diff --git "a/trainer_state.json" "b/trainer_state.json"
new file mode 100644--- /dev/null
+++ "b/trainer_state.json"
@@ -0,0 +1,125869 @@
+{
+  "best_metric": null,
+  "best_model_checkpoint": null,
+  "epoch": 1.0,
+  "eval_steps": 500,
+  "global_step": 17977,
+  "is_hyper_param_search": false,
+  "is_local_process_zero": true,
+  "is_world_process_zero": true,
+  "log_history": [
+    {
+      "epoch": 0.0,
+      "grad_norm": 31.671595567958246,
+      "learning_rate": 1.851851851851852e-08,
+      "loss": 0.9836,
+      "step": 1
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 21.590568341208314,
+      "learning_rate": 3.703703703703704e-08,
+      "loss": 0.6951,
+      "step": 2
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 33.43736925901718,
+      "learning_rate": 5.555555555555556e-08,
+      "loss": 1.0552,
+      "step": 3
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 25.76674648270905,
+      "learning_rate": 7.407407407407409e-08,
+      "loss": 0.9979,
+      "step": 4
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 29.142254478570287,
+      "learning_rate": 9.259259259259259e-08,
+      "loss": 1.1524,
+      "step": 5
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 128.44118605473278,
+      "learning_rate": 1.1111111111111112e-07,
+      "loss": 0.9627,
+      "step": 6
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 22.844065944506312,
+      "learning_rate": 1.2962962962962964e-07,
+      "loss": 0.7732,
+      "step": 7
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 26.662421778739866,
+      "learning_rate": 1.4814814814814817e-07,
+      "loss": 0.9373,
+      "step": 8
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 24.8894090181831,
+      "learning_rate": 1.6666666666666668e-07,
+      "loss": 1.1089,
+      "step": 9
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 21.578880921091358,
+      "learning_rate": 1.8518518518518518e-07,
+      "loss": 0.6824,
+      "step": 10
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 28.74029164802737,
+      "learning_rate": 2.0370370370370374e-07,
+      "loss": 0.8942,
+      "step": 11
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 22.14964568326191,
+      "learning_rate": 2.2222222222222224e-07,
+      "loss": 0.8041,
+      "step": 12
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 32.9657912531208,
+      "learning_rate": 2.4074074074074075e-07,
+      "loss": 1.2601,
+      "step": 13
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 17.145724154465036,
+      "learning_rate": 2.592592592592593e-07,
+      "loss": 0.7905,
+      "step": 14
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 24.26835111004837,
+      "learning_rate": 2.7777777777777776e-07,
+      "loss": 0.9586,
+      "step": 15
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 22.011607835789768,
+      "learning_rate": 2.9629629629629634e-07,
+      "loss": 0.8789,
+      "step": 16
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 30.838211196101994,
+      "learning_rate": 3.148148148148148e-07,
+      "loss": 0.9217,
+      "step": 17
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 15.955309084125279,
+      "learning_rate": 3.3333333333333335e-07,
+      "loss": 0.911,
+      "step": 18
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 24.344206309908216,
+      "learning_rate": 3.518518518518519e-07,
+      "loss": 0.9155,
+      "step": 19
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 32.92989138430358,
+      "learning_rate": 3.7037037037037036e-07,
+      "loss": 0.8828,
+      "step": 20
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 22.41057192779469,
+      "learning_rate": 3.8888888888888895e-07,
+      "loss": 0.8469,
+      "step": 21
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 17.46446762413705,
+      "learning_rate": 4.074074074074075e-07,
+      "loss": 0.6855,
+      "step": 22
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 15.433059856938662,
+      "learning_rate": 4.2592592592592596e-07,
+      "loss": 0.7404,
+      "step": 23
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 10.728638383519646,
+      "learning_rate": 4.444444444444445e-07,
+      "loss": 0.6503,
+      "step": 24
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 12.213215386669342,
+      "learning_rate": 4.6296296296296297e-07,
+      "loss": 0.8609,
+      "step": 25
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 18.317938299817154,
+      "learning_rate": 4.814814814814815e-07,
+      "loss": 0.8072,
+      "step": 26
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 19.438962335560756,
+      "learning_rate": 5.000000000000001e-07,
+      "loss": 0.5746,
+      "step": 27
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 23.19453080923279,
+      "learning_rate": 5.185185185185186e-07,
+      "loss": 0.8071,
+      "step": 28
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 17.74196002097125,
+      "learning_rate": 5.37037037037037e-07,
+      "loss": 0.6765,
+      "step": 29
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 13.827465942539503,
+      "learning_rate": 5.555555555555555e-07,
+      "loss": 0.7021,
+      "step": 30
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 15.488021836868372,
+      "learning_rate": 5.740740740740741e-07,
+      "loss": 0.4757,
+      "step": 31
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 13.20162265514205,
+      "learning_rate": 5.925925925925927e-07,
+      "loss": 0.5341,
+      "step": 32
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 9.110026681402434,
+      "learning_rate": 6.111111111111112e-07,
+      "loss": 0.4415,
+      "step": 33
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 9.830477386012433,
+      "learning_rate": 6.296296296296296e-07,
+      "loss": 0.6151,
+      "step": 34
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 11.622213091116294,
+      "learning_rate": 6.481481481481481e-07,
+      "loss": 0.4269,
+      "step": 35
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 8.207427780107848,
+      "learning_rate": 6.666666666666667e-07,
+      "loss": 0.3635,
+      "step": 36
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 7.836791093758771,
+      "learning_rate": 6.851851851851853e-07,
+      "loss": 0.6096,
+      "step": 37
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 7.193647061430912,
+      "learning_rate": 7.037037037037038e-07,
+      "loss": 0.4741,
+      "step": 38
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 9.476468858625738,
+      "learning_rate": 7.222222222222222e-07,
+      "loss": 0.3655,
+      "step": 39
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 7.157861986137023,
+      "learning_rate": 7.407407407407407e-07,
+      "loss": 0.4635,
+      "step": 40
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 7.246782509697589,
+      "learning_rate": 7.592592592592593e-07,
+      "loss": 0.54,
+      "step": 41
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 4.533392537772903,
+      "learning_rate": 7.777777777777779e-07,
+      "loss": 0.1987,
+      "step": 42
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 6.082186481942415,
+      "learning_rate": 7.962962962962964e-07,
+      "loss": 0.3954,
+      "step": 43
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 4.686018035593464,
+      "learning_rate": 8.14814814814815e-07,
+      "loss": 0.2148,
+      "step": 44
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 5.285244734447013,
+      "learning_rate": 8.333333333333333e-07,
+      "loss": 0.312,
+      "step": 45
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 5.426938300900884,
+      "learning_rate": 8.518518518518519e-07,
+      "loss": 0.2549,
+      "step": 46
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 6.275553478285126,
+      "learning_rate": 8.703703703703705e-07,
+      "loss": 0.3357,
+      "step": 47
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 5.135064616590709,
+      "learning_rate": 8.88888888888889e-07,
+      "loss": 0.1967,
+      "step": 48
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 4.486540108880125,
+      "learning_rate": 9.074074074074076e-07,
+      "loss": 0.3064,
+      "step": 49
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 6.190915869898936,
+      "learning_rate": 9.259259259259259e-07,
+      "loss": 0.4127,
+      "step": 50
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 6.080158746554302,
+      "learning_rate": 9.444444444444445e-07,
+      "loss": 0.2437,
+      "step": 51
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 5.0751084508724,
+      "learning_rate": 9.62962962962963e-07,
+      "loss": 0.4358,
+      "step": 52
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 7.434165335252677,
+      "learning_rate": 9.814814814814816e-07,
+      "loss": 0.4203,
+      "step": 53
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 5.253355134896159,
+      "learning_rate": 1.0000000000000002e-06,
+      "loss": 0.3269,
+      "step": 54
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 4.243061853720817,
+      "learning_rate": 1.0185185185185185e-06,
+      "loss": 0.1813,
+      "step": 55
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 4.856767967025552,
+      "learning_rate": 1.0370370370370371e-06,
+      "loss": 0.3047,
+      "step": 56
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 5.127255350560153,
+      "learning_rate": 1.0555555555555557e-06,
+      "loss": 0.2576,
+      "step": 57
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 4.346136439290515,
+      "learning_rate": 1.074074074074074e-06,
+      "loss": 0.2684,
+      "step": 58
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 3.4920134020938165,
+      "learning_rate": 1.0925925925925927e-06,
+      "loss": 0.194,
+      "step": 59
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 4.647235144260062,
+      "learning_rate": 1.111111111111111e-06,
+      "loss": 0.224,
+      "step": 60
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 5.519727656169724,
+      "learning_rate": 1.1296296296296296e-06,
+      "loss": 0.2936,
+      "step": 61
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 8.295031154084294,
+      "learning_rate": 1.1481481481481482e-06,
+      "loss": 0.3286,
+      "step": 62
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 5.722549697873489,
+      "learning_rate": 1.1666666666666668e-06,
+      "loss": 0.4155,
+      "step": 63
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 5.875671754133045,
+      "learning_rate": 1.1851851851851854e-06,
+      "loss": 0.378,
+      "step": 64
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 6.792663865708492,
+      "learning_rate": 1.2037037037037037e-06,
+      "loss": 0.2497,
+      "step": 65
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 7.20501181162727,
+      "learning_rate": 1.2222222222222223e-06,
+      "loss": 0.3962,
+      "step": 66
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 4.794105811510582,
+      "learning_rate": 1.240740740740741e-06,
+      "loss": 0.2524,
+      "step": 67
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 5.064103025788011,
+      "learning_rate": 1.2592592592592593e-06,
+      "loss": 0.3846,
+      "step": 68
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 3.452696967890651,
+      "learning_rate": 1.2777777777777779e-06,
+      "loss": 0.2413,
+      "step": 69
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 2.9764029793033213,
+      "learning_rate": 1.2962962962962962e-06,
+      "loss": 0.1858,
+      "step": 70
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 4.887938203216541,
+      "learning_rate": 1.3148148148148148e-06,
+      "loss": 0.2745,
+      "step": 71
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 5.134226727564933,
+      "learning_rate": 1.3333333333333334e-06,
+      "loss": 0.2885,
+      "step": 72
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 5.590512921407683,
+      "learning_rate": 1.351851851851852e-06,
+      "loss": 0.2419,
+      "step": 73
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 4.674717092878751,
+      "learning_rate": 1.3703703703703706e-06,
+      "loss": 0.3462,
+      "step": 74
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 4.509220346401417,
+      "learning_rate": 1.3888888888888892e-06,
+      "loss": 0.1586,
+      "step": 75
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 3.997696481472079,
+      "learning_rate": 1.4074074074074075e-06,
+      "loss": 0.2175,
+      "step": 76
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 4.966583739405523,
+      "learning_rate": 1.4259259259259261e-06,
+      "loss": 0.3962,
+      "step": 77
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 5.093276066175207,
+      "learning_rate": 1.4444444444444445e-06,
+      "loss": 0.3756,
+      "step": 78
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 6.335763630204078,
+      "learning_rate": 1.4629629629629629e-06,
+      "loss": 0.2768,
+      "step": 79
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 5.52584594274486,
+      "learning_rate": 1.4814814814814815e-06,
+      "loss": 0.2227,
+      "step": 80
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 4.292995296600986,
+      "learning_rate": 1.5e-06,
+      "loss": 0.2931,
+      "step": 81
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 3.3855191650524232,
+      "learning_rate": 1.5185185185185186e-06,
+      "loss": 0.2184,
+      "step": 82
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 6.680707851356181,
+      "learning_rate": 1.5370370370370372e-06,
+      "loss": 0.3002,
+      "step": 83
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 4.157140363964968,
+      "learning_rate": 1.5555555555555558e-06,
+      "loss": 0.2268,
+      "step": 84
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 2.9320222630351624,
+      "learning_rate": 1.5740740740740742e-06,
+      "loss": 0.2046,
+      "step": 85
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 4.54889007696515,
+      "learning_rate": 1.5925925925925927e-06,
+      "loss": 0.2168,
+      "step": 86
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 4.0441417227247705,
+      "learning_rate": 1.6111111111111113e-06,
+      "loss": 0.2149,
+      "step": 87
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 5.7596791320081,
+      "learning_rate": 1.62962962962963e-06,
+      "loss": 0.2557,
+      "step": 88
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 4.4753129689831885,
+      "learning_rate": 1.648148148148148e-06,
+      "loss": 0.2086,
+      "step": 89
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 4.625168152922665,
+      "learning_rate": 1.6666666666666667e-06,
+      "loss": 0.3106,
+      "step": 90
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 4.778566667141804,
+      "learning_rate": 1.6851851851851852e-06,
+      "loss": 0.236,
+      "step": 91
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 3.4927683097649407,
+      "learning_rate": 1.7037037037037038e-06,
+      "loss": 0.228,
+      "step": 92
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 5.238504380858114,
+      "learning_rate": 1.7222222222222224e-06,
+      "loss": 0.2445,
+      "step": 93
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 5.26550356218346,
+      "learning_rate": 1.740740740740741e-06,
+      "loss": 0.335,
+      "step": 94
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 4.084517792100619,
+      "learning_rate": 1.7592592592592594e-06,
+      "loss": 0.2466,
+      "step": 95
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 4.059980476806942,
+      "learning_rate": 1.777777777777778e-06,
+      "loss": 0.2352,
+      "step": 96
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 5.618509680814076,
+      "learning_rate": 1.7962962962962965e-06,
+      "loss": 0.2983,
+      "step": 97
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 3.2752739027377102,
+      "learning_rate": 1.8148148148148151e-06,
+      "loss": 0.2138,
+      "step": 98
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 4.531959879286411,
+      "learning_rate": 1.8333333333333333e-06,
+      "loss": 0.2109,
+      "step": 99
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 3.719039633234832,
+      "learning_rate": 1.8518518518518519e-06,
+      "loss": 0.2742,
+      "step": 100
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 4.791294915525922,
+      "learning_rate": 1.8703703703703705e-06,
+      "loss": 0.2604,
+      "step": 101
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 4.386141106337409,
+      "learning_rate": 1.888888888888889e-06,
+      "loss": 0.2207,
+      "step": 102
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 3.693752709660448,
+      "learning_rate": 1.9074074074074076e-06,
+      "loss": 0.1844,
+      "step": 103
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 3.539126599807116,
+      "learning_rate": 1.925925925925926e-06,
+      "loss": 0.2236,
+      "step": 104
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 5.701328883111506,
+      "learning_rate": 1.944444444444445e-06,
+      "loss": 0.2436,
+      "step": 105
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 3.49144964082344,
+      "learning_rate": 1.962962962962963e-06,
+      "loss": 0.2266,
+      "step": 106
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 3.6072130209912188,
+      "learning_rate": 1.9814814814814815e-06,
+      "loss": 0.2317,
+      "step": 107
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 3.0365130068969792,
+      "learning_rate": 2.0000000000000003e-06,
+      "loss": 0.2223,
+      "step": 108
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 13.354029999428898,
+      "learning_rate": 2.0185185185185187e-06,
+      "loss": 0.2359,
+      "step": 109
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 4.866215102048313,
+      "learning_rate": 2.037037037037037e-06,
+      "loss": 0.2178,
+      "step": 110
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 4.2919768835495775,
+      "learning_rate": 2.0555555555555555e-06,
+      "loss": 0.2053,
+      "step": 111
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 3.0247999172403386,
+      "learning_rate": 2.0740740740740742e-06,
+      "loss": 0.2116,
+      "step": 112
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 4.873767819329223,
+      "learning_rate": 2.0925925925925926e-06,
+      "loss": 0.2542,
+      "step": 113
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 3.290100400539165,
+      "learning_rate": 2.1111111111111114e-06,
+      "loss": 0.2353,
+      "step": 114
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 4.751597713000032,
+      "learning_rate": 2.1296296296296298e-06,
+      "loss": 0.2315,
+      "step": 115
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 4.473212528201221,
+      "learning_rate": 2.148148148148148e-06,
+      "loss": 0.2496,
+      "step": 116
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 5.783412858689247,
+      "learning_rate": 2.166666666666667e-06,
+      "loss": 0.2554,
+      "step": 117
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 4.699102842656633,
+      "learning_rate": 2.1851851851851853e-06,
+      "loss": 0.2953,
+      "step": 118
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 3.1020105552376935,
+      "learning_rate": 2.203703703703704e-06,
+      "loss": 0.2232,
+      "step": 119
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 4.231270542404818,
+      "learning_rate": 2.222222222222222e-06,
+      "loss": 0.2855,
+      "step": 120
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 5.1755404319160485,
+      "learning_rate": 2.240740740740741e-06,
+      "loss": 0.2506,
+      "step": 121
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.832933353259535,
+      "learning_rate": 2.2592592592592592e-06,
+      "loss": 0.2209,
+      "step": 122
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 3.3747496600378026,
+      "learning_rate": 2.277777777777778e-06,
+      "loss": 0.2641,
+      "step": 123
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 3.1614197231660426,
+      "learning_rate": 2.2962962962962964e-06,
+      "loss": 0.2591,
+      "step": 124
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 3.15164363985477,
+      "learning_rate": 2.314814814814815e-06,
+      "loss": 0.2195,
+      "step": 125
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 3.1520741854423773,
+      "learning_rate": 2.3333333333333336e-06,
+      "loss": 0.2327,
+      "step": 126
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 3.612537600311694,
+      "learning_rate": 2.351851851851852e-06,
+      "loss": 0.2544,
+      "step": 127
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.658574949293894,
+      "learning_rate": 2.3703703703703707e-06,
+      "loss": 0.1689,
+      "step": 128
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.521522811253935,
+      "learning_rate": 2.388888888888889e-06,
+      "loss": 0.2293,
+      "step": 129
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 5.564732601796836,
+      "learning_rate": 2.4074074074074075e-06,
+      "loss": 0.2667,
+      "step": 130
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.44070512393425,
+      "learning_rate": 2.425925925925926e-06,
+      "loss": 0.2093,
+      "step": 131
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 3.7275314191060787,
+      "learning_rate": 2.4444444444444447e-06,
+      "loss": 0.2531,
+      "step": 132
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 3.682061533051656,
+      "learning_rate": 2.462962962962963e-06,
+      "loss": 0.2493,
+      "step": 133
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.951520736342321,
+      "learning_rate": 2.481481481481482e-06,
+      "loss": 0.2651,
+      "step": 134
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 3.1993013028962376,
+      "learning_rate": 2.5e-06,
+      "loss": 0.2193,
+      "step": 135
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 3.4100755941983794,
+      "learning_rate": 2.5185185185185186e-06,
+      "loss": 0.2403,
+      "step": 136
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 3.3821935109396386,
+      "learning_rate": 2.5370370370370374e-06,
+      "loss": 0.2758,
+      "step": 137
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.6543322877297006,
+      "learning_rate": 2.5555555555555557e-06,
+      "loss": 0.2147,
+      "step": 138
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 3.416779857404282,
+      "learning_rate": 2.5740740740740745e-06,
+      "loss": 0.2329,
+      "step": 139
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 4.095895132385728,
+      "learning_rate": 2.5925925925925925e-06,
+      "loss": 0.2328,
+      "step": 140
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.802919951884446,
+      "learning_rate": 2.6111111111111113e-06,
+      "loss": 0.2641,
+      "step": 141
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 3.3297063178072643,
+      "learning_rate": 2.6296296296296297e-06,
+      "loss": 0.237,
+      "step": 142
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 3.369922845676044,
+      "learning_rate": 2.6481481481481485e-06,
+      "loss": 0.2538,
+      "step": 143
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 3.7670644169639695,
+      "learning_rate": 2.666666666666667e-06,
+      "loss": 0.2539,
+      "step": 144
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 3.4216695924186804,
+      "learning_rate": 2.6851851851851856e-06,
+      "loss": 0.2053,
+      "step": 145
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 3.221839005854741,
+      "learning_rate": 2.703703703703704e-06,
+      "loss": 0.2127,
+      "step": 146
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 3.0235480946266806,
+      "learning_rate": 2.7222222222222224e-06,
+      "loss": 0.2106,
+      "step": 147
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.834709595232568,
+      "learning_rate": 2.740740740740741e-06,
+      "loss": 0.1946,
+      "step": 148
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 3.015178509009479,
+      "learning_rate": 2.759259259259259e-06,
+      "loss": 0.2388,
+      "step": 149
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 4.409130426513994,
+      "learning_rate": 2.7777777777777783e-06,
+      "loss": 0.2415,
+      "step": 150
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 3.3739352577710386,
+      "learning_rate": 2.7962962962962963e-06,
+      "loss": 0.2457,
+      "step": 151
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 3.4098241760142383,
+      "learning_rate": 2.814814814814815e-06,
+      "loss": 0.251,
+      "step": 152
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 3.333754001338976,
+      "learning_rate": 2.8333333333333335e-06,
+      "loss": 0.1801,
+      "step": 153
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 3.318077097848955,
+      "learning_rate": 2.8518518518518522e-06,
+      "loss": 0.2458,
+      "step": 154
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.9466885051594103,
+      "learning_rate": 2.8703703703703706e-06,
+      "loss": 0.2406,
+      "step": 155
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.731062708528279,
+      "learning_rate": 2.888888888888889e-06,
+      "loss": 0.2001,
+      "step": 156
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.9272308425569653,
+      "learning_rate": 2.907407407407408e-06,
+      "loss": 0.2369,
+      "step": 157
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.481414362366478,
+      "learning_rate": 2.9259259259259257e-06,
+      "loss": 0.1691,
+      "step": 158
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 3.0329718074790053,
+      "learning_rate": 2.944444444444445e-06,
+      "loss": 0.2387,
+      "step": 159
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 3.391297972382365,
+      "learning_rate": 2.962962962962963e-06,
+      "loss": 0.2509,
+      "step": 160
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 3.569378052344309,
+      "learning_rate": 2.9814814814814817e-06,
+      "loss": 0.2017,
+      "step": 161
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 3.7110963646864787,
+      "learning_rate": 3e-06,
+      "loss": 0.2179,
+      "step": 162
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.246489859773125,
+      "learning_rate": 3.018518518518519e-06,
+      "loss": 0.1956,
+      "step": 163
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 4.009805700489273,
+      "learning_rate": 3.0370370370370372e-06,
+      "loss": 0.2621,
+      "step": 164
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 6.603600398514934,
+      "learning_rate": 3.055555555555556e-06,
+      "loss": 0.2286,
+      "step": 165
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 3.2140092102588227,
+      "learning_rate": 3.0740740740740744e-06,
+      "loss": 0.1985,
+      "step": 166
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 3.129528788571764,
+      "learning_rate": 3.0925925925925928e-06,
+      "loss": 0.2672,
+      "step": 167
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 4.012024850134178,
+      "learning_rate": 3.1111111111111116e-06,
+      "loss": 0.217,
+      "step": 168
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 7.820458104991282,
+      "learning_rate": 3.1296296296296295e-06,
+      "loss": 0.301,
+      "step": 169
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 3.0860558209491344,
+      "learning_rate": 3.1481481481481483e-06,
+      "loss": 0.3091,
+      "step": 170
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.7755722534108647,
+      "learning_rate": 3.1666666666666667e-06,
+      "loss": 0.2226,
+      "step": 171
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.5247435245938923,
+      "learning_rate": 3.1851851851851855e-06,
+      "loss": 0.1661,
+      "step": 172
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 3.01004084492849,
+      "learning_rate": 3.203703703703704e-06,
+      "loss": 0.21,
+      "step": 173
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 4.228753627219633,
+      "learning_rate": 3.2222222222222227e-06,
+      "loss": 0.2477,
+      "step": 174
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 3.1752871871583532,
+      "learning_rate": 3.240740740740741e-06,
+      "loss": 0.2079,
+      "step": 175
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 3.152560617770134,
+      "learning_rate": 3.25925925925926e-06,
+      "loss": 0.2665,
+      "step": 176
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.961029589788596,
+      "learning_rate": 3.277777777777778e-06,
+      "loss": 0.1568,
+      "step": 177
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 4.709782042755141,
+      "learning_rate": 3.296296296296296e-06,
+      "loss": 0.2532,
+      "step": 178
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 3.8365927717296775,
+      "learning_rate": 3.314814814814815e-06,
+      "loss": 0.2034,
+      "step": 179
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 3.693611221159195,
+      "learning_rate": 3.3333333333333333e-06,
+      "loss": 0.2731,
+      "step": 180
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 3.9986957942154158,
+      "learning_rate": 3.351851851851852e-06,
+      "loss": 0.2597,
+      "step": 181
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 3.2919659156622347,
+      "learning_rate": 3.3703703703703705e-06,
+      "loss": 0.1949,
+      "step": 182
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.9341893207437875,
+      "learning_rate": 3.3888888888888893e-06,
+      "loss": 0.2386,
+      "step": 183
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 5.102536761269667,
+      "learning_rate": 3.4074074074074077e-06,
+      "loss": 0.2774,
+      "step": 184
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 3.4536820389601535,
+      "learning_rate": 3.4259259259259265e-06,
+      "loss": 0.1826,
+      "step": 185
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 3.0300989423466786,
+      "learning_rate": 3.444444444444445e-06,
+      "loss": 0.2446,
+      "step": 186
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 3.4403011526481473,
+      "learning_rate": 3.4629629629629628e-06,
+      "loss": 0.1901,
+      "step": 187
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 3.6775091247134197,
+      "learning_rate": 3.481481481481482e-06,
+      "loss": 0.2674,
+      "step": 188
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 4.391028178699074,
+      "learning_rate": 3.5e-06,
+      "loss": 0.2829,
+      "step": 189
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.423366192150436,
+      "learning_rate": 3.5185185185185187e-06,
+      "loss": 0.1824,
+      "step": 190
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 3.122177437193447,
+      "learning_rate": 3.537037037037037e-06,
+      "loss": 0.2371,
+      "step": 191
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 3.45841158448972,
+      "learning_rate": 3.555555555555556e-06,
+      "loss": 0.2292,
+      "step": 192
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 3.4469803213869747,
+      "learning_rate": 3.5740740740740743e-06,
+      "loss": 0.1955,
+      "step": 193
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.9771563520791173,
+      "learning_rate": 3.592592592592593e-06,
+      "loss": 0.2377,
+      "step": 194
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 3.8796906691318185,
+      "learning_rate": 3.6111111111111115e-06,
+      "loss": 0.2551,
+      "step": 195
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 3.468676128116348,
+      "learning_rate": 3.6296296296296302e-06,
+      "loss": 0.2039,
+      "step": 196
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.958880079590501,
+      "learning_rate": 3.6481481481481486e-06,
+      "loss": 0.1795,
+      "step": 197
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.6200401360225767,
+      "learning_rate": 3.6666666666666666e-06,
+      "loss": 0.232,
+      "step": 198
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 4.983293081071033,
+      "learning_rate": 3.6851851851851854e-06,
+      "loss": 0.1908,
+      "step": 199
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 3.0050236938894415,
+      "learning_rate": 3.7037037037037037e-06,
+      "loss": 0.2286,
+      "step": 200
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 3.3306140725673634,
+      "learning_rate": 3.7222222222222225e-06,
+      "loss": 0.2619,
+      "step": 201
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 3.4938522953418802,
+      "learning_rate": 3.740740740740741e-06,
+      "loss": 0.1959,
+      "step": 202
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 3.4177239728199855,
+      "learning_rate": 3.7592592592592597e-06,
+      "loss": 0.2236,
+      "step": 203
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 3.7652783689393297,
+      "learning_rate": 3.777777777777778e-06,
+      "loss": 0.2085,
+      "step": 204
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 3.6604435338767782,
+      "learning_rate": 3.796296296296297e-06,
+      "loss": 0.261,
+      "step": 205
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 3.113607597716548,
+      "learning_rate": 3.814814814814815e-06,
+      "loss": 0.2424,
+      "step": 206
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 4.4771669455747585,
+      "learning_rate": 3.833333333333334e-06,
+      "loss": 0.2283,
+      "step": 207
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 3.1589659628880478,
+      "learning_rate": 3.851851851851852e-06,
+      "loss": 0.2219,
+      "step": 208
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.7645376675755027,
+      "learning_rate": 3.87037037037037e-06,
+      "loss": 0.2135,
+      "step": 209
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.980265299201392,
+      "learning_rate": 3.88888888888889e-06,
+      "loss": 0.2686,
+      "step": 210
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 3.1181423283269614,
+      "learning_rate": 3.907407407407408e-06,
+      "loss": 0.223,
+      "step": 211
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.42736974480654,
+      "learning_rate": 3.925925925925926e-06,
+      "loss": 0.187,
+      "step": 212
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.8743075840685295,
+      "learning_rate": 3.944444444444445e-06,
+      "loss": 0.1818,
+      "step": 213
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.343322079058555,
+      "learning_rate": 3.962962962962963e-06,
+      "loss": 0.2371,
+      "step": 214
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.081694878178825,
+      "learning_rate": 3.9814814814814814e-06,
+      "loss": 0.2267,
+      "step": 215
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 4.3576806169853315,
+      "learning_rate": 4.000000000000001e-06,
+      "loss": 0.2567,
+      "step": 216
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 3.4811683148674706,
+      "learning_rate": 4.018518518518519e-06,
+      "loss": 0.2124,
+      "step": 217
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 3.5712780387693677,
+      "learning_rate": 4.037037037037037e-06,
+      "loss": 0.175,
+      "step": 218
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 3.10431155350948,
+      "learning_rate": 4.055555555555556e-06,
+      "loss": 0.2544,
+      "step": 219
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 3.4035525621721305,
+      "learning_rate": 4.074074074074074e-06,
+      "loss": 0.2443,
+      "step": 220
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.497804840268957,
+      "learning_rate": 4.092592592592593e-06,
+      "loss": 0.2103,
+      "step": 221
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 3.453502720676144,
+      "learning_rate": 4.111111111111111e-06,
+      "loss": 0.1928,
+      "step": 222
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.5319715107359095,
+      "learning_rate": 4.12962962962963e-06,
+      "loss": 0.3045,
+      "step": 223
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 3.652891605515027,
+      "learning_rate": 4.1481481481481485e-06,
+      "loss": 0.2066,
+      "step": 224
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 5.197667005680618,
+      "learning_rate": 4.166666666666667e-06,
+      "loss": 0.2867,
+      "step": 225
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 3.1852861083681456,
+      "learning_rate": 4.185185185185185e-06,
+      "loss": 0.2962,
+      "step": 226
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 3.309751449945396,
+      "learning_rate": 4.2037037037037045e-06,
+      "loss": 0.2399,
+      "step": 227
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 5.662600750900519,
+      "learning_rate": 4.222222222222223e-06,
+      "loss": 0.1835,
+      "step": 228
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 3.1627381367040655,
+      "learning_rate": 4.240740740740741e-06,
+      "loss": 0.2256,
+      "step": 229
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.859414772632586,
+      "learning_rate": 4.2592592592592596e-06,
+      "loss": 0.2474,
+      "step": 230
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 3.5930782685464924,
+      "learning_rate": 4.277777777777778e-06,
+      "loss": 0.2853,
+      "step": 231
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.3413935831851864,
+      "learning_rate": 4.296296296296296e-06,
+      "loss": 0.2001,
+      "step": 232
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.1343012773284142,
+      "learning_rate": 4.314814814814815e-06,
+      "loss": 0.1442,
+      "step": 233
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 4.447325533939575,
+      "learning_rate": 4.333333333333334e-06,
+      "loss": 0.279,
+      "step": 234
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.0281971364645366,
+      "learning_rate": 4.351851851851852e-06,
+      "loss": 0.1831,
+      "step": 235
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.399177738362128,
+      "learning_rate": 4.370370370370371e-06,
+      "loss": 0.2159,
+      "step": 236
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 3.9459614890346613,
+      "learning_rate": 4.388888888888889e-06,
+      "loss": 0.2585,
+      "step": 237
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 4.2484957894089685,
+      "learning_rate": 4.407407407407408e-06,
+      "loss": 0.2717,
+      "step": 238
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 3.671599268198715,
+      "learning_rate": 4.425925925925927e-06,
+      "loss": 0.2394,
+      "step": 239
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 3.005918803380453,
+      "learning_rate": 4.444444444444444e-06,
+      "loss": 0.2524,
+      "step": 240
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.6518617846617962,
+      "learning_rate": 4.462962962962963e-06,
+      "loss": 0.2379,
+      "step": 241
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.437999918608316,
+      "learning_rate": 4.481481481481482e-06,
+      "loss": 0.2195,
+      "step": 242
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.6583787466273825,
+      "learning_rate": 4.5e-06,
+      "loss": 0.1729,
+      "step": 243
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.402951258317827,
+      "learning_rate": 4.5185185185185185e-06,
+      "loss": 0.2234,
+      "step": 244
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.7857996129322897,
+      "learning_rate": 4.537037037037038e-06,
+      "loss": 0.2547,
+      "step": 245
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.49488576274901,
+      "learning_rate": 4.555555555555556e-06,
+      "loss": 0.2407,
+      "step": 246
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.868135271583363,
+      "learning_rate": 4.5740740740740745e-06,
+      "loss": 0.2208,
+      "step": 247
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.5562428105913084,
+      "learning_rate": 4.592592592592593e-06,
+      "loss": 0.2166,
+      "step": 248
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.58224768027302,
+      "learning_rate": 4.611111111111112e-06,
+      "loss": 0.2492,
+      "step": 249
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.3433573075972296,
+      "learning_rate": 4.62962962962963e-06,
+      "loss": 0.1936,
+      "step": 250
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.8626887969261507,
+      "learning_rate": 4.648148148148148e-06,
+      "loss": 0.2315,
+      "step": 251
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.5816386699164062,
+      "learning_rate": 4.666666666666667e-06,
+      "loss": 0.2163,
+      "step": 252
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.3400604453390237,
+      "learning_rate": 4.6851851851851855e-06,
+      "loss": 0.2611,
+      "step": 253
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.8458488911384965,
+      "learning_rate": 4.703703703703704e-06,
+      "loss": 0.2489,
+      "step": 254
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.522444603513024,
+      "learning_rate": 4.722222222222222e-06,
+      "loss": 0.2122,
+      "step": 255
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.3783927325650613,
+      "learning_rate": 4.7407407407407415e-06,
+      "loss": 0.1981,
+      "step": 256
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 3.019533613828299,
+      "learning_rate": 4.75925925925926e-06,
+      "loss": 0.2384,
+      "step": 257
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.918759132999941,
+      "learning_rate": 4.777777777777778e-06,
+      "loss": 0.1788,
+      "step": 258
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.4610427470635274,
+      "learning_rate": 4.796296296296297e-06,
+      "loss": 0.2268,
+      "step": 259
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.7848761312000536,
+      "learning_rate": 4.814814814814815e-06,
+      "loss": 0.2325,
+      "step": 260
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 3.4191520473690815,
+      "learning_rate": 4.833333333333333e-06,
+      "loss": 0.2172,
+      "step": 261
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 3.103758968658333,
+      "learning_rate": 4.851851851851852e-06,
+      "loss": 0.2478,
+      "step": 262
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.5482504471380185,
+      "learning_rate": 4.870370370370371e-06,
+      "loss": 0.2497,
+      "step": 263
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 3.4810407020892766,
+      "learning_rate": 4.888888888888889e-06,
+      "loss": 0.1416,
+      "step": 264
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 3.3881716312252794,
+      "learning_rate": 4.907407407407408e-06,
+      "loss": 0.2587,
+      "step": 265
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.3292168482752666,
+      "learning_rate": 4.925925925925926e-06,
+      "loss": 0.2184,
+      "step": 266
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 4.639069216123649,
+      "learning_rate": 4.944444444444445e-06,
+      "loss": 0.2131,
+      "step": 267
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.775550585330974,
+      "learning_rate": 4.962962962962964e-06,
+      "loss": 0.1981,
+      "step": 268
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.6776742740622606,
+      "learning_rate": 4.981481481481482e-06,
+      "loss": 0.18,
+      "step": 269
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 4.811276614758999,
+      "learning_rate": 5e-06,
+      "loss": 0.2509,
+      "step": 270
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 3.316089340902278,
+      "learning_rate": 5.01851851851852e-06,
+      "loss": 0.267,
+      "step": 271
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.212839461001947,
+      "learning_rate": 5.037037037037037e-06,
+      "loss": 0.2075,
+      "step": 272
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.9776739041544913,
+      "learning_rate": 5.0555555555555555e-06,
+      "loss": 0.1698,
+      "step": 273
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 3.1890958082449243,
+      "learning_rate": 5.074074074074075e-06,
+      "loss": 0.2372,
+      "step": 274
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.947525306217703,
+      "learning_rate": 5.092592592592593e-06,
+      "loss": 0.2561,
+      "step": 275
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 4.297391853458588,
+      "learning_rate": 5.1111111111111115e-06,
+      "loss": 0.2481,
+      "step": 276
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 3.322602754472144,
+      "learning_rate": 5.12962962962963e-06,
+      "loss": 0.2631,
+      "step": 277
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 3.050443252029387,
+      "learning_rate": 5.148148148148149e-06,
+      "loss": 0.2261,
+      "step": 278
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.8973907756153363,
+      "learning_rate": 5.1666666666666675e-06,
+      "loss": 0.2428,
+      "step": 279
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 4.0422139591962765,
+      "learning_rate": 5.185185185185185e-06,
+      "loss": 0.237,
+      "step": 280
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 3.3841461829891855,
+      "learning_rate": 5.203703703703704e-06,
+      "loss": 0.2163,
+      "step": 281
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.5529633837409493,
+      "learning_rate": 5.2222222222222226e-06,
+      "loss": 0.1871,
+      "step": 282
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 3.3624447392598986,
+      "learning_rate": 5.240740740740741e-06,
+      "loss": 0.2218,
+      "step": 283
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.2956439501496666,
+      "learning_rate": 5.259259259259259e-06,
+      "loss": 0.2417,
+      "step": 284
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.297097785680975,
+      "learning_rate": 5.2777777777777785e-06,
+      "loss": 0.2266,
+      "step": 285
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.7470113660102002,
+      "learning_rate": 5.296296296296297e-06,
+      "loss": 0.191,
+      "step": 286
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.3084465857781824,
+      "learning_rate": 5.314814814814815e-06,
+      "loss": 0.2177,
+      "step": 287
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.7920950850278485,
+      "learning_rate": 5.333333333333334e-06,
+      "loss": 0.2735,
+      "step": 288
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 5.370310217840505,
+      "learning_rate": 5.351851851851853e-06,
+      "loss": 0.2444,
+      "step": 289
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 5.418966695111219,
+      "learning_rate": 5.370370370370371e-06,
+      "loss": 0.2502,
+      "step": 290
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 5.1345778728764495,
+      "learning_rate": 5.388888888888889e-06,
+      "loss": 0.2527,
+      "step": 291
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 4.809604219423627,
+      "learning_rate": 5.407407407407408e-06,
+      "loss": 0.309,
+      "step": 292
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.1252464263745963,
+      "learning_rate": 5.425925925925926e-06,
+      "loss": 0.1791,
+      "step": 293
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.382409471290975,
+      "learning_rate": 5.444444444444445e-06,
+      "loss": 0.1995,
+      "step": 294
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.78654941635765,
+      "learning_rate": 5.462962962962963e-06,
+      "loss": 0.2433,
+      "step": 295
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 4.162243895055201,
+      "learning_rate": 5.481481481481482e-06,
+      "loss": 0.2002,
+      "step": 296
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 3.020240729949975,
+      "learning_rate": 5.500000000000001e-06,
+      "loss": 0.2292,
+      "step": 297
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 5.455860053216916,
+      "learning_rate": 5.518518518518518e-06,
+      "loss": 0.2278,
+      "step": 298
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.0276007075998423,
+      "learning_rate": 5.5370370370370374e-06,
+      "loss": 0.2324,
+      "step": 299
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.6593618957429763,
+      "learning_rate": 5.555555555555557e-06,
+      "loss": 0.2855,
+      "step": 300
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 3.3040564094529894,
+      "learning_rate": 5.574074074074075e-06,
+      "loss": 0.2893,
+      "step": 301
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.0887879744067774,
+      "learning_rate": 5.5925925925925926e-06,
+      "loss": 0.2725,
+      "step": 302
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.256352232891461,
+      "learning_rate": 5.611111111111112e-06,
+      "loss": 0.2021,
+      "step": 303
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.5862358935281637,
+      "learning_rate": 5.62962962962963e-06,
+      "loss": 0.2192,
+      "step": 304
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.253995397778198,
+      "learning_rate": 5.6481481481481485e-06,
+      "loss": 0.2372,
+      "step": 305
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.7795628348705796,
+      "learning_rate": 5.666666666666667e-06,
+      "loss": 0.2946,
+      "step": 306
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.6514569015090887,
+      "learning_rate": 5.685185185185186e-06,
+      "loss": 0.2368,
+      "step": 307
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.7901136196750858,
+      "learning_rate": 5.7037037037037045e-06,
+      "loss": 0.2328,
+      "step": 308
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.929473431713666,
+      "learning_rate": 5.722222222222222e-06,
+      "loss": 0.2245,
+      "step": 309
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.7336612641838127,
+      "learning_rate": 5.740740740740741e-06,
+      "loss": 0.2357,
+      "step": 310
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.833964247208147,
+      "learning_rate": 5.75925925925926e-06,
+      "loss": 0.2282,
+      "step": 311
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 3.404103521479665,
+      "learning_rate": 5.777777777777778e-06,
+      "loss": 0.2043,
+      "step": 312
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.4901435985101377,
+      "learning_rate": 5.796296296296296e-06,
+      "loss": 0.1827,
+      "step": 313
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.4801730482808333,
+      "learning_rate": 5.814814814814816e-06,
+      "loss": 0.2216,
+      "step": 314
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 3.1306132066915424,
+      "learning_rate": 5.833333333333334e-06,
+      "loss": 0.2295,
+      "step": 315
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.1971949417613255,
+      "learning_rate": 5.8518518518518515e-06,
+      "loss": 0.2182,
+      "step": 316
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.0099244050381335,
+      "learning_rate": 5.870370370370371e-06,
+      "loss": 0.2124,
+      "step": 317
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 10.956187648563132,
+      "learning_rate": 5.88888888888889e-06,
+      "loss": 0.2244,
+      "step": 318
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.237500079384062,
+      "learning_rate": 5.907407407407408e-06,
+      "loss": 0.2017,
+      "step": 319
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.7735158452533324,
+      "learning_rate": 5.925925925925926e-06,
+      "loss": 0.2866,
+      "step": 320
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 3.0634574172078675,
+      "learning_rate": 5.944444444444445e-06,
+      "loss": 0.2174,
+      "step": 321
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.082726231968373,
+      "learning_rate": 5.962962962962963e-06,
+      "loss": 0.221,
+      "step": 322
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.5734571016094367,
+      "learning_rate": 5.981481481481482e-06,
+      "loss": 0.2007,
+      "step": 323
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.6433284734602704,
+      "learning_rate": 6e-06,
+      "loss": 0.2226,
+      "step": 324
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.9103685818787577,
+      "learning_rate": 6.018518518518519e-06,
+      "loss": 0.2193,
+      "step": 325
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.3024757052308895,
+      "learning_rate": 6.037037037037038e-06,
+      "loss": 0.2581,
+      "step": 326
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 3.275150115088954,
+      "learning_rate": 6.055555555555555e-06,
+      "loss": 0.2276,
+      "step": 327
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.9732833180868155,
+      "learning_rate": 6.0740740740740745e-06,
+      "loss": 0.2077,
+      "step": 328
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.7655336812127842,
+      "learning_rate": 6.092592592592593e-06,
+      "loss": 0.2234,
+      "step": 329
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.8087553526925393,
+      "learning_rate": 6.111111111111112e-06,
+      "loss": 0.247,
+      "step": 330
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.171279420768631,
+      "learning_rate": 6.12962962962963e-06,
+      "loss": 0.2527,
+      "step": 331
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.4271539806076436,
+      "learning_rate": 6.148148148148149e-06,
+      "loss": 0.1873,
+      "step": 332
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.2561373188486256,
+      "learning_rate": 6.166666666666667e-06,
+      "loss": 0.2243,
+      "step": 333
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.0341046960919478,
+      "learning_rate": 6.1851851851851856e-06,
+      "loss": 0.263,
+      "step": 334
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.2655957647607807,
+      "learning_rate": 6.203703703703704e-06,
+      "loss": 0.2711,
+      "step": 335
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.8712782002364605,
+      "learning_rate": 6.222222222222223e-06,
+      "loss": 0.2466,
+      "step": 336
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.3390799589142879,
+      "learning_rate": 6.2407407407407415e-06,
+      "loss": 0.1433,
+      "step": 337
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 3.373163217946868,
+      "learning_rate": 6.259259259259259e-06,
+      "loss": 0.1725,
+      "step": 338
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.394808070251205,
+      "learning_rate": 6.277777777777778e-06,
+      "loss": 0.2197,
+      "step": 339
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.755518848297966,
+      "learning_rate": 6.296296296296297e-06,
+      "loss": 0.272,
+      "step": 340
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.8779923323935852,
+      "learning_rate": 6.314814814814816e-06,
+      "loss": 0.2232,
+      "step": 341
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.6353096651260834,
+      "learning_rate": 6.333333333333333e-06,
+      "loss": 0.2798,
+      "step": 342
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.763581485415892,
+      "learning_rate": 6.351851851851853e-06,
+      "loss": 0.1904,
+      "step": 343
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.758094713722052,
+      "learning_rate": 6.370370370370371e-06,
+      "loss": 0.2441,
+      "step": 344
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.9172636589026844,
+      "learning_rate": 6.3888888888888885e-06,
+      "loss": 0.2762,
+      "step": 345
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.394315538755901,
+      "learning_rate": 6.407407407407408e-06,
+      "loss": 0.2145,
+      "step": 346
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.4548469501382484,
+      "learning_rate": 6.425925925925927e-06,
+      "loss": 0.1939,
+      "step": 347
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.2452509086512755,
+      "learning_rate": 6.444444444444445e-06,
+      "loss": 0.2615,
+      "step": 348
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 3.098802544447235,
+      "learning_rate": 6.462962962962963e-06,
+      "loss": 0.2461,
+      "step": 349
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.616470147506204,
+      "learning_rate": 6.481481481481482e-06,
+      "loss": 0.2117,
+      "step": 350
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.2334691592329525,
+      "learning_rate": 6.5000000000000004e-06,
+      "loss": 0.2131,
+      "step": 351
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.5589799662286956,
+      "learning_rate": 6.51851851851852e-06,
+      "loss": 0.242,
+      "step": 352
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.659120456475795,
+      "learning_rate": 6.537037037037037e-06,
+      "loss": 0.1708,
+      "step": 353
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 3.0372077950479426,
+      "learning_rate": 6.555555555555556e-06,
+      "loss": 0.2708,
+      "step": 354
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 3.1775196661741996,
+      "learning_rate": 6.574074074074075e-06,
+      "loss": 0.2874,
+      "step": 355
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.830893410296079,
+      "learning_rate": 6.592592592592592e-06,
+      "loss": 0.2716,
+      "step": 356
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 3.7230523009836536,
+      "learning_rate": 6.6111111111111115e-06,
+      "loss": 0.1982,
+      "step": 357
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.893792415370512,
+      "learning_rate": 6.62962962962963e-06,
+      "loss": 0.1864,
+      "step": 358
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.666844618467573,
+      "learning_rate": 6.648148148148149e-06,
+      "loss": 0.2428,
+      "step": 359
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 3.608780254558276,
+      "learning_rate": 6.666666666666667e-06,
+      "loss": 0.2006,
+      "step": 360
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 3.302008298067494,
+      "learning_rate": 6.685185185185186e-06,
+      "loss": 0.201,
+      "step": 361
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.7545141320246036,
+      "learning_rate": 6.703703703703704e-06,
+      "loss": 0.2023,
+      "step": 362
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 3.310552294341296,
+      "learning_rate": 6.7222222222222235e-06,
+      "loss": 0.2524,
+      "step": 363
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 10.135064668152676,
+      "learning_rate": 6.740740740740741e-06,
+      "loss": 0.3195,
+      "step": 364
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.9020320884158277,
+      "learning_rate": 6.75925925925926e-06,
+      "loss": 0.2071,
+      "step": 365
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.27841075156875,
+      "learning_rate": 6.777777777777779e-06,
+      "loss": 0.2421,
+      "step": 366
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 3.4100360302965322,
+      "learning_rate": 6.796296296296296e-06,
+      "loss": 0.2714,
+      "step": 367
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.5363251000308034,
+      "learning_rate": 6.814814814814815e-06,
+      "loss": 0.244,
+      "step": 368
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.592834724808257,
+      "learning_rate": 6.833333333333334e-06,
+      "loss": 0.1717,
+      "step": 369
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.667653937916638,
+      "learning_rate": 6.851851851851853e-06,
+      "loss": 0.1939,
+      "step": 370
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.4491773781331814,
+      "learning_rate": 6.8703703703703704e-06,
+      "loss": 0.2883,
+      "step": 371
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.491279484468612,
+      "learning_rate": 6.88888888888889e-06,
+      "loss": 0.284,
+      "step": 372
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 3.0752719398369,
+      "learning_rate": 6.907407407407408e-06,
+      "loss": 0.2466,
+      "step": 373
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.9643781997595835,
+      "learning_rate": 6.9259259259259256e-06,
+      "loss": 0.2628,
+      "step": 374
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.0006146510934153,
+      "learning_rate": 6.944444444444445e-06,
+      "loss": 0.1835,
+      "step": 375
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 3.8235231337722313,
+      "learning_rate": 6.962962962962964e-06,
+      "loss": 0.3059,
+      "step": 376
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.05096254546488,
+      "learning_rate": 6.981481481481482e-06,
+      "loss": 0.2287,
+      "step": 377
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.4587295800172235,
+      "learning_rate": 7e-06,
+      "loss": 0.2654,
+      "step": 378
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.6549597748171516,
+      "learning_rate": 7.018518518518519e-06,
+      "loss": 0.3164,
+      "step": 379
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.9038346967059707,
+      "learning_rate": 7.0370370370370375e-06,
+      "loss": 0.2488,
+      "step": 380
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.3697078179976567,
+      "learning_rate": 7.055555555555557e-06,
+      "loss": 0.2037,
+      "step": 381
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.0954183152096797,
+      "learning_rate": 7.074074074074074e-06,
+      "loss": 0.1585,
+      "step": 382
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.2018921290610343,
+      "learning_rate": 7.0925925925925935e-06,
+      "loss": 0.1813,
+      "step": 383
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.5869216684012442,
+      "learning_rate": 7.111111111111112e-06,
+      "loss": 0.2381,
+      "step": 384
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.5440442501750655,
+      "learning_rate": 7.129629629629629e-06,
+      "loss": 0.1969,
+      "step": 385
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.2971557912660665,
+      "learning_rate": 7.1481481481481486e-06,
+      "loss": 0.2147,
+      "step": 386
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.550293529670386,
+      "learning_rate": 7.166666666666667e-06,
+      "loss": 0.1814,
+      "step": 387
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.8004297380922656,
+      "learning_rate": 7.185185185185186e-06,
+      "loss": 0.2686,
+      "step": 388
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.1173214359781203,
+      "learning_rate": 7.203703703703704e-06,
+      "loss": 0.2166,
+      "step": 389
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.518227044480827,
+      "learning_rate": 7.222222222222223e-06,
+      "loss": 0.1604,
+      "step": 390
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.8852826175316413,
+      "learning_rate": 7.240740740740741e-06,
+      "loss": 0.1795,
+      "step": 391
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.30584778108288,
+      "learning_rate": 7.2592592592592605e-06,
+      "loss": 0.2376,
+      "step": 392
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.5760050357974753,
+      "learning_rate": 7.277777777777778e-06,
+      "loss": 0.1886,
+      "step": 393
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.6085933249919047,
+      "learning_rate": 7.296296296296297e-06,
+      "loss": 0.1625,
+      "step": 394
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.8386472676550367,
+      "learning_rate": 7.314814814814816e-06,
+      "loss": 0.2517,
+      "step": 395
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.6820386575918245,
+      "learning_rate": 7.333333333333333e-06,
+      "loss": 0.253,
+      "step": 396
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.7885866053003387,
+      "learning_rate": 7.351851851851852e-06,
+      "loss": 0.2369,
+      "step": 397
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.2231783865351793,
+      "learning_rate": 7.370370370370371e-06,
+      "loss": 0.2185,
+      "step": 398
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.86403769874136,
+      "learning_rate": 7.38888888888889e-06,
+      "loss": 0.191,
+      "step": 399
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.6112782053430785,
+      "learning_rate": 7.4074074074074075e-06,
+      "loss": 0.2541,
+      "step": 400
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.1721369907794355,
+      "learning_rate": 7.425925925925927e-06,
+      "loss": 0.2131,
+      "step": 401
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.3261842511133697,
+      "learning_rate": 7.444444444444445e-06,
+      "loss": 0.2542,
+      "step": 402
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.4321492348789997,
+      "learning_rate": 7.462962962962964e-06,
+      "loss": 0.2691,
+      "step": 403
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.564152501781726,
+      "learning_rate": 7.481481481481482e-06,
+      "loss": 0.15,
+      "step": 404
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.345816845324918,
+      "learning_rate": 7.500000000000001e-06,
+      "loss": 0.2501,
+      "step": 405
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.015424997549943,
+      "learning_rate": 7.518518518518519e-06,
+      "loss": 0.2193,
+      "step": 406
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.6450364209319477,
+      "learning_rate": 7.537037037037037e-06,
+      "loss": 0.2726,
+      "step": 407
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.665537149427856,
+      "learning_rate": 7.555555555555556e-06,
+      "loss": 0.1735,
+      "step": 408
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.318529079431767,
+      "learning_rate": 7.5740740740740745e-06,
+      "loss": 0.2626,
+      "step": 409
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.9981876819623996,
+      "learning_rate": 7.592592592592594e-06,
+      "loss": 0.2393,
+      "step": 410
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.0379069550854196,
+      "learning_rate": 7.611111111111111e-06,
+      "loss": 0.2069,
+      "step": 411
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.9110119780954062,
+      "learning_rate": 7.62962962962963e-06,
+      "loss": 0.231,
+      "step": 412
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.6627143090730263,
+      "learning_rate": 7.64814814814815e-06,
+      "loss": 0.2442,
+      "step": 413
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.3799133298824615,
+      "learning_rate": 7.666666666666667e-06,
+      "loss": 0.281,
+      "step": 414
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.5833113318773921,
+      "learning_rate": 7.685185185185185e-06,
+      "loss": 0.221,
+      "step": 415
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.9057489268142138,
+      "learning_rate": 7.703703703703704e-06,
+      "loss": 0.2688,
+      "step": 416
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.8121203361980647,
+      "learning_rate": 7.722222222222223e-06,
+      "loss": 0.2347,
+      "step": 417
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 3.0687752410914997,
+      "learning_rate": 7.74074074074074e-06,
+      "loss": 0.2464,
+      "step": 418
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.8131616737058565,
+      "learning_rate": 7.75925925925926e-06,
+      "loss": 0.251,
+      "step": 419
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.9091346676877106,
+      "learning_rate": 7.77777777777778e-06,
+      "loss": 0.2727,
+      "step": 420
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.46372123585346,
+      "learning_rate": 7.796296296296297e-06,
+      "loss": 0.287,
+      "step": 421
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.8823349203033533,
+      "learning_rate": 7.814814814814816e-06,
+      "loss": 0.2292,
+      "step": 422
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.4743929849468036,
+      "learning_rate": 7.833333333333333e-06,
+      "loss": 0.2142,
+      "step": 423
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.2897706352334475,
+      "learning_rate": 7.851851851851853e-06,
+      "loss": 0.2657,
+      "step": 424
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.6830692137954337,
+      "learning_rate": 7.870370370370372e-06,
+      "loss": 0.2479,
+      "step": 425
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.5407121550492335,
+      "learning_rate": 7.88888888888889e-06,
+      "loss": 0.2246,
+      "step": 426
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.412774183574061,
+      "learning_rate": 7.907407407407409e-06,
+      "loss": 0.2763,
+      "step": 427
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.964345565494054,
+      "learning_rate": 7.925925925925926e-06,
+      "loss": 0.2587,
+      "step": 428
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.0450440983284652,
+      "learning_rate": 7.944444444444445e-06,
+      "loss": 0.1765,
+      "step": 429
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.7867699472742486,
+      "learning_rate": 7.962962962962963e-06,
+      "loss": 0.1863,
+      "step": 430
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.5398855665291156,
+      "learning_rate": 7.981481481481482e-06,
+      "loss": 0.3173,
+      "step": 431
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.046668493613578,
+      "learning_rate": 8.000000000000001e-06,
+      "loss": 0.2758,
+      "step": 432
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.6390293672556615,
+      "learning_rate": 8.018518518518519e-06,
+      "loss": 0.2715,
+      "step": 433
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.4504855224681408,
+      "learning_rate": 8.037037037037038e-06,
+      "loss": 0.2455,
+      "step": 434
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.599909976400243,
+      "learning_rate": 8.055555555555557e-06,
+      "loss": 0.2278,
+      "step": 435
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.234139456704803,
+      "learning_rate": 8.074074074074075e-06,
+      "loss": 0.2529,
+      "step": 436
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.7070969104141578,
+      "learning_rate": 8.092592592592592e-06,
+      "loss": 0.2011,
+      "step": 437
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.8321578382667267,
+      "learning_rate": 8.111111111111112e-06,
+      "loss": 0.2051,
+      "step": 438
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.9908165571547292,
+      "learning_rate": 8.12962962962963e-06,
+      "loss": 0.1968,
+      "step": 439
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.4006332821096077,
+      "learning_rate": 8.148148148148148e-06,
+      "loss": 0.2538,
+      "step": 440
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.6387770354396216,
+      "learning_rate": 8.166666666666668e-06,
+      "loss": 0.1666,
+      "step": 441
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.292311358079653,
+      "learning_rate": 8.185185185185187e-06,
+      "loss": 0.2268,
+      "step": 442
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.9779324740690347,
+      "learning_rate": 8.203703703703704e-06,
+      "loss": 0.2034,
+      "step": 443
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.9459054495158559,
+      "learning_rate": 8.222222222222222e-06,
+      "loss": 0.1963,
+      "step": 444
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.727768704983483,
+      "learning_rate": 8.240740740740741e-06,
+      "loss": 0.248,
+      "step": 445
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.762347648339636,
+      "learning_rate": 8.25925925925926e-06,
+      "loss": 0.2705,
+      "step": 446
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.0538712972668325,
+      "learning_rate": 8.277777777777778e-06,
+      "loss": 0.2157,
+      "step": 447
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.8976656510690924,
+      "learning_rate": 8.296296296296297e-06,
+      "loss": 0.2878,
+      "step": 448
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.917974747844571,
+      "learning_rate": 8.314814814814816e-06,
+      "loss": 0.2703,
+      "step": 449
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.225709818814864,
+      "learning_rate": 8.333333333333334e-06,
+      "loss": 0.2594,
+      "step": 450
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.1943687433117205,
+      "learning_rate": 8.351851851851851e-06,
+      "loss": 0.2184,
+      "step": 451
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.918405805463593,
+      "learning_rate": 8.37037037037037e-06,
+      "loss": 0.2855,
+      "step": 452
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.723128091977472,
+      "learning_rate": 8.38888888888889e-06,
+      "loss": 0.1916,
+      "step": 453
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.958157567411811,
+      "learning_rate": 8.407407407407409e-06,
+      "loss": 0.1901,
+      "step": 454
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.4763912525479064,
+      "learning_rate": 8.425925925925926e-06,
+      "loss": 0.273,
+      "step": 455
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.5981461116737503,
+      "learning_rate": 8.444444444444446e-06,
+      "loss": 0.2516,
+      "step": 456
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.1663828771437823,
+      "learning_rate": 8.462962962962963e-06,
+      "loss": 0.1716,
+      "step": 457
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.4781190549476864,
+      "learning_rate": 8.481481481481482e-06,
+      "loss": 0.2412,
+      "step": 458
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.6193928401336815,
+      "learning_rate": 8.5e-06,
+      "loss": 0.2716,
+      "step": 459
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.6339004624552358,
+      "learning_rate": 8.518518518518519e-06,
+      "loss": 0.1968,
+      "step": 460
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.260078140056335,
+      "learning_rate": 8.537037037037038e-06,
+      "loss": 0.2227,
+      "step": 461
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.7074133840265566,
+      "learning_rate": 8.555555555555556e-06,
+      "loss": 0.219,
+      "step": 462
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.5611418448064829,
+      "learning_rate": 8.574074074074075e-06,
+      "loss": 0.1963,
+      "step": 463
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.6694050753250427,
+      "learning_rate": 8.592592592592593e-06,
+      "loss": 0.205,
+      "step": 464
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.5774347382027096,
+      "learning_rate": 8.611111111111112e-06,
+      "loss": 0.2314,
+      "step": 465
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.707162173333124,
+      "learning_rate": 8.62962962962963e-06,
+      "loss": 0.2333,
+      "step": 466
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.586900107102287,
+      "learning_rate": 8.648148148148149e-06,
+      "loss": 0.1828,
+      "step": 467
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.8205988008032974,
+      "learning_rate": 8.666666666666668e-06,
+      "loss": 0.2781,
+      "step": 468
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.011013273961918,
+      "learning_rate": 8.685185185185185e-06,
+      "loss": 0.2364,
+      "step": 469
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.347424978140611,
+      "learning_rate": 8.703703703703705e-06,
+      "loss": 0.2951,
+      "step": 470
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.5832688941725674,
+      "learning_rate": 8.722222222222224e-06,
+      "loss": 0.1645,
+      "step": 471
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.0163279823952933,
+      "learning_rate": 8.740740740740741e-06,
+      "loss": 0.2116,
+      "step": 472
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.599090443091723,
+      "learning_rate": 8.759259259259259e-06,
+      "loss": 0.2338,
+      "step": 473
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.5651329244699874,
+      "learning_rate": 8.777777777777778e-06,
+      "loss": 0.2141,
+      "step": 474
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.644723516265944,
+      "learning_rate": 8.796296296296297e-06,
+      "loss": 0.2205,
+      "step": 475
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.5697078465922618,
+      "learning_rate": 8.814814814814817e-06,
+      "loss": 0.1948,
+      "step": 476
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.4385062614115038,
+      "learning_rate": 8.833333333333334e-06,
+      "loss": 0.2373,
+      "step": 477
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.911017504582339,
+      "learning_rate": 8.851851851851853e-06,
+      "loss": 0.205,
+      "step": 478
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.069808868113705,
+      "learning_rate": 8.87037037037037e-06,
+      "loss": 0.2354,
+      "step": 479
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.159608690490603,
+      "learning_rate": 8.888888888888888e-06,
+      "loss": 0.2083,
+      "step": 480
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.4701455010284599,
+      "learning_rate": 8.907407407407408e-06,
+      "loss": 0.2135,
+      "step": 481
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.8772804379750925,
+      "learning_rate": 8.925925925925927e-06,
+      "loss": 0.1787,
+      "step": 482
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.3222927854219297,
+      "learning_rate": 8.944444444444446e-06,
+      "loss": 0.2602,
+      "step": 483
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.9651963465511846,
+      "learning_rate": 8.962962962962963e-06,
+      "loss": 0.1747,
+      "step": 484
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.9266834331682001,
+      "learning_rate": 8.981481481481483e-06,
+      "loss": 0.1919,
+      "step": 485
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.3278082610152648,
+      "learning_rate": 9e-06,
+      "loss": 0.2168,
+      "step": 486
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.0527498297286506,
+      "learning_rate": 9.01851851851852e-06,
+      "loss": 0.2822,
+      "step": 487
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.4196503456996559,
+      "learning_rate": 9.037037037037037e-06,
+      "loss": 0.2282,
+      "step": 488
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.7355838892748336,
+      "learning_rate": 9.055555555555556e-06,
+      "loss": 0.2613,
+      "step": 489
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.3124993169628887,
+      "learning_rate": 9.074074074074075e-06,
+      "loss": 0.2076,
+      "step": 490
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.55742161398934,
+      "learning_rate": 9.092592592592593e-06,
+      "loss": 0.1866,
+      "step": 491
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 4.12557886859245,
+      "learning_rate": 9.111111111111112e-06,
+      "loss": 0.1925,
+      "step": 492
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.1746743501029098,
+      "learning_rate": 9.12962962962963e-06,
+      "loss": 0.1506,
+      "step": 493
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.524886716057685,
+      "learning_rate": 9.148148148148149e-06,
+      "loss": 0.2345,
+      "step": 494
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.179892451724111,
+      "learning_rate": 9.166666666666666e-06,
+      "loss": 0.2059,
+      "step": 495
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.913187582169457,
+      "learning_rate": 9.185185185185186e-06,
+      "loss": 0.2595,
+      "step": 496
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.6506808324873237,
+      "learning_rate": 9.203703703703705e-06,
+      "loss": 0.2009,
+      "step": 497
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 7.353623914997798,
+      "learning_rate": 9.222222222222224e-06,
+      "loss": 0.2359,
+      "step": 498
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.73397872452955,
+      "learning_rate": 9.240740740740742e-06,
+      "loss": 0.1601,
+      "step": 499
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.363800054935029,
+      "learning_rate": 9.25925925925926e-06,
+      "loss": 0.2178,
+      "step": 500
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 3.174926117165057,
+      "learning_rate": 9.277777777777778e-06,
+      "loss": 0.2262,
+      "step": 501
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.6754495456894998,
+      "learning_rate": 9.296296296296296e-06,
+      "loss": 0.2648,
+      "step": 502
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.1212048896630415,
+      "learning_rate": 9.314814814814815e-06,
+      "loss": 0.3119,
+      "step": 503
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.845141499967797,
+      "learning_rate": 9.333333333333334e-06,
+      "loss": 0.214,
+      "step": 504
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.6034717530580787,
+      "learning_rate": 9.351851851851854e-06,
+      "loss": 0.1861,
+      "step": 505
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.494876834777987,
+      "learning_rate": 9.370370370370371e-06,
+      "loss": 0.1951,
+      "step": 506
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.567749735780204,
+      "learning_rate": 9.38888888888889e-06,
+      "loss": 0.2589,
+      "step": 507
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.156161180339284,
+      "learning_rate": 9.407407407407408e-06,
+      "loss": 0.2119,
+      "step": 508
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.1757434074631155,
+      "learning_rate": 9.425925925925925e-06,
+      "loss": 0.2806,
+      "step": 509
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.6893535871344756,
+      "learning_rate": 9.444444444444445e-06,
+      "loss": 0.2379,
+      "step": 510
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.7028334459930194,
+      "learning_rate": 9.462962962962964e-06,
+      "loss": 0.2342,
+      "step": 511
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.573367833466672,
