    {
+      "epoch": 0.642110744662385,
+      "grad_norm": 0.1533203125,
+      "learning_rate": 0.02,
+      "loss": 1.5359,
+      "step": 6936
+    },
+    {
+      "epoch": 0.6422033211826651,
+      "grad_norm": 0.142578125,
+      "learning_rate": 0.02,
+      "loss": 1.5331,
+      "step": 6937
+    },
+    {
+      "epoch": 0.642295897702945,
+      "grad_norm": 0.154296875,
+      "learning_rate": 0.02,
+      "loss": 1.5694,
+      "step": 6938
+    },
+    {
+      "epoch": 0.6423884742232251,
+      "grad_norm": 0.1494140625,
+      "learning_rate": 0.02,
+      "loss": 1.5365,
+      "step": 6939
+    },
+    {
+      "epoch": 0.6424810507435051,
+      "grad_norm": 0.1591796875,
+      "learning_rate": 0.02,
+      "loss": 1.5305,
+      "step": 6940
+    },
+    {
+      "epoch": 0.6425736272637852,
+      "grad_norm": 0.142578125,
+      "learning_rate": 0.02,
+      "loss": 1.5389,
+      "step": 6941
+    },
+    {
+      "epoch": 0.6426662037840652,
+      "grad_norm": 0.1552734375,
+      "learning_rate": 0.02,
+      "loss": 1.4914,
+      "step": 6942
+    },
+    {
+      "epoch": 0.6427587803043453,
+      "grad_norm": 0.1435546875,
+      "learning_rate": 0.02,
+      "loss": 1.5213,
+      "step": 6943
+    },
+    {
+      "epoch": 0.6428513568246254,
+      "grad_norm": 0.154296875,
+      "learning_rate": 0.02,
+      "loss": 1.5706,
+      "step": 6944
+    },
+    {
+      "epoch": 0.6429439333449054,
+      "grad_norm": 0.1494140625,
+      "learning_rate": 0.02,
+      "loss": 1.5141,
+      "step": 6945
+    },
+    {
+      "epoch": 0.6430365098651855,
+      "grad_norm": 0.1513671875,
+      "learning_rate": 0.02,
+      "loss": 1.4908,
+      "step": 6946
+    },
+    {
+      "epoch": 0.6431290863854655,
+      "grad_norm": 0.14453125,
+      "learning_rate": 0.02,
+      "loss": 1.5179,
+      "step": 6947
+    },
+    {
+      "epoch": 0.6432216629057456,
+      "grad_norm": 0.154296875,
+      "learning_rate": 0.02,
+      "loss": 1.509,
+      "step": 6948
+    },
+    {
+      "epoch": 0.6433142394260256,
+      "grad_norm": 0.140625,
+      "learning_rate": 0.02,
+      "loss": 1.5446,
+      "step": 6949
+    },
+    {
+      "epoch": 0.6434068159463057,
+      "grad_norm": 0.1572265625,
+      "learning_rate": 0.02,
+      "loss": 1.5493,
+      "step": 6950
+    },
+    {
+      "epoch": 0.6434993924665856,
+      "grad_norm": 0.1513671875,
+      "learning_rate": 0.02,
+      "loss": 1.5626,
+      "step": 6951
+    },
+    {
+      "epoch": 0.6435919689868657,
+      "grad_norm": 0.1552734375,
+      "learning_rate": 0.02,
+      "loss": 1.5388,
+      "step": 6952
+    },
+    {
+      "epoch": 0.6436845455071457,
+      "grad_norm": 0.140625,
+      "learning_rate": 0.02,
+      "loss": 1.5022,
+      "step": 6953
+    },
+    {
+      "epoch": 0.6437771220274258,
+      "grad_norm": 0.1552734375,
+      "learning_rate": 0.02,
+      "loss": 1.532,
+      "step": 6954
+    },
+    {
+      "epoch": 0.6438696985477058,
+      "grad_norm": 0.146484375,
+      "learning_rate": 0.02,
+      "loss": 1.5086,
+      "step": 6955
+    },
+    {
+      "epoch": 0.6439622750679859,
+      "grad_norm": 0.1435546875,
+      "learning_rate": 0.02,
+      "loss": 1.5256,
+      "step": 6956
+    },
+    {
+      "epoch": 0.6440548515882659,
+      "grad_norm": 0.1357421875,
+      "learning_rate": 0.02,
+      "loss": 1.545,
+      "step": 6957
+    },
+    {
+      "epoch": 0.644147428108546,
+      "grad_norm": 0.140625,
+      "learning_rate": 0.02,
+      "loss": 1.5534,
+      "step": 6958
+    },
+    {
+      "epoch": 0.644240004628826,
+      "grad_norm": 0.142578125,
+      "learning_rate": 0.02,
+      "loss": 1.5293,
+      "step": 6959
+    },
+    {
+      "epoch": 0.6443325811491061,
+      "grad_norm": 0.142578125,
+      "learning_rate": 0.02,
+      "loss": 1.4989,
+      "step": 6960
+    },
+    {
+      "epoch": 0.6444251576693861,
+      "grad_norm": 0.1513671875,
+      "learning_rate": 0.02,
+      "loss": 1.5141,
+      "step": 6961
+    },
+    {
+      "epoch": 0.6445177341896662,
+      "grad_norm": 0.1572265625,
+      "learning_rate": 0.02,
+      "loss": 1.5295,
+      "step": 6962
+    },
+    {
+      "epoch": 0.6446103107099462,
+      "grad_norm": 0.150390625,
+      "learning_rate": 0.02,
+      "loss": 1.4915,
+      "step": 6963
+    },
+    {
+      "epoch": 0.6447028872302263,
+      "grad_norm": 0.140625,
+      "learning_rate": 0.02,
+      "loss": 1.537,
+      "step": 6964
+    },
+    {
+      "epoch": 0.6447954637505062,
+      "grad_norm": 0.1552734375,
+      "learning_rate": 0.02,
+      "loss": 1.6043,
+      "step": 6965
+    },
+    {
+      "epoch": 0.6448880402707863,
+      "grad_norm": 0.1455078125,
+      "learning_rate": 0.02,
+      "loss": 1.513,
+      "step": 6966
+    },
+    {
+      "epoch": 0.6449806167910663,
+      "grad_norm": 0.154296875,
+      "learning_rate": 0.02,
+      "loss": 1.513,
+      "step": 6967
+    },
+    {
+      "epoch": 0.6450731933113464,
+      "grad_norm": 0.1455078125,
+      "learning_rate": 0.02,
+      "loss": 1.5594,
+      "step": 6968
+    },
+    {
+      "epoch": 0.6451657698316264,
+      "grad_norm": 0.1552734375,
+      "learning_rate": 0.02,
+      "loss": 1.5712,
+      "step": 6969
+    },
+    {
+      "epoch": 0.6452583463519065,
+      "grad_norm": 0.15625,
+      "learning_rate": 0.02,
+      "loss": 1.5551,
+      "step": 6970
+    },
+    {
+      "epoch": 0.6453509228721865,
+      "grad_norm": 0.1591796875,
+      "learning_rate": 0.02,
+      "loss": 1.5477,
+      "step": 6971
+    },
+    {
+      "epoch": 0.6454434993924666,
+      "grad_norm": 0.134765625,
+      "learning_rate": 0.02,
+      "loss": 1.4643,
+      "step": 6972
+    },
+    {
+      "epoch": 0.6455360759127466,
+      "grad_norm": 0.1552734375,
+      "learning_rate": 0.02,
+      "loss": 1.5483,
+      "step": 6973
+    },
+    {
+      "epoch": 0.6456286524330267,
+      "grad_norm": 0.138671875,
+      "learning_rate": 0.02,
+      "loss": 1.4795,
+      "step": 6974
+    },
+    {
+      "epoch": 0.6457212289533067,
+      "grad_norm": 0.1474609375,
+      "learning_rate": 0.02,
+      "loss": 1.5788,
+      "step": 6975
+    },
+    {
+      "epoch": 0.6458138054735868,
+      "grad_norm": 0.1455078125,
+      "learning_rate": 0.02,
+      "loss": 1.5272,
+      "step": 6976
+    },
+    {
+      "epoch": 0.6459063819938669,
+      "grad_norm": 0.146484375,
+      "learning_rate": 0.02,
+      "loss": 1.5188,
+      "step": 6977
+    },
+    {
+      "epoch": 0.6459989585141469,
+      "grad_norm": 0.1396484375,
+      "learning_rate": 0.02,
+      "loss": 1.5237,
+      "step": 6978
+    },
+    {
+      "epoch": 0.6460915350344268,
+      "grad_norm": 0.15625,
+      "learning_rate": 0.02,
+      "loss": 1.5556,
+      "step": 6979
+    },
+    {
+      "epoch": 0.6461841115547069,
+      "grad_norm": 0.15625,
+      "learning_rate": 0.02,
+      "loss": 1.6078,
+      "step": 6980
+    },
+    {
+      "epoch": 0.646276688074987,
+      "grad_norm": 0.146484375,
+      "learning_rate": 0.02,
+      "loss": 1.5068,
+      "step": 6981
+    },
+    {
+      "epoch": 0.646369264595267,
+      "grad_norm": 0.150390625,
+      "learning_rate": 0.02,
+      "loss": 1.5594,
+      "step": 6982
+    },
+    {
+      "epoch": 0.646461841115547,
+      "grad_norm": 0.1376953125,
+      "learning_rate": 0.02,
+      "loss": 1.5739,
+      "step": 6983
+    },
+    {
+      "epoch": 0.6465544176358271,
+      "grad_norm": 0.1474609375,
+      "learning_rate": 0.02,
+      "loss": 1.5648,
+      "step": 6984
+    },
+    {
+      "epoch": 0.6466469941561072,
+      "grad_norm": 0.1484375,
+      "learning_rate": 0.02,
+      "loss": 1.5095,
+      "step": 6985
+    },
+    {
+      "epoch": 0.6467395706763872,
+      "grad_norm": 0.150390625,
+      "learning_rate": 0.02,
+      "loss": 1.5381,
+      "step": 6986
+    },
+    {
+      "epoch": 0.6468321471966673,
+      "grad_norm": 0.162109375,
+      "learning_rate": 0.02,
+      "loss": 1.592,
+      "step": 6987
+    },
+    {
+      "epoch": 0.6469247237169473,
+      "grad_norm": 0.1533203125,
+      "learning_rate": 0.02,
+      "loss": 1.5577,
+      "step": 6988
+    },
+    {
+      "epoch": 0.6470173002372274,
+      "grad_norm": 0.158203125,
+      "learning_rate": 0.02,
+      "loss": 1.5469,
+      "step": 6989
+    },
+    {
+      "epoch": 0.6471098767575074,
+      "grad_norm": 0.1416015625,
+      "learning_rate": 0.02,
+      "loss": 1.5236,
+      "step": 6990
+    },
+    {
+      "epoch": 0.6472024532777875,
+      "grad_norm": 0.1474609375,
+      "learning_rate": 0.02,
+      "loss": 1.496,
+      "step": 6991
+    },
+    {
+      "epoch": 0.6472950297980674,
+      "grad_norm": 0.1435546875,
+      "learning_rate": 0.02,
+      "loss": 1.5682,
+      "step": 6992
+    },
+    {
+      "epoch": 0.6473876063183475,
+      "grad_norm": 0.1591796875,
+      "learning_rate": 0.02,
+      "loss": 1.5051,
+      "step": 6993
+    },
+    {
+      "epoch": 0.6474801828386275,
+      "grad_norm": 0.15234375,
+      "learning_rate": 0.02,
+      "loss": 1.4882,
+      "step": 6994
+    },
+    {
+      "epoch": 0.6475727593589076,
+      "grad_norm": 0.1591796875,
+      "learning_rate": 0.02,
+      "loss": 1.5555,
+      "step": 6995
+    },
+    {
+      "epoch": 0.6476653358791876,
+      "grad_norm": 0.158203125,
+      "learning_rate": 0.02,
+      "loss": 1.5144,
+      "step": 6996
+    },
+    {
+      "epoch": 0.6477579123994677,
+      "grad_norm": 0.16015625,
+      "learning_rate": 0.02,
+      "loss": 1.5738,
+      "step": 6997
+    },
+    {
+      "epoch": 0.6478504889197477,
+      "grad_norm": 0.15625,
+      "learning_rate": 0.02,
+      "loss": 1.5481,
+      "step": 6998
+    },
+    {
+      "epoch": 0.6479430654400278,
+      "grad_norm": 0.15234375,
+      "learning_rate": 0.02,
+      "loss": 1.5341,
+      "step": 6999
+    },
+    {
+      "epoch": 0.6480356419603078,
+      "grad_norm": 0.1611328125,
+      "learning_rate": 0.02,
+      "loss": 1.57,
+      "step": 7000
+    },
+    {
+      "epoch": 0.6481282184805879,
+      "grad_norm": 0.1611328125,
+      "learning_rate": 0.02,
+      "loss": 1.5522,
+      "step": 7001
+    },
+    {
+      "epoch": 0.6482207950008679,
+      "grad_norm": 0.142578125,
+      "learning_rate": 0.02,
+      "loss": 1.4615,
+      "step": 7002
+    },
+    {
+      "epoch": 0.648313371521148,
+      "grad_norm": 0.1376953125,
+      "learning_rate": 0.02,
+      "loss": 1.4901,
+      "step": 7003
+    },
+    {
+      "epoch": 0.648405948041428,
+      "grad_norm": 0.142578125,
+      "learning_rate": 0.02,
+      "loss": 1.5291,
+      "step": 7004
+    },
+    {
+      "epoch": 0.6484985245617081,
+      "grad_norm": 0.1455078125,
+      "learning_rate": 0.02,
+      "loss": 1.5162,
+      "step": 7005
+    },
+    {
+      "epoch": 0.648591101081988,
+      "grad_norm": 0.1328125,
+      "learning_rate": 0.02,
+      "loss": 1.55,
+      "step": 7006
+    },
+    {
+      "epoch": 0.6486836776022681,
+      "grad_norm": 0.13671875,
+      "learning_rate": 0.02,
+      "loss": 1.503,
+      "step": 7007
+    },
+    {
+      "epoch": 0.6487762541225481,
+      "grad_norm": 0.14453125,
+      "learning_rate": 0.02,
+      "loss": 1.5728,
+      "step": 7008
+    },
+    {
+      "epoch": 0.6488688306428282,
+      "grad_norm": 0.142578125,
+      "learning_rate": 0.02,
+      "loss": 1.5504,
+      "step": 7009
+    },
+    {
+      "epoch": 0.6489614071631082,
+      "grad_norm": 0.1416015625,
+      "learning_rate": 0.02,
+      "loss": 1.5112,
+      "step": 7010
+    },
+    {
+      "epoch": 0.6490539836833883,
+      "grad_norm": 0.1484375,
+      "learning_rate": 0.02,
+      "loss": 1.498,
+      "step": 7011
+    },
+    {
+      "epoch": 0.6491465602036683,
+      "grad_norm": 0.1533203125,
+      "learning_rate": 0.02,
+      "loss": 1.5504,
+      "step": 7012
+    },
+    {
+      "epoch": 0.6492391367239484,
+      "grad_norm": 0.1484375,
+      "learning_rate": 0.02,
+      "loss": 1.5146,
+      "step": 7013
+    },
+    {
+      "epoch": 0.6493317132442284,
+      "grad_norm": 0.134765625,
+      "learning_rate": 0.02,
+      "loss": 1.5434,
+      "step": 7014
+    },
+    {
+      "epoch": 0.6494242897645085,
+      "grad_norm": 0.142578125,
+      "learning_rate": 0.02,
+      "loss": 1.6207,
+      "step": 7015
+    },
+    {
+      "epoch": 0.6495168662847886,
+      "grad_norm": 0.1474609375,
+      "learning_rate": 0.02,
+      "loss": 1.5758,
+      "step": 7016
+    },
+    {
+      "epoch": 0.6496094428050686,
+      "grad_norm": 0.140625,
+      "learning_rate": 0.02,
+      "loss": 1.5382,
+      "step": 7017
+    },
+    {
+      "epoch": 0.6497020193253487,
+      "grad_norm": 0.142578125,
+      "learning_rate": 0.02,
+      "loss": 1.51,
+      "step": 7018
+    },
+    {
+      "epoch": 0.6497945958456286,
+      "grad_norm": 0.1474609375,
+      "learning_rate": 0.02,
+      "loss": 1.5833,
+      "step": 7019
+    },
+    {
+      "epoch": 0.6498871723659086,
+      "grad_norm": 0.13671875,
+      "learning_rate": 0.02,
+      "loss": 1.5017,
+      "step": 7020
+    },
+    {
+      "epoch": 0.6499797488861887,
+      "grad_norm": 0.14453125,
+      "learning_rate": 0.02,
+      "loss": 1.6002,
+      "step": 7021
+    },
+    {
+      "epoch": 0.6500723254064688,
+      "grad_norm": 0.158203125,
+      "learning_rate": 0.02,
+      "loss": 1.5729,
+      "step": 7022
+    },
+    {
+      "epoch": 0.6501649019267488,
+      "grad_norm": 0.1611328125,
+      "learning_rate": 0.02,
+      "loss": 1.5552,
+      "step": 7023
+    },
+    {
+      "epoch": 0.6502574784470289,
+      "grad_norm": 0.1533203125,
+      "learning_rate": 0.02,
+      "loss": 1.5883,
+      "step": 7024
+    },
+    {
+      "epoch": 0.6503500549673089,
+      "grad_norm": 0.16015625,
+      "learning_rate": 0.02,
+      "loss": 1.5193,
+      "step": 7025
+    },
+    {
+      "epoch": 0.650442631487589,
+      "grad_norm": 0.1640625,
+      "learning_rate": 0.02,
+      "loss": 1.5085,
+      "step": 7026
+    },
+    {
+      "epoch": 0.650535208007869,
+      "grad_norm": 0.146484375,
+      "learning_rate": 0.02,
+      "loss": 1.5219,
+      "step": 7027
+    },
+    {
+      "epoch": 0.6506277845281491,
+      "grad_norm": 0.1455078125,
+      "learning_rate": 0.02,
+      "loss": 1.5924,
+      "step": 7028
+    },
+    {
+      "epoch": 0.6507203610484291,
+      "grad_norm": 0.162109375,
+      "learning_rate": 0.02,
+      "loss": 1.5353,
+      "step": 7029
+    },
+    {
+      "epoch": 0.6508129375687092,
+      "grad_norm": 0.15625,
+      "learning_rate": 0.02,
+      "loss": 1.498,
+      "step": 7030
+    },
+    {
+      "epoch": 0.6509055140889892,
+      "grad_norm": 0.1435546875,
+      "learning_rate": 0.02,
+      "loss": 1.5127,
+      "step": 7031
+    },
+    {
+      "epoch": 0.6509980906092693,
+      "grad_norm": 0.1552734375,
+      "learning_rate": 0.02,
+      "loss": 1.5929,
+      "step": 7032
+    },
+    {
+      "epoch": 0.6510906671295492,
+      "grad_norm": 0.146484375,
+      "learning_rate": 0.02,
+      "loss": 1.5752,
+      "step": 7033
+    },
+    {
+      "epoch": 0.6511832436498293,
+      "grad_norm": 0.1533203125,
+      "learning_rate": 0.02,
+      "loss": 1.5528,
+      "step": 7034
+    },
+    {
+      "epoch": 0.6512758201701093,
+      "grad_norm": 0.1484375,
+      "learning_rate": 0.02,
+      "loss": 1.5595,
+      "step": 7035
+    },
+    {
+      "epoch": 0.6513683966903894,
+      "grad_norm": 0.14453125,
+      "learning_rate": 0.02,
+      "loss": 1.5401,
+      "step": 7036
+    },
+    {
+      "epoch": 0.6514609732106694,
+      "grad_norm": 0.1572265625,
+      "learning_rate": 0.02,
+      "loss": 1.6001,
+      "step": 7037
+    },
+    {
+      "epoch": 0.6515535497309495,
+      "grad_norm": 0.1494140625,
+      "learning_rate": 0.02,
+      "loss": 1.5089,
+      "step": 7038
+    },
+    {
+      "epoch": 0.6516461262512295,
+      "grad_norm": 0.1474609375,
+      "learning_rate": 0.02,
+      "loss": 1.5124,
+      "step": 7039
+    },
+    {
+      "epoch": 0.6517387027715096,
+      "grad_norm": 0.1435546875,
+      "learning_rate": 0.02,
+      "loss": 1.5675,
+      "step": 7040
+    },
+    {
+      "epoch": 0.6518312792917896,
+      "grad_norm": 0.15625,
+      "learning_rate": 0.02,
+      "loss": 1.4882,
+      "step": 7041
+    },
+    {
+      "epoch": 0.6519238558120697,
+      "grad_norm": 0.14453125,
+      "learning_rate": 0.02,
+      "loss": 1.5613,
+      "step": 7042
+    },
+    {
+      "epoch": 0.6520164323323497,
+      "grad_norm": 0.1494140625,
+      "learning_rate": 0.02,
+      "loss": 1.5192,
+      "step": 7043
+    },
+    {
+      "epoch": 0.6521090088526298,
+      "grad_norm": 0.1474609375,
+      "learning_rate": 0.02,
+      "loss": 1.5245,
+      "step": 7044
+    },
+    {
+      "epoch": 0.6522015853729098,
+      "grad_norm": 0.1533203125,
+      "learning_rate": 0.02,
+      "loss": 1.5628,
+      "step": 7045
+    },
+    {
+      "epoch": 0.6522941618931899,
+      "grad_norm": 0.1357421875,
+      "learning_rate": 0.02,
+      "loss": 1.5165,
+      "step": 7046
+    },
+    {
+      "epoch": 0.6523867384134698,
+      "grad_norm": 0.1640625,
+      "learning_rate": 0.02,
+      "loss": 1.4986,
+      "step": 7047
+    },
+    {
+      "epoch": 0.6524793149337499,
+      "grad_norm": 0.1376953125,
+      "learning_rate": 0.02,
+      "loss": 1.5836,
+      "step": 7048
+    },
+    {
+      "epoch": 0.6525718914540299,
+      "grad_norm": 0.1513671875,
+      "learning_rate": 0.02,
+      "loss": 1.5569,
+      "step": 7049
+    },
+    {
+      "epoch": 0.65266446797431,
+      "grad_norm": 0.13671875,
+      "learning_rate": 0.02,
+      "loss": 1.5142,
+      "step": 7050
+    },
+    {
+      "epoch": 0.65275704449459,
+      "grad_norm": 0.1435546875,
+      "learning_rate": 0.02,
+      "loss": 1.6124,
+      "step": 7051
+    },
+    {
+      "epoch": 0.6528496210148701,
+      "grad_norm": 0.1357421875,
+      "learning_rate": 0.02,
+      "loss": 1.4804,
+      "step": 7052
+    },
+    {
+      "epoch": 0.6529421975351501,
+      "grad_norm": 0.1455078125,
+      "learning_rate": 0.02,
+      "loss": 1.534,
+      "step": 7053
+    },
+    {
+      "epoch": 0.6530347740554302,
+      "grad_norm": 0.1376953125,
+      "learning_rate": 0.02,
+      "loss": 1.5432,
+      "step": 7054
+    },
+    {
+      "epoch": 0.6531273505757103,
+      "grad_norm": 0.1455078125,
+      "learning_rate": 0.02,
+      "loss": 1.4559,
+      "step": 7055
+    },
+    {
+      "epoch": 0.6532199270959903,
+      "grad_norm": 0.14453125,
+      "learning_rate": 0.02,
+      "loss": 1.5517,
+      "step": 7056
+    },
+    {
+      "epoch": 0.6533125036162704,
+      "grad_norm": 0.150390625,
+      "learning_rate": 0.02,
+      "loss": 1.5942,
+      "step": 7057
+    },
+    {
+      "epoch": 0.6534050801365504,
+      "grad_norm": 0.14453125,
+      "learning_rate": 0.02,
+      "loss": 1.4725,
+      "step": 7058
+    },
+    {
+      "epoch": 0.6534976566568305,
+      "grad_norm": 0.1533203125,
+      "learning_rate": 0.02,
+      "loss": 1.5225,
+      "step": 7059
+    },
+    {
+      "epoch": 0.6535902331771104,
+      "grad_norm": 0.1513671875,
+      "learning_rate": 0.02,
+      "loss": 1.5129,
+      "step": 7060
+    },
+    {
+      "epoch": 0.6536828096973905,
+      "grad_norm": 0.15234375,
+      "learning_rate": 0.02,
+      "loss": 1.5477,
+      "step": 7061
+    },
+    {
+      "epoch": 0.6537753862176705,
+      "grad_norm": 0.1533203125,
+      "learning_rate": 0.02,
+      "loss": 1.4667,
+      "step": 7062
+    },
+    {
+      "epoch": 0.6538679627379506,
+      "grad_norm": 0.150390625,
+      "learning_rate": 0.02,
+      "loss": 1.5612,
+      "step": 7063
+    },
+    {
+      "epoch": 0.6539605392582306,
+      "grad_norm": 0.1630859375,
+      "learning_rate": 0.02,
+      "loss": 1.524,
+      "step": 7064
+    },
+    {
+      "epoch": 0.6540531157785107,
+      "grad_norm": 0.1630859375,
+      "learning_rate": 0.02,
+      "loss": 1.5737,
+      "step": 7065
+    },
+    {
+      "epoch": 0.6541456922987907,
+      "grad_norm": 0.16015625,
+      "learning_rate": 0.02,
+      "loss": 1.5636,
+      "step": 7066
+    },
+    {
+      "epoch": 0.6542382688190708,
+      "grad_norm": 0.1435546875,
+      "learning_rate": 0.02,
+      "loss": 1.538,
+      "step": 7067
+    },
+    {
+      "epoch": 0.6543308453393508,
+      "grad_norm": 0.1591796875,
+      "learning_rate": 0.02,
+      "loss": 1.4911,
+      "step": 7068
+    },
+    {
+      "epoch": 0.6544234218596309,
+      "grad_norm": 0.146484375,
+      "learning_rate": 0.02,
+      "loss": 1.5825,
+      "step": 7069
+    },
+    {
+      "epoch": 0.6545159983799109,
+      "grad_norm": 0.1474609375,
+      "learning_rate": 0.02,
+      "loss": 1.489,
+      "step": 7070
+    },
+    {
+      "epoch": 0.654608574900191,
+      "grad_norm": 0.13671875,
+      "learning_rate": 0.02,
+      "loss": 1.4667,
+      "step": 7071
+    },
+    {
+      "epoch": 0.654701151420471,
+      "grad_norm": 0.1474609375,
+      "learning_rate": 0.02,
+      "loss": 1.5188,
+      "step": 7072
+    },
+    {
+      "epoch": 0.6547937279407511,
+      "grad_norm": 0.15234375,
+      "learning_rate": 0.02,
+      "loss": 1.4967,
+      "step": 7073
+    },
+    {
+      "epoch": 0.654886304461031,
+      "grad_norm": 0.154296875,
+      "learning_rate": 0.02,
+      "loss": 1.502,
+      "step": 7074
+    },
+    {
+      "epoch": 0.6549788809813111,
+      "grad_norm": 0.1416015625,
+      "learning_rate": 0.02,
+      "loss": 1.538,
+      "step": 7075
+    },
+    {
+      "epoch": 0.6550714575015911,
+      "grad_norm": 0.154296875,
+      "learning_rate": 0.02,
+      "loss": 1.5159,
+      "step": 7076
+    },
+    {
+      "epoch": 0.6551640340218712,
+      "grad_norm": 0.14453125,
+      "learning_rate": 0.02,
+      "loss": 1.5382,
+      "step": 7077
+    },
+    {
+      "epoch": 0.6552566105421512,
+      "grad_norm": 0.1455078125,
+      "learning_rate": 0.02,
+      "loss": 1.5602,
+      "step": 7078
+    },
+    {
+      "epoch": 0.6553491870624313,
+      "grad_norm": 0.1572265625,
+      "learning_rate": 0.02,
+      "loss": 1.4543,
+      "step": 7079
+    },
+    {
+      "epoch": 0.6554417635827113,
+      "grad_norm": 0.1572265625,
+      "learning_rate": 0.02,
+      "loss": 1.543,
+      "step": 7080
+    },
+    {
+      "epoch": 0.6555343401029914,
+      "grad_norm": 0.1533203125,
+      "learning_rate": 0.02,
+      "loss": 1.5178,
+      "step": 7081
+    },
+    {
+      "epoch": 0.6556269166232714,
+      "grad_norm": 0.162109375,
+      "learning_rate": 0.02,
+      "loss": 1.5675,
+      "step": 7082
+    },
+    {
+      "epoch": 0.6557194931435515,
+      "grad_norm": 0.1455078125,
+      "learning_rate": 0.02,
+      "loss": 1.5709,
+      "step": 7083
+    },
+    {
+      "epoch": 0.6558120696638315,
+      "grad_norm": 0.15625,
+      "learning_rate": 0.02,
+      "loss": 1.5332,
+      "step": 7084
+    },
+    {
+      "epoch": 0.6559046461841116,
+      "grad_norm": 0.1455078125,
+      "learning_rate": 0.02,
+      "loss": 1.6065,
+      "step": 7085
+    },
+    {
+      "epoch": 0.6559972227043916,
+      "grad_norm": 0.13671875,
+      "learning_rate": 0.02,
+      "loss": 1.5203,
+      "step": 7086
+    },
+    {
+      "epoch": 0.6560897992246716,
+      "grad_norm": 0.140625,
+      "learning_rate": 0.02,
+      "loss": 1.5032,
+      "step": 7087
+    },
+    {
+      "epoch": 0.6561823757449516,
+      "grad_norm": 0.1474609375,
+      "learning_rate": 0.02,
+      "loss": 1.5191,
+      "step": 7088
+    },
+    {
+      "epoch": 0.6562749522652317,
+      "grad_norm": 0.1572265625,
+      "learning_rate": 0.02,
+      "loss": 1.5865,
+      "step": 7089
+    },
+    {
+      "epoch": 0.6563675287855117,
+      "grad_norm": 0.1572265625,
+      "learning_rate": 0.02,
+      "loss": 1.5691,
+      "step": 7090
+    },
+    {
+      "epoch": 0.6564601053057918,
+      "grad_norm": 0.1484375,
+      "learning_rate": 0.02,
+      "loss": 1.5632,
+      "step": 7091
+    },
+    {
+      "epoch": 0.6565526818260718,
+      "grad_norm": 0.1435546875,
+      "learning_rate": 0.02,
+      "loss": 1.5852,
+      "step": 7092
+    },
+    {
+      "epoch": 0.6566452583463519,
+      "grad_norm": 0.1484375,
+      "learning_rate": 0.02,
+      "loss": 1.5442,
+      "step": 7093
+    },
+    {
+      "epoch": 0.656737834866632,
+      "grad_norm": 0.1513671875,
+      "learning_rate": 0.02,
+      "loss": 1.5066,
+      "step": 7094
+    },
+    {
+      "epoch": 0.656830411386912,
+      "grad_norm": 0.16015625,
+      "learning_rate": 0.02,
+      "loss": 1.5269,
+      "step": 7095
+    },
+    {
+      "epoch": 0.6569229879071921,
+      "grad_norm": 0.1484375,
+      "learning_rate": 0.02,
+      "loss": 1.5695,
+      "step": 7096
+    },
+    {
+      "epoch": 0.6570155644274721,
+      "grad_norm": 0.1474609375,
+      "learning_rate": 0.02,
+      "loss": 1.5025,
+      "step": 7097
+    },
+    {
+      "epoch": 0.6571081409477522,
+      "grad_norm": 0.1513671875,
+      "learning_rate": 0.02,
+      "loss": 1.5173,
+      "step": 7098
+    },
+    {
+      "epoch": 0.6572007174680322,
+      "grad_norm": 0.1455078125,
+      "learning_rate": 0.02,
+      "loss": 1.5416,
+      "step": 7099
+    },
+    {
+      "epoch": 0.6572932939883123,
+      "grad_norm": 0.1357421875,
+      "learning_rate": 0.02,
+      "loss": 1.513,
+      "step": 7100
+    },
+    {
+      "epoch": 0.6573858705085922,
+      "grad_norm": 0.15234375,
+      "learning_rate": 0.02,
+      "loss": 1.5878,
+      "step": 7101
+    },
+    {
+      "epoch": 0.6574784470288723,
+      "grad_norm": 0.14453125,
+      "learning_rate": 0.02,
+      "loss": 1.5321,
+      "step": 7102
+    },
+    {
+      "epoch": 0.6575710235491523,
+      "grad_norm": 0.1552734375,
+      "learning_rate": 0.02,
+      "loss": 1.4953,
+      "step": 7103
+    },
+    {
+      "epoch": 0.6576636000694324,
+      "grad_norm": 0.1474609375,
+      "learning_rate": 0.02,
+      "loss": 1.5587,
+      "step": 7104
+    },
+    {
+      "epoch": 0.6577561765897124,
+      "grad_norm": 0.158203125,
+      "learning_rate": 0.02,
+      "loss": 1.5255,
+      "step": 7105
+    },
+    {
+      "epoch": 0.6578487531099925,
+      "grad_norm": 0.142578125,
+      "learning_rate": 0.02,
+      "loss": 1.5729,
+      "step": 7106
+    },
+    {
+      "epoch": 0.6579413296302725,
+      "grad_norm": 0.14453125,
+      "learning_rate": 0.02,
+      "loss": 1.4722,
+      "step": 7107
+    },
+    {
+      "epoch": 0.6580339061505526,
+      "grad_norm": 0.1396484375,
+      "learning_rate": 0.02,
+      "loss": 1.5141,
+      "step": 7108
+    },
+    {
+      "epoch": 0.6581264826708326,
+      "grad_norm": 0.15234375,
+      "learning_rate": 0.02,
+      "loss": 1.53,
+      "step": 7109
+    },
+    {
+      "epoch": 0.6582190591911127,
+      "grad_norm": 0.14453125,
+      "learning_rate": 0.02,
+      "loss": 1.5049,
+      "step": 7110
+    },
+    {
+      "epoch": 0.6583116357113927,
+      "grad_norm": 0.1552734375,
+      "learning_rate": 0.02,
+      "loss": 1.5695,
+      "step": 7111
+    },
+    {
+      "epoch": 0.6584042122316728,
+      "grad_norm": 0.1474609375,
+      "learning_rate": 0.02,
+      "loss": 1.5702,
+      "step": 7112
+    },
+    {
+      "epoch": 0.6584967887519528,
+      "grad_norm": 0.1474609375,
+      "learning_rate": 0.02,
+      "loss": 1.5384,
+      "step": 7113
+    },
+    {
+      "epoch": 0.6585893652722329,
+      "grad_norm": 0.13671875,
+      "learning_rate": 0.02,
+      "loss": 1.5777,
+      "step": 7114
+    },
+    {
+      "epoch": 0.6586819417925128,
+      "grad_norm": 0.1484375,
+      "learning_rate": 0.02,
+      "loss": 1.5142,
+      "step": 7115
+    },
+    {
+      "epoch": 0.6587745183127929,
+      "grad_norm": 0.146484375,
+      "learning_rate": 0.02,
+      "loss": 1.5431,
+      "step": 7116
+    },
+    {
+      "epoch": 0.6588670948330729,
+      "grad_norm": 0.1455078125,
+      "learning_rate": 0.02,
+      "loss": 1.4982,
+      "step": 7117
+    },
+    {
+      "epoch": 0.658959671353353,
+      "grad_norm": 0.1474609375,
+      "learning_rate": 0.02,
+      "loss": 1.5792,
+      "step": 7118
+    },
+    {
+      "epoch": 0.659052247873633,
+      "grad_norm": 0.1484375,
+      "learning_rate": 0.02,
+      "loss": 1.5506,
+      "step": 7119
+    },
+    {
+      "epoch": 0.6591448243939131,
+      "grad_norm": 0.1552734375,
+      "learning_rate": 0.02,
+      "loss": 1.5503,
+      "step": 7120
+    },
+    {
+      "epoch": 0.6592374009141931,
+      "grad_norm": 0.1474609375,
+      "learning_rate": 0.02,
+      "loss": 1.5406,
+      "step": 7121
+    },
+    {
+      "epoch": 0.6593299774344732,
+      "grad_norm": 0.1435546875,
+      "learning_rate": 0.02,
+      "loss": 1.5276,
+      "step": 7122
+    },
+    {
+      "epoch": 0.6594225539547532,
+      "grad_norm": 0.142578125,
+      "learning_rate": 0.02,
+      "loss": 1.5232,
+      "step": 7123
+    },
+    {
+      "epoch": 0.6595151304750333,
+      "grad_norm": 0.1552734375,
+      "learning_rate": 0.02,
+      "loss": 1.562,
+      "step": 7124
+    },
+    {
+      "epoch": 0.6596077069953133,
+      "grad_norm": 0.14453125,
+      "learning_rate": 0.02,
+      "loss": 1.5392,
+      "step": 7125
+    },
+    {
+      "epoch": 0.6597002835155934,
+      "grad_norm": 0.150390625,
+      "learning_rate": 0.02,
+      "loss": 1.5357,
+      "step": 7126
+    },
+    {
+      "epoch": 0.6597928600358735,
+      "grad_norm": 0.1474609375,
+      "learning_rate": 0.02,
+      "loss": 1.5559,
+      "step": 7127
+    },
+    {
+      "epoch": 0.6598854365561534,
+      "grad_norm": 0.162109375,
+      "learning_rate": 0.02,
+      "loss": 1.4738,
+      "step": 7128
+    },
+    {
+      "epoch": 0.6599780130764334,
+      "grad_norm": 0.1416015625,
+      "learning_rate": 0.02,
+      "loss": 1.5174,
+      "step": 7129
+    },
+    {
+      "epoch": 0.6600705895967135,
+      "grad_norm": 0.1513671875,
+      "learning_rate": 0.02,
+      "loss": 1.4874,
+      "step": 7130
+    },
+    {
+      "epoch": 0.6601631661169935,
+      "grad_norm": 0.13671875,
+      "learning_rate": 0.02,
+      "loss": 1.5331,
+      "step": 7131
+    },
+    {
+      "epoch": 0.6602557426372736,
+      "grad_norm": 0.146484375,
+      "learning_rate": 0.02,
+      "loss": 1.5166,
+      "step": 7132
+    },
+    {
+      "epoch": 0.6603483191575537,
+      "grad_norm": 0.154296875,
+      "learning_rate": 0.02,
+      "loss": 1.5238,
+      "step": 7133
+    },
+    {
+      "epoch": 0.6604408956778337,
+      "grad_norm": 0.16015625,
+      "learning_rate": 0.02,
+      "loss": 1.5368,
+      "step": 7134
+    },
+    {
+      "epoch": 0.6605334721981138,
+      "grad_norm": 0.1484375,
+      "learning_rate": 0.02,
+      "loss": 1.4943,
+      "step": 7135
+    },
+    {
+      "epoch": 0.6606260487183938,
+      "grad_norm": 0.1591796875,
+      "learning_rate": 0.02,
+      "loss": 1.5634,
+      "step": 7136
+    },
+    {
+      "epoch": 0.6607186252386739,
+      "grad_norm": 0.14453125,
+      "learning_rate": 0.02,
+      "loss": 1.5243,
+      "step": 7137
+    },
+    {
+      "epoch": 0.6608112017589539,
+      "grad_norm": 0.158203125,
+      "learning_rate": 0.02,
+      "loss": 1.4649,
+      "step": 7138
+    },
+    {
+      "epoch": 0.660903778279234,
+      "grad_norm": 0.1376953125,
+      "learning_rate": 0.02,
+      "loss": 1.5548,
+      "step": 7139
+    },
+    {
+      "epoch": 0.660996354799514,
+      "grad_norm": 0.15234375,
+      "learning_rate": 0.02,
+      "loss": 1.546,
+      "step": 7140
+    },
+    {
+      "epoch": 0.6610889313197941,
+      "grad_norm": 0.142578125,
+      "learning_rate": 0.02,
+      "loss": 1.5231,
+      "step": 7141
+    },
+    {
+      "epoch": 0.661181507840074,
+      "grad_norm": 0.154296875,
+      "learning_rate": 0.02,
+      "loss": 1.4466,
+      "step": 7142
+    },
+    {
+      "epoch": 0.6612740843603541,
+      "grad_norm": 0.1416015625,
+      "learning_rate": 0.02,
+      "loss": 1.6078,
+      "step": 7143
+    },
+    {
+      "epoch": 0.6613666608806341,
+      "grad_norm": 0.1513671875,
+      "learning_rate": 0.02,
+      "loss": 1.5178,
+      "step": 7144
+    },
+    {
+      "epoch": 0.6614592374009142,
+      "grad_norm": 0.138671875,
+      "learning_rate": 0.02,
+      "loss": 1.5519,
+      "step": 7145
+    },
+    {
+      "epoch": 0.6615518139211942,
+      "grad_norm": 0.1533203125,
+      "learning_rate": 0.02,
+      "loss": 1.5726,
+      "step": 7146
+    },
+    {
+      "epoch": 0.6616443904414743,
+      "grad_norm": 0.14453125,
+      "learning_rate": 0.02,
+      "loss": 1.5071,
+      "step": 7147
+    },
+    {
+      "epoch": 0.6617369669617543,
+      "grad_norm": 0.15625,
+      "learning_rate": 0.02,
+      "loss": 1.5506,
+      "step": 7148
+    },
+    {
+      "epoch": 0.6618295434820344,
+      "grad_norm": 0.15625,
+      "learning_rate": 0.02,
+      "loss": 1.5557,
+      "step": 7149
+    },
+    {
+      "epoch": 0.6619221200023144,
+      "grad_norm": 0.1640625,
+      "learning_rate": 0.02,
+      "loss": 1.598,
+      "step": 7150
+    },
+    {
+      "epoch": 0.6620146965225945,
+      "grad_norm": 0.150390625,
+      "learning_rate": 0.02,
+      "loss": 1.5964,
+      "step": 7151
+    },
+    {
+      "epoch": 0.6621072730428745,
+      "grad_norm": 0.166015625,
+      "learning_rate": 0.02,
+      "loss": 1.5408,
+      "step": 7152
+    },
+    {
+      "epoch": 0.6621998495631546,
+      "grad_norm": 0.1416015625,
+      "learning_rate": 0.02,
+      "loss": 1.5969,
+      "step": 7153
+    },
+    {
+      "epoch": 0.6622924260834346,
+      "grad_norm": 0.166015625,
+      "learning_rate": 0.02,
+      "loss": 1.5948,
+      "step": 7154
+    },
+    {
+      "epoch": 0.6623850026037146,
+      "grad_norm": 0.162109375,
+      "learning_rate": 0.02,
+      "loss": 1.5563,
+      "step": 7155
+    },
+    {
+      "epoch": 0.6624775791239946,
+      "grad_norm": 0.166015625,
+      "learning_rate": 0.02,
+      "loss": 1.5378,
+      "step": 7156
+    },
+    {
+      "epoch": 0.6625701556442747,
+      "grad_norm": 0.1787109375,
+      "learning_rate": 0.02,
+      "loss": 1.5333,
+      "step": 7157
+    },
+    {
+      "epoch": 0.6626627321645547,
+      "grad_norm": 0.1533203125,
+      "learning_rate": 0.02,
+      "loss": 1.5619,
+      "step": 7158
+    },
+    {
+      "epoch": 0.6627553086848348,
+      "grad_norm": 0.1494140625,
+      "learning_rate": 0.02,
+      "loss": 1.5354,
+      "step": 7159
+    },
+    {
+      "epoch": 0.6628478852051148,
+      "grad_norm": 0.1552734375,
+      "learning_rate": 0.02,
+      "loss": 1.5298,
+      "step": 7160
+    },
+    {
+      "epoch": 0.6629404617253949,
+      "grad_norm": 0.1552734375,
+      "learning_rate": 0.02,
+      "loss": 1.5279,
+      "step": 7161
+    },
+    {
+      "epoch": 0.6630330382456749,
+      "grad_norm": 0.1552734375,
+      "learning_rate": 0.02,
+      "loss": 1.5337,
+      "step": 7162
+    },
+    {
+      "epoch": 0.663125614765955,
+      "grad_norm": 0.1572265625,
+      "learning_rate": 0.02,
+      "loss": 1.5593,
+      "step": 7163
+    },
+    {
+      "epoch": 0.663218191286235,
+      "grad_norm": 0.15625,
+      "learning_rate": 0.02,
+      "loss": 1.5441,
+      "step": 7164
+    },
+    {
+      "epoch": 0.6633107678065151,
+      "grad_norm": 0.162109375,
+      "learning_rate": 0.02,
+      "loss": 1.6139,
+      "step": 7165
+    },
+    {
+      "epoch": 0.6634033443267952,
+      "grad_norm": 0.1591796875,
+      "learning_rate": 0.02,
+      "loss": 1.4896,
+      "step": 7166
+    },
+    {
+      "epoch": 0.6634959208470752,
+      "grad_norm": 0.158203125,
+      "learning_rate": 0.02,
+      "loss": 1.6001,
+      "step": 7167
+    },
+    {
+      "epoch": 0.6635884973673553,
+      "grad_norm": 0.1474609375,
+      "learning_rate": 0.02,
+      "loss": 1.5397,
+      "step": 7168
+    },
+    {
+      "epoch": 0.6636810738876352,
+      "grad_norm": 0.1552734375,
+      "learning_rate": 0.02,
+      "loss": 1.5601,
+      "step": 7169
+    },
+    {
+      "epoch": 0.6637736504079152,
+      "grad_norm": 0.1513671875,
+      "learning_rate": 0.02,
+      "loss": 1.5898,
+      "step": 7170
+    },
+    {
+      "epoch": 0.6638662269281953,
+      "grad_norm": 0.1611328125,
+      "learning_rate": 0.02,
+      "loss": 1.5099,
+      "step": 7171
+    },
+    {
+      "epoch": 0.6639588034484754,
+      "grad_norm": 0.1552734375,
+      "learning_rate": 0.02,
+      "loss": 1.5629,
+      "step": 7172
+    },
+    {
+      "epoch": 0.6640513799687554,
+      "grad_norm": 0.1484375,
+      "learning_rate": 0.02,
+      "loss": 1.4783,
+      "step": 7173
+    },
+    {
+      "epoch": 0.6641439564890355,
+      "grad_norm": 0.138671875,
+      "learning_rate": 0.02,
+      "loss": 1.4947,
+      "step": 7174
+    },
+    {
+      "epoch": 0.6642365330093155,
+      "grad_norm": 0.1474609375,
+      "learning_rate": 0.02,
+      "loss": 1.5053,
+      "step": 7175
+    },
+    {
+      "epoch": 0.6643291095295956,
+      "grad_norm": 0.142578125,
+      "learning_rate": 0.02,
+      "loss": 1.5149,