+      "learning_rate": 9.481481481481483e-06,
+      "loss": 0.2728,
+      "step": 512
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.629665728283705,
+      "learning_rate": 9.5e-06,
+      "loss": 0.1883,
+      "step": 513
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 5.469608392104702,
+      "learning_rate": 9.51851851851852e-06,
+      "loss": 0.2699,
+      "step": 514
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.922868529122991,
+      "learning_rate": 9.537037037037037e-06,
+      "loss": 0.2474,
+      "step": 515
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.383420583648189,
+      "learning_rate": 9.555555555555556e-06,
+      "loss": 0.2799,
+      "step": 516
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.7284155714312717,
+      "learning_rate": 9.574074074074074e-06,
+      "loss": 0.2199,
+      "step": 517
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.8042763638037695,
+      "learning_rate": 9.592592592592593e-06,
+      "loss": 0.2357,
+      "step": 518
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.8240537997703765,
+      "learning_rate": 9.611111111111112e-06,
+      "loss": 0.2571,
+      "step": 519
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.240434759867715,
+      "learning_rate": 9.62962962962963e-06,
+      "loss": 0.2571,
+      "step": 520
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.2212595515134352,
+      "learning_rate": 9.64814814814815e-06,
+      "loss": 0.2805,
+      "step": 521
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.6706904432909366,
+      "learning_rate": 9.666666666666667e-06,
+      "loss": 0.2818,
+      "step": 522
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.882622112644092,
+      "learning_rate": 9.685185185185186e-06,
+      "loss": 0.2546,
+      "step": 523
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.003335137994384,
+      "learning_rate": 9.703703703703703e-06,
+      "loss": 0.2216,
+      "step": 524
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.8173690992162865,
+      "learning_rate": 9.722222222222223e-06,
+      "loss": 0.2595,
+      "step": 525
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.4188288081930738,
+      "learning_rate": 9.740740740740742e-06,
+      "loss": 0.2453,
+      "step": 526
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 3.275836038529327,
+      "learning_rate": 9.759259259259261e-06,
+      "loss": 0.2442,
+      "step": 527
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.3602762550601097,
+      "learning_rate": 9.777777777777779e-06,
+      "loss": 0.2221,
+      "step": 528
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.4099576775919647,
+      "learning_rate": 9.796296296296298e-06,
+      "loss": 0.2381,
+      "step": 529
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.948210621819492,
+      "learning_rate": 9.814814814814815e-06,
+      "loss": 0.2942,
+      "step": 530
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.6352317839316701,
+      "learning_rate": 9.833333333333333e-06,
+      "loss": 0.2261,
+      "step": 531
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.07004143541843,
+      "learning_rate": 9.851851851851852e-06,
+      "loss": 0.1908,
+      "step": 532
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.15105594941232,
+      "learning_rate": 9.870370370370371e-06,
+      "loss": 0.2878,
+      "step": 533
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.0665011897929353,
+      "learning_rate": 9.88888888888889e-06,
+      "loss": 0.2648,
+      "step": 534
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.3177102461777124,
+      "learning_rate": 9.907407407407408e-06,
+      "loss": 0.1964,
+      "step": 535
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.6617003227758134,
+      "learning_rate": 9.925925925925927e-06,
+      "loss": 0.2061,
+      "step": 536
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.8712692178466874,
+      "learning_rate": 9.944444444444445e-06,
+      "loss": 0.2097,
+      "step": 537
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.2364405901770519,
+      "learning_rate": 9.962962962962964e-06,
+      "loss": 0.2007,
+      "step": 538
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.60292206138748,
+      "learning_rate": 9.981481481481482e-06,
+      "loss": 0.2364,
+      "step": 539
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.8072231920574293,
+      "learning_rate": 1e-05,
+      "loss": 0.2238,
+      "step": 540
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.4814728889599436,
+      "learning_rate": 9.999999918848563e-06,
+      "loss": 0.2314,
+      "step": 541
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.7318830824208753,
+      "learning_rate": 9.999999675394253e-06,
+      "loss": 0.2389,
+      "step": 542
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.3036697895542924,
+      "learning_rate": 9.999999269637079e-06,
+      "loss": 0.1491,
+      "step": 543
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.8513482106621912,
+      "learning_rate": 9.999998701577052e-06,
+      "loss": 0.2016,
+      "step": 544
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.2647752911372168,
+      "learning_rate": 9.999997971214192e-06,
+      "loss": 0.2791,
+      "step": 545
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.897363047140287,
+      "learning_rate": 9.999997078548524e-06,
+      "loss": 0.2325,
+      "step": 546
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.5670415512729239,
+      "learning_rate": 9.999996023580074e-06,
+      "loss": 0.2251,
+      "step": 547
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.6066296463318857,
+      "learning_rate": 9.99999480630888e-06,
+      "loss": 0.2611,
+      "step": 548
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.4653422867881003,
+      "learning_rate": 9.999993426734977e-06,
+      "loss": 0.2573,
+      "step": 549
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.5471928587812813,
+      "learning_rate": 9.999991884858413e-06,
+      "loss": 0.2317,
+      "step": 550
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.1788321358326717,
+      "learning_rate": 9.999990180679238e-06,
+      "loss": 0.2661,
+      "step": 551
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.9518145246539516,
+      "learning_rate": 9.999988314197507e-06,
+      "loss": 0.1772,
+      "step": 552
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.6427564762279832,
+      "learning_rate": 9.999986285413278e-06,
+      "loss": 0.2232,
+      "step": 553
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.7725487637369426,
+      "learning_rate": 9.99998409432662e-06,
+      "loss": 0.1818,
+      "step": 554
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.7108703016558175,
+      "learning_rate": 9.999981740937604e-06,
+      "loss": 0.1876,
+      "step": 555
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.8711769627044077,
+      "learning_rate": 9.999979225246305e-06,
+      "loss": 0.3121,
+      "step": 556
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.7045900117922645,
+      "learning_rate": 9.999976547252805e-06,
+      "loss": 0.1916,
+      "step": 557
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.048128523173138,
+      "learning_rate": 9.999973706957191e-06,
+      "loss": 0.2373,
+      "step": 558
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.9221990901329062,
+      "learning_rate": 9.999970704359555e-06,
+      "loss": 0.2004,
+      "step": 559
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.1596146848014603,
+      "learning_rate": 9.999967539459996e-06,
+      "loss": 0.2664,
+      "step": 560
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.2490925057696007,
+      "learning_rate": 9.999964212258612e-06,
+      "loss": 0.1911,
+      "step": 561
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.7084579310879742,
+      "learning_rate": 9.999960722755519e-06,
+      "loss": 0.2211,
+      "step": 562
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.8416896032362209,
+      "learning_rate": 9.999957070950824e-06,
+      "loss": 0.2526,
+      "step": 563
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.7889661492301776,
+      "learning_rate": 9.999953256844646e-06,
+      "loss": 0.2224,
+      "step": 564
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.8237790858313632,
+      "learning_rate": 9.999949280437111e-06,
+      "loss": 0.1634,
+      "step": 565
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.4797735469173663,
+      "learning_rate": 9.999945141728348e-06,
+      "loss": 0.2205,
+      "step": 566
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.4426823569233567,
+      "learning_rate": 9.99994084071849e-06,
+      "loss": 0.3041,
+      "step": 567
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.743574684996697,
+      "learning_rate": 9.999936377407677e-06,
+      "loss": 0.2238,
+      "step": 568
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.032513000231454,
+      "learning_rate": 9.999931751796055e-06,
+      "loss": 0.1836,
+      "step": 569
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.299920431087586,
+      "learning_rate": 9.99992696388377e-06,
+      "loss": 0.2829,
+      "step": 570
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.456484249618141,
+      "learning_rate": 9.999922013670984e-06,
+      "loss": 0.1619,
+      "step": 571
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.8247965197914737,
+      "learning_rate": 9.999916901157852e-06,
+      "loss": 0.2324,
+      "step": 572
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.8103485101634382,
+      "learning_rate": 9.999911626344544e-06,
+      "loss": 0.2591,
+      "step": 573
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.4997873254847325,
+      "learning_rate": 9.999906189231228e-06,
+      "loss": 0.2105,
+      "step": 574
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.6577781968518213,
+      "learning_rate": 9.999900589818083e-06,
+      "loss": 0.2225,
+      "step": 575
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.0842225165438453,
+      "learning_rate": 9.99989482810529e-06,
+      "loss": 0.1728,
+      "step": 576
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.4611910074773522,
+      "learning_rate": 9.999888904093035e-06,
+      "loss": 0.193,
+      "step": 577
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.2211246342575959,
+      "learning_rate": 9.99988281778151e-06,
+      "loss": 0.1969,
+      "step": 578
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.8041481599306388,
+      "learning_rate": 9.999876569170916e-06,
+      "loss": 0.291,
+      "step": 579
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.436450340993971,
+      "learning_rate": 9.99987015826145e-06,
+      "loss": 0.3033,
+      "step": 580
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.5594826042435612,
+      "learning_rate": 9.999863585053327e-06,
+      "loss": 0.1409,
+      "step": 581
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.9152897928877035,
+      "learning_rate": 9.999856849546757e-06,
+      "loss": 0.267,
+      "step": 582
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.6390096341345133,
+      "learning_rate": 9.999849951741955e-06,
+      "loss": 0.2145,
+      "step": 583
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.903619272035446,
+      "learning_rate": 9.999842891639151e-06,
+      "loss": 0.2076,
+      "step": 584
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.055727158980303,
+      "learning_rate": 9.999835669238571e-06,
+      "loss": 0.2839,
+      "step": 585
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.464435144020495,
+      "learning_rate": 9.999828284540451e-06,
+      "loss": 0.178,
+      "step": 586
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.2701477725215002,
+      "learning_rate": 9.99982073754503e-06,
+      "loss": 0.2349,
+      "step": 587
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.1983313414618901,
+      "learning_rate": 9.999813028252551e-06,
+      "loss": 0.2407,
+      "step": 588
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.4233421299492355,
+      "learning_rate": 9.999805156663267e-06,
+      "loss": 0.2286,
+      "step": 589
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.4703401938433902,
+      "learning_rate": 9.999797122777433e-06,
+      "loss": 0.1695,
+      "step": 590
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.2974353988467726,
+      "learning_rate": 9.999788926595309e-06,
+      "loss": 0.2403,
+      "step": 591
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.754992726258613,
+      "learning_rate": 9.99978056811716e-06,
+      "loss": 0.2765,
+      "step": 592
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.109167652182334,
+      "learning_rate": 9.999772047343259e-06,
+      "loss": 0.2608,
+      "step": 593
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.7035748167716884,
+      "learning_rate": 9.999763364273884e-06,
+      "loss": 0.2052,
+      "step": 594
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.569435155660177,
+      "learning_rate": 9.999754518909314e-06,
+      "loss": 0.2147,
+      "step": 595
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.4123044469514268,
+      "learning_rate": 9.999745511249837e-06,
+      "loss": 0.2548,
+      "step": 596
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.2550973923976543,
+      "learning_rate": 9.999736341295746e-06,
+      "loss": 0.2043,
+      "step": 597
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.0765398045800105,
+      "learning_rate": 9.999727009047337e-06,
+      "loss": 0.2845,
+      "step": 598
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.6420925960269117,
+      "learning_rate": 9.999717514504917e-06,
+      "loss": 0.2908,
+      "step": 599
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.5955704665399983,
+      "learning_rate": 9.99970785766879e-06,
+      "loss": 0.204,
+      "step": 600
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.9013730574616672,
+      "learning_rate": 9.99969803853927e-06,
+      "loss": 0.2616,
+      "step": 601
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.300116200022206,
+      "learning_rate": 9.999688057116679e-06,
+      "loss": 0.1658,
+      "step": 602
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.1232774516287902,
+      "learning_rate": 9.999677913401335e-06,
+      "loss": 0.2344,
+      "step": 603
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.7319872675478898,
+      "learning_rate": 9.999667607393576e-06,
+      "loss": 0.2378,
+      "step": 604
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.4052874396755204,
+      "learning_rate": 9.999657139093727e-06,
+      "loss": 0.2119,
+      "step": 605
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.8178640032969893,
+      "learning_rate": 9.999646508502134e-06,
+      "loss": 0.2131,
+      "step": 606
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.824809053601352,
+      "learning_rate": 9.99963571561914e-06,
+      "loss": 0.2898,
+      "step": 607
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 18.329759841943982,
+      "learning_rate": 9.999624760445095e-06,
+      "loss": 0.3558,
+      "step": 608
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 4.959488207961797,
+      "learning_rate": 9.999613642980358e-06,
+      "loss": 0.2528,
+      "step": 609
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 10.810279662319726,
+      "learning_rate": 9.999602363225284e-06,
+      "loss": 0.2493,
+      "step": 610
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.752698701205404,
+      "learning_rate": 9.999590921180244e-06,
+      "loss": 0.276,
+      "step": 611
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 43.669199514580086,
+      "learning_rate": 9.999579316845607e-06,
+      "loss": 0.2679,
+      "step": 612
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 20.58903500382208,
+      "learning_rate": 9.99956755022175e-06,
+      "loss": 0.3086,
+      "step": 613
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 17.772749385717987,
+      "learning_rate": 9.999555621309057e-06,
+      "loss": 0.2392,
+      "step": 614
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 39.173913440838724,
+      "learning_rate": 9.999543530107912e-06,
+      "loss": 0.3374,
+      "step": 615
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 12.994662509599983,
+      "learning_rate": 9.99953127661871e-06,
+      "loss": 0.3013,
+      "step": 616
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 81.81398140634857,
+      "learning_rate": 9.999518860841847e-06,
+      "loss": 0.3234,
+      "step": 617
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 41.446118041182025,
+      "learning_rate": 9.999506282777727e-06,
+      "loss": 0.3536,
+      "step": 618
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 29.228570897145367,
+      "learning_rate": 9.999493542426759e-06,
+      "loss": 0.4052,
+      "step": 619
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 15.15354187357157,
+      "learning_rate": 9.999480639789355e-06,
+      "loss": 0.2878,
+      "step": 620
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 6.612631786742226,
+      "learning_rate": 9.999467574865935e-06,
+      "loss": 0.2893,
+      "step": 621
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 24.43898009674187,
+      "learning_rate": 9.999454347656923e-06,
+      "loss": 0.3064,
+      "step": 622
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 14.64972989090464,
+      "learning_rate": 9.999440958162747e-06,
+      "loss": 0.3189,
+      "step": 623
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 3.2441817068017054,
+      "learning_rate": 9.999427406383843e-06,
+      "loss": 0.2497,
+      "step": 624
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 6.36137774749413,
+      "learning_rate": 9.99941369232065e-06,
+      "loss": 0.304,
+      "step": 625
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 5.610408422117223,
+      "learning_rate": 9.999399815973615e-06,
+      "loss": 0.2792,
+      "step": 626
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.644214887304561,
+      "learning_rate": 9.999385777343185e-06,
+      "loss": 0.2389,
+      "step": 627
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 4.995091961958238,
+      "learning_rate": 9.999371576429819e-06,
+      "loss": 0.1999,
+      "step": 628
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 9.846007654361046,
+      "learning_rate": 9.999357213233978e-06,
+      "loss": 0.32,
+      "step": 629
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 14.346899613459236,
+      "learning_rate": 9.999342687756126e-06,
+      "loss": 0.229,
+      "step": 630
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.7492375075104223,
+      "learning_rate": 9.999327999996735e-06,
+      "loss": 0.2714,
+      "step": 631
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.713373966378423,
+      "learning_rate": 9.999313149956282e-06,
+      "loss": 0.2408,
+      "step": 632
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.5228538250927752,
+      "learning_rate": 9.99929813763525e-06,
+      "loss": 0.2474,
+      "step": 633
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.5658819838583038,
+      "learning_rate": 9.999282963034126e-06,
+      "loss": 0.2009,
+      "step": 634
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.3916280188471633,
+      "learning_rate": 9.999267626153401e-06,
+      "loss": 0.2009,
+      "step": 635
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 2.7101261122944824,
+      "learning_rate": 9.999252126993574e-06,
+      "loss": 0.2652,
+      "step": 636
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.7148812535384157,
+      "learning_rate": 9.999236465555152e-06,
+      "loss": 0.2354,
+      "step": 637
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 2.0234141624607167,
+      "learning_rate": 9.999220641838637e-06,
+      "loss": 0.2267,
+      "step": 638
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.3750651300784082,
+      "learning_rate": 9.999204655844543e-06,
+      "loss": 0.2218,
+      "step": 639
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.7894919600454031,
+      "learning_rate": 9.999188507573394e-06,
+      "loss": 0.2294,
+      "step": 640
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.253131242210937,
+      "learning_rate": 9.99917219702571e-06,
+      "loss": 0.2007,
+      "step": 641
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.7178947835267584,
+      "learning_rate": 9.999155724202022e-06,
+      "loss": 0.2208,
+      "step": 642
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.4817840701843323,
+      "learning_rate": 9.999139089102866e-06,
+      "loss": 0.1922,
+      "step": 643
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 2.2511555432970494,
+      "learning_rate": 9.999122291728778e-06,
+      "loss": 0.2371,
+      "step": 644
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.3644123822672882,
+      "learning_rate": 9.999105332080306e-06,
+      "loss": 0.1903,
+      "step": 645
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.2690162200868096,
+      "learning_rate": 9.999088210158001e-06,
+      "loss": 0.217,
+      "step": 646
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.7579445153572213,
+      "learning_rate": 9.999070925962418e-06,
+      "loss": 0.237,
+      "step": 647
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 2.7747401066383834,
+      "learning_rate": 9.999053479494118e-06,
+      "loss": 0.2233,
+      "step": 648
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.625083673597082,
+      "learning_rate": 9.999035870753669e-06,
+      "loss": 0.2504,
+      "step": 649
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 2.455600206556279,
+      "learning_rate": 9.99901809974164e-06,
+      "loss": 0.3197,
+      "step": 650
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.3696478585684608,
+      "learning_rate": 9.999000166458607e-06,
+      "loss": 0.1852,
+      "step": 651
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 2.9877473640982473,
+      "learning_rate": 9.998982070905155e-06,
+      "loss": 0.2738,
+      "step": 652
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.5267221376121667,
+      "learning_rate": 9.99896381308187e-06,
+      "loss": 0.2145,
+      "step": 653
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 2.62593829324627,
+      "learning_rate": 9.998945392989347e-06,
+      "loss": 0.2533,
+      "step": 654
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 2.001065789798318,
+      "learning_rate": 9.99892681062818e-06,
+      "loss": 0.2617,
+      "step": 655
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 2.603263756945268,
+      "learning_rate": 9.998908065998975e-06,
+      "loss": 0.2039,
+      "step": 656
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.956740344387672,
+      "learning_rate": 9.998889159102339e-06,
+      "loss": 0.2122,
+      "step": 657
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.584203293391321,
+      "learning_rate": 9.998870089938886e-06,
+      "loss": 0.1865,
+      "step": 658
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 2.053990771737211,
+      "learning_rate": 9.998850858509237e-06,
+      "loss": 0.2554,
+      "step": 659
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 2.405207222241233,
+      "learning_rate": 9.998831464814014e-06,
+      "loss": 0.254,
+      "step": 660
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 2.808261957851011,
+      "learning_rate": 9.998811908853847e-06,
+      "loss": 0.2548,
+      "step": 661
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.5276001400400812,
+      "learning_rate": 9.998792190629369e-06,
+      "loss": 0.2019,
+      "step": 662
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 2.2381691796438554,
+      "learning_rate": 9.998772310141225e-06,
+      "loss": 0.2737,
+      "step": 663
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 2.413871066209702,
+      "learning_rate": 9.998752267390055e-06,
+      "loss": 0.2401,
+      "step": 664
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.229324476334034,
+      "learning_rate": 9.998732062376513e-06,
+      "loss": 0.1934,
+      "step": 665
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.7869477329215326,
+      "learning_rate": 9.998711695101254e-06,
+      "loss": 0.2307,
+      "step": 666
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.2130204037635122,
+      "learning_rate": 9.998691165564938e-06,
+      "loss": 0.222,
+      "step": 667
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.5596636826017594,
+      "learning_rate": 9.998670473768232e-06,
+      "loss": 0.2044,
+      "step": 668
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.3948911007574005,
+      "learning_rate": 9.998649619711808e-06,
+      "loss": 0.2502,
+      "step": 669
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 2.359095121040132,
+      "learning_rate": 9.998628603396345e-06,
+      "loss": 0.2524,
+      "step": 670
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.7274161942575648,
+      "learning_rate": 9.998607424822522e-06,
+      "loss": 0.2523,
+      "step": 671
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 3.43300454353923,
+      "learning_rate": 9.998586083991028e-06,
+      "loss": 0.2306,
+      "step": 672
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 2.1139468627339784,
+      "learning_rate": 9.998564580902555e-06,
+      "loss": 0.2472,
+      "step": 673
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 2.1596330005523834,
+      "learning_rate": 9.9985429155578e-06,
+      "loss": 0.2363,
+      "step": 674
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 2.280091409544863,
+      "learning_rate": 9.998521087957468e-06,
+      "loss": 0.2438,
+      "step": 675
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.9700134628218318,
+      "learning_rate": 9.998499098102268e-06,
+      "loss": 0.2778,
+      "step": 676
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 2.06506438214983,
+      "learning_rate": 9.998476945992913e-06,
+      "loss": 0.2987,
+      "step": 677
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.8740205073967353,
+      "learning_rate": 9.998454631630123e-06,
+      "loss": 0.2418,
+      "step": 678
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.8872574773562758,
+      "learning_rate": 9.998432155014619e-06,
+      "loss": 0.2837,
+      "step": 679
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.4657243828131423,
+      "learning_rate": 9.998409516147135e-06,
+      "loss": 0.1898,
+      "step": 680
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 2.3584560221400532,
+      "learning_rate": 9.998386715028403e-06,
+      "loss": 0.2326,
+      "step": 681
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.8820926350629803,
+      "learning_rate": 9.998363751659163e-06,
+      "loss": 0.1695,
+      "step": 682
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 2.9141438654363108,
+      "learning_rate": 9.998340626040162e-06,
+      "loss": 0.2036,
+      "step": 683
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.7084189934789658,
+      "learning_rate": 9.99831733817215e-06,
+      "loss": 0.2271,
+      "step": 684
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.3658176248622982,
+      "learning_rate": 9.998293888055882e-06,
+      "loss": 0.2688,
+      "step": 685
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.4298539742233574,
+      "learning_rate": 9.998270275692123e-06,
+      "loss": 0.2012,
+      "step": 686
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 2.1142214501053354,
+      "learning_rate": 9.998246501081635e-06,
+      "loss": 0.2675,
+      "step": 687
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.7491912345971081,
+      "learning_rate": 9.998222564225191e-06,
+      "loss": 0.2313,
+      "step": 688
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.9556946882636561,
+      "learning_rate": 9.998198465123569e-06,
+      "loss": 0.2405,
+      "step": 689
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.733282763422793,
+      "learning_rate": 9.998174203777549e-06,
+      "loss": 0.2521,
+      "step": 690
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.5037118810910108,
+      "learning_rate": 9.998149780187922e-06,
+      "loss": 0.1822,
+      "step": 691
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 2.9634766172075806,
+      "learning_rate": 9.998125194355479e-06,
+      "loss": 0.2111,
+      "step": 692
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.5280912623747476,
+      "learning_rate": 9.998100446281018e-06,
+      "loss": 0.2512,
+      "step": 693
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.7715209591251764,
+      "learning_rate": 9.998075535965342e-06,
+      "loss": 0.2257,
+      "step": 694
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.9308633117820866,
+      "learning_rate": 9.998050463409259e-06,
+      "loss": 0.2138,
+      "step": 695
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.2120358669477806,
+      "learning_rate": 9.998025228613586e-06,
+      "loss": 0.2171,
+      "step": 696
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 3.948725910167626,
+      "learning_rate": 9.997999831579138e-06,
+      "loss": 0.2203,
+      "step": 697
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.3921277895023731,
+      "learning_rate": 9.997974272306742e-06,
+      "loss": 0.2406,
+      "step": 698
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.2562402812026698,
+      "learning_rate": 9.997948550797227e-06,
+      "loss": 0.165,
+      "step": 699
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.610502840874703,
+      "learning_rate": 9.997922667051429e-06,
+      "loss": 0.2431,
+      "step": 700
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.276328494150188,
+      "learning_rate": 9.997896621070187e-06,
+      "loss": 0.1982,
+      "step": 701
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 2.1114722669380392,
+      "learning_rate": 9.997870412854347e-06,
+      "loss": 0.2066,
+      "step": 702
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 2.78201244918868,
+      "learning_rate": 9.997844042404758e-06,
+      "loss": 0.2583,
+      "step": 703
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.535084620835797,
+      "learning_rate": 9.997817509722279e-06,
+      "loss": 0.2257,
+      "step": 704
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.8874994325321177,
+      "learning_rate": 9.997790814807769e-06,
+      "loss": 0.2009,
+      "step": 705
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.6047968049973862,
+      "learning_rate": 9.997763957662094e-06,
+      "loss": 0.2602,
+      "step": 706
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.7570357757703434,
+      "learning_rate": 9.997736938286129e-06,
+      "loss": 0.2485,
+      "step": 707
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 3.2229389968292472,
+      "learning_rate": 9.997709756680746e-06,
+      "loss": 0.2572,
+      "step": 708
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.8492538247369952,
+      "learning_rate": 9.997682412846835e-06,
+      "loss": 0.2053,
+      "step": 709
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.7469708897898746,
+      "learning_rate": 9.997654906785274e-06,
+      "loss": 0.2146,
+      "step": 710
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 2.2291057918531525,
+      "learning_rate": 9.997627238496964e-06,
+      "loss": 0.2099,
+      "step": 711
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.3402541063235494,
+      "learning_rate": 9.9975994079828e-06,
+      "loss": 0.1807,
+      "step": 712
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.922145502701153,
+      "learning_rate": 9.997571415243684e-06,
+      "loss": 0.236,
+      "step": 713
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.4461143048070895,
+      "learning_rate": 9.997543260280527e-06,
+      "loss": 0.1837,
+      "step": 714
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.3119016202921865,
+      "learning_rate": 9.997514943094243e-06,
+      "loss": 0.2211,
+      "step": 715
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 5.0686545931001525,
+      "learning_rate": 9.997486463685748e-06,
+      "loss": 0.1679,
+      "step": 716
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 2.8344728665347683,
+      "learning_rate": 9.99745782205597e-06,
+      "loss": 0.2857,
+      "step": 717
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 2.4148963426375163,
+      "learning_rate": 9.997429018205838e-06,
+      "loss": 0.246,
+      "step": 718
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 2.3087572791826942,
+      "learning_rate": 9.997400052136284e-06,
+      "loss": 0.2206,
+      "step": 719
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.8846658472718998,
+      "learning_rate": 9.997370923848253e-06,
+      "loss": 0.2204,
+      "step": 720
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 2.4933567951325903,
+      "learning_rate": 9.997341633342685e-06,
+      "loss": 0.1732,
+      "step": 721
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.4171284742165549,
+      "learning_rate": 9.997312180620538e-06,
+      "loss": 0.1865,
+      "step": 722
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.7608028746274034,
+      "learning_rate": 9.99728256568276e-06,
+      "loss": 0.2024,
+      "step": 723
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 2.12302828221747,
+      "learning_rate": 9.997252788530318e-06,
+      "loss": 0.2497,
+      "step": 724
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 26.926509571172836,
+      "learning_rate": 9.997222849164179e-06,
+      "loss": 0.2529,
+      "step": 725
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 9.988579785507007,
+      "learning_rate": 9.997192747585309e-06,
+      "loss": 0.2325,
+      "step": 726
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 2.073780616996417,
+      "learning_rate": 9.997162483794692e-06,
+      "loss": 0.2371,
+      "step": 727
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 83.88128606717183,
+      "learning_rate": 9.997132057793306e-06,
+      "loss": 0.2815,
+      "step": 728
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.6923507763339276,
+      "learning_rate": 9.99710146958214e-06,
+      "loss": 0.205,
+      "step": 729
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.6437592540596508,
+      "learning_rate": 9.997070719162185e-06,
+      "loss": 0.2277,
+      "step": 730
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 2.1464775022860074,
+      "learning_rate": 9.997039806534441e-06,
+      "loss": 0.2652,
+      "step": 731
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.7918314370547077,
+      "learning_rate": 9.997008731699914e-06,
+      "loss": 0.1765,
+      "step": 732
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.1126204754928446,
+      "learning_rate": 9.99697749465961e-06,
+      "loss": 0.2301,
+      "step": 733
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.9214132517978426,
+      "learning_rate": 9.99694609541454e-06,
+      "loss": 0.2226,
+      "step": 734
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.3356586968469206,
+      "learning_rate": 9.996914533965729e-06,
+      "loss": 0.2596,
+      "step": 735
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 2.049688458611854,
+      "learning_rate": 9.996882810314198e-06,
+      "loss": 0.2259,
+      "step": 736
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 2.5737350773575582,
+      "learning_rate": 9.996850924460977e-06,
+      "loss": 0.2308,
+      "step": 737
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.5406058289958093,
+      "learning_rate": 9.996818876407101e-06,
+      "loss": 0.2375,
+      "step": 738
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.538738234726375,
+      "learning_rate": 9.996786666153612e-06,
+      "loss": 0.2012,
+      "step": 739
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.453470927255759,
+      "learning_rate": 9.996754293701556e-06,
+      "loss": 0.2343,
+      "step": 740
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.6459359356329737,
+      "learning_rate": 9.99672175905198e-06,
+      "loss": 0.2573,
+      "step": 741
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.7620574473342865,
+      "learning_rate": 9.996689062205942e-06,
+      "loss": 0.2457,
+      "step": 742
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.7957641319017301,
+      "learning_rate": 9.996656203164505e-06,
+      "loss": 0.2719,
+      "step": 743
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.5850265357732467,
+      "learning_rate": 9.996623181928735e-06,
+      "loss": 0.2108,
+      "step": 744
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 6.063883358148109,
+      "learning_rate": 9.996589998499702e-06,
+      "loss": 0.2384,
+      "step": 745
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.7691083322695482,
+      "learning_rate": 9.996556652878486e-06,
+      "loss": 0.233,
+      "step": 746
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.8960295111073695,
+      "learning_rate": 9.996523145066165e-06,
+      "loss": 0.1974,
+      "step": 747
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.6525652491229723,
+      "learning_rate": 9.996489475063832e-06,
+      "loss": 0.2616,
+      "step": 748
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.6148126159936937,
+      "learning_rate": 9.996455642872578e-06,
+      "loss": 0.1963,
+      "step": 749
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.7864077623918613,
+      "learning_rate": 9.996421648493499e-06,
+      "loss": 0.1906,
+      "step": 750
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.830951504661101,
+      "learning_rate": 9.996387491927699e-06,
+      "loss": 0.251,
+      "step": 751
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.1138158384711583,
+      "learning_rate": 9.99635317317629e-06,
+      "loss": 0.1347,
+      "step": 752
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.9654294424616947,
+      "learning_rate": 9.996318692240384e-06,
+      "loss": 0.2566,
+      "step": 753
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.4760254016795331,
+      "learning_rate": 9.996284049121098e-06,
+      "loss": 0.2391,
+      "step": 754
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.5792199953262465,
+      "learning_rate": 9.99624924381956e-06,
+      "loss": 0.2341,
+      "step": 755
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 9.185073220631361,
+      "learning_rate": 9.9962142763369e-06,
+      "loss": 0.2156,
+      "step": 756
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 2.3936048959223153,
+      "learning_rate": 9.99617914667425e-06,
+      "loss": 0.263,
+      "step": 757
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.7616782892152645,
+      "learning_rate": 9.996143854832752e-06,
+      "loss": 0.2508,
+      "step": 758
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.72638969937547,
+      "learning_rate": 9.99610840081355e-06,
+      "loss": 0.2191,
+      "step": 759
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 2.205201257894953,
+      "learning_rate": 9.996072784617799e-06,
+      "loss": 0.2754,
+      "step": 760
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 2.3208722296787774,
+      "learning_rate": 9.996037006246651e-06,
+      "loss": 0.2637,
+      "step": 761
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.4094917998582825,
+      "learning_rate": 9.99600106570127e-06,
+      "loss": 0.1922,
+      "step": 762
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.4576681294561826,
+      "learning_rate": 9.99596496298282e-06,
+      "loss": 0.212,
+      "step": 763
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 4.372567508758709,
+      "learning_rate": 9.995928698092475e-06,
+      "loss": 0.1416,
+      "step": 764
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.5621205047855777,
+      "learning_rate": 9.99589227103141e-06,
+      "loss": 0.171,
+      "step": 765
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.4635162543628049,
+      "learning_rate": 9.995855681800813e-06,
+      "loss": 0.241,
+      "step": 766
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 2.308129291318679,
+      "learning_rate": 9.995818930401865e-06,
+      "loss": 0.252,
+      "step": 767
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.6108515734099502,
+      "learning_rate": 9.995782016835762e-06,
+      "loss": 0.2097,
+      "step": 768
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.5092952335991863,
+      "learning_rate": 9.995744941103702e-06,
+      "loss": 0.2199,
+      "step": 769
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.5698018121380966,
+      "learning_rate": 9.995707703206889e-06,
+      "loss": 0.2054,
+      "step": 770
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 2.0919332949536926,
+      "learning_rate": 9.99567030314653e-06,
+      "loss": 0.2476,
+      "step": 771
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.7983041221681502,
+      "learning_rate": 9.995632740923841e-06,
+      "loss": 0.2442,
+      "step": 772
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.8123833026807628,
+      "learning_rate": 9.99559501654004e-06,
+      "loss": 0.2489,
+      "step": 773
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.4794587109495483,
+      "learning_rate": 9.995557129996354e-06,
+      "loss": 0.2479,
+      "step": 774
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 2.0845419800043787,
+      "learning_rate": 9.995519081294009e-06,
+      "loss": 0.2502,
+      "step": 775
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 2.218294768577398,
+      "learning_rate": 9.995480870434241e-06,
+      "loss": 0.2272,
+      "step": 776
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.3143656873032976,
+      "learning_rate": 9.995442497418295e-06,
+      "loss": 0.2182,
+      "step": 777
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 2.029352181849417,
+      "learning_rate": 9.995403962247411e-06,
+      "loss": 0.2909,
+      "step": 778
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.0203498024423483,
+      "learning_rate": 9.995365264922841e-06,
+      "loss": 0.2256,
+      "step": 779
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 2.1115985241987407,
+      "learning_rate": 9.995326405445842e-06,
+      "loss": 0.2991,
+      "step": 780
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 2.0550399013924285,
+      "learning_rate": 9.995287383817676e-06,
+      "loss": 0.2447,
+      "step": 781
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.2567733646813406,
+      "learning_rate": 9.99524820003961e-06,
+      "loss": 0.1906,
+      "step": 782
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.5716726618425232,
+      "learning_rate": 9.995208854112915e-06,
+      "loss": 0.2231,
+      "step": 783
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.1434299163118575,
+      "learning_rate": 9.995169346038867e-06,
+      "loss": 0.2275,
+      "step": 784
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.7054133877612792,
+      "learning_rate": 9.995129675818748e-06,
+      "loss": 0.2149,
+      "step": 785
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.6301066003282731,
+      "learning_rate": 9.995089843453849e-06,
+      "loss": 0.2611,
+      "step": 786
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.9479567545745378,
+      "learning_rate": 9.995049848945462e-06,
+      "loss": 0.2367,
+      "step": 787
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.7888615018469156,
+      "learning_rate": 9.995009692294884e-06,
+      "loss": 0.2558,
+      "step": 788
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.1355459858878072,
+      "learning_rate": 9.994969373503419e-06,
+      "loss": 0.1715,
+      "step": 789
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.4047483851102234,
+      "learning_rate": 9.994928892572376e-06,
+      "loss": 0.187,
+      "step": 790
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.3005230908061722,
+      "learning_rate": 9.994888249503069e-06,
+      "loss": 0.2253,
+      "step": 791
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.3781808222820444,
+      "learning_rate": 9.994847444296816e-06,
+      "loss": 0.2364,
+      "step": 792
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.1397672458882684,
+      "learning_rate": 9.994806476954943e-06,
+      "loss": 0.1481,
+      "step": 793
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.147034264635771,
+      "learning_rate": 9.99476534747878e-06,
+      "loss": 0.2296,
+      "step": 794
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.2264614868482613,
+      "learning_rate": 9.994724055869663e-06,
+      "loss": 0.2105,
+      "step": 795
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.2980497193169003,
+      "learning_rate": 9.994682602128929e-06,
+      "loss": 0.2495,
+      "step": 796
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.2960108727647346,
+      "learning_rate": 9.994640986257927e-06,
+      "loss": 0.1762,
+      "step": 797
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.6657532801801207,
+      "learning_rate": 9.994599208258005e-06,
+      "loss": 0.2725,
+      "step": 798
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.3637181759625494,
+      "learning_rate": 9.99455726813052e-06,
+      "loss": 0.281,
+      "step": 799
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.004064282176184,
+      "learning_rate": 9.994515165876837e-06,
+      "loss": 0.1306,
+      "step": 800
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.160079748026799,
+      "learning_rate": 9.994472901498318e-06,
+      "loss": 0.2201,
+      "step": 801
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.6866050484001025,
+      "learning_rate": 9.994430474996337e-06,
+      "loss": 0.1441,
+      "step": 802
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.2357125188438258,
+      "learning_rate": 9.994387886372269e-06,
+      "loss": 0.213,
+      "step": 803
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.3157086179306139,
+      "learning_rate": 9.994345135627498e-06,
+      "loss": 0.1935,
+      "step": 804
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 2.088398720819187,
+      "learning_rate": 9.994302222763415e-06,
+      "loss": 0.2422,
+      "step": 805
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.501032900823252,
+      "learning_rate": 9.99425914778141e-06,
+      "loss": 0.2431,
+      "step": 806
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.9161484426335658,
+      "learning_rate": 9.994215910682878e-06,
+      "loss": 0.2639,
+      "step": 807
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.4251877267893638,
+      "learning_rate": 9.994172511469227e-06,
+      "loss": 0.1807,
+      "step": 808
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.585305987844208,
+      "learning_rate": 9.994128950141866e-06,
+      "loss": 0.1572,
+      "step": 809
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.568296769722774,
+      "learning_rate": 9.994085226702206e-06,
+      "loss": 0.218,
+      "step": 810
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 4.279281497790149,
+      "learning_rate": 9.994041341151667e-06,
+      "loss": 0.2671,
+      "step": 811
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.6708094329583227,
+      "learning_rate": 9.993997293491674e-06,
+      "loss": 0.2519,
+      "step": 812
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.8452941885547565,
+      "learning_rate": 9.99395308372366e-06,
+      "loss": 0.2501,
+      "step": 813
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.313985562878818,
+      "learning_rate": 9.993908711849055e-06,
+      "loss": 0.2488,
+      "step": 814
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.329512446431142,
+      "learning_rate": 9.993864177869302e-06,
+      "loss": 0.2596,
+      "step": 815
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.2622987695407983,
+      "learning_rate": 9.993819481785846e-06,
+      "loss": 0.1941,
+      "step": 816
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.629491924865182,
+      "learning_rate": 9.993774623600138e-06,
+      "loss": 0.129,
+      "step": 817
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.8711916222778915,
+      "learning_rate": 9.993729603313633e-06,
+      "loss": 0.2262,
+      "step": 818
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.445170470295729,
+      "learning_rate": 9.993684420927795e-06,
+      "loss": 0.2237,
+      "step": 819
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 2.5140740488568305,
+      "learning_rate": 9.993639076444089e-06,
+      "loss": 0.3494,
+      "step": 820
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.9510408076807002,
+      "learning_rate": 9.993593569863987e-06,
+      "loss": 0.2524,
+      "step": 821
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.8622112240714863,
+      "learning_rate": 9.993547901188966e-06,
+      "loss": 0.2383,
+      "step": 822
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.7631780539261597,
+      "learning_rate": 9.993502070420509e-06,
+      "loss": 0.2936,
+      "step": 823
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 2.1521535190488015,
+      "learning_rate": 9.993456077560101e-06,
+      "loss": 0.2377,
+      "step": 824
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.4783259714859684,
+      "learning_rate": 9.993409922609241e-06,
+      "loss": 0.215,
+      "step": 825
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 2.625916769383279,
+      "learning_rate": 9.993363605569421e-06,
+      "loss": 0.3517,
+      "step": 826
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.901423682190051,
+      "learning_rate": 9.993317126442149e-06,
+      "loss": 0.2951,
+      "step": 827
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.843750507144535,
+      "learning_rate": 9.993270485228929e-06,
+      "loss": 0.2428,
+      "step": 828
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.410532408497231,
+      "learning_rate": 9.993223681931278e-06,
+      "loss": 0.2409,
+      "step": 829
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.3757206981467418,
+      "learning_rate": 9.993176716550717e-06,
+      "loss": 0.2084,
+      "step": 830
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.3048156115741543,
+      "learning_rate": 9.993129589088769e-06,
+      "loss": 0.2426,
+      "step": 831
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.866408569674609,
+      "learning_rate": 9.993082299546961e-06,
+      "loss": 0.2548,
+      "step": 832
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.7268773095065082,
+      "learning_rate": 9.99303484792683e-06,
+      "loss": 0.217,
+      "step": 833
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.6728730899029136,
+      "learning_rate": 9.992987234229918e-06,
+      "loss": 0.211,
+      "step": 834
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.4376750015422304,
+      "learning_rate": 9.992939458457769e-06,
+      "loss": 0.2419,
+      "step": 835
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.9700654850317781,
+      "learning_rate": 9.992891520611934e-06,
+      "loss": 0.203,
+      "step": 836
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.5474833703380089,
+      "learning_rate": 9.99284342069397e-06,
+      "loss": 0.1894,
+      "step": 837
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.2430915601614154,
+      "learning_rate": 9.992795158705434e-06,
+      "loss": 0.229,
+      "step": 838
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.1066378706181776,
+      "learning_rate": 9.9927467346479e-06,
+      "loss": 0.138,
+      "step": 839
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.2442485052912813,
+      "learning_rate": 9.992698148522934e-06,
+      "loss": 0.2188,
+      "step": 840
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.503172641026925,
+      "learning_rate": 9.992649400332117e-06,
+      "loss": 0.1977,
+      "step": 841
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.3095615053840808,
+      "learning_rate": 9.992600490077027e-06,
+      "loss": 0.2181,
+      "step": 842
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.4209163953033057,
+      "learning_rate": 9.992551417759256e-06,
+      "loss": 0.2142,
+      "step": 843
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.8985850478946014,
+      "learning_rate": 9.992502183380395e-06,
+      "loss": 0.231,
+      "step": 844
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.4331381850919158,
+      "learning_rate": 9.992452786942041e-06,
+      "loss": 0.2471,
+      "step": 845
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.4496724917035175,
+      "learning_rate": 9.9924032284458e-06,
+      "loss": 0.1898,
+      "step": 846
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.0682342637360658,
+      "learning_rate": 9.992353507893279e-06,
+      "loss": 0.1589,
+      "step": 847
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.811711592041193,
+      "learning_rate": 9.992303625286092e-06,
+      "loss": 0.2533,
+      "step": 848
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.5553208671215994,
+      "learning_rate": 9.992253580625858e-06,
+      "loss": 0.2926,
+      "step": 849
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.3665783765002752,
+      "learning_rate": 9.992203373914203e-06,
+      "loss": 0.2051,
+      "step": 850
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 2.278038370450238,
+      "learning_rate": 9.992153005152754e-06,
+      "loss": 0.2622,
+      "step": 851
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 2.244775285609426,
+      "learning_rate": 9.99210247434315e-06,
+      "loss": 0.2412,
+      "step": 852
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.460435046526099,
+      "learning_rate": 9.992051781487029e-06,
+      "loss": 0.1935,
+      "step": 853
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.3709198713766988,
+      "learning_rate": 9.992000926586035e-06,
+      "loss": 0.1716,
+      "step": 854
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.486706223187014,
+      "learning_rate": 9.99194990964182e-06,
+      "loss": 0.2501,
+      "step": 855
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.5583334970814147,
+      "learning_rate": 9.991898730656043e-06,
+      "loss": 0.1993,
+      "step": 856
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.437353572436606,
+      "learning_rate": 9.99184738963036e-06,
+      "loss": 0.2311,
+      "step": 857
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.677794414329712,
+      "learning_rate": 9.991795886566443e-06,
+      "loss": 0.2295,
+      "step": 858
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.9166710040874402,
+      "learning_rate": 9.991744221465958e-06,
+      "loss": 0.2475,
+      "step": 859
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.8914227024451598,
+      "learning_rate": 9.991692394330586e-06,
+      "loss": 0.2385,
+      "step": 860
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.3891599678615987,
+      "learning_rate": 9.991640405162009e-06,
+      "loss": 0.2042,
+      "step": 861
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 4.536812150590555,
+      "learning_rate": 9.991588253961914e-06,
+      "loss": 0.2438,
+      "step": 862
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.8795405722864267,
+      "learning_rate": 9.991535940731993e-06,
+      "loss": 0.2164,
+      "step": 863
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.8151272218590928,
+      "learning_rate": 9.991483465473945e-06,
+      "loss": 0.2886,
+      "step": 864
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.9047040318216781,
+      "learning_rate": 9.991430828189476e-06,
+      "loss": 0.218,
+      "step": 865
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.5089580256101758,
+      "learning_rate": 9.99137802888029e-06,
+      "loss": 0.238,
+      "step": 866
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.5526845456171137,
+      "learning_rate": 9.991325067548103e-06,
+      "loss": 0.2379,
+      "step": 867
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.8013273763946662,
+      "learning_rate": 9.991271944194634e-06,
+      "loss": 0.2145,
+      "step": 868
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.8404494582374002,
+      "learning_rate": 9.991218658821609e-06,
+      "loss": 0.1939,
+      "step": 869
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 2.08758554825834,
+      "learning_rate": 9.991165211430753e-06,
+      "loss": 0.2482,
+      "step": 870
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 3.7042416204901096,
+      "learning_rate": 9.991111602023807e-06,
+      "loss": 0.2697,
+      "step": 871
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.766875578430197,
+      "learning_rate": 9.991057830602508e-06,
+      "loss": 0.2023,
+      "step": 872
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 2.1795710221482967,
+      "learning_rate": 9.9910038971686e-06,
+      "loss": 0.2272,
+      "step": 873
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.6079584572606918,
+      "learning_rate": 9.990949801723837e-06,
+      "loss": 0.1936,
+      "step": 874
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.9629808232771345,
+      "learning_rate": 9.990895544269973e-06,
+      "loss": 0.214,
+      "step": 875
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.9922395499161029,
+      "learning_rate": 9.99084112480877e-06,
+      "loss": 0.1953,
+      "step": 876
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.7868187756680873,
+      "learning_rate": 9.990786543341994e-06,
+      "loss": 0.268,
+      "step": 877
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.4733962873437458,
+      "learning_rate": 9.990731799871415e-06,
+      "loss": 0.185,
+      "step": 878
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.1905698216535727,
+      "learning_rate": 9.990676894398815e-06,
+      "loss": 0.1686,
+      "step": 879
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.429458474747954,
+      "learning_rate": 9.990621826925972e-06,
+      "loss": 0.1919,
+      "step": 880
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 2.214006551656187,
+      "learning_rate": 9.990566597454673e-06,
+      "loss": 0.2069,
+      "step": 881
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.9883021684174255,
+      "learning_rate": 9.990511205986716e-06,
+      "loss": 0.1796,
+      "step": 882
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.6791241432970971,
+      "learning_rate": 9.990455652523893e-06,
+      "loss": 0.3108,
+      "step": 883
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.6095545191777498,
+      "learning_rate": 9.99039993706801e-06,
+      "loss": 0.2519,
+      "step": 884
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.3335942051042562,
+      "learning_rate": 9.990344059620876e-06,
+      "loss": 0.2722,
+      "step": 885
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.1723544475154293,
+      "learning_rate": 9.990288020184303e-06,
+      "loss": 0.2113,
+      "step": 886
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.5290073057293918,
+      "learning_rate": 9.990231818760112e-06,
+      "loss": 0.2109,
+      "step": 887
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 3.4602109607153175,
+      "learning_rate": 9.990175455350126e-06,
+      "loss": 0.2651,
+      "step": 888
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.1566241374285202,
+      "learning_rate": 9.990118929956175e-06,
+      "loss": 0.2727,
+      "step": 889
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 2.052044855199194,
+      "learning_rate": 9.990062242580096e-06,
+      "loss": 0.2241,
+      "step": 890
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.4177814076146151,
+      "learning_rate": 9.990005393223726e-06,
+      "loss": 0.2151,
+      "step": 891
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.7672564303364384,
+      "learning_rate": 9.989948381888913e-06,
+      "loss": 0.2271,
+      "step": 892
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.3500005649194595,
+      "learning_rate": 9.989891208577502e-06,
+      "loss": 0.1935,
+      "step": 893
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.641037130816496,
+      "learning_rate": 9.989833873291357e-06,
+      "loss": 0.2766,
+      "step": 894
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.7856500556157282,
+      "learning_rate": 9.989776376032333e-06,
+      "loss": 0.2342,
+      "step": 895
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.2303187324263358,
+      "learning_rate": 9.989718716802299e-06,
+      "loss": 0.2207,
+      "step": 896
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.407066620705677,
+      "learning_rate": 9.989660895603127e-06,
+      "loss": 0.2123,
+      "step": 897
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.2960471216727876,
+      "learning_rate": 9.989602912436692e-06,
+      "loss": 0.2201,
+      "step": 898
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.8542379372718563,
+      "learning_rate": 9.989544767304877e-06,
+      "loss": 0.2902,
+      "step": 899
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.4525291697975822,
+      "learning_rate": 9.98948646020957e-06,
+      "loss": 0.2387,
+      "step": 900
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.107567794376534,
+      "learning_rate": 9.989427991152664e-06,
+      "loss": 0.1903,
+      "step": 901
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.7098400872801431,
+      "learning_rate": 9.989369360136056e-06,
+      "loss": 0.249,
+      "step": 902
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.5465166478789543,
+      "learning_rate": 9.989310567161649e-06,
+      "loss": 0.1837,
+      "step": 903
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.161074838241732,
+      "learning_rate": 9.989251612231352e-06,
+      "loss": 0.2363,
+      "step": 904
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.4277623196018425,
+      "learning_rate": 9.989192495347079e-06,
+      "loss": 0.2542,
+      "step": 905
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.6174152462686229,
+      "learning_rate": 9.98913321651075e-06,
+      "loss": 0.2556,
+      "step": 906
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.9159902946205454,
+      "learning_rate": 9.989073775724284e-06,
+      "loss": 0.2401,
+      "step": 907
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.3589066158898169,
+      "learning_rate": 9.989014172989617e-06,
+      "loss": 0.2285,
+      "step": 908
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.4265811244170847,
+      "learning_rate": 9.988954408308681e-06,
+      "loss": 0.2512,
+      "step": 909
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.402832397051935,
+      "learning_rate": 9.988894481683416e-06,
+      "loss": 0.2234,
+      "step": 910
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 2.708595913604761,
+      "learning_rate": 9.988834393115768e-06,
+      "loss": 0.2364,
+      "step": 911
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.4567256011175753,
+      "learning_rate": 9.988774142607684e-06,
+      "loss": 0.2889,
+      "step": 912
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.5397321144795744,
+      "learning_rate": 9.988713730161125e-06,
+      "loss": 0.2027,
+      "step": 913
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 2.225814318854797,
+      "learning_rate": 9.988653155778049e-06,
+      "loss": 0.2534,
+      "step": 914
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.5422927510367952,
+      "learning_rate": 9.988592419460423e-06,
+      "loss": 0.2823,
+      "step": 915
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.391193557624271,
+      "learning_rate": 9.988531521210219e-06,
+      "loss": 0.2625,
+      "step": 916
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.3037852405419272,
+      "learning_rate": 9.988470461029412e-06,
+      "loss": 0.1961,
+      "step": 917
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 2.726029439589459,
+      "learning_rate": 9.988409238919985e-06,
+      "loss": 0.2418,
+      "step": 918
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.1114674030042353,
+      "learning_rate": 9.988347854883927e-06,
+      "loss": 0.2316,
+      "step": 919
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.208037956558638,
+      "learning_rate": 9.988286308923228e-06,
+      "loss": 0.196,
+      "step": 920
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.4125976364710648,
+      "learning_rate": 9.988224601039888e-06,
+      "loss": 0.2762,
+      "step": 921
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.596446395302791,
+      "learning_rate": 9.988162731235908e-06,
+      "loss": 0.2864,
+      "step": 922
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.5402656538455302,
+      "learning_rate": 9.988100699513298e-06,
+      "loss": 0.1571,
+      "step": 923
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.9300669652218219,
+      "learning_rate": 9.988038505874072e-06,
+      "loss": 0.2278,
+      "step": 924
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.3783263102699999,
+      "learning_rate": 9.987976150320246e-06,
+      "loss": 0.2505,
+      "step": 925
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.7210174607924251,
+      "learning_rate": 9.987913632853845e-06,
+      "loss": 0.2132,
+      "step": 926
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 2.088656656963556,
+      "learning_rate": 9.9878509534769e-06,
+      "loss": 0.2803,
+      "step": 927
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.2339367873340437,
+      "learning_rate": 9.987788112191447e-06,
+      "loss": 0.2691,
+      "step": 928
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.303307064234305,
+      "learning_rate": 9.987725108999523e-06,
+      "loss": 0.214,
+      "step": 929
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.5709334137535031,
+      "learning_rate": 9.987661943903172e-06,
+      "loss": 0.1954,
+      "step": 930
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.5057379330995977,
+      "learning_rate": 9.987598616904446e-06,
+      "loss": 0.1227,
+      "step": 931
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.7730160245965954,
+      "learning_rate": 9.987535128005404e-06,
+      "loss": 0.1864,
+      "step": 932
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.3758593388176685,
+      "learning_rate": 9.987471477208101e-06,
+      "loss": 0.2178,
+      "step": 933
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.111303493448027,
+      "learning_rate": 9.987407664514606e-06,
+      "loss": 0.177,
+      "step": 934
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.4100832299429573,
+      "learning_rate": 9.98734368992699e-06,
+      "loss": 0.2389,
+      "step": 935
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 3.491723219626732,
+      "learning_rate": 9.98727955344733e-06,
+      "loss": 0.2795,
+      "step": 936
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.788121406957569,
+      "learning_rate": 9.987215255077708e-06,
+      "loss": 0.2841,
+      "step": 937
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.8608540227093385,
+      "learning_rate": 9.987150794820213e-06,
+      "loss": 0.1929,
+      "step": 938
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.3179591934929455,
+      "learning_rate": 9.987086172676934e-06,
+      "loss": 0.2007,
+      "step": 939
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 2.0060940524422732,
+      "learning_rate": 9.98702138864997e-06,
+      "loss": 0.2028,
+      "step": 940
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.544763192805912,
+      "learning_rate": 9.986956442741424e-06,
+      "loss": 0.2767,
+      "step": 941
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.579922007049674,
+      "learning_rate": 9.986891334953404e-06,
+      "loss": 0.2125,
+      "step": 942
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.5379047044822187,
+      "learning_rate": 9.986826065288024e-06,
+      "loss": 0.2404,
+      "step": 943
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.2091100231504772,
+      "learning_rate": 9.986760633747401e-06,
+      "loss": 0.1808,
+      "step": 944
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.3489025255083558,
+      "learning_rate": 9.986695040333661e-06,
+      "loss": 0.2095,
+      "step": 945
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.7426656229455786,
+      "learning_rate": 9.986629285048933e-06,
+      "loss": 0.1903,
+      "step": 946
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.7732170635972555,
+      "learning_rate": 9.98656336789535e-06,
+      "loss": 0.2418,
+      "step": 947
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 2.1766471535171377,
+      "learning_rate": 9.986497288875054e-06,
+      "loss": 0.2252,
+      "step": 948
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.9042980174525819,
+      "learning_rate": 9.986431047990188e-06,
+      "loss": 0.1937,
+      "step": 949
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.0773368421669207,
+      "learning_rate": 9.9863646452429e-06,
+      "loss": 0.2179,
+      "step": 950
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.2613252566491926,
+      "learning_rate": 9.986298080635352e-06,
+      "loss": 0.1558,
+      "step": 951
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 2.777151409630218,
+      "learning_rate": 9.986231354169699e-06,
+      "loss": 0.2041,
+      "step": 952
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.3013455710461423,
+      "learning_rate": 9.98616446584811e-06,
+      "loss": 0.2094,
+      "step": 953
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.3425800690568885,
+      "learning_rate": 9.986097415672755e-06,
+      "loss": 0.2093,
+      "step": 954
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.1973478030803608,
+      "learning_rate": 9.986030203645812e-06,
+      "loss": 0.2496,
+      "step": 955
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 2.7112634116615357,
+      "learning_rate": 9.985962829769457e-06,
+      "loss": 0.2504,
+      "step": 956
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.724628998654484,
+      "learning_rate": 9.985895294045885e-06,
+      "loss": 0.2121,
+      "step": 957
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.5906423980908053,
+      "learning_rate": 9.985827596477285e-06,
+      "loss": 0.1802,
+      "step": 958
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.4000098270684438,
+      "learning_rate": 9.985759737065853e-06,
+      "loss": 0.2129,
+      "step": 959
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.5633062142395335,
+      "learning_rate": 9.985691715813792e-06,
+      "loss": 0.2261,
+      "step": 960
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.2664589400864432,
+      "learning_rate": 9.985623532723313e-06,
+      "loss": 0.1809,
+      "step": 961
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 8.009300785859471,
+      "learning_rate": 9.985555187796625e-06,
+      "loss": 0.2434,
+      "step": 962
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.4531394114343694,
+      "learning_rate": 9.98548668103595e-06,
+      "loss": 0.2355,
+      "step": 963
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.3856378506609577,
+      "learning_rate": 9.98541801244351e-06,
+      "loss": 0.1993,
+      "step": 964
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.39277996831488,
+      "learning_rate": 9.985349182021536e-06,
+      "loss": 0.2082,
+      "step": 965
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 3.105960322406658,
+      "learning_rate": 9.98528018977226e-06,
+      "loss": 0.2322,
+      "step": 966
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.3446049957343713,
+      "learning_rate": 9.985211035697921e-06,
+      "loss": 0.2616,
+      "step": 967
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.4718135502008964,
+      "learning_rate": 9.985141719800766e-06,
+      "loss": 0.2464,
+      "step": 968
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.1873764174553603,
+      "learning_rate": 9.985072242083045e-06,
+      "loss": 0.2077,
+      "step": 969
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.599956986451696,
+      "learning_rate": 9.98500260254701e-06,
+      "loss": 0.2659,
+      "step": 970
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.1697543339498704,
+      "learning_rate": 9.984932801194927e-06,
+      "loss": 0.209,
+      "step": 971
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.238303089550941,
+      "learning_rate": 9.984862838029057e-06,
+      "loss": 0.2155,
+      "step": 972
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.5609784335228798,
+      "learning_rate": 9.984792713051672e-06,
+      "loss": 0.1656,
+      "step": 973
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 2.156906591182393,
+      "learning_rate": 9.98472242626505e-06,
+      "loss": 0.2568,
+      "step": 974
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.0959353192435246,
+      "learning_rate": 9.984651977671471e-06,
+      "loss": 0.1709,
+      "step": 975
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.1125309925771782,
+      "learning_rate": 9.984581367273222e-06,
+      "loss": 0.2534,
+      "step": 976
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.5918457210979915,
+      "learning_rate": 9.984510595072599e-06,
+      "loss": 0.2306,
+      "step": 977
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.490419401329037,
+      "learning_rate": 9.984439661071892e-06,
+      "loss": 0.1957,
+      "step": 978
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.4420343080042421,
+      "learning_rate": 9.984368565273409e-06,
+      "loss": 0.2393,
+      "step": 979
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.6885920231947016,
+      "learning_rate": 9.984297307679455e-06,
+      "loss": 0.2326,
+      "step": 980
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.334491033110002,
+      "learning_rate": 9.984225888292346e-06,
+      "loss": 0.2536,
+      "step": 981
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.1817211502002398,
+      "learning_rate": 9.984154307114398e-06,
+      "loss": 0.2059,
+      "step": 982
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.7102879266595847,
+      "learning_rate": 9.984082564147935e-06,
+      "loss": 0.252,
+      "step": 983
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.2358043820807088,
+      "learning_rate": 9.984010659395285e-06,
+      "loss": 0.2067,
+      "step": 984
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.1077678537754339,
+      "learning_rate": 9.983938592858783e-06,
+      "loss": 0.2,
+      "step": 985
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.3227772532867288,
+      "learning_rate": 9.98386636454077e-06,
+      "loss": 0.1774,
+      "step": 986
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.675970791499035,
+      "learning_rate": 9.983793974443587e-06,
+      "loss": 0.2707,
+      "step": 987
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.6790284471787886,
+      "learning_rate": 9.983721422569585e-06,
+      "loss": 0.2455,
+      "step": 988
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.8196197918993409,
+      "learning_rate": 9.983648708921123e-06,
+      "loss": 0.2302,
+      "step": 989
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.3500481649919593,
+      "learning_rate": 9.983575833500556e-06,
+      "loss": 0.2694,
+      "step": 990
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.3586135847454621,
+      "learning_rate": 9.983502796310252e-06,
+      "loss": 0.2187,
+      "step": 991
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.5726381843768615,
+      "learning_rate": 9.98342959735258e-06,
+      "loss": 0.2718,
+      "step": 992
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.4075071649385964,
+      "learning_rate": 9.983356236629919e-06,
+      "loss": 0.2217,
+      "step": 993
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.963694407196175,
+      "learning_rate": 9.98328271414465e-06,
+      "loss": 0.2068,
+      "step": 994
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.3379321725338948,
+      "learning_rate": 9.983209029899158e-06,
+      "loss": 0.1902,
+      "step": 995
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.4019286303693481,
+      "learning_rate": 9.983135183895835e-06,
+      "loss": 0.2352,
+      "step": 996
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.3446962869819388,
+      "learning_rate": 9.983061176137079e-06,
+      "loss": 0.2242,
+      "step": 997
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.6288738156023739,
+      "learning_rate": 9.98298700662529e-06,
+      "loss": 0.2494,
+      "step": 998
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 3.631970379754838,
+      "learning_rate": 9.98291267536288e-06,
+      "loss": 0.2025,
+      "step": 999
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.4028370767855172,
+      "learning_rate": 9.982838182352257e-06,
+      "loss": 0.2134,
+      "step": 1000
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.2612395365614641,
+      "learning_rate": 9.982763527595841e-06,
+      "loss": 0.2125,
+      "step": 1001
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.1018481357705399,
+      "learning_rate": 9.982688711096056e-06,
+      "loss": 0.1776,
+      "step": 1002
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.6733297342707991,
+      "learning_rate": 9.98261373285533e-06,
+      "loss": 0.1784,
+      "step": 1003
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.483247157827921,
+      "learning_rate": 9.982538592876099e-06,
+      "loss": 0.2525,
+      "step": 1004
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.3897791926297314,
+      "learning_rate": 9.9824632911608e-06,
+      "loss": 0.2314,
+      "step": 1005
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.4408984271811753,
+      "learning_rate": 9.982387827711874e-06,
+      "loss": 0.22,
+      "step": 1006
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 2.008263822957917,
+      "learning_rate": 9.982312202531777e-06,
+      "loss": 0.2636,
+      "step": 1007
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.1482339016803096,
+      "learning_rate": 9.98223641562296e-06,
+      "loss": 0.1518,
+      "step": 1008
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.1761861481017488,
+      "learning_rate": 9.982160466987885e-06,
+      "loss": 0.1561,
+      "step": 1009
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.0828725611693262,
+      "learning_rate": 9.982084356629014e-06,
+      "loss": 0.1837,
+      "step": 1010
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.1709146832571171,
+      "learning_rate": 9.982008084548822e-06,
+      "loss": 0.1954,
+      "step": 1011
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.3596966845167775,
+      "learning_rate": 9.981931650749782e-06,
+      "loss": 0.2841,
+      "step": 1012
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.5602116696233383,
+      "learning_rate": 9.981855055234375e-06,
+      "loss": 0.2601,
+      "step": 1013
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.3779024463450789,
+      "learning_rate": 9.98177829800509e-06,
+      "loss": 0.2278,
+      "step": 1014
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.2277743357612954,
+      "learning_rate": 9.981701379064414e-06,
+      "loss": 0.2206,
+      "step": 1015
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.4654456669422902,
+      "learning_rate": 9.981624298414849e-06,
+      "loss": 0.2239,
+      "step": 1016
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.7127379134034206,
+      "learning_rate": 9.981547056058893e-06,
+      "loss": 0.2437,
+      "step": 1017
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.525082967457466,
+      "learning_rate": 9.981469651999057e-06,
+      "loss": 0.2055,
+      "step": 1018
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.5774340970246645,
+      "learning_rate": 9.981392086237851e-06,
+      "loss": 0.2501,
+      "step": 1019
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.5134016767960998,
+      "learning_rate": 9.981314358777792e-06,
+      "loss": 0.2124,
+      "step": 1020
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.7651890283961635,
+      "learning_rate": 9.981236469621404e-06,
+      "loss": 0.2555,
+      "step": 1021
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.2555344073652324,
+      "learning_rate": 9.981158418771218e-06,
+      "loss": 0.225,
+      "step": 1022
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.5026501382580701,
+      "learning_rate": 9.981080206229765e-06,
+      "loss": 0.2321,
+      "step": 1023
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.4043536512970234,
+      "learning_rate": 9.981001831999582e-06,
+      "loss": 0.2214,
+      "step": 1024
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.258302208244986,
+      "learning_rate": 9.980923296083219e-06,
+      "loss": 0.194,
+      "step": 1025
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.7705610196786978,
+      "learning_rate": 9.980844598483217e-06,
+      "loss": 0.2515,
+      "step": 1026
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.598332739580166,