+      "step": 7176
+    },
+    {
+      "epoch": 0.6644216860498756,
+      "grad_norm": 0.1435546875,
+      "learning_rate": 0.02,
+      "loss": 1.5256,
+      "step": 7177
+    },
+    {
+      "epoch": 0.6645142625701557,
+      "grad_norm": 0.14453125,
+      "learning_rate": 0.02,
+      "loss": 1.5664,
+      "step": 7178
+    },
+    {
+      "epoch": 0.6646068390904357,
+      "grad_norm": 0.1552734375,
+      "learning_rate": 0.02,
+      "loss": 1.5148,
+      "step": 7179
+    },
+    {
+      "epoch": 0.6646994156107158,
+      "grad_norm": 0.15234375,
+      "learning_rate": 0.02,
+      "loss": 1.5602,
+      "step": 7180
+    },
+    {
+      "epoch": 0.6647919921309958,
+      "grad_norm": 0.146484375,
+      "learning_rate": 0.02,
+      "loss": 1.5878,
+      "step": 7181
+    },
+    {
+      "epoch": 0.6648845686512758,
+      "grad_norm": 0.150390625,
+      "learning_rate": 0.02,
+      "loss": 1.5171,
+      "step": 7182
+    },
+    {
+      "epoch": 0.6649771451715558,
+      "grad_norm": 0.140625,
+      "learning_rate": 0.02,
+      "loss": 1.5691,
+      "step": 7183
+    },
+    {
+      "epoch": 0.6650697216918359,
+      "grad_norm": 0.1474609375,
+      "learning_rate": 0.02,
+      "loss": 1.5745,
+      "step": 7184
+    },
+    {
+      "epoch": 0.6651622982121159,
+      "grad_norm": 0.1484375,
+      "learning_rate": 0.02,
+      "loss": 1.4759,
+      "step": 7185
+    },
+    {
+      "epoch": 0.665254874732396,
+      "grad_norm": 0.1494140625,
+      "learning_rate": 0.02,
+      "loss": 1.5892,
+      "step": 7186
+    },
+    {
+      "epoch": 0.665347451252676,
+      "grad_norm": 0.1435546875,
+      "learning_rate": 0.02,
+      "loss": 1.5543,
+      "step": 7187
+    },
+    {
+      "epoch": 0.6654400277729561,
+      "grad_norm": 0.154296875,
+      "learning_rate": 0.02,
+      "loss": 1.502,
+      "step": 7188
+    },
+    {
+      "epoch": 0.6655326042932361,
+      "grad_norm": 0.1513671875,
+      "learning_rate": 0.02,
+      "loss": 1.474,
+      "step": 7189
+    },
+    {
+      "epoch": 0.6656251808135162,
+      "grad_norm": 0.146484375,
+      "learning_rate": 0.02,
+      "loss": 1.5121,
+      "step": 7190
+    },
+    {
+      "epoch": 0.6657177573337962,
+      "grad_norm": 0.14453125,
+      "learning_rate": 0.02,
+      "loss": 1.4957,
+      "step": 7191
+    },
+    {
+      "epoch": 0.6658103338540763,
+      "grad_norm": 0.140625,
+      "learning_rate": 0.02,
+      "loss": 1.4405,
+      "step": 7192
+    },
+    {
+      "epoch": 0.6659029103743563,
+      "grad_norm": 0.1416015625,
+      "learning_rate": 0.02,
+      "loss": 1.5286,
+      "step": 7193
+    },
+    {
+      "epoch": 0.6659954868946364,
+      "grad_norm": 0.146484375,
+      "learning_rate": 0.02,
+      "loss": 1.5325,
+      "step": 7194
+    },
+    {
+      "epoch": 0.6660880634149164,
+      "grad_norm": 0.1591796875,
+      "learning_rate": 0.02,
+      "loss": 1.5783,
+      "step": 7195
+    },
+    {
+      "epoch": 0.6661806399351964,
+      "grad_norm": 0.150390625,
+      "learning_rate": 0.02,
+      "loss": 1.5453,
+      "step": 7196
+    },
+    {
+      "epoch": 0.6662732164554764,
+      "grad_norm": 0.150390625,
+      "learning_rate": 0.02,
+      "loss": 1.5387,
+      "step": 7197
+    },
+    {
+      "epoch": 0.6663657929757565,
+      "grad_norm": 0.142578125,
+      "learning_rate": 0.02,
+      "loss": 1.571,
+      "step": 7198
+    },
+    {
+      "epoch": 0.6664583694960365,
+      "grad_norm": 0.158203125,
+      "learning_rate": 0.02,
+      "loss": 1.5366,
+      "step": 7199
+    },
+    {
+      "epoch": 0.6665509460163166,
+      "grad_norm": 0.1455078125,
+      "learning_rate": 0.02,
+      "loss": 1.5379,
+      "step": 7200
+    },
+    {
+      "epoch": 0.6666435225365966,
+      "grad_norm": 0.1474609375,
+      "learning_rate": 0.02,
+      "loss": 1.4347,
+      "step": 7201
+    },
+    {
+      "epoch": 0.6667360990568767,
+      "grad_norm": 0.1455078125,
+      "learning_rate": 0.02,
+      "loss": 1.5078,
+      "step": 7202
+    },
+    {
+      "epoch": 0.6668286755771567,
+      "grad_norm": 0.1591796875,
+      "learning_rate": 0.02,
+      "loss": 1.4877,
+      "step": 7203
+    },
+    {
+      "epoch": 0.6669212520974368,
+      "grad_norm": 0.154296875,
+      "learning_rate": 0.02,
+      "loss": 1.577,
+      "step": 7204
+    },
+    {
+      "epoch": 0.6670138286177169,
+      "grad_norm": 0.1533203125,
+      "learning_rate": 0.02,
+      "loss": 1.5395,
+      "step": 7205
+    },
+    {
+      "epoch": 0.6671064051379969,
+      "grad_norm": 0.154296875,
+      "learning_rate": 0.02,
+      "loss": 1.5582,
+      "step": 7206
+    },
+    {
+      "epoch": 0.667198981658277,
+      "grad_norm": 0.162109375,
+      "learning_rate": 0.02,
+      "loss": 1.6177,
+      "step": 7207
+    },
+    {
+      "epoch": 0.667291558178557,
+      "grad_norm": 0.1572265625,
+      "learning_rate": 0.02,
+      "loss": 1.5635,
+      "step": 7208
+    },
+    {
+      "epoch": 0.6673841346988371,
+      "grad_norm": 0.138671875,
+      "learning_rate": 0.02,
+      "loss": 1.4763,
+      "step": 7209
+    },
+    {
+      "epoch": 0.667476711219117,
+      "grad_norm": 0.1640625,
+      "learning_rate": 0.02,
+      "loss": 1.5029,
+      "step": 7210
+    },
+    {
+      "epoch": 0.667569287739397,
+      "grad_norm": 0.1533203125,
+      "learning_rate": 0.02,
+      "loss": 1.5508,
+      "step": 7211
+    },
+    {
+      "epoch": 0.6676618642596771,
+      "grad_norm": 0.16015625,
+      "learning_rate": 0.02,
+      "loss": 1.5336,
+      "step": 7212
+    },
+    {
+      "epoch": 0.6677544407799572,
+      "grad_norm": 0.14453125,
+      "learning_rate": 0.02,
+      "loss": 1.4986,
+      "step": 7213
+    },
+    {
+      "epoch": 0.6678470173002372,
+      "grad_norm": 0.162109375,
+      "learning_rate": 0.02,
+      "loss": 1.5035,
+      "step": 7214
+    },
+    {
+      "epoch": 0.6679395938205173,
+      "grad_norm": 0.1513671875,
+      "learning_rate": 0.02,
+      "loss": 1.56,
+      "step": 7215
+    },
+    {
+      "epoch": 0.6680321703407973,
+      "grad_norm": 0.15234375,
+      "learning_rate": 0.02,
+      "loss": 1.5499,
+      "step": 7216
+    },
+    {
+      "epoch": 0.6681247468610774,
+      "grad_norm": 0.140625,
+      "learning_rate": 0.02,
+      "loss": 1.6123,
+      "step": 7217
+    },
+    {
+      "epoch": 0.6682173233813574,
+      "grad_norm": 0.158203125,
+      "learning_rate": 0.02,
+      "loss": 1.5434,
+      "step": 7218
+    },
+    {
+      "epoch": 0.6683098999016375,
+      "grad_norm": 0.1455078125,
+      "learning_rate": 0.02,
+      "loss": 1.4776,
+      "step": 7219
+    },
+    {
+      "epoch": 0.6684024764219175,
+      "grad_norm": 0.1533203125,
+      "learning_rate": 0.02,
+      "loss": 1.524,
+      "step": 7220
+    },
+    {
+      "epoch": 0.6684950529421976,
+      "grad_norm": 0.15625,
+      "learning_rate": 0.02,
+      "loss": 1.525,
+      "step": 7221
+    },
+    {
+      "epoch": 0.6685876294624776,
+      "grad_norm": 0.15625,
+      "learning_rate": 0.02,
+      "loss": 1.5257,
+      "step": 7222
+    },
+    {
+      "epoch": 0.6686802059827576,
+      "grad_norm": 0.1513671875,
+      "learning_rate": 0.02,
+      "loss": 1.5541,
+      "step": 7223
+    },
+    {
+      "epoch": 0.6687727825030376,
+      "grad_norm": 0.15625,
+      "learning_rate": 0.02,
+      "loss": 1.5456,
+      "step": 7224
+    },
+    {
+      "epoch": 0.6688653590233177,
+      "grad_norm": 0.1591796875,
+      "learning_rate": 0.02,
+      "loss": 1.4785,
+      "step": 7225
+    },
+    {
+      "epoch": 0.6689579355435977,
+      "grad_norm": 0.15625,
+      "learning_rate": 0.02,
+      "loss": 1.5146,
+      "step": 7226
+    },
+    {
+      "epoch": 0.6690505120638778,
+      "grad_norm": 0.142578125,
+      "learning_rate": 0.02,
+      "loss": 1.4814,
+      "step": 7227
+    },
+    {
+      "epoch": 0.6691430885841578,
+      "grad_norm": 0.15234375,
+      "learning_rate": 0.02,
+      "loss": 1.4842,
+      "step": 7228
+    },
+    {
+      "epoch": 0.6692356651044379,
+      "grad_norm": 0.146484375,
+      "learning_rate": 0.02,
+      "loss": 1.4826,
+      "step": 7229
+    },
+    {
+      "epoch": 0.6693282416247179,
+      "grad_norm": 0.140625,
+      "learning_rate": 0.02,
+      "loss": 1.5381,
+      "step": 7230
+    },
+    {
+      "epoch": 0.669420818144998,
+      "grad_norm": 0.154296875,
+      "learning_rate": 0.02,
+      "loss": 1.6165,
+      "step": 7231
+    },
+    {
+      "epoch": 0.669513394665278,
+      "grad_norm": 0.15234375,
+      "learning_rate": 0.02,
+      "loss": 1.4714,
+      "step": 7232
+    },
+    {
+      "epoch": 0.6696059711855581,
+      "grad_norm": 0.1494140625,
+      "learning_rate": 0.02,
+      "loss": 1.5282,
+      "step": 7233
+    },
+    {
+      "epoch": 0.6696985477058381,
+      "grad_norm": 0.1435546875,
+      "learning_rate": 0.02,
+      "loss": 1.6161,
+      "step": 7234
+    },
+    {
+      "epoch": 0.6697911242261182,
+      "grad_norm": 0.1494140625,
+      "learning_rate": 0.02,
+      "loss": 1.498,
+      "step": 7235
+    },
+    {
+      "epoch": 0.6698837007463982,
+      "grad_norm": 0.1474609375,
+      "learning_rate": 0.02,
+      "loss": 1.465,
+      "step": 7236
+    },
+    {
+      "epoch": 0.6699762772666782,
+      "grad_norm": 0.1513671875,
+      "learning_rate": 0.02,
+      "loss": 1.5868,
+      "step": 7237
+    },
+    {
+      "epoch": 0.6700688537869582,
+      "grad_norm": 0.1484375,
+      "learning_rate": 0.02,
+      "loss": 1.5231,
+      "step": 7238
+    },
+    {
+      "epoch": 0.6701614303072383,
+      "grad_norm": 0.1513671875,
+      "learning_rate": 0.02,
+      "loss": 1.5378,
+      "step": 7239
+    },
+    {
+      "epoch": 0.6702540068275183,
+      "grad_norm": 0.1435546875,
+      "learning_rate": 0.02,
+      "loss": 1.5046,
+      "step": 7240
+    },
+    {
+      "epoch": 0.6703465833477984,
+      "grad_norm": 0.1494140625,
+      "learning_rate": 0.02,
+      "loss": 1.5005,
+      "step": 7241
+    },
+    {
+      "epoch": 0.6704391598680784,
+      "grad_norm": 0.1416015625,
+      "learning_rate": 0.02,
+      "loss": 1.4253,
+      "step": 7242
+    },
+    {
+      "epoch": 0.6705317363883585,
+      "grad_norm": 0.142578125,
+      "learning_rate": 0.02,
+      "loss": 1.4826,
+      "step": 7243
+    },
+    {
+      "epoch": 0.6706243129086386,
+      "grad_norm": 0.142578125,
+      "learning_rate": 0.02,
+      "loss": 1.4633,
+      "step": 7244
+    },
+    {
+      "epoch": 0.6707168894289186,
+      "grad_norm": 0.1416015625,
+      "learning_rate": 0.02,
+      "loss": 1.4599,
+      "step": 7245
+    },
+    {
+      "epoch": 0.6708094659491987,
+      "grad_norm": 0.1513671875,
+      "learning_rate": 0.02,
+      "loss": 1.5515,
+      "step": 7246
+    },
+    {
+      "epoch": 0.6709020424694787,
+      "grad_norm": 0.14453125,
+      "learning_rate": 0.02,
+      "loss": 1.5312,
+      "step": 7247
+    },
+    {
+      "epoch": 0.6709946189897588,
+      "grad_norm": 0.140625,
+      "learning_rate": 0.02,
+      "loss": 1.488,
+      "step": 7248
+    },
+    {
+      "epoch": 0.6710871955100388,
+      "grad_norm": 0.1484375,
+      "learning_rate": 0.02,
+      "loss": 1.5267,
+      "step": 7249
+    },
+    {
+      "epoch": 0.6711797720303188,
+      "grad_norm": 0.1455078125,
+      "learning_rate": 0.02,
+      "loss": 1.4875,
+      "step": 7250
+    },
+    {
+      "epoch": 0.6712723485505988,
+      "grad_norm": 0.14453125,
+      "learning_rate": 0.02,
+      "loss": 1.5055,
+      "step": 7251
+    },
+    {
+      "epoch": 0.6713649250708789,
+      "grad_norm": 0.1416015625,
+      "learning_rate": 0.02,
+      "loss": 1.5251,
+      "step": 7252
+    },
+    {
+      "epoch": 0.6714575015911589,
+      "grad_norm": 0.1591796875,
+      "learning_rate": 0.02,
+      "loss": 1.4931,
+      "step": 7253
+    },
+    {
+      "epoch": 0.671550078111439,
+      "grad_norm": 0.154296875,
+      "learning_rate": 0.02,
+      "loss": 1.5863,
+      "step": 7254
+    },
+    {
+      "epoch": 0.671642654631719,
+      "grad_norm": 0.1611328125,
+      "learning_rate": 0.02,
+      "loss": 1.5178,
+      "step": 7255
+    },
+    {
+      "epoch": 0.6717352311519991,
+      "grad_norm": 0.142578125,
+      "learning_rate": 0.02,
+      "loss": 1.5436,
+      "step": 7256
+    },
+    {
+      "epoch": 0.6718278076722791,
+      "grad_norm": 0.1484375,
+      "learning_rate": 0.02,
+      "loss": 1.4701,
+      "step": 7257
+    },
+    {
+      "epoch": 0.6719203841925592,
+      "grad_norm": 0.1494140625,
+      "learning_rate": 0.02,
+      "loss": 1.5031,
+      "step": 7258
+    },
+    {
+      "epoch": 0.6720129607128392,
+      "grad_norm": 0.150390625,
+      "learning_rate": 0.02,
+      "loss": 1.5146,
+      "step": 7259
+    },
+    {
+      "epoch": 0.6721055372331193,
+      "grad_norm": 0.154296875,
+      "learning_rate": 0.02,
+      "loss": 1.5435,
+      "step": 7260
+    },
+    {
+      "epoch": 0.6721981137533993,
+      "grad_norm": 0.15234375,
+      "learning_rate": 0.02,
+      "loss": 1.4942,
+      "step": 7261
+    },
+    {
+      "epoch": 0.6722906902736794,
+      "grad_norm": 0.1484375,
+      "learning_rate": 0.02,
+      "loss": 1.4986,
+      "step": 7262
+    },
+    {
+      "epoch": 0.6723832667939594,
+      "grad_norm": 0.1484375,
+      "learning_rate": 0.02,
+      "loss": 1.4925,
+      "step": 7263
+    },
+    {
+      "epoch": 0.6724758433142394,
+      "grad_norm": 0.150390625,
+      "learning_rate": 0.02,
+      "loss": 1.5009,
+      "step": 7264
+    },
+    {
+      "epoch": 0.6725684198345194,
+      "grad_norm": 0.154296875,
+      "learning_rate": 0.02,
+      "loss": 1.5492,
+      "step": 7265
+    },
+    {
+      "epoch": 0.6726609963547995,
+      "grad_norm": 0.1552734375,
+      "learning_rate": 0.02,
+      "loss": 1.529,
+      "step": 7266
+    },
+    {
+      "epoch": 0.6727535728750795,
+      "grad_norm": 0.154296875,
+      "learning_rate": 0.02,
+      "loss": 1.4179,
+      "step": 7267
+    },
+    {
+      "epoch": 0.6728461493953596,
+      "grad_norm": 0.1611328125,
+      "learning_rate": 0.02,
+      "loss": 1.4933,
+      "step": 7268
+    },
+    {
+      "epoch": 0.6729387259156396,
+      "grad_norm": 0.1435546875,
+      "learning_rate": 0.02,
+      "loss": 1.5538,
+      "step": 7269
+    },
+    {
+      "epoch": 0.6730313024359197,
+      "grad_norm": 0.1572265625,
+      "learning_rate": 0.02,
+      "loss": 1.5521,
+      "step": 7270
+    },
+    {
+      "epoch": 0.6731238789561997,
+      "grad_norm": 0.14453125,
+      "learning_rate": 0.02,
+      "loss": 1.5315,
+      "step": 7271
+    },
+    {
+      "epoch": 0.6732164554764798,
+      "grad_norm": 0.1455078125,
+      "learning_rate": 0.02,
+      "loss": 1.5084,
+      "step": 7272
+    },
+    {
+      "epoch": 0.6733090319967598,
+      "grad_norm": 0.1435546875,
+      "learning_rate": 0.02,
+      "loss": 1.5228,
+      "step": 7273
+    },
+    {
+      "epoch": 0.6734016085170399,
+      "grad_norm": 0.1513671875,
+      "learning_rate": 0.02,
+      "loss": 1.5578,
+      "step": 7274
+    },
+    {
+      "epoch": 0.67349418503732,
+      "grad_norm": 0.150390625,
+      "learning_rate": 0.02,
+      "loss": 1.4632,
+      "step": 7275
+    },
+    {
+      "epoch": 0.6735867615576,
+      "grad_norm": 0.154296875,
+      "learning_rate": 0.02,
+      "loss": 1.5971,
+      "step": 7276
+    },
+    {
+      "epoch": 0.67367933807788,
+      "grad_norm": 0.150390625,
+      "learning_rate": 0.02,
+      "loss": 1.5007,
+      "step": 7277
+    },
+    {
+      "epoch": 0.67377191459816,
+      "grad_norm": 0.162109375,
+      "learning_rate": 0.02,
+      "loss": 1.5587,
+      "step": 7278
+    },
+    {
+      "epoch": 0.67386449111844,
+      "grad_norm": 0.1513671875,
+      "learning_rate": 0.02,
+      "loss": 1.4848,
+      "step": 7279
+    },
+    {
+      "epoch": 0.6739570676387201,
+      "grad_norm": 0.1494140625,
+      "learning_rate": 0.02,
+      "loss": 1.4997,
+      "step": 7280
+    },
+    {
+      "epoch": 0.6740496441590001,
+      "grad_norm": 0.158203125,
+      "learning_rate": 0.02,
+      "loss": 1.5154,
+      "step": 7281
+    },
+    {
+      "epoch": 0.6741422206792802,
+      "grad_norm": 0.15625,
+      "learning_rate": 0.02,
+      "loss": 1.5275,
+      "step": 7282
+    },
+    {
+      "epoch": 0.6742347971995603,
+      "grad_norm": 0.1513671875,
+      "learning_rate": 0.02,
+      "loss": 1.5133,
+      "step": 7283
+    },
+    {
+      "epoch": 0.6743273737198403,
+      "grad_norm": 0.162109375,
+      "learning_rate": 0.02,
+      "loss": 1.502,
+      "step": 7284
+    },
+    {
+      "epoch": 0.6744199502401204,
+      "grad_norm": 0.1484375,
+      "learning_rate": 0.02,
+      "loss": 1.5601,
+      "step": 7285
+    },
+    {
+      "epoch": 0.6745125267604004,
+      "grad_norm": 0.1533203125,
+      "learning_rate": 0.02,
+      "loss": 1.5227,
+      "step": 7286
+    },
+    {
+      "epoch": 0.6746051032806805,
+      "grad_norm": 0.1416015625,
+      "learning_rate": 0.02,
+      "loss": 1.4801,
+      "step": 7287
+    },
+    {
+      "epoch": 0.6746976798009605,
+      "grad_norm": 0.1484375,
+      "learning_rate": 0.02,
+      "loss": 1.5126,
+      "step": 7288
+    },
+    {
+      "epoch": 0.6747902563212406,
+      "grad_norm": 0.1474609375,
+      "learning_rate": 0.02,
+      "loss": 1.5353,
+      "step": 7289
+    },
+    {
+      "epoch": 0.6748828328415206,
+      "grad_norm": 0.142578125,
+      "learning_rate": 0.02,
+      "loss": 1.5646,
+      "step": 7290
+    },
+    {
+      "epoch": 0.6749754093618006,
+      "grad_norm": 0.1435546875,
+      "learning_rate": 0.02,
+      "loss": 1.4658,
+      "step": 7291
+    },
+    {
+      "epoch": 0.6750679858820806,
+      "grad_norm": 0.15234375,
+      "learning_rate": 0.02,
+      "loss": 1.5118,
+      "step": 7292
+    },
+    {
+      "epoch": 0.6751605624023607,
+      "grad_norm": 0.146484375,
+      "learning_rate": 0.02,
+      "loss": 1.596,
+      "step": 7293
+    },
+    {
+      "epoch": 0.6752531389226407,
+      "grad_norm": 0.142578125,
+      "learning_rate": 0.02,
+      "loss": 1.5276,
+      "step": 7294
+    },
+    {
+      "epoch": 0.6753457154429208,
+      "grad_norm": 0.150390625,
+      "learning_rate": 0.02,
+      "loss": 1.5659,
+      "step": 7295
+    },
+    {
+      "epoch": 0.6754382919632008,
+      "grad_norm": 0.15234375,
+      "learning_rate": 0.02,
+      "loss": 1.5002,
+      "step": 7296
+    },
+    {
+      "epoch": 0.6755308684834809,
+      "grad_norm": 0.1572265625,
+      "learning_rate": 0.02,
+      "loss": 1.5878,
+      "step": 7297
+    },
+    {
+      "epoch": 0.6756234450037609,
+      "grad_norm": 0.142578125,
+      "learning_rate": 0.02,
+      "loss": 1.5248,
+      "step": 7298
+    },
+    {
+      "epoch": 0.675716021524041,
+      "grad_norm": 0.1552734375,
+      "learning_rate": 0.02,
+      "loss": 1.4757,
+      "step": 7299
+    },
+    {
+      "epoch": 0.675808598044321,
+      "grad_norm": 0.140625,
+      "learning_rate": 0.02,
+      "loss": 1.477,
+      "step": 7300
+    },
+    {
+      "epoch": 0.6759011745646011,
+      "grad_norm": 0.150390625,
+      "learning_rate": 0.02,
+      "loss": 1.5331,
+      "step": 7301
+    },
+    {
+      "epoch": 0.6759937510848811,
+      "grad_norm": 0.146484375,
+      "learning_rate": 0.02,
+      "loss": 1.4789,
+      "step": 7302
+    },
+    {
+      "epoch": 0.6760863276051612,
+      "grad_norm": 0.1533203125,
+      "learning_rate": 0.02,
+      "loss": 1.5093,
+      "step": 7303
+    },
+    {
+      "epoch": 0.6761789041254412,
+      "grad_norm": 0.1435546875,
+      "learning_rate": 0.02,
+      "loss": 1.5175,
+      "step": 7304
+    },
+    {
+      "epoch": 0.6762714806457212,
+      "grad_norm": 0.150390625,
+      "learning_rate": 0.02,
+      "loss": 1.4642,
+      "step": 7305
+    },
+    {
+      "epoch": 0.6763640571660012,
+      "grad_norm": 0.15234375,
+      "learning_rate": 0.02,
+      "loss": 1.5742,
+      "step": 7306
+    },
+    {
+      "epoch": 0.6764566336862813,
+      "grad_norm": 0.15625,
+      "learning_rate": 0.02,
+      "loss": 1.6081,
+      "step": 7307
+    },
+    {
+      "epoch": 0.6765492102065613,
+      "grad_norm": 0.1513671875,
+      "learning_rate": 0.02,
+      "loss": 1.5655,
+      "step": 7308
+    },
+    {
+      "epoch": 0.6766417867268414,
+      "grad_norm": 0.1572265625,
+      "learning_rate": 0.02,
+      "loss": 1.5124,
+      "step": 7309
+    },
+    {
+      "epoch": 0.6767343632471214,
+      "grad_norm": 0.1533203125,
+      "learning_rate": 0.02,
+      "loss": 1.4942,
+      "step": 7310
+    },
+    {
+      "epoch": 0.6768269397674015,
+      "grad_norm": 0.1533203125,
+      "learning_rate": 0.02,
+      "loss": 1.5616,
+      "step": 7311
+    },
+    {
+      "epoch": 0.6769195162876815,
+      "grad_norm": 0.15625,
+      "learning_rate": 0.02,
+      "loss": 1.5361,
+      "step": 7312
+    },
+    {
+      "epoch": 0.6770120928079616,
+      "grad_norm": 0.1484375,
+      "learning_rate": 0.02,
+      "loss": 1.5343,
+      "step": 7313
+    },
+    {
+      "epoch": 0.6771046693282416,
+      "grad_norm": 0.1513671875,
+      "learning_rate": 0.02,
+      "loss": 1.5188,
+      "step": 7314
+    },
+    {
+      "epoch": 0.6771972458485217,
+      "grad_norm": 0.14453125,
+      "learning_rate": 0.02,
+      "loss": 1.56,
+      "step": 7315
+    },
+    {
+      "epoch": 0.6772898223688018,
+      "grad_norm": 0.1650390625,
+      "learning_rate": 0.02,
+      "loss": 1.5331,
+      "step": 7316
+    },
+    {
+      "epoch": 0.6773823988890818,
+      "grad_norm": 0.1416015625,
+      "learning_rate": 0.02,
+      "loss": 1.5047,
+      "step": 7317
+    },
+    {
+      "epoch": 0.6774749754093617,
+      "grad_norm": 0.1474609375,
+      "learning_rate": 0.02,
+      "loss": 1.4803,
+      "step": 7318
+    },
+    {
+      "epoch": 0.6775675519296418,
+      "grad_norm": 0.150390625,
+      "learning_rate": 0.02,
+      "loss": 1.5665,
+      "step": 7319
+    },
+    {
+      "epoch": 0.6776601284499219,
+      "grad_norm": 0.1572265625,
+      "learning_rate": 0.02,
+      "loss": 1.5727,
+      "step": 7320
+    },
+    {
+      "epoch": 0.6777527049702019,
+      "grad_norm": 0.1533203125,
+      "learning_rate": 0.02,
+      "loss": 1.5347,
+      "step": 7321
+    },
+    {
+      "epoch": 0.677845281490482,
+      "grad_norm": 0.15625,
+      "learning_rate": 0.02,
+      "loss": 1.4872,
+      "step": 7322
+    },
+    {
+      "epoch": 0.677937858010762,
+      "grad_norm": 0.1591796875,
+      "learning_rate": 0.02,
+      "loss": 1.5092,
+      "step": 7323
+    },
+    {
+      "epoch": 0.6780304345310421,
+      "grad_norm": 0.1591796875,
+      "learning_rate": 0.02,
+      "loss": 1.4781,
+      "step": 7324
+    },
+    {
+      "epoch": 0.6781230110513221,
+      "grad_norm": 0.1533203125,
+      "learning_rate": 0.02,
+      "loss": 1.5573,
+      "step": 7325
+    },
+    {
+      "epoch": 0.6782155875716022,
+      "grad_norm": 0.1611328125,
+      "learning_rate": 0.02,
+      "loss": 1.4438,
+      "step": 7326
+    },
+    {
+      "epoch": 0.6783081640918822,
+      "grad_norm": 0.162109375,
+      "learning_rate": 0.02,
+      "loss": 1.552,
+      "step": 7327
+    },
+    {
+      "epoch": 0.6784007406121623,
+      "grad_norm": 0.1591796875,
+      "learning_rate": 0.02,
+      "loss": 1.5302,
+      "step": 7328
+    },
+    {
+      "epoch": 0.6784933171324423,
+      "grad_norm": 0.154296875,
+      "learning_rate": 0.02,
+      "loss": 1.5429,
+      "step": 7329
+    },
+    {
+      "epoch": 0.6785858936527224,
+      "grad_norm": 0.146484375,
+      "learning_rate": 0.02,
+      "loss": 1.4792,
+      "step": 7330
+    },
+    {
+      "epoch": 0.6786784701730024,
+      "grad_norm": 0.1552734375,
+      "learning_rate": 0.02,
+      "loss": 1.5632,
+      "step": 7331
+    },
+    {
+      "epoch": 0.6787710466932824,
+      "grad_norm": 0.1533203125,
+      "learning_rate": 0.02,
+      "loss": 1.5327,
+      "step": 7332
+    },
+    {
+      "epoch": 0.6788636232135624,
+      "grad_norm": 0.158203125,
+      "learning_rate": 0.02,
+      "loss": 1.5624,
+      "step": 7333
+    },
+    {
+      "epoch": 0.6789561997338425,
+      "grad_norm": 0.1533203125,
+      "learning_rate": 0.02,
+      "loss": 1.5891,
+      "step": 7334
+    },
+    {
+      "epoch": 0.6790487762541225,
+      "grad_norm": 0.1494140625,
+      "learning_rate": 0.02,
+      "loss": 1.5289,
+      "step": 7335
+    },
+    {
+      "epoch": 0.6791413527744026,
+      "grad_norm": 0.1416015625,
+      "learning_rate": 0.02,
+      "loss": 1.4554,
+      "step": 7336
+    },
+    {
+      "epoch": 0.6792339292946826,
+      "grad_norm": 0.158203125,
+      "learning_rate": 0.02,
+      "loss": 1.525,
+      "step": 7337
+    },
+    {
+      "epoch": 0.6793265058149627,
+      "grad_norm": 0.1552734375,
+      "learning_rate": 0.02,
+      "loss": 1.5914,
+      "step": 7338
+    },
+    {
+      "epoch": 0.6794190823352427,
+      "grad_norm": 0.1533203125,
+      "learning_rate": 0.02,
+      "loss": 1.5697,
+      "step": 7339
+    },
+    {
+      "epoch": 0.6795116588555228,
+      "grad_norm": 0.142578125,
+      "learning_rate": 0.02,
+      "loss": 1.4911,
+      "step": 7340
+    },
+    {
+      "epoch": 0.6796042353758028,
+      "grad_norm": 0.154296875,
+      "learning_rate": 0.02,
+      "loss": 1.5078,
+      "step": 7341
+    },
+    {
+      "epoch": 0.6796968118960829,
+      "grad_norm": 0.1484375,
+      "learning_rate": 0.02,
+      "loss": 1.5713,
+      "step": 7342
+    },
+    {
+      "epoch": 0.6797893884163629,
+      "grad_norm": 0.166015625,
+      "learning_rate": 0.02,
+      "loss": 1.4377,
+      "step": 7343
+    },
+    {
+      "epoch": 0.679881964936643,
+      "grad_norm": 0.1533203125,
+      "learning_rate": 0.02,
+      "loss": 1.4304,
+      "step": 7344
+    },
+    {
+      "epoch": 0.679974541456923,
+      "grad_norm": 0.1611328125,
+      "learning_rate": 0.02,
+      "loss": 1.6018,
+      "step": 7345
+    },
+    {
+      "epoch": 0.680067117977203,
+      "grad_norm": 0.1513671875,
+      "learning_rate": 0.02,
+      "loss": 1.6202,
+      "step": 7346
+    },
+    {
+      "epoch": 0.680159694497483,
+      "grad_norm": 0.16015625,
+      "learning_rate": 0.02,
+      "loss": 1.489,
+      "step": 7347
+    },
+    {
+      "epoch": 0.6802522710177631,
+      "grad_norm": 0.146484375,
+      "learning_rate": 0.02,
+      "loss": 1.5354,
+      "step": 7348
+    },
+    {
+      "epoch": 0.6803448475380431,
+      "grad_norm": 0.1494140625,
+      "learning_rate": 0.02,
+      "loss": 1.4541,
+      "step": 7349
+    },
+    {
+      "epoch": 0.6804374240583232,
+      "grad_norm": 0.150390625,
+      "learning_rate": 0.02,
+      "loss": 1.5599,
+      "step": 7350
+    },
+    {
+      "epoch": 0.6805300005786032,
+      "grad_norm": 0.14453125,
+      "learning_rate": 0.02,
+      "loss": 1.5078,
+      "step": 7351
+    },
+    {
+      "epoch": 0.6806225770988833,
+      "grad_norm": 0.1455078125,
+      "learning_rate": 0.02,
+      "loss": 1.4506,
+      "step": 7352
+    },
+    {
+      "epoch": 0.6807151536191633,
+      "grad_norm": 0.1455078125,
+      "learning_rate": 0.02,
+      "loss": 1.5107,
+      "step": 7353
+    },
+    {
+      "epoch": 0.6808077301394434,
+      "grad_norm": 0.1513671875,
+      "learning_rate": 0.02,
+      "loss": 1.5214,
+      "step": 7354
+    },
+    {
+      "epoch": 0.6809003066597235,
+      "grad_norm": 0.154296875,
+      "learning_rate": 0.02,
+      "loss": 1.4983,
+      "step": 7355
+    },
+    {
+      "epoch": 0.6809928831800035,
+      "grad_norm": 0.1494140625,
+      "learning_rate": 0.02,
+      "loss": 1.4641,
+      "step": 7356
+    },
+    {
+      "epoch": 0.6810854597002836,
+      "grad_norm": 0.142578125,
+      "learning_rate": 0.02,
+      "loss": 1.5502,
+      "step": 7357
+    },
+    {
+      "epoch": 0.6811780362205636,
+      "grad_norm": 0.146484375,
+      "learning_rate": 0.02,
+      "loss": 1.5399,
+      "step": 7358
+    },
+    {
+      "epoch": 0.6812706127408436,
+      "grad_norm": 0.1435546875,
+      "learning_rate": 0.02,
+      "loss": 1.5021,
+      "step": 7359
+    },
+    {
+      "epoch": 0.6813631892611236,
+      "grad_norm": 0.1416015625,
+      "learning_rate": 0.02,
+      "loss": 1.469,
+      "step": 7360
+    },
+    {
+      "epoch": 0.6814557657814037,
+      "grad_norm": 0.14453125,
+      "learning_rate": 0.02,
+      "loss": 1.5576,
+      "step": 7361
+    },
+    {
+      "epoch": 0.6815483423016837,
+      "grad_norm": 0.142578125,
+      "learning_rate": 0.02,
+      "loss": 1.5761,
+      "step": 7362
+    },
+    {
+      "epoch": 0.6816409188219638,
+      "grad_norm": 0.150390625,
+      "learning_rate": 0.02,
+      "loss": 1.488,
+      "step": 7363
+    },
+    {
+      "epoch": 0.6817334953422438,
+      "grad_norm": 0.1533203125,
+      "learning_rate": 0.02,
+      "loss": 1.4215,
+      "step": 7364
+    },
+    {
+      "epoch": 0.6818260718625239,
+      "grad_norm": 0.1533203125,
+      "learning_rate": 0.02,
+      "loss": 1.5353,
+      "step": 7365
+    },
+    {
+      "epoch": 0.6819186483828039,
+      "grad_norm": 0.1533203125,
+      "learning_rate": 0.02,
+      "loss": 1.5073,
+      "step": 7366
+    },
+    {
+      "epoch": 0.682011224903084,
+      "grad_norm": 0.158203125,
+      "learning_rate": 0.02,
+      "loss": 1.5642,
+      "step": 7367
+    },
+    {
+      "epoch": 0.682103801423364,
+      "grad_norm": 0.158203125,
+      "learning_rate": 0.02,
+      "loss": 1.4755,
+      "step": 7368
+    },
+    {
+      "epoch": 0.6821963779436441,
+      "grad_norm": 0.1435546875,
+      "learning_rate": 0.02,
+      "loss": 1.5806,
+      "step": 7369
+    },
+    {
+      "epoch": 0.6822889544639241,
+      "grad_norm": 0.1494140625,
+      "learning_rate": 0.02,
+      "loss": 1.5347,
+      "step": 7370
+    },
+    {
+      "epoch": 0.6823815309842042,
+      "grad_norm": 0.1513671875,
+      "learning_rate": 0.02,
+      "loss": 1.5303,
+      "step": 7371
+    },
+    {
+      "epoch": 0.6824741075044842,
+      "grad_norm": 0.15234375,
+      "learning_rate": 0.02,
+      "loss": 1.5137,
+      "step": 7372
+    },
+    {
+      "epoch": 0.6825666840247642,
+      "grad_norm": 0.15234375,
+      "learning_rate": 0.02,
+      "loss": 1.5442,
+      "step": 7373
+    },
+    {
+      "epoch": 0.6826592605450442,
+      "grad_norm": 0.1455078125,
+      "learning_rate": 0.02,
+      "loss": 1.5166,
+      "step": 7374
+    },
+    {
+      "epoch": 0.6827518370653243,
+      "grad_norm": 0.1552734375,
+      "learning_rate": 0.02,
+      "loss": 1.4808,
+      "step": 7375
+    },
+    {
+      "epoch": 0.6828444135856043,
+      "grad_norm": 0.142578125,
+      "learning_rate": 0.02,
+      "loss": 1.497,
+      "step": 7376
+    },
+    {
+      "epoch": 0.6829369901058844,
+      "grad_norm": 0.1572265625,
+      "learning_rate": 0.02,
+      "loss": 1.4695,
+      "step": 7377
+    },
+    {
+      "epoch": 0.6830295666261644,
+      "grad_norm": 0.1533203125,
+      "learning_rate": 0.02,
+      "loss": 1.521,
+      "step": 7378
+    },
+    {
+      "epoch": 0.6831221431464445,
+      "grad_norm": 0.162109375,
+      "learning_rate": 0.02,
+      "loss": 1.4887,
+      "step": 7379
+    },
+    {
+      "epoch": 0.6832147196667245,
+      "grad_norm": 0.1357421875,
+      "learning_rate": 0.02,
+      "loss": 1.4984,
+      "step": 7380
+    },
+    {
+      "epoch": 0.6833072961870046,
+      "grad_norm": 0.146484375,
+      "learning_rate": 0.02,
+      "loss": 1.5001,
+      "step": 7381
+    },
+    {
+      "epoch": 0.6833998727072846,
+      "grad_norm": 0.146484375,
+      "learning_rate": 0.02,
+      "loss": 1.5183,
+      "step": 7382
+    },
+    {
+      "epoch": 0.6834924492275647,
+      "grad_norm": 0.1494140625,
+      "learning_rate": 0.02,
+      "loss": 1.495,
+      "step": 7383
+    },
+    {
+      "epoch": 0.6835850257478447,
+      "grad_norm": 0.1474609375,
+      "learning_rate": 0.02,
+      "loss": 1.5766,
+      "step": 7384
+    },
+    {
+      "epoch": 0.6836776022681248,
+      "grad_norm": 0.1572265625,
+      "learning_rate": 0.02,
+      "loss": 1.5788,
+      "step": 7385
+    },
+    {
+      "epoch": 0.6837701787884047,
+      "grad_norm": 0.154296875,
+      "learning_rate": 0.02,
+      "loss": 1.5495,
+      "step": 7386
+    },
+    {
+      "epoch": 0.6838627553086848,
+      "grad_norm": 0.1513671875,
+      "learning_rate": 0.02,
+      "loss": 1.4891,
+      "step": 7387
+    },
+    {
+      "epoch": 0.6839553318289648,
+      "grad_norm": 0.1533203125,
+      "learning_rate": 0.02,
+      "loss": 1.5912,
+      "step": 7388
+    },
+    {
+      "epoch": 0.6840479083492449,
+      "grad_norm": 0.1494140625,
+      "learning_rate": 0.02,
+      "loss": 1.5436,
+      "step": 7389
+    },
+    {
+      "epoch": 0.684140484869525,
+      "grad_norm": 0.1611328125,
+      "learning_rate": 0.02,
+      "loss": 1.5307,
+      "step": 7390
+    },
+    {
+      "epoch": 0.684233061389805,
+      "grad_norm": 0.142578125,
+      "learning_rate": 0.02,
+      "loss": 1.5441,
+      "step": 7391
+    },
+    {
+      "epoch": 0.684325637910085,
+      "grad_norm": 0.15625,
+      "learning_rate": 0.02,
+      "loss": 1.5218,
+      "step": 7392
+    },
+    {
+      "epoch": 0.6844182144303651,
+      "grad_norm": 0.1533203125,
+      "learning_rate": 0.02,
+      "loss": 1.6006,
+      "step": 7393
+    },
+    {
+      "epoch": 0.6845107909506452,
+      "grad_norm": 0.16015625,
+      "learning_rate": 0.02,
+      "loss": 1.5419,
+      "step": 7394
+    },
+    {
+      "epoch": 0.6846033674709252,
+      "grad_norm": 0.1494140625,
+      "learning_rate": 0.02,
+      "loss": 1.4775,
+      "step": 7395
+    },
+    {
+      "epoch": 0.6846959439912053,
+      "grad_norm": 0.154296875,
+      "learning_rate": 0.02,
+      "loss": 1.545,
+      "step": 7396
+    },
+    {
+      "epoch": 0.6847885205114853,
+      "grad_norm": 0.150390625,
+      "learning_rate": 0.02,
+      "loss": 1.5628,
+      "step": 7397
+    },
+    {
+      "epoch": 0.6848810970317654,
+      "grad_norm": 0.14453125,
+      "learning_rate": 0.02,
+      "loss": 1.5521,
+      "step": 7398
+    },
+    {
+      "epoch": 0.6849736735520454,
+      "grad_norm": 0.1513671875,
+      "learning_rate": 0.02,
+      "loss": 1.514,
+      "step": 7399
+    },
+    {
+      "epoch": 0.6850662500723254,
+      "grad_norm": 0.15234375,
+      "learning_rate": 0.02,
+      "loss": 1.5556,
+      "step": 7400
+    },
+    {
+      "epoch": 0.6851588265926054,
+      "grad_norm": 0.14453125,
+      "learning_rate": 0.02,
+      "loss": 1.5129,
+      "step": 7401
+    },
+    {
+      "epoch": 0.6852514031128855,
+      "grad_norm": 0.1484375,
+      "learning_rate": 0.02,
+      "loss": 1.5262,
+      "step": 7402
+    },
+    {
+      "epoch": 0.6853439796331655,
+      "grad_norm": 0.1611328125,
+      "learning_rate": 0.02,
+      "loss": 1.5519,
+      "step": 7403
+    },
+    {
+      "epoch": 0.6854365561534456,
+      "grad_norm": 0.16015625,
+      "learning_rate": 0.02,
+      "loss": 1.5364,
+      "step": 7404
+    },
+    {
+      "epoch": 0.6855291326737256,
+      "grad_norm": 0.16015625,
+      "learning_rate": 0.02,
+      "loss": 1.4768,
+      "step": 7405
+    },
+    {
+      "epoch": 0.6856217091940057,
+      "grad_norm": 0.162109375,
+      "learning_rate": 0.02,
+      "loss": 1.5421,
+      "step": 7406
+    },
+    {
+      "epoch": 0.6857142857142857,
+      "grad_norm": 0.1591796875,
+      "learning_rate": 0.02,
+      "loss": 1.5112,
+      "step": 7407
+    },
+    {
+      "epoch": 0.6858068622345658,
+      "grad_norm": 0.15625,
+      "learning_rate": 0.02,
+      "loss": 1.5195,
+      "step": 7408
+    },
+    {
+      "epoch": 0.6858994387548458,
+      "grad_norm": 0.15625,
+      "learning_rate": 0.02,
+      "loss": 1.5407,
+      "step": 7409
+    },
+    {
+      "epoch": 0.6859920152751259,
+      "grad_norm": 0.1728515625,
+      "learning_rate": 0.02,
+      "loss": 1.4589,
+      "step": 7410
+    },
+    {
+      "epoch": 0.6860845917954059,
+      "grad_norm": 0.1455078125,
+      "learning_rate": 0.02,
+      "loss": 1.5148,
+      "step": 7411
+    },
+    {
+      "epoch": 0.686177168315686,
+      "grad_norm": 0.158203125,
+      "learning_rate": 0.02,
+      "loss": 1.4883,
+      "step": 7412
+    },
+    {
+      "epoch": 0.686269744835966,
+      "grad_norm": 0.14453125,
+      "learning_rate": 0.02,
+      "loss": 1.5213,
+      "step": 7413
+    },
+    {
+      "epoch": 0.686362321356246,
+      "grad_norm": 0.1572265625,
+      "learning_rate": 0.02,
+      "loss": 1.5104,
+      "step": 7414
+    },
+    {
+      "epoch": 0.686454897876526,
+      "grad_norm": 0.1494140625,
+      "learning_rate": 0.02,
+      "loss": 1.548,
+      "step": 7415
+    },
+    {
+      "epoch": 0.6865474743968061,
+      "grad_norm": 0.1552734375,
+      "learning_rate": 0.02,
+      "loss": 1.5317,
+      "step": 7416
+    },
+    {
+      "epoch": 0.6866400509170861,
+      "grad_norm": 0.1416015625,
+      "learning_rate": 0.02,
+      "loss": 1.516,
+      "step": 7417
+    },
+    {
+      "epoch": 0.6867326274373662,
+      "grad_norm": 0.146484375,
+      "learning_rate": 0.02,
+      "loss": 1.5046,
+      "step": 7418
+    },
+    {
+      "epoch": 0.6868252039576462,
+      "grad_norm": 0.1484375,
+      "learning_rate": 0.02,
+      "loss": 1.4812,
+      "step": 7419
+    },
+    {
+      "epoch": 0.6869177804779263,
+      "grad_norm": 0.146484375,
+      "learning_rate": 0.02,
+      "loss": 1.5018,
+      "step": 7420
+    },
+    {
+      "epoch": 0.6870103569982063,
+      "grad_norm": 0.142578125,
+      "learning_rate": 0.02,
+      "loss": 1.5047,
+      "step": 7421
+    },
+    {
+      "epoch": 0.6871029335184864,
+      "grad_norm": 0.1552734375,
+      "learning_rate": 0.02,
+      "loss": 1.5227,
+      "step": 7422
+    },
+    {
+      "epoch": 0.6871955100387664,
+      "grad_norm": 0.1474609375,
+      "learning_rate": 0.02,
+      "loss": 1.5506,
+      "step": 7423
+    },
+    {
+      "epoch": 0.6872880865590465,
+      "grad_norm": 0.15234375,
+      "learning_rate": 0.02,
+      "loss": 1.4786,
+      "step": 7424
+    },
+    {
+      "epoch": 0.6873806630793265,
+      "grad_norm": 0.1513671875,
+      "learning_rate": 0.02,
+      "loss": 1.4945,
+      "step": 7425
+    },
+    {
+      "epoch": 0.6874732395996066,
+      "grad_norm": 0.142578125,
+      "learning_rate": 0.02,
+      "loss": 1.4975,
+      "step": 7426
+    },
+    {
+      "epoch": 0.6875658161198865,
+      "grad_norm": 0.150390625,
+      "learning_rate": 0.02,
+      "loss": 1.6003,
+      "step": 7427
+    },
+    {
+      "epoch": 0.6876583926401666,
+      "grad_norm": 0.14453125,
+      "learning_rate": 0.02,
+      "loss": 1.4866,
+      "step": 7428
+    },
+    {
+      "epoch": 0.6877509691604466,
+      "grad_norm": 0.16015625,
+      "learning_rate": 0.02,
+      "loss": 1.4946,
+      "step": 7429
+    },
+    {
+      "epoch": 0.6878435456807267,
+      "grad_norm": 0.1337890625,
+      "learning_rate": 0.02,
+      "loss": 1.501,
+      "step": 7430
+    },
+    {
+      "epoch": 0.6879361222010068,
+      "grad_norm": 0.146484375,
+      "learning_rate": 0.02,
+      "loss": 1.5101,
+      "step": 7431
+    },
+    {
+      "epoch": 0.6880286987212868,
+      "grad_norm": 0.150390625,
+      "learning_rate": 0.02,
+      "loss": 1.4444,
+      "step": 7432
+    },
+    {
+      "epoch": 0.6881212752415669,
+      "grad_norm": 0.1494140625,
+      "learning_rate": 0.02,
+      "loss": 1.5034,
+      "step": 7433
+    },
+    {
+      "epoch": 0.6882138517618469,
+      "grad_norm": 0.146484375,
+      "learning_rate": 0.02,
+      "loss": 1.5182,
+      "step": 7434
+    },
+    {
+      "epoch": 0.688306428282127,
+      "grad_norm": 0.154296875,
+      "learning_rate": 0.02,
+      "loss": 1.6076,
+      "step": 7435
+    },
+    {
+      "epoch": 0.688399004802407,
+      "grad_norm": 0.1494140625,
+      "learning_rate": 0.02,
+      "loss": 1.4451,
+      "step": 7436
+    },
+    {
+      "epoch": 0.6884915813226871,
+      "grad_norm": 0.15234375,
+      "learning_rate": 0.02,
+      "loss": 1.513,
+      "step": 7437
+    },
+    {
+      "epoch": 0.6885841578429671,
+      "grad_norm": 0.15234375,
+      "learning_rate": 0.02,
+      "loss": 1.5914,
+      "step": 7438
+    },
+    {
+      "epoch": 0.6886767343632472,
+      "grad_norm": 0.1630859375,
+      "learning_rate": 0.02,
+      "loss": 1.5765,
+      "step": 7439
+    },
+    {
+      "epoch": 0.6887693108835272,
+      "grad_norm": 0.1357421875,
+      "learning_rate": 0.02,
+      "loss": 1.5209,
+      "step": 7440
+    },
+    {
+      "epoch": 0.6888618874038072,
+      "grad_norm": 0.1435546875,
+      "learning_rate": 0.02,
+      "loss": 1.5498,
+      "step": 7441
+    },
+    {
+      "epoch": 0.6889544639240872,
+      "grad_norm": 0.1494140625,
+      "learning_rate": 0.02,
+      "loss": 1.5108,
+      "step": 7442
+    },
+    {
+      "epoch": 0.6890470404443673,
+      "grad_norm": 0.15625,
+      "learning_rate": 0.02,
+      "loss": 1.5532,
+      "step": 7443
+    },
+    {
+      "epoch": 0.6891396169646473,
+      "grad_norm": 0.138671875,
+      "learning_rate": 0.02,
+      "loss": 1.5476,
+      "step": 7444
+    },
+    {
+      "epoch": 0.6892321934849274,
+      "grad_norm": 0.15625,
+      "learning_rate": 0.02,
+      "loss": 1.5146,
+      "step": 7445
+    },
+    {
+      "epoch": 0.6893247700052074,
+      "grad_norm": 0.130859375,
+      "learning_rate": 0.02,
+      "loss": 1.6135,
+      "step": 7446
+    },
+    {
+      "epoch": 0.6894173465254875,
+      "grad_norm": 0.1484375,
+      "learning_rate": 0.02,
+      "loss": 1.5087,
+      "step": 7447
+    },
+    {
+      "epoch": 0.6895099230457675,
+      "grad_norm": 0.1435546875,
+      "learning_rate": 0.02,
+      "loss": 1.4749,
+      "step": 7448
+    },
+    {
+      "epoch": 0.6896024995660476,
+      "grad_norm": 0.1513671875,
+      "learning_rate": 0.02,
+      "loss": 1.4974,
+      "step": 7449
+    },
+    {
+      "epoch": 0.6896950760863276,
+      "grad_norm": 0.1484375,
+      "learning_rate": 0.02,
+      "loss": 1.4655,
+      "step": 7450
+    },
+    {
+      "epoch": 0.6897876526066077,
+      "grad_norm": 0.1474609375,
+      "learning_rate": 0.02,
+      "loss": 1.5119,
+      "step": 7451
+    },
+    {
+      "epoch": 0.6898802291268877,
+      "grad_norm": 0.150390625,
+      "learning_rate": 0.02,
+      "loss": 1.5145,
+      "step": 7452
+    },
+    {
+      "epoch": 0.6899728056471678,
+      "grad_norm": 0.1357421875,
+      "learning_rate": 0.02,
+      "loss": 1.5053,
+      "step": 7453
+    },
+    {
+      "epoch": 0.6900653821674477,
+      "grad_norm": 0.1552734375,
+      "learning_rate": 0.02,
+      "loss": 1.5263,
+      "step": 7454
+    },
+    {
+      "epoch": 0.6901579586877278,
+      "grad_norm": 0.158203125,
+      "learning_rate": 0.02,
+      "loss": 1.5702,
+      "step": 7455
+    },
+    {
+      "epoch": 0.6902505352080078,
+      "grad_norm": 0.1611328125,
+      "learning_rate": 0.02,
+      "loss": 1.563,
+      "step": 7456
+    },
+    {
+      "epoch": 0.6903431117282879,
+      "grad_norm": 0.1455078125,
+      "learning_rate": 0.02,
+      "loss": 1.4994,
+      "step": 7457
+    },
+    {
+      "epoch": 0.6904356882485679,
+      "grad_norm": 0.1494140625,
+      "learning_rate": 0.02,
+      "loss": 1.4944,
+      "step": 7458
+    },
+    {
+      "epoch": 0.690528264768848,
+      "grad_norm": 0.15234375,
+      "learning_rate": 0.02,
+      "loss": 1.5524,
+      "step": 7459
+    },
+    {
+      "epoch": 0.690620841289128,
+      "grad_norm": 0.158203125,
+      "learning_rate": 0.02,
+      "loss": 1.5919,
+      "step": 7460
+    },
+    {
+      "epoch": 0.6907134178094081,
+      "grad_norm": 0.146484375,
+      "learning_rate": 0.02,
+      "loss": 1.5794,
+      "step": 7461
+    },
+    {
+      "epoch": 0.6908059943296881,
+      "grad_norm": 0.1484375,
+      "learning_rate": 0.02,
+      "loss": 1.5688,
+      "step": 7462
+    },
+    {
+      "epoch": 0.6908985708499682,
+      "grad_norm": 0.1474609375,
+      "learning_rate": 0.02,
+      "loss": 1.4746,
+      "step": 7463
+    },
+    {
+      "epoch": 0.6909911473702482,
+      "grad_norm": 0.1533203125,
+      "learning_rate": 0.02,
+      "loss": 1.518,
+      "step": 7464
+    },
+    {
+      "epoch": 0.6910837238905283,
+      "grad_norm": 0.154296875,
+      "learning_rate": 0.02,
+      "loss": 1.5837,
+      "step": 7465
+    },
+    {
+      "epoch": 0.6911763004108084,
+      "grad_norm": 0.1494140625,
+      "learning_rate": 0.02,
+      "loss": 1.4904,
+      "step": 7466
+    },
+    {
+      "epoch": 0.6912688769310884,
+      "grad_norm": 0.142578125,
+      "learning_rate": 0.02,
+      "loss": 1.4974,
+      "step": 7467
+    },
+    {
+      "epoch": 0.6913614534513683,
+      "grad_norm": 0.150390625,
+      "learning_rate": 0.02,
+      "loss": 1.5821,
+      "step": 7468
+    },
+    {
+      "epoch": 0.6914540299716484,
+      "grad_norm": 0.1455078125,
+      "learning_rate": 0.02,
+      "loss": 1.5613,
+      "step": 7469
+    },
+    {
+      "epoch": 0.6915466064919285,
+      "grad_norm": 0.142578125,
+      "learning_rate": 0.02,
+      "loss": 1.5714,
+      "step": 7470
+    },
+    {
+      "epoch": 0.6916391830122085,
+      "grad_norm": 0.1533203125,
+      "learning_rate": 0.02,
+      "loss": 1.4991,
+      "step": 7471
+    },
+    {
+      "epoch": 0.6917317595324886,
+      "grad_norm": 0.14453125,
+      "learning_rate": 0.02,
+      "loss": 1.5148,
+      "step": 7472
+    },
+    {
+      "epoch": 0.6918243360527686,
+      "grad_norm": 0.1455078125,
+      "learning_rate": 0.02,
+      "loss": 1.5261,
+      "step": 7473
+    },
+    {
+      "epoch": 0.6919169125730487,
+      "grad_norm": 0.1484375,
+      "learning_rate": 0.02,
+      "loss": 1.4809,
+      "step": 7474
+    },
+    {
+      "epoch": 0.6920094890933287,
+      "grad_norm": 0.1513671875,
+      "learning_rate": 0.02,
+      "loss": 1.5217,
+      "step": 7475
+    },
+    {
+      "epoch": 0.6921020656136088,
+      "grad_norm": 0.130859375,
+      "learning_rate": 0.02,
+      "loss": 1.498,
+      "step": 7476
+    },
+    {
+      "epoch": 0.6921946421338888,
+      "grad_norm": 0.1533203125,
+      "learning_rate": 0.02,
+      "loss": 1.5627,
+      "step": 7477
+    },
+    {
+      "epoch": 0.6922872186541689,
+      "grad_norm": 0.1416015625,
+      "learning_rate": 0.02,
+      "loss": 1.4821,
+      "step": 7478
+    },
+    {
+      "epoch": 0.6923797951744489,
+      "grad_norm": 0.15234375,
+      "learning_rate": 0.02,
+      "loss": 1.49,
+      "step": 7479
+    },
+    {
+      "epoch": 0.692472371694729,
+      "grad_norm": 0.150390625,
+      "learning_rate": 0.02,
+      "loss": 1.4513,
+      "step": 7480
+    },
+    {
+      "epoch": 0.6925649482150089,
+      "grad_norm": 0.1611328125,
+      "learning_rate": 0.02,
+      "loss": 1.5825,
+      "step": 7481
+    },
+    {
+      "epoch": 0.692657524735289,
+      "grad_norm": 0.1494140625,
+      "learning_rate": 0.02,
+      "loss": 1.487,
+      "step": 7482
+    },
+    {
+      "epoch": 0.692750101255569,
+      "grad_norm": 0.16796875,
+      "learning_rate": 0.02,
+      "loss": 1.5491,
+      "step": 7483
+    },
+    {
+      "epoch": 0.6928426777758491,
+      "grad_norm": 0.158203125,
+      "learning_rate": 0.02,
+      "loss": 1.5566,
+      "step": 7484
+    },
+    {
+      "epoch": 0.6929352542961291,
+      "grad_norm": 0.150390625,
+      "learning_rate": 0.02,
+      "loss": 1.5274,
+      "step": 7485
+    },
+    {
+      "epoch": 0.6930278308164092,
+      "grad_norm": 0.1591796875,
+      "learning_rate": 0.02,
+      "loss": 1.4931,
+      "step": 7486
+    },
+    {
+      "epoch": 0.6931204073366892,
+      "grad_norm": 0.1416015625,
+      "learning_rate": 0.02,
+      "loss": 1.4059,
+      "step": 7487
+    },
+    {
+      "epoch": 0.6932129838569693,
+      "grad_norm": 0.1396484375,
+      "learning_rate": 0.02,
+      "loss": 1.5341,
+      "step": 7488
+    },
+    {
+      "epoch": 0.6933055603772493,
+      "grad_norm": 0.15234375,
+      "learning_rate": 0.02,
+      "loss": 1.4593,
+      "step": 7489
+    },
+    {
+      "epoch": 0.6933981368975294,
+      "grad_norm": 0.1494140625,
+      "learning_rate": 0.02,
+      "loss": 1.5187,
+      "step": 7490
+    },
+    {
+      "epoch": 0.6934907134178094,
+      "grad_norm": 0.15234375,
+      "learning_rate": 0.02,
+      "loss": 1.4973,
+      "step": 7491
+    },
+    {
+      "epoch": 0.6935832899380895,
+      "grad_norm": 0.166015625,
+      "learning_rate": 0.02,
+      "loss": 1.5263,
+      "step": 7492
+    },
+    {
+      "epoch": 0.6936758664583695,
+      "grad_norm": 0.1650390625,
+      "learning_rate": 0.02,
+      "loss": 1.5353,
+      "step": 7493
+    },
+    {
+      "epoch": 0.6937684429786496,
+      "grad_norm": 0.158203125,
+      "learning_rate": 0.02,
+      "loss": 1.5267,
+      "step": 7494
+    },
+    {
+      "epoch": 0.6938610194989295,
+      "grad_norm": 0.1552734375,
+      "learning_rate": 0.02,
+      "loss": 1.5473,
+      "step": 7495
+    },
+    {
+      "epoch": 0.6939535960192096,
+      "grad_norm": 0.1533203125,
+      "learning_rate": 0.02,
+      "loss": 1.4518,
+      "step": 7496
+    },
+    {
+      "epoch": 0.6940461725394896,
+      "grad_norm": 0.16015625,
+      "learning_rate": 0.02,
+      "loss": 1.5179,
+      "step": 7497
+    },
+    {
+      "epoch": 0.6941387490597697,
+      "grad_norm": 0.1572265625,
+      "learning_rate": 0.02,
+      "loss": 1.5151,
+      "step": 7498
+    },
+    {
+      "epoch": 0.6942313255800497,
+      "grad_norm": 0.154296875,
+      "learning_rate": 0.02,
+      "loss": 1.5052,
+      "step": 7499
+    },
+    {
+      "epoch": 0.6943239021003298,
+      "grad_norm": 0.162109375,
+      "learning_rate": 0.02,
+      "loss": 1.5426,
+      "step": 7500
+    },
+    {
+      "epoch": 0.6944164786206098,
+      "grad_norm": 0.1396484375,
+      "learning_rate": 0.02,
+      "loss": 1.5119,
+      "step": 7501
+    },
+    {
+      "epoch": 0.6945090551408899,
+      "grad_norm": 0.1494140625,
+      "learning_rate": 0.02,
+      "loss": 1.5393,
+      "step": 7502
+    },
+    {
+      "epoch": 0.69460163166117,
+      "grad_norm": 0.142578125,
+      "learning_rate": 0.02,
+      "loss": 1.546,
+      "step": 7503
+    },
+    {
+      "epoch": 0.69469420818145,
+      "grad_norm": 0.15234375,
+      "learning_rate": 0.02,
+      "loss": 1.5237,
+      "step": 7504
+    },
+    {
+      "epoch": 0.69478678470173,
+      "grad_norm": 0.1513671875,
+      "learning_rate": 0.02,
+      "loss": 1.5001,
+      "step": 7505
+    },
+    {
+      "epoch": 0.6948793612220101,
+      "grad_norm": 0.146484375,
+      "learning_rate": 0.02,
+      "loss": 1.4796,
+      "step": 7506
+    },
+    {
+      "epoch": 0.6949719377422902,
+      "grad_norm": 0.1455078125,
+      "learning_rate": 0.02,
+      "loss": 1.5413,
+      "step": 7507
+    },
+    {
+      "epoch": 0.6950645142625702,
+      "grad_norm": 0.1533203125,
+      "learning_rate": 0.02,
+      "loss": 1.5172,
+      "step": 7508
+    },
+    {
+      "epoch": 0.6951570907828502,
+      "grad_norm": 0.1494140625,
+      "learning_rate": 0.02,
+      "loss": 1.4144,
+      "step": 7509
+    },
+    {
+      "epoch": 0.6952496673031302,
+      "grad_norm": 0.154296875,
+      "learning_rate": 0.02,
+      "loss": 1.4986,
+      "step": 7510
+    },
+    {
+      "epoch": 0.6953422438234103,
+      "grad_norm": 0.146484375,
+      "learning_rate": 0.02,
+      "loss": 1.4773,
+      "step": 7511
+    },
+    {
+      "epoch": 0.6954348203436903,
+      "grad_norm": 0.1640625,
+      "learning_rate": 0.02,
+      "loss": 1.5456,
+      "step": 7512
+    },
+    {
+      "epoch": 0.6955273968639704,
+      "grad_norm": 0.1455078125,
+      "learning_rate": 0.02,
+      "loss": 1.5528,
+      "step": 7513
+    },
+    {
+      "epoch": 0.6956199733842504,
+      "grad_norm": 0.15625,
+      "learning_rate": 0.02,
+      "loss": 1.5635,
+      "step": 7514
+    },
+    {
+      "epoch": 0.6957125499045305,
+      "grad_norm": 0.1435546875,
+      "learning_rate": 0.02,
+      "loss": 1.5663,
+      "step": 7515
+    },
+    {
+      "epoch": 0.6958051264248105,
+      "grad_norm": 0.158203125,
+      "learning_rate": 0.02,
+      "loss": 1.5133,
+      "step": 7516
+    },
+    {
+      "epoch": 0.6958977029450906,
+      "grad_norm": 0.1494140625,
+      "learning_rate": 0.02,
+      "loss": 1.4846,
+      "step": 7517
+    },
+    {
+      "epoch": 0.6959902794653706,
+      "grad_norm": 0.1552734375,
+      "learning_rate": 0.02,
+      "loss": 1.4764,
+      "step": 7518
+    },
+    {
+      "epoch": 0.6960828559856507,
+      "grad_norm": 0.1533203125,
+      "learning_rate": 0.02,
+      "loss": 1.5359,
+      "step": 7519
+    },
+    {
+      "epoch": 0.6961754325059307,
+      "grad_norm": 0.169921875,
+      "learning_rate": 0.02,
+      "loss": 1.5553,
+      "step": 7520
+    },
+    {
+      "epoch": 0.6962680090262108,
+      "grad_norm": 0.1357421875,
+      "learning_rate": 0.02,
+      "loss": 1.4681,
+      "step": 7521
+    },
+    {
+      "epoch": 0.6963605855464907,
+      "grad_norm": 0.1533203125,
+      "learning_rate": 0.02,
+      "loss": 1.5074,
+      "step": 7522
+    },
+    {
+      "epoch": 0.6964531620667708,
+      "grad_norm": 0.146484375,
+      "learning_rate": 0.02,
+      "loss": 1.4757,
+      "step": 7523
+    },
+    {
+      "epoch": 0.6965457385870508,
+      "grad_norm": 0.169921875,
+      "learning_rate": 0.02,
+      "loss": 1.5168,
+      "step": 7524
+    },
+    {
+      "epoch": 0.6966383151073309,
+      "grad_norm": 0.1513671875,
+      "learning_rate": 0.02,
+      "loss": 1.5201,
+      "step": 7525
+    },
+    {
+      "epoch": 0.6967308916276109,
+      "grad_norm": 0.1650390625,
+      "learning_rate": 0.02,
+      "loss": 1.5197,
+      "step": 7526
+    },
+    {
+      "epoch": 0.696823468147891,
+      "grad_norm": 0.1455078125,
+      "learning_rate": 0.02,
+      "loss": 1.5281,
+      "step": 7527
+    },
+    {
+      "epoch": 0.696916044668171,
+      "grad_norm": 0.1552734375,
+      "learning_rate": 0.02,
+      "loss": 1.5335,
+      "step": 7528
+    },
+    {
+      "epoch": 0.6970086211884511,
+      "grad_norm": 0.1435546875,
+      "learning_rate": 0.02,
+      "loss": 1.6124,
+      "step": 7529
+    },
+    {
+      "epoch": 0.6971011977087311,
+      "grad_norm": 0.1591796875,
+      "learning_rate": 0.02,
+      "loss": 1.5772,
+      "step": 7530
+    },
+    {
+      "epoch": 0.6971937742290112,
+      "grad_norm": 0.1552734375,
+      "learning_rate": 0.02,
+      "loss": 1.4507,
+      "step": 7531
+    },
+    {
+      "epoch": 0.6972863507492912,
+      "grad_norm": 0.1591796875,
+      "learning_rate": 0.02,
+      "loss": 1.544,
+      "step": 7532
+    },
+    {
+      "epoch": 0.6973789272695713,
+      "grad_norm": 0.1435546875,
+      "learning_rate": 0.02,
+      "loss": 1.5214,
+      "step": 7533
+    },
+    {
+      "epoch": 0.6974715037898513,
+      "grad_norm": 0.1484375,
+      "learning_rate": 0.02,
+      "loss": 1.5598,
+      "step": 7534
+    },
+    {
+      "epoch": 0.6975640803101314,
+      "grad_norm": 0.140625,
+      "learning_rate": 0.02,
+      "loss": 1.5613,
+      "step": 7535
+    },
+    {
+      "epoch": 0.6976566568304113,
+      "grad_norm": 0.1484375,
+      "learning_rate": 0.02,
+      "loss": 1.5836,
+      "step": 7536
+    },
+    {
+      "epoch": 0.6977492333506914,
+      "grad_norm": 0.1435546875,
+      "learning_rate": 0.02,
+      "loss": 1.5146,
+      "step": 7537
+    },
+    {
+      "epoch": 0.6978418098709714,
+      "grad_norm": 0.1513671875,
+      "learning_rate": 0.02,
+      "loss": 1.4865,
+      "step": 7538
+    },
+    {
+      "epoch": 0.6979343863912515,
+      "grad_norm": 0.1513671875,
+      "learning_rate": 0.02,
+      "loss": 1.523,
+      "step": 7539
+    },
+    {
+      "epoch": 0.6980269629115315,
+      "grad_norm": 0.1650390625,
+      "learning_rate": 0.02,
+      "loss": 1.5572,
+      "step": 7540
+    },
+    {
+      "epoch": 0.6981195394318116,
+      "grad_norm": 0.166015625,
+      "learning_rate": 0.02,
+      "loss": 1.5795,
+      "step": 7541
+    },
+    {
+      "epoch": 0.6982121159520916,
+      "grad_norm": 0.1533203125,
+      "learning_rate": 0.02,
+      "loss": 1.5246,
+      "step": 7542
+    },
+    {
+      "epoch": 0.6983046924723717,
+      "grad_norm": 0.1552734375,
+      "learning_rate": 0.02,
+      "loss": 1.5507,
+      "step": 7543
+    },
+    {
+      "epoch": 0.6983972689926518,
+      "grad_norm": 0.1640625,
+      "learning_rate": 0.02,
+      "loss": 1.5741,
+      "step": 7544
+    },
+    {
+      "epoch": 0.6984898455129318,
+      "grad_norm": 0.173828125,
+      "learning_rate": 0.02,
+      "loss": 1.5025,
+      "step": 7545
+    },
+    {
+      "epoch": 0.6985824220332119,
+      "grad_norm": 0.1552734375,
+      "learning_rate": 0.02,
+      "loss": 1.537,
+      "step": 7546
+    },
+    {
+      "epoch": 0.6986749985534919,
+      "grad_norm": 0.1689453125,
+      "learning_rate": 0.02,
+      "loss": 1.5144,
+      "step": 7547
+    },
+    {
+      "epoch": 0.698767575073772,
+      "grad_norm": 0.1513671875,
+      "learning_rate": 0.02,
+      "loss": 1.5189,
+      "step": 7548
+    },
+    {
+      "epoch": 0.6988601515940519,
+      "grad_norm": 0.16015625,
+      "learning_rate": 0.02,
+      "loss": 1.5359,
+      "step": 7549
+    },
+    {
+      "epoch": 0.698952728114332,
+      "grad_norm": 0.1513671875,
+      "learning_rate": 0.02,
+      "loss": 1.5593,
+      "step": 7550
+    },
+    {
+      "epoch": 0.699045304634612,
+      "grad_norm": 0.158203125,
+      "learning_rate": 0.02,
+      "loss": 1.576,
+      "step": 7551
+    },
+    {
+      "epoch": 0.6991378811548921,
+      "grad_norm": 0.158203125,
+      "learning_rate": 0.02,
+      "loss": 1.5345,
+      "step": 7552
+    },
+    {
+      "epoch": 0.6992304576751721,
+      "grad_norm": 0.1650390625,
+      "learning_rate": 0.02,
+      "loss": 1.4853,
+      "step": 7553
+    },
+    {
+      "epoch": 0.6993230341954522,
+      "grad_norm": 0.1650390625,
+      "learning_rate": 0.02,
+      "loss": 1.5315,
+      "step": 7554
+    },
+    {
+      "epoch": 0.6994156107157322,
+      "grad_norm": 0.154296875,
+      "learning_rate": 0.02,
+      "loss": 1.5508,
+      "step": 7555
+    },
+    {
+      "epoch": 0.6995081872360123,
+      "grad_norm": 0.16015625,
+      "learning_rate": 0.02,
+      "loss": 1.4804,
+      "step": 7556
+    },
+    {
+      "epoch": 0.6996007637562923,
+      "grad_norm": 0.1494140625,
+      "learning_rate": 0.02,
+      "loss": 1.4959,
+      "step": 7557
+    },
+    {
+      "epoch": 0.6996933402765724,
+      "grad_norm": 0.1533203125,
+      "learning_rate": 0.02,
+      "loss": 1.5545,
+      "step": 7558
+    },
+    {
+      "epoch": 0.6997859167968524,
+      "grad_norm": 0.1455078125,
+      "learning_rate": 0.02,
+      "loss": 1.4717,
+      "step": 7559
+    },
+    {
+      "epoch": 0.6998784933171325,
+      "grad_norm": 0.1376953125,
+      "learning_rate": 0.02,
+      "loss": 1.5235,
+      "step": 7560
+    },
+    {
+      "epoch": 0.6999710698374125,
+      "grad_norm": 0.150390625,
+      "learning_rate": 0.02,
+      "loss": 1.4385,
+      "step": 7561
+    },
+    {
+      "epoch": 0.7000636463576926,
+      "grad_norm": 0.14453125,
+      "learning_rate": 0.02,
+      "loss": 1.5338,
+      "step": 7562
+    },
+    {
+      "epoch": 0.7001562228779725,
+      "grad_norm": 0.1435546875,
+      "learning_rate": 0.02,
+      "loss": 1.5149,
+      "step": 7563
+    },
+    {
+      "epoch": 0.7002487993982526,
+      "grad_norm": 0.150390625,
+      "learning_rate": 0.02,
+      "loss": 1.4353,
+      "step": 7564
+    },
+    {
+      "epoch": 0.7003413759185326,
+      "grad_norm": 0.1484375,
+      "learning_rate": 0.02,
+      "loss": 1.5843,
+      "step": 7565
+    },
+    {
+      "epoch": 0.7004339524388127,
+      "grad_norm": 0.1396484375,
+      "learning_rate": 0.02,
+      "loss": 1.4983,
+      "step": 7566
+    },
+    {
+      "epoch": 0.7005265289590927,
+      "grad_norm": 0.1513671875,
+      "learning_rate": 0.02,
+      "loss": 1.4845,
+      "step": 7567
+    },
+    {
+      "epoch": 0.7006191054793728,
+      "grad_norm": 0.1376953125,
+      "learning_rate": 0.02,
+      "loss": 1.4864,
+      "step": 7568
+    },
+    {
+      "epoch": 0.7007116819996528,
+      "grad_norm": 0.1533203125,
+      "learning_rate": 0.02,
+      "loss": 1.5612,
+      "step": 7569
+    },
+    {
+      "epoch": 0.7008042585199329,
+      "grad_norm": 0.150390625,
+      "learning_rate": 0.02,
+      "loss": 1.4832,
+      "step": 7570
+    },
+    {
+      "epoch": 0.7008968350402129,
+      "grad_norm": 0.1533203125,
+      "learning_rate": 0.02,
+      "loss": 1.5358,
+      "step": 7571
+    },
+    {
+      "epoch": 0.700989411560493,
+      "grad_norm": 0.1474609375,
+      "learning_rate": 0.02,
+      "loss": 1.5197,
+      "step": 7572
+    },
+    {
+      "epoch": 0.701081988080773,
+      "grad_norm": 0.154296875,
+      "learning_rate": 0.02,
+      "loss": 1.4998,
+      "step": 7573
+    },
+    {
+      "epoch": 0.7011745646010531,
+      "grad_norm": 0.14453125,
+      "learning_rate": 0.02,
+      "loss": 1.4766,
+      "step": 7574
+    },
+    {
+      "epoch": 0.7012671411213331,
+      "grad_norm": 0.15234375,
+      "learning_rate": 0.02,
+      "loss": 1.5925,
+      "step": 7575
+    },
+    {
+      "epoch": 0.7013597176416132,
+      "grad_norm": 0.1416015625,
+      "learning_rate": 0.02,
+      "loss": 1.4765,
+      "step": 7576
+    },
+    {
+      "epoch": 0.7014522941618931,
+      "grad_norm": 0.140625,
+      "learning_rate": 0.02,
+      "loss": 1.518,
+      "step": 7577
+    },
+    {
+      "epoch": 0.7015448706821732,
+      "grad_norm": 0.1494140625,
+      "learning_rate": 0.02,
+      "loss": 1.5397,
+      "step": 7578
+    },
+    {
+      "epoch": 0.7016374472024532,
+      "grad_norm": 0.15234375,
+      "learning_rate": 0.02,
+      "loss": 1.5347,
+      "step": 7579
+    },
+    {
+      "epoch": 0.7017300237227333,
+      "grad_norm": 0.1513671875,
+      "learning_rate": 0.02,
+      "loss": 1.4671,
+      "step": 7580
+    },
+    {
+      "epoch": 0.7018226002430134,
+      "grad_norm": 0.154296875,
+      "learning_rate": 0.02,
+      "loss": 1.5348,
+      "step": 7581
+    },
+    {
+      "epoch": 0.7019151767632934,
+      "grad_norm": 0.14453125,
+      "learning_rate": 0.02,
+      "loss": 1.5384,
+      "step": 7582
+    },
+    {
+      "epoch": 0.7020077532835735,
+      "grad_norm": 0.14453125,
+      "learning_rate": 0.02,
+      "loss": 1.519,
+      "step": 7583
+    },
+    {
+      "epoch": 0.7021003298038535,
+      "grad_norm": 0.1533203125,
+      "learning_rate": 0.02,
+      "loss": 1.5263,
+      "step": 7584
+    },
+    {
+      "epoch": 0.7021929063241336,
+      "grad_norm": 0.1435546875,
+      "learning_rate": 0.02,
+      "loss": 1.5024,
+      "step": 7585
+    },
+    {
+      "epoch": 0.7022854828444136,
+      "grad_norm": 0.15234375,
+      "learning_rate": 0.02,
+      "loss": 1.4819,
+      "step": 7586
+    },
+    {
+      "epoch": 0.7023780593646937,
+      "grad_norm": 0.1513671875,
+      "learning_rate": 0.02,
+      "loss": 1.5073,
+      "step": 7587
+    },
+    {
+      "epoch": 0.7024706358849737,
+      "grad_norm": 0.158203125,
+      "learning_rate": 0.02,
+      "loss": 1.5192,
+      "step": 7588
+    },
+    {
+      "epoch": 0.7025632124052538,
+      "grad_norm": 0.1484375,
+      "learning_rate": 0.02,
+      "loss": 1.5205,
+      "step": 7589
+    },
+    {
+      "epoch": 0.7026557889255337,
+      "grad_norm": 0.1708984375,
+      "learning_rate": 0.02,
+      "loss": 1.5366,
+      "step": 7590
+    },
+    {
+      "epoch": 0.7027483654458138,
+      "grad_norm": 0.1484375,
+      "learning_rate": 0.02,
+      "loss": 1.5637,
+      "step": 7591
+    },
+    {
+      "epoch": 0.7028409419660938,
+      "grad_norm": 0.1572265625,
+      "learning_rate": 0.02,
+      "loss": 1.5897,
+      "step": 7592
+    },
+    {
+      "epoch": 0.7029335184863739,
+      "grad_norm": 0.14453125,
+      "learning_rate": 0.02,
+      "loss": 1.5577,
+      "step": 7593
+    },
+    {
+      "epoch": 0.7030260950066539,
+      "grad_norm": 0.15234375,
+      "learning_rate": 0.02,
+      "loss": 1.5472,
+      "step": 7594
+    },
+    {
+      "epoch": 0.703118671526934,
+      "grad_norm": 0.1533203125,
+      "learning_rate": 0.02,
+      "loss": 1.4924,
+      "step": 7595
+    },
+    {
+      "epoch": 0.703211248047214,
+      "grad_norm": 0.1494140625,
+      "learning_rate": 0.02,
+      "loss": 1.5079,
+      "step": 7596
+    },
+    {
+      "epoch": 0.7033038245674941,
+      "grad_norm": 0.1435546875,
+      "learning_rate": 0.02,
+      "loss": 1.4745,
+      "step": 7597
+    },
+    {
+      "epoch": 0.7033964010877741,
+      "grad_norm": 0.1572265625,
+      "learning_rate": 0.02,
+      "loss": 1.487,
+      "step": 7598
+    },
+    {
+      "epoch": 0.7034889776080542,
+      "grad_norm": 0.1474609375,
+      "learning_rate": 0.02,
+      "loss": 1.5354,
+      "step": 7599
+    },
+    {
+      "epoch": 0.7035815541283342,
+      "grad_norm": 0.15625,
+      "learning_rate": 0.02,
+      "loss": 1.4648,
+      "step": 7600
+    },
+    {
+      "epoch": 0.7036741306486143,
+      "grad_norm": 0.1533203125,
+      "learning_rate": 0.02,
+      "loss": 1.5083,
+      "step": 7601
+    },
+    {
+      "epoch": 0.7037667071688943,
+      "grad_norm": 0.154296875,
+      "learning_rate": 0.02,
+      "loss": 1.5227,
+      "step": 7602
+    },
+    {
+      "epoch": 0.7038592836891744,
+      "grad_norm": 0.1376953125,
+      "learning_rate": 0.02,
+      "loss": 1.5672,
+      "step": 7603
+    },
+    {
+      "epoch": 0.7039518602094543,
+      "grad_norm": 0.1416015625,
+      "learning_rate": 0.02,
+      "loss": 1.4859,
+      "step": 7604
+    },
+    {
+      "epoch": 0.7040444367297344,
+      "grad_norm": 0.138671875,
+      "learning_rate": 0.02,
+      "loss": 1.4818,
+      "step": 7605
+    },
+    {
+      "epoch": 0.7041370132500144,
+      "grad_norm": 0.150390625,
+      "learning_rate": 0.02,
+      "loss": 1.4952,
+      "step": 7606
+    },
+    {
+      "epoch": 0.7042295897702945,
+      "grad_norm": 0.14453125,
+      "learning_rate": 0.02,
+      "loss": 1.5472,
+      "step": 7607
+    },
+    {
+      "epoch": 0.7043221662905745,
+      "grad_norm": 0.1455078125,
+      "learning_rate": 0.02,
+      "loss": 1.5417,
+      "step": 7608
+    },
+    {
+      "epoch": 0.7044147428108546,
+      "grad_norm": 0.14453125,
+      "learning_rate": 0.02,
+      "loss": 1.4839,
+      "step": 7609
+    },
+    {
+      "epoch": 0.7045073193311346,
+      "grad_norm": 0.1435546875,
+      "learning_rate": 0.02,
+      "loss": 1.476,
+      "step": 7610
+    },
+    {
+      "epoch": 0.7045998958514147,
+      "grad_norm": 0.14453125,
+      "learning_rate": 0.02,
+      "loss": 1.5358,
+      "step": 7611
+    },
+    {
+      "epoch": 0.7046924723716947,
+      "grad_norm": 0.142578125,
+      "learning_rate": 0.02,
+      "loss": 1.4664,
+      "step": 7612
+    },
+    {
+      "epoch": 0.7047850488919748,
+      "grad_norm": 0.1611328125,
+      "learning_rate": 0.02,
+      "loss": 1.5275,
+      "step": 7613
+    },
+    {
+      "epoch": 0.7048776254122548,
+      "grad_norm": 0.1484375,
+      "learning_rate": 0.02,
+      "loss": 1.5479,
+      "step": 7614
+    },
+    {
+      "epoch": 0.7049702019325349,
+      "grad_norm": 0.154296875,
+      "learning_rate": 0.02,
+      "loss": 1.4961,
+      "step": 7615
+    },
+    {
+      "epoch": 0.705062778452815,
+      "grad_norm": 0.146484375,
+      "learning_rate": 0.02,
+      "loss": 1.5144,
+      "step": 7616
+    },
+    {
+      "epoch": 0.7051553549730949,
+      "grad_norm": 0.15234375,
+      "learning_rate": 0.02,
+      "loss": 1.5764,
+      "step": 7617
+    },
+    {
+      "epoch": 0.705247931493375,
+      "grad_norm": 0.15625,
+      "learning_rate": 0.02,
+      "loss": 1.5745,
+      "step": 7618
+    },
+    {
+      "epoch": 0.705340508013655,
+      "grad_norm": 0.15234375,
+      "learning_rate": 0.02,
+      "loss": 1.5031,
+      "step": 7619
+    },
+    {
+      "epoch": 0.705433084533935,
+      "grad_norm": 0.1591796875,
+      "learning_rate": 0.02,
+      "loss": 1.5554,
+      "step": 7620
+    },
+    {
+      "epoch": 0.7055256610542151,
+      "grad_norm": 0.1552734375,
+      "learning_rate": 0.02,
+      "loss": 1.4711,
+      "step": 7621
+    },
+    {
+      "epoch": 0.7056182375744952,
+      "grad_norm": 0.15625,
+      "learning_rate": 0.02,
+      "loss": 1.5018,
+      "step": 7622
+    },
+    {
+      "epoch": 0.7057108140947752,
+      "grad_norm": 0.142578125,
+      "learning_rate": 0.02,
+      "loss": 1.5085,
+      "step": 7623
+    },
+    {
+      "epoch": 0.7058033906150553,
+      "grad_norm": 0.15625,
+      "learning_rate": 0.02,
+      "loss": 1.518,
+      "step": 7624
+    },
+    {
+      "epoch": 0.7058959671353353,
+      "grad_norm": 0.1494140625,
+      "learning_rate": 0.02,
+      "loss": 1.5571,
+      "step": 7625
+    },
+    {
+      "epoch": 0.7059885436556154,
+      "grad_norm": 0.14453125,
+      "learning_rate": 0.02,
+      "loss": 1.4966,
+      "step": 7626
+    },
+    {
+      "epoch": 0.7060811201758954,
+      "grad_norm": 0.1513671875,
+      "learning_rate": 0.02,
+      "loss": 1.5498,
+      "step": 7627
+    },
+    {
+      "epoch": 0.7061736966961755,
+      "grad_norm": 0.158203125,
+      "learning_rate": 0.02,
+      "loss": 1.5213,
+      "step": 7628
+    },
+    {
+      "epoch": 0.7062662732164555,
+      "grad_norm": 0.1572265625,
+      "learning_rate": 0.02,
+      "loss": 1.4703,
+      "step": 7629
+    },
+    {
+      "epoch": 0.7063588497367356,
+      "grad_norm": 0.166015625,
+      "learning_rate": 0.02,
+      "loss": 1.5612,
+      "step": 7630
+    },
+    {
+      "epoch": 0.7064514262570155,
+      "grad_norm": 0.158203125,
+      "learning_rate": 0.02,
+      "loss": 1.4863,
+      "step": 7631
+    },
+    {
+      "epoch": 0.7065440027772956,
+      "grad_norm": 0.16015625,
+      "learning_rate": 0.02,
+      "loss": 1.5141,
+      "step": 7632
+    },
+    {
+      "epoch": 0.7066365792975756,
+      "grad_norm": 0.1572265625,
+      "learning_rate": 0.02,
+      "loss": 1.4938,
+      "step": 7633
+    },
+    {
+      "epoch": 0.7067291558178557,
+      "grad_norm": 0.1552734375,
+      "learning_rate": 0.02,
+      "loss": 1.4808,
+      "step": 7634
+    },
+    {
+      "epoch": 0.7068217323381357,
+      "grad_norm": 0.1640625,
+      "learning_rate": 0.02,
+      "loss": 1.4867,
+      "step": 7635
+    },
+    {
+      "epoch": 0.7069143088584158,
+      "grad_norm": 0.1513671875,
+      "learning_rate": 0.02,
+      "loss": 1.5196,
+      "step": 7636
+    },
+    {
+      "epoch": 0.7070068853786958,
+      "grad_norm": 0.162109375,
+      "learning_rate": 0.02,
+      "loss": 1.5439,
+      "step": 7637
+    },
+    {
+      "epoch": 0.7070994618989759,
+      "grad_norm": 0.15234375,
+      "learning_rate": 0.02,
+      "loss": 1.5287,
+      "step": 7638
+    },
+    {
+      "epoch": 0.7071920384192559,
+      "grad_norm": 0.1630859375,
+      "learning_rate": 0.02,
+      "loss": 1.5582,
+      "step": 7639
+    },
+    {
+      "epoch": 0.707284614939536,
+      "grad_norm": 0.154296875,
+      "learning_rate": 0.02,
+      "loss": 1.5241,
+      "step": 7640
+    },
+    {
+      "epoch": 0.707377191459816,
+      "grad_norm": 0.162109375,
+      "learning_rate": 0.02,
+      "loss": 1.4695,
+      "step": 7641
+    },
+    {
+      "epoch": 0.7074697679800961,
+      "grad_norm": 0.1455078125,
+      "learning_rate": 0.02,
+      "loss": 1.5513,
+      "step": 7642
+    },
+    {
+      "epoch": 0.7075623445003761,
+      "grad_norm": 0.158203125,
+      "learning_rate": 0.02,
+      "loss": 1.5265,
+      "step": 7643
+    },
+    {
+      "epoch": 0.7076549210206562,
+      "grad_norm": 0.1552734375,
+      "learning_rate": 0.02,
+      "loss": 1.4906,
+      "step": 7644
+    },
+    {
+      "epoch": 0.7077474975409361,
+      "grad_norm": 0.1640625,
+      "learning_rate": 0.02,
+      "loss": 1.5162,
+      "step": 7645
+    },
+    {
+      "epoch": 0.7078400740612162,
+      "grad_norm": 0.166015625,
+      "learning_rate": 0.02,
+      "loss": 1.5679,
+      "step": 7646
+    },
+    {
+      "epoch": 0.7079326505814962,
+      "grad_norm": 0.16796875,
+      "learning_rate": 0.02,
+      "loss": 1.5079,
+      "step": 7647
+    },
+    {
+      "epoch": 0.7080252271017763,
+      "grad_norm": 0.1669921875,
+      "learning_rate": 0.02,
+      "loss": 1.4981,
+      "step": 7648
+    },
+    {
+      "epoch": 0.7081178036220563,
+      "grad_norm": 0.1484375,
+      "learning_rate": 0.02,
+      "loss": 1.4469,
+      "step": 7649
+    },
+    {
+      "epoch": 0.7082103801423364,
+      "grad_norm": 0.14453125,
+      "learning_rate": 0.02,
+      "loss": 1.5071,
+      "step": 7650
+    },
+    {
+      "epoch": 0.7083029566626164,
+      "grad_norm": 0.146484375,
+      "learning_rate": 0.02,
+      "loss": 1.5151,
+      "step": 7651
+    },
+    {
+      "epoch": 0.7083955331828965,
+      "grad_norm": 0.162109375,
+      "learning_rate": 0.02,
+      "loss": 1.6101,
+      "step": 7652
+    },
+    {
+      "epoch": 0.7084881097031765,
+      "grad_norm": 0.142578125,
+      "learning_rate": 0.02,
+      "loss": 1.4234,
+      "step": 7653
+    },
+    {
+      "epoch": 0.7085806862234566,
+      "grad_norm": 0.1591796875,
+      "learning_rate": 0.02,
+      "loss": 1.5569,
+      "step": 7654
+    },
+    {
+      "epoch": 0.7086732627437367,
+      "grad_norm": 0.1533203125,
+      "learning_rate": 0.02,
+      "loss": 1.4798,
+      "step": 7655
+    },
+    {
+      "epoch": 0.7087658392640167,
+      "grad_norm": 0.158203125,
+      "learning_rate": 0.02,
+      "loss": 1.4985,
+      "step": 7656
+    },
+    {
+      "epoch": 0.7088584157842968,
+      "grad_norm": 0.134765625,
+      "learning_rate": 0.02,
+      "loss": 1.4918,
+      "step": 7657
+    },
+    {
+      "epoch": 0.7089509923045767,