+      "learning_rate": 9.980765739202138e-06,
+      "loss": 0.2171,
+      "step": 1027
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.0787654715806567,
+      "learning_rate": 9.980686718242539e-06,
+      "loss": 0.2462,
+      "step": 1028
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.5956307234850093,
+      "learning_rate": 9.980607535606984e-06,
+      "loss": 0.2063,
+      "step": 1029
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.284701712646174,
+      "learning_rate": 9.980528191298045e-06,
+      "loss": 0.1744,
+      "step": 1030
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.3371605398601,
+      "learning_rate": 9.980448685318296e-06,
+      "loss": 0.2014,
+      "step": 1031
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.4221399343386596,
+      "learning_rate": 9.980369017670319e-06,
+      "loss": 0.2174,
+      "step": 1032
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.1243029221271443,
+      "learning_rate": 9.980289188356702e-06,
+      "loss": 0.2689,
+      "step": 1033
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.2531696648735722,
+      "learning_rate": 9.980209197380031e-06,
+      "loss": 0.2014,
+      "step": 1034
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.1166276512681408,
+      "learning_rate": 9.980129044742908e-06,
+      "loss": 0.191,
+      "step": 1035
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.5782207516619255,
+      "learning_rate": 9.980048730447931e-06,
+      "loss": 0.2333,
+      "step": 1036
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.356467923112551,
+      "learning_rate": 9.979968254497709e-06,
+      "loss": 0.2496,
+      "step": 1037
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.1586807080083221,
+      "learning_rate": 9.979887616894853e-06,
+      "loss": 0.1681,
+      "step": 1038
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.5044882824691215,
+      "learning_rate": 9.979806817641982e-06,
+      "loss": 0.2715,
+      "step": 1039
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.608116558321884,
+      "learning_rate": 9.979725856741717e-06,
+      "loss": 0.1615,
+      "step": 1040
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.7380444183174493,
+      "learning_rate": 9.979644734196689e-06,
+      "loss": 0.2722,
+      "step": 1041
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.212400279694366,
+      "learning_rate": 9.979563450009529e-06,
+      "loss": 0.1864,
+      "step": 1042
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.3972279629685254,
+      "learning_rate": 9.979482004182875e-06,
+      "loss": 0.2434,
+      "step": 1043
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.9651048071585169,
+      "learning_rate": 9.979400396719371e-06,
+      "loss": 0.2157,
+      "step": 1044
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.3254184858259093,
+      "learning_rate": 9.97931862762167e-06,
+      "loss": 0.179,
+      "step": 1045
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.2126011487651382,
+      "learning_rate": 9.97923669689242e-06,
+      "loss": 0.1982,
+      "step": 1046
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.0638866649684466,
+      "learning_rate": 9.979154604534284e-06,
+      "loss": 0.2111,
+      "step": 1047
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.2915385611198724,
+      "learning_rate": 9.979072350549927e-06,
+      "loss": 0.2082,
+      "step": 1048
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.460339173788245,
+      "learning_rate": 9.978989934942018e-06,
+      "loss": 0.1828,
+      "step": 1049
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.5152332731234583,
+      "learning_rate": 9.978907357713233e-06,
+      "loss": 0.2307,
+      "step": 1050
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.3680527578491752,
+      "learning_rate": 9.978824618866252e-06,
+      "loss": 0.2365,
+      "step": 1051
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.6014565874760736,
+      "learning_rate": 9.978741718403761e-06,
+      "loss": 0.2028,
+      "step": 1052
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.2004489180676827,
+      "learning_rate": 9.978658656328451e-06,
+      "loss": 0.197,
+      "step": 1053
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.5499932908097547,
+      "learning_rate": 9.978575432643018e-06,
+      "loss": 0.2085,
+      "step": 1054
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.5193147676706682,
+      "learning_rate": 9.978492047350164e-06,
+      "loss": 0.2529,
+      "step": 1055
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.2343401330538368,
+      "learning_rate": 9.978408500452593e-06,
+      "loss": 0.1798,
+      "step": 1056
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.17362880442648,
+      "learning_rate": 9.97832479195302e-06,
+      "loss": 0.2344,
+      "step": 1057
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.8033892577551598,
+      "learning_rate": 9.978240921854163e-06,
+      "loss": 0.2735,
+      "step": 1058
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.1623977046996727,
+      "learning_rate": 9.978156890158744e-06,
+      "loss": 0.1886,
+      "step": 1059
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.2457845817787145,
+      "learning_rate": 9.978072696869488e-06,
+      "loss": 0.194,
+      "step": 1060
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 2.983520864918462,
+      "learning_rate": 9.97798834198913e-06,
+      "loss": 0.2074,
+      "step": 1061
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.4231519519463496,
+      "learning_rate": 9.977903825520406e-06,
+      "loss": 0.2406,
+      "step": 1062
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.1931269086809513,
+      "learning_rate": 9.977819147466064e-06,
+      "loss": 0.2525,
+      "step": 1063
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.7768871779415316,
+      "learning_rate": 9.977734307828848e-06,
+      "loss": 0.1975,
+      "step": 1064
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.5985628606770355,
+      "learning_rate": 9.977649306611514e-06,
+      "loss": 0.2274,
+      "step": 1065
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.4786693815560086,
+      "learning_rate": 9.977564143816822e-06,
+      "loss": 0.2568,
+      "step": 1066
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.3039492671446387,
+      "learning_rate": 9.977478819447538e-06,
+      "loss": 0.2579,
+      "step": 1067
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.4226615532781977,
+      "learning_rate": 9.977393333506424e-06,
+      "loss": 0.2085,
+      "step": 1068
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.0142143227699938,
+      "learning_rate": 9.977307685996264e-06,
+      "loss": 0.1719,
+      "step": 1069
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.5551789401993938,
+      "learning_rate": 9.977221876919833e-06,
+      "loss": 0.2348,
+      "step": 1070
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.274074854230643,
+      "learning_rate": 9.977135906279919e-06,
+      "loss": 0.2379,
+      "step": 1071
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.1426826934566265,
+      "learning_rate": 9.977049774079311e-06,
+      "loss": 0.2135,
+      "step": 1072
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.3485706157554151,
+      "learning_rate": 9.976963480320805e-06,
+      "loss": 0.1698,
+      "step": 1073
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.4279162842887603,
+      "learning_rate": 9.9768770250072e-06,
+      "loss": 0.2187,
+      "step": 1074
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.5395495883248618,
+      "learning_rate": 9.97679040814131e-06,
+      "loss": 0.2046,
+      "step": 1075
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.375191207888285,
+      "learning_rate": 9.976703629725939e-06,
+      "loss": 0.1631,
+      "step": 1076
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.4872179788823656,
+      "learning_rate": 9.976616689763905e-06,
+      "loss": 0.26,
+      "step": 1077
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 2.0683896257963816,
+      "learning_rate": 9.976529588258032e-06,
+      "loss": 0.2537,
+      "step": 1078
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.725141898828764,
+      "learning_rate": 9.97644232521115e-06,
+      "loss": 0.2319,
+      "step": 1079
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.5301394498259802,
+      "learning_rate": 9.976354900626085e-06,
+      "loss": 0.2308,
+      "step": 1080
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.5084850091330002,
+      "learning_rate": 9.976267314505679e-06,
+      "loss": 0.2506,
+      "step": 1081
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.3275396291202055,
+      "learning_rate": 9.976179566852773e-06,
+      "loss": 0.2363,
+      "step": 1082
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.620963095949561,
+      "learning_rate": 9.976091657670219e-06,
+      "loss": 0.2497,
+      "step": 1083
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.5403741775519166,
+      "learning_rate": 9.976003586960867e-06,
+      "loss": 0.2511,
+      "step": 1084
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.302273730344846,
+      "learning_rate": 9.97591535472758e-06,
+      "loss": 0.2463,
+      "step": 1085
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.9751420763086212,
+      "learning_rate": 9.975826960973217e-06,
+      "loss": 0.2089,
+      "step": 1086
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.255645155248849,
+      "learning_rate": 9.975738405700649e-06,
+      "loss": 0.2556,
+      "step": 1087
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.4553224038525743,
+      "learning_rate": 9.975649688912752e-06,
+      "loss": 0.2143,
+      "step": 1088
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.3479339285894436,
+      "learning_rate": 9.975560810612405e-06,
+      "loss": 0.2481,
+      "step": 1089
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.136295944576476,
+      "learning_rate": 9.975471770802491e-06,
+      "loss": 0.1905,
+      "step": 1090
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.0786396643337717,
+      "learning_rate": 9.975382569485906e-06,
+      "loss": 0.2233,
+      "step": 1091
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.2482098486430877,
+      "learning_rate": 9.975293206665538e-06,
+      "loss": 0.2392,
+      "step": 1092
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.4142840716343987,
+      "learning_rate": 9.975203682344295e-06,
+      "loss": 0.1945,
+      "step": 1093
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.687878474646679,
+      "learning_rate": 9.975113996525077e-06,
+      "loss": 0.2276,
+      "step": 1094
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.569324551525009,
+      "learning_rate": 9.975024149210797e-06,
+      "loss": 0.2575,
+      "step": 1095
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.0269006001798315,
+      "learning_rate": 9.974934140404374e-06,
+      "loss": 0.2567,
+      "step": 1096
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.7396789696287105,
+      "learning_rate": 9.974843970108728e-06,
+      "loss": 0.2455,
+      "step": 1097
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.291552293227994,
+      "learning_rate": 9.974753638326787e-06,
+      "loss": 0.2183,
+      "step": 1098
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.0802537928719138,
+      "learning_rate": 9.97466314506148e-06,
+      "loss": 0.2078,
+      "step": 1099
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.5885525012558797,
+      "learning_rate": 9.974572490315748e-06,
+      "loss": 0.2491,
+      "step": 1100
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.5450829959626144,
+      "learning_rate": 9.974481674092532e-06,
+      "loss": 0.2204,
+      "step": 1101
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 3.1398201001167174,
+      "learning_rate": 9.97439069639478e-06,
+      "loss": 0.2879,
+      "step": 1102
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.7635262303503902,
+      "learning_rate": 9.974299557225445e-06,
+      "loss": 0.2706,
+      "step": 1103
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.9778437163526165,
+      "learning_rate": 9.974208256587488e-06,
+      "loss": 0.1851,
+      "step": 1104
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.9963944168339308,
+      "learning_rate": 9.974116794483867e-06,
+      "loss": 0.217,
+      "step": 1105
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.5551605271769557,
+      "learning_rate": 9.974025170917559e-06,
+      "loss": 0.2887,
+      "step": 1106
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.6296051919191152,
+      "learning_rate": 9.97393338589153e-06,
+      "loss": 0.2193,
+      "step": 1107
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.6648401538697264,
+      "learning_rate": 9.973841439408765e-06,
+      "loss": 0.2625,
+      "step": 1108
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.8883196625334552,
+      "learning_rate": 9.973749331472247e-06,
+      "loss": 0.1687,
+      "step": 1109
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.4129830746756131,
+      "learning_rate": 9.973657062084965e-06,
+      "loss": 0.2089,
+      "step": 1110
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.5961235501069702,
+      "learning_rate": 9.973564631249915e-06,
+      "loss": 0.2486,
+      "step": 1111
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.2755840966779854,
+      "learning_rate": 9.973472038970096e-06,
+      "loss": 0.2187,
+      "step": 1112
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.5905142431726182,
+      "learning_rate": 9.973379285248515e-06,
+      "loss": 0.2721,
+      "step": 1113
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.6272583568692174,
+      "learning_rate": 9.973286370088183e-06,
+      "loss": 0.2757,
+      "step": 1114
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.4034509379011764,
+      "learning_rate": 9.973193293492116e-06,
+      "loss": 0.2348,
+      "step": 1115
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.2640838423201548,
+      "learning_rate": 9.973100055463334e-06,
+      "loss": 0.2091,
+      "step": 1116
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.733951557722844,
+      "learning_rate": 9.973006656004865e-06,
+      "loss": 0.2693,
+      "step": 1117
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.8540303557676574,
+      "learning_rate": 9.972913095119741e-06,
+      "loss": 0.2122,
+      "step": 1118
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.162211813539026,
+      "learning_rate": 9.972819372810995e-06,
+      "loss": 0.2669,
+      "step": 1119
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.1191814017398378,
+      "learning_rate": 9.972725489081676e-06,
+      "loss": 0.1838,
+      "step": 1120
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.7908737975096738,
+      "learning_rate": 9.972631443934828e-06,
+      "loss": 0.2648,
+      "step": 1121
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.8151183084477929,
+      "learning_rate": 9.972537237373504e-06,
+      "loss": 0.2286,
+      "step": 1122
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.9877329292851007,
+      "learning_rate": 9.97244286940076e-06,
+      "loss": 0.2467,
+      "step": 1123
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.0101496678187325,
+      "learning_rate": 9.972348340019661e-06,
+      "loss": 0.2285,
+      "step": 1124
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.794749516294568,
+      "learning_rate": 9.972253649233277e-06,
+      "loss": 0.2413,
+      "step": 1125
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.3388058114817674,
+      "learning_rate": 9.97215879704468e-06,
+      "loss": 0.2615,
+      "step": 1126
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.541568252095337,
+      "learning_rate": 9.972063783456949e-06,
+      "loss": 0.2176,
+      "step": 1127
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.092783968775016,
+      "learning_rate": 9.971968608473168e-06,
+      "loss": 0.2207,
+      "step": 1128
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.1739509457965303,
+      "learning_rate": 9.971873272096427e-06,
+      "loss": 0.2189,
+      "step": 1129
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.9050224198477936,
+      "learning_rate": 9.971777774329821e-06,
+      "loss": 0.215,
+      "step": 1130
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.5660719383964075,
+      "learning_rate": 9.971682115176449e-06,
+      "loss": 0.2319,
+      "step": 1131
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.6148626183384251,
+      "learning_rate": 9.971586294639418e-06,
+      "loss": 0.2024,
+      "step": 1132
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.2279040113519024,
+      "learning_rate": 9.971490312721837e-06,
+      "loss": 0.238,
+      "step": 1133
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 2.326344775893484,
+      "learning_rate": 9.97139416942682e-06,
+      "loss": 0.1964,
+      "step": 1134
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.5633242495398034,
+      "learning_rate": 9.97129786475749e-06,
+      "loss": 0.2694,
+      "step": 1135
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.9358990384484664,
+      "learning_rate": 9.971201398716972e-06,
+      "loss": 0.1894,
+      "step": 1136
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.9689453460641065,
+      "learning_rate": 9.9711047713084e-06,
+      "loss": 0.1829,
+      "step": 1137
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.1911443578824656,
+      "learning_rate": 9.971007982534906e-06,
+      "loss": 0.2163,
+      "step": 1138
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.1511987258774994,
+      "learning_rate": 9.970911032399637e-06,
+      "loss": 0.212,
+      "step": 1139
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.6614171630791912,
+      "learning_rate": 9.970813920905735e-06,
+      "loss": 0.2954,
+      "step": 1140
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.3018148685895115,
+      "learning_rate": 9.970716648056356e-06,
+      "loss": 0.2078,
+      "step": 1141
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.9700106716490696,
+      "learning_rate": 9.970619213854655e-06,
+      "loss": 0.245,
+      "step": 1142
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.3256875145223679,
+      "learning_rate": 9.970521618303798e-06,
+      "loss": 0.2101,
+      "step": 1143
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.4619544755628442,
+      "learning_rate": 9.97042386140695e-06,
+      "loss": 0.2301,
+      "step": 1144
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.5978720068687378,
+      "learning_rate": 9.970325943167285e-06,
+      "loss": 0.2588,
+      "step": 1145
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.583913140629568,
+      "learning_rate": 9.970227863587983e-06,
+      "loss": 0.228,
+      "step": 1146
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.432970350612137,
+      "learning_rate": 9.970129622672225e-06,
+      "loss": 0.2215,
+      "step": 1147
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.6365386231024384,
+      "learning_rate": 9.970031220423202e-06,
+      "loss": 0.2379,
+      "step": 1148
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.1763369587233856,
+      "learning_rate": 9.969932656844106e-06,
+      "loss": 0.2186,
+      "step": 1149
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.6031551333387646,
+      "learning_rate": 9.96983393193814e-06,
+      "loss": 0.2084,
+      "step": 1150
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.44545184829755,
+      "learning_rate": 9.969735045708506e-06,
+      "loss": 0.1937,
+      "step": 1151
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.6201716063812184,
+      "learning_rate": 9.969635998158415e-06,
+      "loss": 0.2682,
+      "step": 1152
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.296238800633165,
+      "learning_rate": 9.969536789291082e-06,
+      "loss": 0.2258,
+      "step": 1153
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.120666938541898,
+      "learning_rate": 9.969437419109726e-06,
+      "loss": 0.2177,
+      "step": 1154
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.8720273225372333,
+      "learning_rate": 9.969337887617575e-06,
+      "loss": 0.218,
+      "step": 1155
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.2822775979500782,
+      "learning_rate": 9.969238194817859e-06,
+      "loss": 0.2424,
+      "step": 1156
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.6663800407848786,
+      "learning_rate": 9.969138340713812e-06,
+      "loss": 0.2108,
+      "step": 1157
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.5279340554513465,
+      "learning_rate": 9.969038325308678e-06,
+      "loss": 0.2424,
+      "step": 1158
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.3855930554588178,
+      "learning_rate": 9.968938148605703e-06,
+      "loss": 0.2061,
+      "step": 1159
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 2.1246209823367077,
+      "learning_rate": 9.968837810608137e-06,
+      "loss": 0.2509,
+      "step": 1160
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.3187173936528143,
+      "learning_rate": 9.96873731131924e-06,
+      "loss": 0.2071,
+      "step": 1161
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.1811594432603254,
+      "learning_rate": 9.96863665074227e-06,
+      "loss": 0.1889,
+      "step": 1162
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.20653708073265,
+      "learning_rate": 9.9685358288805e-06,
+      "loss": 0.2407,
+      "step": 1163
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.0489369874005263,
+      "learning_rate": 9.968434845737198e-06,
+      "loss": 0.1911,
+      "step": 1164
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 2.005586444929045,
+      "learning_rate": 9.968333701315644e-06,
+      "loss": 0.2535,
+      "step": 1165
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.8373086111149095,
+      "learning_rate": 9.968232395619121e-06,
+      "loss": 0.1946,
+      "step": 1166
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.8352536977437872,
+      "learning_rate": 9.96813092865092e-06,
+      "loss": 0.212,
+      "step": 1167
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 10.533414623896471,
+      "learning_rate": 9.96802930041433e-06,
+      "loss": 0.2433,
+      "step": 1168
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.0594594871116747,
+      "learning_rate": 9.96792751091265e-06,
+      "loss": 0.1699,
+      "step": 1169
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.2073656902736285,
+      "learning_rate": 9.967825560149191e-06,
+      "loss": 0.1534,
+      "step": 1170
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.724766370943612,
+      "learning_rate": 9.967723448127255e-06,
+      "loss": 0.1709,
+      "step": 1171
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.605598461869892,
+      "learning_rate": 9.96762117485016e-06,
+      "loss": 0.1792,
+      "step": 1172
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.2858774891937985,
+      "learning_rate": 9.967518740321224e-06,
+      "loss": 0.1835,
+      "step": 1173
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.5327312362878311,
+      "learning_rate": 9.967416144543773e-06,
+      "loss": 0.1769,
+      "step": 1174
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.3358248752391468,
+      "learning_rate": 9.967313387521139e-06,
+      "loss": 0.2008,
+      "step": 1175
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.8299385432831197,
+      "learning_rate": 9.967210469256657e-06,
+      "loss": 0.2346,
+      "step": 1176
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.1703193065919026,
+      "learning_rate": 9.967107389753663e-06,
+      "loss": 0.2066,
+      "step": 1177
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 2.0022700510143925,
+      "learning_rate": 9.967004149015508e-06,
+      "loss": 0.2555,
+      "step": 1178
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.4813826314481562,
+      "learning_rate": 9.966900747045545e-06,
+      "loss": 0.2709,
+      "step": 1179
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.884743456141887,
+      "learning_rate": 9.966797183847125e-06,
+      "loss": 0.2442,
+      "step": 1180
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 2.064939080807794,
+      "learning_rate": 9.966693459423613e-06,
+      "loss": 0.252,
+      "step": 1181
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.331739228644009,
+      "learning_rate": 9.966589573778376e-06,
+      "loss": 0.1924,
+      "step": 1182
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.6388789240508113,
+      "learning_rate": 9.966485526914784e-06,
+      "loss": 0.2244,
+      "step": 1183
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 2.1134589520693603,
+      "learning_rate": 9.966381318836217e-06,
+      "loss": 0.2254,
+      "step": 1184
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 2.33596388377724,
+      "learning_rate": 9.966276949546054e-06,
+      "loss": 0.2408,
+      "step": 1185
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.1757510599588439,
+      "learning_rate": 9.966172419047688e-06,
+      "loss": 0.2024,
+      "step": 1186
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.4900783442135732,
+      "learning_rate": 9.96606772734451e-06,
+      "loss": 0.2195,
+      "step": 1187
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.5433092924549787,
+      "learning_rate": 9.965962874439916e-06,
+      "loss": 0.2065,
+      "step": 1188
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.6079502233939778,
+      "learning_rate": 9.965857860337312e-06,
+      "loss": 0.2225,
+      "step": 1189
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.2695019322457997,
+      "learning_rate": 9.965752685040106e-06,
+      "loss": 0.233,
+      "step": 1190
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.3211265211648366,
+      "learning_rate": 9.965647348551713e-06,
+      "loss": 0.2087,
+      "step": 1191
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 6.3737631701258355,
+      "learning_rate": 9.965541850875551e-06,
+      "loss": 0.2571,
+      "step": 1192
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.550481660428357,
+      "learning_rate": 9.965436192015047e-06,
+      "loss": 0.2024,
+      "step": 1193
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.6879537246163894,
+      "learning_rate": 9.965330371973627e-06,
+      "loss": 0.2243,
+      "step": 1194
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.6184137505105205,
+      "learning_rate": 9.965224390754728e-06,
+      "loss": 0.2253,
+      "step": 1195
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.3625814320843634,
+      "learning_rate": 9.965118248361791e-06,
+      "loss": 0.2639,
+      "step": 1196
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.3113752552661642,
+      "learning_rate": 9.96501194479826e-06,
+      "loss": 0.25,
+      "step": 1197
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.4997644040738192,
+      "learning_rate": 9.964905480067585e-06,
+      "loss": 0.2184,
+      "step": 1198
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.1841394799837206,
+      "learning_rate": 9.964798854173226e-06,
+      "loss": 0.2479,
+      "step": 1199
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.23311765529864,
+      "learning_rate": 9.964692067118638e-06,
+      "loss": 0.1735,
+      "step": 1200
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.9638356065455886,
+      "learning_rate": 9.964585118907293e-06,
+      "loss": 0.2424,
+      "step": 1201
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 2.0383486978199263,
+      "learning_rate": 9.964478009542659e-06,
+      "loss": 0.2224,
+      "step": 1202
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 3.1660810648551974,
+      "learning_rate": 9.964370739028213e-06,
+      "loss": 0.2383,
+      "step": 1203
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.3042102621875598,
+      "learning_rate": 9.96426330736744e-06,
+      "loss": 0.1548,
+      "step": 1204
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.3229613748702798,
+      "learning_rate": 9.964155714563825e-06,
+      "loss": 0.1977,
+      "step": 1205
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.303266364458382,
+      "learning_rate": 9.96404796062086e-06,
+      "loss": 0.2115,
+      "step": 1206
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.4822427618810599,
+      "learning_rate": 9.963940045542045e-06,
+      "loss": 0.2579,
+      "step": 1207
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.3540751964291668,
+      "learning_rate": 9.96383196933088e-06,
+      "loss": 0.2132,
+      "step": 1208
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.6814753781294958,
+      "learning_rate": 9.963723731990876e-06,
+      "loss": 0.2121,
+      "step": 1209
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.264337773993661,
+      "learning_rate": 9.963615333525545e-06,
+      "loss": 0.2229,
+      "step": 1210
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.2407482674517611,
+      "learning_rate": 9.963506773938407e-06,
+      "loss": 0.1514,
+      "step": 1211
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.0263612458468134,
+      "learning_rate": 9.963398053232983e-06,
+      "loss": 0.2047,
+      "step": 1212
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.1170530213267242,
+      "learning_rate": 9.963289171412805e-06,
+      "loss": 0.2308,
+      "step": 1213
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.2222907766288316,
+      "learning_rate": 9.963180128481407e-06,
+      "loss": 0.1959,
+      "step": 1214
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.9512990188413355,
+      "learning_rate": 9.963070924442327e-06,
+      "loss": 0.2044,
+      "step": 1215
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.8226421923857645,
+      "learning_rate": 9.962961559299112e-06,
+      "loss": 0.2589,
+      "step": 1216
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.2292193745637818,
+      "learning_rate": 9.96285203305531e-06,
+      "loss": 0.2249,
+      "step": 1217
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.366735045633654,
+      "learning_rate": 9.962742345714478e-06,
+      "loss": 0.2723,
+      "step": 1218
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.4177191254496748,
+      "learning_rate": 9.962632497280173e-06,
+      "loss": 0.2404,
+      "step": 1219
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.1159206189716409,
+      "learning_rate": 9.962522487755967e-06,
+      "loss": 0.2305,
+      "step": 1220
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.2420738801036055,
+      "learning_rate": 9.962412317145425e-06,
+      "loss": 0.1768,
+      "step": 1221
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.9856528699547278,
+      "learning_rate": 9.962301985452128e-06,
+      "loss": 0.2182,
+      "step": 1222
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.69177014442735,
+      "learning_rate": 9.962191492679651e-06,
+      "loss": 0.2212,
+      "step": 1223
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.6762725236678755,
+      "learning_rate": 9.962080838831587e-06,
+      "loss": 0.2021,
+      "step": 1224
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.1975803241688392,
+      "learning_rate": 9.961970023911527e-06,
+      "loss": 0.1698,
+      "step": 1225
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.6352117122378995,
+      "learning_rate": 9.961859047923066e-06,
+      "loss": 0.2514,
+      "step": 1226
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.4827589109261874,
+      "learning_rate": 9.961747910869806e-06,
+      "loss": 0.2412,
+      "step": 1227
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.3689301324896115,
+      "learning_rate": 9.961636612755355e-06,
+      "loss": 0.2007,
+      "step": 1228
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.1162307750844793,
+      "learning_rate": 9.961525153583327e-06,
+      "loss": 0.1994,
+      "step": 1229
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.2635035336372953,
+      "learning_rate": 9.961413533357341e-06,
+      "loss": 0.1718,
+      "step": 1230
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.142958390299116,
+      "learning_rate": 9.961301752081018e-06,
+      "loss": 0.1994,
+      "step": 1231
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.6424752360811217,
+      "learning_rate": 9.961189809757985e-06,
+      "loss": 0.2404,
+      "step": 1232
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.521669472048194,
+      "learning_rate": 9.96107770639188e-06,
+      "loss": 0.1776,
+      "step": 1233
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.6049726513288376,
+      "learning_rate": 9.96096544198634e-06,
+      "loss": 0.25,
+      "step": 1234
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 2.3419313956877437,
+      "learning_rate": 9.96085301654501e-06,
+      "loss": 0.2622,
+      "step": 1235
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.667985659207615,
+      "learning_rate": 9.960740430071537e-06,
+      "loss": 0.2577,
+      "step": 1236
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.3878485133994307,
+      "learning_rate": 9.960627682569574e-06,
+      "loss": 0.1936,
+      "step": 1237
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.3813732484649426,
+      "learning_rate": 9.960514774042787e-06,
+      "loss": 0.2353,
+      "step": 1238
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.2648490555164826,
+      "learning_rate": 9.960401704494837e-06,
+      "loss": 0.2376,
+      "step": 1239
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 5.93126487609477,
+      "learning_rate": 9.960288473929397e-06,
+      "loss": 0.2317,
+      "step": 1240
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.7374729506013138,
+      "learning_rate": 9.96017508235014e-06,
+      "loss": 0.2769,
+      "step": 1241
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.0267814395501615,
+      "learning_rate": 9.960061529760748e-06,
+      "loss": 0.2041,
+      "step": 1242
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.671547815358573,
+      "learning_rate": 9.959947816164905e-06,
+      "loss": 0.1681,
+      "step": 1243
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.291475989431219,
+      "learning_rate": 9.959833941566304e-06,
+      "loss": 0.2079,
+      "step": 1244
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.1124200041076264,
+      "learning_rate": 9.959719905968642e-06,
+      "loss": 0.2128,
+      "step": 1245
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.0320198651522543,
+      "learning_rate": 9.95960570937562e-06,
+      "loss": 0.2012,
+      "step": 1246
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.9434107851350768,
+      "learning_rate": 9.959491351790942e-06,
+      "loss": 0.2029,
+      "step": 1247
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.004121744279179,
+      "learning_rate": 9.959376833218325e-06,
+      "loss": 0.2406,
+      "step": 1248
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.5059661592355678,
+      "learning_rate": 9.959262153661484e-06,
+      "loss": 0.234,
+      "step": 1249
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.4129993218408412,
+      "learning_rate": 9.959147313124143e-06,
+      "loss": 0.2122,
+      "step": 1250
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.3093720772495516,
+      "learning_rate": 9.959032311610026e-06,
+      "loss": 0.1951,
+      "step": 1251
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.1662756998827455,
+      "learning_rate": 9.95891714912287e-06,
+      "loss": 0.2174,
+      "step": 1252
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.3896017542575558,
+      "learning_rate": 9.95880182566641e-06,
+      "loss": 0.2634,
+      "step": 1253
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.3822224066460136,
+      "learning_rate": 9.958686341244394e-06,
+      "loss": 0.2132,
+      "step": 1254
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 2.1524623277404693,
+      "learning_rate": 9.958570695860566e-06,
+      "loss": 0.1989,
+      "step": 1255
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.3145134686680187,
+      "learning_rate": 9.958454889518682e-06,
+      "loss": 0.1949,
+      "step": 1256
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.1229722338211459,
+      "learning_rate": 9.958338922222503e-06,
+      "loss": 0.2164,
+      "step": 1257
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.4771108870453908,
+      "learning_rate": 9.958222793975789e-06,
+      "loss": 0.2461,
+      "step": 1258
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.3426861349682295,
+      "learning_rate": 9.958106504782312e-06,
+      "loss": 0.2417,
+      "step": 1259
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.397093274676967,
+      "learning_rate": 9.957990054645847e-06,
+      "loss": 0.2517,
+      "step": 1260
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.042955135942394,
+      "learning_rate": 9.957873443570175e-06,
+      "loss": 0.2421,
+      "step": 1261
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.3357947667875154,
+      "learning_rate": 9.95775667155908e-06,
+      "loss": 0.1659,
+      "step": 1262
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 10.789818761397644,
+      "learning_rate": 9.957639738616351e-06,
+      "loss": 0.2178,
+      "step": 1263
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.2971355911782692,
+      "learning_rate": 9.957522644745787e-06,
+      "loss": 0.2198,
+      "step": 1264
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.6878141004911116,
+      "learning_rate": 9.957405389951185e-06,
+      "loss": 0.2493,
+      "step": 1265
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.9401519453749483,
+      "learning_rate": 9.957287974236355e-06,
+      "loss": 0.2459,
+      "step": 1266
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.0284068260498436,
+      "learning_rate": 9.957170397605105e-06,
+      "loss": 0.1921,
+      "step": 1267
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.1738073693089135,
+      "learning_rate": 9.957052660061255e-06,
+      "loss": 0.2326,
+      "step": 1268
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 2.8473907263187663,
+      "learning_rate": 9.956934761608625e-06,
+      "loss": 0.2097,
+      "step": 1269
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.3509803383666765,
+      "learning_rate": 9.956816702251042e-06,
+      "loss": 0.2554,
+      "step": 1270
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.9484346146550722,
+      "learning_rate": 9.956698481992339e-06,
+      "loss": 0.2437,
+      "step": 1271
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.103761525733859,
+      "learning_rate": 9.956580100836351e-06,
+      "loss": 0.2092,
+      "step": 1272
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 2.0892314673505945,
+      "learning_rate": 9.956461558786926e-06,
+      "loss": 0.2204,
+      "step": 1273
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.5686282691041467,
+      "learning_rate": 9.956342855847905e-06,
+      "loss": 0.238,
+      "step": 1274
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.4123511575664331,
+      "learning_rate": 9.956223992023145e-06,
+      "loss": 0.21,
+      "step": 1275
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.2917595579665087,
+      "learning_rate": 9.956104967316507e-06,
+      "loss": 0.2105,
+      "step": 1276
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.7462893041557859,
+      "learning_rate": 9.95598578173185e-06,
+      "loss": 0.2134,
+      "step": 1277
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.4425413552088115,
+      "learning_rate": 9.955866435273045e-06,
+      "loss": 0.1861,
+      "step": 1278
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.9783880827183091,
+      "learning_rate": 9.955746927943965e-06,
+      "loss": 0.2163,
+      "step": 1279
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 2.386231642742895,
+      "learning_rate": 9.955627259748493e-06,
+      "loss": 0.3036,
+      "step": 1280
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.9232775033192736,
+      "learning_rate": 9.955507430690508e-06,
+      "loss": 0.1612,
+      "step": 1281
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.3909334178265584,
+      "learning_rate": 9.955387440773902e-06,
+      "loss": 0.2079,
+      "step": 1282
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.385585852710067,
+      "learning_rate": 9.95526729000257e-06,
+      "loss": 0.204,
+      "step": 1283
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.0908995661892074,
+      "learning_rate": 9.955146978380412e-06,
+      "loss": 0.2227,
+      "step": 1284
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 2.2701917104231217,
+      "learning_rate": 9.955026505911334e-06,
+      "loss": 0.2973,
+      "step": 1285
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.1477740473372868,
+      "learning_rate": 9.954905872599248e-06,
+      "loss": 0.2612,
+      "step": 1286
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.332133065524745,
+      "learning_rate": 9.954785078448065e-06,
+      "loss": 0.1926,
+      "step": 1287
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.4570795311995,
+      "learning_rate": 9.954664123461713e-06,
+      "loss": 0.2416,
+      "step": 1288
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.572419854412675,
+      "learning_rate": 9.95454300764411e-06,
+      "loss": 0.2193,
+      "step": 1289
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.3232022719327843,
+      "learning_rate": 9.954421730999195e-06,
+      "loss": 0.2544,
+      "step": 1290
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.2791423806239195,
+      "learning_rate": 9.954300293530901e-06,
+      "loss": 0.2226,
+      "step": 1291
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.379618436482356,
+      "learning_rate": 9.954178695243171e-06,
+      "loss": 0.2087,
+      "step": 1292
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.9485328476618322,
+      "learning_rate": 9.95405693613995e-06,
+      "loss": 0.2492,
+      "step": 1293
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.1716694953654485,
+      "learning_rate": 9.953935016225194e-06,
+      "loss": 0.2192,
+      "step": 1294
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.402774383390787,
+      "learning_rate": 9.953812935502858e-06,
+      "loss": 0.2273,
+      "step": 1295
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.083085200964765,
+      "learning_rate": 9.953690693976904e-06,
+      "loss": 0.193,
+      "step": 1296
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.2707276619560053,
+      "learning_rate": 9.953568291651302e-06,
+      "loss": 0.2255,
+      "step": 1297
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.0974408291293098,
+      "learning_rate": 9.953445728530028e-06,
+      "loss": 0.2249,
+      "step": 1298
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.3238291434528646,
+      "learning_rate": 9.953323004617054e-06,
+      "loss": 0.2317,
+      "step": 1299
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.4860295799177319,
+      "learning_rate": 9.953200119916368e-06,
+      "loss": 0.2665,
+      "step": 1300
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.277053867192546,
+      "learning_rate": 9.953077074431956e-06,
+      "loss": 0.2216,
+      "step": 1301
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.9182660480371374,
+      "learning_rate": 9.952953868167816e-06,
+      "loss": 0.2312,
+      "step": 1302
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.9658080168694501,
+      "learning_rate": 9.952830501127945e-06,
+      "loss": 0.1912,
+      "step": 1303
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.711948657733515,
+      "learning_rate": 9.952706973316348e-06,
+      "loss": 0.2112,
+      "step": 1304
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.3920087646882506,
+      "learning_rate": 9.952583284737034e-06,
+      "loss": 0.2189,
+      "step": 1305
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.5237842409747588,
+      "learning_rate": 9.95245943539402e-06,
+      "loss": 0.2224,
+      "step": 1306
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.3562791254282667,
+      "learning_rate": 9.952335425291321e-06,
+      "loss": 0.2423,
+      "step": 1307
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.257638058798033,
+      "learning_rate": 9.95221125443297e-06,
+      "loss": 0.2102,
+      "step": 1308
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.63113765055615,
+      "learning_rate": 9.952086922822993e-06,
+      "loss": 0.2229,
+      "step": 1309
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.3367001409032362,
+      "learning_rate": 9.951962430465427e-06,
+      "loss": 0.2057,
+      "step": 1310
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.3333175606713592,
+      "learning_rate": 9.951837777364313e-06,
+      "loss": 0.2416,
+      "step": 1311
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.0052374870967802,
+      "learning_rate": 9.951712963523697e-06,
+      "loss": 0.1875,
+      "step": 1312
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.1334056435933932,
+      "learning_rate": 9.951587988947631e-06,
+      "loss": 0.1977,
+      "step": 1313
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.5264084554744755,
+      "learning_rate": 9.95146285364017e-06,
+      "loss": 0.2628,
+      "step": 1314
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.5413319271554495,
+      "learning_rate": 9.95133755760538e-06,
+      "loss": 0.2729,
+      "step": 1315
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.544504526434848,
+      "learning_rate": 9.951212100847324e-06,
+      "loss": 0.2037,
+      "step": 1316
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.2609050115195732,
+      "learning_rate": 9.951086483370075e-06,
+      "loss": 0.2364,
+      "step": 1317
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.5855154124106023,
+      "learning_rate": 9.950960705177714e-06,
+      "loss": 0.2317,
+      "step": 1318
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.740806768025055,
+      "learning_rate": 9.95083476627432e-06,
+      "loss": 0.209,
+      "step": 1319
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.540022908611639,
+      "learning_rate": 9.950708666663984e-06,
+      "loss": 0.1629,
+      "step": 1320
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.75534631873462,
+      "learning_rate": 9.950582406350797e-06,
+      "loss": 0.2989,
+      "step": 1321
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.9845363464992838,
+      "learning_rate": 9.950455985338858e-06,
+      "loss": 0.2006,
+      "step": 1322
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.4813110476802542,
+      "learning_rate": 9.950329403632271e-06,
+      "loss": 0.242,
+      "step": 1323
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.2189362670924673,
+      "learning_rate": 9.950202661235146e-06,
+      "loss": 0.1875,
+      "step": 1324
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.1770590527176679,
+      "learning_rate": 9.950075758151595e-06,
+      "loss": 0.1936,
+      "step": 1325
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.249276026044338,
+      "learning_rate": 9.949948694385739e-06,
+      "loss": 0.2439,
+      "step": 1326
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.153297579135089,
+      "learning_rate": 9.9498214699417e-06,
+      "loss": 0.2089,
+      "step": 1327
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.4586405061604863,
+      "learning_rate": 9.949694084823612e-06,
+      "loss": 0.2333,
+      "step": 1328
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.4073584638376784,
+      "learning_rate": 9.949566539035607e-06,
+      "loss": 0.2186,
+      "step": 1329
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.221203904238583,
+      "learning_rate": 9.949438832581825e-06,
+      "loss": 0.1805,
+      "step": 1330
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.4531822372772065,
+      "learning_rate": 9.949310965466414e-06,
+      "loss": 0.235,
+      "step": 1331
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.3207062861966896,
+      "learning_rate": 9.949182937693522e-06,
+      "loss": 0.2227,
+      "step": 1332
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.6043480212221013,
+      "learning_rate": 9.949054749267307e-06,
+      "loss": 0.1974,
+      "step": 1333
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.4942762911670193,
+      "learning_rate": 9.948926400191927e-06,
+      "loss": 0.23,
+      "step": 1334
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.4938888537141437,
+      "learning_rate": 9.948797890471552e-06,
+      "loss": 0.1747,
+      "step": 1335
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.0753424783689003,
+      "learning_rate": 9.948669220110351e-06,
+      "loss": 0.1549,
+      "step": 1336
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.6532677936718372,
+      "learning_rate": 9.948540389112502e-06,
+      "loss": 0.2379,
+      "step": 1337
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.9361034578491436,
+      "learning_rate": 9.948411397482185e-06,
+      "loss": 0.2344,
+      "step": 1338
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.5441670171140602,
+      "learning_rate": 9.94828224522359e-06,
+      "loss": 0.2837,
+      "step": 1339
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.7279342190801754,
+      "learning_rate": 9.948152932340906e-06,
+      "loss": 0.2622,
+      "step": 1340
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.5261193169984717,
+      "learning_rate": 9.948023458838334e-06,
+      "loss": 0.2219,
+      "step": 1341
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.5614468792534895,
+      "learning_rate": 9.947893824720076e-06,
+      "loss": 0.2559,
+      "step": 1342
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.184188992103389,
+      "learning_rate": 9.947764029990336e-06,
+      "loss": 0.2141,
+      "step": 1343
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.0646445146039465,
+      "learning_rate": 9.947634074653334e-06,
+      "loss": 0.2024,
+      "step": 1344
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.1759162521166031,
+      "learning_rate": 9.947503958713283e-06,
+      "loss": 0.1787,
+      "step": 1345
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.6675339574319505,
+      "learning_rate": 9.947373682174408e-06,
+      "loss": 0.2568,
+      "step": 1346
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.3892255341979134,
+      "learning_rate": 9.94724324504094e-06,
+      "loss": 0.1884,
+      "step": 1347
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.6085484914725872,
+      "learning_rate": 9.94711264731711e-06,
+      "loss": 0.2303,
+      "step": 1348
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.5877687048470224,
+      "learning_rate": 9.94698188900716e-06,
+      "loss": 0.2415,
+      "step": 1349
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.5258051105202044,
+      "learning_rate": 9.946850970115331e-06,
+      "loss": 0.2425,
+      "step": 1350
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 2.1545654502974383,
+      "learning_rate": 9.946719890645877e-06,
+      "loss": 0.2699,
+      "step": 1351
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.2825368373629518,
+      "learning_rate": 9.94658865060305e-06,
+      "loss": 0.2397,
+      "step": 1352
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.2983529878189863,
+      "learning_rate": 9.946457249991111e-06,
+      "loss": 0.191,
+      "step": 1353
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.2835750737517753,
+      "learning_rate": 9.946325688814325e-06,
+      "loss": 0.249,
+      "step": 1354
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.2483322932049243,
+      "learning_rate": 9.946193967076963e-06,
+      "loss": 0.255,
+      "step": 1355
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.0370508606705902,
+      "learning_rate": 9.946062084783301e-06,
+      "loss": 0.187,
+      "step": 1356
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.2995529059916984,
+      "learning_rate": 9.945930041937619e-06,
+      "loss": 0.1895,
+      "step": 1357
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 2.05945379724387,
+      "learning_rate": 9.945797838544204e-06,
+      "loss": 0.2547,
+      "step": 1358
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.7581240939430562,
+      "learning_rate": 9.945665474607347e-06,
+      "loss": 0.24,
+      "step": 1359
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.1517184031631802,
+      "learning_rate": 9.945532950131344e-06,
+      "loss": 0.2199,
+      "step": 1360
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.1363434762006293,
+      "learning_rate": 9.945400265120497e-06,
+      "loss": 0.2206,
+      "step": 1361
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.897448528404319,
+      "learning_rate": 9.945267419579115e-06,
+      "loss": 0.25,
+      "step": 1362
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.3102391893740681,
+      "learning_rate": 9.945134413511509e-06,
+      "loss": 0.2354,
+      "step": 1363
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.5377930280525893,
+      "learning_rate": 9.945001246921996e-06,
+      "loss": 0.2257,
+      "step": 1364
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.8462730650068777,
+      "learning_rate": 9.944867919814899e-06,
+      "loss": 0.1714,
+      "step": 1365
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.6075831850302882,
+      "learning_rate": 9.944734432194544e-06,
+      "loss": 0.1934,
+      "step": 1366
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.1877622957792544,
+      "learning_rate": 9.94460078406527e-06,
+      "loss": 0.2346,
+      "step": 1367
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.7551123161729707,
+      "learning_rate": 9.944466975431409e-06,
+      "loss": 0.2381,
+      "step": 1368
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.1077430064379408,
+      "learning_rate": 9.944333006297305e-06,
+      "loss": 0.2215,
+      "step": 1369
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.0165811567796186,
+      "learning_rate": 9.94419887666731e-06,
+      "loss": 0.1865,
+      "step": 1370
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.119640726259161,
+      "learning_rate": 9.944064586545776e-06,
+      "loss": 0.1799,
+      "step": 1371
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.3376888693833824,
+      "learning_rate": 9.943930135937064e-06,
+      "loss": 0.2013,
+      "step": 1372
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.6808923924186634,
+      "learning_rate": 9.943795524845534e-06,
+      "loss": 0.2159,
+      "step": 1373
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.544555965548068,
+      "learning_rate": 9.94366075327556e-06,
+      "loss": 0.2203,
+      "step": 1374
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.283266753966333,
+      "learning_rate": 9.943525821231516e-06,
+      "loss": 0.2334,
+      "step": 1375
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.470960918695552,
+      "learning_rate": 9.943390728717779e-06,
+      "loss": 0.2202,
+      "step": 1376
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.2960172597286255,
+      "learning_rate": 9.943255475738739e-06,
+      "loss": 0.2388,
+      "step": 1377
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.562513137106505,
+      "learning_rate": 9.943120062298781e-06,
+      "loss": 0.2179,
+      "step": 1378
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.4123173582943298,
+      "learning_rate": 9.942984488402305e-06,
+      "loss": 0.187,
+      "step": 1379
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.737074117815268,
+      "learning_rate": 9.942848754053709e-06,
+      "loss": 0.254,
+      "step": 1380
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.2577759251696856,
+      "learning_rate": 9.942712859257401e-06,
+      "loss": 0.2633,
+      "step": 1381
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.4286484342867316,
+      "learning_rate": 9.94257680401779e-06,
+      "loss": 0.1826,
+      "step": 1382
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.1153928206571069,
+      "learning_rate": 9.942440588339294e-06,
+      "loss": 0.253,
+      "step": 1383
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.0991696813445084,
+      "learning_rate": 9.942304212226337e-06,
+      "loss": 0.2022,
+      "step": 1384
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.3675015116165425,
+      "learning_rate": 9.942167675683341e-06,
+      "loss": 0.2538,
+      "step": 1385
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.9572844228826728,
+      "learning_rate": 9.94203097871474e-06,
+      "loss": 0.1778,
+      "step": 1386
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.2714968664607538,
+      "learning_rate": 9.941894121324973e-06,
+      "loss": 0.2252,
+      "step": 1387
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.8838669352708726,
+      "learning_rate": 9.94175710351848e-06,
+      "loss": 0.205,
+      "step": 1388
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 2.122189030440291,
+      "learning_rate": 9.94161992529971e-06,
+      "loss": 0.2055,
+      "step": 1389
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.547466208446442,
+      "learning_rate": 9.941482586673114e-06,
+      "loss": 0.3065,
+      "step": 1390
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.4330556727633588,
+      "learning_rate": 9.941345087643155e-06,
+      "loss": 0.2036,
+      "step": 1391
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.5114220320612026,
+      "learning_rate": 9.941207428214291e-06,
+      "loss": 0.2751,
+      "step": 1392
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.3187018444152647,
+      "learning_rate": 9.941069608390992e-06,
+      "loss": 0.1884,
+      "step": 1393
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.173947504355246,
+      "learning_rate": 9.940931628177733e-06,
+      "loss": 0.243,
+      "step": 1394
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.904863513859335,
+      "learning_rate": 9.940793487578992e-06,
+      "loss": 0.2419,
+      "step": 1395
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.5625006973741884,
+      "learning_rate": 9.940655186599252e-06,
+      "loss": 0.2209,
+      "step": 1396
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.1110921401359686,
+      "learning_rate": 9.940516725243006e-06,
+      "loss": 0.1618,
+      "step": 1397
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.181789250130163,
+      "learning_rate": 9.940378103514745e-06,
+      "loss": 0.2193,
+      "step": 1398
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.6314330228536515,
+      "learning_rate": 9.94023932141897e-06,
+      "loss": 0.2225,
+      "step": 1399
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.4851972786996792,
+      "learning_rate": 9.940100378960184e-06,
+      "loss": 0.254,
+      "step": 1400
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.1518260116012338,
+      "learning_rate": 9.939961276142901e-06,
+      "loss": 0.2202,
+      "step": 1401
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.3836100057974914,
+      "learning_rate": 9.939822012971633e-06,
+      "loss": 0.1871,
+      "step": 1402
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.2722214585421814,
+      "learning_rate": 9.939682589450904e-06,
+      "loss": 0.242,
+      "step": 1403
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.9335578703200498,
+      "learning_rate": 9.939543005585235e-06,
+      "loss": 0.2175,
+      "step": 1404
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.6538836515934254,
+      "learning_rate": 9.939403261379162e-06,
+      "loss": 0.2269,
+      "step": 1405
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.7030154212645014,
+      "learning_rate": 9.939263356837217e-06,
+      "loss": 0.229,
+      "step": 1406
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.7182635572618217,
+      "learning_rate": 9.939123291963943e-06,
+      "loss": 0.256,
+      "step": 1407
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.9920163476311024,
+      "learning_rate": 9.938983066763887e-06,
+      "loss": 0.2265,
+      "step": 1408
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.8060054817315856,
+      "learning_rate": 9.938842681241603e-06,
+      "loss": 0.2794,
+      "step": 1409
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.9743513442282203,
+      "learning_rate": 9.938702135401643e-06,
+      "loss": 0.207,
+      "step": 1410
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.9105397835964328,
+      "learning_rate": 9.938561429248573e-06,
+      "loss": 0.2575,
+      "step": 1411
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.6439283538918967,
+      "learning_rate": 9.938420562786957e-06,
+      "loss": 0.1678,
+      "step": 1412
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.3070187527594355,
+      "learning_rate": 9.938279536021372e-06,
+      "loss": 0.1575,
+      "step": 1413
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.292882441034289,
+      "learning_rate": 9.938138348956393e-06,
+      "loss": 0.2196,
+      "step": 1414
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.558598169401782,
+      "learning_rate": 9.937997001596602e-06,
+      "loss": 0.2597,
+      "step": 1415
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.632995575812348,
+      "learning_rate": 9.937855493946591e-06,
+      "loss": 0.2171,
+      "step": 1416
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 2.444298762588584,
+      "learning_rate": 9.93771382601095e-06,
+      "loss": 0.2368,
+      "step": 1417
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.344527490933099,
+      "learning_rate": 9.937571997794278e-06,
+      "loss": 0.2174,
+      "step": 1418
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.298435461591275,
+      "learning_rate": 9.937430009301182e-06,
+      "loss": 0.2298,
+      "step": 1419
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.3214595484508733,
+      "learning_rate": 9.937287860536267e-06,
+      "loss": 0.2147,
+      "step": 1420
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 2.8712013626691517,
+      "learning_rate": 9.937145551504148e-06,
+      "loss": 0.2133,
+      "step": 1421
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 2.0668684905728023,
+      "learning_rate": 9.937003082209448e-06,
+      "loss": 0.2109,
+      "step": 1422
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.4536391394650237,
+      "learning_rate": 9.936860452656787e-06,
+      "loss": 0.1841,
+      "step": 1423
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.018939687030773,
+      "learning_rate": 9.936717662850797e-06,
+      "loss": 0.1581,
+      "step": 1424
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.4600950626015665,
+      "learning_rate": 9.936574712796114e-06,
+      "loss": 0.2144,
+      "step": 1425
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.2381614302176873,
+      "learning_rate": 9.936431602497376e-06,
+      "loss": 0.1782,
+      "step": 1426
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.2979502385661914,
+      "learning_rate": 9.93628833195923e-06,
+      "loss": 0.1919,
+      "step": 1427
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.2507821884247061,
+      "learning_rate": 9.936144901186327e-06,
+      "loss": 0.2077,
+      "step": 1428
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.1520703330809725,
+      "learning_rate": 9.93600131018332e-06,
+      "loss": 0.2051,
+      "step": 1429
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.422919900754583,
+      "learning_rate": 9.935857558954873e-06,
+      "loss": 0.2243,
+      "step": 1430
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.2917285889605794,
+      "learning_rate": 9.935713647505652e-06,
+      "loss": 0.2649,
+      "step": 1431
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.176144033841741,
+      "learning_rate": 9.935569575840327e-06,
+      "loss": 0.2844,
+      "step": 1432
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.1105960284937233,
+      "learning_rate": 9.935425343963574e-06,
+      "loss": 0.2093,
+      "step": 1433
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.580679968665866,
+      "learning_rate": 9.935280951880078e-06,
+      "loss": 0.1966,
+      "step": 1434
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.2566399514098965,
+      "learning_rate": 9.935136399594524e-06,
+      "loss": 0.2419,
+      "step": 1435
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.438045349024953,
+      "learning_rate": 9.934991687111607e-06,
+      "loss": 0.1835,
+      "step": 1436
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.3021264917055766,
+      "learning_rate": 9.934846814436019e-06,
+      "loss": 0.2191,
+      "step": 1437
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.7284553905272237,
+      "learning_rate": 9.934701781572465e-06,
+      "loss": 0.2478,
+      "step": 1438
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.6458279646311087,
+      "learning_rate": 9.934556588525655e-06,
+      "loss": 0.2499,
+      "step": 1439
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.0028507532186248,
+      "learning_rate": 9.9344112353003e-06,
+      "loss": 0.176,
+      "step": 1440
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.6202680622139412,
+      "learning_rate": 9.93426572190112e-06,
+      "loss": 0.2434,
+      "step": 1441
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.1215189831774026,
+      "learning_rate": 9.934120048332837e-06,
+      "loss": 0.2428,
+      "step": 1442
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.1019277922319155,
+      "learning_rate": 9.933974214600179e-06,
+      "loss": 0.2095,
+      "step": 1443
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.7599852337724233,
+      "learning_rate": 9.933828220707881e-06,
+      "loss": 0.1743,
+      "step": 1444
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.2191052725199893,
+      "learning_rate": 9.933682066660683e-06,
+      "loss": 0.2034,
+      "step": 1445
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.3180099045309601,
+      "learning_rate": 9.933535752463326e-06,
+      "loss": 0.2392,
+      "step": 1446
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.2577090665096775,
+      "learning_rate": 9.933389278120563e-06,
+      "loss": 0.2278,
+      "step": 1447
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.283980521650947,
+      "learning_rate": 9.933242643637146e-06,
+      "loss": 0.2444,
+      "step": 1448
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.1273649279798719,
+      "learning_rate": 9.933095849017838e-06,
+      "loss": 0.2285,
+      "step": 1449
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.3534213166732552,
+      "learning_rate": 9.9329488942674e-06,
+      "loss": 0.1995,
+      "step": 1450
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.2907321347436045,
+      "learning_rate": 9.932801779390606e-06,
+      "loss": 0.2208,
+      "step": 1451
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.2839722498456718,
+      "learning_rate": 9.932654504392228e-06,
+      "loss": 0.2427,
+      "step": 1452
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.904910643470907,
+      "learning_rate": 9.932507069277049e-06,
+      "loss": 0.1794,
+      "step": 1453
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.5669195993997176,
+      "learning_rate": 9.932359474049852e-06,
+      "loss": 0.2149,
+      "step": 1454
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.3359998192110696,
+      "learning_rate": 9.932211718715434e-06,
+      "loss": 0.2046,
+      "step": 1455
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.6120266897091369,
+      "learning_rate": 9.932063803278586e-06,
+      "loss": 0.2053,
+      "step": 1456
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.8376242452063587,
+      "learning_rate": 9.93191572774411e-06,
+      "loss": 0.1456,
+      "step": 1457
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.7011007300652667,
+      "learning_rate": 9.931767492116813e-06,
+      "loss": 0.2235,
+      "step": 1458
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.7554097135537259,
+      "learning_rate": 9.93161909640151e-06,
+      "loss": 0.2363,
+      "step": 1459
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.3508268170478546,
+      "learning_rate": 9.931470540603013e-06,
+      "loss": 0.2303,
+      "step": 1460
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.2873324720405162,
+      "learning_rate": 9.931321824726145e-06,
+      "loss": 0.239,
+      "step": 1461
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.3579153494956782,
+      "learning_rate": 9.931172948775739e-06,
+      "loss": 0.2097,
+      "step": 1462
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.3462217353113473,
+      "learning_rate": 9.931023912756619e-06,
+      "loss": 0.2409,
+      "step": 1463
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.2075256620529897,
+      "learning_rate": 9.93087471667363e-06,
+      "loss": 0.1818,
+      "step": 1464
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.257602092073513,
+      "learning_rate": 9.930725360531612e-06,
+      "loss": 0.2297,
+      "step": 1465
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.3914680283040164,
+      "learning_rate": 9.930575844335413e-06,
+      "loss": 0.1992,
+      "step": 1466
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.311119701991059,
+      "learning_rate": 9.930426168089885e-06,
+      "loss": 0.2293,
+      "step": 1467
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.8664334582217428,
+      "learning_rate": 9.93027633179989e-06,
+      "loss": 0.2491,
+      "step": 1468
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.9932090290684732,
+      "learning_rate": 9.930126335470292e-06,
+      "loss": 0.2608,
+      "step": 1469
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.5488989502346024,
+      "learning_rate": 9.929976179105956e-06,
+      "loss": 0.2295,
+      "step": 1470
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.285774105625606,
+      "learning_rate": 9.929825862711759e-06,
+      "loss": 0.1918,
+      "step": 1471
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.207334389320146,
+      "learning_rate": 9.929675386292578e-06,
+      "loss": 0.1495,
+      "step": 1472
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.2493310554007533,
+      "learning_rate": 9.929524749853302e-06,
+      "loss": 0.2067,
+      "step": 1473
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.6259435266569653,
+      "learning_rate": 9.929373953398816e-06,
+      "loss": 0.2093,
+      "step": 1474
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.1367731507382957,
+      "learning_rate": 9.929222996934018e-06,
+      "loss": 0.2437,
+      "step": 1475
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.5065330032747999,
+      "learning_rate": 9.929071880463806e-06,
+      "loss": 0.2575,
+      "step": 1476
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.655571754828104,
+      "learning_rate": 9.928920603993086e-06,
+      "loss": 0.249,
+      "step": 1477
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.8600985497662255,
+      "learning_rate": 9.92876916752677e-06,
+      "loss": 0.2634,
+      "step": 1478
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 2.040487598945859,
+      "learning_rate": 9.928617571069772e-06,
+      "loss": 0.2796,
+      "step": 1479
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.563622510151303,
+      "learning_rate": 9.928465814627016e-06,
+      "loss": 0.2752,
+      "step": 1480
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.2630579494810283,
+      "learning_rate": 9.92831389820342e-06,
+      "loss": 0.2089,
+      "step": 1481
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.358966215099504,
+      "learning_rate": 9.928161821803926e-06,
+      "loss": 0.2064,
+      "step": 1482
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.786686854626152,
+      "learning_rate": 9.928009585433464e-06,
+      "loss": 0.2559,
+      "step": 1483
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.4558620901161794,
+      "learning_rate": 9.927857189096976e-06,
+      "loss": 0.2623,
+      "step": 1484
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.927774937134356,
+      "learning_rate": 9.927704632799412e-06,
+      "loss": 0.1872,
+      "step": 1485
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.247890064077216,
+      "learning_rate": 9.92755191654572e-06,
+      "loss": 0.2275,
+      "step": 1486
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.5040339155629465,
+      "learning_rate": 9.927399040340859e-06,
+      "loss": 0.2609,
+      "step": 1487
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.3982488568899964,
+      "learning_rate": 9.927246004189792e-06,
+      "loss": 0.1606,
+      "step": 1488
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.3871958274617135,
+      "learning_rate": 9.927092808097487e-06,
+      "loss": 0.2078,
+      "step": 1489
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 2.0401428218182565,
+      "learning_rate": 9.926939452068916e-06,
+      "loss": 0.2636,
+      "step": 1490
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.3224710251413991,
+      "learning_rate": 9.926785936109057e-06,
+      "loss": 0.1908,
+      "step": 1491
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.3919545311596058,
+      "learning_rate": 9.926632260222894e-06,
+      "loss": 0.2049,
+      "step": 1492
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 3.126685736885422,
+      "learning_rate": 9.926478424415413e-06,
+      "loss": 0.2296,
+      "step": 1493
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.9963383131299739,
+      "learning_rate": 9.926324428691612e-06,
+      "loss": 0.1904,
+      "step": 1494
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.4432020179031033,
+      "learning_rate": 9.926170273056486e-06,
+      "loss": 0.2054,
+      "step": 1495
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.4810238806443188,
+      "learning_rate": 9.92601595751504e-06,
+      "loss": 0.254,
+      "step": 1496
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.6349128830008455,
+      "learning_rate": 9.925861482072282e-06,
+      "loss": 0.2422,
+      "step": 1497
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.553909857272511,
+      "learning_rate": 9.92570684673323e-06,
+      "loss": 0.2639,
+      "step": 1498
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.472225107124814,
+      "learning_rate": 9.9255520515029e-06,
+      "loss": 0.182,
+      "step": 1499
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.5495215820202455,
+      "learning_rate": 9.92539709638632e-06,
+      "loss": 0.2548,
+      "step": 1500
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.246681333442473,
+      "learning_rate": 9.925241981388515e-06,
+      "loss": 0.185,
+      "step": 1501
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.3056603619245588,
+      "learning_rate": 9.925086706514526e-06,
+      "loss": 0.2229,
+      "step": 1502
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.3398401855332365,
+      "learning_rate": 9.92493127176939e-06,
+      "loss": 0.2582,
+      "step": 1503
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.2401830332183985,
+      "learning_rate": 9.924775677158149e-06,
+      "loss": 0.2117,
+      "step": 1504
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.3605675796238905,
+      "learning_rate": 9.924619922685861e-06,
+      "loss": 0.2413,
+      "step": 1505
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.637726271080243,
+      "learning_rate": 9.924464008357579e-06,
+      "loss": 0.2516,
+      "step": 1506
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.7343361421691201,
+      "learning_rate": 9.924307934178362e-06,
+      "loss": 0.2327,
+      "step": 1507
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.6343117543049266,
+      "learning_rate": 9.924151700153277e-06,
+      "loss": 0.2638,
+      "step": 1508
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.5396637407815261,
+      "learning_rate": 9.923995306287396e-06,
+      "loss": 0.1956,
+      "step": 1509
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.9025192698983319,
+      "learning_rate": 9.9238387525858e-06,
+      "loss": 0.1819,
+      "step": 1510
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.3165198639309392,
+      "learning_rate": 9.923682039053564e-06,
+      "loss": 0.2115,
+      "step": 1511
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.5177354571507038,
+      "learning_rate": 9.923525165695776e-06,
+      "loss": 0.2427,
+      "step": 1512
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.1548434317403464,
+      "learning_rate": 9.923368132517532e-06,
+      "loss": 0.2277,
+      "step": 1513
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.8774870749128998,
+      "learning_rate": 9.923210939523926e-06,
+      "loss": 0.2333,
+      "step": 1514
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.7937855761020567,
+      "learning_rate": 9.923053586720064e-06,
+      "loss": 0.2544,
+      "step": 1515
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.3809664386644378,
+      "learning_rate": 9.92289607411105e-06,
+      "loss": 0.2286,
+      "step": 1516
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.3610573154453982,
+      "learning_rate": 9.922738401701999e-06,
+      "loss": 0.2026,
+      "step": 1517
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 2.011279189595608,
+      "learning_rate": 9.922580569498029e-06,
+      "loss": 0.2515,
+      "step": 1518
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.420027877125141,
+      "learning_rate": 9.922422577504262e-06,
+      "loss": 0.2628,
+      "step": 1519
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.372851047822403,
+      "learning_rate": 9.922264425725827e-06,
+      "loss": 0.2544,
+      "step": 1520
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.2326609610840948,
+      "learning_rate": 9.92210611416786e-06,
+      "loss": 0.1719,
+      "step": 1521
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.4673433720723565,
+      "learning_rate": 9.921947642835497e-06,
+      "loss": 0.2156,
+      "step": 1522
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.9366570358531431,
+      "learning_rate": 9.921789011733885e-06,
+      "loss": 0.1512,
+      "step": 1523
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.217399868722497,
+      "learning_rate": 9.921630220868168e-06,
+      "loss": 0.2243,
+      "step": 1524
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.1912031156874492,
+      "learning_rate": 9.921471270243508e-06,
+      "loss": 0.1855,
+      "step": 1525
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.957343026388638,
+      "learning_rate": 9.921312159865057e-06,
+      "loss": 0.1477,
+      "step": 1526
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.6468762192255644,
+      "learning_rate": 9.921152889737985e-06,
+      "loss": 0.2339,
+      "step": 1527
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.1824135530202262,
+      "learning_rate": 9.920993459867461e-06,
+      "loss": 0.2046,
+      "step": 1528
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.5344947913853038,
+      "learning_rate": 9.920833870258657e-06,
+      "loss": 0.2545,
+      "step": 1529
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.7349035643290134,
+      "learning_rate": 9.920674120916757e-06,
+      "loss": 0.3199,
+      "step": 1530
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.9240071917540796,
+      "learning_rate": 9.920514211846948e-06,
+      "loss": 0.1617,
+      "step": 1531
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.8677659635026298,
+      "learning_rate": 9.920354143054415e-06,
+      "loss": 0.2548,
+      "step": 1532
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.3452984100250807,
+      "learning_rate": 9.920193914544358e-06,
+      "loss": 0.2031,
+      "step": 1533
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.405232326179258,
+      "learning_rate": 9.920033526321979e-06,
+      "loss": 0.2022,
+      "step": 1534
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.2810562382703359,
+      "learning_rate": 9.91987297839248e-06,
+      "loss": 0.2581,
+      "step": 1535
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.715513441839788,
+      "learning_rate": 9.919712270761077e-06,
+      "loss": 0.2456,
+      "step": 1536
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.3720471906240994,
+      "learning_rate": 9.919551403432984e-06,
+      "loss": 0.2732,
+      "step": 1537
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.3479825495558215,
+      "learning_rate": 9.919390376413423e-06,
+      "loss": 0.2247,
+      "step": 1538
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.3472761723920377,