+      "grad_norm": 0.146484375,
+      "learning_rate": 0.02,
+      "loss": 1.528,
+      "step": 7658
+    },
+    {
+      "epoch": 0.7090435688248568,
+      "grad_norm": 0.1494140625,
+      "learning_rate": 0.02,
+      "loss": 1.5206,
+      "step": 7659
+    },
+    {
+      "epoch": 0.7091361453451368,
+      "grad_norm": 0.1484375,
+      "learning_rate": 0.02,
+      "loss": 1.4869,
+      "step": 7660
+    },
+    {
+      "epoch": 0.7092287218654169,
+      "grad_norm": 0.1552734375,
+      "learning_rate": 0.02,
+      "loss": 1.5163,
+      "step": 7661
+    },
+    {
+      "epoch": 0.7093212983856969,
+      "grad_norm": 0.15234375,
+      "learning_rate": 0.02,
+      "loss": 1.474,
+      "step": 7662
+    },
+    {
+      "epoch": 0.709413874905977,
+      "grad_norm": 0.150390625,
+      "learning_rate": 0.02,
+      "loss": 1.6379,
+      "step": 7663
+    },
+    {
+      "epoch": 0.709506451426257,
+      "grad_norm": 0.1552734375,
+      "learning_rate": 0.02,
+      "loss": 1.5027,
+      "step": 7664
+    },
+    {
+      "epoch": 0.7095990279465371,
+      "grad_norm": 0.1494140625,
+      "learning_rate": 0.02,
+      "loss": 1.551,
+      "step": 7665
+    },
+    {
+      "epoch": 0.7096916044668171,
+      "grad_norm": 0.1416015625,
+      "learning_rate": 0.02,
+      "loss": 1.5175,
+      "step": 7666
+    },
+    {
+      "epoch": 0.7097841809870972,
+      "grad_norm": 0.1474609375,
+      "learning_rate": 0.02,
+      "loss": 1.5063,
+      "step": 7667
+    },
+    {
+      "epoch": 0.7098767575073772,
+      "grad_norm": 0.154296875,
+      "learning_rate": 0.02,
+      "loss": 1.5561,
+      "step": 7668
+    },
+    {
+      "epoch": 0.7099693340276573,
+      "grad_norm": 0.16015625,
+      "learning_rate": 0.02,
+      "loss": 1.556,
+      "step": 7669
+    },
+    {
+      "epoch": 0.7100619105479373,
+      "grad_norm": 0.1591796875,
+      "learning_rate": 0.02,
+      "loss": 1.4556,
+      "step": 7670
+    },
+    {
+      "epoch": 0.7101544870682174,
+      "grad_norm": 0.1572265625,
+      "learning_rate": 0.02,
+      "loss": 1.4426,
+      "step": 7671
+    },
+    {
+      "epoch": 0.7102470635884973,
+      "grad_norm": 0.150390625,
+      "learning_rate": 0.02,
+      "loss": 1.5091,
+      "step": 7672
+    },
+    {
+      "epoch": 0.7103396401087774,
+      "grad_norm": 0.1611328125,
+      "learning_rate": 0.02,
+      "loss": 1.5336,
+      "step": 7673
+    },
+    {
+      "epoch": 0.7104322166290574,
+      "grad_norm": 0.1513671875,
+      "learning_rate": 0.02,
+      "loss": 1.5234,
+      "step": 7674
+    },
+    {
+      "epoch": 0.7105247931493375,
+      "grad_norm": 0.15625,
+      "learning_rate": 0.02,
+      "loss": 1.5755,
+      "step": 7675
+    },
+    {
+      "epoch": 0.7106173696696175,
+      "grad_norm": 0.1494140625,
+      "learning_rate": 0.02,
+      "loss": 1.4785,
+      "step": 7676
+    },
+    {
+      "epoch": 0.7107099461898976,
+      "grad_norm": 0.1552734375,
+      "learning_rate": 0.02,
+      "loss": 1.545,
+      "step": 7677
+    },
+    {
+      "epoch": 0.7108025227101776,
+      "grad_norm": 0.1435546875,
+      "learning_rate": 0.02,
+      "loss": 1.5005,
+      "step": 7678
+    },
+    {
+      "epoch": 0.7108950992304577,
+      "grad_norm": 0.15234375,
+      "learning_rate": 0.02,
+      "loss": 1.4959,
+      "step": 7679
+    },
+    {
+      "epoch": 0.7109876757507377,
+      "grad_norm": 0.130859375,
+      "learning_rate": 0.02,
+      "loss": 1.5445,
+      "step": 7680
+    },
+    {
+      "epoch": 0.7110802522710178,
+      "grad_norm": 0.1533203125,
+      "learning_rate": 0.02,
+      "loss": 1.5081,
+      "step": 7681
+    },
+    {
+      "epoch": 0.7111728287912978,
+      "grad_norm": 0.142578125,
+      "learning_rate": 0.02,
+      "loss": 1.5176,
+      "step": 7682
+    },
+    {
+      "epoch": 0.7112654053115779,
+      "grad_norm": 0.1376953125,
+      "learning_rate": 0.02,
+      "loss": 1.532,
+      "step": 7683
+    },
+    {
+      "epoch": 0.7113579818318579,
+      "grad_norm": 0.1357421875,
+      "learning_rate": 0.02,
+      "loss": 1.4641,
+      "step": 7684
+    },
+    {
+      "epoch": 0.7114505583521379,
+      "grad_norm": 0.154296875,
+      "learning_rate": 0.02,
+      "loss": 1.5018,
+      "step": 7685
+    },
+    {
+      "epoch": 0.7115431348724179,
+      "grad_norm": 0.1513671875,
+      "learning_rate": 0.02,
+      "loss": 1.4807,
+      "step": 7686
+    },
+    {
+      "epoch": 0.711635711392698,
+      "grad_norm": 0.1455078125,
+      "learning_rate": 0.02,
+      "loss": 1.5729,
+      "step": 7687
+    },
+    {
+      "epoch": 0.711728287912978,
+      "grad_norm": 0.1474609375,
+      "learning_rate": 0.02,
+      "loss": 1.4545,
+      "step": 7688
+    },
+    {
+      "epoch": 0.7118208644332581,
+      "grad_norm": 0.1494140625,
+      "learning_rate": 0.02,
+      "loss": 1.4703,
+      "step": 7689
+    },
+    {
+      "epoch": 0.7119134409535381,
+      "grad_norm": 0.1435546875,
+      "learning_rate": 0.02,
+      "loss": 1.4842,
+      "step": 7690
+    },
+    {
+      "epoch": 0.7120060174738182,
+      "grad_norm": 0.1494140625,
+      "learning_rate": 0.02,
+      "loss": 1.5716,
+      "step": 7691
+    },
+    {
+      "epoch": 0.7120985939940983,
+      "grad_norm": 0.150390625,
+      "learning_rate": 0.02,
+      "loss": 1.4861,
+      "step": 7692
+    },
+    {
+      "epoch": 0.7121911705143783,
+      "grad_norm": 0.1611328125,
+      "learning_rate": 0.02,
+      "loss": 1.5383,
+      "step": 7693
+    },
+    {
+      "epoch": 0.7122837470346584,
+      "grad_norm": 0.15234375,
+      "learning_rate": 0.02,
+      "loss": 1.5068,
+      "step": 7694
+    },
+    {
+      "epoch": 0.7123763235549384,
+      "grad_norm": 0.150390625,
+      "learning_rate": 0.02,
+      "loss": 1.5329,
+      "step": 7695
+    },
+    {
+      "epoch": 0.7124689000752185,
+      "grad_norm": 0.146484375,
+      "learning_rate": 0.02,
+      "loss": 1.4361,
+      "step": 7696
+    },
+    {
+      "epoch": 0.7125614765954985,
+      "grad_norm": 0.16796875,
+      "learning_rate": 0.02,
+      "loss": 1.4743,
+      "step": 7697
+    },
+    {
+      "epoch": 0.7126540531157786,
+      "grad_norm": 0.1259765625,
+      "learning_rate": 0.02,
+      "loss": 1.5451,
+      "step": 7698
+    },
+    {
+      "epoch": 0.7127466296360585,
+      "grad_norm": 0.1513671875,
+      "learning_rate": 0.02,
+      "loss": 1.5074,
+      "step": 7699
+    },
+    {
+      "epoch": 0.7128392061563386,
+      "grad_norm": 0.146484375,
+      "learning_rate": 0.02,
+      "loss": 1.5297,
+      "step": 7700
+    },
+    {
+      "epoch": 0.7129317826766186,
+      "grad_norm": 0.1455078125,
+      "learning_rate": 0.02,
+      "loss": 1.4687,
+      "step": 7701
+    },
+    {
+      "epoch": 0.7130243591968987,
+      "grad_norm": 0.1474609375,
+      "learning_rate": 0.02,
+      "loss": 1.4661,
+      "step": 7702
+    },
+    {
+      "epoch": 0.7131169357171787,
+      "grad_norm": 0.1455078125,
+      "learning_rate": 0.02,
+      "loss": 1.5149,
+      "step": 7703
+    },
+    {
+      "epoch": 0.7132095122374588,
+      "grad_norm": 0.146484375,
+      "learning_rate": 0.02,
+      "loss": 1.5816,
+      "step": 7704
+    },
+    {
+      "epoch": 0.7133020887577388,
+      "grad_norm": 0.1513671875,
+      "learning_rate": 0.02,
+      "loss": 1.4707,
+      "step": 7705
+    },
+    {
+      "epoch": 0.7133946652780189,
+      "grad_norm": 0.1591796875,
+      "learning_rate": 0.02,
+      "loss": 1.5322,
+      "step": 7706
+    },
+    {
+      "epoch": 0.7134872417982989,
+      "grad_norm": 0.1455078125,
+      "learning_rate": 0.02,
+      "loss": 1.5004,
+      "step": 7707
+    },
+    {
+      "epoch": 0.713579818318579,
+      "grad_norm": 0.1435546875,
+      "learning_rate": 0.02,
+      "loss": 1.4875,
+      "step": 7708
+    },
+    {
+      "epoch": 0.713672394838859,
+      "grad_norm": 0.1435546875,
+      "learning_rate": 0.02,
+      "loss": 1.4713,
+      "step": 7709
+    },
+    {
+      "epoch": 0.7137649713591391,
+      "grad_norm": 0.146484375,
+      "learning_rate": 0.02,
+      "loss": 1.4796,
+      "step": 7710
+    },
+    {
+      "epoch": 0.7138575478794191,
+      "grad_norm": 0.140625,
+      "learning_rate": 0.02,
+      "loss": 1.4667,
+      "step": 7711
+    },
+    {
+      "epoch": 0.7139501243996991,
+      "grad_norm": 0.150390625,
+      "learning_rate": 0.02,
+      "loss": 1.4786,
+      "step": 7712
+    },
+    {
+      "epoch": 0.7140427009199791,
+      "grad_norm": 0.140625,
+      "learning_rate": 0.02,
+      "loss": 1.5322,
+      "step": 7713
+    },
+    {
+      "epoch": 0.7141352774402592,
+      "grad_norm": 0.150390625,
+      "learning_rate": 0.02,
+      "loss": 1.5119,
+      "step": 7714
+    },
+    {
+      "epoch": 0.7142278539605392,
+      "grad_norm": 0.1416015625,
+      "learning_rate": 0.02,
+      "loss": 1.5048,
+      "step": 7715
+    },
+    {
+      "epoch": 0.7143204304808193,
+      "grad_norm": 0.1591796875,
+      "learning_rate": 0.02,
+      "loss": 1.4805,
+      "step": 7716
+    },
+    {
+      "epoch": 0.7144130070010993,
+      "grad_norm": 0.146484375,
+      "learning_rate": 0.02,
+      "loss": 1.5309,
+      "step": 7717
+    },
+    {
+      "epoch": 0.7145055835213794,
+      "grad_norm": 0.15234375,
+      "learning_rate": 0.02,
+      "loss": 1.476,
+      "step": 7718
+    },
+    {
+      "epoch": 0.7145981600416594,
+      "grad_norm": 0.15234375,
+      "learning_rate": 0.02,
+      "loss": 1.4991,
+      "step": 7719
+    },
+    {
+      "epoch": 0.7146907365619395,
+      "grad_norm": 0.1611328125,
+      "learning_rate": 0.02,
+      "loss": 1.593,
+      "step": 7720
+    },
+    {
+      "epoch": 0.7147833130822195,
+      "grad_norm": 0.1513671875,
+      "learning_rate": 0.02,
+      "loss": 1.5004,
+      "step": 7721
+    },
+    {
+      "epoch": 0.7148758896024996,
+      "grad_norm": 0.15625,
+      "learning_rate": 0.02,
+      "loss": 1.4668,
+      "step": 7722
+    },
+    {
+      "epoch": 0.7149684661227796,
+      "grad_norm": 0.1474609375,
+      "learning_rate": 0.02,
+      "loss": 1.5306,
+      "step": 7723
+    },
+    {
+      "epoch": 0.7150610426430597,
+      "grad_norm": 0.16015625,
+      "learning_rate": 0.02,
+      "loss": 1.5046,
+      "step": 7724
+    },
+    {
+      "epoch": 0.7151536191633397,
+      "grad_norm": 0.1513671875,
+      "learning_rate": 0.02,
+      "loss": 1.5265,
+      "step": 7725
+    },
+    {
+      "epoch": 0.7152461956836197,
+      "grad_norm": 0.15234375,
+      "learning_rate": 0.02,
+      "loss": 1.4846,
+      "step": 7726
+    },
+    {
+      "epoch": 0.7153387722038997,
+      "grad_norm": 0.16015625,
+      "learning_rate": 0.02,
+      "loss": 1.5151,
+      "step": 7727
+    },
+    {
+      "epoch": 0.7154313487241798,
+      "grad_norm": 0.1572265625,
+      "learning_rate": 0.02,
+      "loss": 1.608,
+      "step": 7728
+    },
+    {
+      "epoch": 0.7155239252444598,
+      "grad_norm": 0.1650390625,
+      "learning_rate": 0.02,
+      "loss": 1.4923,
+      "step": 7729
+    },
+    {
+      "epoch": 0.7156165017647399,
+      "grad_norm": 0.1630859375,
+      "learning_rate": 0.02,
+      "loss": 1.4585,
+      "step": 7730
+    },
+    {
+      "epoch": 0.71570907828502,
+      "grad_norm": 0.1611328125,
+      "learning_rate": 0.02,
+      "loss": 1.5827,
+      "step": 7731
+    },
+    {
+      "epoch": 0.7158016548053,
+      "grad_norm": 0.1494140625,
+      "learning_rate": 0.02,
+      "loss": 1.4826,
+      "step": 7732
+    },
+    {
+      "epoch": 0.71589423132558,
+      "grad_norm": 0.158203125,
+      "learning_rate": 0.02,
+      "loss": 1.5527,
+      "step": 7733
+    },
+    {
+      "epoch": 0.7159868078458601,
+      "grad_norm": 0.15234375,
+      "learning_rate": 0.02,
+      "loss": 1.5321,
+      "step": 7734
+    },
+    {
+      "epoch": 0.7160793843661402,
+      "grad_norm": 0.142578125,
+      "learning_rate": 0.02,
+      "loss": 1.4897,
+      "step": 7735
+    },
+    {
+      "epoch": 0.7161719608864202,
+      "grad_norm": 0.1435546875,
+      "learning_rate": 0.02,
+      "loss": 1.5067,
+      "step": 7736
+    },
+    {
+      "epoch": 0.7162645374067003,
+      "grad_norm": 0.1533203125,
+      "learning_rate": 0.02,
+      "loss": 1.5434,
+      "step": 7737
+    },
+    {
+      "epoch": 0.7163571139269803,
+      "grad_norm": 0.14453125,
+      "learning_rate": 0.02,
+      "loss": 1.5174,
+      "step": 7738
+    },
+    {
+      "epoch": 0.7164496904472604,
+      "grad_norm": 0.1416015625,
+      "learning_rate": 0.02,
+      "loss": 1.5022,
+      "step": 7739
+    },
+    {
+      "epoch": 0.7165422669675403,
+      "grad_norm": 0.1416015625,
+      "learning_rate": 0.02,
+      "loss": 1.5707,
+      "step": 7740
+    },
+    {
+      "epoch": 0.7166348434878204,
+      "grad_norm": 0.1435546875,
+      "learning_rate": 0.02,
+      "loss": 1.4918,
+      "step": 7741
+    },
+    {
+      "epoch": 0.7167274200081004,
+      "grad_norm": 0.134765625,
+      "learning_rate": 0.02,
+      "loss": 1.4215,
+      "step": 7742
+    },
+    {
+      "epoch": 0.7168199965283805,
+      "grad_norm": 0.14453125,
+      "learning_rate": 0.02,
+      "loss": 1.536,
+      "step": 7743
+    },
+    {
+      "epoch": 0.7169125730486605,
+      "grad_norm": 0.1396484375,
+      "learning_rate": 0.02,
+      "loss": 1.5155,
+      "step": 7744
+    },
+    {
+      "epoch": 0.7170051495689406,
+      "grad_norm": 0.15234375,
+      "learning_rate": 0.02,
+      "loss": 1.5018,
+      "step": 7745
+    },
+    {
+      "epoch": 0.7170977260892206,
+      "grad_norm": 0.154296875,
+      "learning_rate": 0.02,
+      "loss": 1.4814,
+      "step": 7746
+    },
+    {
+      "epoch": 0.7171903026095007,
+      "grad_norm": 0.158203125,
+      "learning_rate": 0.02,
+      "loss": 1.5417,
+      "step": 7747
+    },
+    {
+      "epoch": 0.7172828791297807,
+      "grad_norm": 0.1416015625,
+      "learning_rate": 0.02,
+      "loss": 1.5389,
+      "step": 7748
+    },
+    {
+      "epoch": 0.7173754556500608,
+      "grad_norm": 0.16015625,
+      "learning_rate": 0.02,
+      "loss": 1.5322,
+      "step": 7749
+    },
+    {
+      "epoch": 0.7174680321703408,
+      "grad_norm": 0.154296875,
+      "learning_rate": 0.02,
+      "loss": 1.5215,
+      "step": 7750
+    },
+    {
+      "epoch": 0.7175606086906209,
+      "grad_norm": 0.1591796875,
+      "learning_rate": 0.02,
+      "loss": 1.5235,
+      "step": 7751
+    },
+    {
+      "epoch": 0.7176531852109009,
+      "grad_norm": 0.1630859375,
+      "learning_rate": 0.02,
+      "loss": 1.5468,
+      "step": 7752
+    },
+    {
+      "epoch": 0.7177457617311809,
+      "grad_norm": 0.1650390625,
+      "learning_rate": 0.02,
+      "loss": 1.5513,
+      "step": 7753
+    },
+    {
+      "epoch": 0.7178383382514609,
+      "grad_norm": 0.1494140625,
+      "learning_rate": 0.02,
+      "loss": 1.5472,
+      "step": 7754
+    },
+    {
+      "epoch": 0.717930914771741,
+      "grad_norm": 0.15625,
+      "learning_rate": 0.02,
+      "loss": 1.4885,
+      "step": 7755
+    },
+    {
+      "epoch": 0.718023491292021,
+      "grad_norm": 0.1533203125,
+      "learning_rate": 0.02,
+      "loss": 1.5136,
+      "step": 7756
+    },
+    {
+      "epoch": 0.7181160678123011,
+      "grad_norm": 0.1650390625,
+      "learning_rate": 0.02,
+      "loss": 1.5534,
+      "step": 7757
+    },
+    {
+      "epoch": 0.7182086443325811,
+      "grad_norm": 0.150390625,
+      "learning_rate": 0.02,
+      "loss": 1.4913,
+      "step": 7758
+    },
+    {
+      "epoch": 0.7183012208528612,
+      "grad_norm": 0.150390625,
+      "learning_rate": 0.02,
+      "loss": 1.5235,
+      "step": 7759
+    },
+    {
+      "epoch": 0.7183937973731412,
+      "grad_norm": 0.1455078125,
+      "learning_rate": 0.02,
+      "loss": 1.5198,
+      "step": 7760
+    },
+    {
+      "epoch": 0.7184863738934213,
+      "grad_norm": 0.1533203125,
+      "learning_rate": 0.02,
+      "loss": 1.4641,
+      "step": 7761
+    },
+    {
+      "epoch": 0.7185789504137013,
+      "grad_norm": 0.15234375,
+      "learning_rate": 0.02,
+      "loss": 1.6061,
+      "step": 7762
+    },
+    {
+      "epoch": 0.7186715269339814,
+      "grad_norm": 0.1591796875,
+      "learning_rate": 0.02,
+      "loss": 1.5469,
+      "step": 7763
+    },
+    {
+      "epoch": 0.7187641034542614,
+      "grad_norm": 0.1484375,
+      "learning_rate": 0.02,
+      "loss": 1.545,
+      "step": 7764
+    },
+    {
+      "epoch": 0.7188566799745415,
+      "grad_norm": 0.1611328125,
+      "learning_rate": 0.02,
+      "loss": 1.5707,
+      "step": 7765
+    },
+    {
+      "epoch": 0.7189492564948216,
+      "grad_norm": 0.1474609375,
+      "learning_rate": 0.02,
+      "loss": 1.4845,
+      "step": 7766
+    },
+    {
+      "epoch": 0.7190418330151015,
+      "grad_norm": 0.154296875,
+      "learning_rate": 0.02,
+      "loss": 1.5237,
+      "step": 7767
+    },
+    {
+      "epoch": 0.7191344095353815,
+      "grad_norm": 0.162109375,
+      "learning_rate": 0.02,
+      "loss": 1.4818,
+      "step": 7768
+    },
+    {
+      "epoch": 0.7192269860556616,
+      "grad_norm": 0.15625,
+      "learning_rate": 0.02,
+      "loss": 1.5811,
+      "step": 7769
+    },
+    {
+      "epoch": 0.7193195625759417,
+      "grad_norm": 0.1494140625,
+      "learning_rate": 0.02,
+      "loss": 1.544,
+      "step": 7770
+    },
+    {
+      "epoch": 0.7194121390962217,
+      "grad_norm": 0.15625,
+      "learning_rate": 0.02,
+      "loss": 1.4944,
+      "step": 7771
+    },
+    {
+      "epoch": 0.7195047156165018,
+      "grad_norm": 0.1474609375,
+      "learning_rate": 0.02,
+      "loss": 1.5235,
+      "step": 7772
+    },
+    {
+      "epoch": 0.7195972921367818,
+      "grad_norm": 0.146484375,
+      "learning_rate": 0.02,
+      "loss": 1.4933,
+      "step": 7773
+    },
+    {
+      "epoch": 0.7196898686570619,
+      "grad_norm": 0.1435546875,
+      "learning_rate": 0.02,
+      "loss": 1.5034,
+      "step": 7774
+    },
+    {
+      "epoch": 0.7197824451773419,
+      "grad_norm": 0.1494140625,
+      "learning_rate": 0.02,
+      "loss": 1.4852,
+      "step": 7775
+    },
+    {
+      "epoch": 0.719875021697622,
+      "grad_norm": 0.150390625,
+      "learning_rate": 0.02,
+      "loss": 1.4817,
+      "step": 7776
+    },
+    {
+      "epoch": 0.719967598217902,
+      "grad_norm": 0.1484375,
+      "learning_rate": 0.02,
+      "loss": 1.4551,
+      "step": 7777
+    },
+    {
+      "epoch": 0.7200601747381821,
+      "grad_norm": 0.15625,
+      "learning_rate": 0.02,
+      "loss": 1.5408,
+      "step": 7778
+    },
+    {
+      "epoch": 0.7201527512584621,
+      "grad_norm": 0.1455078125,
+      "learning_rate": 0.02,
+      "loss": 1.5264,
+      "step": 7779
+    },
+    {
+      "epoch": 0.7202453277787421,
+      "grad_norm": 0.158203125,
+      "learning_rate": 0.02,
+      "loss": 1.5561,
+      "step": 7780
+    },
+    {
+      "epoch": 0.7203379042990221,
+      "grad_norm": 0.158203125,
+      "learning_rate": 0.02,
+      "loss": 1.483,
+      "step": 7781
+    },
+    {
+      "epoch": 0.7204304808193022,
+      "grad_norm": 0.1767578125,
+      "learning_rate": 0.02,
+      "loss": 1.5742,
+      "step": 7782
+    },
+    {
+      "epoch": 0.7205230573395822,
+      "grad_norm": 0.150390625,
+      "learning_rate": 0.02,
+      "loss": 1.5362,
+      "step": 7783
+    },
+    {
+      "epoch": 0.7206156338598623,
+      "grad_norm": 0.1484375,
+      "learning_rate": 0.02,
+      "loss": 1.5801,
+      "step": 7784
+    },
+    {
+      "epoch": 0.7207082103801423,
+      "grad_norm": 0.1484375,
+      "learning_rate": 0.02,
+      "loss": 1.4563,
+      "step": 7785
+    },
+    {
+      "epoch": 0.7208007869004224,
+      "grad_norm": 0.1474609375,
+      "learning_rate": 0.02,
+      "loss": 1.507,
+      "step": 7786
+    },
+    {
+      "epoch": 0.7208933634207024,
+      "grad_norm": 0.150390625,
+      "learning_rate": 0.02,
+      "loss": 1.5333,
+      "step": 7787
+    },
+    {
+      "epoch": 0.7209859399409825,
+      "grad_norm": 0.1611328125,
+      "learning_rate": 0.02,
+      "loss": 1.5059,
+      "step": 7788
+    },
+    {
+      "epoch": 0.7210785164612625,
+      "grad_norm": 0.1572265625,
+      "learning_rate": 0.02,
+      "loss": 1.4816,
+      "step": 7789
+    },
+    {
+      "epoch": 0.7211710929815426,
+      "grad_norm": 0.150390625,
+      "learning_rate": 0.02,
+      "loss": 1.5343,
+      "step": 7790
+    },
+    {
+      "epoch": 0.7212636695018226,
+      "grad_norm": 0.1416015625,
+      "learning_rate": 0.02,
+      "loss": 1.53,
+      "step": 7791
+    },
+    {
+      "epoch": 0.7213562460221027,
+      "grad_norm": 0.1435546875,
+      "learning_rate": 0.02,
+      "loss": 1.5379,
+      "step": 7792
+    },
+    {
+      "epoch": 0.7214488225423827,
+      "grad_norm": 0.1494140625,
+      "learning_rate": 0.02,
+      "loss": 1.4633,
+      "step": 7793
+    },
+    {
+      "epoch": 0.7215413990626627,
+      "grad_norm": 0.1484375,
+      "learning_rate": 0.02,
+      "loss": 1.4981,
+      "step": 7794
+    },
+    {
+      "epoch": 0.7216339755829427,
+      "grad_norm": 0.1611328125,
+      "learning_rate": 0.02,
+      "loss": 1.553,
+      "step": 7795
+    },
+    {
+      "epoch": 0.7217265521032228,
+      "grad_norm": 0.1630859375,
+      "learning_rate": 0.02,
+      "loss": 1.4959,
+      "step": 7796
+    },
+    {
+      "epoch": 0.7218191286235028,
+      "grad_norm": 0.1796875,
+      "learning_rate": 0.02,
+      "loss": 1.5682,
+      "step": 7797
+    },
+    {
+      "epoch": 0.7219117051437829,
+      "grad_norm": 0.150390625,
+      "learning_rate": 0.02,
+      "loss": 1.5873,
+      "step": 7798
+    },
+    {
+      "epoch": 0.7220042816640629,
+      "grad_norm": 0.1640625,
+      "learning_rate": 0.02,
+      "loss": 1.5551,
+      "step": 7799
+    },
+    {
+      "epoch": 0.722096858184343,
+      "grad_norm": 0.1494140625,
+      "learning_rate": 0.02,
+      "loss": 1.5049,
+      "step": 7800
+    },
+    {
+      "epoch": 0.722189434704623,
+      "grad_norm": 0.1533203125,
+      "learning_rate": 0.02,
+      "loss": 1.4968,
+      "step": 7801
+    },
+    {
+      "epoch": 0.7222820112249031,
+      "grad_norm": 0.1611328125,
+      "learning_rate": 0.02,
+      "loss": 1.5309,
+      "step": 7802
+    },
+    {
+      "epoch": 0.7223745877451831,
+      "grad_norm": 0.18359375,
+      "learning_rate": 0.02,
+      "loss": 1.5329,
+      "step": 7803
+    },
+    {
+      "epoch": 0.7224671642654632,
+      "grad_norm": 0.1533203125,
+      "learning_rate": 0.02,
+      "loss": 1.5736,
+      "step": 7804
+    },
+    {
+      "epoch": 0.7225597407857433,
+      "grad_norm": 0.166015625,
+      "learning_rate": 0.02,
+      "loss": 1.5255,
+      "step": 7805
+    },
+    {
+      "epoch": 0.7226523173060233,
+      "grad_norm": 0.1484375,
+      "learning_rate": 0.02,
+      "loss": 1.5113,
+      "step": 7806
+    },
+    {
+      "epoch": 0.7227448938263034,
+      "grad_norm": 0.1572265625,
+      "learning_rate": 0.02,
+      "loss": 1.4451,
+      "step": 7807
+    },
+    {
+      "epoch": 0.7228374703465833,
+      "grad_norm": 0.138671875,
+      "learning_rate": 0.02,
+      "loss": 1.4951,
+      "step": 7808
+    },
+    {
+      "epoch": 0.7229300468668634,
+      "grad_norm": 0.150390625,
+      "learning_rate": 0.02,
+      "loss": 1.5084,
+      "step": 7809
+    },
+    {
+      "epoch": 0.7230226233871434,
+      "grad_norm": 0.140625,
+      "learning_rate": 0.02,
+      "loss": 1.5297,
+      "step": 7810
+    },
+    {
+      "epoch": 0.7231151999074235,
+      "grad_norm": 0.14453125,
+      "learning_rate": 0.02,
+      "loss": 1.5034,
+      "step": 7811
+    },
+    {
+      "epoch": 0.7232077764277035,
+      "grad_norm": 0.1455078125,
+      "learning_rate": 0.02,
+      "loss": 1.4859,
+      "step": 7812
+    },
+    {
+      "epoch": 0.7233003529479836,
+      "grad_norm": 0.169921875,
+      "learning_rate": 0.02,
+      "loss": 1.5245,
+      "step": 7813
+    },
+    {
+      "epoch": 0.7233929294682636,
+      "grad_norm": 0.1484375,
+      "learning_rate": 0.02,
+      "loss": 1.5223,
+      "step": 7814
+    },
+    {
+      "epoch": 0.7234855059885437,
+      "grad_norm": 0.162109375,
+      "learning_rate": 0.02,
+      "loss": 1.4869,
+      "step": 7815
+    },
+    {
+      "epoch": 0.7235780825088237,
+      "grad_norm": 0.154296875,
+      "learning_rate": 0.02,
+      "loss": 1.5236,
+      "step": 7816
+    },
+    {
+      "epoch": 0.7236706590291038,
+      "grad_norm": 0.1484375,
+      "learning_rate": 0.02,
+      "loss": 1.4638,
+      "step": 7817
+    },
+    {
+      "epoch": 0.7237632355493838,
+      "grad_norm": 0.1552734375,
+      "learning_rate": 0.02,
+      "loss": 1.5083,
+      "step": 7818
+    },
+    {
+      "epoch": 0.7238558120696639,
+      "grad_norm": 0.142578125,
+      "learning_rate": 0.02,
+      "loss": 1.5798,
+      "step": 7819
+    },
+    {
+      "epoch": 0.7239483885899439,
+      "grad_norm": 0.14453125,
+      "learning_rate": 0.02,
+      "loss": 1.5488,
+      "step": 7820
+    },
+    {
+      "epoch": 0.7240409651102239,
+      "grad_norm": 0.1484375,
+      "learning_rate": 0.02,
+      "loss": 1.4672,
+      "step": 7821
+    },
+    {
+      "epoch": 0.7241335416305039,
+      "grad_norm": 0.1552734375,
+      "learning_rate": 0.02,
+      "loss": 1.5566,
+      "step": 7822
+    },
+    {
+      "epoch": 0.724226118150784,
+      "grad_norm": 0.15625,
+      "learning_rate": 0.02,
+      "loss": 1.5925,
+      "step": 7823
+    },
+    {
+      "epoch": 0.724318694671064,
+      "grad_norm": 0.1552734375,
+      "learning_rate": 0.02,
+      "loss": 1.4139,
+      "step": 7824
+    },
+    {
+      "epoch": 0.7244112711913441,
+      "grad_norm": 0.1474609375,
+      "learning_rate": 0.02,
+      "loss": 1.5333,
+      "step": 7825
+    },
+    {
+      "epoch": 0.7245038477116241,
+      "grad_norm": 0.16015625,
+      "learning_rate": 0.02,
+      "loss": 1.5217,
+      "step": 7826
+    },
+    {
+      "epoch": 0.7245964242319042,
+      "grad_norm": 0.158203125,
+      "learning_rate": 0.02,
+      "loss": 1.4961,
+      "step": 7827
+    },
+    {
+      "epoch": 0.7246890007521842,
+      "grad_norm": 0.162109375,
+      "learning_rate": 0.02,
+      "loss": 1.5876,
+      "step": 7828
+    },
+    {
+      "epoch": 0.7247815772724643,
+      "grad_norm": 0.1484375,
+      "learning_rate": 0.02,
+      "loss": 1.5225,
+      "step": 7829
+    },
+    {
+      "epoch": 0.7248741537927443,
+      "grad_norm": 0.1611328125,
+      "learning_rate": 0.02,
+      "loss": 1.5505,
+      "step": 7830
+    },
+    {
+      "epoch": 0.7249667303130244,
+      "grad_norm": 0.1572265625,
+      "learning_rate": 0.02,
+      "loss": 1.5127,
+      "step": 7831
+    },
+    {
+      "epoch": 0.7250593068333044,
+      "grad_norm": 0.15625,
+      "learning_rate": 0.02,
+      "loss": 1.5529,
+      "step": 7832
+    },
+    {
+      "epoch": 0.7251518833535845,
+      "grad_norm": 0.1474609375,
+      "learning_rate": 0.02,
+      "loss": 1.4527,
+      "step": 7833
+    },
+    {
+      "epoch": 0.7252444598738645,
+      "grad_norm": 0.1552734375,
+      "learning_rate": 0.02,
+      "loss": 1.4986,
+      "step": 7834
+    },
+    {
+      "epoch": 0.7253370363941445,
+      "grad_norm": 0.1669921875,
+      "learning_rate": 0.02,
+      "loss": 1.5328,
+      "step": 7835
+    },
+    {
+      "epoch": 0.7254296129144245,
+      "grad_norm": 0.154296875,
+      "learning_rate": 0.02,
+      "loss": 1.4828,
+      "step": 7836
+    },
+    {
+      "epoch": 0.7255221894347046,
+      "grad_norm": 0.1484375,
+      "learning_rate": 0.02,
+      "loss": 1.5196,
+      "step": 7837
+    },
+    {
+      "epoch": 0.7256147659549846,
+      "grad_norm": 0.1513671875,
+      "learning_rate": 0.02,
+      "loss": 1.4595,
+      "step": 7838
+    },
+    {
+      "epoch": 0.7257073424752647,
+      "grad_norm": 0.1572265625,
+      "learning_rate": 0.02,
+      "loss": 1.5328,
+      "step": 7839
+    },
+    {
+      "epoch": 0.7257999189955447,
+      "grad_norm": 0.166015625,
+      "learning_rate": 0.02,
+      "loss": 1.4363,
+      "step": 7840
+    },
+    {
+      "epoch": 0.7258924955158248,
+      "grad_norm": 0.1513671875,
+      "learning_rate": 0.02,
+      "loss": 1.4657,
+      "step": 7841
+    },
+    {
+      "epoch": 0.7259850720361049,
+      "grad_norm": 0.1533203125,
+      "learning_rate": 0.02,
+      "loss": 1.5563,
+      "step": 7842
+    },
+    {
+      "epoch": 0.7260776485563849,
+      "grad_norm": 0.1416015625,
+      "learning_rate": 0.02,
+      "loss": 1.4653,
+      "step": 7843
+    },
+    {
+      "epoch": 0.726170225076665,
+      "grad_norm": 0.1494140625,
+      "learning_rate": 0.02,
+      "loss": 1.5136,
+      "step": 7844
+    },
+    {
+      "epoch": 0.726262801596945,
+      "grad_norm": 0.1513671875,
+      "learning_rate": 0.02,
+      "loss": 1.5195,
+      "step": 7845
+    },
+    {
+      "epoch": 0.7263553781172251,
+      "grad_norm": 0.1533203125,
+      "learning_rate": 0.02,
+      "loss": 1.4731,
+      "step": 7846
+    },
+    {
+      "epoch": 0.7264479546375051,
+      "grad_norm": 0.1484375,
+      "learning_rate": 0.02,
+      "loss": 1.5116,
+      "step": 7847
+    },
+    {
+      "epoch": 0.726540531157785,
+      "grad_norm": 0.16015625,
+      "learning_rate": 0.02,
+      "loss": 1.5646,
+      "step": 7848
+    },
+    {
+      "epoch": 0.7266331076780651,
+      "grad_norm": 0.1513671875,
+      "learning_rate": 0.02,
+      "loss": 1.416,
+      "step": 7849
+    },
+    {
+      "epoch": 0.7267256841983452,
+      "grad_norm": 0.146484375,
+      "learning_rate": 0.02,
+      "loss": 1.5085,
+      "step": 7850
+    },
+    {
+      "epoch": 0.7268182607186252,
+      "grad_norm": 0.1552734375,
+      "learning_rate": 0.02,
+      "loss": 1.5269,
+      "step": 7851
+    },
+    {
+      "epoch": 0.7269108372389053,
+      "grad_norm": 0.1474609375,
+      "learning_rate": 0.02,
+      "loss": 1.5921,
+      "step": 7852
+    },
+    {
+      "epoch": 0.7270034137591853,
+      "grad_norm": 0.1484375,
+      "learning_rate": 0.02,
+      "loss": 1.495,
+      "step": 7853
+    },
+    {
+      "epoch": 0.7270959902794654,
+      "grad_norm": 0.1513671875,
+      "learning_rate": 0.02,
+      "loss": 1.4774,
+      "step": 7854
+    },
+    {
+      "epoch": 0.7271885667997454,
+      "grad_norm": 0.1484375,
+      "learning_rate": 0.02,
+      "loss": 1.535,
+      "step": 7855
+    },
+    {
+      "epoch": 0.7272811433200255,
+      "grad_norm": 0.1630859375,
+      "learning_rate": 0.02,
+      "loss": 1.5507,
+      "step": 7856
+    },
+    {
+      "epoch": 0.7273737198403055,
+      "grad_norm": 0.1435546875,
+      "learning_rate": 0.02,
+      "loss": 1.4898,
+      "step": 7857
+    },
+    {
+      "epoch": 0.7274662963605856,
+      "grad_norm": 0.158203125,
+      "learning_rate": 0.02,
+      "loss": 1.5132,
+      "step": 7858
+    },
+    {
+      "epoch": 0.7275588728808656,
+      "grad_norm": 0.1572265625,
+      "learning_rate": 0.02,
+      "loss": 1.5128,
+      "step": 7859
+    },
+    {
+      "epoch": 0.7276514494011457,
+      "grad_norm": 0.162109375,
+      "learning_rate": 0.02,
+      "loss": 1.5744,
+      "step": 7860
+    },
+    {
+      "epoch": 0.7277440259214257,
+      "grad_norm": 0.1494140625,
+      "learning_rate": 0.02,
+      "loss": 1.5549,
+      "step": 7861
+    },
+    {
+      "epoch": 0.7278366024417057,
+      "grad_norm": 0.1533203125,
+      "learning_rate": 0.02,
+      "loss": 1.5236,
+      "step": 7862
+    },
+    {
+      "epoch": 0.7279291789619857,
+      "grad_norm": 0.146484375,
+      "learning_rate": 0.02,
+      "loss": 1.5549,
+      "step": 7863
+    },
+    {
+      "epoch": 0.7280217554822658,
+      "grad_norm": 0.140625,
+      "learning_rate": 0.02,
+      "loss": 1.5159,
+      "step": 7864
+    },
+    {
+      "epoch": 0.7281143320025458,
+      "grad_norm": 0.14453125,
+      "learning_rate": 0.02,
+      "loss": 1.4893,
+      "step": 7865
+    },
+    {
+      "epoch": 0.7282069085228259,
+      "grad_norm": 0.15625,
+      "learning_rate": 0.02,
+      "loss": 1.4658,
+      "step": 7866
+    },
+    {
+      "epoch": 0.7282994850431059,
+      "grad_norm": 0.1513671875,
+      "learning_rate": 0.02,
+      "loss": 1.4724,
+      "step": 7867
+    },
+    {
+      "epoch": 0.728392061563386,
+      "grad_norm": 0.1484375,
+      "learning_rate": 0.02,
+      "loss": 1.512,
+      "step": 7868
+    },
+    {
+      "epoch": 0.728484638083666,
+      "grad_norm": 0.1552734375,
+      "learning_rate": 0.02,
+      "loss": 1.5526,
+      "step": 7869
+    },
+    {
+      "epoch": 0.7285772146039461,
+      "grad_norm": 0.146484375,
+      "learning_rate": 0.02,
+      "loss": 1.5615,
+      "step": 7870
+    },
+    {
+      "epoch": 0.7286697911242261,
+      "grad_norm": 0.16015625,
+      "learning_rate": 0.02,
+      "loss": 1.5445,
+      "step": 7871
+    },
+    {
+      "epoch": 0.7287623676445062,
+      "grad_norm": 0.1376953125,
+      "learning_rate": 0.02,
+      "loss": 1.5582,
+      "step": 7872
+    },
+    {
+      "epoch": 0.7288549441647862,
+      "grad_norm": 0.150390625,
+      "learning_rate": 0.02,
+      "loss": 1.4417,
+      "step": 7873
+    },
+    {
+      "epoch": 0.7289475206850663,
+      "grad_norm": 0.14453125,
+      "learning_rate": 0.02,
+      "loss": 1.4782,
+      "step": 7874
+    },
+    {
+      "epoch": 0.7290400972053463,
+      "grad_norm": 0.1484375,
+      "learning_rate": 0.02,
+      "loss": 1.4925,
+      "step": 7875
+    },
+    {
+      "epoch": 0.7291326737256263,
+      "grad_norm": 0.1357421875,
+      "learning_rate": 0.02,
+      "loss": 1.502,
+      "step": 7876
+    },
+    {
+      "epoch": 0.7292252502459063,
+      "grad_norm": 0.1474609375,
+      "learning_rate": 0.02,
+      "loss": 1.4626,
+      "step": 7877
+    },
+    {
+      "epoch": 0.7293178267661864,
+      "grad_norm": 0.1357421875,
+      "learning_rate": 0.02,
+      "loss": 1.4428,
+      "step": 7878
+    },
+    {
+      "epoch": 0.7294104032864664,
+      "grad_norm": 0.1416015625,
+      "learning_rate": 0.02,
+      "loss": 1.5016,
+      "step": 7879
+    },
+    {
+      "epoch": 0.7295029798067465,
+      "grad_norm": 0.1416015625,
+      "learning_rate": 0.02,
+      "loss": 1.5248,
+      "step": 7880
+    },
+    {
+      "epoch": 0.7295955563270266,
+      "grad_norm": 0.1484375,
+      "learning_rate": 0.02,
+      "loss": 1.5355,
+      "step": 7881
+    },
+    {
+      "epoch": 0.7296881328473066,
+      "grad_norm": 0.146484375,
+      "learning_rate": 0.02,
+      "loss": 1.5324,
+      "step": 7882
+    },
+    {
+      "epoch": 0.7297807093675867,
+      "grad_norm": 0.14453125,
+      "learning_rate": 0.02,
+      "loss": 1.516,
+      "step": 7883
+    },
+    {
+      "epoch": 0.7298732858878667,
+      "grad_norm": 0.1494140625,
+      "learning_rate": 0.02,
+      "loss": 1.5075,
+      "step": 7884
+    },
+    {
+      "epoch": 0.7299658624081468,
+      "grad_norm": 0.1494140625,
+      "learning_rate": 0.02,
+      "loss": 1.5009,
+      "step": 7885
+    },
+    {
+      "epoch": 0.7300584389284268,
+      "grad_norm": 0.1494140625,
+      "learning_rate": 0.02,
+      "loss": 1.55,
+      "step": 7886
+    },
+    {
+      "epoch": 0.7301510154487069,
+      "grad_norm": 0.154296875,
+      "learning_rate": 0.02,
+      "loss": 1.5433,
+      "step": 7887
+    },
+    {
+      "epoch": 0.7302435919689869,
+      "grad_norm": 0.1552734375,
+      "learning_rate": 0.02,
+      "loss": 1.5471,
+      "step": 7888
+    },
+    {
+      "epoch": 0.7303361684892669,
+      "grad_norm": 0.1552734375,
+      "learning_rate": 0.02,
+      "loss": 1.5421,
+      "step": 7889
+    },
+    {
+      "epoch": 0.7304287450095469,
+      "grad_norm": 0.15234375,
+      "learning_rate": 0.02,
+      "loss": 1.6102,
+      "step": 7890
+    },
+    {
+      "epoch": 0.730521321529827,
+      "grad_norm": 0.146484375,
+      "learning_rate": 0.02,
+      "loss": 1.496,
+      "step": 7891
+    },
+    {
+      "epoch": 0.730613898050107,
+      "grad_norm": 0.150390625,
+      "learning_rate": 0.02,
+      "loss": 1.5763,
+      "step": 7892
+    },
+    {
+      "epoch": 0.7307064745703871,
+      "grad_norm": 0.14453125,
+      "learning_rate": 0.02,
+      "loss": 1.4969,
+      "step": 7893
+    },
+    {
+      "epoch": 0.7307990510906671,
+      "grad_norm": 0.146484375,
+      "learning_rate": 0.02,
+      "loss": 1.5403,
+      "step": 7894
+    },
+    {
+      "epoch": 0.7308916276109472,
+      "grad_norm": 0.134765625,
+      "learning_rate": 0.02,
+      "loss": 1.4978,
+      "step": 7895
+    },
+    {
+      "epoch": 0.7309842041312272,
+      "grad_norm": 0.154296875,
+      "learning_rate": 0.02,
+      "loss": 1.4424,
+      "step": 7896
+    },
+    {
+      "epoch": 0.7310767806515073,
+      "grad_norm": 0.16796875,
+      "learning_rate": 0.02,
+      "loss": 1.482,
+      "step": 7897
+    },
+    {
+      "epoch": 0.7311693571717873,
+      "grad_norm": 0.1494140625,
+      "learning_rate": 0.02,
+      "loss": 1.5416,
+      "step": 7898
+    },
+    {