+      "learning_rate": 9.91922918970762e-06,
+      "loss": 0.222,
+      "step": 1539
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.3891760867098744,
+      "learning_rate": 9.919067843320811e-06,
+      "loss": 0.2444,
+      "step": 1540
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.2882551323606894,
+      "learning_rate": 9.91890633725823e-06,
+      "loss": 0.2541,
+      "step": 1541
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.6948195391177954,
+      "learning_rate": 9.918744671525123e-06,
+      "loss": 0.2449,
+      "step": 1542
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.4570364568479226,
+      "learning_rate": 9.918582846126731e-06,
+      "loss": 0.2763,
+      "step": 1543
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.2650961654093873,
+      "learning_rate": 9.918420861068317e-06,
+      "loss": 0.1437,
+      "step": 1544
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.6880087394677383,
+      "learning_rate": 9.91825871635513e-06,
+      "loss": 0.2696,
+      "step": 1545
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.4752327012109798,
+      "learning_rate": 9.918096411992438e-06,
+      "loss": 0.1895,
+      "step": 1546
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.0949051308959203,
+      "learning_rate": 9.917933947985508e-06,
+      "loss": 0.225,
+      "step": 1547
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.1988786718427447,
+      "learning_rate": 9.917771324339613e-06,
+      "loss": 0.2069,
+      "step": 1548
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.6885420729734515,
+      "learning_rate": 9.917608541060034e-06,
+      "loss": 0.2096,
+      "step": 1549
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.427295178911265,
+      "learning_rate": 9.917445598152054e-06,
+      "loss": 0.2302,
+      "step": 1550
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.1164077775462906,
+      "learning_rate": 9.917282495620961e-06,
+      "loss": 0.1617,
+      "step": 1551
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.590144201658293,
+      "learning_rate": 9.917119233472052e-06,
+      "loss": 0.2591,
+      "step": 1552
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.2893890259910648,
+      "learning_rate": 9.916955811710625e-06,
+      "loss": 0.2314,
+      "step": 1553
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.1762617213083757,
+      "learning_rate": 9.916792230341985e-06,
+      "loss": 0.2372,
+      "step": 1554
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.1382599113207952,
+      "learning_rate": 9.91662848937144e-06,
+      "loss": 0.1926,
+      "step": 1555
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.1966329306135752,
+      "learning_rate": 9.916464588804308e-06,
+      "loss": 0.2646,
+      "step": 1556
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.8538996958787552,
+      "learning_rate": 9.91630052864591e-06,
+      "loss": 0.2801,
+      "step": 1557
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.715895620179582,
+      "learning_rate": 9.916136308901569e-06,
+      "loss": 0.2105,
+      "step": 1558
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.1777713589531162,
+      "learning_rate": 9.915971929576615e-06,
+      "loss": 0.1825,
+      "step": 1559
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.528773954663678,
+      "learning_rate": 9.915807390676386e-06,
+      "loss": 0.1919,
+      "step": 1560
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.0058661229175772,
+      "learning_rate": 9.915642692206223e-06,
+      "loss": 0.1826,
+      "step": 1561
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.0919220704172021,
+      "learning_rate": 9.91547783417147e-06,
+      "loss": 0.2043,
+      "step": 1562
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.2250835224794383,
+      "learning_rate": 9.915312816577482e-06,
+      "loss": 0.2394,
+      "step": 1563
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.4504462657563646,
+      "learning_rate": 9.915147639429612e-06,
+      "loss": 0.2492,
+      "step": 1564
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.4227609860842831,
+      "learning_rate": 9.914982302733224e-06,
+      "loss": 0.2353,
+      "step": 1565
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.1837313465045478,
+      "learning_rate": 9.914816806493684e-06,
+      "loss": 0.2477,
+      "step": 1566
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.5693781929522945,
+      "learning_rate": 9.914651150716363e-06,
+      "loss": 0.1849,
+      "step": 1567
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 2.1608239947427523,
+      "learning_rate": 9.914485335406641e-06,
+      "loss": 0.2458,
+      "step": 1568
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.0808538682836635,
+      "learning_rate": 9.914319360569897e-06,
+      "loss": 0.1992,
+      "step": 1569
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.0453176372840898,
+      "learning_rate": 9.914153226211523e-06,
+      "loss": 0.2071,
+      "step": 1570
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.5249636732722944,
+      "learning_rate": 9.913986932336909e-06,
+      "loss": 0.1976,
+      "step": 1571
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.4217518029970395,
+      "learning_rate": 9.913820478951452e-06,
+      "loss": 0.2456,
+      "step": 1572
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.8945467368210644,
+      "learning_rate": 9.913653866060558e-06,
+      "loss": 0.254,
+      "step": 1573
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.5048231203393894,
+      "learning_rate": 9.913487093669633e-06,
+      "loss": 0.2398,
+      "step": 1574
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.2836737763945232,
+      "learning_rate": 9.913320161784093e-06,
+      "loss": 0.1973,
+      "step": 1575
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.0769966239847508,
+      "learning_rate": 9.913153070409355e-06,
+      "loss": 0.1876,
+      "step": 1576
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.1921483175593324,
+      "learning_rate": 9.91298581955084e-06,
+      "loss": 0.2168,
+      "step": 1577
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.3395543800812357,
+      "learning_rate": 9.912818409213982e-06,
+      "loss": 0.2104,
+      "step": 1578
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.471638649822063,
+      "learning_rate": 9.912650839404213e-06,
+      "loss": 0.2377,
+      "step": 1579
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.3044254131107167,
+      "learning_rate": 9.912483110126976e-06,
+      "loss": 0.1476,
+      "step": 1580
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.427648573044106,
+      "learning_rate": 9.91231522138771e-06,
+      "loss": 0.2633,
+      "step": 1581
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.3047241345704457,
+      "learning_rate": 9.912147173191866e-06,
+      "loss": 0.1876,
+      "step": 1582
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.3327436304131612,
+      "learning_rate": 9.911978965544902e-06,
+      "loss": 0.1841,
+      "step": 1583
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.4994599045521857,
+      "learning_rate": 9.911810598452273e-06,
+      "loss": 0.198,
+      "step": 1584
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.2977323742883762,
+      "learning_rate": 9.911642071919452e-06,
+      "loss": 0.2114,
+      "step": 1585
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.6543989245747812,
+      "learning_rate": 9.911473385951902e-06,
+      "loss": 0.2011,
+      "step": 1586
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.261435785539633,
+      "learning_rate": 9.911304540555102e-06,
+      "loss": 0.2099,
+      "step": 1587
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.7162252609403654,
+      "learning_rate": 9.911135535734532e-06,
+      "loss": 0.2519,
+      "step": 1588
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.9884412978928792,
+      "learning_rate": 9.91096637149568e-06,
+      "loss": 0.1788,
+      "step": 1589
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.5732535581456069,
+      "learning_rate": 9.910797047844034e-06,
+      "loss": 0.2444,
+      "step": 1590
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.5866171879009165,
+      "learning_rate": 9.910627564785093e-06,
+      "loss": 0.2478,
+      "step": 1591
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.190539027223919,
+      "learning_rate": 9.910457922324358e-06,
+      "loss": 0.226,
+      "step": 1592
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.457318244466312,
+      "learning_rate": 9.910288120467333e-06,
+      "loss": 0.2694,
+      "step": 1593
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.5193782572474022,
+      "learning_rate": 9.910118159219532e-06,
+      "loss": 0.2966,
+      "step": 1594
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.0236178632320416,
+      "learning_rate": 9.909948038586475e-06,
+      "loss": 0.1762,
+      "step": 1595
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.0591613560433264,
+      "learning_rate": 9.909777758573678e-06,
+      "loss": 0.1945,
+      "step": 1596
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.2645060116605937,
+      "learning_rate": 9.909607319186673e-06,
+      "loss": 0.2217,
+      "step": 1597
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.3696109410159985,
+      "learning_rate": 9.909436720430992e-06,
+      "loss": 0.2172,
+      "step": 1598
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.1822294911730062,
+      "learning_rate": 9.90926596231217e-06,
+      "loss": 0.2295,
+      "step": 1599
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.2366571937611224,
+      "learning_rate": 9.909095044835755e-06,
+      "loss": 0.2453,
+      "step": 1600
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.4107092007040734,
+      "learning_rate": 9.90892396800729e-06,
+      "loss": 0.2044,
+      "step": 1601
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.2197687341807661,
+      "learning_rate": 9.908752731832329e-06,
+      "loss": 0.1984,
+      "step": 1602
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.2063565698796312,
+      "learning_rate": 9.908581336316432e-06,
+      "loss": 0.19,
+      "step": 1603
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.139865895357509,
+      "learning_rate": 9.908409781465163e-06,
+      "loss": 0.235,
+      "step": 1604
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.7908886039727343,
+      "learning_rate": 9.908238067284091e-06,
+      "loss": 0.2004,
+      "step": 1605
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.3019005431777204,
+      "learning_rate": 9.908066193778786e-06,
+      "loss": 0.215,
+      "step": 1606
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.6281935564368304,
+      "learning_rate": 9.907894160954833e-06,
+      "loss": 0.2347,
+      "step": 1607
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.4064567202049925,
+      "learning_rate": 9.907721968817812e-06,
+      "loss": 0.2747,
+      "step": 1608
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.502611155028155,
+      "learning_rate": 9.907549617373313e-06,
+      "loss": 0.2013,
+      "step": 1609
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.4854692616877623,
+      "learning_rate": 9.907377106626931e-06,
+      "loss": 0.202,
+      "step": 1610
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.4386679579644968,
+      "learning_rate": 9.907204436584269e-06,
+      "loss": 0.289,
+      "step": 1611
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.5741035606230924,
+      "learning_rate": 9.907031607250927e-06,
+      "loss": 0.2313,
+      "step": 1612
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.3988851624561889,
+      "learning_rate": 9.906858618632517e-06,
+      "loss": 0.1904,
+      "step": 1613
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.143613359361613,
+      "learning_rate": 9.906685470734656e-06,
+      "loss": 0.2136,
+      "step": 1614
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.4740058737648951,
+      "learning_rate": 9.906512163562962e-06,
+      "loss": 0.2315,
+      "step": 1615
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.9284956368624752,
+      "learning_rate": 9.906338697123063e-06,
+      "loss": 0.183,
+      "step": 1616
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.2263777010497836,
+      "learning_rate": 9.906165071420588e-06,
+      "loss": 0.1757,
+      "step": 1617
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.1336272976102564,
+      "learning_rate": 9.905991286461173e-06,
+      "loss": 0.1846,
+      "step": 1618
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.4775165119958964,
+      "learning_rate": 9.90581734225046e-06,
+      "loss": 0.2429,
+      "step": 1619
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.5337520430709828,
+      "learning_rate": 9.905643238794093e-06,
+      "loss": 0.2337,
+      "step": 1620
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.1734260213115257,
+      "learning_rate": 9.90546897609773e-06,
+      "loss": 0.1756,
+      "step": 1621
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.5767993121318982,
+      "learning_rate": 9.90529455416702e-06,
+      "loss": 0.2129,
+      "step": 1622
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.3378911791055466,
+      "learning_rate": 9.905119973007629e-06,
+      "loss": 0.2106,
+      "step": 1623
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.8840323202976306,
+      "learning_rate": 9.904945232625223e-06,
+      "loss": 0.125,
+      "step": 1624
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.2393843133088807,
+      "learning_rate": 9.904770333025472e-06,
+      "loss": 0.2145,
+      "step": 1625
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.2816503817981297,
+      "learning_rate": 9.904595274214058e-06,
+      "loss": 0.2465,
+      "step": 1626
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.3770327876428807,
+      "learning_rate": 9.904420056196661e-06,
+      "loss": 0.1614,
+      "step": 1627
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.4846644721616993,
+      "learning_rate": 9.904244678978969e-06,
+      "loss": 0.2168,
+      "step": 1628
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.580759208169585,
+      "learning_rate": 9.904069142566673e-06,
+      "loss": 0.1792,
+      "step": 1629
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.5428605778114428,
+      "learning_rate": 9.903893446965474e-06,
+      "loss": 0.2386,
+      "step": 1630
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.8462390989344641,
+      "learning_rate": 9.903717592181072e-06,
+      "loss": 0.2143,
+      "step": 1631
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.5302870797898451,
+      "learning_rate": 9.903541578219177e-06,
+      "loss": 0.2436,
+      "step": 1632
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.093626895856789,
+      "learning_rate": 9.903365405085506e-06,
+      "loss": 0.1715,
+      "step": 1633
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.5850972453679864,
+      "learning_rate": 9.903189072785772e-06,
+      "loss": 0.2854,
+      "step": 1634
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.4067587885740565,
+      "learning_rate": 9.903012581325702e-06,
+      "loss": 0.248,
+      "step": 1635
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.2928478466796218,
+      "learning_rate": 9.902835930711024e-06,
+      "loss": 0.2362,
+      "step": 1636
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.1143150014017473,
+      "learning_rate": 9.90265912094747e-06,
+      "loss": 0.2542,
+      "step": 1637
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.442235754493191,
+      "learning_rate": 9.902482152040784e-06,
+      "loss": 0.2018,
+      "step": 1638
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.6810589314838535,
+      "learning_rate": 9.902305023996708e-06,
+      "loss": 0.2731,
+      "step": 1639
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.4264891970165567,
+      "learning_rate": 9.902127736820993e-06,
+      "loss": 0.2249,
+      "step": 1640
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.6712761635397917,
+      "learning_rate": 9.901950290519391e-06,
+      "loss": 0.2085,
+      "step": 1641
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.0403794926930563,
+      "learning_rate": 9.901772685097663e-06,
+      "loss": 0.1787,
+      "step": 1642
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.0712339826975497,
+      "learning_rate": 9.901594920561577e-06,
+      "loss": 0.1991,
+      "step": 1643
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.4143429999581194,
+      "learning_rate": 9.9014169969169e-06,
+      "loss": 0.233,
+      "step": 1644
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.1515479686032428,
+      "learning_rate": 9.901238914169409e-06,
+      "loss": 0.1807,
+      "step": 1645
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 2.0659715108620906,
+      "learning_rate": 9.901060672324885e-06,
+      "loss": 0.2182,
+      "step": 1646
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.5988116481816466,
+      "learning_rate": 9.900882271389113e-06,
+      "loss": 0.2169,
+      "step": 1647
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.396532554062132,
+      "learning_rate": 9.900703711367885e-06,
+      "loss": 0.2772,
+      "step": 1648
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.087201240761928,
+      "learning_rate": 9.900524992266994e-06,
+      "loss": 0.1924,
+      "step": 1649
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.2061869753360521,
+      "learning_rate": 9.900346114092245e-06,
+      "loss": 0.2235,
+      "step": 1650
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.627878788985903,
+      "learning_rate": 9.900167076849445e-06,
+      "loss": 0.2123,
+      "step": 1651
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.8123799748800036,
+      "learning_rate": 9.899987880544401e-06,
+      "loss": 0.2357,
+      "step": 1652
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.414946345733292,
+      "learning_rate": 9.899808525182935e-06,
+      "loss": 0.1928,
+      "step": 1653
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.1616386694458425,
+      "learning_rate": 9.899629010770866e-06,
+      "loss": 0.2191,
+      "step": 1654
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.0976989303382094,
+      "learning_rate": 9.899449337314022e-06,
+      "loss": 0.1945,
+      "step": 1655
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.8792491151921282,
+      "learning_rate": 9.899269504818234e-06,
+      "loss": 0.2686,
+      "step": 1656
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.597729289523498,
+      "learning_rate": 9.899089513289342e-06,
+      "loss": 0.227,
+      "step": 1657
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.7081946635565561,
+      "learning_rate": 9.898909362733187e-06,
+      "loss": 0.2596,
+      "step": 1658
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.9249918675870155,
+      "learning_rate": 9.898729053155615e-06,
+      "loss": 0.2815,
+      "step": 1659
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.1303831178095678,
+      "learning_rate": 9.898548584562483e-06,
+      "loss": 0.2049,
+      "step": 1660
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.0012268348995361,
+      "learning_rate": 9.898367956959646e-06,
+      "loss": 0.2391,
+      "step": 1661
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.9500336935944784,
+      "learning_rate": 9.89818717035297e-06,
+      "loss": 0.2467,
+      "step": 1662
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.8698905281410947,
+      "learning_rate": 9.89800622474832e-06,
+      "loss": 0.1679,
+      "step": 1663
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.911193095397332,
+      "learning_rate": 9.897825120151573e-06,
+      "loss": 0.1738,
+      "step": 1664
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.1126493886102076,
+      "learning_rate": 9.897643856568604e-06,
+      "loss": 0.1863,
+      "step": 1665
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.5681700823838538,
+      "learning_rate": 9.897462434005302e-06,
+      "loss": 0.2383,
+      "step": 1666
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.2465476357436147,
+      "learning_rate": 9.89728085246755e-06,
+      "loss": 0.1816,
+      "step": 1667
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.1137406840052522,
+      "learning_rate": 9.897099111961248e-06,
+      "loss": 0.2306,
+      "step": 1668
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.483630719058229,
+      "learning_rate": 9.896917212492292e-06,
+      "loss": 0.1914,
+      "step": 1669
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.4074615822303802,
+      "learning_rate": 9.896735154066586e-06,
+      "loss": 0.26,
+      "step": 1670
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.3559745345120964,
+      "learning_rate": 9.896552936690043e-06,
+      "loss": 0.2089,
+      "step": 1671
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.5126464151220935,
+      "learning_rate": 9.896370560368575e-06,
+      "loss": 0.1736,
+      "step": 1672
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 3.3059613813970503,
+      "learning_rate": 9.896188025108102e-06,
+      "loss": 0.2297,
+      "step": 1673
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.0390121799130037,
+      "learning_rate": 9.896005330914553e-06,
+      "loss": 0.1659,
+      "step": 1674
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.815547732023845,
+      "learning_rate": 9.895822477793853e-06,
+      "loss": 0.272,
+      "step": 1675
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.4900571760265948,
+      "learning_rate": 9.895639465751941e-06,
+      "loss": 0.2377,
+      "step": 1676
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.297702904217964,
+      "learning_rate": 9.895456294794757e-06,
+      "loss": 0.2599,
+      "step": 1677
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.2733904765521593,
+      "learning_rate": 9.895272964928246e-06,
+      "loss": 0.2078,
+      "step": 1678
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.2375022178025554,
+      "learning_rate": 9.89508947615836e-06,
+      "loss": 0.2336,
+      "step": 1679
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.7160843031122286,
+      "learning_rate": 9.894905828491055e-06,
+      "loss": 0.2145,
+      "step": 1680
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.3465409653482754,
+      "learning_rate": 9.89472202193229e-06,
+      "loss": 0.1528,
+      "step": 1681
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.507647831478488,
+      "learning_rate": 9.894538056488037e-06,
+      "loss": 0.1814,
+      "step": 1682
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.0484543832394548,
+      "learning_rate": 9.89435393216426e-06,
+      "loss": 0.2393,
+      "step": 1683
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.305269141663191,
+      "learning_rate": 9.894169648966943e-06,
+      "loss": 0.2167,
+      "step": 1684
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.9097603900950448,
+      "learning_rate": 9.893985206902064e-06,
+      "loss": 0.1978,
+      "step": 1685
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.4882844151440702,
+      "learning_rate": 9.89380060597561e-06,
+      "loss": 0.2315,
+      "step": 1686
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.971012526696035,
+      "learning_rate": 9.893615846193576e-06,
+      "loss": 0.13,
+      "step": 1687
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.6274726560232238,
+      "learning_rate": 9.893430927561958e-06,
+      "loss": 0.2557,
+      "step": 1688
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.1209391145027217,
+      "learning_rate": 9.893245850086755e-06,
+      "loss": 0.2091,
+      "step": 1689
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.1287838100574243,
+      "learning_rate": 9.893060613773981e-06,
+      "loss": 0.2222,
+      "step": 1690
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.2267955084907711,
+      "learning_rate": 9.892875218629644e-06,
+      "loss": 0.2448,
+      "step": 1691
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.9048463742633372,
+      "learning_rate": 9.892689664659765e-06,
+      "loss": 0.2399,
+      "step": 1692
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.5325019059006009,
+      "learning_rate": 9.892503951870365e-06,
+      "loss": 0.2517,
+      "step": 1693
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.3756049678493016,
+      "learning_rate": 9.892318080267474e-06,
+      "loss": 0.2607,
+      "step": 1694
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.7999399615740268,
+      "learning_rate": 9.892132049857124e-06,
+      "loss": 0.157,
+      "step": 1695
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.4620180061199533,
+      "learning_rate": 9.891945860645355e-06,
+      "loss": 0.2851,
+      "step": 1696
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.7518374299597974,
+      "learning_rate": 9.891759512638211e-06,
+      "loss": 0.2487,
+      "step": 1697
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.5261192169165805,
+      "learning_rate": 9.891573005841739e-06,
+      "loss": 0.1691,
+      "step": 1698
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.076936064506501,
+      "learning_rate": 9.891386340261995e-06,
+      "loss": 0.2156,
+      "step": 1699
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.138729185621811,
+      "learning_rate": 9.891199515905037e-06,
+      "loss": 0.1962,
+      "step": 1700
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.5524818370992117,
+      "learning_rate": 9.89101253277693e-06,
+      "loss": 0.2314,
+      "step": 1701
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.367547952275,
+      "learning_rate": 9.890825390883745e-06,
+      "loss": 0.2438,
+      "step": 1702
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.2676099147248858,
+      "learning_rate": 9.890638090231554e-06,
+      "loss": 0.2425,
+      "step": 1703
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.3004314865952842,
+      "learning_rate": 9.890450630826438e-06,
+      "loss": 0.1594,
+      "step": 1704
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.3047512061732416,
+      "learning_rate": 9.890263012674484e-06,
+      "loss": 0.2663,
+      "step": 1705
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.5853065941149977,
+      "learning_rate": 9.89007523578178e-06,
+      "loss": 0.2298,
+      "step": 1706
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.1549666731844201,
+      "learning_rate": 9.889887300154422e-06,
+      "loss": 0.2022,
+      "step": 1707
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.4216960228422901,
+      "learning_rate": 9.88969920579851e-06,
+      "loss": 0.1929,
+      "step": 1708
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.657250495787279,
+      "learning_rate": 9.88951095272015e-06,
+      "loss": 0.2337,
+      "step": 1709
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.4961882927406989,
+      "learning_rate": 9.889322540925452e-06,
+      "loss": 0.2038,
+      "step": 1710
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.2265796720918478,
+      "learning_rate": 9.889133970420535e-06,
+      "loss": 0.2238,
+      "step": 1711
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.6795147385576568,
+      "learning_rate": 9.888945241211516e-06,
+      "loss": 0.2244,
+      "step": 1712
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.405723887945155,
+      "learning_rate": 9.888756353304522e-06,
+      "loss": 0.1915,
+      "step": 1713
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.1675607485353976,
+      "learning_rate": 9.888567306705687e-06,
+      "loss": 0.2364,
+      "step": 1714
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.619588364377495,
+      "learning_rate": 9.888378101421148e-06,
+      "loss": 0.1946,
+      "step": 1715
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.5275045588295768,
+      "learning_rate": 9.888188737457042e-06,
+      "loss": 0.199,
+      "step": 1716
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.1814263673689545,
+      "learning_rate": 9.887999214819519e-06,
+      "loss": 0.2357,
+      "step": 1717
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.8116018602202473,
+      "learning_rate": 9.887809533514732e-06,
+      "loss": 0.2537,
+      "step": 1718
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.3698775760981636,
+      "learning_rate": 9.887619693548836e-06,
+      "loss": 0.2196,
+      "step": 1719
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.521326155999429,
+      "learning_rate": 9.887429694927994e-06,
+      "loss": 0.1627,
+      "step": 1720
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.0299627188074232,
+      "learning_rate": 9.887239537658374e-06,
+      "loss": 0.1839,
+      "step": 1721
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.3013293125527106,
+      "learning_rate": 9.887049221746147e-06,
+      "loss": 0.1927,
+      "step": 1722
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.40548074928153,
+      "learning_rate": 9.886858747197494e-06,
+      "loss": 0.1979,
+      "step": 1723
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.2232587627741776,
+      "learning_rate": 9.886668114018596e-06,
+      "loss": 0.2364,
+      "step": 1724
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.2906507749847074,
+      "learning_rate": 9.88647732221564e-06,
+      "loss": 0.2168,
+      "step": 1725
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.3457798015525126,
+      "learning_rate": 9.88628637179482e-06,
+      "loss": 0.2458,
+      "step": 1726
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.4526520107971028,
+      "learning_rate": 9.886095262762336e-06,
+      "loss": 0.271,
+      "step": 1727
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.2512026090476662,
+      "learning_rate": 9.88590399512439e-06,
+      "loss": 0.179,
+      "step": 1728
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.9293594623092893,
+      "learning_rate": 9.885712568887187e-06,
+      "loss": 0.1921,
+      "step": 1729
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.219580526291137,
+      "learning_rate": 9.885520984056947e-06,
+      "loss": 0.2472,
+      "step": 1730
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.2555131153084786,
+      "learning_rate": 9.885329240639888e-06,
+      "loss": 0.2097,
+      "step": 1731
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.5204600263570143,
+      "learning_rate": 9.88513733864223e-06,
+      "loss": 0.2493,
+      "step": 1732
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.9361514274251053,
+      "learning_rate": 9.884945278070206e-06,
+      "loss": 0.2668,
+      "step": 1733
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.200490708776527,
+      "learning_rate": 9.88475305893005e-06,
+      "loss": 0.2771,
+      "step": 1734
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.3088065308903862,
+      "learning_rate": 9.884560681228e-06,
+      "loss": 0.2284,
+      "step": 1735
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.224862329619738,
+      "learning_rate": 9.884368144970303e-06,
+      "loss": 0.213,
+      "step": 1736
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.159766623065194,
+      "learning_rate": 9.884175450163204e-06,
+      "loss": 0.1946,
+      "step": 1737
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.4271338761619043,
+      "learning_rate": 9.883982596812965e-06,
+      "loss": 0.2321,
+      "step": 1738
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.0992048834066395,
+      "learning_rate": 9.883789584925839e-06,
+      "loss": 0.2548,
+      "step": 1739
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.5688228539917384,
+      "learning_rate": 9.883596414508096e-06,
+      "loss": 0.227,
+      "step": 1740
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.3577833696012682,
+      "learning_rate": 9.883403085566005e-06,
+      "loss": 0.2092,
+      "step": 1741
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.1127769164416903,
+      "learning_rate": 9.883209598105843e-06,
+      "loss": 0.1862,
+      "step": 1742
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.984370040502619,
+      "learning_rate": 9.883015952133887e-06,
+      "loss": 0.2001,
+      "step": 1743
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.4974418161767777,
+      "learning_rate": 9.882822147656428e-06,
+      "loss": 0.2434,
+      "step": 1744
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.244435435628048,
+      "learning_rate": 9.882628184679753e-06,
+      "loss": 0.2305,
+      "step": 1745
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.1194184958597817,
+      "learning_rate": 9.88243406321016e-06,
+      "loss": 0.2114,
+      "step": 1746
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.0546898841830963,
+      "learning_rate": 9.882239783253948e-06,
+      "loss": 0.1551,
+      "step": 1747
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.5371146098996586,
+      "learning_rate": 9.882045344817425e-06,
+      "loss": 0.2724,
+      "step": 1748
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.5150029292094047,
+      "learning_rate": 9.881850747906905e-06,
+      "loss": 0.1952,
+      "step": 1749
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.3067918584456732,
+      "learning_rate": 9.881655992528702e-06,
+      "loss": 0.21,
+      "step": 1750
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.3188195407416994,
+      "learning_rate": 9.881461078689139e-06,
+      "loss": 0.2286,
+      "step": 1751
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.4530329431579367,
+      "learning_rate": 9.88126600639454e-06,
+      "loss": 0.2016,
+      "step": 1752
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.1845725327586527,
+      "learning_rate": 9.88107077565124e-06,
+      "loss": 0.1918,
+      "step": 1753
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.8732821256394637,
+      "learning_rate": 9.880875386465576e-06,
+      "loss": 0.1722,
+      "step": 1754
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.2781801809929834,
+      "learning_rate": 9.880679838843891e-06,
+      "loss": 0.2058,
+      "step": 1755
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.3741762202384582,
+      "learning_rate": 9.880484132792532e-06,
+      "loss": 0.2064,
+      "step": 1756
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.95563193024941,
+      "learning_rate": 9.88028826831785e-06,
+      "loss": 0.2549,
+      "step": 1757
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.0423602729491916,
+      "learning_rate": 9.880092245426208e-06,
+      "loss": 0.2642,
+      "step": 1758
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.3034312873400522,
+      "learning_rate": 9.879896064123961e-06,
+      "loss": 0.2132,
+      "step": 1759
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.2906421306508868,
+      "learning_rate": 9.879699724417485e-06,
+      "loss": 0.2604,
+      "step": 1760
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.4308116459156384,
+      "learning_rate": 9.879503226313147e-06,
+      "loss": 0.2209,
+      "step": 1761
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.5330362046852186,
+      "learning_rate": 9.879306569817331e-06,
+      "loss": 0.2551,
+      "step": 1762
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.6211711416959333,
+      "learning_rate": 9.879109754936415e-06,
+      "loss": 0.217,
+      "step": 1763
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.0839489561600715,
+      "learning_rate": 9.878912781676794e-06,
+      "loss": 0.2253,
+      "step": 1764
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.0116303103318307,
+      "learning_rate": 9.878715650044856e-06,
+      "loss": 0.2054,
+      "step": 1765
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.3992526845794648,
+      "learning_rate": 9.878518360047005e-06,
+      "loss": 0.2651,
+      "step": 1766
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.0755684159339116,
+      "learning_rate": 9.87832091168964e-06,
+      "loss": 0.2672,
+      "step": 1767
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.3751576655657667,
+      "learning_rate": 9.878123304979175e-06,
+      "loss": 0.1764,
+      "step": 1768
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.522887614851957,
+      "learning_rate": 9.877925539922022e-06,
+      "loss": 0.2537,
+      "step": 1769
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.1498699349697856,
+      "learning_rate": 9.8777276165246e-06,
+      "loss": 0.181,
+      "step": 1770
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.0883288844761507,
+      "learning_rate": 9.877529534793336e-06,
+      "loss": 0.2475,
+      "step": 1771
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.42443348138927,
+      "learning_rate": 9.877331294734656e-06,
+      "loss": 0.2113,
+      "step": 1772
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.1376760996886681,
+      "learning_rate": 9.877132896355e-06,
+      "loss": 0.199,
+      "step": 1773
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.2399717833905057,
+      "learning_rate": 9.876934339660805e-06,
+      "loss": 0.2053,
+      "step": 1774
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.8961566303094868,
+      "learning_rate": 9.876735624658517e-06,
+      "loss": 0.2343,
+      "step": 1775
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.1639873757219033,
+      "learning_rate": 9.876536751354587e-06,
+      "loss": 0.2263,
+      "step": 1776
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.3976387881088541,
+      "learning_rate": 9.876337719755468e-06,
+      "loss": 0.255,
+      "step": 1777
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.9979445550579843,
+      "learning_rate": 9.876138529867622e-06,
+      "loss": 0.1797,
+      "step": 1778
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.0902363837280684,
+      "learning_rate": 9.875939181697518e-06,
+      "loss": 0.2379,
+      "step": 1779
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.0510436186057164,
+      "learning_rate": 9.875739675251624e-06,
+      "loss": 0.1945,
+      "step": 1780
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.0385731032149867,
+      "learning_rate": 9.875540010536416e-06,
+      "loss": 0.1674,
+      "step": 1781
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.5745053468227608,
+      "learning_rate": 9.875340187558373e-06,
+      "loss": 0.2521,
+      "step": 1782
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.4079134831055498,
+      "learning_rate": 9.875140206323986e-06,
+      "loss": 0.1743,
+      "step": 1783
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.3096037112990082,
+      "learning_rate": 9.874940066839746e-06,
+      "loss": 0.2062,
+      "step": 1784
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.939039492206251,
+      "learning_rate": 9.874739769112146e-06,
+      "loss": 0.2156,
+      "step": 1785
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.3146676700416753,
+      "learning_rate": 9.874539313147691e-06,
+      "loss": 0.2227,
+      "step": 1786
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.377674803758694,
+      "learning_rate": 9.874338698952886e-06,
+      "loss": 0.2398,
+      "step": 1787
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.1611645167251792,
+      "learning_rate": 9.874137926534244e-06,
+      "loss": 0.2162,
+      "step": 1788
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.6639216681332147,
+      "learning_rate": 9.873936995898284e-06,
+      "loss": 0.2896,
+      "step": 1789
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.3963375734649746,
+      "learning_rate": 9.873735907051524e-06,
+      "loss": 0.2492,
+      "step": 1790
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.5431778500511577,
+      "learning_rate": 9.873534660000494e-06,
+      "loss": 0.1855,
+      "step": 1791
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.2781020707473048,
+      "learning_rate": 9.873333254751728e-06,
+      "loss": 0.2171,
+      "step": 1792
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.229866510179902,
+      "learning_rate": 9.873131691311761e-06,
+      "loss": 0.1693,
+      "step": 1793
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.2943029816246796,
+      "learning_rate": 9.872929969687137e-06,
+      "loss": 0.1771,
+      "step": 1794
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.7444459620126302,
+      "learning_rate": 9.872728089884405e-06,
+      "loss": 0.2549,
+      "step": 1795
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.2964388949999073,
+      "learning_rate": 9.872526051910116e-06,
+      "loss": 0.2358,
+      "step": 1796
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.7440555466303383,
+      "learning_rate": 9.87232385577083e-06,
+      "loss": 0.1636,
+      "step": 1797
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.5196112216998843,
+      "learning_rate": 9.872121501473111e-06,
+      "loss": 0.2273,
+      "step": 1798
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.0951274432182925,
+      "learning_rate": 9.871918989023525e-06,
+      "loss": 0.1819,
+      "step": 1799
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.5131995023941762,
+      "learning_rate": 9.871716318428648e-06,
+      "loss": 0.2127,
+      "step": 1800
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.088748654298739,
+      "learning_rate": 9.871513489695057e-06,
+      "loss": 0.2115,
+      "step": 1801
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.0018267198582487,
+      "learning_rate": 9.871310502829336e-06,
+      "loss": 0.2259,
+      "step": 1802
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.284441409981342,
+      "learning_rate": 9.871107357838076e-06,
+      "loss": 0.2144,
+      "step": 1803
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.4463535126409426,
+      "learning_rate": 9.87090405472787e-06,
+      "loss": 0.2564,
+      "step": 1804
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.1196558937017727,
+      "learning_rate": 9.870700593505317e-06,
+      "loss": 0.2313,
+      "step": 1805
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.1038749531779486,
+      "learning_rate": 9.870496974177025e-06,
+      "loss": 0.2335,
+      "step": 1806
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.1654046533042925,
+      "learning_rate": 9.870293196749596e-06,
+      "loss": 0.1998,
+      "step": 1807
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.274693500093655,
+      "learning_rate": 9.870089261229652e-06,
+      "loss": 0.251,
+      "step": 1808
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.0588455323059431,
+      "learning_rate": 9.869885167623808e-06,
+      "loss": 0.207,
+      "step": 1809
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.3084489329471551,
+      "learning_rate": 9.869680915938693e-06,
+      "loss": 0.216,
+      "step": 1810
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.9819041626634492,
+      "learning_rate": 9.869476506180933e-06,
+      "loss": 0.1933,
+      "step": 1811
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.9400550763644402,
+      "learning_rate": 9.869271938357168e-06,
+      "loss": 0.1957,
+      "step": 1812
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.9686712109920115,
+      "learning_rate": 9.869067212474034e-06,
+      "loss": 0.1807,
+      "step": 1813
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.0034934630383907,
+      "learning_rate": 9.868862328538179e-06,
+      "loss": 0.1502,
+      "step": 1814
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.1645535315075273,
+      "learning_rate": 9.868657286556254e-06,
+      "loss": 0.2009,
+      "step": 1815
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.4493560430369399,
+      "learning_rate": 9.868452086534913e-06,
+      "loss": 0.2172,
+      "step": 1816
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.1532627579090646,
+      "learning_rate": 9.868246728480817e-06,
+      "loss": 0.1763,
+      "step": 1817
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.6729530486295232,
+      "learning_rate": 9.868041212400633e-06,
+      "loss": 0.2863,
+      "step": 1818
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.247073344538385,
+      "learning_rate": 9.867835538301032e-06,
+      "loss": 0.2106,
+      "step": 1819
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.9596068834370137,
+      "learning_rate": 9.867629706188691e-06,
+      "loss": 0.1744,
+      "step": 1820
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.5866839268411586,
+      "learning_rate": 9.86742371607029e-06,
+      "loss": 0.1934,
+      "step": 1821
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.0914920541734678,
+      "learning_rate": 9.867217567952516e-06,
+      "loss": 0.2047,
+      "step": 1822
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.1781412310709074,
+      "learning_rate": 9.86701126184206e-06,
+      "loss": 0.2244,
+      "step": 1823
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.317144854569088,
+      "learning_rate": 9.866804797745622e-06,
+      "loss": 0.2597,
+      "step": 1824
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.8944035083925033,
+      "learning_rate": 9.866598175669901e-06,
+      "loss": 0.2493,
+      "step": 1825
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.5419037583973996,
+      "learning_rate": 9.866391395621605e-06,
+      "loss": 0.2327,
+      "step": 1826
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.5131210567800357,
+      "learning_rate": 9.866184457607444e-06,
+      "loss": 0.2014,
+      "step": 1827
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.4022687670125078,
+      "learning_rate": 9.865977361634138e-06,
+      "loss": 0.2134,
+      "step": 1828
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.3296305551785854,
+      "learning_rate": 9.86577010770841e-06,
+      "loss": 0.233,
+      "step": 1829
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.508821726571975,
+      "learning_rate": 9.865562695836985e-06,
+      "loss": 0.2179,
+      "step": 1830
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 2.0565175728787928,
+      "learning_rate": 9.865355126026597e-06,
+      "loss": 0.2349,
+      "step": 1831
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.6533326105154993,
+      "learning_rate": 9.865147398283986e-06,
+      "loss": 0.1953,
+      "step": 1832
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.0449461426100577,
+      "learning_rate": 9.864939512615889e-06,
+      "loss": 0.217,
+      "step": 1833
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.1655482133292208,
+      "learning_rate": 9.86473146902906e-06,
+      "loss": 0.1721,
+      "step": 1834
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.376148839850353,
+      "learning_rate": 9.86452326753025e-06,
+      "loss": 0.2558,
+      "step": 1835
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.180312559497964,
+      "learning_rate": 9.864314908126218e-06,
+      "loss": 0.1946,
+      "step": 1836
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.0796586622494257,
+      "learning_rate": 9.864106390823727e-06,
+      "loss": 0.2584,
+      "step": 1837
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.0194581926412667,
+      "learning_rate": 9.863897715629545e-06,
+      "loss": 0.1738,
+      "step": 1838
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.1214054352267164,
+      "learning_rate": 9.863688882550445e-06,
+      "loss": 0.1857,
+      "step": 1839
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.120715823588555,
+      "learning_rate": 9.86347989159321e-06,
+      "loss": 0.1936,
+      "step": 1840
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.4139887358462642,
+      "learning_rate": 9.863270742764618e-06,
+      "loss": 0.2132,
+      "step": 1841
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.1869956532893258,
+      "learning_rate": 9.863061436071462e-06,
+      "loss": 0.1929,
+      "step": 1842
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.4026395570752448,
+      "learning_rate": 9.862851971520537e-06,
+      "loss": 0.2266,
+      "step": 1843
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.6464617400689423,
+      "learning_rate": 9.862642349118638e-06,
+      "loss": 0.2448,
+      "step": 1844
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.9690436775614729,
+      "learning_rate": 9.862432568872574e-06,
+      "loss": 0.2125,
+      "step": 1845
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.3050609171308658,
+      "learning_rate": 9.862222630789153e-06,
+      "loss": 0.2155,
+      "step": 1846
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.0437748433485783,
+      "learning_rate": 9.862012534875186e-06,
+      "loss": 0.2007,
+      "step": 1847
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.0448321758523125,
+      "learning_rate": 9.861802281137501e-06,
+      "loss": 0.1584,
+      "step": 1848
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.4916074634059155,
+      "learning_rate": 9.861591869582917e-06,
+      "loss": 0.2301,
+      "step": 1849
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.111558256808792,
+      "learning_rate": 9.861381300218263e-06,
+      "loss": 0.1728,
+      "step": 1850
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.1412364483537423,
+      "learning_rate": 9.861170573050379e-06,
+      "loss": 0.2071,
+      "step": 1851
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.3624794755601568,
+      "learning_rate": 9.8609596880861e-06,
+      "loss": 0.2116,
+      "step": 1852
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.939131664976067,
+      "learning_rate": 9.860748645332276e-06,
+      "loss": 0.2221,
+      "step": 1853
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.4005825969893995,
+      "learning_rate": 9.860537444795755e-06,
+      "loss": 0.2006,
+      "step": 1854
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.7355037530164987,
+      "learning_rate": 9.860326086483394e-06,
+      "loss": 0.1956,
+      "step": 1855
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.5330778228679534,
+      "learning_rate": 9.860114570402055e-06,
+      "loss": 0.2336,
+      "step": 1856
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.3501135806201687,
+      "learning_rate": 9.8599028965586e-06,
+      "loss": 0.2075,
+      "step": 1857
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.9995470064062374,
+      "learning_rate": 9.859691064959904e-06,
+      "loss": 0.2164,
+      "step": 1858
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.0269459273273627,
+      "learning_rate": 9.85947907561284e-06,
+      "loss": 0.1917,
+      "step": 1859
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.2376459090775227,
+      "learning_rate": 9.859266928524292e-06,
+      "loss": 0.1896,
+      "step": 1860
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.3776261274417783,
+      "learning_rate": 9.859054623701145e-06,
+      "loss": 0.2343,
+      "step": 1861
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.296597378527818,
+      "learning_rate": 9.858842161150291e-06,
+      "loss": 0.193,
+      "step": 1862
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.4691021286622696,
+      "learning_rate": 9.858629540878626e-06,
+      "loss": 0.2395,
+      "step": 1863
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 2.0357017902493304,
+      "learning_rate": 9.858416762893052e-06,
+      "loss": 0.2087,
+      "step": 1864
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.3325144822530681,
+      "learning_rate": 9.858203827200477e-06,
+      "loss": 0.2301,
+      "step": 1865
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.0480621273865995,
+      "learning_rate": 9.857990733807809e-06,
+      "loss": 0.1832,
+      "step": 1866
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.5861577411933727,
+      "learning_rate": 9.857777482721972e-06,
+      "loss": 0.2615,
+      "step": 1867
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.9420566172181066,
+      "learning_rate": 9.857564073949883e-06,
+      "loss": 0.1975,
+      "step": 1868
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.327975463863681,
+      "learning_rate": 9.85735050749847e-06,
+      "loss": 0.2431,
+      "step": 1869
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.6140036431888023,
+      "learning_rate": 9.857136783374668e-06,
+      "loss": 0.2209,
+      "step": 1870
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.5449447340787226,
+      "learning_rate": 9.856922901585414e-06,
+      "loss": 0.2232,
+      "step": 1871
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.0703060017687818,
+      "learning_rate": 9.856708862137646e-06,
+      "loss": 0.1865,
+      "step": 1872
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.1626042778791055,
+      "learning_rate": 9.856494665038318e-06,
+      "loss": 0.2048,
+      "step": 1873
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.54499633937544,
+      "learning_rate": 9.856280310294379e-06,
+      "loss": 0.2195,
+      "step": 1874
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.1870469245413027,
+      "learning_rate": 9.85606579791279e-06,
+      "loss": 0.2598,
+      "step": 1875
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 2.1944818180827697,
+      "learning_rate": 9.855851127900512e-06,
+      "loss": 0.2816,
+      "step": 1876
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.7485332302582504,
+      "learning_rate": 9.855636300264516e-06,
+      "loss": 0.242,
+      "step": 1877
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.5623243590817903,
+      "learning_rate": 9.855421315011771e-06,
+      "loss": 0.2461,
+      "step": 1878
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.0685394177246128,
+      "learning_rate": 9.855206172149258e-06,
+      "loss": 0.204,
+      "step": 1879
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.9405179878524462,
+      "learning_rate": 9.854990871683963e-06,
+      "loss": 0.2182,
+      "step": 1880
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.8846817566319038,
+      "learning_rate": 9.85477541362287e-06,
+      "loss": 0.1692,
+      "step": 1881
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.053808177742528,
+      "learning_rate": 9.854559797972975e-06,
+      "loss": 0.208,
+      "step": 1882
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.3788625721051944,
+      "learning_rate": 9.85434402474128e-06,
+      "loss": 0.2116,
+      "step": 1883
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.3298770350614926,
+      "learning_rate": 9.854128093934784e-06,
+      "loss": 0.1557,
+      "step": 1884
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.299239134470896,
+      "learning_rate": 9.853912005560502e-06,
+      "loss": 0.2601,
+      "step": 1885
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.0430332406782732,
+      "learning_rate": 9.853695759625444e-06,
+      "loss": 0.2169,
+      "step": 1886
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.2933279616061635,
+      "learning_rate": 9.853479356136628e-06,
+      "loss": 0.2602,
+      "step": 1887
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.3197464011478004,
+      "learning_rate": 9.853262795101083e-06,
+      "loss": 0.1549,
+      "step": 1888
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.3637877878619036,
+      "learning_rate": 9.853046076525837e-06,
+      "loss": 0.2097,
+      "step": 1889
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.7298056836919289,
+      "learning_rate": 9.852829200417925e-06,
+      "loss": 0.2398,
+      "step": 1890
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.1951925896979718,
+      "learning_rate": 9.852612166784386e-06,
+      "loss": 0.2331,
+      "step": 1891
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.6479087067322198,
+      "learning_rate": 9.852394975632265e-06,
+      "loss": 0.2239,
+      "step": 1892
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.4448570255475326,
+      "learning_rate": 9.852177626968613e-06,
+      "loss": 0.2047,
+      "step": 1893
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.4861920052555586,
+      "learning_rate": 9.851960120800485e-06,
+      "loss": 0.2023,
+      "step": 1894
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.351869019165777,
+      "learning_rate": 9.851742457134943e-06,
+      "loss": 0.2107,
+      "step": 1895
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.4347973183658185,
+      "learning_rate": 9.85152463597905e-06,
+      "loss": 0.1982,
+      "step": 1896
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.0426406415688283,
+      "learning_rate": 9.851306657339876e-06,
+      "loss": 0.1689,
+      "step": 1897
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.7343024392242576,
+      "learning_rate": 9.8510885212245e-06,
+      "loss": 0.2312,
+      "step": 1898
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.4123525027681112,
+      "learning_rate": 9.850870227640002e-06,
+      "loss": 0.252,
+      "step": 1899
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.5648094276324622,
+      "learning_rate": 9.850651776593463e-06,
+      "loss": 0.2781,
+      "step": 1900
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.5863766696380999,
+      "learning_rate": 9.850433168091982e-06,
+      "loss": 0.2512,
+      "step": 1901
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.2601913161699356,
+      "learning_rate": 9.85021440214265e-06,
+      "loss": 0.226,
+      "step": 1902
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.3105784155152902,
+      "learning_rate": 9.849995478752568e-06,
+      "loss": 0.2465,
+      "step": 1903
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.4127698016111332,
+      "learning_rate": 9.849776397928844e-06,
+      "loss": 0.2337,
+      "step": 1904
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.603588557936979,
+      "learning_rate": 9.849557159678591e-06,
+      "loss": 0.2082,
+      "step": 1905
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.5938658999393518,
+      "learning_rate": 9.849337764008924e-06,
+      "loss": 0.2143,
+      "step": 1906
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.22503761852901,
+      "learning_rate": 9.849118210926962e-06,
+      "loss": 0.1421,
+      "step": 1907
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.339214588728953,
+      "learning_rate": 9.848898500439836e-06,
+      "loss": 0.1367,
+      "step": 1908
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.1955922771717884,
+      "learning_rate": 9.848678632554677e-06,
+      "loss": 0.1961,
+      "step": 1909
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.1463416862075608,
+      "learning_rate": 9.848458607278622e-06,
+      "loss": 0.1778,
+      "step": 1910
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.5795151287491627,
+      "learning_rate": 9.84823842461881e-06,
+      "loss": 0.2266,
+      "step": 1911
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.3296751694718294,
+      "learning_rate": 9.848018084582393e-06,
+      "loss": 0.2534,
+      "step": 1912
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.5837973191411479,
+      "learning_rate": 9.84779758717652e-06,
+      "loss": 0.2529,
+      "step": 1913
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.0974449364968384,
+      "learning_rate": 9.847576932408349e-06,
+      "loss": 0.2433,
+      "step": 1914
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.3946143881541164,
+      "learning_rate": 9.847356120285046e-06,
+      "loss": 0.2777,
+      "step": 1915
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.3358168519626126,
+      "learning_rate": 9.847135150813774e-06,
+      "loss": 0.1957,
+      "step": 1916
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.2750289175521037,
+      "learning_rate": 9.846914024001706e-06,
+      "loss": 0.1918,
+      "step": 1917
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.397581355186976,
+      "learning_rate": 9.846692739856023e-06,
+      "loss": 0.2094,
+      "step": 1918
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.0131432692994407,
+      "learning_rate": 9.846471298383909e-06,
+      "loss": 0.2038,
+      "step": 1919
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.5412460645589063,
+      "learning_rate": 9.846249699592548e-06,
+      "loss": 0.1781,
+      "step": 1920
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.102300940884773,
+      "learning_rate": 9.846027943489133e-06,
+      "loss": 0.1923,
+      "step": 1921
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.313627220220629,
+      "learning_rate": 9.845806030080867e-06,
+      "loss": 0.2303,
+      "step": 1922
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.2338670307093749,
+      "learning_rate": 9.845583959374949e-06,
+      "loss": 0.2092,
+      "step": 1923
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.6008946794207435,
+      "learning_rate": 9.84536173137859e-06,
+      "loss": 0.2158,
+      "step": 1924
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.1774514587585752,
+      "learning_rate": 9.845139346099002e-06,
+      "loss": 0.1584,
+      "step": 1925
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.9166145075061338,
+      "learning_rate": 9.844916803543408e-06,
+      "loss": 0.2121,
+      "step": 1926
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.3061971560482752,
+      "learning_rate": 9.844694103719025e-06,
+      "loss": 0.1824,
+      "step": 1927
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.9249196997992166,
+      "learning_rate": 9.844471246633086e-06,
+      "loss": 0.29,
+      "step": 1928
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.2781942376780904,
+      "learning_rate": 9.844248232292825e-06,
+      "loss": 0.2534,
+      "step": 1929
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.8539152850072766,
+      "learning_rate": 9.844025060705482e-06,
+      "loss": 0.2397,
+      "step": 1930
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.407938016944049,
+      "learning_rate": 9.843801731878298e-06,
+      "loss": 0.1914,
+      "step": 1931
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.4027932271559747,
+      "learning_rate": 9.843578245818526e-06,
+      "loss": 0.218,
+      "step": 1932
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.0951409385123487,
+      "learning_rate": 9.843354602533416e-06,
+      "loss": 0.2449,
+      "step": 1933
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.9972295755339863,
+      "learning_rate": 9.843130802030234e-06,
+      "loss": 0.2295,
+      "step": 1934
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.91561589334145,
+      "learning_rate": 9.84290684431624e-06,
+      "loss": 0.178,
+      "step": 1935
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.6630989041721733,
+      "learning_rate": 9.842682729398704e-06,
+      "loss": 0.1989,
+      "step": 1936
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.2633477639951622,
+      "learning_rate": 9.842458457284904e-06,
+      "loss": 0.1748,
+      "step": 1937
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.3630747754883066,
+      "learning_rate": 9.842234027982116e-06,
+      "loss": 0.1943,
+      "step": 1938
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.7656724450436007,
+      "learning_rate": 9.842009441497627e-06,
+      "loss": 0.1431,
+      "step": 1939
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.1900157631302897,
+      "learning_rate": 9.841784697838728e-06,
+      "loss": 0.1795,
+      "step": 1940
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.2020147576211626,
+      "learning_rate": 9.841559797012714e-06,
+      "loss": 0.1791,
+      "step": 1941
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.0851325320871283,
+      "learning_rate": 9.841334739026885e-06,
+      "loss": 0.2432,
+      "step": 1942
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.6016433520959308,
+      "learning_rate": 9.841109523888544e-06,
+      "loss": 0.2731,
+      "step": 1943
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.3365090908332455,
+      "learning_rate": 9.840884151605005e-06,
+      "loss": 0.1941,
+      "step": 1944
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.3808891503605345,
+      "learning_rate": 9.840658622183583e-06,
+      "loss": 0.236,
+      "step": 1945
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.0567041755222844,
+      "learning_rate": 9.840432935631598e-06,
+      "loss": 0.16,
+      "step": 1946
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.1730275494548914,
+      "learning_rate": 9.840207091956379e-06,
+      "loss": 0.2447,
+      "step": 1947
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.8927245687722486,
+      "learning_rate": 9.839981091165252e-06,
+      "loss": 0.2153,
+      "step": 1948
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.687404141970708,
+      "learning_rate": 9.839754933265558e-06,
+      "loss": 0.242,
+      "step": 1949
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.4601367168436465,
+      "learning_rate": 9.839528618264633e-06,
+      "loss": 0.1448,
+      "step": 1950
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.2354911808229734,
+      "learning_rate": 9.839302146169827e-06,
+      "loss": 0.2062,
+      "step": 1951
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.5001118762296735,
+      "learning_rate": 9.83907551698849e-06,
+      "loss": 0.2931,
+      "step": 1952
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.3880677042330278,
+      "learning_rate": 9.838848730727981e-06,
+      "loss": 0.2091,
+      "step": 1953
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.328854419453257,
+      "learning_rate": 9.838621787395658e-06,
+      "loss": 0.1829,
+      "step": 1954
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.9671114592788194,
+      "learning_rate": 9.838394686998888e-06,
+      "loss": 0.1934,
+      "step": 1955
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.2636543844625137,
+      "learning_rate": 9.838167429545047e-06,
+      "loss": 0.2305,
+      "step": 1956
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.3206073837982897,
+      "learning_rate": 9.83794001504151e-06,
+      "loss": 0.2136,
+      "step": 1957
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.3154787197476068,
+      "learning_rate": 9.837712443495655e-06,
+      "loss": 0.2456,
+      "step": 1958
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.0925670289071823,
+      "learning_rate": 9.837484714914874e-06,
+      "loss": 0.2091,
+      "step": 1959
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.2199742315431037,
+      "learning_rate": 9.837256829306556e-06,
+      "loss": 0.2403,
+      "step": 1960
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.6676523390925773,
+      "learning_rate": 9.8370287866781e-06,
+      "loss": 0.2389,
+      "step": 1961
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.1382238052838638,
+      "learning_rate": 9.83680058703691e-06,
+      "loss": 0.2083,
+      "step": 1962
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.109865337123217,
+      "learning_rate": 9.836572230390388e-06,
+      "loss": 0.2201,
+      "step": 1963
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.1157440083336647,
+      "learning_rate": 9.836343716745955e-06,
+      "loss": 0.2084,
+      "step": 1964
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.5087413545738497,
+      "learning_rate": 9.83611504611102e-06,
+      "loss": 0.2332,
+      "step": 1965
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.098648436541192,
+      "learning_rate": 9.835886218493011e-06,
+      "loss": 0.1912,
+      "step": 1966
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.0278545531208831,
+      "learning_rate": 9.835657233899353e-06,
+      "loss": 0.1423,
+      "step": 1967
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.2959680114559653,
+      "learning_rate": 9.835428092337483e-06,
+      "loss": 0.2,
+      "step": 1968
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.1182284840512278,
+      "learning_rate": 9.835198793814836e-06,
+      "loss": 0.1948,
+      "step": 1969
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.1449223849276888,
+      "learning_rate": 9.834969338338854e-06,
+      "loss": 0.1438,
+      "step": 1970
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.2009800532934378,
+      "learning_rate": 9.834739725916988e-06,
+      "loss": 0.2407,
+      "step": 1971
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.4993800748615558,
+      "learning_rate": 9.834509956556691e-06,
+      "loss": 0.196,
+      "step": 1972
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.5662077132235972,
+      "learning_rate": 9.834280030265421e-06,
+      "loss": 0.2394,
+      "step": 1973
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.1321581101400022,
+      "learning_rate": 9.83404994705064e-06,
+      "loss": 0.1711,
+      "step": 1974
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.624743196021069,
+      "learning_rate": 9.833819706919818e-06,
+      "loss": 0.2371,
+      "step": 1975
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.2538911864602889,
+      "learning_rate": 9.83358930988043e-06,
+      "loss": 0.1976,
+      "step": 1976
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.0399902887504193,
+      "learning_rate": 9.833358755939953e-06,
+      "loss": 0.2064,
+      "step": 1977
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.1393466916101376,
+      "learning_rate": 9.83312804510587e-06,
+      "loss": 0.2079,
+      "step": 1978
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.026319273094316,
+      "learning_rate": 9.832897177385672e-06,
+      "loss": 0.1481,
+      "step": 1979
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.1416638816616216,
+      "learning_rate": 9.832666152786855e-06,
+      "loss": 0.2018,
+      "step": 1980
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.496635128653734,
+      "learning_rate": 9.832434971316911e-06,
+      "loss": 0.1938,
+      "step": 1981
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.243686991562038,
+      "learning_rate": 9.832203632983353e-06,
+      "loss": 0.1859,
+      "step": 1982
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.270743149608712,
+      "learning_rate": 9.831972137793684e-06,
+      "loss": 0.1813,
+      "step": 1983
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.3135347091113407,
+      "learning_rate": 9.831740485755422e-06,
+      "loss": 0.1873,
+      "step": 1984
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.1485622072306139,
+      "learning_rate": 9.831508676876084e-06,
+      "loss": 0.2399,
+      "step": 1985
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.2433918686532925,
+      "learning_rate": 9.831276711163197e-06,
+      "loss": 0.204,
+      "step": 1986
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.429782410895124,
+      "learning_rate": 9.831044588624293e-06,
+      "loss": 0.2151,
+      "step": 1987
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 2.6404065826192955,
+      "learning_rate": 9.8308123092669e-06,
+      "loss": 0.2673,
+      "step": 1988
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.9300063652287329,
+      "learning_rate": 9.830579873098561e-06,
+      "loss": 0.1652,
+      "step": 1989
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.5730980274872273,
+      "learning_rate": 9.830347280126823e-06,
+      "loss": 0.2261,
+      "step": 1990
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.087153309884529,
+      "learning_rate": 9.830114530359234e-06,
+      "loss": 0.2043,
+      "step": 1991
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 2.0144720254528843,
+      "learning_rate": 9.82988162380335e-06,
+      "loss": 0.1902,
+      "step": 1992
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.411968063206239,
+      "learning_rate": 9.829648560466732e-06,
+      "loss": 0.2247,
+      "step": 1993
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.3235743074179114,
+      "learning_rate": 9.829415340356943e-06,
+      "loss": 0.2147,
+      "step": 1994
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.426146480080312,
+      "learning_rate": 9.829181963481555e-06,
+      "loss": 0.2397,
+      "step": 1995
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.2177117446041812,
+      "learning_rate": 9.828948429848145e-06,
+      "loss": 0.2061,
+      "step": 1996
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.0048223538464576,
+      "learning_rate": 9.82871473946429e-06,
+      "loss": 0.1306,
+      "step": 1997
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.2939252801460095,
+      "learning_rate": 9.82848089233758e-06,
+      "loss": 0.2051,
+      "step": 1998
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.5375439903122352,
+      "learning_rate": 9.828246888475604e-06,
+      "loss": 0.1885,
+      "step": 1999
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.7308755357763106,
+      "learning_rate": 9.828012727885957e-06,
+      "loss": 0.2162,
+      "step": 2000
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.2989618332955144,
+      "learning_rate": 9.827778410576239e-06,
+      "loss": 0.2211,
+      "step": 2001
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.4040835116334207,
+      "learning_rate": 9.827543936554058e-06,
+      "loss": 0.2096,
+      "step": 2002
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.323058563948907,
+      "learning_rate": 9.827309305827025e-06,
+      "loss": 0.2123,
+      "step": 2003
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.352928507680311,
+      "learning_rate": 9.827074518402756e-06,
+      "loss": 0.1714,
+      "step": 2004
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 3.1311720360588575,
+      "learning_rate": 9.826839574288874e-06,
+      "loss": 0.2391,
+      "step": 2005
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.9954902952739995,
+      "learning_rate": 9.826604473493002e-06,
+      "loss": 0.2066,
+      "step": 2006
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.4320520665871646,
+      "learning_rate": 9.826369216022773e-06,
+      "loss": 0.2634,
+      "step": 2007
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.1119465724521775,
+      "learning_rate": 9.826133801885824e-06,
+      "loss": 0.2132,
+      "step": 2008
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.0736407233354994,
+      "learning_rate": 9.825898231089798e-06,
+      "loss": 0.2021,
+      "step": 2009
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.6146722343539737,
+      "learning_rate": 9.825662503642339e-06,
+      "loss": 0.2093,
+      "step": 2010
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.1267549738895328,
+      "learning_rate": 9.825426619551101e-06,
+      "loss": 0.193,
+      "step": 2011
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.0238457349669432,
+      "learning_rate": 9.82519057882374e-06,
+      "loss": 0.2474,
+      "step": 2012
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.0721060385704595,
+      "learning_rate": 9.824954381467917e-06,
+      "loss": 0.238,
+      "step": 2013
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.3632637408096746,
+      "learning_rate": 9.824718027491302e-06,
+      "loss": 0.233,
+      "step": 2014
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.9437925911760011,
+      "learning_rate": 9.824481516901565e-06,
+      "loss": 0.2403,
+      "step": 2015
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.0851838391932191,
+      "learning_rate": 9.824244849706383e-06,
+      "loss": 0.1888,
+      "step": 2016
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.4220137318851493,
+      "learning_rate": 9.824008025913438e-06,
+      "loss": 0.2379,
+      "step": 2017
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.243241775912471,
+      "learning_rate": 9.823771045530421e-06,
+      "loss": 0.229,
+      "step": 2018
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.873910587445407,
+      "learning_rate": 9.823533908565022e-06,
+      "loss": 0.3347,
+      "step": 2019
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.5967414727345393,
+      "learning_rate": 9.823296615024936e-06,
+      "loss": 0.2648,
+      "step": 2020
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.4453344704272635,
+      "learning_rate": 9.82305916491787e-06,
+      "loss": 0.216,
+      "step": 2021
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.0145242301102317,
+      "learning_rate": 9.822821558251529e-06,
+      "loss": 0.2163,
+      "step": 2022
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.2908045109909951,
+      "learning_rate": 9.822583795033628e-06,
+      "loss": 0.2304,
+      "step": 2023
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.0213211541852383,
+      "learning_rate": 9.822345875271884e-06,
+      "loss": 0.2075,
+      "step": 2024
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.9332064385192063,
+      "learning_rate": 9.822107798974019e-06,
+      "loss": 0.2502,
+      "step": 2025
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.1948364921285537,
+      "learning_rate": 9.821869566147762e-06,
+      "loss": 0.2113,
+      "step": 2026
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.1495113607109044,
+      "learning_rate": 9.821631176800847e-06,
+      "loss": 0.2469,
+      "step": 2027
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.7425975787358081,
+      "learning_rate": 9.82139263094101e-06,
+      "loss": 0.2789,
+      "step": 2028
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.0835769832856836,
+      "learning_rate": 9.821153928575997e-06,
+      "loss": 0.2213,
+      "step": 2029
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.3106535180274557,
+      "learning_rate": 9.820915069713555e-06,
+      "loss": 0.2238,
+      "step": 2030
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.4327204484396372,
+      "learning_rate": 9.820676054361436e-06,
+      "loss": 0.1841,
+      "step": 2031
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.104873971952608,
+      "learning_rate": 9.820436882527402e-06,
+      "loss": 0.1719,
+      "step": 2032
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.2083465946089147,
+      "learning_rate": 9.820197554219214e-06,
+      "loss": 0.2146,
+      "step": 2033
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.3779161582352881,
+      "learning_rate": 9.819958069444641e-06,
+      "loss": 0.2142,
+      "step": 2034
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.382926903857554,
+      "learning_rate": 9.819718428211458e-06,
+      "loss": 0.2164,
+      "step": 2035
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.0525437573490848,
+      "learning_rate": 9.819478630527444e-06,
+      "loss": 0.2509,
+      "step": 2036
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.7281124420640992,
+      "learning_rate": 9.819238676400382e-06,
+      "loss": 0.2392,
+      "step": 2037
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.9143589957898055,
+      "learning_rate": 9.81899856583806e-06,
+      "loss": 0.2492,
+      "step": 2038
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.1253177290177623,
+      "learning_rate": 9.818758298848274e-06,
+      "loss": 0.2005,
+      "step": 2039
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.0351009525068295,
+      "learning_rate": 9.818517875438825e-06,
+      "loss": 0.1965,
+      "step": 2040
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.5116130437044522,
+      "learning_rate": 9.818277295617512e-06,
+      "loss": 0.2469,
+      "step": 2041
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.1402293247295325,
+      "learning_rate": 9.818036559392147e-06,
+      "loss": 0.221,
+      "step": 2042
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.4968451525334507,
+      "learning_rate": 9.817795666770547e-06,
+      "loss": 0.2099,
+      "step": 2043
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.28026903211358,
+      "learning_rate": 9.817554617760529e-06,
+      "loss": 0.2096,
+      "step": 2044
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.9241995980490223,
+      "learning_rate": 9.817313412369917e-06,
+      "loss": 0.2327,
+      "step": 2045
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.1796712368987285,
+      "learning_rate": 9.81707205060654e-06,
+      "loss": 0.1981,
+      "step": 2046
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.0054465659716831,
+      "learning_rate": 9.816830532478237e-06,
+      "loss": 0.1983,
+      "step": 2047
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.0005614578347204,
+      "learning_rate": 9.816588857992844e-06,
+      "loss": 0.2442,
+      "step": 2048
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 3.005736508996977,
+      "learning_rate": 9.816347027158205e-06,
+      "loss": 0.3137,
+      "step": 2049
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.4272686554896783,
+      "learning_rate": 9.816105039982174e-06,
+      "loss": 0.236,
+      "step": 2050
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.0055788222960937,
+      "learning_rate": 9.815862896472603e-06,
+      "loss": 0.2435,