+      "epoch": 0.7312619336920674,
+      "grad_norm": 0.16015625,
+      "learning_rate": 0.02,
+      "loss": 1.4786,
+      "step": 7899
+    },
+    {
+      "epoch": 0.7313545102123474,
+      "grad_norm": 0.14453125,
+      "learning_rate": 0.02,
+      "loss": 1.4786,
+      "step": 7900
+    },
+    {
+      "epoch": 0.7314470867326275,
+      "grad_norm": 0.15625,
+      "learning_rate": 0.02,
+      "loss": 1.4553,
+      "step": 7901
+    },
+    {
+      "epoch": 0.7315396632529075,
+      "grad_norm": 0.142578125,
+      "learning_rate": 0.02,
+      "loss": 1.498,
+      "step": 7902
+    },
+    {
+      "epoch": 0.7316322397731875,
+      "grad_norm": 0.1640625,
+      "learning_rate": 0.02,
+      "loss": 1.5535,
+      "step": 7903
+    },
+    {
+      "epoch": 0.7317248162934675,
+      "grad_norm": 0.1396484375,
+      "learning_rate": 0.02,
+      "loss": 1.5517,
+      "step": 7904
+    },
+    {
+      "epoch": 0.7318173928137476,
+      "grad_norm": 0.146484375,
+      "learning_rate": 0.02,
+      "loss": 1.553,
+      "step": 7905
+    },
+    {
+      "epoch": 0.7319099693340276,
+      "grad_norm": 0.1435546875,
+      "learning_rate": 0.02,
+      "loss": 1.5343,
+      "step": 7906
+    },
+    {
+      "epoch": 0.7320025458543077,
+      "grad_norm": 0.1591796875,
+      "learning_rate": 0.02,
+      "loss": 1.5876,
+      "step": 7907
+    },
+    {
+      "epoch": 0.7320951223745877,
+      "grad_norm": 0.150390625,
+      "learning_rate": 0.02,
+      "loss": 1.5066,
+      "step": 7908
+    },
+    {
+      "epoch": 0.7321876988948678,
+      "grad_norm": 0.1533203125,
+      "learning_rate": 0.02,
+      "loss": 1.4519,
+      "step": 7909
+    },
+    {
+      "epoch": 0.7322802754151478,
+      "grad_norm": 0.1484375,
+      "learning_rate": 0.02,
+      "loss": 1.5326,
+      "step": 7910
+    },
+    {
+      "epoch": 0.7323728519354279,
+      "grad_norm": 0.1552734375,
+      "learning_rate": 0.02,
+      "loss": 1.5173,
+      "step": 7911
+    },
+    {
+      "epoch": 0.732465428455708,
+      "grad_norm": 0.15234375,
+      "learning_rate": 0.02,
+      "loss": 1.605,
+      "step": 7912
+    },
+    {
+      "epoch": 0.732558004975988,
+      "grad_norm": 0.15234375,
+      "learning_rate": 0.02,
+      "loss": 1.528,
+      "step": 7913
+    },
+    {
+      "epoch": 0.732650581496268,
+      "grad_norm": 0.1640625,
+      "learning_rate": 0.02,
+      "loss": 1.4924,
+      "step": 7914
+    },
+    {
+      "epoch": 0.7327431580165481,
+      "grad_norm": 0.1455078125,
+      "learning_rate": 0.02,
+      "loss": 1.5685,
+      "step": 7915
+    },
+    {
+      "epoch": 0.732835734536828,
+      "grad_norm": 0.1572265625,
+      "learning_rate": 0.02,
+      "loss": 1.5577,
+      "step": 7916
+    },
+    {
+      "epoch": 0.7329283110571081,
+      "grad_norm": 0.154296875,
+      "learning_rate": 0.02,
+      "loss": 1.5342,
+      "step": 7917
+    },
+    {
+      "epoch": 0.7330208875773881,
+      "grad_norm": 0.1494140625,
+      "learning_rate": 0.02,
+      "loss": 1.4942,
+      "step": 7918
+    },
+    {
+      "epoch": 0.7331134640976682,
+      "grad_norm": 0.15625,
+      "learning_rate": 0.02,
+      "loss": 1.495,
+      "step": 7919
+    },
+    {
+      "epoch": 0.7332060406179483,
+      "grad_norm": 0.158203125,
+      "learning_rate": 0.02,
+      "loss": 1.5058,
+      "step": 7920
+    },
+    {
+      "epoch": 0.7332986171382283,
+      "grad_norm": 0.1552734375,
+      "learning_rate": 0.02,
+      "loss": 1.4517,
+      "step": 7921
+    },
+    {
+      "epoch": 0.7333911936585084,
+      "grad_norm": 0.1591796875,
+      "learning_rate": 0.02,
+      "loss": 1.4969,
+      "step": 7922
+    },
+    {
+      "epoch": 0.7334837701787884,
+      "grad_norm": 0.1533203125,
+      "learning_rate": 0.02,
+      "loss": 1.4831,
+      "step": 7923
+    },
+    {
+      "epoch": 0.7335763466990685,
+      "grad_norm": 0.15625,
+      "learning_rate": 0.02,
+      "loss": 1.4828,
+      "step": 7924
+    },
+    {
+      "epoch": 0.7336689232193485,
+      "grad_norm": 0.154296875,
+      "learning_rate": 0.02,
+      "loss": 1.498,
+      "step": 7925
+    },
+    {
+      "epoch": 0.7337614997396286,
+      "grad_norm": 0.1669921875,
+      "learning_rate": 0.02,
+      "loss": 1.5567,
+      "step": 7926
+    },
+    {
+      "epoch": 0.7338540762599086,
+      "grad_norm": 0.1552734375,
+      "learning_rate": 0.02,
+      "loss": 1.4741,
+      "step": 7927
+    },
+    {
+      "epoch": 0.7339466527801887,
+      "grad_norm": 0.1748046875,
+      "learning_rate": 0.02,
+      "loss": 1.5315,
+      "step": 7928
+    },
+    {
+      "epoch": 0.7340392293004687,
+      "grad_norm": 0.1416015625,
+      "learning_rate": 0.02,
+      "loss": 1.5776,
+      "step": 7929
+    },
+    {
+      "epoch": 0.7341318058207487,
+      "grad_norm": 0.1494140625,
+      "learning_rate": 0.02,
+      "loss": 1.5339,
+      "step": 7930
+    },
+    {
+      "epoch": 0.7342243823410287,
+      "grad_norm": 0.146484375,
+      "learning_rate": 0.02,
+      "loss": 1.5447,
+      "step": 7931
+    },
+    {
+      "epoch": 0.7343169588613088,
+      "grad_norm": 0.158203125,
+      "learning_rate": 0.02,
+      "loss": 1.5062,
+      "step": 7932
+    },
+    {
+      "epoch": 0.7344095353815888,
+      "grad_norm": 0.140625,
+      "learning_rate": 0.02,
+      "loss": 1.5496,
+      "step": 7933
+    },
+    {
+      "epoch": 0.7345021119018689,
+      "grad_norm": 0.1689453125,
+      "learning_rate": 0.02,
+      "loss": 1.618,
+      "step": 7934
+    },
+    {
+      "epoch": 0.7345946884221489,
+      "grad_norm": 0.1533203125,
+      "learning_rate": 0.02,
+      "loss": 1.4796,
+      "step": 7935
+    },
+    {
+      "epoch": 0.734687264942429,
+      "grad_norm": 0.1650390625,
+      "learning_rate": 0.02,
+      "loss": 1.494,
+      "step": 7936
+    },
+    {
+      "epoch": 0.734779841462709,
+      "grad_norm": 0.1591796875,
+      "learning_rate": 0.02,
+      "loss": 1.5211,
+      "step": 7937
+    },
+    {
+      "epoch": 0.7348724179829891,
+      "grad_norm": 0.162109375,
+      "learning_rate": 0.02,
+      "loss": 1.4557,
+      "step": 7938
+    },
+    {
+      "epoch": 0.7349649945032691,
+      "grad_norm": 0.150390625,
+      "learning_rate": 0.02,
+      "loss": 1.6116,
+      "step": 7939
+    },
+    {
+      "epoch": 0.7350575710235492,
+      "grad_norm": 0.1630859375,
+      "learning_rate": 0.02,
+      "loss": 1.5292,
+      "step": 7940
+    },
+    {
+      "epoch": 0.7351501475438292,
+      "grad_norm": 0.1357421875,
+      "learning_rate": 0.02,
+      "loss": 1.5363,
+      "step": 7941
+    },
+    {
+      "epoch": 0.7352427240641093,
+      "grad_norm": 0.166015625,
+      "learning_rate": 0.02,
+      "loss": 1.555,
+      "step": 7942
+    },
+    {
+      "epoch": 0.7353353005843893,
+      "grad_norm": 0.1484375,
+      "learning_rate": 0.02,
+      "loss": 1.4862,
+      "step": 7943
+    },
+    {
+      "epoch": 0.7354278771046693,
+      "grad_norm": 0.1533203125,
+      "learning_rate": 0.02,
+      "loss": 1.479,
+      "step": 7944
+    },
+    {
+      "epoch": 0.7355204536249493,
+      "grad_norm": 0.1396484375,
+      "learning_rate": 0.02,
+      "loss": 1.5242,
+      "step": 7945
+    },
+    {
+      "epoch": 0.7356130301452294,
+      "grad_norm": 0.1572265625,
+      "learning_rate": 0.02,
+      "loss": 1.5565,
+      "step": 7946
+    },
+    {
+      "epoch": 0.7357056066655094,
+      "grad_norm": 0.1552734375,
+      "learning_rate": 0.02,
+      "loss": 1.5733,
+      "step": 7947
+    },
+    {
+      "epoch": 0.7357981831857895,
+      "grad_norm": 0.1513671875,
+      "learning_rate": 0.02,
+      "loss": 1.5185,
+      "step": 7948
+    },
+    {
+      "epoch": 0.7358907597060695,
+      "grad_norm": 0.1416015625,
+      "learning_rate": 0.02,
+      "loss": 1.5685,
+      "step": 7949
+    },
+    {
+      "epoch": 0.7359833362263496,
+      "grad_norm": 0.146484375,
+      "learning_rate": 0.02,
+      "loss": 1.543,
+      "step": 7950
+    },
+    {
+      "epoch": 0.7360759127466296,
+      "grad_norm": 0.1494140625,
+      "learning_rate": 0.02,
+      "loss": 1.4818,
+      "step": 7951
+    },
+    {
+      "epoch": 0.7361684892669097,
+      "grad_norm": 0.138671875,
+      "learning_rate": 0.02,
+      "loss": 1.4629,
+      "step": 7952
+    },
+    {
+      "epoch": 0.7362610657871898,
+      "grad_norm": 0.154296875,
+      "learning_rate": 0.02,
+      "loss": 1.5001,
+      "step": 7953
+    },
+    {
+      "epoch": 0.7363536423074698,
+      "grad_norm": 0.1396484375,
+      "learning_rate": 0.02,
+      "loss": 1.5416,
+      "step": 7954
+    },
+    {
+      "epoch": 0.7364462188277499,
+      "grad_norm": 0.1494140625,
+      "learning_rate": 0.02,
+      "loss": 1.4574,
+      "step": 7955
+    },
+    {
+      "epoch": 0.7365387953480299,
+      "grad_norm": 0.1474609375,
+      "learning_rate": 0.02,
+      "loss": 1.517,
+      "step": 7956
+    },
+    {
+      "epoch": 0.7366313718683098,
+      "grad_norm": 0.1513671875,
+      "learning_rate": 0.02,
+      "loss": 1.4874,
+      "step": 7957
+    },
+    {
+      "epoch": 0.7367239483885899,
+      "grad_norm": 0.15234375,
+      "learning_rate": 0.02,
+      "loss": 1.5042,
+      "step": 7958
+    },
+    {
+      "epoch": 0.73681652490887,
+      "grad_norm": 0.15234375,
+      "learning_rate": 0.02,
+      "loss": 1.5283,
+      "step": 7959
+    },
+    {
+      "epoch": 0.73690910142915,
+      "grad_norm": 0.1552734375,
+      "learning_rate": 0.02,
+      "loss": 1.5311,
+      "step": 7960
+    },
+    {
+      "epoch": 0.7370016779494301,
+      "grad_norm": 0.150390625,
+      "learning_rate": 0.02,
+      "loss": 1.6066,
+      "step": 7961
+    },
+    {
+      "epoch": 0.7370942544697101,
+      "grad_norm": 0.15625,
+      "learning_rate": 0.02,
+      "loss": 1.501,
+      "step": 7962
+    },
+    {
+      "epoch": 0.7371868309899902,
+      "grad_norm": 0.146484375,
+      "learning_rate": 0.02,
+      "loss": 1.555,
+      "step": 7963
+    },
+    {
+      "epoch": 0.7372794075102702,
+      "grad_norm": 0.154296875,
+      "learning_rate": 0.02,
+      "loss": 1.5314,
+      "step": 7964
+    },
+    {
+      "epoch": 0.7373719840305503,
+      "grad_norm": 0.142578125,
+      "learning_rate": 0.02,
+      "loss": 1.4733,
+      "step": 7965
+    },
+    {
+      "epoch": 0.7374645605508303,
+      "grad_norm": 0.142578125,
+      "learning_rate": 0.02,
+      "loss": 1.5177,
+      "step": 7966
+    },
+    {
+      "epoch": 0.7375571370711104,
+      "grad_norm": 0.146484375,
+      "learning_rate": 0.02,
+      "loss": 1.5421,
+      "step": 7967
+    },
+    {
+      "epoch": 0.7376497135913904,
+      "grad_norm": 0.1474609375,
+      "learning_rate": 0.02,
+      "loss": 1.5286,
+      "step": 7968
+    },
+    {
+      "epoch": 0.7377422901116705,
+      "grad_norm": 0.1484375,
+      "learning_rate": 0.02,
+      "loss": 1.576,
+      "step": 7969
+    },
+    {
+      "epoch": 0.7378348666319505,
+      "grad_norm": 0.166015625,
+      "learning_rate": 0.02,
+      "loss": 1.4917,
+      "step": 7970
+    },
+    {
+      "epoch": 0.7379274431522305,
+      "grad_norm": 0.1552734375,
+      "learning_rate": 0.02,
+      "loss": 1.6344,
+      "step": 7971
+    },
+    {
+      "epoch": 0.7380200196725105,
+      "grad_norm": 0.16015625,
+      "learning_rate": 0.02,
+      "loss": 1.5245,
+      "step": 7972
+    },
+    {
+      "epoch": 0.7381125961927906,
+      "grad_norm": 0.146484375,
+      "learning_rate": 0.02,
+      "loss": 1.5145,
+      "step": 7973
+    },
+    {
+      "epoch": 0.7382051727130706,
+      "grad_norm": 0.1494140625,
+      "learning_rate": 0.02,
+      "loss": 1.5032,
+      "step": 7974
+    },
+    {
+      "epoch": 0.7382977492333507,
+      "grad_norm": 0.1474609375,
+      "learning_rate": 0.02,
+      "loss": 1.4666,
+      "step": 7975
+    },
+    {
+      "epoch": 0.7383903257536307,
+      "grad_norm": 0.1494140625,
+      "learning_rate": 0.02,
+      "loss": 1.5164,
+      "step": 7976
+    },
+    {
+      "epoch": 0.7384829022739108,
+      "grad_norm": 0.1552734375,
+      "learning_rate": 0.02,
+      "loss": 1.585,
+      "step": 7977
+    },
+    {
+      "epoch": 0.7385754787941908,
+      "grad_norm": 0.15234375,
+      "learning_rate": 0.02,
+      "loss": 1.5353,
+      "step": 7978
+    },
+    {
+      "epoch": 0.7386680553144709,
+      "grad_norm": 0.1474609375,
+      "learning_rate": 0.02,
+      "loss": 1.4924,
+      "step": 7979
+    },
+    {
+      "epoch": 0.7387606318347509,
+      "grad_norm": 0.1572265625,
+      "learning_rate": 0.02,
+      "loss": 1.5184,
+      "step": 7980
+    },
+    {
+      "epoch": 0.738853208355031,
+      "grad_norm": 0.1494140625,
+      "learning_rate": 0.02,
+      "loss": 1.5061,
+      "step": 7981
+    },
+    {
+      "epoch": 0.738945784875311,
+      "grad_norm": 0.1611328125,
+      "learning_rate": 0.02,
+      "loss": 1.5123,
+      "step": 7982
+    },
+    {
+      "epoch": 0.7390383613955911,
+      "grad_norm": 0.1484375,
+      "learning_rate": 0.02,
+      "loss": 1.5293,
+      "step": 7983
+    },
+    {
+      "epoch": 0.739130937915871,
+      "grad_norm": 0.1640625,
+      "learning_rate": 0.02,
+      "loss": 1.4974,
+      "step": 7984
+    },
+    {
+      "epoch": 0.7392235144361511,
+      "grad_norm": 0.1513671875,
+      "learning_rate": 0.02,
+      "loss": 1.5402,
+      "step": 7985
+    },
+    {
+      "epoch": 0.7393160909564311,
+      "grad_norm": 0.1572265625,
+      "learning_rate": 0.02,
+      "loss": 1.5311,
+      "step": 7986
+    },
+    {
+      "epoch": 0.7394086674767112,
+      "grad_norm": 0.142578125,
+      "learning_rate": 0.02,
+      "loss": 1.4867,
+      "step": 7987
+    },
+    {
+      "epoch": 0.7395012439969912,
+      "grad_norm": 0.15625,
+      "learning_rate": 0.02,
+      "loss": 1.5648,
+      "step": 7988
+    },
+    {
+      "epoch": 0.7395938205172713,
+      "grad_norm": 0.1494140625,
+      "learning_rate": 0.02,
+      "loss": 1.5031,
+      "step": 7989
+    },
+    {
+      "epoch": 0.7396863970375513,
+      "grad_norm": 0.1474609375,
+      "learning_rate": 0.02,
+      "loss": 1.4394,
+      "step": 7990
+    },
+    {
+      "epoch": 0.7397789735578314,
+      "grad_norm": 0.1435546875,
+      "learning_rate": 0.02,
+      "loss": 1.6166,
+      "step": 7991
+    },
+    {
+      "epoch": 0.7398715500781115,
+      "grad_norm": 0.150390625,
+      "learning_rate": 0.02,
+      "loss": 1.5365,
+      "step": 7992
+    },
+    {
+      "epoch": 0.7399641265983915,
+      "grad_norm": 0.1455078125,
+      "learning_rate": 0.02,
+      "loss": 1.494,
+      "step": 7993
+    },
+    {
+      "epoch": 0.7400567031186716,
+      "grad_norm": 0.1533203125,
+      "learning_rate": 0.02,
+      "loss": 1.5367,
+      "step": 7994
+    },
+    {
+      "epoch": 0.7401492796389516,
+      "grad_norm": 0.142578125,
+      "learning_rate": 0.02,
+      "loss": 1.515,
+      "step": 7995
+    },
+    {
+      "epoch": 0.7402418561592317,
+      "grad_norm": 0.154296875,
+      "learning_rate": 0.02,
+      "loss": 1.5471,
+      "step": 7996
+    },
+    {
+      "epoch": 0.7403344326795117,
+      "grad_norm": 0.154296875,
+      "learning_rate": 0.02,
+      "loss": 1.5172,
+      "step": 7997
+    },
+    {
+      "epoch": 0.7404270091997917,
+      "grad_norm": 0.1455078125,
+      "learning_rate": 0.02,
+      "loss": 1.4884,
+      "step": 7998
+    },
+    {
+      "epoch": 0.7405195857200717,
+      "grad_norm": 0.1376953125,
+      "learning_rate": 0.02,
+      "loss": 1.4913,
+      "step": 7999
+    },
+    {
+      "epoch": 0.7406121622403518,
+      "grad_norm": 0.14453125,
+      "learning_rate": 0.02,
+      "loss": 1.4811,
+      "step": 8000
+    },
+    {
+      "epoch": 0.7407047387606318,
+      "grad_norm": 0.138671875,
+      "learning_rate": 0.02,
+      "loss": 1.4943,
+      "step": 8001
+    },
+    {
+      "epoch": 0.7407973152809119,
+      "grad_norm": 0.1513671875,
+      "learning_rate": 0.02,
+      "loss": 1.5129,
+      "step": 8002
+    },
+    {
+      "epoch": 0.7408898918011919,
+      "grad_norm": 0.146484375,
+      "learning_rate": 0.02,
+      "loss": 1.5279,
+      "step": 8003
+    },
+    {
+      "epoch": 0.740982468321472,
+      "grad_norm": 0.1484375,
+      "learning_rate": 0.02,
+      "loss": 1.4991,
+      "step": 8004
+    },
+    {
+      "epoch": 0.741075044841752,
+      "grad_norm": 0.150390625,
+      "learning_rate": 0.02,
+      "loss": 1.5416,
+      "step": 8005
+    },
+    {
+      "epoch": 0.7411676213620321,
+      "grad_norm": 0.1494140625,
+      "learning_rate": 0.02,
+      "loss": 1.5098,
+      "step": 8006
+    },
+    {
+      "epoch": 0.7412601978823121,
+      "grad_norm": 0.1650390625,
+      "learning_rate": 0.02,
+      "loss": 1.4704,
+      "step": 8007
+    },
+    {
+      "epoch": 0.7413527744025922,
+      "grad_norm": 0.14453125,
+      "learning_rate": 0.02,
+      "loss": 1.5872,
+      "step": 8008
+    },
+    {
+      "epoch": 0.7414453509228722,
+      "grad_norm": 0.154296875,
+      "learning_rate": 0.02,
+      "loss": 1.4781,
+      "step": 8009
+    },
+    {
+      "epoch": 0.7415379274431523,
+      "grad_norm": 0.1328125,
+      "learning_rate": 0.02,
+      "loss": 1.5151,
+      "step": 8010
+    },
+    {
+      "epoch": 0.7416305039634322,
+      "grad_norm": 0.154296875,
+      "learning_rate": 0.02,
+      "loss": 1.5681,
+      "step": 8011
+    },
+    {
+      "epoch": 0.7417230804837123,
+      "grad_norm": 0.1494140625,
+      "learning_rate": 0.02,
+      "loss": 1.4912,
+      "step": 8012
+    },
+    {
+      "epoch": 0.7418156570039923,
+      "grad_norm": 0.16015625,
+      "learning_rate": 0.02,
+      "loss": 1.5609,
+      "step": 8013
+    },
+    {
+      "epoch": 0.7419082335242724,
+      "grad_norm": 0.138671875,
+      "learning_rate": 0.02,
+      "loss": 1.5582,
+      "step": 8014
+    },
+    {
+      "epoch": 0.7420008100445524,
+      "grad_norm": 0.158203125,
+      "learning_rate": 0.02,
+      "loss": 1.5393,
+      "step": 8015
+    },
+    {
+      "epoch": 0.7420933865648325,
+      "grad_norm": 0.1484375,
+      "learning_rate": 0.02,
+      "loss": 1.5473,
+      "step": 8016
+    },
+    {
+      "epoch": 0.7421859630851125,
+      "grad_norm": 0.1552734375,
+      "learning_rate": 0.02,
+      "loss": 1.5199,
+      "step": 8017
+    },
+    {
+      "epoch": 0.7422785396053926,
+      "grad_norm": 0.1572265625,
+      "learning_rate": 0.02,
+      "loss": 1.5782,
+      "step": 8018
+    },
+    {
+      "epoch": 0.7423711161256726,
+      "grad_norm": 0.1611328125,
+      "learning_rate": 0.02,
+      "loss": 1.5046,
+      "step": 8019
+    },
+    {
+      "epoch": 0.7424636926459527,
+      "grad_norm": 0.1669921875,
+      "learning_rate": 0.02,
+      "loss": 1.4432,
+      "step": 8020
+    },
+    {
+      "epoch": 0.7425562691662327,
+      "grad_norm": 0.1455078125,
+      "learning_rate": 0.02,
+      "loss": 1.468,
+      "step": 8021
+    },
+    {
+      "epoch": 0.7426488456865128,
+      "grad_norm": 0.1572265625,
+      "learning_rate": 0.02,
+      "loss": 1.496,
+      "step": 8022
+    },
+    {
+      "epoch": 0.7427414222067928,
+      "grad_norm": 0.1552734375,
+      "learning_rate": 0.02,
+      "loss": 1.5223,
+      "step": 8023
+    },
+    {
+      "epoch": 0.7428339987270729,
+      "grad_norm": 0.162109375,
+      "learning_rate": 0.02,
+      "loss": 1.6112,
+      "step": 8024
+    },
+    {
+      "epoch": 0.7429265752473528,
+      "grad_norm": 0.1552734375,
+      "learning_rate": 0.02,
+      "loss": 1.5021,
+      "step": 8025
+    },
+    {
+      "epoch": 0.7430191517676329,
+      "grad_norm": 0.162109375,
+      "learning_rate": 0.02,
+      "loss": 1.4417,
+      "step": 8026
+    },
+    {
+      "epoch": 0.7431117282879129,
+      "grad_norm": 0.15234375,
+      "learning_rate": 0.02,
+      "loss": 1.4619,
+      "step": 8027
+    },
+    {
+      "epoch": 0.743204304808193,
+      "grad_norm": 0.1572265625,
+      "learning_rate": 0.02,
+      "loss": 1.4378,
+      "step": 8028
+    },
+    {
+      "epoch": 0.743296881328473,
+      "grad_norm": 0.1474609375,
+      "learning_rate": 0.02,
+      "loss": 1.4815,
+      "step": 8029
+    },
+    {
+      "epoch": 0.7433894578487531,
+      "grad_norm": 0.1552734375,
+      "learning_rate": 0.02,
+      "loss": 1.4758,
+      "step": 8030
+    },
+    {
+      "epoch": 0.7434820343690332,
+      "grad_norm": 0.1611328125,
+      "learning_rate": 0.02,
+      "loss": 1.5233,
+      "step": 8031
+    },
+    {
+      "epoch": 0.7435746108893132,
+      "grad_norm": 0.1474609375,
+      "learning_rate": 0.02,
+      "loss": 1.4837,
+      "step": 8032
+    },
+    {
+      "epoch": 0.7436671874095933,
+      "grad_norm": 0.146484375,
+      "learning_rate": 0.02,
+      "loss": 1.4834,
+      "step": 8033
+    },
+    {
+      "epoch": 0.7437597639298733,
+      "grad_norm": 0.1494140625,
+      "learning_rate": 0.02,
+      "loss": 1.4326,
+      "step": 8034
+    },
+    {
+      "epoch": 0.7438523404501534,
+      "grad_norm": 0.1572265625,
+      "learning_rate": 0.02,
+      "loss": 1.4893,
+      "step": 8035
+    },
+    {
+      "epoch": 0.7439449169704334,
+      "grad_norm": 0.146484375,
+      "learning_rate": 0.02,
+      "loss": 1.471,
+      "step": 8036
+    },
+    {
+      "epoch": 0.7440374934907135,
+      "grad_norm": 0.1435546875,
+      "learning_rate": 0.02,
+      "loss": 1.5032,
+      "step": 8037
+    },
+    {
+      "epoch": 0.7441300700109935,
+      "grad_norm": 0.1494140625,
+      "learning_rate": 0.02,
+      "loss": 1.5257,
+      "step": 8038
+    },
+    {
+      "epoch": 0.7442226465312735,
+      "grad_norm": 0.1552734375,
+      "learning_rate": 0.02,
+      "loss": 1.4998,
+      "step": 8039
+    },
+    {
+      "epoch": 0.7443152230515535,
+      "grad_norm": 0.146484375,
+      "learning_rate": 0.02,
+      "loss": 1.5254,
+      "step": 8040
+    },
+    {
+      "epoch": 0.7444077995718336,
+      "grad_norm": 0.1455078125,
+      "learning_rate": 0.02,
+      "loss": 1.4864,
+      "step": 8041
+    },
+    {
+      "epoch": 0.7445003760921136,
+      "grad_norm": 0.14453125,
+      "learning_rate": 0.02,
+      "loss": 1.4754,
+      "step": 8042
+    },
+    {
+      "epoch": 0.7445929526123937,
+      "grad_norm": 0.142578125,
+      "learning_rate": 0.02,
+      "loss": 1.4889,
+      "step": 8043
+    },
+    {
+      "epoch": 0.7446855291326737,
+      "grad_norm": 0.1533203125,
+      "learning_rate": 0.02,
+      "loss": 1.4961,
+      "step": 8044
+    },
+    {
+      "epoch": 0.7447781056529538,
+      "grad_norm": 0.15234375,
+      "learning_rate": 0.02,
+      "loss": 1.5138,
+      "step": 8045
+    },
+    {
+      "epoch": 0.7448706821732338,
+      "grad_norm": 0.1552734375,
+      "learning_rate": 0.02,
+      "loss": 1.5353,
+      "step": 8046
+    },
+    {
+      "epoch": 0.7449632586935139,
+      "grad_norm": 0.14453125,
+      "learning_rate": 0.02,
+      "loss": 1.459,
+      "step": 8047
+    },
+    {
+      "epoch": 0.7450558352137939,
+      "grad_norm": 0.15625,
+      "learning_rate": 0.02,
+      "loss": 1.5512,
+      "step": 8048
+    },
+    {
+      "epoch": 0.745148411734074,
+      "grad_norm": 0.1513671875,
+      "learning_rate": 0.02,
+      "loss": 1.4723,
+      "step": 8049
+    },
+    {
+      "epoch": 0.745240988254354,
+      "grad_norm": 0.1552734375,
+      "learning_rate": 0.02,
+      "loss": 1.493,
+      "step": 8050
+    },
+    {
+      "epoch": 0.7453335647746341,
+      "grad_norm": 0.1494140625,
+      "learning_rate": 0.02,
+      "loss": 1.5356,
+      "step": 8051
+    },
+    {
+      "epoch": 0.745426141294914,
+      "grad_norm": 0.1552734375,
+      "learning_rate": 0.02,
+      "loss": 1.4729,
+      "step": 8052
+    },
+    {
+      "epoch": 0.7455187178151941,
+      "grad_norm": 0.17578125,
+      "learning_rate": 0.02,
+      "loss": 1.5199,
+      "step": 8053
+    },
+    {
+      "epoch": 0.7456112943354741,
+      "grad_norm": 0.154296875,
+      "learning_rate": 0.02,
+      "loss": 1.5121,
+      "step": 8054
+    },
+    {
+      "epoch": 0.7457038708557542,
+      "grad_norm": 0.150390625,
+      "learning_rate": 0.02,
+      "loss": 1.4869,
+      "step": 8055
+    },
+    {
+      "epoch": 0.7457964473760342,
+      "grad_norm": 0.1484375,
+      "learning_rate": 0.02,
+      "loss": 1.5313,
+      "step": 8056
+    },
+    {
+      "epoch": 0.7458890238963143,
+      "grad_norm": 0.1513671875,
+      "learning_rate": 0.02,
+      "loss": 1.5439,
+      "step": 8057
+    },
+    {
+      "epoch": 0.7459816004165943,
+      "grad_norm": 0.1611328125,
+      "learning_rate": 0.02,
+      "loss": 1.4931,
+      "step": 8058
+    },
+    {
+      "epoch": 0.7460741769368744,
+      "grad_norm": 0.1533203125,
+      "learning_rate": 0.02,
+      "loss": 1.5148,
+      "step": 8059
+    },
+    {
+      "epoch": 0.7461667534571544,
+      "grad_norm": 0.1552734375,
+      "learning_rate": 0.02,
+      "loss": 1.5292,
+      "step": 8060
+    },
+    {
+      "epoch": 0.7462593299774345,
+      "grad_norm": 0.1455078125,
+      "learning_rate": 0.02,
+      "loss": 1.5504,
+      "step": 8061
+    },
+    {
+      "epoch": 0.7463519064977145,
+      "grad_norm": 0.1640625,
+      "learning_rate": 0.02,
+      "loss": 1.4837,
+      "step": 8062
+    },
+    {
+      "epoch": 0.7464444830179946,
+      "grad_norm": 0.1396484375,
+      "learning_rate": 0.02,
+      "loss": 1.5026,
+      "step": 8063
+    },
+    {
+      "epoch": 0.7465370595382746,
+      "grad_norm": 0.1474609375,
+      "learning_rate": 0.02,
+      "loss": 1.5427,
+      "step": 8064
+    },
+    {
+      "epoch": 0.7466296360585547,
+      "grad_norm": 0.1396484375,
+      "learning_rate": 0.02,
+      "loss": 1.485,
+      "step": 8065
+    },
+    {
+      "epoch": 0.7467222125788346,
+      "grad_norm": 0.154296875,
+      "learning_rate": 0.02,
+      "loss": 1.5366,
+      "step": 8066
+    },
+    {
+      "epoch": 0.7468147890991147,
+      "grad_norm": 0.1357421875,
+      "learning_rate": 0.02,
+      "loss": 1.4742,
+      "step": 8067
+    },
+    {
+      "epoch": 0.7469073656193947,
+      "grad_norm": 0.1435546875,
+      "learning_rate": 0.02,
+      "loss": 1.555,
+      "step": 8068
+    },
+    {
+      "epoch": 0.7469999421396748,
+      "grad_norm": 0.1376953125,
+      "learning_rate": 0.02,
+      "loss": 1.5594,
+      "step": 8069
+    },
+    {
+      "epoch": 0.7470925186599549,
+      "grad_norm": 0.150390625,
+      "learning_rate": 0.02,
+      "loss": 1.4916,
+      "step": 8070
+    },
+    {
+      "epoch": 0.7471850951802349,
+      "grad_norm": 0.1455078125,
+      "learning_rate": 0.02,
+      "loss": 1.4876,
+      "step": 8071
+    },
+    {
+      "epoch": 0.747277671700515,
+      "grad_norm": 0.1533203125,
+      "learning_rate": 0.02,
+      "loss": 1.5325,
+      "step": 8072
+    },
+    {
+      "epoch": 0.747370248220795,
+      "grad_norm": 0.1494140625,
+      "learning_rate": 0.02,
+      "loss": 1.4698,
+      "step": 8073
+    },
+    {
+      "epoch": 0.7474628247410751,
+      "grad_norm": 0.1494140625,
+      "learning_rate": 0.02,
+      "loss": 1.4938,
+      "step": 8074
+    },
+    {
+      "epoch": 0.7475554012613551,
+      "grad_norm": 0.142578125,
+      "learning_rate": 0.02,
+      "loss": 1.4718,
+      "step": 8075
+    },
+    {
+      "epoch": 0.7476479777816352,
+      "grad_norm": 0.1494140625,
+      "learning_rate": 0.02,
+      "loss": 1.5174,
+      "step": 8076
+    },
+    {
+      "epoch": 0.7477405543019152,
+      "grad_norm": 0.1474609375,
+      "learning_rate": 0.02,
+      "loss": 1.5115,
+      "step": 8077
+    },
+    {
+      "epoch": 0.7478331308221953,
+      "grad_norm": 0.15234375,
+      "learning_rate": 0.02,
+      "loss": 1.5752,
+      "step": 8078
+    },
+    {
+      "epoch": 0.7479257073424752,
+      "grad_norm": 0.1552734375,
+      "learning_rate": 0.02,
+      "loss": 1.5383,
+      "step": 8079
+    },
+    {
+      "epoch": 0.7480182838627553,
+      "grad_norm": 0.146484375,
+      "learning_rate": 0.02,
+      "loss": 1.4665,
+      "step": 8080
+    },
+    {
+      "epoch": 0.7481108603830353,
+      "grad_norm": 0.1455078125,
+      "learning_rate": 0.02,
+      "loss": 1.4864,
+      "step": 8081
+    },
+    {
+      "epoch": 0.7482034369033154,
+      "grad_norm": 0.150390625,
+      "learning_rate": 0.02,
+      "loss": 1.4994,
+      "step": 8082
+    },
+    {
+      "epoch": 0.7482960134235954,
+      "grad_norm": 0.15625,
+      "learning_rate": 0.02,
+      "loss": 1.4863,
+      "step": 8083
+    },
+    {
+      "epoch": 0.7483885899438755,
+      "grad_norm": 0.1572265625,
+      "learning_rate": 0.02,
+      "loss": 1.4767,
+      "step": 8084
+    },
+    {
+      "epoch": 0.7484811664641555,
+      "grad_norm": 0.1435546875,
+      "learning_rate": 0.02,
+      "loss": 1.5111,
+      "step": 8085
+    },
+    {
+      "epoch": 0.7485737429844356,
+      "grad_norm": 0.1572265625,
+      "learning_rate": 0.02,
+      "loss": 1.5092,
+      "step": 8086
+    },
+    {
+      "epoch": 0.7486663195047156,
+      "grad_norm": 0.146484375,
+      "learning_rate": 0.02,
+      "loss": 1.5814,
+      "step": 8087
+    },
+    {
+      "epoch": 0.7487588960249957,
+      "grad_norm": 0.1572265625,
+      "learning_rate": 0.02,
+      "loss": 1.5062,
+      "step": 8088
+    },
+    {
+      "epoch": 0.7488514725452757,
+      "grad_norm": 0.1435546875,
+      "learning_rate": 0.02,
+      "loss": 1.5279,
+      "step": 8089
+    },
+    {
+      "epoch": 0.7489440490655558,
+      "grad_norm": 0.15234375,
+      "learning_rate": 0.02,
+      "loss": 1.4762,
+      "step": 8090
+    },
+    {
+      "epoch": 0.7490366255858358,
+      "grad_norm": 0.1435546875,
+      "learning_rate": 0.02,
+      "loss": 1.5336,
+      "step": 8091
+    },
+    {
+      "epoch": 0.7491292021061159,
+      "grad_norm": 0.1513671875,
+      "learning_rate": 0.02,
+      "loss": 1.537,
+      "step": 8092
+    },
+    {
+      "epoch": 0.7492217786263958,
+      "grad_norm": 0.146484375,
+      "learning_rate": 0.02,
+      "loss": 1.4763,
+      "step": 8093
+    },
+    {
+      "epoch": 0.7493143551466759,
+      "grad_norm": 0.15625,
+      "learning_rate": 0.02,
+      "loss": 1.5438,
+      "step": 8094
+    },
+    {
+      "epoch": 0.7494069316669559,
+      "grad_norm": 0.1611328125,
+      "learning_rate": 0.02,
+      "loss": 1.4911,
+      "step": 8095
+    },
+    {
+      "epoch": 0.749499508187236,
+      "grad_norm": 0.142578125,
+      "learning_rate": 0.02,
+      "loss": 1.4907,
+      "step": 8096
+    },
+    {
+      "epoch": 0.749592084707516,
+      "grad_norm": 0.158203125,
+      "learning_rate": 0.02,
+      "loss": 1.5589,
+      "step": 8097
+    },
+    {
+      "epoch": 0.7496846612277961,
+      "grad_norm": 0.1572265625,
+      "learning_rate": 0.02,
+      "loss": 1.5207,
+      "step": 8098
+    },
+    {
+      "epoch": 0.7497772377480761,
+      "grad_norm": 0.1513671875,
+      "learning_rate": 0.02,
+      "loss": 1.5081,
+      "step": 8099
+    },
+    {
+      "epoch": 0.7498698142683562,
+      "grad_norm": 0.1533203125,
+      "learning_rate": 0.02,
+      "loss": 1.5473,
+      "step": 8100
+    },
+    {
+      "epoch": 0.7499623907886362,
+      "grad_norm": 0.16015625,
+      "learning_rate": 0.02,
+      "loss": 1.5483,
+      "step": 8101
+    },
+    {
+      "epoch": 0.7500549673089163,
+      "grad_norm": 0.154296875,
+      "learning_rate": 0.02,
+      "loss": 1.5493,
+      "step": 8102
+    },
+    {
+      "epoch": 0.7501475438291964,
+      "grad_norm": 0.15234375,
+      "learning_rate": 0.02,
+      "loss": 1.5257,
+      "step": 8103
+    },
+    {
+      "epoch": 0.7502401203494764,
+      "grad_norm": 0.1396484375,
+      "learning_rate": 0.02,
+      "loss": 1.4808,
+      "step": 8104
+    },
+    {
+      "epoch": 0.7503326968697565,
+      "grad_norm": 0.15234375,
+      "learning_rate": 0.02,
+      "loss": 1.4977,
+      "step": 8105
+    },
+    {
+      "epoch": 0.7504252733900365,
+      "grad_norm": 0.14453125,
+      "learning_rate": 0.02,
+      "loss": 1.4527,
+      "step": 8106
+    },
+    {
+      "epoch": 0.7505178499103164,
+      "grad_norm": 0.14453125,
+      "learning_rate": 0.02,
+      "loss": 1.511,
+      "step": 8107
+    },
+    {
+      "epoch": 0.7506104264305965,
+      "grad_norm": 0.142578125,
+      "learning_rate": 0.02,
+      "loss": 1.4498,
+      "step": 8108
+    },
+    {
+      "epoch": 0.7507030029508766,
+      "grad_norm": 0.150390625,
+      "learning_rate": 0.02,
+      "loss": 1.4318,
+      "step": 8109
+    },
+    {
+      "epoch": 0.7507955794711566,
+      "grad_norm": 0.146484375,
+      "learning_rate": 0.02,
+      "loss": 1.5185,
+      "step": 8110
+    },
+    {
+      "epoch": 0.7508881559914367,
+      "grad_norm": 0.16015625,
+      "learning_rate": 0.02,
+      "loss": 1.4431,
+      "step": 8111
+    },
+    {
+      "epoch": 0.7509807325117167,
+      "grad_norm": 0.1494140625,
+      "learning_rate": 0.02,
+      "loss": 1.5008,
+      "step": 8112
+    },
+    {
+      "epoch": 0.7510733090319968,
+      "grad_norm": 0.166015625,
+      "learning_rate": 0.02,
+      "loss": 1.5544,
+      "step": 8113
+    },
+    {
+      "epoch": 0.7511658855522768,
+      "grad_norm": 0.1455078125,
+      "learning_rate": 0.02,
+      "loss": 1.5408,
+      "step": 8114
+    },
+    {
+      "epoch": 0.7512584620725569,
+      "grad_norm": 0.1630859375,
+      "learning_rate": 0.02,
+      "loss": 1.4719,
+      "step": 8115
+    },
+    {
+      "epoch": 0.7513510385928369,
+      "grad_norm": 0.1494140625,
+      "learning_rate": 0.02,
+      "loss": 1.5083,
+      "step": 8116
+    },
+    {
+      "epoch": 0.751443615113117,
+      "grad_norm": 0.1591796875,
+      "learning_rate": 0.02,
+      "loss": 1.4902,
+      "step": 8117
+    },
+    {
+      "epoch": 0.751536191633397,
+      "grad_norm": 0.146484375,
+      "learning_rate": 0.02,
+      "loss": 1.4508,
+      "step": 8118
+    },
+    {
+      "epoch": 0.7516287681536771,
+      "grad_norm": 0.158203125,
+      "learning_rate": 0.02,
+      "loss": 1.4425,
+      "step": 8119
+    },
+    {
+      "epoch": 0.751721344673957,
+      "grad_norm": 0.15234375,
+      "learning_rate": 0.02,
+      "loss": 1.5232,
+      "step": 8120
+    },
+    {
+      "epoch": 0.7518139211942371,
+      "grad_norm": 0.15234375,
+      "learning_rate": 0.02,
+      "loss": 1.4922,
+      "step": 8121
+    },
+    {
+      "epoch": 0.7519064977145171,
+      "grad_norm": 0.1591796875,
+      "learning_rate": 0.02,
+      "loss": 1.4992,
+      "step": 8122
+    },
+    {
+      "epoch": 0.7519990742347972,
+      "grad_norm": 0.1494140625,
+      "learning_rate": 0.02,
+      "loss": 1.5077,
+      "step": 8123
+    },
+    {
+      "epoch": 0.7520916507550772,
+      "grad_norm": 0.166015625,
+      "learning_rate": 0.02,
+      "loss": 1.5431,
+      "step": 8124
+    },
+    {
+      "epoch": 0.7521842272753573,
+      "grad_norm": 0.1630859375,
+      "learning_rate": 0.02,
+      "loss": 1.5064,
+      "step": 8125
+    },
+    {
+      "epoch": 0.7522768037956373,
+      "grad_norm": 0.1611328125,
+      "learning_rate": 0.02,
+      "loss": 1.5282,
+      "step": 8126
+    },
+    {
+      "epoch": 0.7523693803159174,
+      "grad_norm": 0.1591796875,
+      "learning_rate": 0.02,
+      "loss": 1.5068,
+      "step": 8127
+    },
+    {
+      "epoch": 0.7524619568361974,
+      "grad_norm": 0.154296875,
+      "learning_rate": 0.02,
+      "loss": 1.4802,
+      "step": 8128
+    },
+    {
+      "epoch": 0.7525545333564775,
+      "grad_norm": 0.150390625,
+      "learning_rate": 0.02,
+      "loss": 1.5544,
+      "step": 8129
+    },
+    {
+      "epoch": 0.7526471098767575,
+      "grad_norm": 0.1611328125,
+      "learning_rate": 0.02,
+      "loss": 1.5678,
+      "step": 8130
+    },
+    {
+      "epoch": 0.7527396863970376,
+      "grad_norm": 0.154296875,
+      "learning_rate": 0.02,
+      "loss": 1.5428,
+      "step": 8131
+    },
+    {
+      "epoch": 0.7528322629173176,
+      "grad_norm": 0.15625,
+      "learning_rate": 0.02,
+      "loss": 1.4732,
+      "step": 8132
+    },
+    {
+      "epoch": 0.7529248394375977,
+      "grad_norm": 0.1533203125,
+      "learning_rate": 0.02,
+      "loss": 1.4581,
+      "step": 8133
+    },
+    {
+      "epoch": 0.7530174159578776,
+      "grad_norm": 0.1630859375,
+      "learning_rate": 0.02,
+      "loss": 1.4846,
+      "step": 8134
+    },
+    {
+      "epoch": 0.7531099924781577,
+      "grad_norm": 0.1533203125,
+      "learning_rate": 0.02,
+      "loss": 1.4853,
+      "step": 8135
+    },
+    {
+      "epoch": 0.7532025689984377,
+      "grad_norm": 0.1611328125,
+      "learning_rate": 0.02,
+      "loss": 1.4582,
+      "step": 8136
+    },
+    {
+      "epoch": 0.7532951455187178,
+      "grad_norm": 0.162109375,
+      "learning_rate": 0.02,
+      "loss": 1.545,
+      "step": 8137
+    },
+    {
+      "epoch": 0.7533877220389978,
+      "grad_norm": 0.1630859375,
+      "learning_rate": 0.02,
+      "loss": 1.5124,
+      "step": 8138
+    },
+    {
+      "epoch": 0.7534802985592779,
+      "grad_norm": 0.1689453125,
+      "learning_rate": 0.02,