+      "step": 2051
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.9164165112438034,
+      "learning_rate": 9.815620596637354e-06,
+      "loss": 0.2653,
+      "step": 2052
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.1071389121049042,
+      "learning_rate": 9.815378140484291e-06,
+      "loss": 0.2269,
+      "step": 2053
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.9621436837414099,
+      "learning_rate": 9.815135528021284e-06,
+      "loss": 0.1682,
+      "step": 2054
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.4353290600218984,
+      "learning_rate": 9.81489275925621e-06,
+      "loss": 0.2231,
+      "step": 2055
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.0279281208853406,
+      "learning_rate": 9.814649834196947e-06,
+      "loss": 0.1857,
+      "step": 2056
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.2543498024217714,
+      "learning_rate": 9.814406752851383e-06,
+      "loss": 0.2698,
+      "step": 2057
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.989784653998838,
+      "learning_rate": 9.814163515227406e-06,
+      "loss": 0.1752,
+      "step": 2058
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 40.907474508237364,
+      "learning_rate": 9.813920121332913e-06,
+      "loss": 0.2673,
+      "step": 2059
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.7603344691195895,
+      "learning_rate": 9.813676571175805e-06,
+      "loss": 0.2762,
+      "step": 2060
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.6528881970643015,
+      "learning_rate": 9.813432864763989e-06,
+      "loss": 0.2056,
+      "step": 2061
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 15.039415897329155,
+      "learning_rate": 9.813189002105373e-06,
+      "loss": 0.323,
+      "step": 2062
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.5581335207482954,
+      "learning_rate": 9.812944983207872e-06,
+      "loss": 0.1861,
+      "step": 2063
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.0355984966688794,
+      "learning_rate": 9.812700808079413e-06,
+      "loss": 0.2114,
+      "step": 2064
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.0808394828664396,
+      "learning_rate": 9.812456476727916e-06,
+      "loss": 0.2704,
+      "step": 2065
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.9466981111988161,
+      "learning_rate": 9.812211989161316e-06,
+      "loss": 0.1807,
+      "step": 2066
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.4411453706214181,
+      "learning_rate": 9.811967345387545e-06,
+      "loss": 0.197,
+      "step": 2067
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.5492375663657278,
+      "learning_rate": 9.811722545414548e-06,
+      "loss": 0.2217,
+      "step": 2068
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.149360413218925,
+      "learning_rate": 9.811477589250272e-06,
+      "loss": 0.1736,
+      "step": 2069
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.2999769694209382,
+      "learning_rate": 9.811232476902666e-06,
+      "loss": 0.202,
+      "step": 2070
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.4525727278775793,
+      "learning_rate": 9.810987208379686e-06,
+      "loss": 0.2696,
+      "step": 2071
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.26325056520044,
+      "learning_rate": 9.810741783689295e-06,
+      "loss": 0.2759,
+      "step": 2072
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.5039662367759534,
+      "learning_rate": 9.810496202839459e-06,
+      "loss": 0.2385,
+      "step": 2073
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.4943085544935684,
+      "learning_rate": 9.810250465838151e-06,
+      "loss": 0.2083,
+      "step": 2074
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.2825613783162293,
+      "learning_rate": 9.810004572693347e-06,
+      "loss": 0.2477,
+      "step": 2075
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.3409991509736587,
+      "learning_rate": 9.809758523413028e-06,
+      "loss": 0.2072,
+      "step": 2076
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.3476656899605335,
+      "learning_rate": 9.80951231800518e-06,
+      "loss": 0.2328,
+      "step": 2077
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.3676419415657564,
+      "learning_rate": 9.809265956477799e-06,
+      "loss": 0.2092,
+      "step": 2078
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.0973038192462587,
+      "learning_rate": 9.809019438838879e-06,
+      "loss": 0.2137,
+      "step": 2079
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.593898677435856,
+      "learning_rate": 9.808772765096421e-06,
+      "loss": 0.2256,
+      "step": 2080
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.3047278306632313,
+      "learning_rate": 9.808525935258436e-06,
+      "loss": 0.2135,
+      "step": 2081
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.6517919922308018,
+      "learning_rate": 9.808278949332932e-06,
+      "loss": 0.1878,
+      "step": 2082
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.2934670518117595,
+      "learning_rate": 9.80803180732793e-06,
+      "loss": 0.2274,
+      "step": 2083
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.9698452026997929,
+      "learning_rate": 9.80778450925145e-06,
+      "loss": 0.2211,
+      "step": 2084
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.314851963926465,
+      "learning_rate": 9.80753705511152e-06,
+      "loss": 0.2075,
+      "step": 2085
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 2.155862662867026,
+      "learning_rate": 9.807289444916173e-06,
+      "loss": 0.2363,
+      "step": 2086
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.9400380013113335,
+      "learning_rate": 9.807041678673446e-06,
+      "loss": 0.1987,
+      "step": 2087
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.9526925209943289,
+      "learning_rate": 9.806793756391382e-06,
+      "loss": 0.1724,
+      "step": 2088
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.2452453240676038,
+      "learning_rate": 9.80654567807803e-06,
+      "loss": 0.2123,
+      "step": 2089
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.9492149205778136,
+      "learning_rate": 9.80629744374144e-06,
+      "loss": 0.2104,
+      "step": 2090
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.1056501480666352,
+      "learning_rate": 9.80604905338967e-06,
+      "loss": 0.2207,
+      "step": 2091
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.6091700326153253,
+      "learning_rate": 9.805800507030785e-06,
+      "loss": 0.2278,
+      "step": 2092
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.4172570239386935,
+      "learning_rate": 9.805551804672852e-06,
+      "loss": 0.2151,
+      "step": 2093
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 3.1823670065661176,
+      "learning_rate": 9.805302946323945e-06,
+      "loss": 0.228,
+      "step": 2094
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.561914596567066,
+      "learning_rate": 9.80505393199214e-06,
+      "loss": 0.2031,
+      "step": 2095
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.1573726400897617,
+      "learning_rate": 9.80480476168552e-06,
+      "loss": 0.2094,
+      "step": 2096
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.0952015985239851,
+      "learning_rate": 9.804555435412177e-06,
+      "loss": 0.252,
+      "step": 2097
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.6537401958385445,
+      "learning_rate": 9.8043059531802e-06,
+      "loss": 0.2182,
+      "step": 2098
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 2.0870102827625816,
+      "learning_rate": 9.80405631499769e-06,
+      "loss": 0.2103,
+      "step": 2099
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.1615503729786811,
+      "learning_rate": 9.803806520872749e-06,
+      "loss": 0.2236,
+      "step": 2100
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.284597592834828,
+      "learning_rate": 9.803556570813486e-06,
+      "loss": 0.229,
+      "step": 2101
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.2350367329613479,
+      "learning_rate": 9.803306464828015e-06,
+      "loss": 0.2612,
+      "step": 2102
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.4427424546164922,
+      "learning_rate": 9.803056202924455e-06,
+      "loss": 0.197,
+      "step": 2103
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.2411942919823786,
+      "learning_rate": 9.802805785110926e-06,
+      "loss": 0.2098,
+      "step": 2104
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.0957449056781972,
+      "learning_rate": 9.802555211395561e-06,
+      "loss": 0.2407,
+      "step": 2105
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.2684676916526312,
+      "learning_rate": 9.802304481786493e-06,
+      "loss": 0.1896,
+      "step": 2106
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.4047975151783294,
+      "learning_rate": 9.802053596291859e-06,
+      "loss": 0.2482,
+      "step": 2107
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.9823335860075298,
+      "learning_rate": 9.801802554919804e-06,
+      "loss": 0.2143,
+      "step": 2108
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.9691241814549489,
+      "learning_rate": 9.801551357678477e-06,
+      "loss": 0.1785,
+      "step": 2109
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.203269819733646,
+      "learning_rate": 9.801300004576032e-06,
+      "loss": 0.1861,
+      "step": 2110
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.1623224431744015,
+      "learning_rate": 9.801048495620627e-06,
+      "loss": 0.2179,
+      "step": 2111
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.1608903180901922,
+      "learning_rate": 9.800796830820427e-06,
+      "loss": 0.2122,
+      "step": 2112
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 2.4100717192777434,
+      "learning_rate": 9.800545010183601e-06,
+      "loss": 0.3622,
+      "step": 2113
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.2489227898603876,
+      "learning_rate": 9.800293033718324e-06,
+      "loss": 0.2002,
+      "step": 2114
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.2495015730153396,
+      "learning_rate": 9.800040901432776e-06,
+      "loss": 0.1892,
+      "step": 2115
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.0521550964518642,
+      "learning_rate": 9.799788613335139e-06,
+      "loss": 0.2597,
+      "step": 2116
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.537349878421653,
+      "learning_rate": 9.799536169433603e-06,
+      "loss": 0.2114,
+      "step": 2117
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.3016205133604217,
+      "learning_rate": 9.799283569736364e-06,
+      "loss": 0.2399,
+      "step": 2118
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.9038158243344518,
+      "learning_rate": 9.79903081425162e-06,
+      "loss": 0.1835,
+      "step": 2119
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.4997184682932725,
+      "learning_rate": 9.798777902987577e-06,
+      "loss": 0.2563,
+      "step": 2120
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.7642759564442133,
+      "learning_rate": 9.798524835952444e-06,
+      "loss": 0.1849,
+      "step": 2121
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.7465165370982225,
+      "learning_rate": 9.798271613154433e-06,
+      "loss": 0.2526,
+      "step": 2122
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.0488639457808533,
+      "learning_rate": 9.798018234601768e-06,
+      "loss": 0.1709,
+      "step": 2123
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.2789101380083456,
+      "learning_rate": 9.797764700302673e-06,
+      "loss": 0.288,
+      "step": 2124
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.9043788234251565,
+      "learning_rate": 9.797511010265374e-06,
+      "loss": 0.2749,
+      "step": 2125
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.2862901714458839,
+      "learning_rate": 9.79725716449811e-06,
+      "loss": 0.2475,
+      "step": 2126
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.5114773500886252,
+      "learning_rate": 9.797003163009122e-06,
+      "loss": 0.24,
+      "step": 2127
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.192719583027965,
+      "learning_rate": 9.796749005806651e-06,
+      "loss": 0.2125,
+      "step": 2128
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.90346233282495,
+      "learning_rate": 9.796494692898948e-06,
+      "loss": 0.2193,
+      "step": 2129
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.3156665816667608,
+      "learning_rate": 9.79624022429427e-06,
+      "loss": 0.1931,
+      "step": 2130
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.3840497833997607,
+      "learning_rate": 9.795985600000879e-06,
+      "loss": 0.2555,
+      "step": 2131
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.291640257533365,
+      "learning_rate": 9.795730820027034e-06,
+      "loss": 0.2372,
+      "step": 2132
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.856290508801194,
+      "learning_rate": 9.79547588438101e-06,
+      "loss": 0.2753,
+      "step": 2133
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.9159415729579217,
+      "learning_rate": 9.79522079307108e-06,
+      "loss": 0.2295,
+      "step": 2134
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.6459116816482942,
+      "learning_rate": 9.794965546105529e-06,
+      "loss": 0.264,
+      "step": 2135
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.1778848150274086,
+      "learning_rate": 9.794710143492637e-06,
+      "loss": 0.2098,
+      "step": 2136
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.402649973532007,
+      "learning_rate": 9.794454585240696e-06,
+      "loss": 0.2614,
+      "step": 2137
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.016342804414424,
+      "learning_rate": 9.794198871358002e-06,
+      "loss": 0.1665,
+      "step": 2138
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.476212370449032,
+      "learning_rate": 9.793943001852856e-06,
+      "loss": 0.1773,
+      "step": 2139
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.1114806437629163,
+      "learning_rate": 9.793686976733564e-06,
+      "loss": 0.211,
+      "step": 2140
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.5318533416716547,
+      "learning_rate": 9.793430796008437e-06,
+      "loss": 0.2449,
+      "step": 2141
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.07504991800825,
+      "learning_rate": 9.793174459685789e-06,
+      "loss": 0.1762,
+      "step": 2142
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.1811258250772687,
+      "learning_rate": 9.79291796777394e-06,
+      "loss": 0.1815,
+      "step": 2143
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.2874796136612416,
+      "learning_rate": 9.79266132028122e-06,
+      "loss": 0.1941,
+      "step": 2144
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.1455025310996243,
+      "learning_rate": 9.792404517215957e-06,
+      "loss": 0.2096,
+      "step": 2145
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.5091161404245905,
+      "learning_rate": 9.792147558586486e-06,
+      "loss": 0.2065,
+      "step": 2146
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.4565414757358062,
+      "learning_rate": 9.791890444401151e-06,
+      "loss": 0.2269,
+      "step": 2147
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.3888273299668208,
+      "learning_rate": 9.791633174668299e-06,
+      "loss": 0.1788,
+      "step": 2148
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.3366155848497936,
+      "learning_rate": 9.791375749396275e-06,
+      "loss": 0.184,
+      "step": 2149
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.5770057164858424,
+      "learning_rate": 9.79111816859344e-06,
+      "loss": 0.2257,
+      "step": 2150
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.6266289739556556,
+      "learning_rate": 9.790860432268154e-06,
+      "loss": 0.243,
+      "step": 2151
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.0542256589173788,
+      "learning_rate": 9.790602540428784e-06,
+      "loss": 0.2191,
+      "step": 2152
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.6483226211798911,
+      "learning_rate": 9.790344493083702e-06,
+      "loss": 0.2077,
+      "step": 2153
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.5462898727555565,
+      "learning_rate": 9.79008629024128e-06,
+      "loss": 0.2181,
+      "step": 2154
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.0076067302284701,
+      "learning_rate": 9.789827931909905e-06,
+      "loss": 0.2337,
+      "step": 2155
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.3918871777211168,
+      "learning_rate": 9.78956941809796e-06,
+      "loss": 0.2493,
+      "step": 2156
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.4555157048283316,
+      "learning_rate": 9.789310748813838e-06,
+      "loss": 0.2288,
+      "step": 2157
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.7676788385688396,
+      "learning_rate": 9.789051924065935e-06,
+      "loss": 0.2079,
+      "step": 2158
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.406460377400993,
+      "learning_rate": 9.788792943862654e-06,
+      "loss": 0.2015,
+      "step": 2159
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.4764862141758457,
+      "learning_rate": 9.788533808212397e-06,
+      "loss": 0.2145,
+      "step": 2160
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.945523199134706,
+      "learning_rate": 9.788274517123582e-06,
+      "loss": 0.2026,
+      "step": 2161
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.152965045188393,
+      "learning_rate": 9.78801507060462e-06,
+      "loss": 0.2217,
+      "step": 2162
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.2126325879530522,
+      "learning_rate": 9.787755468663938e-06,
+      "loss": 0.224,
+      "step": 2163
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.9106585463194519,
+      "learning_rate": 9.787495711309958e-06,
+      "loss": 0.2186,
+      "step": 2164
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.6435923217889081,
+      "learning_rate": 9.787235798551115e-06,
+      "loss": 0.2474,
+      "step": 2165
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.259851961356305,
+      "learning_rate": 9.786975730395848e-06,
+      "loss": 0.1839,
+      "step": 2166
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.301469481173629,
+      "learning_rate": 9.786715506852593e-06,
+      "loss": 0.2057,
+      "step": 2167
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.3328267012864776,
+      "learning_rate": 9.7864551279298e-06,
+      "loss": 0.247,
+      "step": 2168
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.150756311502496,
+      "learning_rate": 9.786194593635922e-06,
+      "loss": 0.1972,
+      "step": 2169
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.0777618778870184,
+      "learning_rate": 9.785933903979415e-06,
+      "loss": 0.2007,
+      "step": 2170
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.429524457994396,
+      "learning_rate": 9.785673058968741e-06,
+      "loss": 0.2236,
+      "step": 2171
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.9506375513861627,
+      "learning_rate": 9.785412058612368e-06,
+      "loss": 0.233,
+      "step": 2172
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.1596632537745948,
+      "learning_rate": 9.785150902918769e-06,
+      "loss": 0.2257,
+      "step": 2173
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.2753358839417828,
+      "learning_rate": 9.784889591896419e-06,
+      "loss": 0.1846,
+      "step": 2174
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.6835455113113988,
+      "learning_rate": 9.784628125553801e-06,
+      "loss": 0.2434,
+      "step": 2175
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.376875340855193,
+      "learning_rate": 9.784366503899403e-06,
+      "loss": 0.1942,
+      "step": 2176
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.3775411689381982,
+      "learning_rate": 9.784104726941716e-06,
+      "loss": 0.216,
+      "step": 2177
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.9883046882473614,
+      "learning_rate": 9.78384279468924e-06,
+      "loss": 0.202,
+      "step": 2178
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.9652682154492089,
+      "learning_rate": 9.783580707150475e-06,
+      "loss": 0.1816,
+      "step": 2179
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.8634233746113571,
+      "learning_rate": 9.78331846433393e-06,
+      "loss": 0.2245,
+      "step": 2180
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.0178959033632162,
+      "learning_rate": 9.783056066248118e-06,
+      "loss": 0.1949,
+      "step": 2181
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.3789886093398815,
+      "learning_rate": 9.782793512901555e-06,
+      "loss": 0.2342,
+      "step": 2182
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.2884244408964576,
+      "learning_rate": 9.782530804302763e-06,
+      "loss": 0.2314,
+      "step": 2183
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.6794758508098202,
+      "learning_rate": 9.782267940460272e-06,
+      "loss": 0.2586,
+      "step": 2184
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.7178747723017238,
+      "learning_rate": 9.782004921382612e-06,
+      "loss": 0.24,
+      "step": 2185
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.7633000992700028,
+      "learning_rate": 9.781741747078325e-06,
+      "loss": 0.1892,
+      "step": 2186
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.1889815813018345,
+      "learning_rate": 9.78147841755595e-06,
+      "loss": 0.2289,
+      "step": 2187
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.3683544328727193,
+      "learning_rate": 9.781214932824036e-06,
+      "loss": 0.2311,
+      "step": 2188
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.3072018180863707,
+      "learning_rate": 9.780951292891136e-06,
+      "loss": 0.21,
+      "step": 2189
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.5983691175717574,
+      "learning_rate": 9.780687497765808e-06,
+      "loss": 0.2686,
+      "step": 2190
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 7.895351724019506,
+      "learning_rate": 9.780423547456615e-06,
+      "loss": 0.2719,
+      "step": 2191
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.0855714297597474,
+      "learning_rate": 9.780159441972123e-06,
+      "loss": 0.1873,
+      "step": 2192
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.0650542355549137,
+      "learning_rate": 9.779895181320908e-06,
+      "loss": 0.2043,
+      "step": 2193
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.7525129200014247,
+      "learning_rate": 9.779630765511548e-06,
+      "loss": 0.2711,
+      "step": 2194
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.473840307058485,
+      "learning_rate": 9.779366194552621e-06,
+      "loss": 0.2475,
+      "step": 2195
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.219397498181828,
+      "learning_rate": 9.779101468452724e-06,
+      "loss": 0.1729,
+      "step": 2196
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.2174468363452817,
+      "learning_rate": 9.778836587220441e-06,
+      "loss": 0.2624,
+      "step": 2197
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.8564293212041352,
+      "learning_rate": 9.778571550864378e-06,
+      "loss": 0.256,
+      "step": 2198
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.6887754635758219,
+      "learning_rate": 9.77830635939313e-06,
+      "loss": 0.2461,
+      "step": 2199
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.349306332010216,
+      "learning_rate": 9.778041012815313e-06,
+      "loss": 0.1577,
+      "step": 2200
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.9101615524956137,
+      "learning_rate": 9.777775511139538e-06,
+      "loss": 0.243,
+      "step": 2201
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.9710606467327595,
+      "learning_rate": 9.777509854374421e-06,
+      "loss": 0.1975,
+      "step": 2202
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.4154722894526843,
+      "learning_rate": 9.777244042528586e-06,
+      "loss": 0.2779,
+      "step": 2203
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.7707736181743171,
+      "learning_rate": 9.776978075610663e-06,
+      "loss": 0.2272,
+      "step": 2204
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.1794894380595844,
+      "learning_rate": 9.776711953629285e-06,
+      "loss": 0.1569,
+      "step": 2205
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.9724457505345215,
+      "learning_rate": 9.776445676593091e-06,
+      "loss": 0.2119,
+      "step": 2206
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.336137601857495,
+      "learning_rate": 9.776179244510725e-06,
+      "loss": 0.2454,
+      "step": 2207
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.187790831915437,
+      "learning_rate": 9.77591265739083e-06,
+      "loss": 0.2118,
+      "step": 2208
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.4358234087269304,
+      "learning_rate": 9.775645915242067e-06,
+      "loss": 0.1829,
+      "step": 2209
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.1457265009350353,
+      "learning_rate": 9.77537901807309e-06,
+      "loss": 0.202,
+      "step": 2210
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.2622450713986773,
+      "learning_rate": 9.775111965892563e-06,
+      "loss": 0.2003,
+      "step": 2211
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.3120829836504329,
+      "learning_rate": 9.774844758709158e-06,
+      "loss": 0.2398,
+      "step": 2212
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.5291053321948807,
+      "learning_rate": 9.774577396531546e-06,
+      "loss": 0.2436,
+      "step": 2213
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.9187640093364611,
+      "learning_rate": 9.774309879368405e-06,
+      "loss": 0.1929,
+      "step": 2214
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.2202715415955736,
+      "learning_rate": 9.77404220722842e-06,
+      "loss": 0.2018,
+      "step": 2215
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.1919595684230648,
+      "learning_rate": 9.77377438012028e-06,
+      "loss": 0.19,
+      "step": 2216
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.9315146185074803,
+      "learning_rate": 9.77350639805268e-06,
+      "loss": 0.2175,
+      "step": 2217
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.4536177962197485,
+      "learning_rate": 9.773238261034315e-06,
+      "loss": 0.1851,
+      "step": 2218
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.1429939002596428,
+      "learning_rate": 9.772969969073893e-06,
+      "loss": 0.1872,
+      "step": 2219
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.2864873537916475,
+      "learning_rate": 9.77270152218012e-06,
+      "loss": 0.2237,
+      "step": 2220
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.7089795968125168,
+      "learning_rate": 9.77243292036171e-06,
+      "loss": 0.1938,
+      "step": 2221
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.3043627713166674,
+      "learning_rate": 9.772164163627385e-06,
+      "loss": 0.1928,
+      "step": 2222
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.503130839211588,
+      "learning_rate": 9.771895251985866e-06,
+      "loss": 0.1893,
+      "step": 2223
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.079125387696875,
+      "learning_rate": 9.771626185445884e-06,
+      "loss": 0.2169,
+      "step": 2224
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.1077966274636268,
+      "learning_rate": 9.771356964016172e-06,
+      "loss": 0.1567,
+      "step": 2225
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.794411641719733,
+      "learning_rate": 9.77108758770547e-06,
+      "loss": 0.2252,
+      "step": 2226
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.521448017415599,
+      "learning_rate": 9.77081805652252e-06,
+      "loss": 0.2394,
+      "step": 2227
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.3343233764096514,
+      "learning_rate": 9.770548370476072e-06,
+      "loss": 0.244,
+      "step": 2228
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.153999548506194,
+      "learning_rate": 9.770278529574882e-06,
+      "loss": 0.2577,
+      "step": 2229
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.1915478293923634,
+      "learning_rate": 9.770008533827707e-06,
+      "loss": 0.1363,
+      "step": 2230
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.4249487131493175,
+      "learning_rate": 9.769738383243313e-06,
+      "loss": 0.2524,
+      "step": 2231
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.9268649173022253,
+      "learning_rate": 9.769468077830467e-06,
+      "loss": 0.2023,
+      "step": 2232
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 2.081496350712433,
+      "learning_rate": 9.769197617597946e-06,
+      "loss": 0.1814,
+      "step": 2233
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.181144890164421,
+      "learning_rate": 9.768927002554527e-06,
+      "loss": 0.1813,
+      "step": 2234
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.0165952999277545,
+      "learning_rate": 9.768656232708995e-06,
+      "loss": 0.23,
+      "step": 2235
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.382554806119626,
+      "learning_rate": 9.768385308070139e-06,
+      "loss": 0.2379,
+      "step": 2236
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.189034352238826,
+      "learning_rate": 9.768114228646755e-06,
+      "loss": 0.244,
+      "step": 2237
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.316417136557641,
+      "learning_rate": 9.76784299444764e-06,
+      "loss": 0.1939,
+      "step": 2238
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.0709806022521937,
+      "learning_rate": 9.7675716054816e-06,
+      "loss": 0.1862,
+      "step": 2239
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.5025205495257399,
+      "learning_rate": 9.767300061757445e-06,
+      "loss": 0.1972,
+      "step": 2240
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.451270464632527,
+      "learning_rate": 9.76702836328399e-06,
+      "loss": 0.2226,
+      "step": 2241
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.7939660990532955,
+      "learning_rate": 9.76675651007005e-06,
+      "loss": 0.236,
+      "step": 2242
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.041336940396367,
+      "learning_rate": 9.766484502124453e-06,
+      "loss": 0.2518,
+      "step": 2243
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.8660261899005157,
+      "learning_rate": 9.76621233945603e-06,
+      "loss": 0.2038,
+      "step": 2244
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.9321605203421958,
+      "learning_rate": 9.765940022073613e-06,
+      "loss": 0.1735,
+      "step": 2245
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.8646685524051195,
+      "learning_rate": 9.765667549986042e-06,
+      "loss": 0.1745,
+      "step": 2246
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.4483325168685877,
+      "learning_rate": 9.765394923202163e-06,
+      "loss": 0.1988,
+      "step": 2247
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.650977623962978,
+      "learning_rate": 9.765122141730822e-06,
+      "loss": 0.1816,
+      "step": 2248
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.2386054617130882,
+      "learning_rate": 9.764849205580877e-06,
+      "loss": 0.2149,
+      "step": 2249
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.154965923266439,
+      "learning_rate": 9.764576114761188e-06,
+      "loss": 0.19,
+      "step": 2250
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.0056225710881586,
+      "learning_rate": 9.764302869280618e-06,
+      "loss": 0.2476,
+      "step": 2251
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.1031293699404907,
+      "learning_rate": 9.764029469148037e-06,
+      "loss": 0.1703,
+      "step": 2252
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.9534190400529895,
+      "learning_rate": 9.76375591437232e-06,
+      "loss": 0.1623,
+      "step": 2253
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.0928576661103884,
+      "learning_rate": 9.763482204962347e-06,
+      "loss": 0.2131,
+      "step": 2254
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.0273793725629528,
+      "learning_rate": 9.763208340927002e-06,
+      "loss": 0.2001,
+      "step": 2255
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.4255493177912628,
+      "learning_rate": 9.762934322275175e-06,
+      "loss": 0.1806,
+      "step": 2256
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.433513101185419,
+      "learning_rate": 9.762660149015761e-06,
+      "loss": 0.2129,
+      "step": 2257
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.1561688043463494,
+      "learning_rate": 9.76238582115766e-06,
+      "loss": 0.1825,
+      "step": 2258
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.4223583364089725,
+      "learning_rate": 9.762111338709778e-06,
+      "loss": 0.2627,
+      "step": 2259
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.1898104063735915,
+      "learning_rate": 9.761836701681022e-06,
+      "loss": 0.1851,
+      "step": 2260
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.927257600042978,
+      "learning_rate": 9.761561910080308e-06,
+      "loss": 0.2282,
+      "step": 2261
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.270665603913988,
+      "learning_rate": 9.76128696391656e-06,
+      "loss": 0.2177,
+      "step": 2262
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.3745495399705712,
+      "learning_rate": 9.761011863198693e-06,
+      "loss": 0.201,
+      "step": 2263
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.0753604457207988,
+      "learning_rate": 9.760736607935648e-06,
+      "loss": 0.1939,
+      "step": 2264
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.327151163082959,
+      "learning_rate": 9.760461198136352e-06,
+      "loss": 0.2168,
+      "step": 2265
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.150664589971099,
+      "learning_rate": 9.76018563380975e-06,
+      "loss": 0.2057,
+      "step": 2266
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.9271045368851405,
+      "learning_rate": 9.759909914964785e-06,
+      "loss": 0.1788,
+      "step": 2267
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.107737338162153,
+      "learning_rate": 9.759634041610406e-06,
+      "loss": 0.2362,
+      "step": 2268
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.4955656262407775,
+      "learning_rate": 9.759358013755571e-06,
+      "loss": 0.2382,
+      "step": 2269
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.9485470579286882,
+      "learning_rate": 9.759081831409236e-06,
+      "loss": 0.1853,
+      "step": 2270
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.5389521506005304,
+      "learning_rate": 9.758805494580369e-06,
+      "loss": 0.2268,
+      "step": 2271
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.4347341845084889,
+      "learning_rate": 9.758529003277939e-06,
+      "loss": 0.2811,
+      "step": 2272
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.0947683379125306,
+      "learning_rate": 9.75825235751092e-06,
+      "loss": 0.2004,
+      "step": 2273
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.1065750854022116,
+      "learning_rate": 9.757975557288296e-06,
+      "loss": 0.2324,
+      "step": 2274
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.0855381148878134,
+      "learning_rate": 9.757698602619047e-06,
+      "loss": 0.1725,
+      "step": 2275
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.1411425203585475,
+      "learning_rate": 9.757421493512168e-06,
+      "loss": 0.2111,
+      "step": 2276
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.9201483141840427,
+      "learning_rate": 9.757144229976647e-06,
+      "loss": 0.2403,
+      "step": 2277
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.7749774898828573,
+      "learning_rate": 9.756866812021493e-06,
+      "loss": 0.171,
+      "step": 2278
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.1869731553159677,
+      "learning_rate": 9.756589239655705e-06,
+      "loss": 0.2441,
+      "step": 2279
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.261283811356082,
+      "learning_rate": 9.756311512888296e-06,
+      "loss": 0.2439,
+      "step": 2280
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.449859208355427,
+      "learning_rate": 9.756033631728278e-06,
+      "loss": 0.2197,
+      "step": 2281
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.512179885931016,
+      "learning_rate": 9.755755596184678e-06,
+      "loss": 0.2285,
+      "step": 2282
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.1324095831602303,
+      "learning_rate": 9.755477406266511e-06,
+      "loss": 0.1993,
+      "step": 2283
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.2953042548898497,
+      "learning_rate": 9.755199061982817e-06,
+      "loss": 0.2379,
+      "step": 2284
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.3700417261965432,
+      "learning_rate": 9.754920563342625e-06,
+      "loss": 0.2609,
+      "step": 2285
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.9573780618648027,
+      "learning_rate": 9.754641910354977e-06,
+      "loss": 0.2463,
+      "step": 2286
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.324627984652208,
+      "learning_rate": 9.754363103028922e-06,
+      "loss": 0.2376,
+      "step": 2287
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.2333934660001158,
+      "learning_rate": 9.754084141373502e-06,
+      "loss": 0.2373,
+      "step": 2288
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.1569616706695338,
+      "learning_rate": 9.75380502539778e-06,
+      "loss": 0.1753,
+      "step": 2289
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.0769708440330974,
+      "learning_rate": 9.753525755110812e-06,
+      "loss": 0.228,
+      "step": 2290
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.650025831908977,
+      "learning_rate": 9.753246330521664e-06,
+      "loss": 0.2313,
+      "step": 2291
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.6505807349507045,
+      "learning_rate": 9.752966751639407e-06,
+      "loss": 0.2537,
+      "step": 2292
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.3686142261870906,
+      "learning_rate": 9.752687018473118e-06,
+      "loss": 0.2041,
+      "step": 2293
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.9671784274179535,
+      "learning_rate": 9.752407131031873e-06,
+      "loss": 0.2101,
+      "step": 2294
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.9944601313715573,
+      "learning_rate": 9.75212708932476e-06,
+      "loss": 0.2514,
+      "step": 2295
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.0968119201019542,
+      "learning_rate": 9.75184689336087e-06,
+      "loss": 0.2042,
+      "step": 2296
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.4727391414434678,
+      "learning_rate": 9.751566543149298e-06,
+      "loss": 0.2085,
+      "step": 2297
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.6506710705814986,
+      "learning_rate": 9.751286038699142e-06,
+      "loss": 0.2131,
+      "step": 2298
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.1921041039512135,
+      "learning_rate": 9.75100538001951e-06,
+      "loss": 0.233,
+      "step": 2299
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.1629444391768615,
+      "learning_rate": 9.750724567119512e-06,
+      "loss": 0.209,
+      "step": 2300
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.821764412663175,
+      "learning_rate": 9.750443600008262e-06,
+      "loss": 0.1903,
+      "step": 2301
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.2638760568316114,
+      "learning_rate": 9.75016247869488e-06,
+      "loss": 0.1909,
+      "step": 2302
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.4593785173590217,
+      "learning_rate": 9.749881203188494e-06,
+      "loss": 0.262,
+      "step": 2303
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.349696407051355,
+      "learning_rate": 9.74959977349823e-06,
+      "loss": 0.2289,
+      "step": 2304
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.9590921797825755,
+      "learning_rate": 9.749318189633229e-06,
+      "loss": 0.2057,
+      "step": 2305
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.634314049683814,
+      "learning_rate": 9.749036451602627e-06,
+      "loss": 0.2063,
+      "step": 2306
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.3516816620031926,
+      "learning_rate": 9.748754559415573e-06,
+      "loss": 0.2487,
+      "step": 2307
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.234390785321574,
+      "learning_rate": 9.748472513081212e-06,
+      "loss": 0.2333,
+      "step": 2308
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.0587850992466001,
+      "learning_rate": 9.748190312608706e-06,
+      "loss": 0.1904,
+      "step": 2309
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.4865488497524555,
+      "learning_rate": 9.74790795800721e-06,
+      "loss": 0.2145,
+      "step": 2310
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.0146912794274803,
+      "learning_rate": 9.74762544928589e-06,
+      "loss": 0.2165,
+      "step": 2311
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.119004409382709,
+      "learning_rate": 9.747342786453922e-06,
+      "loss": 0.1459,
+      "step": 2312
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.0819513880181155,
+      "learning_rate": 9.747059969520474e-06,
+      "loss": 0.2083,
+      "step": 2313
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.0680560792073361,
+      "learning_rate": 9.746776998494731e-06,
+      "loss": 0.1719,
+      "step": 2314
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 4.6425818225427005,
+      "learning_rate": 9.746493873385877e-06,
+      "loss": 0.2386,
+      "step": 2315
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.2480270178204655,
+      "learning_rate": 9.746210594203102e-06,
+      "loss": 0.185,
+      "step": 2316
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.1287551567576615,
+      "learning_rate": 9.745927160955601e-06,
+      "loss": 0.165,
+      "step": 2317
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.3321279185839348,
+      "learning_rate": 9.745643573652577e-06,
+      "loss": 0.1889,
+      "step": 2318
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.9334213950665757,
+      "learning_rate": 9.745359832303233e-06,
+      "loss": 0.2397,
+      "step": 2319
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.80728044675886,
+      "learning_rate": 9.74507593691678e-06,
+      "loss": 0.2224,
+      "step": 2320
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.452473243041798,
+      "learning_rate": 9.744791887502434e-06,
+      "loss": 0.2263,
+      "step": 2321
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.9077508292367742,
+      "learning_rate": 9.744507684069413e-06,
+      "loss": 0.1923,
+      "step": 2322
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.1485540986070126,
+      "learning_rate": 9.744223326626947e-06,
+      "loss": 0.1805,
+      "step": 2323
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.3931027958205573,
+      "learning_rate": 9.743938815184262e-06,
+      "loss": 0.2167,
+      "step": 2324
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.3356157432656208,
+      "learning_rate": 9.743654149750594e-06,
+      "loss": 0.1876,
+      "step": 2325
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.6008910632505253,
+      "learning_rate": 9.743369330335186e-06,
+      "loss": 0.2383,
+      "step": 2326
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.4278112722163223,
+      "learning_rate": 9.743084356947281e-06,
+      "loss": 0.189,
+      "step": 2327
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.7554204920581353,
+      "learning_rate": 9.742799229596132e-06,
+      "loss": 0.2561,
+      "step": 2328
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.597447202105506,
+      "learning_rate": 9.742513948290992e-06,
+      "loss": 0.2199,
+      "step": 2329
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.1277301215450675,
+      "learning_rate": 9.742228513041122e-06,
+      "loss": 0.2305,
+      "step": 2330
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.058946608753291,
+      "learning_rate": 9.741942923855787e-06,
+      "loss": 0.2385,
+      "step": 2331
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.0884882074975866,
+      "learning_rate": 9.741657180744259e-06,
+      "loss": 0.2216,
+      "step": 2332
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.7678634160413538,
+      "learning_rate": 9.74137128371581e-06,
+      "loss": 0.1955,
+      "step": 2333
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.2642280877884349,
+      "learning_rate": 9.741085232779726e-06,
+      "loss": 0.2166,
+      "step": 2334
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.0552542011683002,
+      "learning_rate": 9.740799027945286e-06,
+      "loss": 0.2018,
+      "step": 2335
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.2947574197473148,
+      "learning_rate": 9.740512669221784e-06,
+      "loss": 0.1836,
+      "step": 2336
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.4730410275561983,
+      "learning_rate": 9.740226156618516e-06,
+      "loss": 0.2462,
+      "step": 2337
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.2293113828553905,
+      "learning_rate": 9.739939490144781e-06,
+      "loss": 0.1656,
+      "step": 2338
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.0553527182353046,
+      "learning_rate": 9.739652669809883e-06,
+      "loss": 0.1914,
+      "step": 2339
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.1602305049766197,
+      "learning_rate": 9.739365695623136e-06,
+      "loss": 0.2506,
+      "step": 2340
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.2409488197118739,
+      "learning_rate": 9.739078567593852e-06,
+      "loss": 0.198,
+      "step": 2341
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.6862230425451474,
+      "learning_rate": 9.738791285731353e-06,
+      "loss": 0.1987,
+      "step": 2342
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.2994505946650123,
+      "learning_rate": 9.738503850044962e-06,
+      "loss": 0.2367,
+      "step": 2343
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.7095244682281396,
+      "learning_rate": 9.738216260544014e-06,
+      "loss": 0.1882,
+      "step": 2344
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.3336265008521782,
+      "learning_rate": 9.737928517237839e-06,
+      "loss": 0.2268,
+      "step": 2345
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.1812532250957009,
+      "learning_rate": 9.737640620135781e-06,
+      "loss": 0.1877,
+      "step": 2346
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.1134821961398718,
+      "learning_rate": 9.737352569247184e-06,
+      "loss": 0.1753,
+      "step": 2347
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.3835493753771158,
+      "learning_rate": 9.737064364581399e-06,
+      "loss": 0.1836,
+      "step": 2348
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.043232939688414,
+      "learning_rate": 9.73677600614778e-06,
+      "loss": 0.2138,
+      "step": 2349
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.3745405163636364,
+      "learning_rate": 9.736487493955688e-06,
+      "loss": 0.2104,
+      "step": 2350
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.4341214589877216,
+      "learning_rate": 9.736198828014489e-06,
+      "loss": 0.2054,
+      "step": 2351
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.2509752790219635,
+      "learning_rate": 9.73591000833355e-06,
+      "loss": 0.214,
+      "step": 2352
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.9139246785765774,
+      "learning_rate": 9.735621034922253e-06,
+      "loss": 0.2339,
+      "step": 2353
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.1223423630515856,
+      "learning_rate": 9.735331907789971e-06,
+      "loss": 0.17,
+      "step": 2354
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.1632690557784788,
+      "learning_rate": 9.735042626946094e-06,
+      "loss": 0.2045,
+      "step": 2355
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.4475426054916138,
+      "learning_rate": 9.734753192400009e-06,
+      "loss": 0.1996,
+      "step": 2356
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.2160571775312572,
+      "learning_rate": 9.734463604161114e-06,
+      "loss": 0.2184,
+      "step": 2357
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.1912619057750369,
+      "learning_rate": 9.734173862238807e-06,
+      "loss": 0.1842,
+      "step": 2358
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.3936847767729088,
+      "learning_rate": 9.733883966642494e-06,
+      "loss": 0.2361,
+      "step": 2359
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.5877972114366004,
+      "learning_rate": 9.733593917381585e-06,
+      "loss": 0.2315,
+      "step": 2360
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.1422997795865462,
+      "learning_rate": 9.733303714465496e-06,
+      "loss": 0.1963,
+      "step": 2361
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.3436188994262148,
+      "learning_rate": 9.733013357903647e-06,
+      "loss": 0.1797,
+      "step": 2362
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.5750172710134032,
+      "learning_rate": 9.732722847705462e-06,
+      "loss": 0.2258,
+      "step": 2363
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.766106852044344,
+      "learning_rate": 9.732432183880371e-06,
+      "loss": 0.1611,
+      "step": 2364
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.1806762133052746,
+      "learning_rate": 9.732141366437812e-06,
+      "loss": 0.1939,
+      "step": 2365
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.2356993506073009,
+      "learning_rate": 9.731850395387222e-06,
+      "loss": 0.1939,
+      "step": 2366
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.1873495012835038,
+      "learning_rate": 9.731559270738047e-06,
+      "loss": 0.2317,
+      "step": 2367
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.0520123038376974,
+      "learning_rate": 9.731267992499737e-06,
+      "loss": 0.2019,
+      "step": 2368
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.4854434735454956,
+      "learning_rate": 9.730976560681748e-06,
+      "loss": 0.1729,
+      "step": 2369
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.2707801793726086,
+      "learning_rate": 9.73068497529354e-06,
+      "loss": 0.2394,
+      "step": 2370
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.0776555414461602,
+      "learning_rate": 9.730393236344576e-06,
+      "loss": 0.1513,
+      "step": 2371
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.6549167124890065,
+      "learning_rate": 9.730101343844329e-06,
+      "loss": 0.2548,
+      "step": 2372
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.9312807623691473,
+      "learning_rate": 9.72980929780227e-06,
+      "loss": 0.2054,
+      "step": 2373
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.3747900943493723,
+      "learning_rate": 9.729517098227884e-06,
+      "loss": 0.2558,
+      "step": 2374
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.478053728188869,
+      "learning_rate": 9.729224745130653e-06,
+      "loss": 0.2253,
+      "step": 2375
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.212116828511981,
+      "learning_rate": 9.728932238520065e-06,
+      "loss": 0.1871,
+      "step": 2376
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.0478055362600847,
+      "learning_rate": 9.728639578405618e-06,
+      "loss": 0.157,
+      "step": 2377
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.662635401718828,
+      "learning_rate": 9.728346764796813e-06,
+      "loss": 0.219,
+      "step": 2378
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.0939174592024532,
+      "learning_rate": 9.72805379770315e-06,
+      "loss": 0.1748,
+      "step": 2379
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.6251674483296432,
+      "learning_rate": 9.727760677134144e-06,
+      "loss": 0.2392,
+      "step": 2380
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.1125784479299745,
+      "learning_rate": 9.727467403099306e-06,
+      "loss": 0.2328,
+      "step": 2381
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.1988607712604704,
+      "learning_rate": 9.72717397560816e-06,
+      "loss": 0.1952,
+      "step": 2382
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.1764096073341368,
+      "learning_rate": 9.726880394670226e-06,
+      "loss": 0.1476,
+      "step": 2383
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.9693541066028986,
+      "learning_rate": 9.726586660295038e-06,
+      "loss": 0.2422,
+      "step": 2384
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.5592217626178437,
+      "learning_rate": 9.72629277249213e-06,
+      "loss": 0.3054,
+      "step": 2385
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.0229102898885827,
+      "learning_rate": 9.725998731271037e-06,
+      "loss": 0.2213,
+      "step": 2386
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.036958817331507,
+      "learning_rate": 9.725704536641312e-06,
+      "loss": 0.1708,
+      "step": 2387
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.1927202443988467,
+      "learning_rate": 9.725410188612498e-06,
+      "loss": 0.2526,
+      "step": 2388
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.799601284130631,
+      "learning_rate": 9.725115687194151e-06,
+      "loss": 0.2133,
+      "step": 2389
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.4488307993996077,
+      "learning_rate": 9.724821032395833e-06,
+      "loss": 0.2303,
+      "step": 2390
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.0612783631644696,
+      "learning_rate": 9.724526224227108e-06,
+      "loss": 0.1696,
+      "step": 2391
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.101694302080026,
+      "learning_rate": 9.724231262697544e-06,
+      "loss": 0.1901,
+      "step": 2392
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.9886965769474318,
+      "learning_rate": 9.723936147816718e-06,
+      "loss": 0.1259,
+      "step": 2393
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.636378028042206,
+      "learning_rate": 9.723640879594206e-06,
+      "loss": 0.2877,
+      "step": 2394
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.028528009579887,
+      "learning_rate": 9.723345458039595e-06,
+      "loss": 0.1713,
+      "step": 2395
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.4804995289524494,
+      "learning_rate": 9.723049883162474e-06,
+      "loss": 0.2301,
+      "step": 2396
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.1090477003914248,
+      "learning_rate": 9.72275415497244e-06,
+      "loss": 0.1868,
+      "step": 2397
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.1926861337368462,
+      "learning_rate": 9.722458273479089e-06,
+      "loss": 0.2004,
+      "step": 2398
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.2812775166976003,
+      "learning_rate": 9.722162238692027e-06,
+      "loss": 0.1947,
+      "step": 2399
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.0043866523949203,
+      "learning_rate": 9.721866050620864e-06,
+      "loss": 0.2508,
+      "step": 2400
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.0131584366582163,
+      "learning_rate": 9.72156970927521e-06,
+      "loss": 0.2236,
+      "step": 2401
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.5442111071851299,
+      "learning_rate": 9.721273214664692e-06,
+      "loss": 0.3043,
+      "step": 2402
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.3924164002661956,
+      "learning_rate": 9.72097656679893e-06,
+      "loss": 0.235,
+      "step": 2403
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.1053136844860545,
+      "learning_rate": 9.720679765687555e-06,
+      "loss": 0.1868,
+      "step": 2404
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.0312077773727188,
+      "learning_rate": 9.720382811340197e-06,
+      "loss": 0.2032,
+      "step": 2405
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.8748220450094591,
+      "learning_rate": 9.720085703766502e-06,
+      "loss": 0.2086,
+      "step": 2406
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.2269047150387424,
+      "learning_rate": 9.719788442976109e-06,
+      "loss": 0.1459,
+      "step": 2407
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.279303590014357,
+      "learning_rate": 9.719491028978669e-06,
+      "loss": 0.2142,
+      "step": 2408
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.1653084523430426,
+      "learning_rate": 9.719193461783836e-06,
+      "loss": 0.2204,
+      "step": 2409
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.02410059206599,
+      "learning_rate": 9.71889574140127e-06,
+      "loss": 0.2115,
+      "step": 2410
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.9048540601121986,
+      "learning_rate": 9.718597867840634e-06,
+      "loss": 0.2588,
+      "step": 2411
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.4379818357572547,
+      "learning_rate": 9.7182998411116e-06,
+      "loss": 0.1957,
+      "step": 2412
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.3358908346737208,
+      "learning_rate": 9.718001661223838e-06,
+      "loss": 0.215,
+      "step": 2413
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.2034367739492726,
+      "learning_rate": 9.717703328187029e-06,
+      "loss": 0.2239,
+      "step": 2414
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.3205468898024606,
+      "learning_rate": 9.717404842010857e-06,
+      "loss": 0.2127,
+      "step": 2415
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.9875724398285852,
+      "learning_rate": 9.717106202705012e-06,
+      "loss": 0.1508,
+      "step": 2416
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.9080587210756762,
+      "learning_rate": 9.716807410279186e-06,
+      "loss": 0.2204,
+      "step": 2417
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.3298884661281627,
+      "learning_rate": 9.716508464743081e-06,
+      "loss": 0.2319,
+      "step": 2418
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.279665018939176,
+      "learning_rate": 9.716209366106396e-06,
+      "loss": 0.2458,
+      "step": 2419
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.0208787563332404,
+      "learning_rate": 9.715910114378845e-06,
+      "loss": 0.2004,
+      "step": 2420
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.2705257980398967,
+      "learning_rate": 9.715610709570142e-06,
+      "loss": 0.2194,
+      "step": 2421
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.281952809137262,
+      "learning_rate": 9.715311151690001e-06,
+      "loss": 0.2548,
+      "step": 2422
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.453250433522594,
+      "learning_rate": 9.715011440748149e-06,
+      "loss": 0.2253,
+      "step": 2423
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.0194312627586466,
+      "learning_rate": 9.714711576754314e-06,
+      "loss": 0.1915,
+      "step": 2424
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.5626918841795092,
+      "learning_rate": 9.714411559718231e-06,
+      "loss": 0.2194,
+      "step": 2425
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.307593609666394,
+      "learning_rate": 9.714111389649636e-06,
+      "loss": 0.1931,
+      "step": 2426
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.4366611515839693,
+      "learning_rate": 9.713811066558277e-06,
+      "loss": 0.212,
+      "step": 2427
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.0749002678969724,
+      "learning_rate": 9.713510590453898e-06,
+      "loss": 0.2547,
+      "step": 2428
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.1896199452344025,
+      "learning_rate": 9.713209961346256e-06,
+      "loss": 0.2269,
+      "step": 2429
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.6315076167840419,
+      "learning_rate": 9.712909179245107e-06,
+      "loss": 0.2515,
+      "step": 2430
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.1338033946311823,
+      "learning_rate": 9.712608244160218e-06,
+      "loss": 0.1947,
+      "step": 2431
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.2910572166495429,
+      "learning_rate": 9.712307156101353e-06,
+      "loss": 0.2372,
+      "step": 2432
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.0049648788943646,
+      "learning_rate": 9.712005915078289e-06,
+      "loss": 0.1942,
+      "step": 2433
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.3455670333612115,
+      "learning_rate": 9.711704521100805e-06,
+      "loss": 0.2499,
+      "step": 2434
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.2042952139722645,
+      "learning_rate": 9.711402974178681e-06,
+      "loss": 0.1912,
+      "step": 2435
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.1124721481421147,
+      "learning_rate": 9.711101274321707e-06,
+      "loss": 0.198,
+      "step": 2436
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.0419626437345402,
+      "learning_rate": 9.710799421539677e-06,
+      "loss": 0.2385,
+      "step": 2437
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.376282946180781,
+      "learning_rate": 9.710497415842389e-06,
+      "loss": 0.1811,
+      "step": 2438
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.331591458026999,
+      "learning_rate": 9.710195257239647e-06,
+      "loss": 0.2403,
+      "step": 2439
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.1811302627700973,
+      "learning_rate": 9.709892945741257e-06,
+      "loss": 0.2481,
+      "step": 2440
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.0994073536543307,
+      "learning_rate": 9.709590481357034e-06,
+      "loss": 0.227,
+      "step": 2441
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.5088431438148313,
+      "learning_rate": 9.709287864096797e-06,
+      "loss": 0.2153,
+      "step": 2442
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.4712927790019703,
+      "learning_rate": 9.708985093970367e-06,
+      "loss": 0.24,
+      "step": 2443
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.0138150529311931,
+      "learning_rate": 9.708682170987573e-06,
+      "loss": 0.169,
+      "step": 2444
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.9832823835457621,
+      "learning_rate": 9.708379095158247e-06,
+      "loss": 0.2156,
+      "step": 2445
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.199388446978477,
+      "learning_rate": 9.70807586649223e-06,
+      "loss": 0.1825,
+      "step": 2446
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.3799859607045455,
+      "learning_rate": 9.70777248499936e-06,
+      "loss": 0.1939,
+      "step": 2447
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.1098492340796646,
+      "learning_rate": 9.70746895068949e-06,
+      "loss": 0.2093,
+      "step": 2448
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.2587820516055173,
+      "learning_rate": 9.70716526357247e-06,
+      "loss": 0.2135,
+      "step": 2449
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.241109506227769,
+      "learning_rate": 9.706861423658162e-06,
+      "loss": 0.2139,
+      "step": 2450
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.1870674973833915,
+      "learning_rate": 9.706557430956423e-06,
+      "loss": 0.2104,
+      "step": 2451
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.052285004712827,
+      "learning_rate": 9.706253285477122e-06,
+      "loss": 0.1933,
+      "step": 2452
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.4880918146956894,
+      "learning_rate": 9.705948987230137e-06,
+      "loss": 0.2051,
+      "step": 2453
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.091071618142612,
+      "learning_rate": 9.705644536225338e-06,
+      "loss": 0.2142,
+      "step": 2454
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.347785285632503,
+      "learning_rate": 9.705339932472614e-06,
+      "loss": 0.2203,
+      "step": 2455
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.1772257638170296,
+      "learning_rate": 9.705035175981847e-06,
+      "loss": 0.2012,
+      "step": 2456
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.316363251886603,
+      "learning_rate": 9.704730266762934e-06,
+      "loss": 0.2211,
+      "step": 2457
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.24950038939062,
+      "learning_rate": 9.704425204825772e-06,
+      "loss": 0.2524,
+      "step": 2458
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.7614225489588904,
+      "learning_rate": 9.704119990180263e-06,
+      "loss": 0.2142,
+      "step": 2459
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.461937883140961,
+      "learning_rate": 9.703814622836311e-06,
+      "loss": 0.293,
+      "step": 2460
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.9366528657295831,
+      "learning_rate": 9.703509102803834e-06,
+      "loss": 0.181,
+      "step": 2461
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.9433927398961255,
+      "learning_rate": 9.703203430092746e-06,
+      "loss": 0.2746,
+      "step": 2462
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.0306575547230274,
+      "learning_rate": 9.702897604712972e-06,
+      "loss": 0.1156,
+      "step": 2463
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.254335490404238,
+      "learning_rate": 9.702591626674436e-06,
+      "loss": 0.2829,
+      "step": 2464
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.139636722037041,
+      "learning_rate": 9.702285495987072e-06,
+      "loss": 0.1658,
+      "step": 2465
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.205894142924929,
+      "learning_rate": 9.701979212660815e-06,
+      "loss": 0.2293,
+      "step": 2466
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.253356343601962,
+      "learning_rate": 9.70167277670561e-06,
+      "loss": 0.1673,
+      "step": 2467
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.0809443812965984,
+      "learning_rate": 9.701366188131405e-06,
+      "loss": 0.2115,
+      "step": 2468
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.681535709245137,
+      "learning_rate": 9.701059446948146e-06,
+      "loss": 0.2397,
+      "step": 2469
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.1736428277992441,
+      "learning_rate": 9.700752553165797e-06,
+      "loss": 0.2632,
+      "step": 2470
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.625049671115452,
+      "learning_rate": 9.700445506794318e-06,
+      "loss": 0.2419,
+      "step": 2471
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.6853154886672987,
+      "learning_rate": 9.700138307843672e-06,
+      "loss": 0.2065,
+      "step": 2472
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.466802726561231,
+      "learning_rate": 9.699830956323835e-06,
+      "loss": 0.2113,
+      "step": 2473
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.9726429080431076,
+      "learning_rate": 9.699523452244784e-06,
+      "loss": 0.187,
+      "step": 2474
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.4190427296455639,
+      "learning_rate": 9.699215795616497e-06,
+      "loss": 0.2109,
+      "step": 2475
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.5289484409052623,
+      "learning_rate": 9.698907986448965e-06,
+      "loss": 0.2827,
+      "step": 2476
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.3123839829462915,
+      "learning_rate": 9.698600024752175e-06,
+      "loss": 0.2259,
+      "step": 2477
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.9960404250887167,
+      "learning_rate": 9.698291910536129e-06,
+      "loss": 0.1952,
+      "step": 2478
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.054629799359429,
+      "learning_rate": 9.697983643810825e-06,
+      "loss": 0.2064,
+      "step": 2479
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.245073542858452,
+      "learning_rate": 9.69767522458627e-06,
+      "loss": 0.2044,
+      "step": 2480
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.9704513415149614,
+      "learning_rate": 9.697366652872476e-06,
+      "loss": 0.2245,
+      "step": 2481
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.9447707365606385,
+      "learning_rate": 9.697057928679459e-06,
+      "loss": 0.1426,
+      "step": 2482
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.942966401050616,
+      "learning_rate": 9.69674905201724e-06,
+      "loss": 0.209,
+      "step": 2483
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 2.0261697181173903,
+      "learning_rate": 9.696440022895848e-06,
+      "loss": 0.1858,
+      "step": 2484
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.752173612960367,
+      "learning_rate": 9.696130841325311e-06,
+      "loss": 0.2547,
+      "step": 2485
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.9475191091790873,
+      "learning_rate": 9.695821507315666e-06,
+      "loss": 0.1857,
+      "step": 2486
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.0878148206602776,
+      "learning_rate": 9.695512020876957e-06,
+      "loss": 0.1876,
+      "step": 2487
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.0951236120518253,
+      "learning_rate": 9.695202382019224e-06,
+      "loss": 0.2107,
+      "step": 2488
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.395680974063356,
+      "learning_rate": 9.694892590752523e-06,
+      "loss": 0.2169,
+      "step": 2489
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.2626418982328504,
+      "learning_rate": 9.69458264708691e-06,
+      "loss": 0.2277,
+      "step": 2490
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.9075988594882396,
+      "learning_rate": 9.694272551032444e-06,
+      "loss": 0.2526,
+      "step": 2491
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.1083246273722729,
+      "learning_rate": 9.693962302599192e-06,
+      "loss": 0.162,
+      "step": 2492
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.0642821437107688,
+      "learning_rate": 9.693651901797223e-06,
+      "loss": 0.2348,
+      "step": 2493
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.319961461684073,
+      "learning_rate": 9.693341348636616e-06,
+      "loss": 0.247,
+      "step": 2494
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.3447916531790574,
+      "learning_rate": 9.693030643127448e-06,
+      "loss": 0.2423,
+      "step": 2495
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.0623740365238559,
+      "learning_rate": 9.69271978527981e-06,
+      "loss": 0.2331,
+      "step": 2496
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.0292377889275737,
+      "learning_rate": 9.692408775103786e-06,
+      "loss": 0.2531,
+      "step": 2497
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.9355765778001437,
+      "learning_rate": 9.692097612609478e-06,
+      "loss": 0.2025,
+      "step": 2498
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 13.303656707812195,
+      "learning_rate": 9.69178629780698e-06,
+      "loss": 0.2206,
+      "step": 2499
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.6390642769333885,
+      "learning_rate": 9.691474830706402e-06,
+      "loss": 0.1848,
+      "step": 2500
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.1439270273381692,
+      "learning_rate": 9.691163211317853e-06,
+      "loss": 0.2161,
+      "step": 2501
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.565578287059732,
+      "learning_rate": 9.690851439651448e-06,
+      "loss": 0.1905,
+      "step": 2502
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.2857259238528598,
+      "learning_rate": 9.690539515717309e-06,
+      "loss": 0.2416,
+      "step": 2503
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.6989789947089187,
+      "learning_rate": 9.69022743952556e-06,
+      "loss": 0.2352,
+      "step": 2504
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.3401239285702724,
+      "learning_rate": 9.689915211086331e-06,
+      "loss": 0.2318,
+      "step": 2505
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.0173238914434,
+      "learning_rate": 9.689602830409756e-06,
+      "loss": 0.1371,
+      "step": 2506
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.2815280080612563,
+      "learning_rate": 9.68929029750598e-06,
+      "loss": 0.2501,
+      "step": 2507
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.1785726205347071,
+      "learning_rate": 9.688977612385141e-06,
+      "loss": 0.2006,
+      "step": 2508
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.3891499591044174,
+      "learning_rate": 9.688664775057394e-06,
+      "loss": 0.1819,
+      "step": 2509
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.3357363376525593,
+      "learning_rate": 9.688351785532893e-06,
+      "loss": 0.2593,
+      "step": 2510
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.4173722943676201,
+      "learning_rate": 9.688038643821795e-06,
+      "loss": 0.2156,
+      "step": 2511
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.242890159237102,
+      "learning_rate": 9.687725349934268e-06,
+      "loss": 0.1869,
+      "step": 2512
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.462097258419511,
+      "learning_rate": 9.68741190388048e-06,
+      "loss": 0.2282,
+      "step": 2513
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.0472293865987228,
+      "learning_rate": 9.687098305670606e-06,
+      "loss": 0.1886,
+      "step": 2514
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.1307720124227256,
+      "learning_rate": 9.686784555314825e-06,
+      "loss": 0.212,
+      "step": 2515
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.0091141918140596,
+      "learning_rate": 9.686470652823323e-06,
+      "loss": 0.162,
+      "step": 2516
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.8031773710000416,
+      "learning_rate": 9.686156598206289e-06,
+      "loss": 0.2302,
+      "step": 2517
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.2377794133333815,
+      "learning_rate": 9.685842391473918e-06,
+      "loss": 0.2052,
+      "step": 2518
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.3358497423493652,
+      "learning_rate": 9.685528032636407e-06,
+      "loss": 0.2213,
+      "step": 2519
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.9681439715828133,
+      "learning_rate": 9.685213521703963e-06,
+      "loss": 0.1959,
+      "step": 2520
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.0470612452789878,
+      "learning_rate": 9.684898858686791e-06,
+      "loss": 0.189,
+      "step": 2521
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.505945215611147,
+      "learning_rate": 9.68458404359511e-06,
+      "loss": 0.2013,
+      "step": 2522
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 3.7155191186652496,
+      "learning_rate": 9.684269076439138e-06,
+      "loss": 0.2536,
+      "step": 2523
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.7564966065976413,
+      "learning_rate": 9.683953957229097e-06,
+      "loss": 0.3218,
+      "step": 2524
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.4927052603204387,
+      "learning_rate": 9.683638685975218e-06,
+      "loss": 0.2476,
+      "step": 2525
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.3878036284941608,
+      "learning_rate": 9.683323262687733e-06,
+      "loss": 0.1907,
+      "step": 2526
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.2911465877065524,
+      "learning_rate": 9.683007687376882e-06,
+      "loss": 0.2048,
+      "step": 2527
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.8206897311975294,
+      "learning_rate": 9.68269196005291e-06,
+      "loss": 0.2714,
+      "step": 2528
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.8059610274261435,
+      "learning_rate": 9.682376080726062e-06,
+      "loss": 0.2343,
+      "step": 2529
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.446359345494707,
+      "learning_rate": 9.682060049406594e-06,
+      "loss": 0.1953,
+      "step": 2530
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.8444750593338144,
+      "learning_rate": 9.681743866104767e-06,
+      "loss": 0.2188,
+      "step": 2531
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.1145317098733456,
+      "learning_rate": 9.681427530830838e-06,
+      "loss": 0.1511,
+      "step": 2532
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.159973868731378,
+      "learning_rate": 9.681111043595081e-06,
+      "loss": 0.2099,
+      "step": 2533
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.1493422243523088,
+      "learning_rate": 9.68079440440777e-06,
+      "loss": 0.1644,
+      "step": 2534
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.5786834238683052,
+      "learning_rate": 9.680477613279178e-06,
+      "loss": 0.2265,
+      "step": 2535
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.7219375637928527,
+      "learning_rate": 9.680160670219593e-06,
+      "loss": 0.2036,
+      "step": 2536
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 2.266076917629858,
+      "learning_rate": 9.6798435752393e-06,
+      "loss": 0.2352,
+      "step": 2537
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.1595570208028823,
+      "learning_rate": 9.679526328348594e-06,
+      "loss": 0.1778,
+      "step": 2538
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.0060984003445637,
+      "learning_rate": 9.679208929557772e-06,
+      "loss": 0.1468,
+      "step": 2539
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.9371355739237126,
+      "learning_rate": 9.67889137887714e-06,
+      "loss": 0.1649,
+      "step": 2540
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.0236895010728502,
+      "learning_rate": 9.678573676317001e-06,
+      "loss": 0.1275,
+      "step": 2541
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.2513278628726106,
+      "learning_rate": 9.67825582188767e-06,
+      "loss": 0.2283,
+      "step": 2542
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.5699813087851888,
+      "learning_rate": 9.677937815599465e-06,
+      "loss": 0.1884,
+      "step": 2543
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.400175240311999,
+      "learning_rate": 9.677619657462709e-06,
+      "loss": 0.2526,
+      "step": 2544
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.3074675311256547,
+      "learning_rate": 9.677301347487729e-06,
+      "loss": 0.2471,
+      "step": 2545
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.2046026778880772,
+      "learning_rate": 9.676982885684856e-06,
+      "loss": 0.2163,
+      "step": 2546
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.9061976471117148,
+      "learning_rate": 9.676664272064431e-06,
+      "loss": 0.1446,
+      "step": 2547
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.2810133215321744,
+      "learning_rate": 9.676345506636795e-06,
+      "loss": 0.2349,
+      "step": 2548
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.519923517446383,
+      "learning_rate": 9.676026589412292e-06,
+      "loss": 0.2155,
+      "step": 2549
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.0085480984758055,
+      "learning_rate": 9.67570752040128e-06,
+      "loss": 0.2403,
+      "step": 2550
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.0588994095170312,
+      "learning_rate": 9.675388299614113e-06,
+      "loss": 0.2003,
+      "step": 2551
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.0009328395319357,
+      "learning_rate": 9.675068927061151e-06,
+      "loss": 0.2546,
+      "step": 2552
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.3112077198085115,
+      "learning_rate": 9.674749402752765e-06,
+      "loss": 0.2262,
+      "step": 2553
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.0042665703265616,
+      "learning_rate": 9.674429726699324e-06,
+      "loss": 0.1928,
+      "step": 2554
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.3780066281472074,
+      "learning_rate": 9.674109898911205e-06,
+      "loss": 0.2569,
+      "step": 2555
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.4634099848199285,
+      "learning_rate": 9.673789919398796e-06,
+      "loss": 0.2421,
+      "step": 2556
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.3728578316630466,
+      "learning_rate": 9.673469788172474e-06,
+      "loss": 0.2472,
+      "step": 2557
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 2.3109247151834276,
+      "learning_rate": 9.673149505242637e-06,
+      "loss": 0.3193,
+      "step": 2558
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.059035518710016,
+      "learning_rate": 9.672829070619679e-06,
+      "loss": 0.1773,
+      "step": 2559
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.0614304418458889,
+      "learning_rate": 9.672508484314001e-06,
+      "loss": 0.1684,
+      "step": 2560
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.4837643195355519,
+      "learning_rate": 9.672187746336012e-06,
+      "loss": 0.196,
+      "step": 2561
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.3874262267660487,
+      "learning_rate": 9.671866856696123e-06,
+      "loss": 0.233,
+      "step": 2562
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.4048925464143889,
+      "learning_rate": 9.671545815404747e-06,
+      "loss": 0.1931,
+      "step": 2563
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.1089185032389957,