+      "loss": 1.5603,
+      "step": 8139
+    },
+    {
+      "epoch": 0.753572875079558,
+      "grad_norm": 0.16796875,
+      "learning_rate": 0.02,
+      "loss": 1.4992,
+      "step": 8140
+    },
+    {
+      "epoch": 0.753665451599838,
+      "grad_norm": 0.1591796875,
+      "learning_rate": 0.02,
+      "loss": 1.4514,
+      "step": 8141
+    },
+    {
+      "epoch": 0.753758028120118,
+      "grad_norm": 0.1611328125,
+      "learning_rate": 0.02,
+      "loss": 1.5118,
+      "step": 8142
+    },
+    {
+      "epoch": 0.7538506046403981,
+      "grad_norm": 0.1708984375,
+      "learning_rate": 0.02,
+      "loss": 1.4915,
+      "step": 8143
+    },
+    {
+      "epoch": 0.7539431811606782,
+      "grad_norm": 0.1572265625,
+      "learning_rate": 0.02,
+      "loss": 1.487,
+      "step": 8144
+    },
+    {
+      "epoch": 0.7540357576809582,
+      "grad_norm": 0.173828125,
+      "learning_rate": 0.02,
+      "loss": 1.5949,
+      "step": 8145
+    },
+    {
+      "epoch": 0.7541283342012383,
+      "grad_norm": 0.154296875,
+      "learning_rate": 0.02,
+      "loss": 1.5074,
+      "step": 8146
+    },
+    {
+      "epoch": 0.7542209107215182,
+      "grad_norm": 0.16796875,
+      "learning_rate": 0.02,
+      "loss": 1.5598,
+      "step": 8147
+    },
+    {
+      "epoch": 0.7543134872417983,
+      "grad_norm": 0.1494140625,
+      "learning_rate": 0.02,
+      "loss": 1.4949,
+      "step": 8148
+    },
+    {
+      "epoch": 0.7544060637620783,
+      "grad_norm": 0.1572265625,
+      "learning_rate": 0.02,
+      "loss": 1.4914,
+      "step": 8149
+    },
+    {
+      "epoch": 0.7544986402823584,
+      "grad_norm": 0.154296875,
+      "learning_rate": 0.02,
+      "loss": 1.5352,
+      "step": 8150
+    },
+    {
+      "epoch": 0.7545912168026384,
+      "grad_norm": 0.1650390625,
+      "learning_rate": 0.02,
+      "loss": 1.4792,
+      "step": 8151
+    },
+    {
+      "epoch": 0.7546837933229185,
+      "grad_norm": 0.1689453125,
+      "learning_rate": 0.02,
+      "loss": 1.5061,
+      "step": 8152
+    },
+    {
+      "epoch": 0.7547763698431985,
+      "grad_norm": 0.166015625,
+      "learning_rate": 0.02,
+      "loss": 1.4956,
+      "step": 8153
+    },
+    {
+      "epoch": 0.7548689463634786,
+      "grad_norm": 0.1416015625,
+      "learning_rate": 0.02,
+      "loss": 1.5073,
+      "step": 8154
+    },
+    {
+      "epoch": 0.7549615228837586,
+      "grad_norm": 0.1787109375,
+      "learning_rate": 0.02,
+      "loss": 1.6031,
+      "step": 8155
+    },
+    {
+      "epoch": 0.7550540994040387,
+      "grad_norm": 0.14453125,
+      "learning_rate": 0.02,
+      "loss": 1.4913,
+      "step": 8156
+    },
+    {
+      "epoch": 0.7551466759243187,
+      "grad_norm": 0.154296875,
+      "learning_rate": 0.02,
+      "loss": 1.5373,
+      "step": 8157
+    },
+    {
+      "epoch": 0.7552392524445988,
+      "grad_norm": 0.1435546875,
+      "learning_rate": 0.02,
+      "loss": 1.5343,
+      "step": 8158
+    },
+    {
+      "epoch": 0.7553318289648788,
+      "grad_norm": 0.1552734375,
+      "learning_rate": 0.02,
+      "loss": 1.5134,
+      "step": 8159
+    },
+    {
+      "epoch": 0.7554244054851589,
+      "grad_norm": 0.1513671875,
+      "learning_rate": 0.02,
+      "loss": 1.5686,
+      "step": 8160
+    },
+    {
+      "epoch": 0.7555169820054388,
+      "grad_norm": 0.15625,
+      "learning_rate": 0.02,
+      "loss": 1.4731,
+      "step": 8161
+    },
+    {
+      "epoch": 0.7556095585257189,
+      "grad_norm": 0.150390625,
+      "learning_rate": 0.02,
+      "loss": 1.5275,
+      "step": 8162
+    },
+    {
+      "epoch": 0.7557021350459989,
+      "grad_norm": 0.1494140625,
+      "learning_rate": 0.02,
+      "loss": 1.5052,
+      "step": 8163
+    },
+    {
+      "epoch": 0.755794711566279,
+      "grad_norm": 0.158203125,
+      "learning_rate": 0.02,
+      "loss": 1.5172,
+      "step": 8164
+    },
+    {
+      "epoch": 0.755887288086559,
+      "grad_norm": 0.14453125,
+      "learning_rate": 0.02,
+      "loss": 1.5189,
+      "step": 8165
+    },
+    {
+      "epoch": 0.7559798646068391,
+      "grad_norm": 0.1484375,
+      "learning_rate": 0.02,
+      "loss": 1.4845,
+      "step": 8166
+    },
+    {
+      "epoch": 0.7560724411271191,
+      "grad_norm": 0.15234375,
+      "learning_rate": 0.02,
+      "loss": 1.4847,
+      "step": 8167
+    },
+    {
+      "epoch": 0.7561650176473992,
+      "grad_norm": 0.171875,
+      "learning_rate": 0.02,
+      "loss": 1.4846,
+      "step": 8168
+    },
+    {
+      "epoch": 0.7562575941676792,
+      "grad_norm": 0.1484375,
+      "learning_rate": 0.02,
+      "loss": 1.4483,
+      "step": 8169
+    },
+    {
+      "epoch": 0.7563501706879593,
+      "grad_norm": 0.1552734375,
+      "learning_rate": 0.02,
+      "loss": 1.5637,
+      "step": 8170
+    },
+    {
+      "epoch": 0.7564427472082393,
+      "grad_norm": 0.1572265625,
+      "learning_rate": 0.02,
+      "loss": 1.5178,
+      "step": 8171
+    },
+    {
+      "epoch": 0.7565353237285194,
+      "grad_norm": 0.1611328125,
+      "learning_rate": 0.02,
+      "loss": 1.472,
+      "step": 8172
+    },
+    {
+      "epoch": 0.7566279002487994,
+      "grad_norm": 0.1513671875,
+      "learning_rate": 0.02,
+      "loss": 1.4877,
+      "step": 8173
+    },
+    {
+      "epoch": 0.7567204767690795,
+      "grad_norm": 0.154296875,
+      "learning_rate": 0.02,
+      "loss": 1.5999,
+      "step": 8174
+    },
+    {
+      "epoch": 0.7568130532893594,
+      "grad_norm": 0.138671875,
+      "learning_rate": 0.02,
+      "loss": 1.4639,
+      "step": 8175
+    },
+    {
+      "epoch": 0.7569056298096395,
+      "grad_norm": 0.1552734375,
+      "learning_rate": 0.02,
+      "loss": 1.5869,
+      "step": 8176
+    },
+    {
+      "epoch": 0.7569982063299195,
+      "grad_norm": 0.1416015625,
+      "learning_rate": 0.02,
+      "loss": 1.5212,
+      "step": 8177
+    },
+    {
+      "epoch": 0.7570907828501996,
+      "grad_norm": 0.1484375,
+      "learning_rate": 0.02,
+      "loss": 1.5336,
+      "step": 8178
+    },
+    {
+      "epoch": 0.7571833593704796,
+      "grad_norm": 0.1474609375,
+      "learning_rate": 0.02,
+      "loss": 1.5473,
+      "step": 8179
+    },
+    {
+      "epoch": 0.7572759358907597,
+      "grad_norm": 0.15234375,
+      "learning_rate": 0.02,
+      "loss": 1.4917,
+      "step": 8180
+    },
+    {
+      "epoch": 0.7573685124110398,
+      "grad_norm": 0.14453125,
+      "learning_rate": 0.02,
+      "loss": 1.4869,
+      "step": 8181
+    },
+    {
+      "epoch": 0.7574610889313198,
+      "grad_norm": 0.15234375,
+      "learning_rate": 0.02,
+      "loss": 1.4723,
+      "step": 8182
+    },
+    {
+      "epoch": 0.7575536654515999,
+      "grad_norm": 0.140625,
+      "learning_rate": 0.02,
+      "loss": 1.4544,
+      "step": 8183
+    },
+    {
+      "epoch": 0.7576462419718799,
+      "grad_norm": 0.15625,
+      "learning_rate": 0.02,
+      "loss": 1.5777,
+      "step": 8184
+    },
+    {
+      "epoch": 0.75773881849216,
+      "grad_norm": 0.1572265625,
+      "learning_rate": 0.02,
+      "loss": 1.5585,
+      "step": 8185
+    },
+    {
+      "epoch": 0.75783139501244,
+      "grad_norm": 0.154296875,
+      "learning_rate": 0.02,
+      "loss": 1.5663,
+      "step": 8186
+    },
+    {
+      "epoch": 0.7579239715327201,
+      "grad_norm": 0.146484375,
+      "learning_rate": 0.02,
+      "loss": 1.5259,
+      "step": 8187
+    },
+    {
+      "epoch": 0.758016548053,
+      "grad_norm": 0.1572265625,
+      "learning_rate": 0.02,
+      "loss": 1.5013,
+      "step": 8188
+    },
+    {
+      "epoch": 0.7581091245732801,
+      "grad_norm": 0.146484375,
+      "learning_rate": 0.02,
+      "loss": 1.5585,
+      "step": 8189
+    },
+    {
+      "epoch": 0.7582017010935601,
+      "grad_norm": 0.1494140625,
+      "learning_rate": 0.02,
+      "loss": 1.5613,
+      "step": 8190
+    },
+    {
+      "epoch": 0.7582942776138402,
+      "grad_norm": 0.140625,
+      "learning_rate": 0.02,
+      "loss": 1.4974,
+      "step": 8191
+    },
+    {
+      "epoch": 0.7583868541341202,
+      "grad_norm": 0.1533203125,
+      "learning_rate": 0.02,
+      "loss": 1.4867,
+      "step": 8192
+    },
+    {
+      "epoch": 0.7584794306544003,
+      "grad_norm": 0.142578125,
+      "learning_rate": 0.02,
+      "loss": 1.5131,
+      "step": 8193
+    },
+    {
+      "epoch": 0.7585720071746803,
+      "grad_norm": 0.1533203125,
+      "learning_rate": 0.02,
+      "loss": 1.5088,
+      "step": 8194
+    },
+    {
+      "epoch": 0.7586645836949604,
+      "grad_norm": 0.1435546875,
+      "learning_rate": 0.02,
+      "loss": 1.4586,
+      "step": 8195
+    },
+    {
+      "epoch": 0.7587571602152404,
+      "grad_norm": 0.1591796875,
+      "learning_rate": 0.02,
+      "loss": 1.5123,
+      "step": 8196
+    },
+    {
+      "epoch": 0.7588497367355205,
+      "grad_norm": 0.1533203125,
+      "learning_rate": 0.02,
+      "loss": 1.5104,
+      "step": 8197
+    },
+    {
+      "epoch": 0.7589423132558005,
+      "grad_norm": 0.1611328125,
+      "learning_rate": 0.02,
+      "loss": 1.4858,
+      "step": 8198
+    },
+    {
+      "epoch": 0.7590348897760806,
+      "grad_norm": 0.1533203125,
+      "learning_rate": 0.02,
+      "loss": 1.5371,
+      "step": 8199
+    },
+    {
+      "epoch": 0.7591274662963606,
+      "grad_norm": 0.162109375,
+      "learning_rate": 0.02,
+      "loss": 1.4156,
+      "step": 8200
+    },
+    {
+      "epoch": 0.7592200428166407,
+      "grad_norm": 0.158203125,
+      "learning_rate": 0.02,
+      "loss": 1.5426,
+      "step": 8201
+    },
+    {
+      "epoch": 0.7593126193369206,
+      "grad_norm": 0.1494140625,
+      "learning_rate": 0.02,
+      "loss": 1.4771,
+      "step": 8202
+    },
+    {
+      "epoch": 0.7594051958572007,
+      "grad_norm": 0.154296875,
+      "learning_rate": 0.02,
+      "loss": 1.3929,
+      "step": 8203
+    },
+    {
+      "epoch": 0.7594977723774807,
+      "grad_norm": 0.158203125,
+      "learning_rate": 0.02,
+      "loss": 1.5829,
+      "step": 8204
+    },
+    {
+      "epoch": 0.7595903488977608,
+      "grad_norm": 0.15625,
+      "learning_rate": 0.02,
+      "loss": 1.4988,
+      "step": 8205
+    },
+    {
+      "epoch": 0.7596829254180408,
+      "grad_norm": 0.1552734375,
+      "learning_rate": 0.02,
+      "loss": 1.5114,
+      "step": 8206
+    },
+    {
+      "epoch": 0.7597755019383209,
+      "grad_norm": 0.1572265625,
+      "learning_rate": 0.02,
+      "loss": 1.5094,
+      "step": 8207
+    },
+    {
+      "epoch": 0.7598680784586009,
+      "grad_norm": 0.1640625,
+      "learning_rate": 0.02,
+      "loss": 1.4798,
+      "step": 8208
+    },
+    {
+      "epoch": 0.759960654978881,
+      "grad_norm": 0.1455078125,
+      "learning_rate": 0.02,
+      "loss": 1.5686,
+      "step": 8209
+    },
+    {
+      "epoch": 0.760053231499161,
+      "grad_norm": 0.1572265625,
+      "learning_rate": 0.02,
+      "loss": 1.5619,
+      "step": 8210
+    },
+    {
+      "epoch": 0.7601458080194411,
+      "grad_norm": 0.158203125,
+      "learning_rate": 0.02,
+      "loss": 1.5077,
+      "step": 8211
+    },
+    {
+      "epoch": 0.7602383845397211,
+      "grad_norm": 0.1572265625,
+      "learning_rate": 0.02,
+      "loss": 1.5189,
+      "step": 8212
+    },
+    {
+      "epoch": 0.7603309610600012,
+      "grad_norm": 0.1474609375,
+      "learning_rate": 0.02,
+      "loss": 1.5532,
+      "step": 8213
+    },
+    {
+      "epoch": 0.7604235375802813,
+      "grad_norm": 0.1533203125,
+      "learning_rate": 0.02,
+      "loss": 1.5039,
+      "step": 8214
+    },
+    {
+      "epoch": 0.7605161141005612,
+      "grad_norm": 0.1435546875,
+      "learning_rate": 0.02,
+      "loss": 1.4836,
+      "step": 8215
+    },
+    {
+      "epoch": 0.7606086906208412,
+      "grad_norm": 0.1484375,
+      "learning_rate": 0.02,
+      "loss": 1.5445,
+      "step": 8216
+    },
+    {
+      "epoch": 0.7607012671411213,
+      "grad_norm": 0.154296875,
+      "learning_rate": 0.02,
+      "loss": 1.5528,
+      "step": 8217
+    },
+    {
+      "epoch": 0.7607938436614013,
+      "grad_norm": 0.1630859375,
+      "learning_rate": 0.02,
+      "loss": 1.5756,
+      "step": 8218
+    },
+    {
+      "epoch": 0.7608864201816814,
+      "grad_norm": 0.1484375,
+      "learning_rate": 0.02,
+      "loss": 1.5368,
+      "step": 8219
+    },
+    {
+      "epoch": 0.7609789967019615,
+      "grad_norm": 0.1533203125,
+      "learning_rate": 0.02,
+      "loss": 1.534,
+      "step": 8220
+    },
+    {
+      "epoch": 0.7610715732222415,
+      "grad_norm": 0.1435546875,
+      "learning_rate": 0.02,
+      "loss": 1.4747,
+      "step": 8221
+    },
+    {
+      "epoch": 0.7611641497425216,
+      "grad_norm": 0.1494140625,
+      "learning_rate": 0.02,
+      "loss": 1.4925,
+      "step": 8222
+    },
+    {
+      "epoch": 0.7612567262628016,
+      "grad_norm": 0.1552734375,
+      "learning_rate": 0.02,
+      "loss": 1.5028,
+      "step": 8223
+    },
+    {
+      "epoch": 0.7613493027830817,
+      "grad_norm": 0.150390625,
+      "learning_rate": 0.02,
+      "loss": 1.479,
+      "step": 8224
+    },
+    {
+      "epoch": 0.7614418793033617,
+      "grad_norm": 0.16796875,
+      "learning_rate": 0.02,
+      "loss": 1.5528,
+      "step": 8225
+    },
+    {
+      "epoch": 0.7615344558236418,
+      "grad_norm": 0.15625,
+      "learning_rate": 0.02,
+      "loss": 1.4414,
+      "step": 8226
+    },
+    {
+      "epoch": 0.7616270323439218,
+      "grad_norm": 0.150390625,
+      "learning_rate": 0.02,
+      "loss": 1.4517,
+      "step": 8227
+    },
+    {
+      "epoch": 0.7617196088642019,
+      "grad_norm": 0.154296875,
+      "learning_rate": 0.02,
+      "loss": 1.5682,
+      "step": 8228
+    },
+    {
+      "epoch": 0.7618121853844818,
+      "grad_norm": 0.1640625,
+      "learning_rate": 0.02,
+      "loss": 1.5274,
+      "step": 8229
+    },
+    {
+      "epoch": 0.7619047619047619,
+      "grad_norm": 0.158203125,
+      "learning_rate": 0.02,
+      "loss": 1.4582,
+      "step": 8230
+    },
+    {
+      "epoch": 0.7619973384250419,
+      "grad_norm": 0.1533203125,
+      "learning_rate": 0.02,
+      "loss": 1.4979,
+      "step": 8231
+    },
+    {
+      "epoch": 0.762089914945322,
+      "grad_norm": 0.1611328125,
+      "learning_rate": 0.02,
+      "loss": 1.5139,
+      "step": 8232
+    },
+    {
+      "epoch": 0.762182491465602,
+      "grad_norm": 0.158203125,
+      "learning_rate": 0.02,
+      "loss": 1.5647,
+      "step": 8233
+    },
+    {
+      "epoch": 0.7622750679858821,
+      "grad_norm": 0.1533203125,
+      "learning_rate": 0.02,
+      "loss": 1.489,
+      "step": 8234
+    },
+    {
+      "epoch": 0.7623676445061621,
+      "grad_norm": 0.15234375,
+      "learning_rate": 0.02,
+      "loss": 1.5781,
+      "step": 8235
+    },
+    {
+      "epoch": 0.7624602210264422,
+      "grad_norm": 0.169921875,
+      "learning_rate": 0.02,
+      "loss": 1.5186,
+      "step": 8236
+    },
+    {
+      "epoch": 0.7625527975467222,
+      "grad_norm": 0.15234375,
+      "learning_rate": 0.02,
+      "loss": 1.4941,
+      "step": 8237
+    },
+    {
+      "epoch": 0.7626453740670023,
+      "grad_norm": 0.1669921875,
+      "learning_rate": 0.02,
+      "loss": 1.4895,
+      "step": 8238
+    },
+    {
+      "epoch": 0.7627379505872823,
+      "grad_norm": 0.1689453125,
+      "learning_rate": 0.02,
+      "loss": 1.4994,
+      "step": 8239
+    },
+    {
+      "epoch": 0.7628305271075624,
+      "grad_norm": 0.1513671875,
+      "learning_rate": 0.02,
+      "loss": 1.4994,
+      "step": 8240
+    },
+    {
+      "epoch": 0.7629231036278424,
+      "grad_norm": 0.1513671875,
+      "learning_rate": 0.02,
+      "loss": 1.5391,
+      "step": 8241
+    },
+    {
+      "epoch": 0.7630156801481224,
+      "grad_norm": 0.1630859375,
+      "learning_rate": 0.02,
+      "loss": 1.6066,
+      "step": 8242
+    },
+    {
+      "epoch": 0.7631082566684024,
+      "grad_norm": 0.1630859375,
+      "learning_rate": 0.02,
+      "loss": 1.4803,
+      "step": 8243
+    },
+    {
+      "epoch": 0.7632008331886825,
+      "grad_norm": 0.1650390625,
+      "learning_rate": 0.02,
+      "loss": 1.478,
+      "step": 8244
+    },
+    {
+      "epoch": 0.7632934097089625,
+      "grad_norm": 0.162109375,
+      "learning_rate": 0.02,
+      "loss": 1.4966,
+      "step": 8245
+    },
+    {
+      "epoch": 0.7633859862292426,
+      "grad_norm": 0.1708984375,
+      "learning_rate": 0.02,
+      "loss": 1.5117,
+      "step": 8246
+    },
+    {
+      "epoch": 0.7634785627495226,
+      "grad_norm": 0.146484375,
+      "learning_rate": 0.02,
+      "loss": 1.4854,
+      "step": 8247
+    },
+    {
+      "epoch": 0.7635711392698027,
+      "grad_norm": 0.1591796875,
+      "learning_rate": 0.02,
+      "loss": 1.4958,
+      "step": 8248
+    },
+    {
+      "epoch": 0.7636637157900827,
+      "grad_norm": 0.146484375,
+      "learning_rate": 0.02,
+      "loss": 1.5368,
+      "step": 8249
+    },
+    {
+      "epoch": 0.7637562923103628,
+      "grad_norm": 0.1572265625,
+      "learning_rate": 0.02,
+      "loss": 1.5358,
+      "step": 8250
+    },
+    {
+      "epoch": 0.7638488688306428,
+      "grad_norm": 0.140625,
+      "learning_rate": 0.02,
+      "loss": 1.5415,
+      "step": 8251
+    },
+    {
+      "epoch": 0.7639414453509229,
+      "grad_norm": 0.1484375,
+      "learning_rate": 0.02,
+      "loss": 1.4952,
+      "step": 8252
+    },
+    {
+      "epoch": 0.764034021871203,
+      "grad_norm": 0.162109375,
+      "learning_rate": 0.02,
+      "loss": 1.5651,
+      "step": 8253
+    },
+    {
+      "epoch": 0.764126598391483,
+      "grad_norm": 0.16015625,
+      "learning_rate": 0.02,
+      "loss": 1.4633,
+      "step": 8254
+    },
+    {
+      "epoch": 0.764219174911763,
+      "grad_norm": 0.1513671875,
+      "learning_rate": 0.02,
+      "loss": 1.4872,
+      "step": 8255
+    },
+    {
+      "epoch": 0.764311751432043,
+      "grad_norm": 0.1591796875,
+      "learning_rate": 0.02,
+      "loss": 1.5656,
+      "step": 8256
+    },
+    {
+      "epoch": 0.764404327952323,
+      "grad_norm": 0.1572265625,
+      "learning_rate": 0.02,
+      "loss": 1.5223,
+      "step": 8257
+    },
+    {
+      "epoch": 0.7644969044726031,
+      "grad_norm": 0.158203125,
+      "learning_rate": 0.02,
+      "loss": 1.4683,
+      "step": 8258
+    },
+    {
+      "epoch": 0.7645894809928832,
+      "grad_norm": 0.146484375,
+      "learning_rate": 0.02,
+      "loss": 1.5163,
+      "step": 8259
+    },
+    {
+      "epoch": 0.7646820575131632,
+      "grad_norm": 0.15625,
+      "learning_rate": 0.02,
+      "loss": 1.5576,
+      "step": 8260
+    },
+    {
+      "epoch": 0.7647746340334433,
+      "grad_norm": 0.150390625,
+      "learning_rate": 0.02,
+      "loss": 1.4957,
+      "step": 8261
+    },
+    {
+      "epoch": 0.7648672105537233,
+      "grad_norm": 0.1572265625,
+      "learning_rate": 0.02,
+      "loss": 1.5283,
+      "step": 8262
+    },
+    {
+      "epoch": 0.7649597870740034,
+      "grad_norm": 0.1513671875,
+      "learning_rate": 0.02,
+      "loss": 1.4794,
+      "step": 8263
+    },
+    {
+      "epoch": 0.7650523635942834,
+      "grad_norm": 0.15234375,
+      "learning_rate": 0.02,
+      "loss": 1.5364,
+      "step": 8264
+    },
+    {
+      "epoch": 0.7651449401145635,
+      "grad_norm": 0.1513671875,
+      "learning_rate": 0.02,
+      "loss": 1.4927,
+      "step": 8265
+    },
+    {
+      "epoch": 0.7652375166348435,
+      "grad_norm": 0.142578125,
+      "learning_rate": 0.02,
+      "loss": 1.4928,
+      "step": 8266
+    },
+    {
+      "epoch": 0.7653300931551236,
+      "grad_norm": 0.1611328125,
+      "learning_rate": 0.02,
+      "loss": 1.5165,
+      "step": 8267
+    },
+    {
+      "epoch": 0.7654226696754036,
+      "grad_norm": 0.14453125,
+      "learning_rate": 0.02,
+      "loss": 1.5218,
+      "step": 8268
+    },
+    {
+      "epoch": 0.7655152461956837,
+      "grad_norm": 0.150390625,
+      "learning_rate": 0.02,
+      "loss": 1.4871,
+      "step": 8269
+    },
+    {
+      "epoch": 0.7656078227159636,
+      "grad_norm": 0.146484375,
+      "learning_rate": 0.02,
+      "loss": 1.5259,
+      "step": 8270
+    },
+    {
+      "epoch": 0.7657003992362437,
+      "grad_norm": 0.1474609375,
+      "learning_rate": 0.02,
+      "loss": 1.4932,
+      "step": 8271
+    },
+    {
+      "epoch": 0.7657929757565237,
+      "grad_norm": 0.158203125,
+      "learning_rate": 0.02,
+      "loss": 1.5313,
+      "step": 8272
+    },
+    {
+      "epoch": 0.7658855522768038,
+      "grad_norm": 0.1591796875,
+      "learning_rate": 0.02,
+      "loss": 1.4526,
+      "step": 8273
+    },
+    {
+      "epoch": 0.7659781287970838,
+      "grad_norm": 0.1552734375,
+      "learning_rate": 0.02,
+      "loss": 1.5295,
+      "step": 8274
+    },
+    {
+      "epoch": 0.7660707053173639,
+      "grad_norm": 0.1494140625,
+      "learning_rate": 0.02,
+      "loss": 1.4714,
+      "step": 8275
+    },
+    {
+      "epoch": 0.7661632818376439,
+      "grad_norm": 0.15234375,
+      "learning_rate": 0.02,
+      "loss": 1.5429,
+      "step": 8276
+    },
+    {
+      "epoch": 0.766255858357924,
+      "grad_norm": 0.158203125,
+      "learning_rate": 0.02,
+      "loss": 1.562,
+      "step": 8277
+    },
+    {
+      "epoch": 0.766348434878204,
+      "grad_norm": 0.1337890625,
+      "learning_rate": 0.02,
+      "loss": 1.4638,
+      "step": 8278
+    },
+    {
+      "epoch": 0.7664410113984841,
+      "grad_norm": 0.142578125,
+      "learning_rate": 0.02,
+      "loss": 1.4323,
+      "step": 8279
+    },
+    {
+      "epoch": 0.7665335879187641,
+      "grad_norm": 0.146484375,
+      "learning_rate": 0.02,
+      "loss": 1.4811,
+      "step": 8280
+    },
+    {
+      "epoch": 0.7666261644390442,
+      "grad_norm": 0.1513671875,
+      "learning_rate": 0.02,
+      "loss": 1.4833,
+      "step": 8281
+    },
+    {
+      "epoch": 0.7667187409593242,
+      "grad_norm": 0.15234375,
+      "learning_rate": 0.02,
+      "loss": 1.5277,
+      "step": 8282
+    },
+    {
+      "epoch": 0.7668113174796042,
+      "grad_norm": 0.1572265625,
+      "learning_rate": 0.02,
+      "loss": 1.5386,
+      "step": 8283
+    },
+    {
+      "epoch": 0.7669038939998842,
+      "grad_norm": 0.1474609375,
+      "learning_rate": 0.02,
+      "loss": 1.4959,
+      "step": 8284
+    },
+    {
+      "epoch": 0.7669964705201643,
+      "grad_norm": 0.1494140625,
+      "learning_rate": 0.02,
+      "loss": 1.5,
+      "step": 8285
+    },
+    {
+      "epoch": 0.7670890470404443,
+      "grad_norm": 0.1396484375,
+      "learning_rate": 0.02,
+      "loss": 1.5193,
+      "step": 8286
+    },
+    {
+      "epoch": 0.7671816235607244,
+      "grad_norm": 0.1572265625,
+      "learning_rate": 0.02,
+      "loss": 1.5044,
+      "step": 8287
+    },
+    {
+      "epoch": 0.7672742000810044,
+      "grad_norm": 0.146484375,
+      "learning_rate": 0.02,
+      "loss": 1.5293,
+      "step": 8288
+    },
+    {
+      "epoch": 0.7673667766012845,
+      "grad_norm": 0.1474609375,
+      "learning_rate": 0.02,
+      "loss": 1.5066,
+      "step": 8289
+    },
+    {
+      "epoch": 0.7674593531215645,
+      "grad_norm": 0.142578125,
+      "learning_rate": 0.02,
+      "loss": 1.56,
+      "step": 8290
+    },
+    {
+      "epoch": 0.7675519296418446,
+      "grad_norm": 0.1552734375,
+      "learning_rate": 0.02,
+      "loss": 1.4439,
+      "step": 8291
+    },
+    {
+      "epoch": 0.7676445061621247,
+      "grad_norm": 0.1435546875,
+      "learning_rate": 0.02,
+      "loss": 1.5637,
+      "step": 8292
+    },
+    {
+      "epoch": 0.7677370826824047,
+      "grad_norm": 0.1474609375,
+      "learning_rate": 0.02,
+      "loss": 1.4088,
+      "step": 8293
+    },
+    {
+      "epoch": 0.7678296592026848,
+      "grad_norm": 0.154296875,
+      "learning_rate": 0.02,
+      "loss": 1.5042,
+      "step": 8294
+    },
+    {
+      "epoch": 0.7679222357229648,
+      "grad_norm": 0.15625,
+      "learning_rate": 0.02,
+      "loss": 1.45,
+      "step": 8295
+    },
+    {
+      "epoch": 0.7680148122432449,
+      "grad_norm": 0.1494140625,
+      "learning_rate": 0.02,
+      "loss": 1.4992,
+      "step": 8296
+    },
+    {
+      "epoch": 0.7681073887635248,
+      "grad_norm": 0.15234375,
+      "learning_rate": 0.02,
+      "loss": 1.4777,
+      "step": 8297
+    },
+    {
+      "epoch": 0.7681999652838049,
+      "grad_norm": 0.15625,
+      "learning_rate": 0.02,
+      "loss": 1.5171,
+      "step": 8298
+    },
+    {
+      "epoch": 0.7682925418040849,
+      "grad_norm": 0.15234375,
+      "learning_rate": 0.02,
+      "loss": 1.5121,
+      "step": 8299
+    },
+    {
+      "epoch": 0.768385118324365,
+      "grad_norm": 0.1640625,
+      "learning_rate": 0.02,
+      "loss": 1.5012,
+      "step": 8300
+    },
+    {
+      "epoch": 0.768477694844645,
+      "grad_norm": 0.1494140625,
+      "learning_rate": 0.02,
+      "loss": 1.5265,
+      "step": 8301
+    },
+    {
+      "epoch": 0.7685702713649251,
+      "grad_norm": 0.158203125,
+      "learning_rate": 0.02,
+      "loss": 1.4737,
+      "step": 8302
+    },
+    {
+      "epoch": 0.7686628478852051,
+      "grad_norm": 0.150390625,
+      "learning_rate": 0.02,
+      "loss": 1.4794,
+      "step": 8303
+    },
+    {
+      "epoch": 0.7687554244054852,
+      "grad_norm": 0.1494140625,
+      "learning_rate": 0.02,
+      "loss": 1.425,
+      "step": 8304
+    },
+    {
+      "epoch": 0.7688480009257652,
+      "grad_norm": 0.1552734375,
+      "learning_rate": 0.02,
+      "loss": 1.558,
+      "step": 8305
+    },
+    {
+      "epoch": 0.7689405774460453,
+      "grad_norm": 0.14453125,
+      "learning_rate": 0.02,
+      "loss": 1.5002,
+      "step": 8306
+    },
+    {
+      "epoch": 0.7690331539663253,
+      "grad_norm": 0.142578125,
+      "learning_rate": 0.02,
+      "loss": 1.4415,
+      "step": 8307
+    },
+    {
+      "epoch": 0.7691257304866054,
+      "grad_norm": 0.138671875,
+      "learning_rate": 0.02,
+      "loss": 1.4883,
+      "step": 8308
+    },
+    {
+      "epoch": 0.7692183070068854,
+      "grad_norm": 0.140625,
+      "learning_rate": 0.02,
+      "loss": 1.5649,
+      "step": 8309
+    },
+    {
+      "epoch": 0.7693108835271654,
+      "grad_norm": 0.142578125,
+      "learning_rate": 0.02,
+      "loss": 1.4815,
+      "step": 8310
+    },
+    {
+      "epoch": 0.7694034600474454,
+      "grad_norm": 0.146484375,
+      "learning_rate": 0.02,
+      "loss": 1.5594,
+      "step": 8311
+    },
+    {
+      "epoch": 0.7694960365677255,
+      "grad_norm": 0.138671875,
+      "learning_rate": 0.02,
+      "loss": 1.5476,
+      "step": 8312
+    },
+    {
+      "epoch": 0.7695886130880055,
+      "grad_norm": 0.1455078125,
+      "learning_rate": 0.02,
+      "loss": 1.4531,
+      "step": 8313
+    },
+    {
+      "epoch": 0.7696811896082856,
+      "grad_norm": 0.1552734375,
+      "learning_rate": 0.02,
+      "loss": 1.4617,
+      "step": 8314
+    },
+    {
+      "epoch": 0.7697737661285656,
+      "grad_norm": 0.154296875,
+      "learning_rate": 0.02,
+      "loss": 1.4397,
+      "step": 8315
+    },
+    {
+      "epoch": 0.7698663426488457,
+      "grad_norm": 0.1552734375,
+      "learning_rate": 0.02,
+      "loss": 1.4935,
+      "step": 8316
+    },
+    {
+      "epoch": 0.7699589191691257,
+      "grad_norm": 0.1591796875,
+      "learning_rate": 0.02,
+      "loss": 1.521,
+      "step": 8317
+    },
+    {
+      "epoch": 0.7700514956894058,
+      "grad_norm": 0.1533203125,
+      "learning_rate": 0.02,
+      "loss": 1.5152,
+      "step": 8318
+    },
+    {
+      "epoch": 0.7701440722096858,
+      "grad_norm": 0.154296875,
+      "learning_rate": 0.02,
+      "loss": 1.5036,
+      "step": 8319
+    },
+    {
+      "epoch": 0.7702366487299659,
+      "grad_norm": 0.1357421875,
+      "learning_rate": 0.02,
+      "loss": 1.4854,
+      "step": 8320
+    },
+    {
+      "epoch": 0.7703292252502459,
+      "grad_norm": 0.154296875,
+      "learning_rate": 0.02,
+      "loss": 1.4959,
+      "step": 8321
+    },
+    {
+      "epoch": 0.770421801770526,
+      "grad_norm": 0.150390625,
+      "learning_rate": 0.02,
+      "loss": 1.4967,
+      "step": 8322
+    },
+    {
+      "epoch": 0.770514378290806,
+      "grad_norm": 0.1494140625,
+      "learning_rate": 0.02,
+      "loss": 1.4781,
+      "step": 8323
+    },
+    {
+      "epoch": 0.770606954811086,
+      "grad_norm": 0.1435546875,
+      "learning_rate": 0.02,
+      "loss": 1.4436,
+      "step": 8324
+    },
+    {
+      "epoch": 0.770699531331366,
+      "grad_norm": 0.146484375,
+      "learning_rate": 0.02,
+      "loss": 1.5513,
+      "step": 8325
+    },
+    {
+      "epoch": 0.7707921078516461,
+      "grad_norm": 0.154296875,
+      "learning_rate": 0.02,
+      "loss": 1.5134,
+      "step": 8326
+    },
+    {
+      "epoch": 0.7708846843719261,
+      "grad_norm": 0.162109375,
+      "learning_rate": 0.02,
+      "loss": 1.5658,
+      "step": 8327
+    },
+    {
+      "epoch": 0.7709772608922062,
+      "grad_norm": 0.1494140625,
+      "learning_rate": 0.02,
+      "loss": 1.4923,
+      "step": 8328
+    },
+    {
+      "epoch": 0.7710698374124862,
+      "grad_norm": 0.1513671875,
+      "learning_rate": 0.02,
+      "loss": 1.5298,
+      "step": 8329
+    },
+    {
+      "epoch": 0.7711624139327663,
+      "grad_norm": 0.1474609375,
+      "learning_rate": 0.02,
+      "loss": 1.5486,
+      "step": 8330
+    },
+    {
+      "epoch": 0.7712549904530464,
+      "grad_norm": 0.15234375,
+      "learning_rate": 0.02,
+      "loss": 1.5024,
+      "step": 8331
+    },
+    {
+      "epoch": 0.7713475669733264,
+      "grad_norm": 0.158203125,
+      "learning_rate": 0.02,
+      "loss": 1.5533,
+      "step": 8332
+    },
+    {
+      "epoch": 0.7714401434936065,
+      "grad_norm": 0.1572265625,
+      "learning_rate": 0.02,
+      "loss": 1.5054,
+      "step": 8333
+    },
+    {
+      "epoch": 0.7715327200138865,
+      "grad_norm": 0.1591796875,
+      "learning_rate": 0.02,
+      "loss": 1.5323,
+      "step": 8334
+    },
+    {
+      "epoch": 0.7716252965341666,
+      "grad_norm": 0.154296875,
+      "learning_rate": 0.02,
+      "loss": 1.4798,
+      "step": 8335
+    },
+    {
+      "epoch": 0.7717178730544466,
+      "grad_norm": 0.1494140625,
+      "learning_rate": 0.02,
+      "loss": 1.5659,
+      "step": 8336
+    },
+    {
+      "epoch": 0.7718104495747267,
+      "grad_norm": 0.16015625,
+      "learning_rate": 0.02,
+      "loss": 1.5482,
+      "step": 8337
+    },
+    {
+      "epoch": 0.7719030260950066,
+      "grad_norm": 0.1494140625,
+      "learning_rate": 0.02,
+      "loss": 1.5392,
+      "step": 8338
+    },
+    {
+      "epoch": 0.7719956026152867,
+      "grad_norm": 0.1474609375,
+      "learning_rate": 0.02,
+      "loss": 1.4855,
+      "step": 8339
+    },
+    {
+      "epoch": 0.7720881791355667,
+      "grad_norm": 0.1572265625,
+      "learning_rate": 0.02,
+      "loss": 1.5189,
+      "step": 8340
+    },
+    {
+      "epoch": 0.7721807556558468,
+      "grad_norm": 0.15234375,
+      "learning_rate": 0.02,
+      "loss": 1.5509,
+      "step": 8341
+    },
+    {
+      "epoch": 0.7722733321761268,
+      "grad_norm": 0.16015625,
+      "learning_rate": 0.02,
+      "loss": 1.5038,
+      "step": 8342
+    },
+    {
+      "epoch": 0.7723659086964069,
+      "grad_norm": 0.1474609375,
+      "learning_rate": 0.02,
+      "loss": 1.5035,
+      "step": 8343
+    },
+    {
+      "epoch": 0.7724584852166869,
+      "grad_norm": 0.1484375,
+      "learning_rate": 0.02,
+      "loss": 1.4943,
+      "step": 8344
+    },
+    {
+      "epoch": 0.772551061736967,
+      "grad_norm": 0.162109375,
+      "learning_rate": 0.02,
+      "loss": 1.5159,
+      "step": 8345
+    },
+    {
+      "epoch": 0.772643638257247,
+      "grad_norm": 0.1591796875,
+      "learning_rate": 0.02,
+      "loss": 1.4708,
+      "step": 8346
+    },
+    {
+      "epoch": 0.7727362147775271,
+      "grad_norm": 0.1455078125,
+      "learning_rate": 0.02,
+      "loss": 1.4942,
+      "step": 8347
+    },
+    {
+      "epoch": 0.7728287912978071,
+      "grad_norm": 0.162109375,
+      "learning_rate": 0.02,
+      "loss": 1.4962,
+      "step": 8348
+    },
+    {
+      "epoch": 0.7729213678180872,
+      "grad_norm": 0.1484375,
+      "learning_rate": 0.02,
+      "loss": 1.4967,
+      "step": 8349
+    },
+    {
+      "epoch": 0.7730139443383672,
+      "grad_norm": 0.1572265625,
+      "learning_rate": 0.02,
+      "loss": 1.5142,
+      "step": 8350
+    },
+    {
+      "epoch": 0.7731065208586472,
+      "grad_norm": 0.14453125,
+      "learning_rate": 0.02,
+      "loss": 1.5225,
+      "step": 8351
+    },
+    {
+      "epoch": 0.7731990973789272,
+      "grad_norm": 0.1611328125,
+      "learning_rate": 0.02,
+      "loss": 1.4935,
+      "step": 8352
+    },
+    {
+      "epoch": 0.7732916738992073,
+      "grad_norm": 0.1513671875,
+      "learning_rate": 0.02,
+      "loss": 1.5258,
+      "step": 8353
+    },
+    {
+      "epoch": 0.7733842504194873,
+      "grad_norm": 0.1533203125,
+      "learning_rate": 0.02,
+      "loss": 1.4974,
+      "step": 8354
+    },
+    {
+      "epoch": 0.7734768269397674,
+      "grad_norm": 0.140625,
+      "learning_rate": 0.02,
+      "loss": 1.5185,
+      "step": 8355
+    },
+    {
+      "epoch": 0.7735694034600474,
+      "grad_norm": 0.1611328125,
+      "learning_rate": 0.02,
+      "loss": 1.4641,
+      "step": 8356
+    },
+    {
+      "epoch": 0.7736619799803275,
+      "grad_norm": 0.1474609375,
+      "learning_rate": 0.02,
+      "loss": 1.4758,
+      "step": 8357
+    },
+    {
+      "epoch": 0.7737545565006075,
+      "grad_norm": 0.1640625,
+      "learning_rate": 0.02,
+      "loss": 1.4909,
+      "step": 8358
+    },
+    {
+      "epoch": 0.7738471330208876,
+      "grad_norm": 0.1494140625,
+      "learning_rate": 0.02,
+      "loss": 1.5109,
+      "step": 8359
+    },
+    {
+      "epoch": 0.7739397095411676,
+      "grad_norm": 0.1689453125,
+      "learning_rate": 0.02,
+      "loss": 1.478,
+      "step": 8360
+    },
+    {
+      "epoch": 0.7740322860614477,
+      "grad_norm": 0.146484375,
+      "learning_rate": 0.02,
+      "loss": 1.4703,
+      "step": 8361
+    },
+    {
+      "epoch": 0.7741248625817277,
+      "grad_norm": 0.158203125,
+      "learning_rate": 0.02,
+      "loss": 1.5065,
+      "step": 8362
+    },
+    {
+      "epoch": 0.7742174391020078,
+      "grad_norm": 0.1474609375,
+      "learning_rate": 0.02,
+      "loss": 1.5053,
+      "step": 8363
+    },
+    {
+      "epoch": 0.7743100156222879,
+      "grad_norm": 0.1552734375,
+      "learning_rate": 0.02,
+      "loss": 1.5372,
+      "step": 8364
+    },
+    {
+      "epoch": 0.7744025921425678,
+      "grad_norm": 0.1494140625,
+      "learning_rate": 0.02,
+      "loss": 1.5704,
+      "step": 8365
+    },
+    {
+      "epoch": 0.7744951686628478,
+      "grad_norm": 0.1494140625,
+      "learning_rate": 0.02,
+      "loss": 1.4128,
+      "step": 8366
+    },
+    {
+      "epoch": 0.7745877451831279,
+      "grad_norm": 0.1494140625,
+      "learning_rate": 0.02,
+      "loss": 1.5471,
+      "step": 8367
+    },
+    {
+      "epoch": 0.774680321703408,
+      "grad_norm": 0.1611328125,
+      "learning_rate": 0.02,
+      "loss": 1.4766,
+      "step": 8368
+    },
+    {
+      "epoch": 0.774772898223688,
+      "grad_norm": 0.15234375,
+      "learning_rate": 0.02,
+      "loss": 1.4788,
+      "step": 8369
+    },
+    {
+      "epoch": 0.774865474743968,
+      "grad_norm": 0.1650390625,
+      "learning_rate": 0.02,
+      "loss": 1.5282,
+      "step": 8370
+    },
+    {
+      "epoch": 0.7749580512642481,
+      "grad_norm": 0.1513671875,
+      "learning_rate": 0.02,
+      "loss": 1.4357,
+      "step": 8371
+    },
+    {
+      "epoch": 0.7750506277845282,
+      "grad_norm": 0.1572265625,
+      "learning_rate": 0.02,
+      "loss": 1.4804,
+      "step": 8372
+    },
+    {
+      "epoch": 0.7751432043048082,
+      "grad_norm": 0.1484375,
+      "learning_rate": 0.02,
+      "loss": 1.4374,
+      "step": 8373
+    },
+    {
+      "epoch": 0.7752357808250883,
+      "grad_norm": 0.154296875,
+      "learning_rate": 0.02,
+      "loss": 1.4957,
+      "step": 8374
+    },
+    {
+      "epoch": 0.7753283573453683,
+      "grad_norm": 0.15234375,
+      "learning_rate": 0.02,
+      "loss": 1.463,
+      "step": 8375
+    },
+    {
+      "epoch": 0.7754209338656484,
+      "grad_norm": 0.1689453125,
+      "learning_rate": 0.02,
+      "loss": 1.53,
+      "step": 8376
+    },
+    {
+      "epoch": 0.7755135103859284,
+      "grad_norm": 0.1572265625,
+      "learning_rate": 0.02,
+      "loss": 1.526,
+      "step": 8377
+    },
+    {
+      "epoch": 0.7756060869062084,
+      "grad_norm": 0.3515625,
+      "learning_rate": 0.02,
+      "loss": 1.4427,
+      "step": 8378
+    },
+    {
+      "epoch": 0.7756986634264884,
+      "grad_norm": 0.12255859375,
+      "learning_rate": 0.02,
+      "loss": 1.4627,
+      "step": 8379
+    },
+    {
+      "epoch": 0.7757912399467685,
+      "grad_norm": 0.146484375,
+      "learning_rate": 0.02,
+      "loss": 1.537,
+      "step": 8380
+    },
+    {
+      "epoch": 0.7758838164670485,
+      "grad_norm": 0.154296875,
+      "learning_rate": 0.02,
+      "loss": 1.5075,
+      "step": 8381
+    },
+    {
+      "epoch": 0.7759763929873286,
+      "grad_norm": 0.1474609375,
+      "learning_rate": 0.02,
+      "loss": 1.5419,
+      "step": 8382
+    },
+    {
+      "epoch": 0.7760689695076086,
+      "grad_norm": 0.140625,
+      "learning_rate": 0.02,
+      "loss": 1.4672,
+      "step": 8383
+    },
+    {
+      "epoch": 0.7761615460278887,
+      "grad_norm": 0.15234375,
+      "learning_rate": 0.02,
+      "loss": 1.4756,
+      "step": 8384
+    },
+    {
+      "epoch": 0.7762541225481687,
+      "grad_norm": 0.1435546875,
+      "learning_rate": 0.02,
+      "loss": 1.4722,
+      "step": 8385
+    },
+    {
+      "epoch": 0.7763466990684488,
+      "grad_norm": 0.1572265625,
+      "learning_rate": 0.02,
+      "loss": 1.4154,
+      "step": 8386
+    },
+    {
+      "epoch": 0.7764392755887288,
+      "grad_norm": 0.1513671875,
+      "learning_rate": 0.02,
+      "loss": 1.4436,
+      "step": 8387
+    },
+    {
+      "epoch": 0.7765318521090089,
+      "grad_norm": 0.146484375,
+      "learning_rate": 0.02,
+      "loss": 1.4344,
+      "step": 8388
+    },
+    {
+      "epoch": 0.7766244286292889,
+      "grad_norm": 0.13671875,
+      "learning_rate": 0.02,
+      "loss": 1.448,
+      "step": 8389
+    },
+    {
+      "epoch": 0.776717005149569,
+      "grad_norm": 0.142578125,
+      "learning_rate": 0.02,
+      "loss": 1.5021,
+      "step": 8390
+    },
+    {
+      "epoch": 0.776809581669849,
+      "grad_norm": 0.1513671875,
+      "learning_rate": 0.02,
+      "loss": 1.5086,
+      "step": 8391
+    },
+    {
+      "epoch": 0.776902158190129,
+      "grad_norm": 0.150390625,
+      "learning_rate": 0.02,
+      "loss": 1.5089,
+      "step": 8392
+    },
+    {
+      "epoch": 0.776994734710409,
+      "grad_norm": 0.162109375,
+      "learning_rate": 0.02,
+      "loss": 1.5172,
+      "step": 8393
+    },
+    {
+      "epoch": 0.7770873112306891,
+      "grad_norm": 0.1533203125,
+      "learning_rate": 0.02,
+      "loss": 1.4369,
+      "step": 8394
+    },
+    {
+      "epoch": 0.7771798877509691,
+      "grad_norm": 0.154296875,
+      "learning_rate": 0.02,
+      "loss": 1.4622,
+      "step": 8395
+    },
+    {
+      "epoch": 0.7772724642712492,
+      "grad_norm": 0.1552734375,
+      "learning_rate": 0.02,
+      "loss": 1.5234,
+      "step": 8396
+    },
+    {
+      "epoch": 0.7773650407915292,
+      "grad_norm": 0.1728515625,
+      "learning_rate": 0.02,
+      "loss": 1.5108,
+      "step": 8397
+    },
+    {
+      "epoch": 0.7774576173118093,
+      "grad_norm": 0.1640625,
+      "learning_rate": 0.02,
+      "loss": 1.4724,
+      "step": 8398
+    },
+    {
+      "epoch": 0.7775501938320893,
+      "grad_norm": 0.17578125,
+      "learning_rate": 0.02,
+      "loss": 1.4967,
+      "step": 8399
+    },
+    {
+      "epoch": 0.7776427703523694,
+      "grad_norm": 0.1650390625,
+      "learning_rate": 0.02,
+      "loss": 1.4884,
+      "step": 8400
+    },
+    {
+      "epoch": 0.7777353468726494,
+      "grad_norm": 0.1640625,
+      "learning_rate": 0.02,
+      "loss": 1.5055,
+      "step": 8401
+    },
+    {
+      "epoch": 0.7778279233929295,
+      "grad_norm": 0.150390625,
+      "learning_rate": 0.02,
+      "loss": 1.5102,
+      "step": 8402
+    },
+    {
+      "epoch": 0.7779204999132096,
+      "grad_norm": 0.162109375,
+      "learning_rate": 0.02,
+      "loss": 1.4891,
+      "step": 8403
+    },
+    {
+      "epoch": 0.7780130764334896,
+      "grad_norm": 0.16015625,
+      "learning_rate": 0.02,
+      "loss": 1.5209,
+      "step": 8404
+    },
+    {
+      "epoch": 0.7781056529537697,
+      "grad_norm": 0.1533203125,
+      "learning_rate": 0.02,
+      "loss": 1.4553,
+      "step": 8405
+    },
+    {
+      "epoch": 0.7781982294740496,
+      "grad_norm": 0.154296875,
+      "learning_rate": 0.02,
+      "loss": 1.4831,
+      "step": 8406
+    },
+    {
+      "epoch": 0.7782908059943296,
+      "grad_norm": 0.1494140625,
+      "learning_rate": 0.02,
+      "loss": 1.4431,
+      "step": 8407
+    },
+    {
+      "epoch": 0.7783833825146097,
+      "grad_norm": 0.388671875,
+      "learning_rate": 0.02,
+      "loss": 1.5673,
+      "step": 8408
+    },
+    {
+      "epoch": 0.7784759590348898,
+      "grad_norm": 0.146484375,
+      "learning_rate": 0.02,
+      "loss": 1.5235,
+      "step": 8409
+    },
+    {
+      "epoch": 0.7785685355551698,
+      "grad_norm": 0.1533203125,
+      "learning_rate": 0.02,
+      "loss": 1.5116,
+      "step": 8410
+    },
+    {
+      "epoch": 0.7786611120754499,
+      "grad_norm": 0.138671875,
+      "learning_rate": 0.02,
+      "loss": 1.5374,
+      "step": 8411
+    },
+    {
+      "epoch": 0.7787536885957299,
+      "grad_norm": 0.1552734375,
+      "learning_rate": 0.02,
+      "loss": 1.4902,
+      "step": 8412
+    },
+    {
+      "epoch": 0.77884626511601,
+      "grad_norm": 0.154296875,
+      "learning_rate": 0.02,
+      "loss": 1.5186,
+      "step": 8413
+    },
+    {
+      "epoch": 0.77893884163629,
+      "grad_norm": 0.15625,
+      "learning_rate": 0.02,
+      "loss": 1.5028,
+      "step": 8414
+    },
+    {
+      "epoch": 0.7790314181565701,
+      "grad_norm": 0.1533203125,
+      "learning_rate": 0.02,
+      "loss": 1.507,
+      "step": 8415
+    },
+    {
+      "epoch": 0.7791239946768501,
+      "grad_norm": 0.158203125,
+      "learning_rate": 0.02,
+      "loss": 1.5279,
+      "step": 8416
+    },
+    {
+      "epoch": 0.7792165711971302,
+      "grad_norm": 0.1474609375,
+      "learning_rate": 0.02,
+      "loss": 1.5554,
+      "step": 8417
+    },
+    {
+      "epoch": 0.7793091477174102,
+      "grad_norm": 0.158203125,
+      "learning_rate": 0.02,
+      "loss": 1.4855,
+      "step": 8418
+    },
+    {
+      "epoch": 0.7794017242376902,
+      "grad_norm": 0.1728515625,
+      "learning_rate": 0.02,
+      "loss": 1.5268,
+      "step": 8419
+    },
+    {
+      "epoch": 0.7794943007579702,
+      "grad_norm": 0.1513671875,
+      "learning_rate": 0.02,
+      "loss": 1.4641,
+      "step": 8420
+    },
+    {
+      "epoch": 0.7795868772782503,
+      "grad_norm": 0.1533203125,
+      "learning_rate": 0.02,
+      "loss": 1.4701,
+      "step": 8421
+    },
+    {
+      "epoch": 0.7796794537985303,
+      "grad_norm": 0.1474609375,
+      "learning_rate": 0.02,
+      "loss": 1.4691,
+      "step": 8422
+    },
+    {
+      "epoch": 0.7797720303188104,
+      "grad_norm": 0.1533203125,
+      "learning_rate": 0.02,
+      "loss": 1.5484,
+      "step": 8423
+    },
+    {
+      "epoch": 0.7798646068390904,
+      "grad_norm": 0.150390625,
+      "learning_rate": 0.02,
+      "loss": 1.508,
+      "step": 8424
+    },
+    {
+      "epoch": 0.7799571833593705,
+      "grad_norm": 0.158203125,
+      "learning_rate": 0.02,
+      "loss": 1.5084,
+      "step": 8425
+    },
+    {
+      "epoch": 0.7800497598796505,
+      "grad_norm": 0.1513671875,
+      "learning_rate": 0.02,
+      "loss": 1.48,
+      "step": 8426
+    },
+    {
+      "epoch": 0.7801423363999306,
+      "grad_norm": 0.16015625,
+      "learning_rate": 0.02,
+      "loss": 1.5545,
+      "step": 8427
+    },
+    {
+      "epoch": 0.7802349129202106,
+      "grad_norm": 0.154296875,
+      "learning_rate": 0.02,
+      "loss": 1.4726,
+      "step": 8428
+    },
+    {
+      "epoch": 0.7803274894404907,
+      "grad_norm": 0.1640625,
+      "learning_rate": 0.02,
+      "loss": 1.5056,
+      "step": 8429
+    },
+    {
+      "epoch": 0.7804200659607707,
+      "grad_norm": 0.1396484375,
+      "learning_rate": 0.02,
+      "loss": 1.4729,
+      "step": 8430
+    },
+    {
+      "epoch": 0.7805126424810508,
+      "grad_norm": 0.1591796875,
+      "learning_rate": 0.02,
+      "loss": 1.4941,
+      "step": 8431
+    },
+    {
+      "epoch": 0.7806052190013308,
+      "grad_norm": 0.14453125,
+      "learning_rate": 0.02,
+      "loss": 1.5018,
+      "step": 8432
+    },
+    {
+      "epoch": 0.7806977955216108,
+      "grad_norm": 0.1611328125,
+      "learning_rate": 0.02,
+      "loss": 1.5646,
+      "step": 8433
+    },
+    {
+      "epoch": 0.7807903720418908,
+      "grad_norm": 0.146484375,
+      "learning_rate": 0.02,
+      "loss": 1.5323,
+      "step": 8434
+    },
+    {
+      "epoch": 0.7808829485621709,
+      "grad_norm": 0.16015625,
+      "learning_rate": 0.02,
+      "loss": 1.5292,
+      "step": 8435
+    },
+    {
+      "epoch": 0.7809755250824509,
+      "grad_norm": 0.14453125,
+      "learning_rate": 0.02,
+      "loss": 1.4703,
+      "step": 8436
+    },
+    {
+      "epoch": 0.781068101602731,
+      "grad_norm": 0.1494140625,
+      "learning_rate": 0.02,
+      "loss": 1.4407,
+      "step": 8437
+    },
+    {
+      "epoch": 0.781160678123011,
+      "grad_norm": 0.150390625,
+      "learning_rate": 0.02,
+      "loss": 1.4526,
+      "step": 8438
+    },
+    {
+      "epoch": 0.7812532546432911,
+      "grad_norm": 0.1748046875,
+      "learning_rate": 0.02,
+      "loss": 1.477,
+      "step": 8439
+    },
+    {
+      "epoch": 0.7813458311635711,
+      "grad_norm": 0.154296875,
+      "learning_rate": 0.02,
+      "loss": 1.4643,
+      "step": 8440
+    },
+    {
+      "epoch": 0.7814384076838512,
+      "grad_norm": 0.1650390625,
+      "learning_rate": 0.02,
+      "loss": 1.5659,
+      "step": 8441
+    },
+    {
+      "epoch": 0.7815309842041313,
+      "grad_norm": 0.14453125,
+      "learning_rate": 0.02,
+      "loss": 1.5273,
+      "step": 8442
+    },
+    {
+      "epoch": 0.7816235607244113,
+      "grad_norm": 0.1689453125,
+      "learning_rate": 0.02,
+      "loss": 1.4503,
+      "step": 8443
+    },
+    {
+      "epoch": 0.7817161372446914,
+      "grad_norm": 0.14453125,
+      "learning_rate": 0.02,
+      "loss": 1.561,
+      "step": 8444
+    },
+    {
+      "epoch": 0.7818087137649714,
+      "grad_norm": 0.1572265625,
+      "learning_rate": 0.02,
+      "loss": 1.4825,
+      "step": 8445
+    },
+    {
+      "epoch": 0.7819012902852513,
+      "grad_norm": 0.154296875,
+      "learning_rate": 0.02,
+      "loss": 1.5399,
+      "step": 8446
+    },
+    {
+      "epoch": 0.7819938668055314,
+      "grad_norm": 0.171875,
+      "learning_rate": 0.02,
+      "loss": 1.565,
+      "step": 8447
+    },
+    {
+      "epoch": 0.7820864433258115,
+      "grad_norm": 0.1611328125,
+      "learning_rate": 0.02,
+      "loss": 1.5671,
+      "step": 8448
+    },
+    {
+      "epoch": 0.7821790198460915,
+      "grad_norm": 0.158203125,
+      "learning_rate": 0.02,
+      "loss": 1.482,
+      "step": 8449
+    },
+    {
+      "epoch": 0.7822715963663716,
+      "grad_norm": 0.1552734375,
+      "learning_rate": 0.02,
+      "loss": 1.4606,
+      "step": 8450
+    },
+    {
+      "epoch": 0.7823641728866516,
+      "grad_norm": 0.1513671875,
+      "learning_rate": 0.02,
+      "loss": 1.4869,
+      "step": 8451
+    },
+    {
+      "epoch": 0.7824567494069317,
+      "grad_norm": 0.1630859375,
+      "learning_rate": 0.02,
+      "loss": 1.4784,
+      "step": 8452
+    },
+    {
+      "epoch": 0.7825493259272117,
+      "grad_norm": 0.146484375,
+      "learning_rate": 0.02,
+      "loss": 1.5203,
+      "step": 8453
+    },
+    {
+      "epoch": 0.7826419024474918,
+      "grad_norm": 0.1494140625,
+      "learning_rate": 0.02,
+      "loss": 1.5424,
+      "step": 8454
+    },
+    {
+      "epoch": 0.7827344789677718,
+      "grad_norm": 0.146484375,
+      "learning_rate": 0.02,
+      "loss": 1.5288,
+      "step": 8455
+    },
+    {
+      "epoch": 0.7828270554880519,
+      "grad_norm": 0.1533203125,
+      "learning_rate": 0.02,
+      "loss": 1.5281,
+      "step": 8456
+    },
+    {
+      "epoch": 0.7829196320083319,
+      "grad_norm": 0.1484375,
+      "learning_rate": 0.02,
+      "loss": 1.4796,
+      "step": 8457
+    },
+    {
+      "epoch": 0.783012208528612,
+      "grad_norm": 0.146484375,
+      "learning_rate": 0.02,
+      "loss": 1.5499,
+      "step": 8458
+    },
+    {
+      "epoch": 0.783104785048892,
+      "grad_norm": 0.1474609375,
+      "learning_rate": 0.02,
+      "loss": 1.4776,
+      "step": 8459
+    },
+    {
+      "epoch": 0.783197361569172,
+      "grad_norm": 0.150390625,
+      "learning_rate": 0.02,
+      "loss": 1.4744,
+      "step": 8460
+    },
+    {
+      "epoch": 0.783289938089452,
+      "grad_norm": 0.1455078125,
+      "learning_rate": 0.02,
+      "loss": 1.5027,
+      "step": 8461
+    },
+    {
+      "epoch": 0.7833825146097321,
+      "grad_norm": 0.1474609375,
+      "learning_rate": 0.02,
+      "loss": 1.4789,
+      "step": 8462
+    },
+    {
+      "epoch": 0.7834750911300121,
+      "grad_norm": 0.1611328125,
+      "learning_rate": 0.02,
+      "loss": 1.4892,
+      "step": 8463
+    },
+    {
+      "epoch": 0.7835676676502922,
+      "grad_norm": 0.1474609375,
+      "learning_rate": 0.02,
+      "loss": 1.4736,
+      "step": 8464
+    },
+    {
+      "epoch": 0.7836602441705722,
+      "grad_norm": 0.1591796875,
+      "learning_rate": 0.02,
+      "loss": 1.5268,
+      "step": 8465
+    },
+    {
+      "epoch": 0.7837528206908523,
+      "grad_norm": 0.1572265625,
+      "learning_rate": 0.02,
+      "loss": 1.5043,
+      "step": 8466
+    },
+    {
+      "epoch": 0.7838453972111323,
+      "grad_norm": 0.1591796875,
+      "learning_rate": 0.02,
+      "loss": 1.4441,
+      "step": 8467
+    },
+    {
+      "epoch": 0.7839379737314124,
+      "grad_norm": 0.140625,
+      "learning_rate": 0.02,
+      "loss": 1.5263,
+      "step": 8468
+    },
+    {
+      "epoch": 0.7840305502516924,
+      "grad_norm": 0.1591796875,
+      "learning_rate": 0.02,
+      "loss": 1.4286,
+      "step": 8469
+    },
+    {
+      "epoch": 0.7841231267719725,
+      "grad_norm": 0.1484375,
+      "learning_rate": 0.02,
+      "loss": 1.5384,
+      "step": 8470
+    },
+    {
+      "epoch": 0.7842157032922525,
+      "grad_norm": 0.1591796875,
+      "learning_rate": 0.02,
+      "loss": 1.4539,
+      "step": 8471
+    },
+    {
+      "epoch": 0.7843082798125326,
+      "grad_norm": 0.146484375,
+      "learning_rate": 0.02,
+      "loss": 1.4904,
+      "step": 8472
+    },
+    {
+      "epoch": 0.7844008563328125,
+      "grad_norm": 0.158203125,
+      "learning_rate": 0.02,
+      "loss": 1.5283,
+      "step": 8473
+    },
+    {
+      "epoch": 0.7844934328530926,
+      "grad_norm": 0.14453125,
+      "learning_rate": 0.02,
+      "loss": 1.5529,
+      "step": 8474
+    },
+    {
+      "epoch": 0.7845860093733726,
+      "grad_norm": 0.1533203125,
+      "learning_rate": 0.02,
+      "loss": 1.5209,
+      "step": 8475
+    },
+    {
+      "epoch": 0.7846785858936527,
+      "grad_norm": 0.150390625,
+      "learning_rate": 0.02,
+      "loss": 1.4676,
+      "step": 8476
+    },
+    {
+      "epoch": 0.7847711624139327,
+      "grad_norm": 0.17578125,
+      "learning_rate": 0.02,
+      "loss": 1.4809,
+      "step": 8477
+    },
+    {
+      "epoch": 0.7848637389342128,
+      "grad_norm": 0.1455078125,
+      "learning_rate": 0.02,
+      "loss": 1.5709,
+      "step": 8478
+    },
+    {
+      "epoch": 0.7849563154544928,
+      "grad_norm": 0.1650390625,
+      "learning_rate": 0.02,
+      "loss": 1.4902,
+      "step": 8479
+    },
+    {
+      "epoch": 0.7850488919747729,
+      "grad_norm": 0.140625,
+      "learning_rate": 0.02,
+      "loss": 1.5139,
+      "step": 8480
+    },
+    {
+      "epoch": 0.785141468495053,
+      "grad_norm": 0.15625,
+      "learning_rate": 0.02,
+      "loss": 1.5508,
+      "step": 8481
+    },
+    {
+      "epoch": 0.785234045015333,
+      "grad_norm": 0.1357421875,
+      "learning_rate": 0.02,
+      "loss": 1.4955,
+      "step": 8482
+    },
+    {
+      "epoch": 0.7853266215356131,
+      "grad_norm": 0.154296875,
+      "learning_rate": 0.02,
+      "loss": 1.4918,
+      "step": 8483
+    },
+    {
+      "epoch": 0.7854191980558931,
+      "grad_norm": 0.146484375,
+      "learning_rate": 0.02,
+      "loss": 1.4883,
+      "step": 8484
+    },
+    {
+      "epoch": 0.7855117745761732,
+      "grad_norm": 0.1533203125,
+      "learning_rate": 0.02,
+      "loss": 1.5406,
+      "step": 8485
+    },
+    {
+      "epoch": 0.7856043510964532,
+      "grad_norm": 0.1455078125,
+      "learning_rate": 0.02,
+      "loss": 1.4865,
+      "step": 8486
+    },
+    {
+      "epoch": 0.7856969276167332,
+      "grad_norm": 0.158203125,
+      "learning_rate": 0.02,
+      "loss": 1.642,
+      "step": 8487
+    },
+    {
+      "epoch": 0.7857895041370132,
+      "grad_norm": 0.1572265625,
+      "learning_rate": 0.02,
+      "loss": 1.4572,
+      "step": 8488
+    },
+    {
+      "epoch": 0.7858820806572933,
+      "grad_norm": 0.1728515625,
+      "learning_rate": 0.02,
+      "loss": 1.5633,
+      "step": 8489
+    },
+    {
+      "epoch": 0.7859746571775733,
+      "grad_norm": 0.1484375,
+      "learning_rate": 0.02,
+      "loss": 1.4455,
+      "step": 8490
+    },
+    {
+      "epoch": 0.7860672336978534,
+      "grad_norm": 0.1513671875,
+      "learning_rate": 0.02,
+      "loss": 1.5091,
+      "step": 8491
+    },
+    {
+      "epoch": 0.7861598102181334,
+      "grad_norm": 0.146484375,
+      "learning_rate": 0.02,
+      "loss": 1.4661,
+      "step": 8492
+    },
+    {
+      "epoch": 0.7862523867384135,
+      "grad_norm": 0.1533203125,
+      "learning_rate": 0.02,
+      "loss": 1.4389,
+      "step": 8493
+    },
+    {
+      "epoch": 0.7863449632586935,
+      "grad_norm": 0.1552734375,
+      "learning_rate": 0.02,
+      "loss": 1.604,
+      "step": 8494
+    },
+    {
+      "epoch": 0.7864375397789736,
+      "grad_norm": 0.1572265625,
+      "learning_rate": 0.02,
+      "loss": 1.5328,
+      "step": 8495
+    },
+    {
+      "epoch": 0.7865301162992536,
+      "grad_norm": 0.1533203125,
+      "learning_rate": 0.02,
+      "loss": 1.5301,
+      "step": 8496
+    },
+    {
+      "epoch": 0.7866226928195337,
+      "grad_norm": 0.1484375,
+      "learning_rate": 0.02,
+      "loss": 1.4721,
+      "step": 8497
+    },
+    {
+      "epoch": 0.7867152693398137,
+      "grad_norm": 0.154296875,
+      "learning_rate": 0.02,
+      "loss": 1.5106,
+      "step": 8498
+    },
+    {
+      "epoch": 0.7868078458600938,
+      "grad_norm": 0.1474609375,
+      "learning_rate": 0.02,
+      "loss": 1.4761,
+      "step": 8499
+    },
+    {
+      "epoch": 0.7869004223803738,
+      "grad_norm": 0.1474609375,
+      "learning_rate": 0.02,
+      "loss": 1.5231,
+      "step": 8500
+    },
+    {
+      "epoch": 0.7869929989006538,
+      "grad_norm": 0.1513671875,
+      "learning_rate": 0.02,
+      "loss": 1.4817,
+      "step": 8501
+    },
+    {
+      "epoch": 0.7870855754209338,
+      "grad_norm": 0.1611328125,
+      "learning_rate": 0.02,
+      "loss": 1.4869,
+      "step": 8502
+    },
+    {
+      "epoch": 0.7871781519412139,
+      "grad_norm": 0.154296875,
+      "learning_rate": 0.02,
+      "loss": 1.5372,
+      "step": 8503
+    },
+    {
+      "epoch": 0.7872707284614939,
+      "grad_norm": 0.15625,
+      "learning_rate": 0.02,
+      "loss": 1.5038,
+      "step": 8504
+    },
+    {
+      "epoch": 0.787363304981774,
+      "grad_norm": 0.16796875,
+      "learning_rate": 0.02,
+      "loss": 1.5184,
+      "step": 8505
+    },
+    {
+      "epoch": 0.787455881502054,
+      "grad_norm": 0.1591796875,
+      "learning_rate": 0.02,
+      "loss": 1.5266,
+      "step": 8506
+    },
+    {
+      "epoch": 0.7875484580223341,
+      "grad_norm": 0.1484375,
+      "learning_rate": 0.02,
+      "loss": 1.4656,
+      "step": 8507
+    },
+    {
+      "epoch": 0.7876410345426141,
+      "grad_norm": 0.1533203125,
+      "learning_rate": 0.02,
+      "loss": 1.492,
+      "step": 8508
+    },
+    {
+      "epoch": 0.7877336110628942,
+      "grad_norm": 0.1572265625,
+      "learning_rate": 0.02,
+      "loss": 1.5209,
+      "step": 8509
+    },
+    {
+      "epoch": 0.7878261875831742,
+      "grad_norm": 0.1533203125,
+      "learning_rate": 0.02,
+      "loss": 1.549,
+      "step": 8510
+    },
+    {
+      "epoch": 0.7879187641034543,
+      "grad_norm": 0.1484375,
+      "learning_rate": 0.02,
+      "loss": 1.5816,
+      "step": 8511
+    },
+    {
+      "epoch": 0.7880113406237343,
+      "grad_norm": 0.142578125,
+      "learning_rate": 0.02,
+      "loss": 1.4582,
+      "step": 8512
+    },
+    {
+      "epoch": 0.7881039171440144,
+      "grad_norm": 0.158203125,
+      "learning_rate": 0.02,
+      "loss": 1.453,
+      "step": 8513
+    },
+    {
+      "epoch": 0.7881964936642943,
+      "grad_norm": 0.1484375,
+      "learning_rate": 0.02,
+      "loss": 1.468,
+      "step": 8514
+    },
+    {
+      "epoch": 0.7882890701845744,
+      "grad_norm": 0.1611328125,
+      "learning_rate": 0.02,
+      "loss": 1.5476,
+      "step": 8515
+    },
+    {
+      "epoch": 0.7883816467048544,
+      "grad_norm": 0.154296875,
+      "learning_rate": 0.02,
+      "loss": 1.4411,
+      "step": 8516
+    },
+    {
+      "epoch": 0.7884742232251345,
+      "grad_norm": 0.15625,
+      "learning_rate": 0.02,
+      "loss": 1.4738,
+      "step": 8517
+    },
+    {
+      "epoch": 0.7885667997454145,
+      "grad_norm": 0.1513671875,
+      "learning_rate": 0.02,
+      "loss": 1.4833,
+      "step": 8518
+    },
+    {
+      "epoch": 0.7886593762656946,
+      "grad_norm": 0.1513671875,
+      "learning_rate": 0.02,
+      "loss": 1.5047,
+      "step": 8519
+    },
+    {
+      "epoch": 0.7887519527859747,
+      "grad_norm": 0.1474609375,
+      "learning_rate": 0.02,
+      "loss": 1.5728,
+      "step": 8520
+    },
+    {
+      "epoch": 0.7888445293062547,
+      "grad_norm": 0.154296875,
+      "learning_rate": 0.02,
+      "loss": 1.5038,
+      "step": 8521
+    },
+    {
+      "epoch": 0.7889371058265348,
+      "grad_norm": 0.1533203125,
+      "learning_rate": 0.02,
+      "loss": 1.4594,
+      "step": 8522
+    },
+    {
+      "epoch": 0.7890296823468148,
+      "grad_norm": 0.1572265625,
+      "learning_rate": 0.02,
+      "loss": 1.4664,
+      "step": 8523
+    },
+    {
+      "epoch": 0.7891222588670949,
+      "grad_norm": 0.1435546875,
+      "learning_rate": 0.02,
+      "loss": 1.495,
+      "step": 8524
+    },
+    {
+      "epoch": 0.7892148353873749,
+      "grad_norm": 0.15234375,
+      "learning_rate": 0.02,
+      "loss": 1.5313,
+      "step": 8525
+    },
+    {
+      "epoch": 0.789307411907655,
+      "grad_norm": 0.1591796875,
+      "learning_rate": 0.02,
+      "loss": 1.5138,
+      "step": 8526
+    },
+    {
+      "epoch": 0.789399988427935,
+      "grad_norm": 0.169921875,
+      "learning_rate": 0.02,
+      "loss": 1.4777,
+      "step": 8527
+    },
+    {
+      "epoch": 0.789492564948215,
+      "grad_norm": 0.1650390625,
+      "learning_rate": 0.02,
+      "loss": 1.4908,
+      "step": 8528
+    },
+    {
+      "epoch": 0.789585141468495,
+      "grad_norm": 0.1650390625,
+      "learning_rate": 0.02,
+      "loss": 1.4378,
+      "step": 8529
+    },
+    {
+      "epoch": 0.7896777179887751,
+      "grad_norm": 0.1767578125,
+      "learning_rate": 0.02,
+      "loss": 1.51,
+      "step": 8530
+    },
+    {
+      "epoch": 0.7897702945090551,
+      "grad_norm": 0.1435546875,
+      "learning_rate": 0.02,
+      "loss": 1.5421,
+      "step": 8531
+    },
+    {
+      "epoch": 0.7898628710293352,
+      "grad_norm": 0.1591796875,
+      "learning_rate": 0.02,
+      "loss": 1.5396,
+      "step": 8532
+    },
+    {
+      "epoch": 0.7899554475496152,
+      "grad_norm": 0.15625,
+      "learning_rate": 0.02,
+      "loss": 1.564,
+      "step": 8533
+    },
+    {
+      "epoch": 0.7900480240698953,
+      "grad_norm": 0.16015625,
+      "learning_rate": 0.02,
+      "loss": 1.526,
+      "step": 8534
+    },
+    {
+      "epoch": 0.7901406005901753,
+      "grad_norm": 0.15234375,
+      "learning_rate": 0.02,
+      "loss": 1.4869,
+      "step": 8535
+    },
+    {
+      "epoch": 0.7902331771104554,
+      "grad_norm": 0.166015625,
+      "learning_rate": 0.02,
+      "loss": 1.5717,
+      "step": 8536
+    },
+    {
+      "epoch": 0.7903257536307354,
+      "grad_norm": 0.1337890625,
+      "learning_rate": 0.02,
+      "loss": 1.4467,
+      "step": 8537
+    },
+    {
+      "epoch": 0.7904183301510155,
+      "grad_norm": 0.1572265625,
+      "learning_rate": 0.02,
+      "loss": 1.5365,
+      "step": 8538
+    },
+    {
+      "epoch": 0.7905109066712955,
+      "grad_norm": 0.1474609375,
+      "learning_rate": 0.02,
+      "loss": 1.4843,
+      "step": 8539
+    },
+    {
+      "epoch": 0.7906034831915756,
+      "grad_norm": 0.1572265625,
+      "learning_rate": 0.02,
+      "loss": 1.4832,
+      "step": 8540
+    },
+    {
+      "epoch": 0.7906960597118555,
+      "grad_norm": 0.1513671875,
+      "learning_rate": 0.02,
+      "loss": 1.4904,
+      "step": 8541
+    },
+    {
+      "epoch": 0.7907886362321356,
+      "grad_norm": 0.1455078125,
+      "learning_rate": 0.02,
+      "loss": 1.5024,
+      "step": 8542
+    },
+    {
+      "epoch": 0.7908812127524156,
+      "grad_norm": 0.1484375,
+      "learning_rate": 0.02,
+      "loss": 1.4519,
+      "step": 8543
+    },
+    {
+      "epoch": 0.7909737892726957,
+      "grad_norm": 0.1513671875,
+      "learning_rate": 0.02,
+      "loss": 1.4444,
+      "step": 8544
+    },
+    {
+      "epoch": 0.7910663657929757,
+      "grad_norm": 0.162109375,
+      "learning_rate": 0.02,
+      "loss": 1.4894,
+      "step": 8545
+    },
+    {
+      "epoch": 0.7911589423132558,
+      "grad_norm": 0.1484375,
+      "learning_rate": 0.02,
+      "loss": 1.5045,
+      "step": 8546
+    },
+    {
+      "epoch": 0.7912515188335358,
+      "grad_norm": 0.1494140625,
+      "learning_rate": 0.02,
+      "loss": 1.4772,
+      "step": 8547
+    },
+    {
+      "epoch": 0.7913440953538159,
+      "grad_norm": 0.15234375,
+      "learning_rate": 0.02,
+      "loss": 1.5379,
+      "step": 8548
+    },
+    {
+      "epoch": 0.7914366718740959,
+      "grad_norm": 0.1416015625,
+      "learning_rate": 0.02,
+      "loss": 1.5206,
+      "step": 8549
+    },
+    {
+      "epoch": 0.791529248394376,
+      "grad_norm": 0.1552734375,
+      "learning_rate": 0.02,
+      "loss": 1.5636,
+      "step": 8550
+    },
+    {
+      "epoch": 0.791621824914656,
+      "grad_norm": 0.150390625,
+      "learning_rate": 0.02,
+      "loss": 1.5059,
+      "step": 8551
+    },
+    {
+      "epoch": 0.7917144014349361,
+      "grad_norm": 0.1484375,
+      "learning_rate": 0.02,
+      "loss": 1.4397,
+      "step": 8552
+    },
+    {
+      "epoch": 0.7918069779552162,
+      "grad_norm": 0.1552734375,
+      "learning_rate": 0.02,
+      "loss": 1.4993,
+      "step": 8553
+    },
+    {
+      "epoch": 0.7918995544754962,
+      "grad_norm": 0.1455078125,
+      "learning_rate": 0.02,
+      "loss": 1.5401,
+      "step": 8554
+    },
+    {
+      "epoch": 0.7919921309957761,
+      "grad_norm": 0.1513671875,
+      "learning_rate": 0.02,
+      "loss": 1.4867,
+      "step": 8555
+    },
+    {
+      "epoch": 0.7920847075160562,
+      "grad_norm": 0.1396484375,
+      "learning_rate": 0.02,
+      "loss": 1.493,
+      "step": 8556
+    },
+    {
+      "epoch": 0.7921772840363362,
+      "grad_norm": 0.1455078125,
+      "learning_rate": 0.02,
+      "loss": 1.5202,
+      "step": 8557
+    },
+    {
+      "epoch": 0.7922698605566163,
+      "grad_norm": 0.1611328125,
+      "learning_rate": 0.02,
+      "loss": 1.5474,
+      "step": 8558
+    },
+    {
+      "epoch": 0.7923624370768964,
+      "grad_norm": 0.1513671875,
+      "learning_rate": 0.02,
+      "loss": 1.4925,
+      "step": 8559
+    },
+    {
+      "epoch": 0.7924550135971764,
+      "grad_norm": 0.1494140625,
+      "learning_rate": 0.02,
+      "loss": 1.5524,
+      "step": 8560
+    },
+    {
+      "epoch": 0.7925475901174565,
+      "grad_norm": 0.1572265625,
+      "learning_rate": 0.02,
+      "loss": 1.5671,
+      "step": 8561
+    },
+    {
+      "epoch": 0.7926401666377365,
+      "grad_norm": 0.1533203125,
+      "learning_rate": 0.02,
+      "loss": 1.5474,
+      "step": 8562
+    },
+    {
+      "epoch": 0.7927327431580166,
+      "grad_norm": 0.16796875,
+      "learning_rate": 0.02,
+      "loss": 1.5382,
+      "step": 8563
+    },
+    {
+      "epoch": 0.7928253196782966,
+      "grad_norm": 0.1630859375,
+      "learning_rate": 0.02,
+      "loss": 1.5123,
+      "step": 8564
+    },
+    {
+      "epoch": 0.7929178961985767,
+      "grad_norm": 0.1474609375,
+      "learning_rate": 0.02,
+      "loss": 1.5525,
+      "step": 8565
+    },
+    {
+      "epoch": 0.7930104727188567,
+      "grad_norm": 0.158203125,
+      "learning_rate": 0.02,
+      "loss": 1.511,
+      "step": 8566
+    },
+    {
+      "epoch": 0.7931030492391368,
+      "grad_norm": 0.150390625,
+      "learning_rate": 0.02,
+      "loss": 1.499,
+      "step": 8567
+    },
+    {
+      "epoch": 0.7931956257594168,
+      "grad_norm": 0.1591796875,
+      "learning_rate": 0.02,
+      "loss": 1.5413,
+      "step": 8568
+    },
+    {
+      "epoch": 0.7932882022796968,
+      "grad_norm": 0.15625,
+      "learning_rate": 0.02,
+      "loss": 1.4364,
+      "step": 8569
+    },
+    {
+      "epoch": 0.7933807787999768,
+      "grad_norm": 0.1474609375,
+      "learning_rate": 0.02,
+      "loss": 1.5175,
+      "step": 8570
+    },
+    {
+      "epoch": 0.7934733553202569,
+      "grad_norm": 0.142578125,
+      "learning_rate": 0.02,
+      "loss": 1.5087,
+      "step": 8571
+    },
+    {
+      "epoch": 0.7935659318405369,
+      "grad_norm": 0.15625,
+      "learning_rate": 0.02,
+      "loss": 1.547,
+      "step": 8572
+    },
+    {
+      "epoch": 0.793658508360817,
+      "grad_norm": 0.1484375,
+      "learning_rate": 0.02,
+      "loss": 1.4624,
+      "step": 8573
+    },
+    {
+      "epoch": 0.793751084881097,
+      "grad_norm": 0.1494140625,
+      "learning_rate": 0.02,
+      "loss": 1.4784,
+      "step": 8574
+    },
+    {
+      "epoch": 0.7938436614013771,
+      "grad_norm": 0.16015625,
+      "learning_rate": 0.02,
+      "loss": 1.5007,
+      "step": 8575
+    },
+    {
+      "epoch": 0.7939362379216571,
+      "grad_norm": 0.1640625,
+      "learning_rate": 0.02,
+      "loss": 1.5125,
+      "step": 8576
+    },
+    {
+      "epoch": 0.7940288144419372,
+      "grad_norm": 0.1650390625,
+      "learning_rate": 0.02,
+      "loss": 1.4361,
+      "step": 8577
+    },
+    {
+      "epoch": 0.7941213909622172,
+      "grad_norm": 0.150390625,
+      "learning_rate": 0.02,
+      "loss": 1.5507,
+      "step": 8578
+    },
+    {
+      "epoch": 0.7942139674824973,
+      "grad_norm": 0.173828125,
+      "learning_rate": 0.02,
+      "loss": 1.4885,
+      "step": 8579
+    },
+    {
+      "epoch": 0.7943065440027773,
+      "grad_norm": 0.1640625,
+      "learning_rate": 0.02,
+      "loss": 1.4621,
+      "step": 8580
+    },
+    {
+      "epoch": 0.7943991205230574,
+      "grad_norm": 0.158203125,
+      "learning_rate": 0.02,
+      "loss": 1.4636,
+      "step": 8581
+    },
+    {
+      "epoch": 0.7944916970433373,
+      "grad_norm": 0.16796875,
+      "learning_rate": 0.02,
+      "loss": 1.5266,
+      "step": 8582
+    },
+    {
+      "epoch": 0.7945842735636174,
+      "grad_norm": 0.15625,
+      "learning_rate": 0.02,
+      "loss": 1.4677,
+      "step": 8583
+    },
+    {
+      "epoch": 0.7946768500838974,
+      "grad_norm": 0.1728515625,
+      "learning_rate": 0.02,
+      "loss": 1.4936,
+      "step": 8584
+    },
+    {
+      "epoch": 0.7947694266041775,
+      "grad_norm": 0.1591796875,
+      "learning_rate": 0.02,
+      "loss": 1.5195,
+      "step": 8585
+    },
+    {
+      "epoch": 0.7948620031244575,
+      "grad_norm": 0.154296875,
+      "learning_rate": 0.02,
+      "loss": 1.5194,
+      "step": 8586
+    },
+    {
+      "epoch": 0.7949545796447376,
+      "grad_norm": 0.1513671875,
+      "learning_rate": 0.02,
+      "loss": 1.5527,
+      "step": 8587
+    },
+    {
+      "epoch": 0.7950471561650176,
+      "grad_norm": 0.1552734375,
+      "learning_rate": 0.02,
+      "loss": 1.5489,
+      "step": 8588
+    },
+    {
+      "epoch": 0.7951397326852977,
+      "grad_norm": 0.154296875,
+      "learning_rate": 0.02,
+      "loss": 1.4678,
+      "step": 8589
+    },
+    {
+      "epoch": 0.7952323092055777,
+      "grad_norm": 0.1484375,
+      "learning_rate": 0.02,
+      "loss": 1.4598,
+      "step": 8590
+    },
+    {
+      "epoch": 0.7953248857258578,
+      "grad_norm": 0.1552734375,
+      "learning_rate": 0.02,
+      "loss": 1.4641,
+      "step": 8591
+    },
+    {
+      "epoch": 0.7954174622461379,
+      "grad_norm": 0.14453125,
+      "learning_rate": 0.02,
+      "loss": 1.4682,
+      "step": 8592
+    },
+    {
+      "epoch": 0.7955100387664179,
+      "grad_norm": 0.15234375,
+      "learning_rate": 0.02,
+      "loss": 1.4844,
+      "step": 8593
+    },
+    {
+      "epoch": 0.795602615286698,
+      "grad_norm": 0.150390625,
+      "learning_rate": 0.02,
+      "loss": 1.5641,
+      "step": 8594
+    },
+    {
+      "epoch": 0.795695191806978,
+      "grad_norm": 0.1591796875,
+      "learning_rate": 0.02,
+      "loss": 1.4832,
+      "step": 8595
+    },
+    {
+      "epoch": 0.795787768327258,
+      "grad_norm": 0.1572265625,
+      "learning_rate": 0.02,
+      "loss": 1.5207,
+      "step": 8596
+    },
+    {
+      "epoch": 0.795880344847538,
+      "grad_norm": 0.162109375,
+      "learning_rate": 0.02,
+      "loss": 1.4706,
+      "step": 8597
+    },
+    {
+      "epoch": 0.795972921367818,
+      "grad_norm": 0.1630859375,
+      "learning_rate": 0.02,
+      "loss": 1.4458,
+      "step": 8598
+    },
+    {
+      "epoch": 0.7960654978880981,
+      "grad_norm": 0.16015625,
+      "learning_rate": 0.02,
+      "loss": 1.4729,
+      "step": 8599
+    },
+    {
+      "epoch": 0.7961580744083782,
+      "grad_norm": 0.162109375,
+      "learning_rate": 0.02,
+      "loss": 1.5676,
+      "step": 8600
+    },
+    {
+      "epoch": 0.7962506509286582,
+      "grad_norm": 0.150390625,
+      "learning_rate": 0.02,
+      "loss": 1.4467,
+      "step": 8601
+    },
+    {
+      "epoch": 0.7963432274489383,
+      "grad_norm": 0.15625,
+      "learning_rate": 0.02,
+      "loss": 1.4929,
+      "step": 8602
+    },
+    {
+      "epoch": 0.7964358039692183,
+      "grad_norm": 0.1650390625,
+      "learning_rate": 0.02,
+      "loss": 1.5606,
+      "step": 8603
+    },
+    {
+      "epoch": 0.7965283804894984,
+      "grad_norm": 0.15625,
+      "learning_rate": 0.02,
+      "loss": 1.5207,
+      "step": 8604
+    },
+    {
+      "epoch": 0.7966209570097784,
+      "grad_norm": 0.158203125,
+      "learning_rate": 0.02,
+      "loss": 1.5294,
+      "step": 8605
+    },
+    {
+      "epoch": 0.7967135335300585,
+      "grad_norm": 0.1650390625,
+      "learning_rate": 0.02,
+      "loss": 1.4831,
+      "step": 8606
+    },
+    {
+      "epoch": 0.7968061100503385,
+      "grad_norm": 0.142578125,
+      "learning_rate": 0.02,
+      "loss": 1.4912,
+      "step": 8607
+    },
+    {
+      "epoch": 0.7968986865706186,
+      "grad_norm": 0.1494140625,
+      "learning_rate": 0.02,
+      "loss": 1.4567,
+      "step": 8608
+    },
+    {
+      "epoch": 0.7969912630908985,
+      "grad_norm": 0.1474609375,
+      "learning_rate": 0.02,
+      "loss": 1.4984,
+      "step": 8609
+    },
+    {
+      "epoch": 0.7970838396111786,
+      "grad_norm": 0.1572265625,
+      "learning_rate": 0.02,
+      "loss": 1.5227,
+      "step": 8610
+    },
+    {
+      "epoch": 0.7971764161314586,
+      "grad_norm": 0.14453125,
+      "learning_rate": 0.02,
+      "loss": 1.5217,
+      "step": 8611
+    },
+    {
+      "epoch": 0.7972689926517387,
+      "grad_norm": 0.154296875,
+      "learning_rate": 0.02,
+      "loss": 1.5749,
+      "step": 8612
+    },
+    {
+      "epoch": 0.7973615691720187,
+      "grad_norm": 0.1630859375,
+      "learning_rate": 0.02,
+      "loss": 1.4527,
+      "step": 8613
+    },
+    {
+      "epoch": 0.7974541456922988,
+      "grad_norm": 0.16015625,
+      "learning_rate": 0.02,
+      "loss": 1.5387,
+      "step": 8614
+    },
+    {
+      "epoch": 0.7975467222125788,
+      "grad_norm": 0.16015625,
+      "learning_rate": 0.02,
+      "loss": 1.4824,
+      "step": 8615
+    },
+    {
+      "epoch": 0.7976392987328589,
+      "grad_norm": 0.1552734375,
+      "learning_rate": 0.02,
+      "loss": 1.4987,
+      "step": 8616
+    },
+    {
+      "epoch": 0.7977318752531389,
+      "grad_norm": 0.15234375,
+      "learning_rate": 0.02,
+      "loss": 1.4364,
+      "step": 8617
+    },
+    {
+      "epoch": 0.797824451773419,
+      "grad_norm": 0.15625,
+      "learning_rate": 0.02,
+      "loss": 1.5124,
+      "step": 8618
+    },
+    {
+      "epoch": 0.797917028293699,
+      "grad_norm": 0.1455078125,
+      "learning_rate": 0.02,
+      "loss": 1.4753,
+      "step": 8619
+    },
+    {
+      "epoch": 0.7980096048139791,
+      "grad_norm": 0.1611328125,
+      "learning_rate": 0.02,
+      "loss": 1.4762,
+      "step": 8620
+    },