+      "learning_rate": 9.671224622472311e-06,
+      "loss": 0.2202,
+      "step": 2564
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.4968900727566383,
+      "learning_rate": 9.670903277909236e-06,
+      "loss": 0.2652,
+      "step": 2565
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.2602614558103522,
+      "learning_rate": 9.670581781725955e-06,
+      "loss": 0.194,
+      "step": 2566
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.2636214840259217,
+      "learning_rate": 9.670260133932902e-06,
+      "loss": 0.2441,
+      "step": 2567
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.1867630233944062,
+      "learning_rate": 9.66993833454052e-06,
+      "loss": 0.1773,
+      "step": 2568
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.789362719803808,
+      "learning_rate": 9.669616383559255e-06,
+      "loss": 0.2547,
+      "step": 2569
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.3865863568949108,
+      "learning_rate": 9.669294280999558e-06,
+      "loss": 0.1918,
+      "step": 2570
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.0051159935706528,
+      "learning_rate": 9.668972026871882e-06,
+      "loss": 0.2179,
+      "step": 2571
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.5635618607116848,
+      "learning_rate": 9.668649621186687e-06,
+      "loss": 0.2608,
+      "step": 2572
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.3520490523052078,
+      "learning_rate": 9.668327063954444e-06,
+      "loss": 0.2489,
+      "step": 2573
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.5361323577918988,
+      "learning_rate": 9.668004355185618e-06,
+      "loss": 0.2245,
+      "step": 2574
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.5042868008866892,
+      "learning_rate": 9.667681494890689e-06,
+      "loss": 0.2463,
+      "step": 2575
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.0734852965859443,
+      "learning_rate": 9.667358483080132e-06,
+      "loss": 0.1863,
+      "step": 2576
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.0799805940323914,
+      "learning_rate": 9.667035319764436e-06,
+      "loss": 0.1804,
+      "step": 2577
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.9754286611593694,
+      "learning_rate": 9.666712004954089e-06,
+      "loss": 0.1981,
+      "step": 2578
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.1638188355118595,
+      "learning_rate": 9.666388538659588e-06,
+      "loss": 0.2219,
+      "step": 2579
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.1071278116059777,
+      "learning_rate": 9.66606492089143e-06,
+      "loss": 0.1995,
+      "step": 2580
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.9020146177249857,
+      "learning_rate": 9.665741151660123e-06,
+      "loss": 0.2175,
+      "step": 2581
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.0551372438954871,
+      "learning_rate": 9.665417230976175e-06,
+      "loss": 0.1777,
+      "step": 2582
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.0005134255417816,
+      "learning_rate": 9.6650931588501e-06,
+      "loss": 0.2375,
+      "step": 2583
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.0702368890549152,
+      "learning_rate": 9.664768935292418e-06,
+      "loss": 0.1899,
+      "step": 2584
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.3199803672433645,
+      "learning_rate": 9.664444560313655e-06,
+      "loss": 0.2644,
+      "step": 2585
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.0159451548605665,
+      "learning_rate": 9.66412003392434e-06,
+      "loss": 0.2508,
+      "step": 2586
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.3267863595834801,
+      "learning_rate": 9.663795356135006e-06,
+      "loss": 0.182,
+      "step": 2587
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.3752181356969027,
+      "learning_rate": 9.663470526956193e-06,
+      "loss": 0.2396,
+      "step": 2588
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.093559967289896,
+      "learning_rate": 9.663145546398444e-06,
+      "loss": 0.2504,
+      "step": 2589
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 2.397039496833645,
+      "learning_rate": 9.66282041447231e-06,
+      "loss": 0.2243,
+      "step": 2590
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.4787365359074687,
+      "learning_rate": 9.662495131188343e-06,
+      "loss": 0.2312,
+      "step": 2591
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.4307833553482852,
+      "learning_rate": 9.662169696557104e-06,
+      "loss": 0.2559,
+      "step": 2592
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.1455104718905098,
+      "learning_rate": 9.661844110589155e-06,
+      "loss": 0.2179,
+      "step": 2593
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.0206174253158136,
+      "learning_rate": 9.661518373295067e-06,
+      "loss": 0.2555,
+      "step": 2594
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.9019001720937813,
+      "learning_rate": 9.66119248468541e-06,
+      "loss": 0.2122,
+      "step": 2595
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.9953005194130091,
+      "learning_rate": 9.660866444770766e-06,
+      "loss": 0.2533,
+      "step": 2596
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.0437655494867968,
+      "learning_rate": 9.660540253561716e-06,
+      "loss": 0.1892,
+      "step": 2597
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.1282020107993123,
+      "learning_rate": 9.66021391106885e-06,
+      "loss": 0.2195,
+      "step": 2598
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.1967912421787148,
+      "learning_rate": 9.65988741730276e-06,
+      "loss": 0.2301,
+      "step": 2599
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.2396169627527138,
+      "learning_rate": 9.659560772274045e-06,
+      "loss": 0.3022,
+      "step": 2600
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.1391982944087196,
+      "learning_rate": 9.659233975993308e-06,
+      "loss": 0.2134,
+      "step": 2601
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.9829002148674573,
+      "learning_rate": 9.658907028471158e-06,
+      "loss": 0.184,
+      "step": 2602
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.2323306574367217,
+      "learning_rate": 9.658579929718207e-06,
+      "loss": 0.2171,
+      "step": 2603
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.6194938152511802,
+      "learning_rate": 9.65825267974507e-06,
+      "loss": 0.2157,
+      "step": 2604
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.2480396813795736,
+      "learning_rate": 9.657925278562375e-06,
+      "loss": 0.2042,
+      "step": 2605
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.516878563319148,
+      "learning_rate": 9.657597726180743e-06,
+      "loss": 0.224,
+      "step": 2606
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.7412720507446215,
+      "learning_rate": 9.657270022610814e-06,
+      "loss": 0.2059,
+      "step": 2607
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 5.14574418653523,
+      "learning_rate": 9.656942167863221e-06,
+      "loss": 0.2516,
+      "step": 2608
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 3.0358068193629864,
+      "learning_rate": 9.656614161948608e-06,
+      "loss": 0.1962,
+      "step": 2609
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.6473583716375269,
+      "learning_rate": 9.656286004877622e-06,
+      "loss": 0.2328,
+      "step": 2610
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.7794751141880196,
+      "learning_rate": 9.655957696660914e-06,
+      "loss": 0.1931,
+      "step": 2611
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.7616111739443059,
+      "learning_rate": 9.65562923730914e-06,
+      "loss": 0.1991,
+      "step": 2612
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.030042808717151,
+      "learning_rate": 9.655300626832965e-06,
+      "loss": 0.1538,
+      "step": 2613
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.3956152979338006,
+      "learning_rate": 9.654971865243055e-06,
+      "loss": 0.2063,
+      "step": 2614
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.3602504281859416,
+      "learning_rate": 9.654642952550082e-06,
+      "loss": 0.2215,
+      "step": 2615
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.069044366913865,
+      "learning_rate": 9.654313888764721e-06,
+      "loss": 0.204,
+      "step": 2616
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.140834602928686,
+      "learning_rate": 9.653984673897657e-06,
+      "loss": 0.1958,
+      "step": 2617
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.2088638693941811,
+      "learning_rate": 9.653655307959572e-06,
+      "loss": 0.1786,
+      "step": 2618
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.379193416551652,
+      "learning_rate": 9.65332579096116e-06,
+      "loss": 0.2151,
+      "step": 2619
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.2166099787500753,
+      "learning_rate": 9.652996122913117e-06,
+      "loss": 0.2176,
+      "step": 2620
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.3653590441471446,
+      "learning_rate": 9.652666303826144e-06,
+      "loss": 0.2177,
+      "step": 2621
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.361295349405193,
+      "learning_rate": 9.652336333710948e-06,
+      "loss": 0.2271,
+      "step": 2622
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.3411643343331454,
+      "learning_rate": 9.652006212578239e-06,
+      "loss": 0.1809,
+      "step": 2623
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.8957033786199372,
+      "learning_rate": 9.651675940438734e-06,
+      "loss": 0.2004,
+      "step": 2624
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.8534105692353996,
+      "learning_rate": 9.651345517303151e-06,
+      "loss": 0.2073,
+      "step": 2625
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.453733468076527,
+      "learning_rate": 9.65101494318222e-06,
+      "loss": 0.2423,
+      "step": 2626
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.1793295999480329,
+      "learning_rate": 9.650684218086667e-06,
+      "loss": 0.2404,
+      "step": 2627
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.8215269522124454,
+      "learning_rate": 9.650353342027232e-06,
+      "loss": 0.1805,
+      "step": 2628
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.8358260100581245,
+      "learning_rate": 9.650022315014652e-06,
+      "loss": 0.2686,
+      "step": 2629
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.9906156340165325,
+      "learning_rate": 9.649691137059676e-06,
+      "loss": 0.221,
+      "step": 2630
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 2.6866416447549426,
+      "learning_rate": 9.649359808173052e-06,
+      "loss": 0.2501,
+      "step": 2631
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.751847804112577,
+      "learning_rate": 9.649028328365533e-06,
+      "loss": 0.1748,
+      "step": 2632
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.3619064897605426,
+      "learning_rate": 9.648696697647881e-06,
+      "loss": 0.1931,
+      "step": 2633
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.3918373363265477,
+      "learning_rate": 9.64836491603086e-06,
+      "loss": 0.2533,
+      "step": 2634
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.127333231511418,
+      "learning_rate": 9.648032983525245e-06,
+      "loss": 0.2067,
+      "step": 2635
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.277212760345078,
+      "learning_rate": 9.647700900141804e-06,
+      "loss": 0.2712,
+      "step": 2636
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.549735848352821,
+      "learning_rate": 9.647368665891319e-06,
+      "loss": 0.2384,
+      "step": 2637
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.3346887532873393,
+      "learning_rate": 9.647036280784574e-06,
+      "loss": 0.2227,
+      "step": 2638
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.9519083336555962,
+      "learning_rate": 9.64670374483236e-06,
+      "loss": 0.1946,
+      "step": 2639
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.2860511965333925,
+      "learning_rate": 9.64637105804547e-06,
+      "loss": 0.2317,
+      "step": 2640
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.980381738555721,
+      "learning_rate": 9.646038220434705e-06,
+      "loss": 0.1583,
+      "step": 2641
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.1117950725560914,
+      "learning_rate": 9.645705232010868e-06,
+      "loss": 0.2145,
+      "step": 2642
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.9989074292885287,
+      "learning_rate": 9.645372092784764e-06,
+      "loss": 0.1918,
+      "step": 2643
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.151553917804969,
+      "learning_rate": 9.645038802767213e-06,
+      "loss": 0.2161,
+      "step": 2644
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.3616951429910975,
+      "learning_rate": 9.644705361969032e-06,
+      "loss": 0.2198,
+      "step": 2645
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.0460571966415746,
+      "learning_rate": 9.644371770401042e-06,
+      "loss": 0.1702,
+      "step": 2646
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.5585535828235522,
+      "learning_rate": 9.644038028074077e-06,
+      "loss": 0.2529,
+      "step": 2647
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.1153650190590199,
+      "learning_rate": 9.643704134998964e-06,
+      "loss": 0.2059,
+      "step": 2648
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.080907239567574,
+      "learning_rate": 9.643370091186545e-06,
+      "loss": 0.218,
+      "step": 2649
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.9240632322567802,
+      "learning_rate": 9.643035896647663e-06,
+      "loss": 0.1628,
+      "step": 2650
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.144562024298424,
+      "learning_rate": 9.642701551393168e-06,
+      "loss": 0.2244,
+      "step": 2651
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.4780566512227675,
+      "learning_rate": 9.642367055433908e-06,
+      "loss": 0.1869,
+      "step": 2652
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.3202718553550823,
+      "learning_rate": 9.642032408780744e-06,
+      "loss": 0.2364,
+      "step": 2653
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.4301795907326365,
+      "learning_rate": 9.641697611444541e-06,
+      "loss": 0.2563,
+      "step": 2654
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.9845731210394244,
+      "learning_rate": 9.641362663436162e-06,
+      "loss": 0.1835,
+      "step": 2655
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.9695652331007132,
+      "learning_rate": 9.641027564766484e-06,
+      "loss": 0.2145,
+      "step": 2656
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.6276948824715345,
+      "learning_rate": 9.640692315446381e-06,
+      "loss": 0.2474,
+      "step": 2657
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.0979762976772298,
+      "learning_rate": 9.640356915486738e-06,
+      "loss": 0.2174,
+      "step": 2658
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.9952643476036547,
+      "learning_rate": 9.64002136489844e-06,
+      "loss": 0.2313,
+      "step": 2659
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.080197857088931,
+      "learning_rate": 9.63968566369238e-06,
+      "loss": 0.2058,
+      "step": 2660
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.814714275544913,
+      "learning_rate": 9.639349811879455e-06,
+      "loss": 0.1523,
+      "step": 2661
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.7127419088001645,
+      "learning_rate": 9.639013809470567e-06,
+      "loss": 0.1944,
+      "step": 2662
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.5517883441819047,
+      "learning_rate": 9.638677656476625e-06,
+      "loss": 0.207,
+      "step": 2663
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.4209690758099751,
+      "learning_rate": 9.638341352908538e-06,
+      "loss": 0.2135,
+      "step": 2664
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.8714966390005268,
+      "learning_rate": 9.638004898777225e-06,
+      "loss": 0.1932,
+      "step": 2665
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.1323040018254946,
+      "learning_rate": 9.637668294093604e-06,
+      "loss": 0.1969,
+      "step": 2666
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.0568607827239263,
+      "learning_rate": 9.637331538868603e-06,
+      "loss": 0.1974,
+      "step": 2667
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.2688988152032383,
+      "learning_rate": 9.636994633113155e-06,
+      "loss": 0.2025,
+      "step": 2668
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.5709152583135046,
+      "learning_rate": 9.636657576838195e-06,
+      "loss": 0.2167,
+      "step": 2669
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.2935216231615152,
+      "learning_rate": 9.636320370054662e-06,
+      "loss": 0.1821,
+      "step": 2670
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.519792328399277,
+      "learning_rate": 9.635983012773505e-06,
+      "loss": 0.2057,
+      "step": 2671
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.2035562683592307,
+      "learning_rate": 9.635645505005672e-06,
+      "loss": 0.1556,
+      "step": 2672
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.3533094037101192,
+      "learning_rate": 9.635307846762124e-06,
+      "loss": 0.2026,
+      "step": 2673
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.474518175652278,
+      "learning_rate": 9.634970038053815e-06,
+      "loss": 0.1437,
+      "step": 2674
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.4687365343612893,
+      "learning_rate": 9.634632078891712e-06,
+      "loss": 0.1986,
+      "step": 2675
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.2046522040088528,
+      "learning_rate": 9.634293969286789e-06,
+      "loss": 0.202,
+      "step": 2676
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.2432954843101074,
+      "learning_rate": 9.633955709250018e-06,
+      "loss": 0.2667,
+      "step": 2677
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 5.947363801593701,
+      "learning_rate": 9.63361729879238e-06,
+      "loss": 0.2556,
+      "step": 2678
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.257231381886958,
+      "learning_rate": 9.63327873792486e-06,
+      "loss": 0.2015,
+      "step": 2679
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.1104884268790123,
+      "learning_rate": 9.632940026658448e-06,
+      "loss": 0.1885,
+      "step": 2680
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.0584745762314318,
+      "learning_rate": 9.632601165004137e-06,
+      "loss": 0.2348,
+      "step": 2681
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.2449265756748724,
+      "learning_rate": 9.63226215297293e-06,
+      "loss": 0.1781,
+      "step": 2682
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.1910484276918225,
+      "learning_rate": 9.63192299057583e-06,
+      "loss": 0.1938,
+      "step": 2683
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.575543428421885,
+      "learning_rate": 9.631583677823846e-06,
+      "loss": 0.2025,
+      "step": 2684
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.3158222963449258,
+      "learning_rate": 9.63124421472799e-06,
+      "loss": 0.2049,
+      "step": 2685
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.943624553128122,
+      "learning_rate": 9.630904601299286e-06,
+      "loss": 0.2195,
+      "step": 2686
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.8424365407931331,
+      "learning_rate": 9.630564837548755e-06,
+      "loss": 0.1197,
+      "step": 2687
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.3528201117624161,
+      "learning_rate": 9.630224923487427e-06,
+      "loss": 0.1943,
+      "step": 2688
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.3109859279717349,
+      "learning_rate": 9.629884859126337e-06,
+      "loss": 0.2111,
+      "step": 2689
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.142184908768706,
+      "learning_rate": 9.629544644476521e-06,
+      "loss": 0.2546,
+      "step": 2690
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.462525213871941,
+      "learning_rate": 9.629204279549023e-06,
+      "loss": 0.2979,
+      "step": 2691
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.033261472356134,
+      "learning_rate": 9.62886376435489e-06,
+      "loss": 0.1852,
+      "step": 2692
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.1962653550893483,
+      "learning_rate": 9.62852309890518e-06,
+      "loss": 0.2151,
+      "step": 2693
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.2524286429366491,
+      "learning_rate": 9.628182283210949e-06,
+      "loss": 0.2186,
+      "step": 2694
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.2566337926771973,
+      "learning_rate": 9.62784131728326e-06,
+      "loss": 0.2447,
+      "step": 2695
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.37588608974293,
+      "learning_rate": 9.627500201133178e-06,
+      "loss": 0.2257,
+      "step": 2696
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.5471050207027377,
+      "learning_rate": 9.627158934771781e-06,
+      "loss": 0.2532,
+      "step": 2697
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.2748165693470144,
+      "learning_rate": 9.626817518210143e-06,
+      "loss": 0.228,
+      "step": 2698
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.284627857256252,
+      "learning_rate": 9.626475951459348e-06,
+      "loss": 0.172,
+      "step": 2699
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.1839406973265456,
+      "learning_rate": 9.626134234530483e-06,
+      "loss": 0.2244,
+      "step": 2700
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.9262565648081801,
+      "learning_rate": 9.62579236743464e-06,
+      "loss": 0.1929,
+      "step": 2701
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.7672573325393182,
+      "learning_rate": 9.625450350182919e-06,
+      "loss": 0.1903,
+      "step": 2702
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.210937503076369,
+      "learning_rate": 9.625108182786417e-06,
+      "loss": 0.2027,
+      "step": 2703
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.3129104613894267,
+      "learning_rate": 9.624765865256245e-06,
+      "loss": 0.2322,
+      "step": 2704
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.1590051591923318,
+      "learning_rate": 9.624423397603513e-06,
+      "loss": 0.1891,
+      "step": 2705
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.0260308120083985,
+      "learning_rate": 9.62408077983934e-06,
+      "loss": 0.2177,
+      "step": 2706
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.1581529249405664,
+      "learning_rate": 9.623738011974845e-06,
+      "loss": 0.217,
+      "step": 2707
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.021751368003881,
+      "learning_rate": 9.623395094021155e-06,
+      "loss": 0.2487,
+      "step": 2708
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.3941760425552578,
+      "learning_rate": 9.623052025989402e-06,
+      "loss": 0.2101,
+      "step": 2709
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.5973680324848765,
+      "learning_rate": 9.622708807890722e-06,
+      "loss": 0.2336,
+      "step": 2710
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.2590968848414268,
+      "learning_rate": 9.622365439736254e-06,
+      "loss": 0.2118,
+      "step": 2711
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.8609680960890429,
+      "learning_rate": 9.622021921537147e-06,
+      "loss": 0.204,
+      "step": 2712
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.9204253602732607,
+      "learning_rate": 9.62167825330455e-06,
+      "loss": 0.1995,
+      "step": 2713
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.2478143985914407,
+      "learning_rate": 9.62133443504962e-06,
+      "loss": 0.186,
+      "step": 2714
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.2043367142325423,
+      "learning_rate": 9.620990466783517e-06,
+      "loss": 0.1922,
+      "step": 2715
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.0679565765256323,
+      "learning_rate": 9.620646348517407e-06,
+      "loss": 0.216,
+      "step": 2716
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.7070376136364374,
+      "learning_rate": 9.620302080262457e-06,
+      "loss": 0.2598,
+      "step": 2717
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.9920844039520026,
+      "learning_rate": 9.619957662029847e-06,
+      "loss": 0.201,
+      "step": 2718
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.0509513695998631,
+      "learning_rate": 9.619613093830753e-06,
+      "loss": 0.202,
+      "step": 2719
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.1746406115573491,
+      "learning_rate": 9.619268375676361e-06,
+      "loss": 0.1746,
+      "step": 2720
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.613498774681418,
+      "learning_rate": 9.618923507577863e-06,
+      "loss": 0.208,
+      "step": 2721
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.1968143983456152,
+      "learning_rate": 9.61857848954645e-06,
+      "loss": 0.1872,
+      "step": 2722
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.0706607746208925,
+      "learning_rate": 9.618233321593325e-06,
+      "loss": 0.1388,
+      "step": 2723
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.011098290337284,
+      "learning_rate": 9.617888003729691e-06,
+      "loss": 0.1914,
+      "step": 2724
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.236461209641587,
+      "learning_rate": 9.617542535966756e-06,
+      "loss": 0.188,
+      "step": 2725
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.2484725739537783,
+      "learning_rate": 9.617196918315736e-06,
+      "loss": 0.2019,
+      "step": 2726
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.1213376057477242,
+      "learning_rate": 9.616851150787847e-06,
+      "loss": 0.2427,
+      "step": 2727
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.3966370745850232,
+      "learning_rate": 9.616505233394317e-06,
+      "loss": 0.188,
+      "step": 2728
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.2572580620110294,
+      "learning_rate": 9.61615916614637e-06,
+      "loss": 0.2463,
+      "step": 2729
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.2648502085793567,
+      "learning_rate": 9.615812949055243e-06,
+      "loss": 0.2439,
+      "step": 2730
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.1657971252497583,
+      "learning_rate": 9.615466582132175e-06,
+      "loss": 0.2746,
+      "step": 2731
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.291345568579551,
+      "learning_rate": 9.615120065388406e-06,
+      "loss": 0.2718,
+      "step": 2732
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.3147937183188325,
+      "learning_rate": 9.614773398835186e-06,
+      "loss": 0.2422,
+      "step": 2733
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.1819284325034451,
+      "learning_rate": 9.614426582483768e-06,
+      "loss": 0.2014,
+      "step": 2734
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.4581177012787785,
+      "learning_rate": 9.61407961634541e-06,
+      "loss": 0.2017,
+      "step": 2735
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.338485769175894,
+      "learning_rate": 9.613732500431375e-06,
+      "loss": 0.1872,
+      "step": 2736
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.1959531195582214,
+      "learning_rate": 9.613385234752927e-06,
+      "loss": 0.1778,
+      "step": 2737
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.8626031086410426,
+      "learning_rate": 9.613037819321343e-06,
+      "loss": 0.2322,
+      "step": 2738
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.084680710766898,
+      "learning_rate": 9.6126902541479e-06,
+      "loss": 0.2151,
+      "step": 2739
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.1547068401983103,
+      "learning_rate": 9.612342539243877e-06,
+      "loss": 0.2074,
+      "step": 2740
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.3933135093281817,
+      "learning_rate": 9.611994674620564e-06,
+      "loss": 0.2077,
+      "step": 2741
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.1424782815330743,
+      "learning_rate": 9.61164666028925e-06,
+      "loss": 0.2259,
+      "step": 2742
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.9878915764867048,
+      "learning_rate": 9.611298496261237e-06,
+      "loss": 0.1518,
+      "step": 2743
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.4552718673603382,
+      "learning_rate": 9.610950182547821e-06,
+      "loss": 0.2133,
+      "step": 2744
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.3875063271528614,
+      "learning_rate": 9.61060171916031e-06,
+      "loss": 0.1581,
+      "step": 2745
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.3724259319539696,
+      "learning_rate": 9.610253106110018e-06,
+      "loss": 0.256,
+      "step": 2746
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.0638185972179424,
+      "learning_rate": 9.609904343408257e-06,
+      "loss": 0.1963,
+      "step": 2747
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.0053403163235985,
+      "learning_rate": 9.609555431066352e-06,
+      "loss": 0.184,
+      "step": 2748
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.9787434371762317,
+      "learning_rate": 9.609206369095626e-06,
+      "loss": 0.2052,
+      "step": 2749
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.1552221894687313,
+      "learning_rate": 9.608857157507409e-06,
+      "loss": 0.1873,
+      "step": 2750
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.2554726427892642,
+      "learning_rate": 9.60850779631304e-06,
+      "loss": 0.2349,
+      "step": 2751
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.3518657158152763,
+      "learning_rate": 9.608158285523858e-06,
+      "loss": 0.2418,
+      "step": 2752
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.1702948693703263,
+      "learning_rate": 9.60780862515121e-06,
+      "loss": 0.211,
+      "step": 2753
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.2316411881304905,
+      "learning_rate": 9.60745881520644e-06,
+      "loss": 0.2143,
+      "step": 2754
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.7151564386862386,
+      "learning_rate": 9.60710885570091e-06,
+      "loss": 0.2057,
+      "step": 2755
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.2213990691931491,
+      "learning_rate": 9.60675874664598e-06,
+      "loss": 0.2474,
+      "step": 2756
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.1432089620975738,
+      "learning_rate": 9.606408488053007e-06,
+      "loss": 0.1852,
+      "step": 2757
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.4866750515295433,
+      "learning_rate": 9.606058079933369e-06,
+      "loss": 0.2309,
+      "step": 2758
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.417409184812272,
+      "learning_rate": 9.605707522298434e-06,
+      "loss": 0.2019,
+      "step": 2759
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.428061368263401,
+      "learning_rate": 9.605356815159588e-06,
+      "loss": 0.2298,
+      "step": 2760
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.3355118829227262,
+      "learning_rate": 9.605005958528208e-06,
+      "loss": 0.206,
+      "step": 2761
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.1143422274496098,
+      "learning_rate": 9.604654952415688e-06,
+      "loss": 0.1929,
+      "step": 2762
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.004414607817747,
+      "learning_rate": 9.604303796833421e-06,
+      "loss": 0.1434,
+      "step": 2763
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.977140147074685,
+      "learning_rate": 9.603952491792805e-06,
+      "loss": 0.209,
+      "step": 2764
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.3161928325075258,
+      "learning_rate": 9.603601037305244e-06,
+      "loss": 0.2009,
+      "step": 2765
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.882154610972088,
+      "learning_rate": 9.603249433382145e-06,
+      "loss": 0.1489,
+      "step": 2766
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 3.9027421994169518,
+      "learning_rate": 9.602897680034922e-06,
+      "loss": 0.1971,
+      "step": 2767
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.3804549268344564,
+      "learning_rate": 9.602545777274995e-06,
+      "loss": 0.2839,
+      "step": 2768
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.9453811433397182,
+      "learning_rate": 9.602193725113785e-06,
+      "loss": 0.2044,
+      "step": 2769
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.9111640532310861,
+      "learning_rate": 9.601841523562719e-06,
+      "loss": 0.2026,
+      "step": 2770
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.3187242914186588,
+      "learning_rate": 9.601489172633232e-06,
+      "loss": 0.2354,
+      "step": 2771
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.2212191139686663,
+      "learning_rate": 9.60113667233676e-06,
+      "loss": 0.2098,
+      "step": 2772
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.708205912265209,
+      "learning_rate": 9.600784022684746e-06,
+      "loss": 0.2434,
+      "step": 2773
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.6517439761466945,
+      "learning_rate": 9.600431223688635e-06,
+      "loss": 0.2697,
+      "step": 2774
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.1573453689302247,
+      "learning_rate": 9.600078275359884e-06,
+      "loss": 0.2298,
+      "step": 2775
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.0006785549268884,
+      "learning_rate": 9.599725177709946e-06,
+      "loss": 0.2199,
+      "step": 2776
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.2681545774809262,
+      "learning_rate": 9.599371930750283e-06,
+      "loss": 0.2224,
+      "step": 2777
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.2023782030204804,
+      "learning_rate": 9.599018534492364e-06,
+      "loss": 0.175,
+      "step": 2778
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.6424944943650754,
+      "learning_rate": 9.598664988947658e-06,
+      "loss": 0.2312,
+      "step": 2779
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.0777641812136816,
+      "learning_rate": 9.59831129412764e-06,
+      "loss": 0.2296,
+      "step": 2780
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.9170506695348456,
+      "learning_rate": 9.597957450043795e-06,
+      "loss": 0.2169,
+      "step": 2781
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.8740180287113871,
+      "learning_rate": 9.597603456707606e-06,
+      "loss": 0.2562,
+      "step": 2782
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.1326892127988295,
+      "learning_rate": 9.597249314130569e-06,
+      "loss": 0.2141,
+      "step": 2783
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.8372065145548138,
+      "learning_rate": 9.596895022324171e-06,
+      "loss": 0.1608,
+      "step": 2784
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.1981669047116146,
+      "learning_rate": 9.596540581299918e-06,
+      "loss": 0.1673,
+      "step": 2785
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.2520702920203846,
+      "learning_rate": 9.596185991069316e-06,
+      "loss": 0.1815,
+      "step": 2786
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.5810369710889793,
+      "learning_rate": 9.595831251643872e-06,
+      "loss": 0.2331,
+      "step": 2787
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.3650470244906434,
+      "learning_rate": 9.595476363035105e-06,
+      "loss": 0.2132,
+      "step": 2788
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.1829449855325547,
+      "learning_rate": 9.595121325254531e-06,
+      "loss": 0.2204,
+      "step": 2789
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.020321307867218,
+      "learning_rate": 9.594766138313677e-06,
+      "loss": 0.2019,
+      "step": 2790
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.0990131618357666,
+      "learning_rate": 9.594410802224072e-06,
+      "loss": 0.1697,
+      "step": 2791
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.1510590504143432,
+      "learning_rate": 9.594055316997251e-06,
+      "loss": 0.2078,
+      "step": 2792
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.241904547017199,
+      "learning_rate": 9.59369968264475e-06,
+      "loss": 0.2092,
+      "step": 2793
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.67103545108166,
+      "learning_rate": 9.593343899178118e-06,
+      "loss": 0.1686,
+      "step": 2794
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.3221584597819793,
+      "learning_rate": 9.592987966608902e-06,
+      "loss": 0.1628,
+      "step": 2795
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.0164411878188768,
+      "learning_rate": 9.592631884948654e-06,
+      "loss": 0.1598,
+      "step": 2796
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.9724192541244493,
+      "learning_rate": 9.592275654208936e-06,
+      "loss": 0.2477,
+      "step": 2797
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.2048800806505235,
+      "learning_rate": 9.591919274401306e-06,
+      "loss": 0.1908,
+      "step": 2798
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.472637053701014,
+      "learning_rate": 9.591562745537339e-06,
+      "loss": 0.25,
+      "step": 2799
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.4150374315777654,
+      "learning_rate": 9.591206067628605e-06,
+      "loss": 0.1943,
+      "step": 2800
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.9743224669619072,
+      "learning_rate": 9.59084924068668e-06,
+      "loss": 0.1759,
+      "step": 2801
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.3380110303619415,
+      "learning_rate": 9.59049226472315e-06,
+      "loss": 0.2626,
+      "step": 2802
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.411943673928723,
+      "learning_rate": 9.590135139749602e-06,
+      "loss": 0.1846,
+      "step": 2803
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.0362240511158307,
+      "learning_rate": 9.589777865777627e-06,
+      "loss": 0.1846,
+      "step": 2804
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.3904848081666923,
+      "learning_rate": 9.589420442818823e-06,
+      "loss": 0.2474,
+      "step": 2805
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.160714992684107,
+      "learning_rate": 9.58906287088479e-06,
+      "loss": 0.1479,
+      "step": 2806
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.236048832603948,
+      "learning_rate": 9.588705149987142e-06,
+      "loss": 0.2469,
+      "step": 2807
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.3005419147961954,
+      "learning_rate": 9.588347280137483e-06,
+      "loss": 0.25,
+      "step": 2808
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.4338640171335746,
+      "learning_rate": 9.587989261347435e-06,
+      "loss": 0.247,
+      "step": 2809
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.2106314749266385,
+      "learning_rate": 9.587631093628616e-06,
+      "loss": 0.1895,
+      "step": 2810
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.9929435995414861,
+      "learning_rate": 9.587272776992654e-06,
+      "loss": 0.1905,
+      "step": 2811
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.2649428330572947,
+      "learning_rate": 9.586914311451177e-06,
+      "loss": 0.2207,
+      "step": 2812
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.2517633480247825,
+      "learning_rate": 9.586555697015827e-06,
+      "loss": 0.2159,
+      "step": 2813
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.9066938534207228,
+      "learning_rate": 9.58619693369824e-06,
+      "loss": 0.1748,
+      "step": 2814
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.1109744659006495,
+      "learning_rate": 9.585838021510064e-06,
+      "loss": 0.2171,
+      "step": 2815
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.471150125151635,
+      "learning_rate": 9.585478960462949e-06,
+      "loss": 0.2199,
+      "step": 2816
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.609778571221423,
+      "learning_rate": 9.58511975056855e-06,
+      "loss": 0.2256,
+      "step": 2817
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.0247430345904516,
+      "learning_rate": 9.584760391838528e-06,
+      "loss": 0.1911,
+      "step": 2818
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.3612356127113507,
+      "learning_rate": 9.584400884284546e-06,
+      "loss": 0.2252,
+      "step": 2819
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.451511084135647,
+      "learning_rate": 9.584041227918274e-06,
+      "loss": 0.2174,
+      "step": 2820
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.2402086600694056,
+      "learning_rate": 9.58368142275139e-06,
+      "loss": 0.1961,
+      "step": 2821
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.2304557012441324,
+      "learning_rate": 9.583321468795571e-06,
+      "loss": 0.1785,
+      "step": 2822
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.1845341682423296,
+      "learning_rate": 9.582961366062501e-06,
+      "loss": 0.1835,
+      "step": 2823
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.2718146923480176,
+      "learning_rate": 9.582601114563869e-06,
+      "loss": 0.1572,
+      "step": 2824
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.273725936262716,
+      "learning_rate": 9.58224071431137e-06,
+      "loss": 0.2045,
+      "step": 2825
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.3664935285329158,
+      "learning_rate": 9.581880165316704e-06,
+      "loss": 0.2291,
+      "step": 2826
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.0883337655852794,
+      "learning_rate": 9.58151946759157e-06,
+      "loss": 0.2148,
+      "step": 2827
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.3853211210104368,
+      "learning_rate": 9.581158621147682e-06,
+      "loss": 0.1849,
+      "step": 2828
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.7018135023598497,
+      "learning_rate": 9.580797625996752e-06,
+      "loss": 0.2469,
+      "step": 2829
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.1366524054315148,
+      "learning_rate": 9.580436482150494e-06,
+      "loss": 0.1993,
+      "step": 2830
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.4813529320858736,
+      "learning_rate": 9.580075189620634e-06,
+      "loss": 0.2576,
+      "step": 2831
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.4350811572668203,
+      "learning_rate": 9.579713748418901e-06,
+      "loss": 0.2213,
+      "step": 2832
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.3160923759089151,
+      "learning_rate": 9.579352158557026e-06,
+      "loss": 0.2353,
+      "step": 2833
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.0742065931846072,
+      "learning_rate": 9.578990420046746e-06,
+      "loss": 0.1601,
+      "step": 2834
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.2595378122678131,
+      "learning_rate": 9.578628532899805e-06,
+      "loss": 0.1642,
+      "step": 2835
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.1863355354376581,
+      "learning_rate": 9.578266497127948e-06,
+      "loss": 0.1937,
+      "step": 2836
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.4254557799828353,
+      "learning_rate": 9.577904312742927e-06,
+      "loss": 0.2487,
+      "step": 2837
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.2677996916426002,
+      "learning_rate": 9.577541979756501e-06,
+      "loss": 0.2209,
+      "step": 2838
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.2371538788403387,
+      "learning_rate": 9.577179498180429e-06,
+      "loss": 0.2089,
+      "step": 2839
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.9433065403582148,
+      "learning_rate": 9.57681686802648e-06,
+      "loss": 0.1951,
+      "step": 2840
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.9990211348819392,
+      "learning_rate": 9.576454089306422e-06,
+      "loss": 0.1936,
+      "step": 2841
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.9293315169847516,
+      "learning_rate": 9.576091162032033e-06,
+      "loss": 0.1902,
+      "step": 2842
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.9598947659384963,
+      "learning_rate": 9.575728086215093e-06,
+      "loss": 0.2299,
+      "step": 2843
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.1797428607582183,
+      "learning_rate": 9.575364861867389e-06,
+      "loss": 0.2178,
+      "step": 2844
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.0540504544958382,
+      "learning_rate": 9.575001489000711e-06,
+      "loss": 0.1971,
+      "step": 2845
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 5.571165642804733,
+      "learning_rate": 9.574637967626852e-06,
+      "loss": 0.2395,
+      "step": 2846
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.488686363772311,
+      "learning_rate": 9.574274297757615e-06,
+      "loss": 0.1947,
+      "step": 2847
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.289470098095448,
+      "learning_rate": 9.573910479404804e-06,
+      "loss": 0.2286,
+      "step": 2848
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.095056468790037,
+      "learning_rate": 9.573546512580228e-06,
+      "loss": 0.2269,
+      "step": 2849
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.913160523004979,
+      "learning_rate": 9.573182397295704e-06,
+      "loss": 0.2099,
+      "step": 2850
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.0228458980558608,
+      "learning_rate": 9.572818133563047e-06,
+      "loss": 0.2378,
+      "step": 2851
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.317312378693635,
+      "learning_rate": 9.572453721394087e-06,
+      "loss": 0.2613,
+      "step": 2852
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.8854725431477469,
+      "learning_rate": 9.572089160800648e-06,
+      "loss": 0.1896,
+      "step": 2853
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.3876797865147386,
+      "learning_rate": 9.571724451794564e-06,
+      "loss": 0.2301,
+      "step": 2854
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.042907571962822,
+      "learning_rate": 9.571359594387679e-06,
+      "loss": 0.1567,
+      "step": 2855
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.8446017002148332,
+      "learning_rate": 9.57099458859183e-06,
+      "loss": 0.1865,
+      "step": 2856
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.264395855383602,
+      "learning_rate": 9.57062943441887e-06,
+      "loss": 0.2303,
+      "step": 2857
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 3.3386418639494853,
+      "learning_rate": 9.570264131880648e-06,
+      "loss": 0.2092,
+      "step": 2858
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.6318149633788965,
+      "learning_rate": 9.569898680989027e-06,
+      "loss": 0.1751,
+      "step": 2859
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.086843169540333,
+      "learning_rate": 9.569533081755865e-06,
+      "loss": 0.2246,
+      "step": 2860
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.1540024247893983,
+      "learning_rate": 9.569167334193032e-06,
+      "loss": 0.2172,
+      "step": 2861
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.198026768316968,
+      "learning_rate": 9.568801438312401e-06,
+      "loss": 0.2037,
+      "step": 2862
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.9973627421946214,
+      "learning_rate": 9.568435394125845e-06,
+      "loss": 0.22,
+      "step": 2863
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.3416458361377916,
+      "learning_rate": 9.568069201645252e-06,
+      "loss": 0.2673,
+      "step": 2864
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.1583630996534104,
+      "learning_rate": 9.567702860882504e-06,
+      "loss": 0.198,
+      "step": 2865
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.3809217576499815,
+      "learning_rate": 9.567336371849496e-06,
+      "loss": 0.2003,
+      "step": 2866
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.626466718459044,
+      "learning_rate": 9.566969734558122e-06,
+      "loss": 0.2726,
+      "step": 2867
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.9999884366654859,
+      "learning_rate": 9.566602949020284e-06,
+      "loss": 0.2557,
+      "step": 2868
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.1275630619303945,
+      "learning_rate": 9.566236015247889e-06,
+      "loss": 0.2258,
+      "step": 2869
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.280184677969596,
+      "learning_rate": 9.56586893325285e-06,
+      "loss": 0.2415,
+      "step": 2870
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.0039336037527247,
+      "learning_rate": 9.565501703047075e-06,
+      "loss": 0.1575,
+      "step": 2871
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.58103891497939,
+      "learning_rate": 9.565134324642491e-06,
+      "loss": 0.2438,
+      "step": 2872
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.4057520276100302,
+      "learning_rate": 9.564766798051023e-06,
+      "loss": 0.1764,
+      "step": 2873
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.288589564178207,
+      "learning_rate": 9.564399123284599e-06,
+      "loss": 0.2169,
+      "step": 2874
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.4230799115388382,
+      "learning_rate": 9.564031300355154e-06,
+      "loss": 0.2226,
+      "step": 2875
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.1139978260967396,
+      "learning_rate": 9.563663329274629e-06,
+      "loss": 0.2434,
+      "step": 2876
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.1019038827976224,
+      "learning_rate": 9.563295210054969e-06,
+      "loss": 0.2056,
+      "step": 2877
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.68871538448999,
+      "learning_rate": 9.562926942708121e-06,
+      "loss": 0.247,
+      "step": 2878
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.4190574970516294,
+      "learning_rate": 9.56255852724604e-06,
+      "loss": 0.1947,
+      "step": 2879
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.3818766530646864,
+      "learning_rate": 9.562189963680686e-06,
+      "loss": 0.2835,
+      "step": 2880
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.9173791975657121,
+      "learning_rate": 9.561821252024024e-06,
+      "loss": 0.1927,
+      "step": 2881
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.529559786377707,
+      "learning_rate": 9.561452392288019e-06,
+      "loss": 0.2016,
+      "step": 2882
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.0689092525127182,
+      "learning_rate": 9.561083384484649e-06,
+      "loss": 0.2173,
+      "step": 2883
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.2591255691138483,
+      "learning_rate": 9.560714228625886e-06,
+      "loss": 0.1829,
+      "step": 2884
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.3788616278563393,
+      "learning_rate": 9.560344924723718e-06,
+      "loss": 0.2279,
+      "step": 2885
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.266473129072635,
+      "learning_rate": 9.559975472790133e-06,
+      "loss": 0.2182,
+      "step": 2886
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.0005809406846013,
+      "learning_rate": 9.55960587283712e-06,
+      "loss": 0.1944,
+      "step": 2887
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.1310176458602237,
+      "learning_rate": 9.559236124876678e-06,
+      "loss": 0.2243,
+      "step": 2888
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.590730883553457,
+      "learning_rate": 9.558866228920811e-06,
+      "loss": 0.2516,
+      "step": 2889
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.8936417354854107,
+      "learning_rate": 9.558496184981525e-06,
+      "loss": 0.1863,
+      "step": 2890
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.8261501412959822,
+      "learning_rate": 9.55812599307083e-06,
+      "loss": 0.2041,
+      "step": 2891
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.0155440738421306,
+      "learning_rate": 9.557755653200747e-06,
+      "loss": 0.201,
+      "step": 2892
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.173082046349905,
+      "learning_rate": 9.557385165383291e-06,
+      "loss": 0.175,
+      "step": 2893
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.5471940096966894,
+      "learning_rate": 9.557014529630495e-06,
+      "loss": 0.2171,
+      "step": 2894
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.3712162645057888,
+      "learning_rate": 9.556643745954383e-06,
+      "loss": 0.2377,
+      "step": 2895
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.3902422506822631,
+      "learning_rate": 9.556272814366998e-06,
+      "loss": 0.217,
+      "step": 2896
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.4679976474177272,
+      "learning_rate": 9.555901734880378e-06,
+      "loss": 0.2245,
+      "step": 2897
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.3379312665711216,
+      "learning_rate": 9.555530507506566e-06,
+      "loss": 0.2024,
+      "step": 2898
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.1140472470586262,
+      "learning_rate": 9.555159132257612e-06,
+      "loss": 0.2516,
+      "step": 2899
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.2710921968932416,
+      "learning_rate": 9.554787609145577e-06,
+      "loss": 0.186,
+      "step": 2900
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.5537369508321912,
+      "learning_rate": 9.554415938182516e-06,
+      "loss": 0.2138,
+      "step": 2901
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.2257468395314464,
+      "learning_rate": 9.554044119380494e-06,
+      "loss": 0.2196,
+      "step": 2902
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.4247627827349878,
+      "learning_rate": 9.55367215275158e-06,
+      "loss": 0.224,
+      "step": 2903
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.3250908244691555,
+      "learning_rate": 9.553300038307852e-06,
+      "loss": 0.2281,
+      "step": 2904
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.142675399715007,
+      "learning_rate": 9.552927776061385e-06,
+      "loss": 0.1452,
+      "step": 2905
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.3323915978589722,
+      "learning_rate": 9.552555366024266e-06,
+      "loss": 0.2559,
+      "step": 2906
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.9736660709650402,
+      "learning_rate": 9.55218280820858e-06,
+      "loss": 0.1851,
+      "step": 2907
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.9257141362582549,
+      "learning_rate": 9.551810102626422e-06,
+      "loss": 0.1809,
+      "step": 2908
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.1935794755337041,
+      "learning_rate": 9.551437249289892e-06,
+      "loss": 0.2277,
+      "step": 2909
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.1725061536906543,
+      "learning_rate": 9.55106424821109e-06,
+      "loss": 0.2294,
+      "step": 2910
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.9881244655682483,
+      "learning_rate": 9.550691099402127e-06,
+      "loss": 0.1935,
+      "step": 2911
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.2875813095102464,
+      "learning_rate": 9.550317802875113e-06,
+      "loss": 0.253,
+      "step": 2912
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.05401293763874,
+      "learning_rate": 9.549944358642166e-06,
+      "loss": 0.2133,
+      "step": 2913
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.2569263947367926,
+      "learning_rate": 9.54957076671541e-06,
+      "loss": 0.15,
+      "step": 2914
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.6593602302849433,
+      "learning_rate": 9.54919702710697e-06,
+      "loss": 0.2552,
+      "step": 2915
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.8845742799784067,
+      "learning_rate": 9.54882313982898e-06,
+      "loss": 0.1856,
+      "step": 2916
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.2391712423017027,
+      "learning_rate": 9.548449104893574e-06,
+      "loss": 0.1874,
+      "step": 2917
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.6560170846587032,
+      "learning_rate": 9.548074922312894e-06,
+      "loss": 0.22,
+      "step": 2918
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.9678996347393991,
+      "learning_rate": 9.547700592099086e-06,
+      "loss": 0.2254,
+      "step": 2919
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.890445675489873,
+      "learning_rate": 9.547326114264304e-06,
+      "loss": 0.1635,
+      "step": 2920
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.1296478792157052,
+      "learning_rate": 9.5469514888207e-06,
+      "loss": 0.2367,
+      "step": 2921
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.8738150737181716,
+      "learning_rate": 9.546576715780437e-06,
+      "loss": 0.1921,
+      "step": 2922
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.2717957650043934,
+      "learning_rate": 9.546201795155679e-06,
+      "loss": 0.1977,
+      "step": 2923
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.2811583683928651,
+      "learning_rate": 9.545826726958596e-06,
+      "loss": 0.2255,
+      "step": 2924
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.6951990743347856,
+      "learning_rate": 9.545451511201365e-06,
+      "loss": 0.2546,
+      "step": 2925
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.9544474727288081,
+      "learning_rate": 9.545076147896161e-06,
+      "loss": 0.1534,
+      "step": 2926
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.6881587897163495,
+      "learning_rate": 9.544700637055174e-06,
+      "loss": 0.2165,
+      "step": 2927
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.9059097734556013,
+      "learning_rate": 9.54432497869059e-06,
+      "loss": 0.1957,
+      "step": 2928
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.0095684884693792,
+      "learning_rate": 9.543949172814604e-06,
+      "loss": 0.1909,
+      "step": 2929
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.238689815495979,
+      "learning_rate": 9.543573219439415e-06,
+      "loss": 0.1776,
+      "step": 2930
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.2236574540393286,
+      "learning_rate": 9.543197118577227e-06,
+      "loss": 0.174,
+      "step": 2931
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.249655512190264,
+      "learning_rate": 9.542820870240248e-06,
+      "loss": 0.2928,
+      "step": 2932
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.1626646342203997,
+      "learning_rate": 9.542444474440691e-06,
+      "loss": 0.1839,
+      "step": 2933
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.2445893788337212,
+      "learning_rate": 9.542067931190774e-06,
+      "loss": 0.2053,
+      "step": 2934
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.4599400955804875,
+      "learning_rate": 9.541691240502721e-06,
+      "loss": 0.2109,
+      "step": 2935
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.2511835366288304,
+      "learning_rate": 9.541314402388758e-06,
+      "loss": 0.1792,
+      "step": 2936
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.226633587678485,
+      "learning_rate": 9.540937416861117e-06,
+      "loss": 0.2346,
+      "step": 2937
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.9764541645290232,
+      "learning_rate": 9.540560283932038e-06,
+      "loss": 0.2112,
+      "step": 2938
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.3984432593285652,
+      "learning_rate": 9.540183003613762e-06,
+      "loss": 0.2115,
+      "step": 2939
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.6385126229743738,
+      "learning_rate": 9.539805575918531e-06,
+      "loss": 0.2494,
+      "step": 2940
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.998662721278172,
+      "learning_rate": 9.539428000858604e-06,
+      "loss": 0.2206,
+      "step": 2941
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.6396566984912548,
+      "learning_rate": 9.539050278446231e-06,
+      "loss": 0.1894,
+      "step": 2942
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.4414256696762646,
+      "learning_rate": 9.538672408693679e-06,
+      "loss": 0.2221,
+      "step": 2943
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.3065863414145715,
+      "learning_rate": 9.538294391613208e-06,
+      "loss": 0.2373,
+      "step": 2944
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.050378545008575,
+      "learning_rate": 9.537916227217092e-06,
+      "loss": 0.1457,
+      "step": 2945
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.1536941826751494,
+      "learning_rate": 9.537537915517607e-06,
+      "loss": 0.232,
+      "step": 2946
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.1541343129124273,
+      "learning_rate": 9.53715945652703e-06,
+      "loss": 0.1935,
+      "step": 2947
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.7840900534344486,
+      "learning_rate": 9.53678085025765e-06,
+      "loss": 0.1998,
+      "step": 2948
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.4401910551864372,
+      "learning_rate": 9.536402096721755e-06,
+      "loss": 0.2586,
+      "step": 2949
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.253599785998592,
+      "learning_rate": 9.536023195931637e-06,
+      "loss": 0.1819,
+      "step": 2950
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.0117746967326484,
+      "learning_rate": 9.535644147899599e-06,
+      "loss": 0.2218,
+      "step": 2951
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.1796530156432152,
+      "learning_rate": 9.535264952637944e-06,
+      "loss": 0.2016,
+      "step": 2952
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.2653518620443187,
+      "learning_rate": 9.53488561015898e-06,
+      "loss": 0.2142,
+      "step": 2953
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.574353858318175,
+      "learning_rate": 9.534506120475022e-06,
+      "loss": 0.2005,
+      "step": 2954
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.4150858753535254,
+      "learning_rate": 9.534126483598388e-06,
+      "loss": 0.2543,
+      "step": 2955
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.3278445347402552,
+      "learning_rate": 9.5337466995414e-06,
+      "loss": 0.2444,
+      "step": 2956
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.3129428303811432,
+      "learning_rate": 9.533366768316388e-06,
+      "loss": 0.2444,
+      "step": 2957
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.1863117924063054,
+      "learning_rate": 9.532986689935683e-06,
+      "loss": 0.1392,
+      "step": 2958
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.330580421863898,
+      "learning_rate": 9.532606464411623e-06,
+      "loss": 0.2527,
+      "step": 2959
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.9152200819031796,
+      "learning_rate": 9.532226091756553e-06,
+      "loss": 0.2005,
+      "step": 2960
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.2345838822710016,
+      "learning_rate": 9.531845571982817e-06,
+      "loss": 0.2143,
+      "step": 2961
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.0269849072555577,
+      "learning_rate": 9.531464905102765e-06,
+      "loss": 0.2054,
+      "step": 2962
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.4303592689209252,
+      "learning_rate": 9.53108409112876e-06,
+      "loss": 0.2136,
+      "step": 2963
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.5057812112592688,
+      "learning_rate": 9.530703130073158e-06,
+      "loss": 0.1924,
+      "step": 2964
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.1630847764349714,
+      "learning_rate": 9.530322021948326e-06,
+      "loss": 0.2179,
+      "step": 2965
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.3950389958173,
+      "learning_rate": 9.52994076676664e-06,
+      "loss": 0.2236,
+      "step": 2966
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.1008233370292184,
+      "learning_rate": 9.529559364540469e-06,
+      "loss": 0.1991,
+      "step": 2967
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.3908909358546175,
+      "learning_rate": 9.529177815282196e-06,
+      "loss": 0.296,
+      "step": 2968
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.2256134744415434,
+      "learning_rate": 9.528796119004207e-06,
+      "loss": 0.2388,
+      "step": 2969
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.8390877825552614,
+      "learning_rate": 9.52841427571889e-06,
+      "loss": 0.1952,
+      "step": 2970
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.8347367560161315,
+      "learning_rate": 9.528032285438645e-06,
+      "loss": 0.1311,
+      "step": 2971
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.2247969667741563,
+      "learning_rate": 9.527650148175866e-06,
+      "loss": 0.1858,
+      "step": 2972
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.0830301363323687,
+      "learning_rate": 9.527267863942961e-06,
+      "loss": 0.1596,
+      "step": 2973
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.0606586246381289,
+      "learning_rate": 9.526885432752337e-06,
+      "loss": 0.1742,
+      "step": 2974
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.9541970188645297,
+      "learning_rate": 9.526502854616408e-06,
+      "loss": 0.2186,
+      "step": 2975
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.919477957021421,
+      "learning_rate": 9.526120129547596e-06,
+      "loss": 0.2186,
+      "step": 2976
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.2571075845399458,
+      "learning_rate": 9.52573725755832e-06,
+      "loss": 0.2293,
+      "step": 2977
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.3883552861047486,
+      "learning_rate": 9.52535423866101e-06,
+      "loss": 0.186,
+      "step": 2978
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 2.2630965593674266,
+      "learning_rate": 9.524971072868098e-06,
+      "loss": 0.1703,
+      "step": 2979
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.2252324844673652,
+      "learning_rate": 9.524587760192026e-06,
+      "loss": 0.2081,
+      "step": 2980
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.3842262810333104,
+      "learning_rate": 9.524204300645231e-06,
+      "loss": 0.2282,
+      "step": 2981
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.488999341203686,
+      "learning_rate": 9.523820694240165e-06,
+      "loss": 0.1669,
+      "step": 2982
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.6110655784567163,
+      "learning_rate": 9.523436940989275e-06,
+      "loss": 0.2453,
+      "step": 2983
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.4162178986872427,
+      "learning_rate": 9.523053040905024e-06,
+      "loss": 0.1997,
+      "step": 2984
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.9621288931040425,
+      "learning_rate": 9.522668993999867e-06,
+      "loss": 0.2212,
+      "step": 2985
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.2934269914202492,
+      "learning_rate": 9.522284800286276e-06,
+      "loss": 0.1677,
+      "step": 2986
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.5795501373819127,
+      "learning_rate": 9.521900459776719e-06,
+      "loss": 0.1791,
+      "step": 2987
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.3111183807838047,
+      "learning_rate": 9.521515972483673e-06,
+      "loss": 0.186,
+      "step": 2988
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.298174126971471,
+      "learning_rate": 9.521131338419618e-06,
+      "loss": 0.2134,
+      "step": 2989
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.7745581627689544,
+      "learning_rate": 9.52074655759704e-06,
+      "loss": 0.2612,
+      "step": 2990
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.0532714786452613,
+      "learning_rate": 9.52036163002843e-06,
+      "loss": 0.2078,
+      "step": 2991
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.0330828829278897,
+      "learning_rate": 9.519976555726282e-06,
+      "loss": 0.1962,
+      "step": 2992
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.4378372190755404,
+      "learning_rate": 9.519591334703095e-06,
+      "loss": 0.2215,
+      "step": 2993
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.1994989126904343,
+      "learning_rate": 9.519205966971375e-06,
+      "loss": 0.2617,
+      "step": 2994
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.2209856484582746,
+      "learning_rate": 9.518820452543631e-06,
+      "loss": 0.2059,
+      "step": 2995
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.91828173922165,
+      "learning_rate": 9.518434791432376e-06,
+      "loss": 0.2135,
+      "step": 2996
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.1521356281552044,
+      "learning_rate": 9.51804898365013e-06,
+      "loss": 0.2353,
+      "step": 2997
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.171903174379535,
+      "learning_rate": 9.517663029209414e-06,
+      "loss": 0.2332,
+      "step": 2998
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.1848569870613954,
+      "learning_rate": 9.517276928122761e-06,
+      "loss": 0.203,
+      "step": 2999
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.295059701653088,
+      "learning_rate": 9.5168906804027e-06,
+      "loss": 0.2256,
+      "step": 3000
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.3794539568768331,
+      "learning_rate": 9.51650428606177e-06,
+      "loss": 0.1936,
+      "step": 3001
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.1775007965227944,
+      "learning_rate": 9.516117745112512e-06,
+      "loss": 0.2039,
+      "step": 3002
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.348573085337978,
+      "learning_rate": 9.515731057567477e-06,
+      "loss": 0.2362,
+      "step": 3003
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.01472585013804,
+      "learning_rate": 9.515344223439216e-06,
+      "loss": 0.267,
+      "step": 3004
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.918703308832875,
+      "learning_rate": 9.514957242740283e-06,
+      "loss": 0.2032,
+      "step": 3005
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.559428295149764,
+      "learning_rate": 9.514570115483242e-06,
+      "loss": 0.2387,
+      "step": 3006
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.2608708523575913,
+      "learning_rate": 9.51418284168066e-06,
+      "loss": 0.2333,
+      "step": 3007
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.8756637439964591,
+      "learning_rate": 9.513795421345106e-06,
+      "loss": 0.1634,
+      "step": 3008
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.4695519720661745,
+      "learning_rate": 9.513407854489158e-06,
+      "loss": 0.2332,
+      "step": 3009
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.049988733171573,
+      "learning_rate": 9.513020141125396e-06,
+      "loss": 0.1669,
+      "step": 3010
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.1652008753253742,
+      "learning_rate": 9.512632281266403e-06,
+      "loss": 0.2073,
+      "step": 3011
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.4949582748626602,
+      "learning_rate": 9.512244274924774e-06,
+      "loss": 0.2564,
+      "step": 3012
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.9129262611181551,
+      "learning_rate": 9.511856122113099e-06,
+      "loss": 0.2201,
+      "step": 3013
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.3726513229769113,
+      "learning_rate": 9.51146782284398e-06,
+      "loss": 0.2166,
+      "step": 3014
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.8857732833610704,
+      "learning_rate": 9.51107937713002e-06,
+      "loss": 0.1878,
+      "step": 3015
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.1594145921671435,
+      "learning_rate": 9.510690784983831e-06,
+      "loss": 0.1703,
+      "step": 3016
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.8174245707589436,
+      "learning_rate": 9.510302046418025e-06,
+      "loss": 0.218,
+      "step": 3017
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.2769199420705253,
+      "learning_rate": 9.509913161445223e-06,
+      "loss": 0.1894,
+      "step": 3018
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.2294614491093345,
+      "learning_rate": 9.509524130078043e-06,
+      "loss": 0.1939,
+      "step": 3019
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.6442175076599705,
+      "learning_rate": 9.509134952329119e-06,
+      "loss": 0.2367,
+      "step": 3020
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.4000264755740743,
+      "learning_rate": 9.50874562821108e-06,
+      "loss": 0.239,
+      "step": 3021
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.3259668033654834,
+      "learning_rate": 9.508356157736568e-06,
+      "loss": 0.1642,
+      "step": 3022
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.304737469864063,
+      "learning_rate": 9.507966540918218e-06,
+      "loss": 0.1872,
+      "step": 3023
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.1930084961351222,
+      "learning_rate": 9.507576777768686e-06,
+      "loss": 0.2451,
+      "step": 3024
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.0299877186385595,
+      "learning_rate": 9.507186868300618e-06,
+      "loss": 0.1893,
+      "step": 3025
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.89223699290459,
+      "learning_rate": 9.50679681252667e-06,
+      "loss": 0.1942,
+      "step": 3026
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.3212254262910839,
+      "learning_rate": 9.50640661045951e-06,
+      "loss": 0.1884,
+      "step": 3027
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.6204169647334925,
+      "learning_rate": 9.506016262111798e-06,
+      "loss": 0.2412,
+      "step": 3028
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.318079832604575,
+      "learning_rate": 9.505625767496207e-06,
+      "loss": 0.2184,
+      "step": 3029
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.834650814174236,
+      "learning_rate": 9.505235126625414e-06,
+      "loss": 0.2143,
+      "step": 3030
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.1295550323449075,
+      "learning_rate": 9.504844339512096e-06,
+      "loss": 0.1916,
+      "step": 3031
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.368385878286387,
+      "learning_rate": 9.504453406168942e-06,
+      "loss": 0.2482,
+      "step": 3032
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.6047705271338522,
+      "learning_rate": 9.504062326608639e-06,
+      "loss": 0.1954,
+      "step": 3033
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.9606839510349122,
+      "learning_rate": 9.503671100843883e-06,
+      "loss": 0.2227,
+      "step": 3034
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.5136552674397654,
+      "learning_rate": 9.503279728887374e-06,
+      "loss": 0.2459,
+      "step": 3035
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.6420906502469046,
+      "learning_rate": 9.502888210751815e-06,
+      "loss": 0.2534,
+      "step": 3036
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.1093866858068369,
+      "learning_rate": 9.502496546449916e-06,
+      "loss": 0.2011,
+      "step": 3037
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.2478059825731789,
+      "learning_rate": 9.50210473599439e-06,
+      "loss": 0.2398,
+      "step": 3038
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.2135737594636777,
+      "learning_rate": 9.501712779397955e-06,
+      "loss": 0.2127,
+      "step": 3039
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.2825561486258843,
+      "learning_rate": 9.501320676673334e-06,
+      "loss": 0.1956,
+      "step": 3040
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.1280243623311197,
+      "learning_rate": 9.500928427833256e-06,
+      "loss": 0.1941,
+      "step": 3041
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.176598504074112,
+      "learning_rate": 9.500536032890452e-06,
+      "loss": 0.1926,
+      "step": 3042
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.290654416130069,
+      "learning_rate": 9.50014349185766e-06,
+      "loss": 0.2162,
+      "step": 3043
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.4964866520668416,
+      "learning_rate": 9.499750804747624e-06,
+      "loss": 0.2126,
+      "step": 3044
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.5251694493269248,
+      "learning_rate": 9.499357971573089e-06,
+      "loss": 0.2438,
+      "step": 3045
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.4636332125783982,
+      "learning_rate": 9.498964992346808e-06,
+      "loss": 0.2142,
+      "step": 3046
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.500467991094329,
+      "learning_rate": 9.498571867081534e-06,
+      "loss": 0.1961,
+      "step": 3047
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.564805380493184,
+      "learning_rate": 9.498178595790032e-06,
+      "loss": 0.209,
+      "step": 3048
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.6172178284017242,
+      "learning_rate": 9.497785178485066e-06,
+      "loss": 0.2662,
+      "step": 3049
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.1185466261836827,
+      "learning_rate": 9.497391615179405e-06,
+      "loss": 0.1928,
+      "step": 3050
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.6924671695770666,
+      "learning_rate": 9.496997905885829e-06,
+      "loss": 0.2211,
+      "step": 3051
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.7200388194363756,
+      "learning_rate": 9.496604050617113e-06,
+      "loss": 0.1939,
+      "step": 3052
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.3040272418296532,
+      "learning_rate": 9.496210049386044e-06,
+      "loss": 0.2375,
+      "step": 3053
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 2.167072066029197,
+      "learning_rate": 9.495815902205412e-06,
+      "loss": 0.2332,
+      "step": 3054
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.241152049902959,
+      "learning_rate": 9.495421609088009e-06,
+      "loss": 0.2414,
+      "step": 3055
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.0754625073320603,
+      "learning_rate": 9.495027170046637e-06,
+      "loss": 0.1857,
+      "step": 3056
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.183038064923483,
+      "learning_rate": 9.494632585094097e-06,
+      "loss": 0.2498,
+      "step": 3057
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.2607833240986952,
+      "learning_rate": 9.494237854243201e-06,
+      "loss": 0.2421,
+      "step": 3058
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.7890076962035462,
+      "learning_rate": 9.493842977506757e-06,
+      "loss": 0.2252,
+      "step": 3059
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.6776643145568921,
+      "learning_rate": 9.493447954897587e-06,
+      "loss": 0.3131,
+      "step": 3060
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.1613804270043022,
+      "learning_rate": 9.493052786428512e-06,
+      "loss": 0.213,
+      "step": 3061
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.3133859057143857,
+      "learning_rate": 9.49265747211236e-06,
+      "loss": 0.1814,
+      "step": 3062
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.3984924524048687,
+      "learning_rate": 9.492262011961964e-06,
+      "loss": 0.1669,
+      "step": 3063
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.1625214767651273,
+      "learning_rate": 9.49186640599016e-06,
+      "loss": 0.1793,
+      "step": 3064
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.4412137500028803,
+      "learning_rate": 9.491470654209788e-06,
+      "loss": 0.2441,
+      "step": 3065
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.201915510777354,
+      "learning_rate": 9.491074756633697e-06,
+      "loss": 0.2457,
+      "step": 3066
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.0297978971515234,
+      "learning_rate": 9.490678713274736e-06,
+      "loss": 0.2228,
+      "step": 3067
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.0862319156109284,
+      "learning_rate": 9.490282524145762e-06,
+      "loss": 0.266,
+      "step": 3068
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.8460754209384096,
+      "learning_rate": 9.489886189259634e-06,
+      "loss": 0.2172,
+      "step": 3069
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.9435512454223389,
+      "learning_rate": 9.48948970862922e-06,
+      "loss": 0.1816,
+      "step": 3070
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.5561000359143289,
+      "learning_rate": 9.489093082267388e-06,
+      "loss": 0.2069,
+      "step": 3071
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.2050008980907956,
+      "learning_rate": 9.488696310187013e-06,
+      "loss": 0.213,
+      "step": 3072
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.6249288763473806,
+      "learning_rate": 9.488299392400975e-06,
+      "loss": 0.2039,
+      "step": 3073
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.3965359551567451,
+      "learning_rate": 9.487902328922157e-06,
+      "loss": 0.2284,
+      "step": 3074
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.4453507341018736,
+      "learning_rate": 9.48750511976345e-06,
+      "loss": 0.205,
+      "step": 3075
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 2.0632286374173967,
+      "learning_rate": 9.487107764937745e-06,
+      "loss": 0.2465,
+      "step": 3076
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.867004884384756,
+      "learning_rate": 9.486710264457943e-06,
+      "loss": 0.2265,
+      "step": 3077
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.2751846293163889,
+      "learning_rate": 9.486312618336944e-06,
+      "loss": 0.2537,
+      "step": 3078
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.217909445123223,
+      "learning_rate": 9.485914826587659e-06,
+      "loss": 0.1589,
+      "step": 3079
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.3215295991367464,
+      "learning_rate": 9.485516889222999e-06,
+      "loss": 0.208,
+      "step": 3080
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.441391394313769,
+      "learning_rate": 9.48511880625588e-06,
+      "loss": 0.1909,
+      "step": 3081
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.2151124172636405,
+      "learning_rate": 9.484720577699227e-06,
+      "loss": 0.2027,
+      "step": 3082
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.0772353662048864,
+      "learning_rate": 9.484322203565965e-06,
+      "loss": 0.1284,
+      "step": 3083
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.1365724253122773,
+      "learning_rate": 9.483923683869025e-06,
+      "loss": 0.2552,
+      "step": 3084
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.2834156272664685,
+      "learning_rate": 9.483525018621343e-06,
+      "loss": 0.1812,
+      "step": 3085
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.2193446420422709,
+      "learning_rate": 9.483126207835863e-06,
+      "loss": 0.161,
+      "step": 3086
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.4654941996625883,
+      "learning_rate": 9.482727251525526e-06,
+      "loss": 0.2111,
+      "step": 3087
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.2248216746166165,
+      "learning_rate": 9.482328149703286e-06,
+      "loss": 0.2231,
+      "step": 3088
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.0138876158027335,
+      "learning_rate": 9.481928902382096e-06,
+      "loss": 0.2199,
+      "step": 3089
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.0878177632119868,
+      "learning_rate": 9.481529509574916e-06,
+      "loss": 0.1889,
+      "step": 3090
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.1950474520014636,
+      "learning_rate": 9.481129971294714e-06,
+      "loss": 0.2101,
+      "step": 3091
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.3554068999237128,
+      "learning_rate": 9.480730287554454e-06,
+      "loss": 0.2116,
+      "step": 3092
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.9880225102972272,
+      "learning_rate": 9.480330458367112e-06,
+      "loss": 0.1923,
+      "step": 3093
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.5432883112793079,
+      "learning_rate": 9.479930483745667e-06,
+      "loss": 0.2637,
+      "step": 3094
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.2333451332782186,
+      "learning_rate": 9.479530363703101e-06,
+      "loss": 0.1977,
+      "step": 3095
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.1904710194498107,
+      "learning_rate": 9.479130098252407e-06,
+      "loss": 0.189,
+      "step": 3096
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.4147799394816463,
+      "learning_rate": 9.478729687406572e-06,
+      "loss": 0.2314,
+      "step": 3097
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 8.541241059903635,
+      "learning_rate": 9.478329131178598e-06,
+      "loss": 0.2547,
+      "step": 3098
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.9724719148634194,
+      "learning_rate": 9.477928429581483e-06,
+      "loss": 0.2163,
+      "step": 3099
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.3775199764957202,
+      "learning_rate": 9.477527582628236e-06,
+      "loss": 0.2171,
+      "step": 3100
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.2142214300390899,
+      "learning_rate": 9.477126590331871e-06,
+      "loss": 0.2395,
+      "step": 3101
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.857217065407973,
+      "learning_rate": 9.4767254527054e-06,
+      "loss": 0.2265,
+      "step": 3102
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.3480019402455514,
+      "learning_rate": 9.476324169761849e-06,
+      "loss": 0.2561,
+      "step": 3103
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.8560082384134242,
+      "learning_rate": 9.47592274151424e-06,
+      "loss": 0.2549,
+      "step": 3104
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.1308206376999632,
+      "learning_rate": 9.475521167975605e-06,
+      "loss": 0.1658,
+      "step": 3105
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.155282930672112,
+      "learning_rate": 9.475119449158979e-06,
+      "loss": 0.1454,
+      "step": 3106
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.393390529563395,
+      "learning_rate": 9.474717585077403e-06,
+      "loss": 0.2019,
+      "step": 3107
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.4290263518118709,
+      "learning_rate": 9.47431557574392e-06,
+      "loss": 0.249,
+      "step": 3108
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.35763767091426,
+      "learning_rate": 9.473913421171581e-06,
+      "loss": 0.1969,
+      "step": 3109
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.242779232433093,
+      "learning_rate": 9.47351112137344e-06,
+      "loss": 0.2294,
+      "step": 3110
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.9303229748665075,
+      "learning_rate": 9.473108676362555e-06,
+      "loss": 0.2386,
+      "step": 3111
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.8503031725987009,
+      "learning_rate": 9.47270608615199e-06,
+      "loss": 0.1804,
+      "step": 3112
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.0098953624882945,
+      "learning_rate": 9.472303350754814e-06,
+      "loss": 0.175,
+      "step": 3113
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.1800209057423905,
+      "learning_rate": 9.471900470184101e-06,
+      "loss": 0.1985,
+      "step": 3114
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.2614048303050134,
+      "learning_rate": 9.471497444452925e-06,
+      "loss": 0.2518,
+      "step": 3115
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.8981917978259172,
+      "learning_rate": 9.471094273574371e-06,
+      "loss": 0.1638,
+      "step": 3116
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.9440749155806507,
+      "learning_rate": 9.470690957561526e-06,
+      "loss": 0.2539,
+      "step": 3117
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.2482442848953152,
+      "learning_rate": 9.470287496427482e-06,
+      "loss": 0.1846,
+      "step": 3118
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.992469388522179,
+      "learning_rate": 9.469883890185335e-06,
+      "loss": 0.2484,
+      "step": 3119
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.4996626270892064,
+      "learning_rate": 9.469480138848187e-06,
+      "loss": 0.2062,
+      "step": 3120
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.357236096135525,
+      "learning_rate": 9.469076242429142e-06,
+      "loss": 0.2595,
+      "step": 3121
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.3627457701089563,
+      "learning_rate": 9.468672200941314e-06,
+      "loss": 0.2061,
+      "step": 3122
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.0648573116801885,
+      "learning_rate": 9.468268014397816e-06,
+      "loss": 0.2209,
+      "step": 3123
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.909536259127003,
+      "learning_rate": 9.467863682811768e-06,
+      "loss": 0.2827,
+      "step": 3124
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.0172623292585263,
+      "learning_rate": 9.467459206196298e-06,
+      "loss": 0.2029,
+      "step": 3125
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.9903449376774335,
+      "learning_rate": 9.467054584564532e-06,
+      "loss": 0.2332,
+      "step": 3126
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.0798843655788095,
+      "learning_rate": 9.466649817929606e-06,
+      "loss": 0.2247,
+      "step": 3127
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.9411804039113429,
+      "learning_rate": 9.466244906304657e-06,
+      "loss": 0.1651,
+      "step": 3128
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.758342704915907,
+      "learning_rate": 9.46583984970283e-06,
+      "loss": 0.183,
+      "step": 3129
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.3654056313411398,
+      "learning_rate": 9.465434648137276e-06,
+      "loss": 0.213,
+      "step": 3130
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.9187531123303228,
+      "learning_rate": 9.465029301621142e-06,
+      "loss": 0.2502,
+      "step": 3131
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.1011724745392564,
+      "learning_rate": 9.464623810167592e-06,
+      "loss": 0.2064,
+      "step": 3132
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.0265365273965545,
+      "learning_rate": 9.464218173789783e-06,
+      "loss": 0.2405,
+      "step": 3133
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.0596644636593127,
+      "learning_rate": 9.463812392500887e-06,
+      "loss": 0.215,
+      "step": 3134
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.0569208376468124,
+      "learning_rate": 9.463406466314073e-06,
+      "loss": 0.2381,
+      "step": 3135
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.9286069751637309,
+      "learning_rate": 9.463000395242521e-06,
+      "loss": 0.2423,
+      "step": 3136
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.2000290435018142,
+      "learning_rate": 9.462594179299408e-06,
+      "loss": 0.2193,
+      "step": 3137
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 2.224684421790633,
+      "learning_rate": 9.462187818497922e-06,
+      "loss": 0.2503,
+      "step": 3138
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.1735585025237378,
+      "learning_rate": 9.461781312851251e-06,
+      "loss": 0.1777,
+      "step": 3139
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.242250842002924,
+      "learning_rate": 9.461374662372595e-06,
+      "loss": 0.179,
+      "step": 3140
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.17148273578566,
+      "learning_rate": 9.460967867075152e-06,
+      "loss": 0.1921,
+      "step": 3141
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.261330627528709,
+      "learning_rate": 9.460560926972127e-06,
+      "loss": 0.2042,
+      "step": 3142
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.19559692913191,
+      "learning_rate": 9.46015384207673e-06,
+      "loss": 0.2004,
+      "step": 3143
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.3253222146832901,
+      "learning_rate": 9.459746612402172e-06,
+      "loss": 0.1953,
+      "step": 3144
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.3992385647732282,
+      "learning_rate": 9.459339237961676e-06,
+      "loss": 0.1564,
+      "step": 3145
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.3269885530366163,
+      "learning_rate": 9.458931718768464e-06,
+      "loss": 0.2153,
+      "step": 3146
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.1910055495086513,
+      "learning_rate": 9.458524054835764e-06,
+      "loss": 0.1408,
+      "step": 3147
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.7627977906762868,
+      "learning_rate": 9.458116246176811e-06,
+      "loss": 0.1925,
+      "step": 3148
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.06647430628944,
+      "learning_rate": 9.457708292804839e-06,
+      "loss": 0.1927,
+      "step": 3149
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.6509751159560542,
+      "learning_rate": 9.457300194733091e-06,
+      "loss": 0.2542,
+      "step": 3150
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.4262231004961177,
+      "learning_rate": 9.45689195197482e-06,
+      "loss": 0.2208,
+      "step": 3151
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.2799823675260156,
+      "learning_rate": 9.45648356454327e-06,
+      "loss": 0.2045,
+      "step": 3152
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.2001872873309563,
+      "learning_rate": 9.456075032451703e-06,
+      "loss": 0.2087,
+      "step": 3153
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.3281534977267744,
+      "learning_rate": 9.455666355713377e-06,
+      "loss": 0.2571,
+      "step": 3154
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.2006143749338827,
+      "learning_rate": 9.455257534341558e-06,
+      "loss": 0.2265,
+      "step": 3155
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.499061908290397,
+      "learning_rate": 9.454848568349518e-06,
+      "loss": 0.1653,
+      "step": 3156
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.2672938878994013,
+      "learning_rate": 9.454439457750532e-06,
+      "loss": 0.1972,
+      "step": 3157
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.0977778689793827,
+      "learning_rate": 9.45403020255788e-06,
+      "loss": 0.2285,
+      "step": 3158
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.1817851064956852,
+      "learning_rate": 9.453620802784846e-06,
+      "loss": 0.2427,
+      "step": 3159
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.3976966151351917,
+      "learning_rate": 9.45321125844472e-06,
+      "loss": 0.2113,
+      "step": 3160
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.8766173169008982,
+      "learning_rate": 9.452801569550796e-06,
+      "loss": 0.2036,
+      "step": 3161
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.2074756298113538,
+      "learning_rate": 9.452391736116371e-06,
+      "loss": 0.2329,
+      "step": 3162
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.1436368658637175,
+      "learning_rate": 9.45198175815475e-06,
+      "loss": 0.2342,
+      "step": 3163
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.0602500466772522,
+      "learning_rate": 9.451571635679245e-06,
+      "loss": 0.2116,
+      "step": 3164
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.6400566740227274,
+      "learning_rate": 9.451161368703161e-06,
+      "loss": 0.209,
+      "step": 3165
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.9536128388387705,
+      "learning_rate": 9.45075095723982e-06,
+      "loss": 0.2268,
+      "step": 3166
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.1457805491792155,
+      "learning_rate": 9.450340401302544e-06,
+      "loss": 0.2154,
+      "step": 3167
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.4158616001941202,
+      "learning_rate": 9.44992970090466e-06,
+      "loss": 0.2576,
+      "step": 3168
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.8537944953537215,
+      "learning_rate": 9.449518856059499e-06,
+      "loss": 0.1884,
+      "step": 3169
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.2074626920027767,
+      "learning_rate": 9.449107866780395e-06,
+      "loss": 0.2198,
+      "step": 3170
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.4197886245729467,
+      "learning_rate": 9.448696733080694e-06,
+      "loss": 0.1743,
+      "step": 3171
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.3253288525038232,
+      "learning_rate": 9.448285454973739e-06,
+      "loss": 0.2334,
+      "step": 3172
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.0881460016777589,
+      "learning_rate": 9.447874032472877e-06,
+      "loss": 0.1668,
+      "step": 3173
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.2255642528059372,
+      "learning_rate": 9.447462465591468e-06,
+      "loss": 0.2714,
+      "step": 3174
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.7627177989930845,
+      "learning_rate": 9.44705075434287e-06,
+      "loss": 0.2179,
+      "step": 3175
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.192462006234564,
+      "learning_rate": 9.446638898740447e-06,
+      "loss": 0.2342,
+      "step": 3176
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.1180427385501182,
+      "learning_rate": 9.446226898797569e-06,
+      "loss": 0.1998,
+      "step": 3177
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.7194999229215386,
+      "learning_rate": 9.445814754527609e-06,
+      "loss": 0.2007,
+      "step": 3178
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.9264232123924822,
+      "learning_rate": 9.445402465943946e-06,
+      "loss": 0.1711,
+      "step": 3179
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.8287128212578976,
+      "learning_rate": 9.444990033059961e-06,
+      "loss": 0.2463,
+      "step": 3180
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.1082152261432996,
+      "learning_rate": 9.444577455889044e-06,
+      "loss": 0.2631,
+      "step": 3181
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.241928535434447,
+      "learning_rate": 9.444164734444588e-06,
+      "loss": 0.2309,
+      "step": 3182
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.1107697902849742,
+      "learning_rate": 9.443751868739988e-06,
+      "loss": 0.2408,
+      "step": 3183
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.9122052356737828,
+      "learning_rate": 9.443338858788647e-06,
+      "loss": 0.2008,
+      "step": 3184
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.0423075101434127,
+      "learning_rate": 9.442925704603973e-06,
+      "loss": 0.2479,
+      "step": 3185
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.3949710756083935,
+      "learning_rate": 9.442512406199374e-06,
+      "loss": 0.2626,
+      "step": 3186
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.4130544995104173,
+      "learning_rate": 9.442098963588269e-06,
+      "loss": 0.254,
+      "step": 3187
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.2525568025903964,
+      "learning_rate": 9.441685376784077e-06,
+      "loss": 0.2177,
+      "step": 3188
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.1805521657215643,
+      "learning_rate": 9.441271645800223e-06,
+      "loss": 0.2163,
+      "step": 3189
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.0917852568306798,
+      "learning_rate": 9.440857770650139e-06,
+      "loss": 0.1883,
+      "step": 3190
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.0996674341082218,
+      "learning_rate": 9.440443751347255e-06,
+      "loss": 0.2046,
+      "step": 3191
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.4328406744030813,
+      "learning_rate": 9.440029587905017e-06,
+      "loss": 0.1625,
+      "step": 3192
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.6185582037199688,
+      "learning_rate": 9.439615280336863e-06,
+      "loss": 0.2307,
+      "step": 3193
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.952008038464336,
+      "learning_rate": 9.439200828656245e-06,
+      "loss": 0.1736,
+      "step": 3194
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.8124612973238672,
+      "learning_rate": 9.438786232876615e-06,
+      "loss": 0.1563,
+      "step": 3195
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.7549188341278665,
+      "learning_rate": 9.43837149301143e-06,
+      "loss": 0.1974,
+      "step": 3196
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.850363163511226,
+      "learning_rate": 9.437956609074157e-06,
+      "loss": 0.2257,
+      "step": 3197
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.253005794791456,
+      "learning_rate": 9.437541581078257e-06,
+      "loss": 0.2265,
+      "step": 3198
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.4584576065609773,
+      "learning_rate": 9.437126409037209e-06,
+      "loss": 0.1846,
+      "step": 3199
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.4592481927436938,
+      "learning_rate": 9.436711092964484e-06,
+      "loss": 0.1644,
+      "step": 3200
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.283626865289488,
+      "learning_rate": 9.436295632873566e-06,
+      "loss": 0.2027,
+      "step": 3201
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.2240306941076458,
+      "learning_rate": 9.435880028777942e-06,
+      "loss": 0.2323,
+      "step": 3202
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.2230125394099478,
+      "learning_rate": 9.4354642806911e-06,
+      "loss": 0.2177,
+      "step": 3203
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.389861253431983,
+      "learning_rate": 9.435048388626538e-06,
+      "loss": 0.2203,
+      "step": 3204
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.2487591664471382,
+      "learning_rate": 9.434632352597754e-06,
+      "loss": 0.2433,
+      "step": 3205
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.3927196695586344,
+      "learning_rate": 9.434216172618253e-06,
+      "loss": 0.2188,
+      "step": 3206
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.1437899431435192,
+      "learning_rate": 9.433799848701546e-06,
+      "loss": 0.157,
+      "step": 3207
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.2317763737385234,
+      "learning_rate": 9.433383380861147e-06,
+      "loss": 0.24,
+      "step": 3208
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 3.323731006741638,
+      "learning_rate": 9.432966769110574e-06,
+      "loss": 0.2279,
+      "step": 3209
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.7601396919556374,
+      "learning_rate": 9.432550013463352e-06,
+      "loss": 0.1698,
+      "step": 3210
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.2348897952985047,
+      "learning_rate": 9.432133113933003e-06,
+      "loss": 0.192,
+      "step": 3211
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.8640128404856529,
+      "learning_rate": 9.431716070533067e-06,
+      "loss": 0.2381,
+      "step": 3212
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.2723518728853989,
+      "learning_rate": 9.43129888327708e-06,
+      "loss": 0.2385,
+      "step": 3213
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.8641736091931441,
+      "learning_rate": 9.430881552178584e-06,
+      "loss": 0.1654,
+      "step": 3214
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 2.979784405737317,
+      "learning_rate": 9.430464077251123e-06,
+      "loss": 0.1675,
+      "step": 3215
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.9911275650064111,
+      "learning_rate": 9.43004645850825e-06,
+      "loss": 0.2332,
+      "step": 3216
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.2094230703292752,
+      "learning_rate": 9.429628695963522e-06,
+      "loss": 0.1911,
+      "step": 3217
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.097194586423753,
+      "learning_rate": 9.429210789630499e-06,
+      "loss": 0.2184,
+      "step": 3218
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.4980553063524655,
+      "learning_rate": 9.428792739522747e-06,
+      "loss": 0.2167,
+      "step": 3219
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.9333743651647057,
+      "learning_rate": 9.428374545653835e-06,
+      "loss": 0.2133,
+      "step": 3220
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.2157256123993065,
+      "learning_rate": 9.427956208037341e-06,
+      "loss": 0.2048,
+      "step": 3221
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.3866920092516222,
+      "learning_rate": 9.427537726686842e-06,
+      "loss": 0.1767,
+      "step": 3222
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.8438634968835794,
+      "learning_rate": 9.42711910161592e-06,
+      "loss": 0.2907,
+      "step": 3223
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.1119601358248832,
+      "learning_rate": 9.426700332838169e-06,
+      "loss": 0.1968,
+      "step": 3224
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.9893826860579937,
+      "learning_rate": 9.426281420367178e-06,
+      "loss": 0.1812,
+      "step": 3225
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.334709110241365,
+      "learning_rate": 9.425862364216547e-06,
+      "loss": 0.2077,
+      "step": 3226
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.06872632316943,
+      "learning_rate": 9.425443164399878e-06,
+      "loss": 0.2245,
+      "step": 3227
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.5279742028764893,
+      "learning_rate": 9.42502382093078e-06,
+      "loss": 0.2283,
+      "step": 3228
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.2050075850406317,
+      "learning_rate": 9.424604333822864e-06,
+      "loss": 0.1975,
+      "step": 3229
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.3262615791511034,
+      "learning_rate": 9.424184703089747e-06,
+      "loss": 0.2557,
+      "step": 3230
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.0777949695344957,
+      "learning_rate": 9.42376492874505e-06,
+      "loss": 0.1868,
+      "step": 3231
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.77326303257876,
+      "learning_rate": 9.4233450108024e-06,
+      "loss": 0.2182,
+      "step": 3232
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.6884029718780367,
+      "learning_rate": 9.422924949275428e-06,
+      "loss": 0.1858,
+      "step": 3233
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.0496184791607364,
+      "learning_rate": 9.422504744177768e-06,
+      "loss": 0.2412,
+      "step": 3234
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.5328311080075332,
+      "learning_rate": 9.422084395523062e-06,
+      "loss": 0.173,
+      "step": 3235
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.0271682879511326,
+      "learning_rate": 9.421663903324951e-06,
+      "loss": 0.2549,
+      "step": 3236
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.1860519100182192,
+      "learning_rate": 9.42124326759709e-06,
+      "loss": 0.2381,
+      "step": 3237
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.6177270744705405,
+      "learning_rate": 9.420822488353127e-06,
+      "loss": 0.2097,
+      "step": 3238
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.293353660199885,
+      "learning_rate": 9.420401565606726e-06,
+      "loss": 0.2221,
+      "step": 3239
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.4891841232329404,
+      "learning_rate": 9.419980499371549e-06,
+      "loss": 0.2379,
+      "step": 3240
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.0637748867310886,
+      "learning_rate": 9.419559289661261e-06,
+      "loss": 0.2049,
+      "step": 3241
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.969446537503596,
+      "learning_rate": 9.419137936489539e-06,
+      "loss": 0.1972,
+      "step": 3242
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.5064359320644394,
+      "learning_rate": 9.418716439870056e-06,
+      "loss": 0.2433,
+      "step": 3243
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.0500476227828563,
+      "learning_rate": 9.4182947998165e-06,
+      "loss": 0.2233,
+      "step": 3244
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.283056400277419,
+      "learning_rate": 9.417873016342552e-06,
+      "loss": 0.1511,
+      "step": 3245
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.0537092652541813,
+      "learning_rate": 9.417451089461906e-06,
+      "loss": 0.1838,
+      "step": 3246
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.1283688322005183,
+      "learning_rate": 9.417029019188257e-06,
+      "loss": 0.1836,
+      "step": 3247
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.5326285239196966,
+      "learning_rate": 9.416606805535306e-06,
+      "loss": 0.2396,
+      "step": 3248
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.3222351718703327,
+      "learning_rate": 9.416184448516759e-06,
+      "loss": 0.1967,
+      "step": 3249
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.4418276865934616,
+      "learning_rate": 9.415761948146323e-06,
+      "loss": 0.2167,
+      "step": 3250
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.3042344803008314,
+      "learning_rate": 9.415339304437717e-06,
+      "loss": 0.1768,
+      "step": 3251
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.0009069390623133,
+      "learning_rate": 9.414916517404659e-06,
+      "loss": 0.2303,
+      "step": 3252
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.059199947899894,
+      "learning_rate": 9.414493587060872e-06,
+      "loss": 0.2073,
+      "step": 3253
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.7497752888344469,
+      "learning_rate": 9.414070513420081e-06,
+      "loss": 0.2176,
+      "step": 3254
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.2943715570262209,
+      "learning_rate": 9.413647296496027e-06,
+      "loss": 0.2037,
+      "step": 3255
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.1902299921589183,
+      "learning_rate": 9.413223936302442e-06,
+      "loss": 0.2192,
+      "step": 3256
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.268153068304471,
+      "learning_rate": 9.41280043285307e-06,
+      "loss": 0.1462,
+      "step": 3257
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.4580853614700728,
+      "learning_rate": 9.412376786161659e-06,
+      "loss": 0.2055,
+      "step": 3258
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.9722548311399911,
+      "learning_rate": 9.41195299624196e-06,
+      "loss": 0.2459,
+      "step": 3259
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.4644266323034592,
+      "learning_rate": 9.41152906310773e-06,
+      "loss": 0.2153,
+      "step": 3260
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.9625364681401054,
+      "learning_rate": 9.41110498677273e-06,
+      "loss": 0.2069,
+      "step": 3261
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.2598016541526025,
+      "learning_rate": 9.410680767250724e-06,
+      "loss": 0.2117,
+      "step": 3262
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.8142346539661697,
+      "learning_rate": 9.410256404555485e-06,
+      "loss": 0.1863,
+      "step": 3263
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.4486717878233748,
+      "learning_rate": 9.409831898700787e-06,
+      "loss": 0.2568,
+      "step": 3264
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.1927506612068217,
+      "learning_rate": 9.409407249700411e-06,
+      "loss": 0.2386,
+      "step": 3265
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.0684436236607988,
+      "learning_rate": 9.408982457568138e-06,
+      "loss": 0.2214,
+      "step": 3266
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.166445235241551,
+      "learning_rate": 9.40855752231776e-06,
+      "loss": 0.1859,
+      "step": 3267
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.282113795177243,
+      "learning_rate": 9.408132443963071e-06,
+      "loss": 0.18,
+      "step": 3268
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.1519133733964457,
+      "learning_rate": 9.407707222517865e-06,
+      "loss": 0.1982,
+      "step": 3269
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.134166797533817,
+      "learning_rate": 9.407281857995951e-06,
+      "loss": 0.1748,
+      "step": 3270
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.2200128358318134,
+      "learning_rate": 9.406856350411133e-06,
+      "loss": 0.2325,
+      "step": 3271
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.8117761643034518,
+      "learning_rate": 9.406430699777222e-06,
+      "loss": 0.1688,
+      "step": 3272
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.052893488577849,
+      "learning_rate": 9.406004906108037e-06,
+      "loss": 0.1808,
+      "step": 3273
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.3743138470690608,
+      "learning_rate": 9.4055789694174e-06,
+      "loss": 0.2245,
+      "step": 3274
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.6063863855064964,
+      "learning_rate": 9.405152889719135e-06,
+      "loss": 0.2323,
+      "step": 3275
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.0984394140620262,
+      "learning_rate": 9.404726667027076e-06,
+      "loss": 0.2094,
+      "step": 3276
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.3094723840520193,
+      "learning_rate": 9.404300301355057e-06,
+      "loss": 0.2206,
+      "step": 3277
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.466925355081902,
+      "learning_rate": 9.403873792716915e-06,
+      "loss": 0.2031,
+      "step": 3278
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.370459858355884,
+      "learning_rate": 9.403447141126498e-06,
+      "loss": 0.2488,
+      "step": 3279
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.2514533088143556,
+      "learning_rate": 9.403020346597655e-06,
+      "loss": 0.1467,
+      "step": 3280
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.8523002324940065,
+      "learning_rate": 9.40259340914424e-06,
+      "loss": 0.2052,
+      "step": 3281
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.8270917628710085,
+      "learning_rate": 9.40216632878011e-06,
+      "loss": 0.188,
+      "step": 3282
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.306186624950565,
+      "learning_rate": 9.401739105519131e-06,
+      "loss": 0.1936,
+      "step": 3283
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.2002947683074896,
+      "learning_rate": 9.401311739375169e-06,
+      "loss": 0.2357,
+      "step": 3284
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.9652881240892959,
+      "learning_rate": 9.400884230362097e-06,
+      "loss": 0.1908,
+      "step": 3285
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.2094876592981503,
+      "learning_rate": 9.400456578493791e-06,
+      "loss": 0.1925,
+      "step": 3286
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.5035367957765657,
+      "learning_rate": 9.400028783784136e-06,
+      "loss": 0.1989,
+      "step": 3287
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.2892806215411166,
+      "learning_rate": 9.399600846247015e-06,
+      "loss": 0.2393,
+      "step": 3288
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.3564906129900676,
+      "learning_rate": 9.399172765896322e-06,
+      "loss": 0.2167,
+      "step": 3289
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.221636118062383,
+      "learning_rate": 9.39874454274595e-06,
+      "loss": 0.1722,
+      "step": 3290
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.9610429760180924,
+      "learning_rate": 9.398316176809801e-06,
+      "loss": 0.1569,
+      "step": 3291
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.208115515074719,
+      "learning_rate": 9.39788766810178e-06,
+      "loss": 0.2025,
+      "step": 3292
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.1204690255933578,
+      "learning_rate": 9.397459016635797e-06,
+      "loss": 0.1701,
+      "step": 3293
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.3434010138440782,
+      "learning_rate": 9.397030222425765e-06,
+      "loss": 0.191,
+      "step": 3294
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.0651715924933463,
+      "learning_rate": 9.396601285485603e-06,
+      "loss": 0.2189,
+      "step": 3295
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.590266412886869,
+      "learning_rate": 9.396172205829235e-06,
+      "loss": 0.2445,
+      "step": 3296
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.2534501978464756,
+      "learning_rate": 9.395742983470589e-06,
+      "loss": 0.241,
+      "step": 3297
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.5697827038533343,
+      "learning_rate": 9.3953136184236e-06,
+      "loss": 0.1904,
+      "step": 3298
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.4560185049922685,
+      "learning_rate": 9.3948841107022e-06,
+      "loss": 0.2443,
+      "step": 3299
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.762260190986907,
+      "learning_rate": 9.394454460320337e-06,
+      "loss": 0.2298,
+      "step": 3300
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.1835697851493159,
+      "learning_rate": 9.394024667291955e-06,
+      "loss": 0.1985,
+      "step": 3301
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.1091834762624069,
+      "learning_rate": 9.393594731631006e-06,
+      "loss": 0.2118,
+      "step": 3302
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.1388635555831577,
+      "learning_rate": 9.393164653351444e-06,
+      "loss": 0.1852,
+      "step": 3303
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.1263910615830721,
+      "learning_rate": 9.392734432467232e-06,
+      "loss": 0.2273,
+      "step": 3304
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.3298834876771541,
+      "learning_rate": 9.392304068992335e-06,
+      "loss": 0.2195,
+      "step": 3305
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.5072510494246556,
+      "learning_rate": 9.391873562940721e-06,
+      "loss": 0.1977,
+      "step": 3306
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.0185562857947241,
+      "learning_rate": 9.391442914326367e-06,
+      "loss": 0.2536,
+      "step": 3307
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.8991015343782708,
+      "learning_rate": 9.391012123163249e-06,
+      "loss": 0.1833,
+      "step": 3308
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.1707969109279301,
+      "learning_rate": 9.390581189465354e-06,
+      "loss": 0.2109,
+      "step": 3309
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.314872766821686,
+      "learning_rate": 9.390150113246667e-06,
+      "loss": 0.1739,
+      "step": 3310
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.2081430200697176,
+      "learning_rate": 9.389718894521186e-06,
+      "loss": 0.1978,
+      "step": 3311
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.997199538671376,
+      "learning_rate": 9.389287533302902e-06,
+      "loss": 0.1961,
+      "step": 3312
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.229366347559321,
+      "learning_rate": 9.388856029605821e-06,
+      "loss": 0.246,
+      "step": 3313
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.0939353713408155,
+      "learning_rate": 9.388424383443952e-06,
+      "loss": 0.2037,
+      "step": 3314
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.208235319882142,
+      "learning_rate": 9.387992594831302e-06,
+      "loss": 0.1839,
+      "step": 3315
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.004817454489758,
+      "learning_rate": 9.38756066378189e-06,
+      "loss": 0.2553,
+      "step": 3316
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.0384751538519483,
+      "learning_rate": 9.387128590309735e-06,
+      "loss": 0.2316,
+      "step": 3317
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.878084782191895,
+      "learning_rate": 9.386696374428863e-06,
+      "loss": 0.1927,
+      "step": 3318
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.2269667313157415,
+      "learning_rate": 9.386264016153306e-06,
+      "loss": 0.1799,
+      "step": 3319
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.1580603690082758,
+      "learning_rate": 9.385831515497095e-06,
+      "loss": 0.1982,
+      "step": 3320
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.382170843510891,
+      "learning_rate": 9.385398872474272e-06,
+      "loss": 0.2222,
+      "step": 3321
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.9844718764333906,
+      "learning_rate": 9.384966087098878e-06,
+      "loss": 0.1699,
+      "step": 3322
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.616412871039775,
+      "learning_rate": 9.384533159384965e-06,
+      "loss": 0.1852,
+      "step": 3323
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.8307192536762128,
+      "learning_rate": 9.384100089346584e-06,
+      "loss": 0.1738,
+      "step": 3324
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.4665023149841296,
+      "learning_rate": 9.383666876997793e-06,
+      "loss": 0.2268,
+      "step": 3325
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.2257866425331785,
+      "learning_rate": 9.383233522352654e-06,
+      "loss": 0.1649,
+      "step": 3326
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.548552529492027,
+      "learning_rate": 9.382800025425235e-06,
+      "loss": 0.2728,
+      "step": 3327
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.0877443400159903,
+      "learning_rate": 9.382366386229606e-06,
+      "loss": 0.1679,
+      "step": 3328
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.814720539861424,
+      "learning_rate": 9.381932604779844e-06,
+      "loss": 0.2349,
+      "step": 3329
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.8983744879062706,
+      "learning_rate": 9.38149868109003e-06,
+      "loss": 0.1853,
+      "step": 3330
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.0781434971494195,
+      "learning_rate": 9.381064615174252e-06,
+      "loss": 0.1643,
+      "step": 3331
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.44790762009699,
+      "learning_rate": 9.380630407046594e-06,
+      "loss": 0.2287,
+      "step": 3332
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.2842657720337298,
+      "learning_rate": 9.380196056721156e-06,
+      "loss": 0.2151,
+      "step": 3333
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.6835440475668269,
+      "learning_rate": 9.379761564212034e-06,
+      "loss": 0.2058,
+      "step": 3334
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.0742363720228738,
+      "learning_rate": 9.379326929533332e-06,
+      "loss": 0.1942,
+      "step": 3335
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.9425722613668038,
+      "learning_rate": 9.378892152699162e-06,
+      "loss": 0.1635,
+      "step": 3336
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.5648551072057528,
+      "learning_rate": 9.378457233723633e-06,
+      "loss": 0.2262,
+      "step": 3337
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 2.7196269758473033,
+      "learning_rate": 9.378022172620865e-06,
+      "loss": 0.2142,
+      "step": 3338
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.430075965281619,
+      "learning_rate": 9.37758696940498e-06,
+      "loss": 0.2294,
+      "step": 3339
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.2974053987293108,
+      "learning_rate": 9.377151624090104e-06,
+      "loss": 0.2343,
+      "step": 3340
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.1454115076713254,
+      "learning_rate": 9.376716136690369e-06,
+      "loss": 0.2638,
+      "step": 3341
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.2750693846971501,
+      "learning_rate": 9.376280507219913e-06,
+      "loss": 0.2132,
+      "step": 3342
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.1779309934211313,
+      "learning_rate": 9.375844735692873e-06,
+      "loss": 0.1744,
+      "step": 3343
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.3314039789392438,
+      "learning_rate": 9.3754088221234e-06,
+      "loss": 0.1859,
+      "step": 3344
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.007564885748386,
+      "learning_rate": 9.374972766525638e-06,
+      "loss": 0.2265,
+      "step": 3345
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.9843393832151374,
+      "learning_rate": 9.374536568913743e-06,
+      "loss": 0.1903,
+      "step": 3346
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.414897238295894,
+      "learning_rate": 9.374100229301877e-06,
+      "loss": 0.216,
+      "step": 3347
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.9477164306778282,
+      "learning_rate": 9.373663747704202e-06,
+      "loss": 0.2318,
+      "step": 3348
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.1938204358513997,
+      "learning_rate": 9.373227124134888e-06,
+      "loss": 0.2283,
+      "step": 3349
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.5204301592549707,
+      "learning_rate": 9.372790358608106e-06,
+      "loss": 0.253,
+      "step": 3350
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.4652276184650341,
+      "learning_rate": 9.372353451138035e-06,
+      "loss": 0.2441,
+      "step": 3351
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.2206280067219304,
+      "learning_rate": 9.371916401738855e-06,
+      "loss": 0.1651,
+      "step": 3352
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.00637925795717,
+      "learning_rate": 9.371479210424755e-06,
+      "loss": 0.1726,
+      "step": 3353
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.2532247216781283,
+      "learning_rate": 9.371041877209927e-06,
+      "loss": 0.2434,
+      "step": 3354
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.967435354311242,
+      "learning_rate": 9.370604402108565e-06,
+      "loss": 0.1471,
+      "step": 3355
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.2381179909529387,
+      "learning_rate": 9.370166785134872e-06,
+      "loss": 0.2043,
+      "step": 3356
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.9358676467996156,
+      "learning_rate": 9.369729026303052e-06,
+      "loss": 0.2422,
+      "step": 3357
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.0384265452119081,
+      "learning_rate": 9.369291125627314e-06,
+      "loss": 0.1892,
+      "step": 3358
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.6463761571252031,
+      "learning_rate": 9.368853083121875e-06,
+      "loss": 0.1787,
+      "step": 3359
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.2575905467797537,
+      "learning_rate": 9.368414898800951e-06,
+      "loss": 0.2433,
+      "step": 3360
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.148965611965069,
+      "learning_rate": 9.367976572678769e-06,
+      "loss": 0.1739,
+      "step": 3361
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.3491546068919977,
+      "learning_rate": 9.367538104769555e-06,
+      "loss": 0.192,
+      "step": 3362
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.6476862673604906,
+      "learning_rate": 9.367099495087542e-06,
+      "loss": 0.2357,
+      "step": 3363
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.098736002844294,
+      "learning_rate": 9.366660743646969e-06,
+      "loss": 0.2202,
+      "step": 3364
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.187087553639016,
+      "learning_rate": 9.366221850462076e-06,
+      "loss": 0.2564,
+      "step": 3365
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.1147071006127478,
+      "learning_rate": 9.365782815547113e-06,
+      "loss": 0.2448,
+      "step": 3366
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.70797078837883,
+      "learning_rate": 9.365343638916328e-06,
+      "loss": 0.1971,
+      "step": 3367
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.2938758687512202,
+      "learning_rate": 9.364904320583978e-06,
+      "loss": 0.2173,
+      "step": 3368
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.9648140196138664,
+      "learning_rate": 9.364464860564324e-06,
+      "loss": 0.2231,
+      "step": 3369
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.2707708879538742,
+      "learning_rate": 9.364025258871631e-06,
+      "loss": 0.2271,
+      "step": 3370
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.0347465676107663,
+      "learning_rate": 9.363585515520168e-06,
+      "loss": 0.1865,
+      "step": 3371
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.4201774786882786,
+      "learning_rate": 9.36314563052421e-06,
+      "loss": 0.2521,
+      "step": 3372
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.9240282368217848,
+      "learning_rate": 9.362705603898038e-06,
+      "loss": 0.197,
+      "step": 3373
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.3319545623093136,
+      "learning_rate": 9.362265435655932e-06,
+      "loss": 0.1728,
+      "step": 3374
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.9675266840020769,
+      "learning_rate": 9.361825125812181e-06,
+      "loss": 0.2201,
+      "step": 3375
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.5756213232313,
+      "learning_rate": 9.361384674381078e-06,
+      "loss": 0.2377,
+      "step": 3376
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.5188138979746864,
+      "learning_rate": 9.360944081376921e-06,
+      "loss": 0.2254,
+      "step": 3377
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.644860821317639,
+      "learning_rate": 9.360503346814011e-06,
+      "loss": 0.1745,
+      "step": 3378
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.1931572031125863,
+      "learning_rate": 9.360062470706656e-06,
+      "loss": 0.2392,
+      "step": 3379
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.9476560731810604,
+      "learning_rate": 9.359621453069165e-06,
+      "loss": 0.1755,
+      "step": 3380
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 2.774313946709789,
+      "learning_rate": 9.359180293915856e-06,
+      "loss": 0.2426,
+      "step": 3381
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.8217427591536367,
+      "learning_rate": 9.358738993261046e-06,
+      "loss": 0.185,
+      "step": 3382
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.4841779552981227,
+      "learning_rate": 9.358297551119061e-06,
+      "loss": 0.2472,
+      "step": 3383
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.2081106199216842,
+      "learning_rate": 9.357855967504235e-06,
+      "loss": 0.2167,
+      "step": 3384
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.4811553583311272,
+      "learning_rate": 9.357414242430895e-06,
+      "loss": 0.1781,
+      "step": 3385
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.8797798267245154,
+      "learning_rate": 9.356972375913385e-06,
+      "loss": 0.2432,
+      "step": 3386
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.3067075415726623,
+      "learning_rate": 9.356530367966045e-06,
+      "loss": 0.2323,
+      "step": 3387
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.7360382782175803,
+      "learning_rate": 9.356088218603226e-06,
+      "loss": 0.2952,
+      "step": 3388
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.1536104836943053,
+      "learning_rate": 9.355645927839276e-06,
+      "loss": 0.1858,
+      "step": 3389
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.9983808542802693,
+      "learning_rate": 9.355203495688556e-06,
+      "loss": 0.1947,
+      "step": 3390
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 2.186700484079255,
+      "learning_rate": 9.354760922165425e-06,
+      "loss": 0.1894,
+      "step": 3391
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.9812613053794326,
+      "learning_rate": 9.354318207284252e-06,
+      "loss": 0.1761,
+      "step": 3392
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.115246370187576,
+      "learning_rate": 9.353875351059405e-06,
+      "loss": 0.2818,
+      "step": 3393
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.0838124611425966,
+      "learning_rate": 9.35343235350526e-06,
+      "loss": 0.1931,
+      "step": 3394
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.3841890929279212,
+      "learning_rate": 9.352989214636197e-06,
+      "loss": 0.1459,
+      "step": 3395
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.4499175521773402,
+      "learning_rate": 9.352545934466602e-06,
+      "loss": 0.2144,
+      "step": 3396
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.0117591533178498,
+      "learning_rate": 9.352102513010863e-06,
+      "loss": 0.1862,
+      "step": 3397
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.9989425909865767,
+      "learning_rate": 9.351658950283374e-06,
+      "loss": 0.1846,
+      "step": 3398
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.1823524868721618,
+      "learning_rate": 9.351215246298533e-06,
+      "loss": 0.2151,
+      "step": 3399
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.8841584145135845,
+      "learning_rate": 9.350771401070744e-06,
+      "loss": 0.2432,
+      "step": 3400
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.9922060476657047,
+      "learning_rate": 9.350327414614411e-06,
+      "loss": 0.1846,
+      "step": 3401
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.0018596940230833,
+      "learning_rate": 9.349883286943951e-06,
+      "loss": 0.1544,
+      "step": 3402
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.8680836647833723,
+      "learning_rate": 9.349439018073776e-06,
+      "loss": 0.1831,
+      "step": 3403
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.036707245188668,
+      "learning_rate": 9.34899460801831e-06,
+      "loss": 0.234,
+      "step": 3404
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.706687570756996,
+      "learning_rate": 9.348550056791979e-06,
+      "loss": 0.1989,
+      "step": 3405
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.248020507641426,
+      "learning_rate": 9.348105364409212e-06,
+      "loss": 0.1842,
+      "step": 3406
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.4854241001693074,
+      "learning_rate": 9.347660530884443e-06,
+      "loss": 0.2307,
+      "step": 3407
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.180371806003346,
+      "learning_rate": 9.347215556232115e-06,
+      "loss": 0.1982,
+      "step": 3408
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.1847817781770318,
+      "learning_rate": 9.34677044046667e-06,
+      "loss": 0.2138,
+      "step": 3409
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.3180200147071006,
+      "learning_rate": 9.346325183602557e-06,
+      "loss": 0.1786,
+      "step": 3410
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.503313760345442,
+      "learning_rate": 9.34587978565423e-06,
+      "loss": 0.2816,
+      "step": 3411
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.350541556091833,
+      "learning_rate": 9.345434246636145e-06,
+      "loss": 0.1855,
+      "step": 3412
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.4049543160218259,
+      "learning_rate": 9.344988566562764e-06,
+      "loss": 0.2629,
+      "step": 3413
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.2104431735399839,
+      "learning_rate": 9.344542745448559e-06,
+      "loss": 0.1977,
+      "step": 3414
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.4098770837437813,
+      "learning_rate": 9.344096783307995e-06,
+      "loss": 0.2188,
+      "step": 3415
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.8052424572534944,
+      "learning_rate": 9.343650680155554e-06,
+      "loss": 0.2237,
+      "step": 3416
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.2691806719970078,
+      "learning_rate": 9.343204436005713e-06,
+      "loss": 0.2151,
+      "step": 3417
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.3440727883038068,
+      "learning_rate": 9.342758050872958e-06,
+      "loss": 0.2558,
+      "step": 3418
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.8691344704521682,
+      "learning_rate": 9.342311524771779e-06,
+      "loss": 0.2327,
+      "step": 3419
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.2100595090799668,
+      "learning_rate": 9.341864857716671e-06,
+      "loss": 0.2117,
+      "step": 3420
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.0245172159572706,
+      "learning_rate": 9.341418049722136e-06,
+      "loss": 0.2741,
+      "step": 3421
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.1691467905057233,
+      "learning_rate": 9.340971100802672e-06,
+      "loss": 0.2047,
+      "step": 3422
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.8932036372096284,
+      "learning_rate": 9.340524010972789e-06,
+      "loss": 0.1972,
+      "step": 3423
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.3581365748402436,
+      "learning_rate": 9.340076780247003e-06,
+      "loss": 0.27,
+      "step": 3424
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.9184511195656214,
+      "learning_rate": 9.339629408639828e-06,
+      "loss": 0.1238,
+      "step": 3425
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.6973609338263328,
+      "learning_rate": 9.339181896165786e-06,
+      "loss": 0.1919,
+      "step": 3426
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.4056312471338301,
+      "learning_rate": 9.338734242839405e-06,
+      "loss": 0.255,
+      "step": 3427
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.1609911611664736,
+      "learning_rate": 9.338286448675216e-06,
+      "loss": 0.2014,
+      "step": 3428
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.2428689652867788,
+      "learning_rate": 9.337838513687754e-06,
+      "loss": 0.1852,
+      "step": 3429
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.332871659979454,
+      "learning_rate": 9.337390437891559e-06,
+      "loss": 0.2165,
+      "step": 3430
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.6246631587271103,
+      "learning_rate": 9.336942221301176e-06,
+      "loss": 0.2522,
+      "step": 3431
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.4933391141602195,
+      "learning_rate": 9.336493863931155e-06,
+      "loss": 0.1616,
+      "step": 3432
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.9770967533709778,
+      "learning_rate": 9.336045365796049e-06,
+      "loss": 0.2091,
+      "step": 3433
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.1257021553903206,
+      "learning_rate": 9.33559672691042e-06,
+      "loss": 0.2151,
+      "step": 3434
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 2.509332119205977,
+      "learning_rate": 9.335147947288824e-06,
+      "loss": 0.2071,
+      "step": 3435
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.3922064059030224,
+      "learning_rate": 9.334699026945835e-06,
+      "loss": 0.2084,
+      "step": 3436
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.3889438279697006,
+      "learning_rate": 9.334249965896022e-06,
+      "loss": 0.2424,
+      "step": 3437
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.1906147297468803,
+      "learning_rate": 9.333800764153964e-06,
+      "loss": 0.1758,
+      "step": 3438
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.2331261700534246,
+      "learning_rate": 9.333351421734239e-06,
+      "loss": 0.1996,
+      "step": 3439
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.1944686616277236,
+      "learning_rate": 9.332901938651437e-06,
+      "loss": 0.2419,
+      "step": 3440
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.0166969987410646,
+      "learning_rate": 9.332452314920144e-06,
+      "loss": 0.1886,
+      "step": 3441
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.2703077076602174,
+      "learning_rate": 9.33200255055496e-06,
+      "loss": 0.2026,
+      "step": 3442
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.3120715003809043,
+      "learning_rate": 9.331552645570482e-06,
+      "loss": 0.1792,
+      "step": 3443
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.150373153521467,
+      "learning_rate": 9.331102599981312e-06,
+      "loss": 0.2367,
+      "step": 3444
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.4854938757422544,
+      "learning_rate": 9.330652413802063e-06,
+      "loss": 0.2283,
+      "step": 3445
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.7036303024338226,
+      "learning_rate": 9.330202087047346e-06,
+      "loss": 0.1874,
+      "step": 3446
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.0113757188957337,
+      "learning_rate": 9.329751619731779e-06,
+      "loss": 0.2351,
+      "step": 3447
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.3716485539373546,
+      "learning_rate": 9.329301011869986e-06,
+      "loss": 0.2584,
+      "step": 3448
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.1535876817736108,
+      "learning_rate": 9.328850263476593e-06,
+      "loss": 0.1909,
+      "step": 3449
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.4242407327818865,
+      "learning_rate": 9.328399374566229e-06,
+      "loss": 0.2074,
+      "step": 3450
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.170626655705753,
+      "learning_rate": 9.327948345153531e-06,
+      "loss": 0.2196,
+      "step": 3451
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.9314774894235901,
+      "learning_rate": 9.327497175253145e-06,
+      "loss": 0.2003,
+      "step": 3452
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.4498643953530983,
+      "learning_rate": 9.327045864879708e-06,
+      "loss": 0.2307,
+      "step": 3453
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.5962218605233303,
+      "learning_rate": 9.326594414047877e-06,
+      "loss": 0.2175,
+      "step": 3454
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.0048740560819251,
+      "learning_rate": 9.326142822772301e-06,
+      "loss": 0.2288,
+      "step": 3455
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.130968782011266,
+      "learning_rate": 9.325691091067643e-06,
+      "loss": 0.1982,
+      "step": 3456
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.0254134148381202,
+      "learning_rate": 9.325239218948566e-06,
+      "loss": 0.1768,
+      "step": 3457
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.3250824051605634,
+      "learning_rate": 9.324787206429736e-06,
+      "loss": 0.2067,
+      "step": 3458
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.2279013627882462,
+      "learning_rate": 9.324335053525823e-06,
+      "loss": 0.1509,
+      "step": 3459
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.0277428359521081,
+      "learning_rate": 9.323882760251511e-06,
+      "loss": 0.1639,
+      "step": 3460
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.049492133875297,
+      "learning_rate": 9.323430326621478e-06,
+      "loss": 0.2058,
+      "step": 3461
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.2394462874999683,
+      "learning_rate": 9.32297775265041e-06,
+      "loss": 0.2101,
+      "step": 3462
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 2.2933789329883894,
+      "learning_rate": 9.322525038352997e-06,
+      "loss": 0.2842,
+      "step": 3463
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.7105489664180398,
+      "learning_rate": 9.322072183743937e-06,
+      "loss": 0.1848,
+      "step": 3464
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.4613008519033883,
+      "learning_rate": 9.321619188837928e-06,
+      "loss": 0.1757,
+      "step": 3465
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.3046655795831388,
+      "learning_rate": 9.321166053649674e-06,
+      "loss": 0.2002,
+      "step": 3466
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.178964443805501,
+      "learning_rate": 9.320712778193888e-06,
+      "loss": 0.2118,
+      "step": 3467
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.194810615465231,
+      "learning_rate": 9.320259362485277e-06,
+      "loss": 0.2427,
+      "step": 3468
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.776906264612395,
+      "learning_rate": 9.319805806538566e-06,
+      "loss": 0.226,
+      "step": 3469
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.5273795885871688,
+      "learning_rate": 9.319352110368474e-06,
+      "loss": 0.2137,
+      "step": 3470
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.5303902499292754,
+      "learning_rate": 9.318898273989726e-06,
+      "loss": 0.2379,
+      "step": 3471
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.1900723259698849,
+      "learning_rate": 9.31844429741706e-06,
+      "loss": 0.2075,
+      "step": 3472
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.3329749805473403,
+      "learning_rate": 9.317990180665206e-06,
+      "loss": 0.2161,
+      "step": 3473
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.0256941573161902,
+      "learning_rate": 9.317535923748909e-06,
+      "loss": 0.2469,
+      "step": 3474
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.6072336291775406,
+      "learning_rate": 9.317081526682912e-06,
+      "loss": 0.1877,
+      "step": 3475
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.2980757857612168,
+      "learning_rate": 9.316626989481967e-06,
+      "loss": 0.2112,
+      "step": 3476
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.231902229070083,
+      "learning_rate": 9.316172312160829e-06,
+      "loss": 0.2092,
+      "step": 3477
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.9236005518912395,
+      "learning_rate": 9.315717494734254e-06,
+      "loss": 0.1585,
+      "step": 3478
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.5841963646139892,
+      "learning_rate": 9.315262537217008e-06,
+      "loss": 0.2395,
+      "step": 3479
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.5003565759328497,
+      "learning_rate": 9.314807439623859e-06,
+      "loss": 0.1691,
+      "step": 3480
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.0632833720458617,
+      "learning_rate": 9.314352201969579e-06,
+      "loss": 0.1743,
+      "step": 3481
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.1489776489378425,
+      "learning_rate": 9.313896824268945e-06,
+      "loss": 0.1979,
+      "step": 3482
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.0395818726283128,
+      "learning_rate": 9.31344130653674e-06,
+      "loss": 0.1723,
+      "step": 3483
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.5826718104988462,
+      "learning_rate": 9.312985648787748e-06,
+      "loss": 0.2469,
+      "step": 3484
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.3477252127099517,
+      "learning_rate": 9.312529851036764e-06,
+      "loss": 0.2517,
+      "step": 3485
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.3564100940655588,
+      "learning_rate": 9.31207391329858e-06,
+      "loss": 0.1888,
+      "step": 3486
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.0381339287273423,
+      "learning_rate": 9.311617835587998e-06,
+      "loss": 0.1968,
+      "step": 3487
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.9048953702019458,
+      "learning_rate": 9.311161617919822e-06,
+      "loss": 0.2334,
+      "step": 3488
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.0229662345253994,
+      "learning_rate": 9.310705260308862e-06,
+      "loss": 0.2366,
+      "step": 3489
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.0739576056977749,
+      "learning_rate": 9.310248762769927e-06,
+      "loss": 0.2344,
+      "step": 3490
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.0712445848841659,
+      "learning_rate": 9.30979212531784e-06,
+      "loss": 0.2008,
+      "step": 3491
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.122302279552585,
+      "learning_rate": 9.309335347967423e-06,
+      "loss": 0.2453,
+      "step": 3492
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.9891664890263973,
+      "learning_rate": 9.3088784307335e-06,
+      "loss": 0.1525,
+      "step": 3493
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.195681344029875,
+      "learning_rate": 9.30842137363091e-06,
+      "loss": 0.2096,
+      "step": 3494
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.8998210588774679,
+      "learning_rate": 9.30796417667448e-06,
+      "loss": 0.1785,
+      "step": 3495
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.42815163003751,
+      "learning_rate": 9.30750683987906e-06,
+      "loss": 0.1815,
+      "step": 3496
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.504099078015984,
+      "learning_rate": 9.307049363259486e-06,
+      "loss": 0.2302,
+      "step": 3497
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.25019340062066,
+      "learning_rate": 9.306591746830617e-06,
+      "loss": 0.2151,
+      "step": 3498
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.055004202128975,
+      "learning_rate": 9.306133990607303e-06,
+      "loss": 0.1461,
+      "step": 3499
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.4369617444395177,
+      "learning_rate": 9.3056760946044e-06,
+      "loss": 0.2103,
+      "step": 3500
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.1219865453941,
+      "learning_rate": 9.305218058836778e-06,
+      "loss": 0.1914,
+      "step": 3501
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.503289732957185,
+      "learning_rate": 9.304759883319302e-06,
+      "loss": 0.2636,
+      "step": 3502
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.9868224727513982,
+      "learning_rate": 9.304301568066845e-06,
+      "loss": 0.2133,
+      "step": 3503
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.4674322552819783,
+      "learning_rate": 9.303843113094282e-06,
+      "loss": 0.2487,
+      "step": 3504
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.1132816532201204,
+      "learning_rate": 9.303384518416498e-06,
+      "loss": 0.1759,
+      "step": 3505
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.660245170455845,
+      "learning_rate": 9.302925784048377e-06,
+      "loss": 0.1989,
+      "step": 3506
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.217632002521164,
+      "learning_rate": 9.302466910004812e-06,
+      "loss": 0.2232,
+      "step": 3507
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.4133079027580142,
+      "learning_rate": 9.302007896300697e-06,
+      "loss": 0.2237,
+      "step": 3508
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.3674722286467274,
+      "learning_rate": 9.301548742950931e-06,
+      "loss": 0.1745,
+      "step": 3509
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.7649693199968244,
+      "learning_rate": 9.30108944997042e-06,
+      "loss": 0.1753,
+      "step": 3510
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.0571563505702504,
+      "learning_rate": 9.300630017374073e-06,
+      "loss": 0.2025,
+      "step": 3511
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.1413999824952383,
+      "learning_rate": 9.300170445176801e-06,
+      "loss": 0.204,
+      "step": 3512
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.0355278931456928,
+      "learning_rate": 9.299710733393524e-06,
+      "loss": 0.2467,
+      "step": 3513
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.3895388229577266,
+      "learning_rate": 9.299250882039163e-06,
+      "loss": 0.192,
+      "step": 3514
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.0479699323052294,
+      "learning_rate": 9.298790891128648e-06,
+      "loss": 0.2087,
+      "step": 3515
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.1294483084601346,
+      "learning_rate": 9.298330760676907e-06,
+      "loss": 0.2616,
+      "step": 3516
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.060934719452845,
+      "learning_rate": 9.297870490698879e-06,
+      "loss": 0.1805,
+      "step": 3517
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.0813042594685407,
+      "learning_rate": 9.297410081209503e-06,
+      "loss": 0.2277,
+      "step": 3518
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.2549444428386791,
+      "learning_rate": 9.296949532223726e-06,
+      "loss": 0.204,
+      "step": 3519
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.0614282763658274,
+      "learning_rate": 9.296488843756495e-06,
+      "loss": 0.1864,
+      "step": 3520
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.1472400923509154,
+      "learning_rate": 9.296028015822764e-06,
+      "loss": 0.227,
+      "step": 3521
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.8996017831737945,
+      "learning_rate": 9.295567048437495e-06,
+      "loss": 0.1946,
+      "step": 3522
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.143068645202973,
+      "learning_rate": 9.29510594161565e-06,
+      "loss": 0.1954,
+      "step": 3523
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.015419374337628,
+      "learning_rate": 9.294644695372196e-06,
+      "loss": 0.2174,
+      "step": 3524
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.160622800752788,
+      "learning_rate": 9.294183309722105e-06,
+      "loss": 0.2311,
+      "step": 3525
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.6503906396719126,
+      "learning_rate": 9.293721784680355e-06,
+      "loss": 0.2789,
+      "step": 3526
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.2152018249331327,
+      "learning_rate": 9.293260120261924e-06,
+      "loss": 0.2689,
+      "step": 3527
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.6194971587056648,
+      "learning_rate": 9.292798316481804e-06,
+      "loss": 0.1577,
+      "step": 3528
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.1031936703325362,
+      "learning_rate": 9.29233637335498e-06,
+      "loss": 0.2296,
+      "step": 3529
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.9199726066450207,
+      "learning_rate": 9.29187429089645e-06,
+      "loss": 0.2177,
+      "step": 3530
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.9101599255270683,
+      "learning_rate": 9.291412069121212e-06,
+      "loss": 0.1769,
+      "step": 3531
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.6566672519306067,
+      "learning_rate": 9.290949708044272e-06,
+      "loss": 0.2276,
+      "step": 3532
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.1963375112882426,
+      "learning_rate": 9.290487207680637e-06,
+      "loss": 0.2554,
+      "step": 3533
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.933450659955093,
+      "learning_rate": 9.290024568045319e-06,
+      "loss": 0.2158,
+      "step": 3534
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.4002480500823111,
+      "learning_rate": 9.289561789153336e-06,
+      "loss": 0.2109,
+      "step": 3535
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.2226003765099889,
+      "learning_rate": 9.28909887101971e-06,
+      "loss": 0.1594,
+      "step": 3536
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.9954299659819928,
+      "learning_rate": 9.288635813659472e-06,
+      "loss": 0.1564,
+      "step": 3537
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.8161115520100044,
+      "learning_rate": 9.288172617087646e-06,
+      "loss": 0.1957,
+      "step": 3538
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.6893343409637167,
+      "learning_rate": 9.287709281319273e-06,
+      "loss": 0.2309,
+      "step": 3539
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.2533437582827733,
+      "learning_rate": 9.287245806369389e-06,
+      "loss": 0.2151,
+      "step": 3540
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.6142149871665632,
+      "learning_rate": 9.286782192253041e-06,
+      "loss": 0.2661,
+      "step": 3541
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.117249143673758,
+      "learning_rate": 9.28631843898528e-06,
+      "loss": 0.235,
+      "step": 3542
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.573966363406278,
+      "learning_rate": 9.285854546581156e-06,
+      "loss": 0.2336,
+      "step": 3543
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.4931053695220595,
+      "learning_rate": 9.28539051505573e-06,
+      "loss": 0.2443,
+      "step": 3544
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.4203105903419608,
+      "learning_rate": 9.284926344424062e-06,
+      "loss": 0.2011,
+      "step": 3545
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.43710421213652,
+      "learning_rate": 9.284462034701222e-06,
+      "loss": 0.2439,
+      "step": 3546
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.2039137807544935,
+      "learning_rate": 9.28399758590228e-06,
+      "loss": 0.194,
+      "step": 3547
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.3128988562532744,
+      "learning_rate": 9.283532998042314e-06,
+      "loss": 0.2113,
+      "step": 3548
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.7185338534646433,
+      "learning_rate": 9.283068271136404e-06,
+      "loss": 0.2022,
+      "step": 3549
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.551993139151028,
+      "learning_rate": 9.282603405199633e-06,
+      "loss": 0.2344,
+      "step": 3550
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.2565812977051494,
+      "learning_rate": 9.282138400247094e-06,
+      "loss": 0.2247,
+      "step": 3551
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.3582082295753855,
+      "learning_rate": 9.28167325629388e-06,
+      "loss": 0.2587,
+      "step": 3552
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.9694772411353068,
+      "learning_rate": 9.281207973355088e-06,
+      "loss": 0.1575,
+      "step": 3553
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.4475980523604266,
+      "learning_rate": 9.280742551445825e-06,
+      "loss": 0.2331,
+      "step": 3554
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.2181104172378814,
+      "learning_rate": 9.280276990581199e-06,
+      "loss": 0.2268,
+      "step": 3555
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.231485441039595,
+      "learning_rate": 9.279811290776318e-06,
+      "loss": 0.174,
+      "step": 3556
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.942181242499757,
+      "learning_rate": 9.279345452046303e-06,
+      "loss": 0.2192,
+      "step": 3557
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.9407080419611651,
+      "learning_rate": 9.278879474406274e-06,
+      "loss": 0.2209,
+      "step": 3558
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.4718552290445062,
+      "learning_rate": 9.278413357871354e-06,
+      "loss": 0.2081,
+      "step": 3559
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 3.7084119052635374,
+      "learning_rate": 9.277947102456678e-06,
+      "loss": 0.2508,
+      "step": 3560
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.668090766714118,
+      "learning_rate": 9.27748070817738e-06,
+      "loss": 0.1935,
+      "step": 3561
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.3182892074279238,
+      "learning_rate": 9.277014175048597e-06,
+      "loss": 0.2006,
+      "step": 3562
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.986637663661987,
+      "learning_rate": 9.276547503085475e-06,
+      "loss": 0.1782,
+      "step": 3563
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.2622819343344807,
+      "learning_rate": 9.276080692303161e-06,
+      "loss": 0.2446,
+      "step": 3564
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.910601138801803,
+      "learning_rate": 9.27561374271681e-06,
+      "loss": 0.181,
+      "step": 3565
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.9570386088341374,
+      "learning_rate": 9.275146654341577e-06,
+      "loss": 0.18,
+      "step": 3566
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.2945617445425697,
+      "learning_rate": 9.274679427192627e-06,
+      "loss": 0.2289,
+      "step": 3567
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.0002609486153493,
+      "learning_rate": 9.274212061285124e-06,
+      "loss": 0.2294,
+      "step": 3568
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.9787369304608944,
+      "learning_rate": 9.27374455663424e-06,
+      "loss": 0.2047,
+      "step": 3569
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.6670065344543394,
+      "learning_rate": 9.27327691325515e-06,
+      "loss": 0.1984,
+      "step": 3570
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.0719842554628876,
+      "learning_rate": 9.272809131163032e-06,
+      "loss": 0.2325,
+      "step": 3571
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.0155601150620386,
+      "learning_rate": 9.272341210373075e-06,
+      "loss": 0.217,
+      "step": 3572
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.0647865630597695,
+      "learning_rate": 9.271873150900466e-06,
+      "loss": 0.1931,
+      "step": 3573
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.0239908764348211,
+      "learning_rate": 9.271404952760394e-06,
+      "loss": 0.1766,
+      "step": 3574
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.5488395486052002,
+      "learning_rate": 9.270936615968065e-06,
+      "loss": 0.2369,
+      "step": 3575
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.1207794287132442,
+      "learning_rate": 9.270468140538676e-06,
+      "loss": 0.2382,
+      "step": 3576
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.0849779404098476,
+      "learning_rate": 9.269999526487436e-06,
+      "loss": 0.2651,
+      "step": 3577
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.2829217333698493,
+      "learning_rate": 9.269530773829557e-06,
+      "loss": 0.2065,
+      "step": 3578
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.143898108538973,
+      "learning_rate": 9.269061882580251e-06,
+      "loss": 0.2442,
+      "step": 3579
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.919556229202662,
+      "learning_rate": 9.268592852754742e-06,
+      "loss": 0.2419,
+      "step": 3580
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.9303493367879768,
+      "learning_rate": 9.268123684368257e-06,
+      "loss": 0.1806,
+      "step": 3581
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.290891700544897,
+      "learning_rate": 9.26765437743602e-06,
+      "loss": 0.2058,
+      "step": 3582
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.9718103263485854,
+      "learning_rate": 9.26718493197327e-06,
+      "loss": 0.2225,
+      "step": 3583
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.3866781591465387,
+      "learning_rate": 9.266715347995241e-06,
+      "loss": 0.2359,
+      "step": 3584
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.9435678058297455,
+      "learning_rate": 9.26624562551718e-06,
+      "loss": 0.206,
+      "step": 3585
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.4530520984861988,
+      "learning_rate": 9.265775764554332e-06,
+      "loss": 0.2472,
+      "step": 3586
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.9567200845777295,
+      "learning_rate": 9.26530576512195e-06,
+      "loss": 0.1069,
+      "step": 3587
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.7825692006215694,
+      "learning_rate": 9.264835627235291e-06,
+      "loss": 0.1622,
+      "step": 3588