+    {
+      "epoch": 0.7981021813342591,
+      "grad_norm": 0.1513671875,
+      "learning_rate": 0.02,
+      "loss": 1.476,
+      "step": 8621
+    },
+    {
+      "epoch": 0.7981947578545392,
+      "grad_norm": 0.1552734375,
+      "learning_rate": 0.02,
+      "loss": 1.5151,
+      "step": 8622
+    },
+    {
+      "epoch": 0.7982873343748191,
+      "grad_norm": 0.1455078125,
+      "learning_rate": 0.02,
+      "loss": 1.511,
+      "step": 8623
+    },
+    {
+      "epoch": 0.7983799108950992,
+      "grad_norm": 0.1669921875,
+      "learning_rate": 0.02,
+      "loss": 1.5151,
+      "step": 8624
+    },
+    {
+      "epoch": 0.7984724874153792,
+      "grad_norm": 0.1552734375,
+      "learning_rate": 0.02,
+      "loss": 1.4323,
+      "step": 8625
+    },
+    {
+      "epoch": 0.7985650639356593,
+      "grad_norm": 0.15234375,
+      "learning_rate": 0.02,
+      "loss": 1.496,
+      "step": 8626
+    },
+    {
+      "epoch": 0.7986576404559393,
+      "grad_norm": 0.1435546875,
+      "learning_rate": 0.02,
+      "loss": 1.4924,
+      "step": 8627
+    },
+    {
+      "epoch": 0.7987502169762194,
+      "grad_norm": 0.1513671875,
+      "learning_rate": 0.02,
+      "loss": 1.4613,
+      "step": 8628
+    },
+    {
+      "epoch": 0.7988427934964994,
+      "grad_norm": 0.134765625,
+      "learning_rate": 0.02,
+      "loss": 1.5182,
+      "step": 8629
+    },
+    {
+      "epoch": 0.7989353700167795,
+      "grad_norm": 0.15625,
+      "learning_rate": 0.02,
+      "loss": 1.4881,
+      "step": 8630
+    },
+    {
+      "epoch": 0.7990279465370596,
+      "grad_norm": 0.1533203125,
+      "learning_rate": 0.02,
+      "loss": 1.5102,
+      "step": 8631
+    },
+    {
+      "epoch": 0.7991205230573396,
+      "grad_norm": 0.1494140625,
+      "learning_rate": 0.02,
+      "loss": 1.5083,
+      "step": 8632
+    },
+    {
+      "epoch": 0.7992130995776197,
+      "grad_norm": 0.15234375,
+      "learning_rate": 0.02,
+      "loss": 1.439,
+      "step": 8633
+    },
+    {
+      "epoch": 0.7993056760978997,
+      "grad_norm": 0.1494140625,
+      "learning_rate": 0.02,
+      "loss": 1.4205,
+      "step": 8634
+    },
+    {
+      "epoch": 0.7993982526181798,
+      "grad_norm": 0.14453125,
+      "learning_rate": 0.02,
+      "loss": 1.5088,
+      "step": 8635
+    },
+    {
+      "epoch": 0.7994908291384598,
+      "grad_norm": 0.154296875,
+      "learning_rate": 0.02,
+      "loss": 1.5026,
+      "step": 8636
+    },
+    {
+      "epoch": 0.7995834056587398,
+      "grad_norm": 0.154296875,
+      "learning_rate": 0.02,
+      "loss": 1.502,
+      "step": 8637
+    },
+    {
+      "epoch": 0.7996759821790198,
+      "grad_norm": 0.1591796875,
+      "learning_rate": 0.02,
+      "loss": 1.5393,
+      "step": 8638
+    },
+    {
+      "epoch": 0.7997685586992999,
+      "grad_norm": 0.158203125,
+      "learning_rate": 0.02,
+      "loss": 1.44,
+      "step": 8639
+    },
+    {
+      "epoch": 0.7998611352195799,
+      "grad_norm": 0.146484375,
+      "learning_rate": 0.02,
+      "loss": 1.5235,
+      "step": 8640
+    },
+    {
+      "epoch": 0.79995371173986,
+      "grad_norm": 0.162109375,
+      "learning_rate": 0.02,
+      "loss": 1.5539,
+      "step": 8641
+    },
+    {
+      "epoch": 0.80004628826014,
+      "grad_norm": 0.1572265625,
+      "learning_rate": 0.02,
+      "loss": 1.4711,
+      "step": 8642
+    },
+    {
+      "epoch": 0.8001388647804201,
+      "grad_norm": 0.1611328125,
+      "learning_rate": 0.02,
+      "loss": 1.4888,
+      "step": 8643
+    },
+    {
+      "epoch": 0.8002314413007001,
+      "grad_norm": 0.1357421875,
+      "learning_rate": 0.02,
+      "loss": 1.517,
+      "step": 8644
+    },
+    {
+      "epoch": 0.8003240178209802,
+      "grad_norm": 0.15234375,
+      "learning_rate": 0.02,
+      "loss": 1.4805,
+      "step": 8645
+    },
+    {
+      "epoch": 0.8004165943412602,
+      "grad_norm": 0.142578125,
+      "learning_rate": 0.02,
+      "loss": 1.4614,
+      "step": 8646
+    },
+    {
+      "epoch": 0.8005091708615403,
+      "grad_norm": 0.1552734375,
+      "learning_rate": 0.02,
+      "loss": 1.5107,
+      "step": 8647
+    },
+    {
+      "epoch": 0.8006017473818203,
+      "grad_norm": 0.154296875,
+      "learning_rate": 0.02,
+      "loss": 1.558,
+      "step": 8648
+    },
+    {
+      "epoch": 0.8006943239021004,
+      "grad_norm": 0.1552734375,
+      "learning_rate": 0.02,
+      "loss": 1.4883,
+      "step": 8649
+    },
+    {
+      "epoch": 0.8007869004223803,
+      "grad_norm": 0.1572265625,
+      "learning_rate": 0.02,
+      "loss": 1.5375,
+      "step": 8650
+    },
+    {
+      "epoch": 0.8008794769426604,
+      "grad_norm": 0.158203125,
+      "learning_rate": 0.02,
+      "loss": 1.4716,
+      "step": 8651
+    },
+    {
+      "epoch": 0.8009720534629404,
+      "grad_norm": 0.1474609375,
+      "learning_rate": 0.02,
+      "loss": 1.4938,
+      "step": 8652
+    },
+    {
+      "epoch": 0.8010646299832205,
+      "grad_norm": 0.15625,
+      "learning_rate": 0.02,
+      "loss": 1.5045,
+      "step": 8653
+    },
+    {
+      "epoch": 0.8011572065035005,
+      "grad_norm": 0.1552734375,
+      "learning_rate": 0.02,
+      "loss": 1.4742,
+      "step": 8654
+    },
+    {
+      "epoch": 0.8012497830237806,
+      "grad_norm": 0.1474609375,
+      "learning_rate": 0.02,
+      "loss": 1.5601,
+      "step": 8655
+    },
+    {
+      "epoch": 0.8013423595440606,
+      "grad_norm": 0.1376953125,
+      "learning_rate": 0.02,
+      "loss": 1.4651,
+      "step": 8656
+    },
+    {
+      "epoch": 0.8014349360643407,
+      "grad_norm": 0.15625,
+      "learning_rate": 0.02,
+      "loss": 1.5175,
+      "step": 8657
+    },
+    {
+      "epoch": 0.8015275125846207,
+      "grad_norm": 0.166015625,
+      "learning_rate": 0.02,
+      "loss": 1.547,
+      "step": 8658
+    },
+    {
+      "epoch": 0.8016200891049008,
+      "grad_norm": 0.140625,
+      "learning_rate": 0.02,
+      "loss": 1.4549,
+      "step": 8659
+    },
+    {
+      "epoch": 0.8017126656251808,
+      "grad_norm": 0.150390625,
+      "learning_rate": 0.02,
+      "loss": 1.4569,
+      "step": 8660
+    },
+    {
+      "epoch": 0.8018052421454609,
+      "grad_norm": 0.1513671875,
+      "learning_rate": 0.02,
+      "loss": 1.4804,
+      "step": 8661
+    },
+    {
+      "epoch": 0.801897818665741,
+      "grad_norm": 0.16015625,
+      "learning_rate": 0.02,
+      "loss": 1.5347,
+      "step": 8662
+    },
+    {
+      "epoch": 0.801990395186021,
+      "grad_norm": 0.142578125,
+      "learning_rate": 0.02,
+      "loss": 1.5213,
+      "step": 8663
+    },
+    {
+      "epoch": 0.8020829717063009,
+      "grad_norm": 0.1455078125,
+      "learning_rate": 0.02,
+      "loss": 1.479,
+      "step": 8664
+    },
+    {
+      "epoch": 0.802175548226581,
+      "grad_norm": 0.1513671875,
+      "learning_rate": 0.02,
+      "loss": 1.5368,
+      "step": 8665
+    },
+    {
+      "epoch": 0.802268124746861,
+      "grad_norm": 0.15625,
+      "learning_rate": 0.02,
+      "loss": 1.4722,
+      "step": 8666
+    },
+    {
+      "epoch": 0.8023607012671411,
+      "grad_norm": 0.15234375,
+      "learning_rate": 0.02,
+      "loss": 1.431,
+      "step": 8667
+    },
+    {
+      "epoch": 0.8024532777874211,
+      "grad_norm": 0.162109375,
+      "learning_rate": 0.02,
+      "loss": 1.5086,
+      "step": 8668
+    },
+    {
+      "epoch": 0.8025458543077012,
+      "grad_norm": 0.1455078125,
+      "learning_rate": 0.02,
+      "loss": 1.59,
+      "step": 8669
+    },
+    {
+      "epoch": 0.8026384308279813,
+      "grad_norm": 0.1572265625,
+      "learning_rate": 0.02,
+      "loss": 1.5521,
+      "step": 8670
+    },
+    {
+      "epoch": 0.8027310073482613,
+      "grad_norm": 0.1484375,
+      "learning_rate": 0.02,
+      "loss": 1.4948,
+      "step": 8671
+    },
+    {
+      "epoch": 0.8028235838685414,
+      "grad_norm": 0.1650390625,
+      "learning_rate": 0.02,
+      "loss": 1.4779,
+      "step": 8672
+    },
+    {
+      "epoch": 0.8029161603888214,
+      "grad_norm": 0.15234375,
+      "learning_rate": 0.02,
+      "loss": 1.582,
+      "step": 8673
+    },
+    {
+      "epoch": 0.8030087369091015,
+      "grad_norm": 0.1591796875,
+      "learning_rate": 0.02,
+      "loss": 1.4809,
+      "step": 8674
+    },
+    {
+      "epoch": 0.8031013134293815,
+      "grad_norm": 0.1494140625,
+      "learning_rate": 0.02,
+      "loss": 1.4354,
+      "step": 8675
+    },
+    {
+      "epoch": 0.8031938899496616,
+      "grad_norm": 0.158203125,
+      "learning_rate": 0.02,
+      "loss": 1.6265,
+      "step": 8676
+    },
+    {
+      "epoch": 0.8032864664699415,
+      "grad_norm": 0.1484375,
+      "learning_rate": 0.02,
+      "loss": 1.5139,
+      "step": 8677
+    },
+    {
+      "epoch": 0.8033790429902216,
+      "grad_norm": 0.1484375,
+      "learning_rate": 0.02,
+      "loss": 1.412,
+      "step": 8678
+    },
+    {
+      "epoch": 0.8034716195105016,
+      "grad_norm": 0.1455078125,
+      "learning_rate": 0.02,
+      "loss": 1.574,
+      "step": 8679
+    },
+    {
+      "epoch": 0.8035641960307817,
+      "grad_norm": 0.1630859375,
+      "learning_rate": 0.02,
+      "loss": 1.4964,
+      "step": 8680
+    },
+    {
+      "epoch": 0.8036567725510617,
+      "grad_norm": 0.1533203125,
+      "learning_rate": 0.02,
+      "loss": 1.4656,
+      "step": 8681
+    },
+    {
+      "epoch": 0.8037493490713418,
+      "grad_norm": 0.1513671875,
+      "learning_rate": 0.02,
+      "loss": 1.5201,
+      "step": 8682
+    },
+    {
+      "epoch": 0.8038419255916218,
+      "grad_norm": 0.15625,
+      "learning_rate": 0.02,
+      "loss": 1.5667,
+      "step": 8683
+    },
+    {
+      "epoch": 0.8039345021119019,
+      "grad_norm": 0.1572265625,
+      "learning_rate": 0.02,
+      "loss": 1.537,
+      "step": 8684
+    },
+    {
+      "epoch": 0.8040270786321819,
+      "grad_norm": 0.1416015625,
+      "learning_rate": 0.02,
+      "loss": 1.4913,
+      "step": 8685
+    },
+    {
+      "epoch": 0.804119655152462,
+      "grad_norm": 0.158203125,
+      "learning_rate": 0.02,
+      "loss": 1.4834,
+      "step": 8686
+    },
+    {
+      "epoch": 0.804212231672742,
+      "grad_norm": 0.158203125,
+      "learning_rate": 0.02,
+      "loss": 1.51,
+      "step": 8687
+    },
+    {
+      "epoch": 0.8043048081930221,
+      "grad_norm": 0.1640625,
+      "learning_rate": 0.02,
+      "loss": 1.4474,
+      "step": 8688
+    },
+    {
+      "epoch": 0.8043973847133021,
+      "grad_norm": 0.154296875,
+      "learning_rate": 0.02,
+      "loss": 1.492,
+      "step": 8689
+    },
+    {
+      "epoch": 0.8044899612335822,
+      "grad_norm": 0.1533203125,
+      "learning_rate": 0.02,
+      "loss": 1.5183,
+      "step": 8690
+    },
+    {
+      "epoch": 0.8045825377538621,
+      "grad_norm": 0.15234375,
+      "learning_rate": 0.02,
+      "loss": 1.5539,
+      "step": 8691
+    },
+    {
+      "epoch": 0.8046751142741422,
+      "grad_norm": 0.15625,
+      "learning_rate": 0.02,
+      "loss": 1.5371,
+      "step": 8692
+    },
+    {
+      "epoch": 0.8047676907944222,
+      "grad_norm": 0.150390625,
+      "learning_rate": 0.02,
+      "loss": 1.5487,
+      "step": 8693
+    },
+    {
+      "epoch": 0.8048602673147023,
+      "grad_norm": 0.154296875,
+      "learning_rate": 0.02,
+      "loss": 1.4823,
+      "step": 8694
+    },
+    {
+      "epoch": 0.8049528438349823,
+      "grad_norm": 0.146484375,
+      "learning_rate": 0.02,
+      "loss": 1.5641,
+      "step": 8695
+    },
+    {
+      "epoch": 0.8050454203552624,
+      "grad_norm": 0.162109375,
+      "learning_rate": 0.02,
+      "loss": 1.4725,
+      "step": 8696
+    },
+    {
+      "epoch": 0.8051379968755424,
+      "grad_norm": 0.162109375,
+      "learning_rate": 0.02,
+      "loss": 1.5625,
+      "step": 8697
+    },
+    {
+      "epoch": 0.8052305733958225,
+      "grad_norm": 0.1494140625,
+      "learning_rate": 0.02,
+      "loss": 1.5501,
+      "step": 8698
+    },
+    {
+      "epoch": 0.8053231499161025,
+      "grad_norm": 0.1572265625,
+      "learning_rate": 0.02,
+      "loss": 1.5083,
+      "step": 8699
+    },
+    {
+      "epoch": 0.8054157264363826,
+      "grad_norm": 0.150390625,
+      "learning_rate": 0.02,
+      "loss": 1.4454,
+      "step": 8700
+    },
+    {
+      "epoch": 0.8055083029566626,
+      "grad_norm": 0.15234375,
+      "learning_rate": 0.02,
+      "loss": 1.5305,
+      "step": 8701
+    },
+    {
+      "epoch": 0.8056008794769427,
+      "grad_norm": 0.1552734375,
+      "learning_rate": 0.02,
+      "loss": 1.4881,
+      "step": 8702
+    },
+    {
+      "epoch": 0.8056934559972228,
+      "grad_norm": 0.158203125,
+      "learning_rate": 0.02,
+      "loss": 1.534,
+      "step": 8703
+    },
+    {
+      "epoch": 0.8057860325175028,
+      "grad_norm": 0.140625,
+      "learning_rate": 0.02,
+      "loss": 1.5011,
+      "step": 8704
+    },
+    {
+      "epoch": 0.8058786090377827,
+      "grad_norm": 0.158203125,
+      "learning_rate": 0.02,
+      "loss": 1.4849,
+      "step": 8705
+    },
+    {
+      "epoch": 0.8059711855580628,
+      "grad_norm": 0.150390625,
+      "learning_rate": 0.02,
+      "loss": 1.5278,
+      "step": 8706
+    },
+    {
+      "epoch": 0.8060637620783428,
+      "grad_norm": 0.162109375,
+      "learning_rate": 0.02,
+      "loss": 1.4961,
+      "step": 8707
+    },
+    {
+      "epoch": 0.8061563385986229,
+      "grad_norm": 0.1494140625,
+      "learning_rate": 0.02,
+      "loss": 1.4799,
+      "step": 8708
+    },
+    {
+      "epoch": 0.806248915118903,
+      "grad_norm": 0.1572265625,
+      "learning_rate": 0.02,
+      "loss": 1.5043,
+      "step": 8709
+    },
+    {
+      "epoch": 0.806341491639183,
+      "grad_norm": 0.15625,
+      "learning_rate": 0.02,
+      "loss": 1.5003,
+      "step": 8710
+    },
+    {
+      "epoch": 0.8064340681594631,
+      "grad_norm": 0.1591796875,
+      "learning_rate": 0.02,
+      "loss": 1.4835,
+      "step": 8711
+    },
+    {
+      "epoch": 0.8065266446797431,
+      "grad_norm": 0.1494140625,
+      "learning_rate": 0.02,
+      "loss": 1.5476,
+      "step": 8712
+    },
+    {
+      "epoch": 0.8066192212000232,
+      "grad_norm": 0.1591796875,
+      "learning_rate": 0.02,
+      "loss": 1.4896,
+      "step": 8713
+    },
+    {
+      "epoch": 0.8067117977203032,
+      "grad_norm": 0.16015625,
+      "learning_rate": 0.02,
+      "loss": 1.4975,
+      "step": 8714
+    },
+    {
+      "epoch": 0.8068043742405833,
+      "grad_norm": 0.1552734375,
+      "learning_rate": 0.02,
+      "loss": 1.5304,
+      "step": 8715
+    },
+    {
+      "epoch": 0.8068969507608633,
+      "grad_norm": 0.1572265625,
+      "learning_rate": 0.02,
+      "loss": 1.5148,
+      "step": 8716
+    },
+    {
+      "epoch": 0.8069895272811434,
+      "grad_norm": 0.16015625,
+      "learning_rate": 0.02,
+      "loss": 1.4856,
+      "step": 8717
+    },
+    {
+      "epoch": 0.8070821038014233,
+      "grad_norm": 0.158203125,
+      "learning_rate": 0.02,
+      "loss": 1.4887,
+      "step": 8718
+    },
+    {
+      "epoch": 0.8071746803217034,
+      "grad_norm": 0.171875,
+      "learning_rate": 0.02,
+      "loss": 1.5232,
+      "step": 8719
+    },
+    {
+      "epoch": 0.8072672568419834,
+      "grad_norm": 0.1689453125,
+      "learning_rate": 0.02,
+      "loss": 1.4757,
+      "step": 8720
+    },
+    {
+      "epoch": 0.8073598333622635,
+      "grad_norm": 0.1552734375,
+      "learning_rate": 0.02,
+      "loss": 1.4111,
+      "step": 8721
+    },
+    {
+      "epoch": 0.8074524098825435,
+      "grad_norm": 0.158203125,
+      "learning_rate": 0.02,
+      "loss": 1.5378,
+      "step": 8722
+    },
+    {
+      "epoch": 0.8075449864028236,
+      "grad_norm": 0.1416015625,
+      "learning_rate": 0.02,
+      "loss": 1.4845,
+      "step": 8723
+    },
+    {
+      "epoch": 0.8076375629231036,
+      "grad_norm": 0.1484375,
+      "learning_rate": 0.02,
+      "loss": 1.4353,
+      "step": 8724
+    },
+    {
+      "epoch": 0.8077301394433837,
+      "grad_norm": 0.154296875,
+      "learning_rate": 0.02,
+      "loss": 1.5321,
+      "step": 8725
+    },
+    {
+      "epoch": 0.8078227159636637,
+      "grad_norm": 0.1591796875,
+      "learning_rate": 0.02,
+      "loss": 1.534,
+      "step": 8726
+    },
+    {
+      "epoch": 0.8079152924839438,
+      "grad_norm": 0.158203125,
+      "learning_rate": 0.02,
+      "loss": 1.4992,
+      "step": 8727
+    },
+    {
+      "epoch": 0.8080078690042238,
+      "grad_norm": 0.154296875,
+      "learning_rate": 0.02,
+      "loss": 1.4832,
+      "step": 8728
+    },
+    {
+      "epoch": 0.8081004455245039,
+      "grad_norm": 0.1513671875,
+      "learning_rate": 0.02,
+      "loss": 1.5136,
+      "step": 8729
+    },
+    {
+      "epoch": 0.8081930220447839,
+      "grad_norm": 0.16015625,
+      "learning_rate": 0.02,
+      "loss": 1.5517,
+      "step": 8730
+    },
+    {
+      "epoch": 0.808285598565064,
+      "grad_norm": 0.158203125,
+      "learning_rate": 0.02,
+      "loss": 1.5363,
+      "step": 8731
+    },
+    {
+      "epoch": 0.8083781750853439,
+      "grad_norm": 0.1650390625,
+      "learning_rate": 0.02,
+      "loss": 1.4175,
+      "step": 8732
+    },
+    {
+      "epoch": 0.808470751605624,
+      "grad_norm": 0.1591796875,
+      "learning_rate": 0.02,
+      "loss": 1.4602,
+      "step": 8733
+    },
+    {
+      "epoch": 0.808563328125904,
+      "grad_norm": 0.16796875,
+      "learning_rate": 0.02,
+      "loss": 1.5061,
+      "step": 8734
+    },
+    {
+      "epoch": 0.8086559046461841,
+      "grad_norm": 0.1484375,
+      "learning_rate": 0.02,
+      "loss": 1.5311,
+      "step": 8735
+    },
+    {
+      "epoch": 0.8087484811664641,
+      "grad_norm": 0.171875,
+      "learning_rate": 0.02,
+      "loss": 1.491,
+      "step": 8736
+    },
+    {
+      "epoch": 0.8088410576867442,
+      "grad_norm": 0.16015625,
+      "learning_rate": 0.02,
+      "loss": 1.5112,
+      "step": 8737
+    },
+    {
+      "epoch": 0.8089336342070242,
+      "grad_norm": 0.1533203125,
+      "learning_rate": 0.02,
+      "loss": 1.4979,
+      "step": 8738
+    },
+    {
+      "epoch": 0.8090262107273043,
+      "grad_norm": 0.150390625,
+      "learning_rate": 0.02,
+      "loss": 1.471,
+      "step": 8739
+    },
+    {
+      "epoch": 0.8091187872475843,
+      "grad_norm": 0.1650390625,
+      "learning_rate": 0.02,
+      "loss": 1.4231,
+      "step": 8740
+    },
+    {
+      "epoch": 0.8092113637678644,
+      "grad_norm": 0.1494140625,
+      "learning_rate": 0.02,
+      "loss": 1.4688,
+      "step": 8741
+    },
+    {
+      "epoch": 0.8093039402881445,
+      "grad_norm": 0.1669921875,
+      "learning_rate": 0.02,
+      "loss": 1.5247,
+      "step": 8742
+    },
+    {
+      "epoch": 0.8093965168084245,
+      "grad_norm": 0.1572265625,
+      "learning_rate": 0.02,
+      "loss": 1.5791,
+      "step": 8743
+    },
+    {
+      "epoch": 0.8094890933287046,
+      "grad_norm": 0.15234375,
+      "learning_rate": 0.02,
+      "loss": 1.4914,
+      "step": 8744
+    },
+    {
+      "epoch": 0.8095816698489845,
+      "grad_norm": 0.15234375,
+      "learning_rate": 0.02,
+      "loss": 1.5413,
+      "step": 8745
+    },
+    {
+      "epoch": 0.8096742463692645,
+      "grad_norm": 0.142578125,
+      "learning_rate": 0.02,
+      "loss": 1.4011,
+      "step": 8746
+    },
+    {
+      "epoch": 0.8097668228895446,
+      "grad_norm": 0.1513671875,
+      "learning_rate": 0.02,
+      "loss": 1.519,
+      "step": 8747
+    },
+    {
+      "epoch": 0.8098593994098247,
+      "grad_norm": 0.1513671875,
+      "learning_rate": 0.02,
+      "loss": 1.4885,
+      "step": 8748
+    },
+    {
+      "epoch": 0.8099519759301047,
+      "grad_norm": 0.1533203125,
+      "learning_rate": 0.02,
+      "loss": 1.5257,
+      "step": 8749
+    },
+    {
+      "epoch": 0.8100445524503848,
+      "grad_norm": 0.1552734375,
+      "learning_rate": 0.02,
+      "loss": 1.4947,
+      "step": 8750
+    },
+    {
+      "epoch": 0.8101371289706648,
+      "grad_norm": 0.1591796875,
+      "learning_rate": 0.02,
+      "loss": 1.4778,
+      "step": 8751
+    },
+    {
+      "epoch": 0.8102297054909449,
+      "grad_norm": 0.1640625,
+      "learning_rate": 0.02,
+      "loss": 1.4809,
+      "step": 8752
+    },
+    {
+      "epoch": 0.8103222820112249,
+      "grad_norm": 0.146484375,
+      "learning_rate": 0.02,
+      "loss": 1.5005,
+      "step": 8753
+    },
+    {
+      "epoch": 0.810414858531505,
+      "grad_norm": 0.1416015625,
+      "learning_rate": 0.02,
+      "loss": 1.5216,
+      "step": 8754
+    },
+    {
+      "epoch": 0.810507435051785,
+      "grad_norm": 0.146484375,
+      "learning_rate": 0.02,
+      "loss": 1.4949,
+      "step": 8755
+    },
+    {
+      "epoch": 0.8106000115720651,
+      "grad_norm": 0.1455078125,
+      "learning_rate": 0.02,
+      "loss": 1.502,
+      "step": 8756
+    },
+    {
+      "epoch": 0.8106925880923451,
+      "grad_norm": 0.154296875,
+      "learning_rate": 0.02,
+      "loss": 1.5048,
+      "step": 8757
+    },
+    {
+      "epoch": 0.8107851646126252,
+      "grad_norm": 0.14453125,
+      "learning_rate": 0.02,
+      "loss": 1.4134,
+      "step": 8758
+    },
+    {
+      "epoch": 0.8108777411329051,
+      "grad_norm": 0.1533203125,
+      "learning_rate": 0.02,
+      "loss": 1.5271,
+      "step": 8759
+    },
+    {
+      "epoch": 0.8109703176531852,
+      "grad_norm": 0.154296875,
+      "learning_rate": 0.02,
+      "loss": 1.4546,
+      "step": 8760
+    },
+    {
+      "epoch": 0.8110628941734652,
+      "grad_norm": 0.16015625,
+      "learning_rate": 0.02,
+      "loss": 1.5116,
+      "step": 8761
+    },
+    {
+      "epoch": 0.8111554706937453,
+      "grad_norm": 0.1455078125,
+      "learning_rate": 0.02,
+      "loss": 1.4621,
+      "step": 8762
+    },
+    {
+      "epoch": 0.8112480472140253,
+      "grad_norm": 0.16015625,
+      "learning_rate": 0.02,
+      "loss": 1.4943,
+      "step": 8763
+    },
+    {
+      "epoch": 0.8113406237343054,
+      "grad_norm": 0.1474609375,
+      "learning_rate": 0.02,
+      "loss": 1.4796,
+      "step": 8764
+    },
+    {
+      "epoch": 0.8114332002545854,
+      "grad_norm": 0.1572265625,
+      "learning_rate": 0.02,
+      "loss": 1.4391,
+      "step": 8765
+    },
+    {
+      "epoch": 0.8115257767748655,
+      "grad_norm": 0.15234375,
+      "learning_rate": 0.02,
+      "loss": 1.561,
+      "step": 8766
+    },
+    {
+      "epoch": 0.8116183532951455,
+      "grad_norm": 0.1572265625,
+      "learning_rate": 0.02,
+      "loss": 1.4435,
+      "step": 8767
+    },
+    {
+      "epoch": 0.8117109298154256,
+      "grad_norm": 0.1552734375,
+      "learning_rate": 0.02,
+      "loss": 1.5199,
+      "step": 8768
+    },
+    {
+      "epoch": 0.8118035063357056,
+      "grad_norm": 0.1474609375,
+      "learning_rate": 0.02,
+      "loss": 1.4757,
+      "step": 8769
+    },
+    {
+      "epoch": 0.8118960828559857,
+      "grad_norm": 0.1591796875,
+      "learning_rate": 0.02,
+      "loss": 1.5279,
+      "step": 8770
+    },
+    {
+      "epoch": 0.8119886593762657,
+      "grad_norm": 0.15625,
+      "learning_rate": 0.02,
+      "loss": 1.5109,
+      "step": 8771
+    },
+    {
+      "epoch": 0.8120812358965457,
+      "grad_norm": 0.15625,
+      "learning_rate": 0.02,
+      "loss": 1.4713,
+      "step": 8772
+    },
+    {
+      "epoch": 0.8121738124168257,
+      "grad_norm": 0.14453125,
+      "learning_rate": 0.02,
+      "loss": 1.5069,
+      "step": 8773
+    },
+    {
+      "epoch": 0.8122663889371058,
+      "grad_norm": 0.1572265625,
+      "learning_rate": 0.02,
+      "loss": 1.5342,
+      "step": 8774
+    },
+    {
+      "epoch": 0.8123589654573858,
+      "grad_norm": 0.1572265625,
+      "learning_rate": 0.02,
+      "loss": 1.4679,
+      "step": 8775
+    },
+    {
+      "epoch": 0.8124515419776659,
+      "grad_norm": 0.15625,
+      "learning_rate": 0.02,
+      "loss": 1.4906,
+      "step": 8776
+    },
+    {
+      "epoch": 0.8125441184979459,
+      "grad_norm": 0.162109375,
+      "learning_rate": 0.02,
+      "loss": 1.5081,
+      "step": 8777
+    },
+    {
+      "epoch": 0.812636695018226,
+      "grad_norm": 0.16015625,
+      "learning_rate": 0.02,
+      "loss": 1.5064,
+      "step": 8778
+    },
+    {
+      "epoch": 0.812729271538506,
+      "grad_norm": 0.1513671875,
+      "learning_rate": 0.02,
+      "loss": 1.5294,
+      "step": 8779
+    },
+    {
+      "epoch": 0.8128218480587861,
+      "grad_norm": 0.1591796875,
+      "learning_rate": 0.02,
+      "loss": 1.4666,
+      "step": 8780
+    },
+    {
+      "epoch": 0.8129144245790662,
+      "grad_norm": 0.1591796875,
+      "learning_rate": 0.02,
+      "loss": 1.5183,
+      "step": 8781
+    },
+    {
+      "epoch": 0.8130070010993462,
+      "grad_norm": 0.15234375,
+      "learning_rate": 0.02,
+      "loss": 1.52,
+      "step": 8782
+    },
+    {
+      "epoch": 0.8130995776196263,
+      "grad_norm": 0.154296875,
+      "learning_rate": 0.02,
+      "loss": 1.595,
+      "step": 8783
+    },
+    {
+      "epoch": 0.8131921541399063,
+      "grad_norm": 0.154296875,
+      "learning_rate": 0.02,
+      "loss": 1.4645,
+      "step": 8784
+    },
+    {
+      "epoch": 0.8132847306601864,
+      "grad_norm": 0.162109375,
+      "learning_rate": 0.02,
+      "loss": 1.4453,
+      "step": 8785
+    },
+    {
+      "epoch": 0.8133773071804663,
+      "grad_norm": 0.1533203125,
+      "learning_rate": 0.02,
+      "loss": 1.5373,
+      "step": 8786
+    },
+    {
+      "epoch": 0.8134698837007464,
+      "grad_norm": 0.1591796875,
+      "learning_rate": 0.02,
+      "loss": 1.5037,
+      "step": 8787
+    },
+    {
+      "epoch": 0.8135624602210264,
+      "grad_norm": 0.1513671875,
+      "learning_rate": 0.02,
+      "loss": 1.5731,
+      "step": 8788
+    },
+    {
+      "epoch": 0.8136550367413065,
+      "grad_norm": 0.146484375,
+      "learning_rate": 0.02,
+      "loss": 1.5029,
+      "step": 8789
+    },
+    {
+      "epoch": 0.8137476132615865,
+      "grad_norm": 0.154296875,
+      "learning_rate": 0.02,
+      "loss": 1.4836,
+      "step": 8790
+    },
+    {
+      "epoch": 0.8138401897818666,
+      "grad_norm": 0.150390625,
+      "learning_rate": 0.02,
+      "loss": 1.4438,
+      "step": 8791
+    },
+    {
+      "epoch": 0.8139327663021466,
+      "grad_norm": 0.1533203125,
+      "learning_rate": 0.02,
+      "loss": 1.5156,
+      "step": 8792
+    },
+    {
+      "epoch": 0.8140253428224267,
+      "grad_norm": 0.1533203125,
+      "learning_rate": 0.02,
+      "loss": 1.5589,
+      "step": 8793
+    },
+    {
+      "epoch": 0.8141179193427067,
+      "grad_norm": 0.162109375,
+      "learning_rate": 0.02,
+      "loss": 1.4849,
+      "step": 8794
+    },
+    {
+      "epoch": 0.8142104958629868,
+      "grad_norm": 0.162109375,
+      "learning_rate": 0.02,
+      "loss": 1.4249,
+      "step": 8795
+    },
+    {
+      "epoch": 0.8143030723832668,
+      "grad_norm": 0.1513671875,
+      "learning_rate": 0.02,
+      "loss": 1.5283,
+      "step": 8796
+    },
+    {
+      "epoch": 0.8143956489035469,
+      "grad_norm": 0.1552734375,
+      "learning_rate": 0.02,
+      "loss": 1.4838,
+      "step": 8797
+    },
+    {
+      "epoch": 0.8144882254238269,
+      "grad_norm": 0.1513671875,
+      "learning_rate": 0.02,
+      "loss": 1.5172,
+      "step": 8798
+    },
+    {
+      "epoch": 0.814580801944107,
+      "grad_norm": 0.1611328125,
+      "learning_rate": 0.02,
+      "loss": 1.434,
+      "step": 8799
+    },
+    {
+      "epoch": 0.8146733784643869,
+      "grad_norm": 0.1630859375,
+      "learning_rate": 0.02,
+      "loss": 1.5244,
+      "step": 8800
+    },
+    {
+      "epoch": 0.814765954984667,
+      "grad_norm": 0.1650390625,
+      "learning_rate": 0.02,
+      "loss": 1.5039,
+      "step": 8801
+    },
+    {
+      "epoch": 0.814858531504947,
+      "grad_norm": 0.1572265625,
+      "learning_rate": 0.02,
+      "loss": 1.4103,
+      "step": 8802
+    },
+    {
+      "epoch": 0.8149511080252271,
+      "grad_norm": 0.1396484375,
+      "learning_rate": 0.02,
+      "loss": 1.547,
+      "step": 8803
+    },
+    {
+      "epoch": 0.8150436845455071,
+      "grad_norm": 0.1533203125,
+      "learning_rate": 0.02,
+      "loss": 1.5255,
+      "step": 8804
+    },
+    {
+      "epoch": 0.8151362610657872,
+      "grad_norm": 0.150390625,
+      "learning_rate": 0.02,
+      "loss": 1.6046,
+      "step": 8805
+    },
+    {
+      "epoch": 0.8152288375860672,
+      "grad_norm": 0.158203125,
+      "learning_rate": 0.02,
+      "loss": 1.4958,
+      "step": 8806
+    },
+    {
+      "epoch": 0.8153214141063473,
+      "grad_norm": 0.16796875,
+      "learning_rate": 0.02,
+      "loss": 1.4829,
+      "step": 8807
+    },
+    {
+      "epoch": 0.8154139906266273,
+      "grad_norm": 0.15234375,
+      "learning_rate": 0.02,
+      "loss": 1.5335,
+      "step": 8808
+    },
+    {
+      "epoch": 0.8155065671469074,
+      "grad_norm": 0.1640625,
+      "learning_rate": 0.02,
+      "loss": 1.5104,
+      "step": 8809
+    },
+    {
+      "epoch": 0.8155991436671874,
+      "grad_norm": 0.1357421875,
+      "learning_rate": 0.02,
+      "loss": 1.5327,
+      "step": 8810
+    },
+    {
+      "epoch": 0.8156917201874675,
+      "grad_norm": 0.1552734375,
+      "learning_rate": 0.02,
+      "loss": 1.4899,
+      "step": 8811
+    },
+    {
+      "epoch": 0.8157842967077475,
+      "grad_norm": 0.142578125,
+      "learning_rate": 0.02,
+      "loss": 1.4959,
+      "step": 8812
+    },
+    {
+      "epoch": 0.8158768732280275,
+      "grad_norm": 0.1552734375,
+      "learning_rate": 0.02,
+      "loss": 1.5353,
+      "step": 8813
+    },
+    {
+      "epoch": 0.8159694497483075,
+      "grad_norm": 0.142578125,
+      "learning_rate": 0.02,
+      "loss": 1.4642,
+      "step": 8814
+    },
+    {
+      "epoch": 0.8160620262685876,
+      "grad_norm": 0.1552734375,
+      "learning_rate": 0.02,
+      "loss": 1.417,
+      "step": 8815
+    },
+    {
+      "epoch": 0.8161546027888676,
+      "grad_norm": 0.1435546875,
+      "learning_rate": 0.02,
+      "loss": 1.4875,
+      "step": 8816
+    },
+    {
+      "epoch": 0.8162471793091477,
+      "grad_norm": 0.162109375,
+      "learning_rate": 0.02,
+      "loss": 1.5526,
+      "step": 8817
+    },
+    {
+      "epoch": 0.8163397558294277,
+      "grad_norm": 0.15625,
+      "learning_rate": 0.02,
+      "loss": 1.5282,
+      "step": 8818
+    },
+    {
+      "epoch": 0.8164323323497078,
+      "grad_norm": 0.1650390625,
+      "learning_rate": 0.02,
+      "loss": 1.5158,
+      "step": 8819
+    },
+    {
+      "epoch": 0.8165249088699879,
+      "grad_norm": 0.1513671875,
+      "learning_rate": 0.02,
+      "loss": 1.4477,
+      "step": 8820
+    },
+    {
+      "epoch": 0.8166174853902679,
+      "grad_norm": 0.171875,
+      "learning_rate": 0.02,
+      "loss": 1.5315,
+      "step": 8821
+    },
+    {
+      "epoch": 0.816710061910548,
+      "grad_norm": 0.150390625,
+      "learning_rate": 0.02,
+      "loss": 1.512,
+      "step": 8822
+    },
+    {
+      "epoch": 0.816802638430828,
+      "grad_norm": 0.1650390625,
+      "learning_rate": 0.02,
+      "loss": 1.5513,
+      "step": 8823
+    },
+    {
+      "epoch": 0.8168952149511081,
+      "grad_norm": 0.1611328125,
+      "learning_rate": 0.02,
+      "loss": 1.4754,
+      "step": 8824
+    },
+    {
+      "epoch": 0.8169877914713881,
+      "grad_norm": 0.1611328125,
+      "learning_rate": 0.02,
+      "loss": 1.4866,
+      "step": 8825
+    },
+    {
+      "epoch": 0.8170803679916682,
+      "grad_norm": 0.150390625,
+      "learning_rate": 0.02,
+      "loss": 1.47,
+      "step": 8826
+    },
+    {
+      "epoch": 0.8171729445119481,
+      "grad_norm": 0.162109375,
+      "learning_rate": 0.02,
+      "loss": 1.5111,
+      "step": 8827
+    },
+    {
+      "epoch": 0.8172655210322282,
+      "grad_norm": 0.1591796875,
+      "learning_rate": 0.02,
+      "loss": 1.4721,
+      "step": 8828
+    },
+    {
+      "epoch": 0.8173580975525082,
+      "grad_norm": 0.1513671875,
+      "learning_rate": 0.02,
+      "loss": 1.4897,
+      "step": 8829
+    },
+    {
+      "epoch": 0.8174506740727883,
+      "grad_norm": 0.1513671875,
+      "learning_rate": 0.02,
+      "loss": 1.4561,
+      "step": 8830
+    },
+    {
+      "epoch": 0.8175432505930683,
+      "grad_norm": 0.1474609375,
+      "learning_rate": 0.02,
+      "loss": 1.4751,
+      "step": 8831
+    },
+    {
+      "epoch": 0.8176358271133484,
+      "grad_norm": 0.1484375,
+      "learning_rate": 0.02,
+      "loss": 1.4762,
+      "step": 8832
+    },
+    {
+      "epoch": 0.8177284036336284,
+      "grad_norm": 0.162109375,
+      "learning_rate": 0.02,
+      "loss": 1.4619,
+      "step": 8833
+    },
+    {
+      "epoch": 0.8178209801539085,
+      "grad_norm": 0.162109375,
+      "learning_rate": 0.02,
+      "loss": 1.5328,
+      "step": 8834
+    },
+    {
+      "epoch": 0.8179135566741885,
+      "grad_norm": 0.150390625,
+      "learning_rate": 0.02,
+      "loss": 1.502,
+      "step": 8835
+    },
+    {
+      "epoch": 0.8180061331944686,
+      "grad_norm": 0.146484375,
+      "learning_rate": 0.02,
+      "loss": 1.5011,
+      "step": 8836
+    },
+    {
+      "epoch": 0.8180987097147486,
+      "grad_norm": 0.1552734375,
+      "learning_rate": 0.02,
+      "loss": 1.5425,
+      "step": 8837
+    },
+    {
+      "epoch": 0.8181912862350287,
+      "grad_norm": 0.154296875,
+      "learning_rate": 0.02,
+      "loss": 1.5511,
+      "step": 8838
+    },
+    {
+      "epoch": 0.8182838627553087,
+      "grad_norm": 0.150390625,
+      "learning_rate": 0.02,
+      "loss": 1.4616,
+      "step": 8839
+    },
+    {
+      "epoch": 0.8183764392755887,
+      "grad_norm": 0.146484375,
+      "learning_rate": 0.02,
+      "loss": 1.5784,
+      "step": 8840
+    },
+    {
+      "epoch": 0.8184690157958687,
+      "grad_norm": 0.1533203125,
+      "learning_rate": 0.02,
+      "loss": 1.4879,
+      "step": 8841
+    },
+    {
+      "epoch": 0.8185615923161488,
+      "grad_norm": 0.1591796875,
+      "learning_rate": 0.02,
+      "loss": 1.4609,
+      "step": 8842
+    },
+    {
+      "epoch": 0.8186541688364288,
+      "grad_norm": 0.1572265625,
+      "learning_rate": 0.02,
+      "loss": 1.4845,
+      "step": 8843
+    },
+    {
+      "epoch": 0.8187467453567089,
+      "grad_norm": 0.15234375,
+      "learning_rate": 0.02,
+      "loss": 1.4934,
+      "step": 8844
+    },
+    {
+      "epoch": 0.8188393218769889,
+      "grad_norm": 0.15625,
+      "learning_rate": 0.02,
+      "loss": 1.4387,
+      "step": 8845
+    },
+    {
+      "epoch": 0.818931898397269,
+      "grad_norm": 0.1640625,
+      "learning_rate": 0.02,
+      "loss": 1.4974,
+      "step": 8846
+    },
+    {
+      "epoch": 0.819024474917549,
+      "grad_norm": 0.162109375,
+      "learning_rate": 0.02,
+      "loss": 1.497,
+      "step": 8847
+    },
+    {
+      "epoch": 0.8191170514378291,
+      "grad_norm": 0.1748046875,
+      "learning_rate": 0.02,
+      "loss": 1.5194,
+      "step": 8848
+    },
+    {
+      "epoch": 0.8192096279581091,
+      "grad_norm": 0.1591796875,
+      "learning_rate": 0.02,
+      "loss": 1.5171,
+      "step": 8849
+    },
+    {
+      "epoch": 0.8193022044783892,
+      "grad_norm": 0.169921875,
+      "learning_rate": 0.02,
+      "loss": 1.5205,
+      "step": 8850
+    },
+    {
+      "epoch": 0.8193947809986692,
+      "grad_norm": 0.150390625,
+      "learning_rate": 0.02,
+      "loss": 1.4459,
+      "step": 8851
+    },
+    {
+      "epoch": 0.8194873575189493,
+      "grad_norm": 0.169921875,
+      "learning_rate": 0.02,
+      "loss": 1.5272,
+      "step": 8852
+    },
+    {
+      "epoch": 0.8195799340392294,
+      "grad_norm": 0.146484375,
+      "learning_rate": 0.02,
+      "loss": 1.4741,
+      "step": 8853
+    },
+    {
+      "epoch": 0.8195799340392294,
+      "eval_loss": 1.4963678121566772,
+      "eval_runtime": 70.3423,
+      "eval_samples_per_second": 21.665,
+      "eval_steps_per_second": 5.416,
+      "step": 8853
+    },
+    {
+      "epoch": 0.8196725105595093,
+      "grad_norm": 0.162109375,
+      "learning_rate": 0.02,
+      "loss": 1.5684,
+      "step": 8854
+    },
+    {
+      "epoch": 0.8197650870797893,
+      "grad_norm": 0.1494140625,
+      "learning_rate": 0.02,
+      "loss": 1.5081,
+      "step": 8855
+    },
+    {
+      "epoch": 0.8198576636000694,
+      "grad_norm": 0.1533203125,
+      "learning_rate": 0.02,
+      "loss": 1.5195,
+      "step": 8856
+    }
+  ],
+  "logging_steps": 1,
+  "max_steps": 16203,
+  "num_input_tokens_seen": 0,
+  "num_train_epochs": 2,
+  "save_steps": 1476,
+  "stateful_callbacks": {
+    "TrainerControl": {
+      "args": {
+        "should_epoch_stop": false,
+        "should_evaluate": false,
+        "should_log": false,
+        "should_save": true,
+        "should_training_stop": false
+      },
+      "attributes": {}
+    }
+  },
+  "total_flos": 7.801846987594136e+18,
+  "train_batch_size": 8,
+  "trial_name": null,
+  "trial_params": null
+}