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.9784864261405227,
+      "learning_rate": 9.264365350909614e-06,
+      "loss": 0.2437,
+      "step": 3589
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.3361461501563399,
+      "learning_rate": 9.263894936160186e-06,
+      "loss": 0.211,
+      "step": 3590
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.1972267026315249,
+      "learning_rate": 9.263424383002275e-06,
+      "loss": 0.2573,
+      "step": 3591
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.2069492898663188,
+      "learning_rate": 9.26295369145116e-06,
+      "loss": 0.2072,
+      "step": 3592
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.9649685014546333,
+      "learning_rate": 9.262482861522114e-06,
+      "loss": 0.1713,
+      "step": 3593
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.3684277440359487,
+      "learning_rate": 9.262011893230423e-06,
+      "loss": 0.1601,
+      "step": 3594
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.1497541198445143,
+      "learning_rate": 9.261540786591377e-06,
+      "loss": 0.2117,
+      "step": 3595
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.8943228083136474,
+      "learning_rate": 9.261069541620267e-06,
+      "loss": 0.2162,
+      "step": 3596
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.029453247613401,
+      "learning_rate": 9.260598158332388e-06,
+      "loss": 0.2277,
+      "step": 3597
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.515757281513324,
+      "learning_rate": 9.260126636743043e-06,
+      "loss": 0.245,
+      "step": 3598
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.5568848972281617,
+      "learning_rate": 9.259654976867536e-06,
+      "loss": 0.2313,
+      "step": 3599
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.1456066781622052,
+      "learning_rate": 9.25918317872118e-06,
+      "loss": 0.1768,
+      "step": 3600
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.9269872065956813,
+      "learning_rate": 9.258711242319289e-06,
+      "loss": 0.1557,
+      "step": 3601
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.084058032239679,
+      "learning_rate": 9.258239167677183e-06,
+      "loss": 0.2191,
+      "step": 3602
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.0132921698429644,
+      "learning_rate": 9.257766954810185e-06,
+      "loss": 0.1661,
+      "step": 3603
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.1996621203837001,
+      "learning_rate": 9.257294603733621e-06,
+      "loss": 0.2176,
+      "step": 3604
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.4257188561256382,
+      "learning_rate": 9.25682211446283e-06,
+      "loss": 0.2908,
+      "step": 3605
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.5877332867494434,
+      "learning_rate": 9.256349487013142e-06,
+      "loss": 0.2517,
+      "step": 3606
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.7715260081251414,
+      "learning_rate": 9.255876721399904e-06,
+      "loss": 0.2584,
+      "step": 3607
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.2367128478684746,
+      "learning_rate": 9.25540381763846e-06,
+      "loss": 0.1922,
+      "step": 3608
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.0038436005484794,
+      "learning_rate": 9.25493077574416e-06,
+      "loss": 0.2052,
+      "step": 3609
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.7004228931010451,
+      "learning_rate": 9.25445759573236e-06,
+      "loss": 0.1862,
+      "step": 3610
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.2272981021764773,
+      "learning_rate": 9.253984277618419e-06,
+      "loss": 0.2021,
+      "step": 3611
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.8622563403650086,
+      "learning_rate": 9.253510821417706e-06,
+      "loss": 0.1844,
+      "step": 3612
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.2860355492486324,
+      "learning_rate": 9.253037227145582e-06,
+      "loss": 0.1827,
+      "step": 3613
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.0064329835790622,
+      "learning_rate": 9.252563494817426e-06,
+      "loss": 0.198,
+      "step": 3614
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.7551844236309382,
+      "learning_rate": 9.252089624448614e-06,
+      "loss": 0.2105,
+      "step": 3615
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.1300542276072911,
+      "learning_rate": 9.251615616054526e-06,
+      "loss": 0.2163,
+      "step": 3616
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.2103903657271156,
+      "learning_rate": 9.25114146965055e-06,
+      "loss": 0.1821,
+      "step": 3617
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.0252316778604322,
+      "learning_rate": 9.25066718525208e-06,
+      "loss": 0.1521,
+      "step": 3618
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.2165985397360106,
+      "learning_rate": 9.250192762874506e-06,
+      "loss": 0.1904,
+      "step": 3619
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.0570103321134365,
+      "learning_rate": 9.249718202533233e-06,
+      "loss": 0.1586,
+      "step": 3620
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.4545578544917708,
+      "learning_rate": 9.249243504243662e-06,
+      "loss": 0.1998,
+      "step": 3621
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.5252254936425895,
+      "learning_rate": 9.248768668021203e-06,
+      "loss": 0.2325,
+      "step": 3622
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.0764045939679863,
+      "learning_rate": 9.248293693881272e-06,
+      "loss": 0.2629,
+      "step": 3623
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.9672914665590175,
+      "learning_rate": 9.247818581839284e-06,
+      "loss": 0.238,
+      "step": 3624
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.6682664782131498,
+      "learning_rate": 9.247343331910661e-06,
+      "loss": 0.1725,
+      "step": 3625
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.454831553055597,
+      "learning_rate": 9.246867944110832e-06,
+      "loss": 0.1884,
+      "step": 3626
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.349672476125566,
+      "learning_rate": 9.246392418455226e-06,
+      "loss": 0.2763,
+      "step": 3627
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.182028542047186,
+      "learning_rate": 9.245916754959281e-06,
+      "loss": 0.2399,
+      "step": 3628
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.4935644852835803,
+      "learning_rate": 9.245440953638436e-06,
+      "loss": 0.206,
+      "step": 3629
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.4426765943357216,
+      "learning_rate": 9.244965014508138e-06,
+      "loss": 0.234,
+      "step": 3630
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.1376663924704116,
+      "learning_rate": 9.244488937583834e-06,
+      "loss": 0.1946,
+      "step": 3631
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.4015567265905164,
+      "learning_rate": 9.244012722880978e-06,
+      "loss": 0.2272,
+      "step": 3632
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.3176807946002547,
+      "learning_rate": 9.243536370415028e-06,
+      "loss": 0.2456,
+      "step": 3633
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.6353468277672574,
+      "learning_rate": 9.243059880201449e-06,
+      "loss": 0.2138,
+      "step": 3634
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.8967800899290342,
+      "learning_rate": 9.242583252255704e-06,
+      "loss": 0.2212,
+      "step": 3635
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.8931936858482628,
+      "learning_rate": 9.242106486593268e-06,
+      "loss": 0.162,
+      "step": 3636
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.8588822057279045,
+      "learning_rate": 9.241629583229617e-06,
+      "loss": 0.1986,
+      "step": 3637
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.2475690520234526,
+      "learning_rate": 9.241152542180229e-06,
+      "loss": 0.2378,
+      "step": 3638
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.4294618835030783,
+      "learning_rate": 9.240675363460592e-06,
+      "loss": 0.24,
+      "step": 3639
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.4103621665366821,
+      "learning_rate": 9.240198047086194e-06,
+      "loss": 0.2184,
+      "step": 3640
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.2723986448672517,
+      "learning_rate": 9.239720593072527e-06,
+      "loss": 0.2172,
+      "step": 3641
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.139900065332311,
+      "learning_rate": 9.239243001435095e-06,
+      "loss": 0.2712,
+      "step": 3642
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.1278648440395145,
+      "learning_rate": 9.238765272189395e-06,
+      "loss": 0.2445,
+      "step": 3643
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.0352577591554388,
+      "learning_rate": 9.238287405350938e-06,
+      "loss": 0.18,
+      "step": 3644
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.0796652154651545,
+      "learning_rate": 9.237809400935234e-06,
+      "loss": 0.2042,
+      "step": 3645
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.6827935262549181,
+      "learning_rate": 9.2373312589578e-06,
+      "loss": 0.1523,
+      "step": 3646
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.2113499054523675,
+      "learning_rate": 9.236852979434156e-06,
+      "loss": 0.2506,
+      "step": 3647
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.909994895900767,
+      "learning_rate": 9.236374562379828e-06,
+      "loss": 0.1902,
+      "step": 3648
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.001261680748606,
+      "learning_rate": 9.235896007810348e-06,
+      "loss": 0.1884,
+      "step": 3649
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.274037847857023,
+      "learning_rate": 9.235417315741244e-06,
+      "loss": 0.2243,
+      "step": 3650
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.8240224222510161,
+      "learning_rate": 9.234938486188061e-06,
+      "loss": 0.22,
+      "step": 3651
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.8363714785076277,
+      "learning_rate": 9.234459519166338e-06,
+      "loss": 0.2029,
+      "step": 3652
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.9823093511023441,
+      "learning_rate": 9.233980414691626e-06,
+      "loss": 0.2189,
+      "step": 3653
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.3427967444044393,
+      "learning_rate": 9.233501172779475e-06,
+      "loss": 0.2264,
+      "step": 3654
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.2851094872219537,
+      "learning_rate": 9.23302179344544e-06,
+      "loss": 0.2146,
+      "step": 3655
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.2380636637763722,
+      "learning_rate": 9.232542276705083e-06,
+      "loss": 0.179,
+      "step": 3656
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.7460503858926325,
+      "learning_rate": 9.232062622573972e-06,
+      "loss": 0.2057,
+      "step": 3657
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.543838806146281,
+      "learning_rate": 9.231582831067673e-06,
+      "loss": 0.2753,
+      "step": 3658
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.1219034062511,
+      "learning_rate": 9.231102902201763e-06,
+      "loss": 0.185,
+      "step": 3659
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.082702626396552,
+      "learning_rate": 9.230622835991818e-06,
+      "loss": 0.1874,
+      "step": 3660
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.2706264496169708,
+      "learning_rate": 9.230142632453423e-06,
+      "loss": 0.216,
+      "step": 3661
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.5567803804752107,
+      "learning_rate": 9.229662291602168e-06,
+      "loss": 0.1675,
+      "step": 3662
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.0065767341206588,
+      "learning_rate": 9.22918181345364e-06,
+      "loss": 0.1401,
+      "step": 3663
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.4286250429544785,
+      "learning_rate": 9.228701198023441e-06,
+      "loss": 0.2153,
+      "step": 3664
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.864250260659538,
+      "learning_rate": 9.228220445327168e-06,
+      "loss": 0.2015,
+      "step": 3665
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.4437868326719552,
+      "learning_rate": 9.227739555380428e-06,
+      "loss": 0.1798,
+      "step": 3666
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.8857813787569582,
+      "learning_rate": 9.227258528198832e-06,
+      "loss": 0.2413,
+      "step": 3667
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.3122031956240663,
+      "learning_rate": 9.226777363797991e-06,
+      "loss": 0.2146,
+      "step": 3668
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.9002976772964382,
+      "learning_rate": 9.226296062193529e-06,
+      "loss": 0.1344,
+      "step": 3669
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.5567945598080268,
+      "learning_rate": 9.225814623401064e-06,
+      "loss": 0.2215,
+      "step": 3670
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.2230147485159562,
+      "learning_rate": 9.225333047436228e-06,
+      "loss": 0.2013,
+      "step": 3671
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.2671921411281124,
+      "learning_rate": 9.224851334314651e-06,
+      "loss": 0.2144,
+      "step": 3672
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.9890769574764441,
+      "learning_rate": 9.22436948405197e-06,
+      "loss": 0.1817,
+      "step": 3673
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.0276006768958952,
+      "learning_rate": 9.223887496663828e-06,
+      "loss": 0.1603,
+      "step": 3674
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.5020009344500596,
+      "learning_rate": 9.223405372165866e-06,
+      "loss": 0.1931,
+      "step": 3675
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 3.1135645106934517,
+      "learning_rate": 9.222923110573738e-06,
+      "loss": 0.189,
+      "step": 3676
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.163876483442885,
+      "learning_rate": 9.222440711903099e-06,
+      "loss": 0.2241,
+      "step": 3677
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.4390030860026117,
+      "learning_rate": 9.221958176169605e-06,
+      "loss": 0.1769,
+      "step": 3678
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.1668169664173522,
+      "learning_rate": 9.22147550338892e-06,
+      "loss": 0.1946,
+      "step": 3679
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.0875096090484184,
+      "learning_rate": 9.220992693576715e-06,
+      "loss": 0.1645,
+      "step": 3680
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.3048345088939843,
+      "learning_rate": 9.22050974674866e-06,
+      "loss": 0.1621,
+      "step": 3681
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.3949158516163331,
+      "learning_rate": 9.220026662920429e-06,
+      "loss": 0.1749,
+      "step": 3682
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.2396965482711895,
+      "learning_rate": 9.219543442107708e-06,
+      "loss": 0.2592,
+      "step": 3683
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.9409546223871756,
+      "learning_rate": 9.219060084326179e-06,
+      "loss": 0.1962,
+      "step": 3684
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.9861513205967557,
+      "learning_rate": 9.218576589591535e-06,
+      "loss": 0.1821,
+      "step": 3685
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.2652507113738511,
+      "learning_rate": 9.218092957919469e-06,
+      "loss": 0.1725,
+      "step": 3686
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.1628951773844303,
+      "learning_rate": 9.217609189325678e-06,
+      "loss": 0.1979,
+      "step": 3687
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.3619798990932075,
+      "learning_rate": 9.21712528382587e-06,
+      "loss": 0.2247,
+      "step": 3688
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.562746126340234,
+      "learning_rate": 9.216641241435748e-06,
+      "loss": 0.1994,
+      "step": 3689
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.0973336825071245,
+      "learning_rate": 9.216157062171028e-06,
+      "loss": 0.2235,
+      "step": 3690
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.0264052938397907,
+      "learning_rate": 9.215672746047423e-06,
+      "loss": 0.2063,
+      "step": 3691
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.7929362053088735,
+      "learning_rate": 9.215188293080659e-06,
+      "loss": 0.2699,
+      "step": 3692
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.0537762979594887,
+      "learning_rate": 9.21470370328646e-06,
+      "loss": 0.2177,
+      "step": 3693
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.1093412044911488,
+      "learning_rate": 9.214218976680554e-06,
+      "loss": 0.1201,
+      "step": 3694
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.3668116175425293,
+      "learning_rate": 9.213734113278675e-06,
+      "loss": 0.2186,
+      "step": 3695
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.3603914163287376,
+      "learning_rate": 9.213249113096565e-06,
+      "loss": 0.2466,
+      "step": 3696
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.1484756593138543,
+      "learning_rate": 9.212763976149965e-06,
+      "loss": 0.1681,
+      "step": 3697
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.9926715584293132,
+      "learning_rate": 9.212278702454626e-06,
+      "loss": 0.2417,
+      "step": 3698
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.3944341795392667,
+      "learning_rate": 9.211793292026297e-06,
+      "loss": 0.2012,
+      "step": 3699
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.033637654020085,
+      "learning_rate": 9.211307744880735e-06,
+      "loss": 0.2062,
+      "step": 3700
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.2883447975530842,
+      "learning_rate": 9.210822061033701e-06,
+      "loss": 0.1418,
+      "step": 3701
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.5107910996590512,
+      "learning_rate": 9.210336240500964e-06,
+      "loss": 0.1748,
+      "step": 3702
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.2040448465471412,
+      "learning_rate": 9.209850283298292e-06,
+      "loss": 0.2066,
+      "step": 3703
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.4353278907820288,
+      "learning_rate": 9.209364189441457e-06,
+      "loss": 0.2172,
+      "step": 3704
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.1586721284454649,
+      "learning_rate": 9.20887795894624e-06,
+      "loss": 0.1731,
+      "step": 3705
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.9290459045842381,
+      "learning_rate": 9.208391591828426e-06,
+      "loss": 0.2211,
+      "step": 3706
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.55597121827741,
+      "learning_rate": 9.207905088103797e-06,
+      "loss": 0.2828,
+      "step": 3707
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.3100570382966692,
+      "learning_rate": 9.207418447788152e-06,
+      "loss": 0.1924,
+      "step": 3708
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.2541103449467763,
+      "learning_rate": 9.206931670897284e-06,
+      "loss": 0.1969,
+      "step": 3709
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.158683367708135,
+      "learning_rate": 9.206444757446995e-06,
+      "loss": 0.2232,
+      "step": 3710
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.7773990297628768,
+      "learning_rate": 9.205957707453092e-06,
+      "loss": 0.2167,
+      "step": 3711
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.3486145109444414,
+      "learning_rate": 9.205470520931381e-06,
+      "loss": 0.2028,
+      "step": 3712
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.1650158255498853,
+      "learning_rate": 9.204983197897679e-06,
+      "loss": 0.2076,
+      "step": 3713
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.2326454800069877,
+      "learning_rate": 9.204495738367805e-06,
+      "loss": 0.2286,
+      "step": 3714
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.3710757509772749,
+      "learning_rate": 9.20400814235758e-06,
+      "loss": 0.2649,
+      "step": 3715
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.205618472112531,
+      "learning_rate": 9.203520409882835e-06,
+      "loss": 0.1578,
+      "step": 3716
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.3473233240192417,
+      "learning_rate": 9.2030325409594e-06,
+      "loss": 0.2054,
+      "step": 3717
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.5863879273542933,
+      "learning_rate": 9.202544535603111e-06,
+      "loss": 0.2508,
+      "step": 3718
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.112969607519301,
+      "learning_rate": 9.20205639382981e-06,
+      "loss": 0.1885,
+      "step": 3719
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.0828145906502495,
+      "learning_rate": 9.201568115655343e-06,
+      "loss": 0.1912,
+      "step": 3720
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.5461932065147703,
+      "learning_rate": 9.201079701095557e-06,
+      "loss": 0.1876,
+      "step": 3721
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.9758903479079541,
+      "learning_rate": 9.20059115016631e-06,
+      "loss": 0.1412,
+      "step": 3722
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.9189903945485726,
+      "learning_rate": 9.200102462883457e-06,
+      "loss": 0.2366,
+      "step": 3723
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.1937866554199787,
+      "learning_rate": 9.199613639262865e-06,
+      "loss": 0.2098,
+      "step": 3724
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.2697898065929178,
+      "learning_rate": 9.199124679320399e-06,
+      "loss": 0.2003,
+      "step": 3725
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.1606327556996516,
+      "learning_rate": 9.198635583071929e-06,
+      "loss": 0.2042,
+      "step": 3726
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.242009648000712,
+      "learning_rate": 9.198146350533335e-06,
+      "loss": 0.2314,
+      "step": 3727
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.3218266872838875,
+      "learning_rate": 9.197656981720498e-06,
+      "loss": 0.1958,
+      "step": 3728
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.1600307595813635,
+      "learning_rate": 9.197167476649299e-06,
+      "loss": 0.2397,
+      "step": 3729
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.0562850015246972,
+      "learning_rate": 9.19667783533563e-06,
+      "loss": 0.1493,
+      "step": 3730
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.8528065999051673,
+      "learning_rate": 9.196188057795388e-06,
+      "loss": 0.1385,
+      "step": 3731
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.9847280856811554,
+      "learning_rate": 9.195698144044467e-06,
+      "loss": 0.1701,
+      "step": 3732
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.8949721012149305,
+      "learning_rate": 9.19520809409877e-06,
+      "loss": 0.1871,
+      "step": 3733
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.5845268440091282,
+      "learning_rate": 9.194717907974208e-06,
+      "loss": 0.2365,
+      "step": 3734
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.397573713101278,
+      "learning_rate": 9.194227585686691e-06,
+      "loss": 0.2119,
+      "step": 3735
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.3108593799812276,
+      "learning_rate": 9.193737127252132e-06,
+      "loss": 0.232,
+      "step": 3736
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.3583821748902263,
+      "learning_rate": 9.193246532686457e-06,
+      "loss": 0.246,
+      "step": 3737
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.112382868044654,
+      "learning_rate": 9.192755802005587e-06,
+      "loss": 0.2288,
+      "step": 3738
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.5543933103642527,
+      "learning_rate": 9.192264935225452e-06,
+      "loss": 0.232,
+      "step": 3739
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.0606153046325562,
+      "learning_rate": 9.191773932361988e-06,
+      "loss": 0.2564,
+      "step": 3740
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.187537948729842,
+      "learning_rate": 9.19128279343113e-06,
+      "loss": 0.1758,
+      "step": 3741
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.3815480784988203,
+      "learning_rate": 9.190791518448823e-06,
+      "loss": 0.2512,
+      "step": 3742
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.2117007864795588,
+      "learning_rate": 9.190300107431016e-06,
+      "loss": 0.2361,
+      "step": 3743
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.1096595316167241,
+      "learning_rate": 9.189808560393654e-06,
+      "loss": 0.2122,
+      "step": 3744
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.0606791499846953,
+      "learning_rate": 9.189316877352699e-06,
+      "loss": 0.2328,
+      "step": 3745
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.0508345551424438,
+      "learning_rate": 9.18882505832411e-06,
+      "loss": 0.1764,
+      "step": 3746
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.0152072267326044,
+      "learning_rate": 9.188333103323849e-06,
+      "loss": 0.2192,
+      "step": 3747
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.0307752968866477,
+      "learning_rate": 9.187841012367886e-06,
+      "loss": 0.2333,
+      "step": 3748
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.1702583100699078,
+      "learning_rate": 9.1873487854722e-06,
+      "loss": 0.2255,
+      "step": 3749
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.0075955758140505,
+      "learning_rate": 9.186856422652759e-06,
+      "loss": 0.2353,
+      "step": 3750
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.0476344044432455,
+      "learning_rate": 9.186363923925554e-06,
+      "loss": 0.1795,
+      "step": 3751
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.3552969664298538,
+      "learning_rate": 9.185871289306568e-06,
+      "loss": 0.242,
+      "step": 3752
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.4676919909460384,
+      "learning_rate": 9.185378518811793e-06,
+      "loss": 0.2364,
+      "step": 3753
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.0705689284903697,
+      "learning_rate": 9.184885612457225e-06,
+      "loss": 0.1902,
+      "step": 3754
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.8047919251914469,
+      "learning_rate": 9.184392570258865e-06,
+      "loss": 0.1879,
+      "step": 3755
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.1445807358390923,
+      "learning_rate": 9.183899392232714e-06,
+      "loss": 0.2156,
+      "step": 3756
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.7237154792219052,
+      "learning_rate": 9.18340607839478e-06,
+      "loss": 0.1732,
+      "step": 3757
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.1015106706397861,
+      "learning_rate": 9.182912628761084e-06,
+      "loss": 0.2385,
+      "step": 3758
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.4117161373416134,
+      "learning_rate": 9.182419043347637e-06,
+      "loss": 0.2119,
+      "step": 3759
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.144941956061877,
+      "learning_rate": 9.181925322170461e-06,
+      "loss": 0.2127,
+      "step": 3760
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.1310407941041019,
+      "learning_rate": 9.181431465245587e-06,
+      "loss": 0.2238,
+      "step": 3761
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.0231717989631302,
+      "learning_rate": 9.180937472589043e-06,
+      "loss": 0.2036,
+      "step": 3762
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.1979308683825118,
+      "learning_rate": 9.180443344216863e-06,
+      "loss": 0.2601,
+      "step": 3763
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.9562061275596669,
+      "learning_rate": 9.179949080145088e-06,
+      "loss": 0.1753,
+      "step": 3764
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.2135213526538595,
+      "learning_rate": 9.179454680389762e-06,
+      "loss": 0.1509,
+      "step": 3765
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 3.179980276837791,
+      "learning_rate": 9.178960144966935e-06,
+      "loss": 0.1912,
+      "step": 3766
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.4143242224968988,
+      "learning_rate": 9.178465473892657e-06,
+      "loss": 0.2865,
+      "step": 3767
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.9802592745265337,
+      "learning_rate": 9.17797066718299e-06,
+      "loss": 0.2106,
+      "step": 3768
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.5152313185678465,
+      "learning_rate": 9.177475724853989e-06,
+      "loss": 0.2497,
+      "step": 3769
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.7005739661343349,
+      "learning_rate": 9.176980646921726e-06,
+      "loss": 0.2394,
+      "step": 3770
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.1858577289314791,
+      "learning_rate": 9.17648543340227e-06,
+      "loss": 0.2783,
+      "step": 3771
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.8663082369202804,
+      "learning_rate": 9.175990084311692e-06,
+      "loss": 0.2025,
+      "step": 3772
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.9846883511380025,
+      "learning_rate": 9.175494599666078e-06,
+      "loss": 0.2106,
+      "step": 3773
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.1889173584371593,
+      "learning_rate": 9.174998979481506e-06,
+      "loss": 0.2178,
+      "step": 3774
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.003057144891507,
+      "learning_rate": 9.174503223774067e-06,
+      "loss": 0.1857,
+      "step": 3775
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.0894931106095713,
+      "learning_rate": 9.174007332559851e-06,
+      "loss": 0.2352,
+      "step": 3776
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.3704882314434885,
+      "learning_rate": 9.173511305854959e-06,
+      "loss": 0.1908,
+      "step": 3777
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.9795081093578969,
+      "learning_rate": 9.17301514367549e-06,
+      "loss": 0.2462,
+      "step": 3778
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.82364206367917,
+      "learning_rate": 9.17251884603755e-06,
+      "loss": 0.2114,
+      "step": 3779
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.0421540099801112,
+      "learning_rate": 9.172022412957247e-06,
+      "loss": 0.2094,
+      "step": 3780
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.479501218107253,
+      "learning_rate": 9.1715258444507e-06,
+      "loss": 0.1784,
+      "step": 3781
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.0824270712219426,
+      "learning_rate": 9.171029140534023e-06,
+      "loss": 0.1661,
+      "step": 3782
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.494134285857273,
+      "learning_rate": 9.170532301223343e-06,
+      "loss": 0.2002,
+      "step": 3783
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.0315773003818536,
+      "learning_rate": 9.170035326534785e-06,
+      "loss": 0.1949,
+      "step": 3784
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.5374668582972932,
+      "learning_rate": 9.169538216484484e-06,
+      "loss": 0.2364,
+      "step": 3785
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.1807361262110336,
+      "learning_rate": 9.169040971088574e-06,
+      "loss": 0.2625,
+      "step": 3786
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.190794043292659,
+      "learning_rate": 9.168543590363195e-06,
+      "loss": 0.1606,
+      "step": 3787
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.9309536702447435,
+      "learning_rate": 9.168046074324497e-06,
+      "loss": 0.2361,
+      "step": 3788
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.5753960738689647,
+      "learning_rate": 9.167548422988624e-06,
+      "loss": 0.2016,
+      "step": 3789
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.854969392766836,
+      "learning_rate": 9.167050636371736e-06,
+      "loss": 0.138,
+      "step": 3790
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.5230540575521936,
+      "learning_rate": 9.166552714489984e-06,
+      "loss": 0.2011,
+      "step": 3791
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.4385794650044244,
+      "learning_rate": 9.166054657359537e-06,
+      "loss": 0.1986,
+      "step": 3792
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.072045718852821,
+      "learning_rate": 9.165556464996562e-06,
+      "loss": 0.2266,
+      "step": 3793
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.2960273201389643,
+      "learning_rate": 9.165058137417226e-06,
+      "loss": 0.1673,
+      "step": 3794
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 4.198220408441488,
+      "learning_rate": 9.164559674637709e-06,
+      "loss": 0.2349,
+      "step": 3795
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.4813438285211347,
+      "learning_rate": 9.164061076674189e-06,
+      "loss": 0.2109,
+      "step": 3796
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.287594114956201,
+      "learning_rate": 9.163562343542853e-06,
+      "loss": 0.2303,
+      "step": 3797
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.7932683532322886,
+      "learning_rate": 9.16306347525989e-06,
+      "loss": 0.1882,
+      "step": 3798
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.0106859864168214,
+      "learning_rate": 9.162564471841491e-06,
+      "loss": 0.1449,
+      "step": 3799
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.1265335062081578,
+      "learning_rate": 9.162065333303856e-06,
+      "loss": 0.2363,
+      "step": 3800
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.9884846251672452,
+      "learning_rate": 9.161566059663187e-06,
+      "loss": 0.2347,
+      "step": 3801
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.4669192906890611,
+      "learning_rate": 9.16106665093569e-06,
+      "loss": 0.2281,
+      "step": 3802
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.5679081877077818,
+      "learning_rate": 9.160567107137579e-06,
+      "loss": 0.2157,
+      "step": 3803
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.2844731651855612,
+      "learning_rate": 9.160067428285064e-06,
+      "loss": 0.2212,
+      "step": 3804
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.0892677972892384,
+      "learning_rate": 9.159567614394372e-06,
+      "loss": 0.221,
+      "step": 3805
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.261625670315417,
+      "learning_rate": 9.15906766548172e-06,
+      "loss": 0.1573,
+      "step": 3806
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.0141889194308862,
+      "learning_rate": 9.15856758156334e-06,
+      "loss": 0.186,
+      "step": 3807
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.5179753980230881,
+      "learning_rate": 9.158067362655467e-06,
+      "loss": 0.181,
+      "step": 3808
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.2102107328213934,
+      "learning_rate": 9.157567008774335e-06,
+      "loss": 0.1923,
+      "step": 3809
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.9835012736338915,
+      "learning_rate": 9.157066519936188e-06,
+      "loss": 0.2373,
+      "step": 3810
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.314340716992824,
+      "learning_rate": 9.15656589615727e-06,
+      "loss": 0.2573,
+      "step": 3811
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.2082311783008728,
+      "learning_rate": 9.156065137453833e-06,
+      "loss": 0.2576,
+      "step": 3812
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.9103381380385706,
+      "learning_rate": 9.155564243842133e-06,
+      "loss": 0.2003,
+      "step": 3813
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.1313264159084362,
+      "learning_rate": 9.155063215338429e-06,
+      "loss": 0.1981,
+      "step": 3814
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.5003633059174875,
+      "learning_rate": 9.154562051958981e-06,
+      "loss": 0.2676,
+      "step": 3815
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.4905788178557413,
+      "learning_rate": 9.15406075372006e-06,
+      "loss": 0.214,
+      "step": 3816
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.2943351086506938,
+      "learning_rate": 9.153559320637938e-06,
+      "loss": 0.1988,
+      "step": 3817
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.8426228230645898,
+      "learning_rate": 9.153057752728894e-06,
+      "loss": 0.1999,
+      "step": 3818
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.2886455555171028,
+      "learning_rate": 9.152556050009207e-06,
+      "loss": 0.2366,
+      "step": 3819
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.180806024439646,
+      "learning_rate": 9.152054212495161e-06,
+      "loss": 0.2274,
+      "step": 3820
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.043734518488835,
+      "learning_rate": 9.15155224020305e-06,
+      "loss": 0.2069,
+      "step": 3821
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 7.128763242358423,
+      "learning_rate": 9.151050133149162e-06,
+      "loss": 0.2397,
+      "step": 3822
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.3750913329568635,
+      "learning_rate": 9.150547891349804e-06,
+      "loss": 0.2765,
+      "step": 3823
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.0602465146290134,
+      "learning_rate": 9.150045514821272e-06,
+      "loss": 0.2175,
+      "step": 3824
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.1349249545728006,
+      "learning_rate": 9.149543003579877e-06,
+      "loss": 0.223,
+      "step": 3825
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.7718329366523977,
+      "learning_rate": 9.14904035764193e-06,
+      "loss": 0.2031,
+      "step": 3826
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.9916971827401129,
+      "learning_rate": 9.148537577023747e-06,
+      "loss": 0.1744,
+      "step": 3827
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.34043917201668,
+      "learning_rate": 9.148034661741648e-06,
+      "loss": 0.2349,
+      "step": 3828
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.4067348280244951,
+      "learning_rate": 9.147531611811959e-06,
+      "loss": 0.2164,
+      "step": 3829
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.1231727148124824,
+      "learning_rate": 9.14702842725101e-06,
+      "loss": 0.2264,
+      "step": 3830
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.4964938462927768,
+      "learning_rate": 9.146525108075132e-06,
+      "loss": 0.2079,
+      "step": 3831
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.302180857502982,
+      "learning_rate": 9.146021654300666e-06,
+      "loss": 0.2211,
+      "step": 3832
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.9847949169301132,
+      "learning_rate": 9.145518065943953e-06,
+      "loss": 0.2334,
+      "step": 3833
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.239306584954857,
+      "learning_rate": 9.145014343021338e-06,
+      "loss": 0.1834,
+      "step": 3834
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.1069728034991757,
+      "learning_rate": 9.144510485549175e-06,
+      "loss": 0.1988,
+      "step": 3835
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.2667994818835588,
+      "learning_rate": 9.144006493543817e-06,
+      "loss": 0.2452,
+      "step": 3836
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.227053240383855,
+      "learning_rate": 9.143502367021628e-06,
+      "loss": 0.2074,
+      "step": 3837
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.2972272601337251,
+      "learning_rate": 9.142998105998968e-06,
+      "loss": 0.1541,
+      "step": 3838
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.5019705812310395,
+      "learning_rate": 9.142493710492207e-06,
+      "loss": 0.1883,
+      "step": 3839
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.6933356242730695,
+      "learning_rate": 9.141989180517717e-06,
+      "loss": 0.1985,
+      "step": 3840
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.2692970500013088,
+      "learning_rate": 9.14148451609188e-06,
+      "loss": 0.254,
+      "step": 3841
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.1803248622611495,
+      "learning_rate": 9.140979717231069e-06,
+      "loss": 0.2153,
+      "step": 3842
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.7940046187393797,
+      "learning_rate": 9.140474783951678e-06,
+      "loss": 0.1723,
+      "step": 3843
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.2965654670201443,
+      "learning_rate": 9.139969716270094e-06,
+      "loss": 0.2519,
+      "step": 3844
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.4935629058128899,
+      "learning_rate": 9.139464514202714e-06,
+      "loss": 0.187,
+      "step": 3845
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.1146509779103568,
+      "learning_rate": 9.138959177765933e-06,
+      "loss": 0.1664,
+      "step": 3846
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 2.4666159360344344,
+      "learning_rate": 9.138453706976158e-06,
+      "loss": 0.2621,
+      "step": 3847
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.4773449716709195,
+      "learning_rate": 9.137948101849797e-06,
+      "loss": 0.19,
+      "step": 3848
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.203809626196583,
+      "learning_rate": 9.13744236240326e-06,
+      "loss": 0.1984,
+      "step": 3849
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.3753117125380414,
+      "learning_rate": 9.136936488652964e-06,
+      "loss": 0.1959,
+      "step": 3850
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.5454606227258028,
+      "learning_rate": 9.136430480615332e-06,
+      "loss": 0.2638,
+      "step": 3851
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.160895928200748,
+      "learning_rate": 9.135924338306788e-06,
+      "loss": 0.2351,
+      "step": 3852
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.4412547952017207,
+      "learning_rate": 9.13541806174376e-06,
+      "loss": 0.2133,
+      "step": 3853
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.120684617183233,
+      "learning_rate": 9.134911650942687e-06,
+      "loss": 0.2536,
+      "step": 3854
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.0543571665698088,
+      "learning_rate": 9.134405105920002e-06,
+      "loss": 0.1478,
+      "step": 3855
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.4128580852154318,
+      "learning_rate": 9.133898426692148e-06,
+      "loss": 0.235,
+      "step": 3856
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.6009064772938082,
+      "learning_rate": 9.133391613275576e-06,
+      "loss": 0.187,
+      "step": 3857
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 2.0569291741079585,
+      "learning_rate": 9.132884665686734e-06,
+      "loss": 0.2767,
+      "step": 3858
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.5795714340434275,
+      "learning_rate": 9.13237758394208e-06,
+      "loss": 0.2291,
+      "step": 3859
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.1732972780683086,
+      "learning_rate": 9.131870368058075e-06,
+      "loss": 0.205,
+      "step": 3860
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.6619823824178854,
+      "learning_rate": 9.131363018051181e-06,
+      "loss": 0.1726,
+      "step": 3861
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.3063835680926739,
+      "learning_rate": 9.130855533937865e-06,
+      "loss": 0.1852,
+      "step": 3862
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.156940062585222,
+      "learning_rate": 9.130347915734607e-06,
+      "loss": 0.2251,
+      "step": 3863
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.7887042324041043,
+      "learning_rate": 9.129840163457877e-06,
+      "loss": 0.1802,
+      "step": 3864
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.5342247729897587,
+      "learning_rate": 9.129332277124164e-06,
+      "loss": 0.2164,
+      "step": 3865
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.0835983052807738,
+      "learning_rate": 9.128824256749947e-06,
+      "loss": 0.2224,
+      "step": 3866
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.4972097601894867,
+      "learning_rate": 9.128316102351723e-06,
+      "loss": 0.2552,
+      "step": 3867
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.5763159287812705,
+      "learning_rate": 9.127807813945984e-06,
+      "loss": 0.2158,
+      "step": 3868
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.3012330527855602,
+      "learning_rate": 9.127299391549229e-06,
+      "loss": 0.1849,
+      "step": 3869
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.9834900958200572,
+      "learning_rate": 9.126790835177962e-06,
+      "loss": 0.1747,
+      "step": 3870
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.213180188000205,
+      "learning_rate": 9.126282144848692e-06,
+      "loss": 0.2286,
+      "step": 3871
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.1763901116248014,
+      "learning_rate": 9.125773320577931e-06,
+      "loss": 0.1954,
+      "step": 3872
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.15005030074977,
+      "learning_rate": 9.125264362382196e-06,
+      "loss": 0.1623,
+      "step": 3873
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.9048848264528946,
+      "learning_rate": 9.124755270278008e-06,
+      "loss": 0.2072,
+      "step": 3874
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.6880578111447992,
+      "learning_rate": 9.12424604428189e-06,
+      "loss": 0.1954,
+      "step": 3875
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.3867839556125503,
+      "learning_rate": 9.123736684410374e-06,
+      "loss": 0.225,
+      "step": 3876
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.9340833373790968,
+      "learning_rate": 9.123227190679995e-06,
+      "loss": 0.178,
+      "step": 3877
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.4046946401276694,
+      "learning_rate": 9.122717563107291e-06,
+      "loss": 0.1708,
+      "step": 3878
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.195919680496284,
+      "learning_rate": 9.122207801708802e-06,
+      "loss": 0.2092,
+      "step": 3879
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.0281902717354772,
+      "learning_rate": 9.121697906501079e-06,
+      "loss": 0.2009,
+      "step": 3880
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.341842110605686,
+      "learning_rate": 9.121187877500671e-06,
+      "loss": 0.2696,
+      "step": 3881
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.2371745391115256,
+      "learning_rate": 9.120677714724135e-06,
+      "loss": 0.1915,
+      "step": 3882
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.3073049604137659,
+      "learning_rate": 9.120167418188031e-06,
+      "loss": 0.2679,
+      "step": 3883
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.3797233483327376,
+      "learning_rate": 9.119656987908924e-06,
+      "loss": 0.2022,
+      "step": 3884
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.49447053246563,
+      "learning_rate": 9.11914642390338e-06,
+      "loss": 0.2067,
+      "step": 3885
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 2.0387706328470045,
+      "learning_rate": 9.118635726187977e-06,
+      "loss": 0.2082,
+      "step": 3886
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.0147021043552964,
+      "learning_rate": 9.118124894779287e-06,
+      "loss": 0.226,
+      "step": 3887
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.4926044489494235,
+      "learning_rate": 9.117613929693898e-06,
+      "loss": 0.1863,
+      "step": 3888
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.949110592068576,
+      "learning_rate": 9.117102830948393e-06,
+      "loss": 0.2248,
+      "step": 3889
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.1131462638566672,
+      "learning_rate": 9.11659159855936e-06,
+      "loss": 0.2143,
+      "step": 3890
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.9772688428347178,
+      "learning_rate": 9.116080232543398e-06,
+      "loss": 0.1859,
+      "step": 3891
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 7.173633409740171,
+      "learning_rate": 9.115568732917105e-06,
+      "loss": 0.2196,
+      "step": 3892
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 2.4299116123278197,
+      "learning_rate": 9.115057099697082e-06,
+      "loss": 0.2411,
+      "step": 3893
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.6166281861669305,
+      "learning_rate": 9.11454533289994e-06,
+      "loss": 0.2378,
+      "step": 3894
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.4130517919879522,
+      "learning_rate": 9.114033432542294e-06,
+      "loss": 0.244,
+      "step": 3895
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.6603315292812778,
+      "learning_rate": 9.113521398640753e-06,
+      "loss": 0.1979,
+      "step": 3896
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.8412934441500879,
+      "learning_rate": 9.113009231211943e-06,
+      "loss": 0.2384,
+      "step": 3897
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.215274644679239,
+      "learning_rate": 9.11249693027249e-06,
+      "loss": 0.1815,
+      "step": 3898
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.6592485212627244,
+      "learning_rate": 9.11198449583902e-06,
+      "loss": 0.2244,
+      "step": 3899
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.3533154982322786,
+      "learning_rate": 9.111471927928168e-06,
+      "loss": 0.2492,
+      "step": 3900
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.3079459565500346,
+      "learning_rate": 9.110959226556574e-06,
+      "loss": 0.2031,
+      "step": 3901
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.3008505141229256,
+      "learning_rate": 9.11044639174088e-06,
+      "loss": 0.2202,
+      "step": 3902
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.5337161232487766,
+      "learning_rate": 9.109933423497731e-06,
+      "loss": 0.2229,
+      "step": 3903
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.4729507168329292,
+      "learning_rate": 9.109420321843783e-06,
+      "loss": 0.2101,
+      "step": 3904
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.2524629646810497,
+      "learning_rate": 9.108907086795686e-06,
+      "loss": 0.2543,
+      "step": 3905
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.6337011716086638,
+      "learning_rate": 9.108393718370101e-06,
+      "loss": 0.2236,
+      "step": 3906
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.0276868441286218,
+      "learning_rate": 9.107880216583695e-06,
+      "loss": 0.1975,
+      "step": 3907
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.4957100311402798,
+      "learning_rate": 9.107366581453135e-06,
+      "loss": 0.2342,
+      "step": 3908
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 2.19289528479406,
+      "learning_rate": 9.106852812995092e-06,
+      "loss": 0.2246,
+      "step": 3909
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.4742410363967449,
+      "learning_rate": 9.106338911226247e-06,
+      "loss": 0.2328,
+      "step": 3910
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.4660578490147425,
+      "learning_rate": 9.10582487616328e-06,
+      "loss": 0.2023,
+      "step": 3911
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.1635108110869625,
+      "learning_rate": 9.105310707822875e-06,
+      "loss": 0.1976,
+      "step": 3912
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.5604971470745634,
+      "learning_rate": 9.104796406221725e-06,
+      "loss": 0.2105,
+      "step": 3913
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.286270861215534,
+      "learning_rate": 9.104281971376523e-06,
+      "loss": 0.2573,
+      "step": 3914
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.60761651938328,
+      "learning_rate": 9.103767403303968e-06,
+      "loss": 0.2228,
+      "step": 3915
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.6213528345651746,
+      "learning_rate": 9.103252702020761e-06,
+      "loss": 0.2972,
+      "step": 3916
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.4989476660052201,
+      "learning_rate": 9.102737867543614e-06,
+      "loss": 0.2388,
+      "step": 3917
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.4452350080181369,
+      "learning_rate": 9.102222899889237e-06,
+      "loss": 0.2967,
+      "step": 3918
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.7792288719469874,
+      "learning_rate": 9.101707799074344e-06,
+      "loss": 0.1498,
+      "step": 3919
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 2.967254663624282,
+      "learning_rate": 9.10119256511566e-06,
+      "loss": 0.3116,
+      "step": 3920
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.095126828361625,
+      "learning_rate": 9.100677198029904e-06,
+      "loss": 0.2287,
+      "step": 3921
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 3.5324259462578826,
+      "learning_rate": 9.100161697833809e-06,
+      "loss": 0.233,
+      "step": 3922
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.9081115286793394,
+      "learning_rate": 9.099646064544108e-06,
+      "loss": 0.2135,
+      "step": 3923
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.10704588310017,
+      "learning_rate": 9.099130298177538e-06,
+      "loss": 0.2398,
+      "step": 3924
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 31.445345041957154,
+      "learning_rate": 9.098614398750841e-06,
+      "loss": 0.3902,
+      "step": 3925
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 13.834206781426637,
+      "learning_rate": 9.098098366280765e-06,
+      "loss": 0.3085,
+      "step": 3926
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 2.5556974135523283,
+      "learning_rate": 9.09758220078406e-06,
+      "loss": 0.3209,
+      "step": 3927
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.888401552597067,
+      "learning_rate": 9.097065902277477e-06,
+      "loss": 0.2477,
+      "step": 3928
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.7730864795027543,
+      "learning_rate": 9.09654947077778e-06,
+      "loss": 0.2076,
+      "step": 3929
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.088822457117521,
+      "learning_rate": 9.096032906301733e-06,
+      "loss": 0.2098,
+      "step": 3930
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.6199266023845145,
+      "learning_rate": 9.095516208866102e-06,
+      "loss": 0.1955,
+      "step": 3931
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.3974632147204642,
+      "learning_rate": 9.094999378487659e-06,
+      "loss": 0.2201,
+      "step": 3932
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.635023153750645,
+      "learning_rate": 9.094482415183182e-06,
+      "loss": 0.2522,
+      "step": 3933
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.3414786279288757,
+      "learning_rate": 9.093965318969452e-06,
+      "loss": 0.2295,
+      "step": 3934
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.9510756273550055,
+      "learning_rate": 9.093448089863253e-06,
+      "loss": 0.1731,
+      "step": 3935
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.112925790624138,
+      "learning_rate": 9.092930727881374e-06,
+      "loss": 0.2333,
+      "step": 3936
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.3946101115605556,
+      "learning_rate": 9.09241323304061e-06,
+      "loss": 0.2042,
+      "step": 3937
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.5091041803755698,
+      "learning_rate": 9.09189560535776e-06,
+      "loss": 0.2126,
+      "step": 3938
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.4297062502606,
+      "learning_rate": 9.091377844849628e-06,
+      "loss": 0.2455,
+      "step": 3939
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.3619828257563242,
+      "learning_rate": 9.090859951533016e-06,
+      "loss": 0.2406,
+      "step": 3940
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.0539155448548942,
+      "learning_rate": 9.090341925424737e-06,
+      "loss": 0.2091,
+      "step": 3941
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.399338477276983,
+      "learning_rate": 9.089823766541607e-06,
+      "loss": 0.2547,
+      "step": 3942
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.1333213847011985,
+      "learning_rate": 9.089305474900447e-06,
+      "loss": 0.1903,
+      "step": 3943
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.2841633523692602,
+      "learning_rate": 9.08878705051808e-06,
+      "loss": 0.1964,
+      "step": 3944
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.0435783769080949,
+      "learning_rate": 9.088268493411333e-06,
+      "loss": 0.2413,
+      "step": 3945
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.3892911289467156,
+      "learning_rate": 9.08774980359704e-06,
+      "loss": 0.2217,
+      "step": 3946
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.1039540356756785,
+      "learning_rate": 9.087230981092039e-06,
+      "loss": 0.19,
+      "step": 3947
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.0485545215308834,
+      "learning_rate": 9.086712025913167e-06,
+      "loss": 0.221,
+      "step": 3948
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.316273057125249,
+      "learning_rate": 9.086192938077276e-06,
+      "loss": 0.181,
+      "step": 3949
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.9269335171200681,
+      "learning_rate": 9.085673717601212e-06,
+      "loss": 0.183,
+      "step": 3950
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.1715524642862496,
+      "learning_rate": 9.085154364501826e-06,
+      "loss": 0.2008,
+      "step": 3951
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.3168969117207805,
+      "learning_rate": 9.084634878795984e-06,
+      "loss": 0.174,
+      "step": 3952
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.3312954632401595,
+      "learning_rate": 9.084115260500544e-06,
+      "loss": 0.2369,
+      "step": 3953
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.2462608978671195,
+      "learning_rate": 9.083595509632374e-06,
+      "loss": 0.2397,
+      "step": 3954
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.5062877575110973,
+      "learning_rate": 9.083075626208345e-06,
+      "loss": 0.236,
+      "step": 3955
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.0334684719881109,
+      "learning_rate": 9.082555610245334e-06,
+      "loss": 0.1484,
+      "step": 3956
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.2703822966025162,
+      "learning_rate": 9.08203546176022e-06,
+      "loss": 0.2341,
+      "step": 3957
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.0816951773771033,
+      "learning_rate": 9.081515180769888e-06,
+      "loss": 0.1464,
+      "step": 3958
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.2583359975035309,
+      "learning_rate": 9.080994767291226e-06,
+      "loss": 0.2399,
+      "step": 3959
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.255369093665982,
+      "learning_rate": 9.080474221341127e-06,
+      "loss": 0.2328,
+      "step": 3960
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.1222297440650777,
+      "learning_rate": 9.079953542936489e-06,
+      "loss": 0.2546,
+      "step": 3961
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.0228227669211951,
+      "learning_rate": 9.07943273209421e-06,
+      "loss": 0.2055,
+      "step": 3962
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.4378145601941417,
+      "learning_rate": 9.078911788831201e-06,
+      "loss": 0.2131,
+      "step": 3963
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.4494783061092709,
+      "learning_rate": 9.07839071316437e-06,
+      "loss": 0.2325,
+      "step": 3964
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.1118147342821767,
+      "learning_rate": 9.07786950511063e-06,
+      "loss": 0.2019,
+      "step": 3965
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.211803569567394,
+      "learning_rate": 9.077348164686903e-06,
+      "loss": 0.2035,
+      "step": 3966
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.8593468011217689,
+      "learning_rate": 9.076826691910109e-06,
+      "loss": 0.2046,
+      "step": 3967
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.2829711419862455,
+      "learning_rate": 9.076305086797174e-06,
+      "loss": 0.1913,
+      "step": 3968
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.147250431312169,
+      "learning_rate": 9.075783349365035e-06,
+      "loss": 0.1937,
+      "step": 3969
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.4233976597764066,
+      "learning_rate": 9.075261479630624e-06,
+      "loss": 0.2026,
+      "step": 3970
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.9318792509376028,
+      "learning_rate": 9.07473947761088e-06,
+      "loss": 0.1822,
+      "step": 3971
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.5620405856890769,
+      "learning_rate": 9.07421734332275e-06,
+      "loss": 0.2295,
+      "step": 3972
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.4736730419656456,
+      "learning_rate": 9.073695076783185e-06,
+      "loss": 0.2275,
+      "step": 3973
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.0827056826230326,
+      "learning_rate": 9.073172678009132e-06,
+      "loss": 0.2495,
+      "step": 3974
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.7198120217261466,
+      "learning_rate": 9.072650147017554e-06,
+      "loss": 0.2412,
+      "step": 3975
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.3398646528209446,
+      "learning_rate": 9.072127483825409e-06,
+      "loss": 0.1765,
+      "step": 3976
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.8074823322278791,
+      "learning_rate": 9.071604688449664e-06,
+      "loss": 0.1842,
+      "step": 3977
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.2411501769794118,
+      "learning_rate": 9.07108176090729e-06,
+      "loss": 0.2252,
+      "step": 3978
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.1978711966163913,
+      "learning_rate": 9.07055870121526e-06,
+      "loss": 0.2004,
+      "step": 3979
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 238.53069032783074,
+      "learning_rate": 9.070035509390556e-06,
+      "loss": 0.2214,
+      "step": 3980
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.252281029975538,
+      "learning_rate": 9.069512185450157e-06,
+      "loss": 0.1727,
+      "step": 3981
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.0767822256975874,
+      "learning_rate": 9.068988729411052e-06,
+      "loss": 0.2126,
+      "step": 3982
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.9993137019976452,
+      "learning_rate": 9.068465141290234e-06,
+      "loss": 0.2242,
+      "step": 3983
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.4743079237356895,
+      "learning_rate": 9.067941421104699e-06,
+      "loss": 0.1906,
+      "step": 3984
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.037514302215466,
+      "learning_rate": 9.067417568871444e-06,
+      "loss": 0.2026,
+      "step": 3985
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.9150868813509297,
+      "learning_rate": 9.066893584607478e-06,
+      "loss": 0.2026,
+      "step": 3986
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.6701308961746648,
+      "learning_rate": 9.066369468329806e-06,
+      "loss": 0.2506,
+      "step": 3987
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.353850021999717,
+      "learning_rate": 9.065845220055443e-06,
+      "loss": 0.2552,
+      "step": 3988
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.137571642461414,
+      "learning_rate": 9.065320839801408e-06,
+      "loss": 0.2377,
+      "step": 3989
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.0285701466331931,
+      "learning_rate": 9.064796327584718e-06,
+      "loss": 0.2095,
+      "step": 3990
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.9376985846241689,
+      "learning_rate": 9.064271683422404e-06,
+      "loss": 0.2181,
+      "step": 3991
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.1768676374162577,
+      "learning_rate": 9.063746907331492e-06,
+      "loss": 0.1993,
+      "step": 3992
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.1911501812518448,
+      "learning_rate": 9.06322199932902e-06,
+      "loss": 0.152,
+      "step": 3993
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.9076609336435363,
+      "learning_rate": 9.062696959432025e-06,
+      "loss": 0.1924,
+      "step": 3994
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.1306405456912385,
+      "learning_rate": 9.06217178765755e-06,
+      "loss": 0.2243,
+      "step": 3995
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.9460483674017397,
+      "learning_rate": 9.061646484022643e-06,
+      "loss": 0.2063,
+      "step": 3996
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.0804868031867108,
+      "learning_rate": 9.061121048544357e-06,
+      "loss": 0.1995,
+      "step": 3997
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.9403434735686592,
+      "learning_rate": 9.060595481239745e-06,
+      "loss": 0.1998,
+      "step": 3998
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.2646486046691547,
+      "learning_rate": 9.060069782125868e-06,
+      "loss": 0.2235,
+      "step": 3999
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.085301223117447,
+      "learning_rate": 9.059543951219793e-06,
+      "loss": 0.1665,
+      "step": 4000
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.3763910431443511,
+      "learning_rate": 9.059017988538586e-06,
+      "loss": 0.1971,
+      "step": 4001
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.2290449048472203,
+      "learning_rate": 9.058491894099322e-06,
+      "loss": 0.2356,
+      "step": 4002
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.9379777128308331,
+      "learning_rate": 9.057965667919075e-06,
+      "loss": 0.2031,
+      "step": 4003
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.316092739636983,
+      "learning_rate": 9.05743931001493e-06,
+      "loss": 0.2569,
+      "step": 4004
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.2828464331651683,
+      "learning_rate": 9.056912820403974e-06,
+      "loss": 0.1881,
+      "step": 4005
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.368104591683577,
+      "learning_rate": 9.056386199103292e-06,
+      "loss": 0.1889,
+      "step": 4006
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 3.354530780686837,
+      "learning_rate": 9.055859446129982e-06,
+      "loss": 0.2695,
+      "step": 4007
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.5751665991325952,
+      "learning_rate": 9.055332561501143e-06,
+      "loss": 0.2245,
+      "step": 4008
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.0019568244315518,
+      "learning_rate": 9.054805545233875e-06,
+      "loss": 0.2386,
+      "step": 4009
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.1656136482800763,
+      "learning_rate": 9.05427839734529e-06,
+      "loss": 0.2439,
+      "step": 4010
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.4410217620065393,
+      "learning_rate": 9.053751117852496e-06,
+      "loss": 0.2726,
+      "step": 4011
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.5906254863925633,
+      "learning_rate": 9.053223706772609e-06,
+      "loss": 0.2797,
+      "step": 4012
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.2443335861090432,
+      "learning_rate": 9.05269616412275e-06,
+      "loss": 0.2294,
+      "step": 4013
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.9039719380931001,
+      "learning_rate": 9.052168489920044e-06,
+      "loss": 0.2194,
+      "step": 4014
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.9787745770813103,
+      "learning_rate": 9.051640684181618e-06,
+      "loss": 0.1841,
+      "step": 4015
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 2.068733102320322,
+      "learning_rate": 9.051112746924603e-06,
+      "loss": 0.1926,
+      "step": 4016
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.3960377118904232,
+      "learning_rate": 9.050584678166144e-06,
+      "loss": 0.2152,
+      "step": 4017
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.2298826827687848,
+      "learning_rate": 9.050056477923373e-06,
+      "loss": 0.22,
+      "step": 4018
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.0030527048708848,
+      "learning_rate": 9.04952814621344e-06,
+      "loss": 0.1822,
+      "step": 4019
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.9490897950539084,
+      "learning_rate": 9.048999683053496e-06,
+      "loss": 0.1329,
+      "step": 4020
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.959354921450795,
+      "learning_rate": 9.048471088460694e-06,
+      "loss": 0.2142,
+      "step": 4021
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.096798609249837,
+      "learning_rate": 9.047942362452191e-06,
+      "loss": 0.1951,
+      "step": 4022
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.9275725062854036,
+      "learning_rate": 9.047413505045153e-06,
+      "loss": 0.1793,
+      "step": 4023
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.3577229934257922,
+      "learning_rate": 9.046884516256745e-06,
+      "loss": 0.2137,
+      "step": 4024
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.4634093850073568,
+      "learning_rate": 9.046355396104137e-06,
+      "loss": 0.1715,
+      "step": 4025
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.6197086529654114,
+      "learning_rate": 9.045826144604508e-06,
+      "loss": 0.2356,
+      "step": 4026
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 2.8240591507976562,
+      "learning_rate": 9.045296761775034e-06,
+      "loss": 0.2796,
+      "step": 4027
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.342602289563393,
+      "learning_rate": 9.044767247632901e-06,
+      "loss": 0.2138,
+      "step": 4028
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.2538284800527373,
+      "learning_rate": 9.044237602195297e-06,
+      "loss": 0.2321,
+      "step": 4029
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.1995229040250317,
+      "learning_rate": 9.043707825479418e-06,
+      "loss": 0.1427,
+      "step": 4030
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.904770313188538,
+      "learning_rate": 9.043177917502455e-06,
+      "loss": 0.2453,
+      "step": 4031
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.4094908576275895,
+      "learning_rate": 9.042647878281611e-06,
+      "loss": 0.2012,
+      "step": 4032
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 2.7939199788486038,
+      "learning_rate": 9.042117707834095e-06,
+      "loss": 0.2479,
+      "step": 4033
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.3255575327361306,
+      "learning_rate": 9.04158740617711e-06,
+      "loss": 0.2227,
+      "step": 4034
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.1095749052020163,
+      "learning_rate": 9.041056973327877e-06,
+      "loss": 0.2297,
+      "step": 4035
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.2424678725995542,
+      "learning_rate": 9.04052640930361e-06,
+      "loss": 0.2376,
+      "step": 4036
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.2894560935498505,
+      "learning_rate": 9.039995714121532e-06,
+      "loss": 0.1634,
+      "step": 4037
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.1512980323063886,
+      "learning_rate": 9.03946488779887e-06,
+      "loss": 0.2116,
+      "step": 4038
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.1883593762438958,
+      "learning_rate": 9.038933930352855e-06,
+      "loss": 0.2554,
+      "step": 4039
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.1501325295916955,
+      "learning_rate": 9.038402841800722e-06,
+      "loss": 0.1882,
+      "step": 4040
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.9915651158428462,
+      "learning_rate": 9.03787162215971e-06,
+      "loss": 0.1999,
+      "step": 4041
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.8379947513178955,
+      "learning_rate": 9.037340271447063e-06,
+      "loss": 0.1697,
+      "step": 4042
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.0616980762588175,
+      "learning_rate": 9.036808789680031e-06,
+      "loss": 0.1761,
+      "step": 4043
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.0084199165363403,
+      "learning_rate": 9.036277176875862e-06,
+      "loss": 0.2116,
+      "step": 4044
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.3270723097905792,
+      "learning_rate": 9.035745433051817e-06,
+      "loss": 0.2005,
+      "step": 4045
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.529804669160465,
+      "learning_rate": 9.035213558225153e-06,
+      "loss": 0.2417,
+      "step": 4046
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.2332791008683466,
+      "learning_rate": 9.034681552413138e-06,
+      "loss": 0.2329,
+      "step": 4047
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.2983340334609832,
+      "learning_rate": 9.034149415633038e-06,
+      "loss": 0.1664,
+      "step": 4048
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.567426273301939,
+      "learning_rate": 9.03361714790213e-06,
+      "loss": 0.23,
+      "step": 4049
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.5984972043973504,
+      "learning_rate": 9.033084749237689e-06,
+      "loss": 0.2252,
+      "step": 4050
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.4911776216054171,
+      "learning_rate": 9.032552219656996e-06,
+      "loss": 0.2485,
+      "step": 4051
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.261329573142838,
+      "learning_rate": 9.032019559177341e-06,
+      "loss": 0.2142,
+      "step": 4052
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.5320400467327793,
+      "learning_rate": 9.031486767816012e-06,
+      "loss": 0.2061,
+      "step": 4053
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.2569776127658827,
+      "learning_rate": 9.030953845590305e-06,
+      "loss": 0.2356,
+      "step": 4054
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.2980693156552328,
+      "learning_rate": 9.030420792517518e-06,
+      "loss": 0.2379,
+      "step": 4055
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.0473619936777305,
+      "learning_rate": 9.029887608614954e-06,
+      "loss": 0.2213,
+      "step": 4056
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.3816580290556972,
+      "learning_rate": 9.02935429389992e-06,
+      "loss": 0.2217,
+      "step": 4057
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.9067069682384277,
+      "learning_rate": 9.02882084838973e-06,
+      "loss": 0.203,
+      "step": 4058
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.3370329258578257,
+      "learning_rate": 9.028287272101698e-06,
+      "loss": 0.2555,
+      "step": 4059
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.309308798989649,
+      "learning_rate": 9.027753565053145e-06,
+      "loss": 0.1619,
+      "step": 4060
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.2836237643279862,
+      "learning_rate": 9.027219727261395e-06,
+      "loss": 0.1921,
+      "step": 4061
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.2151061630178122,
+      "learning_rate": 9.026685758743776e-06,
+      "loss": 0.2706,
+      "step": 4062
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.4948279950139127,
+      "learning_rate": 9.026151659517623e-06,
+      "loss": 0.19,
+      "step": 4063
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.3905586691638634,
+      "learning_rate": 9.025617429600271e-06,
+      "loss": 0.1808,
+      "step": 4064
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.2625290406423224,
+      "learning_rate": 9.025083069009065e-06,
+      "loss": 0.1903,
+      "step": 4065
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.443835029121117,
+      "learning_rate": 9.024548577761344e-06,
+      "loss": 0.2085,
+      "step": 4066
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.0186344181356362,
+      "learning_rate": 9.024013955874464e-06,
+      "loss": 0.2081,
+      "step": 4067
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.2966462531062162,
+      "learning_rate": 9.023479203365777e-06,
+      "loss": 0.2324,
+      "step": 4068
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.0285231706260616,
+      "learning_rate": 9.02294432025264e-06,
+      "loss": 0.1757,
+      "step": 4069
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.1970970422502454,
+      "learning_rate": 9.022409306552418e-06,
+      "loss": 0.1759,
+      "step": 4070
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.596210449282529,
+      "learning_rate": 9.021874162282478e-06,
+      "loss": 0.2375,
+      "step": 4071
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.3550213589383704,
+      "learning_rate": 9.021338887460188e-06,
+      "loss": 0.2446,
+      "step": 4072
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.0521225257063553,
+      "learning_rate": 9.020803482102927e-06,
+      "loss": 0.2137,
+      "step": 4073
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.2471333982900443,
+      "learning_rate": 9.020267946228071e-06,
+      "loss": 0.2423,
+      "step": 4074
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.764783346384473,
+      "learning_rate": 9.019732279853008e-06,
+      "loss": 0.2778,
+      "step": 4075
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.952401165800045,
+      "learning_rate": 9.019196482995121e-06,
+      "loss": 0.2307,
+      "step": 4076
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.1038222891133354,
+      "learning_rate": 9.018660555671807e-06,
+      "loss": 0.2241,
+      "step": 4077
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.1672978308184319,
+      "learning_rate": 9.01812449790046e-06,
+      "loss": 0.1982,
+      "step": 4078
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.0608079248272484,
+      "learning_rate": 9.01758830969848e-06,
+      "loss": 0.2489,
+      "step": 4079
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.1702571764145098,
+      "learning_rate": 9.017051991083275e-06,
+      "loss": 0.2343,
+      "step": 4080
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.0144135985424252,
+      "learning_rate": 9.01651554207225e-06,
+      "loss": 0.2017,
+      "step": 4081
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.4596505584719763,
+      "learning_rate": 9.015978962682822e-06,
+      "loss": 0.1828,
+      "step": 4082
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.1334788072913986,
+      "learning_rate": 9.015442252932408e-06,
+      "loss": 0.2065,
+      "step": 4083
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.3118204503397108,
+      "learning_rate": 9.01490541283843e-06,
+      "loss": 0.2215,
+      "step": 4084
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.0552256676965703,
+      "learning_rate": 9.014368442418312e-06,
+      "loss": 0.1818,
+      "step": 4085
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.5182361133693512,
+      "learning_rate": 9.013831341689489e-06,
+      "loss": 0.2239,
+      "step": 4086
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.3147849249416343,
+      "learning_rate": 9.01329411066939e-06,
+      "loss": 0.1965,
+      "step": 4087
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.9186870483642835,
+      "learning_rate": 9.012756749375457e-06,
+      "loss": 0.2024,
+      "step": 4088
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.1077784905136312,
+      "learning_rate": 9.012219257825132e-06,
+      "loss": 0.2416,
+      "step": 4089
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.1073809705308049,
+      "learning_rate": 9.011681636035863e-06,
+      "loss": 0.1659,
+      "step": 4090
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.9697116416559244,
+      "learning_rate": 9.0111438840251e-06,
+      "loss": 0.2164,
+      "step": 4091
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.1488155858491462,
+      "learning_rate": 9.010606001810302e-06,
+      "loss": 0.2031,
+      "step": 4092
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.1629567206643563,
+      "learning_rate": 9.010067989408927e-06,
+      "loss": 0.2144,
+      "step": 4093
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.7042877773205591,
+      "learning_rate": 9.009529846838436e-06,
+      "loss": 0.2397,
+      "step": 4094
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.2683522060627554,
+      "learning_rate": 9.008991574116304e-06,
+      "loss": 0.1837,
+      "step": 4095
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.336553456706433,
+      "learning_rate": 9.00845317126e-06,
+      "loss": 0.2015,
+      "step": 4096
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.4624962778125574,
+      "learning_rate": 9.007914638286997e-06,
+      "loss": 0.218,
+      "step": 4097
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.5479859723959837,
+      "learning_rate": 9.007375975214783e-06,
+      "loss": 0.226,
+      "step": 4098
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.3392935059525133,
+      "learning_rate": 9.006837182060841e-06,
+      "loss": 0.2485,
+      "step": 4099
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.4190646093197863,
+      "learning_rate": 9.006298258842659e-06,
+      "loss": 0.2275,
+      "step": 4100
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.9046259810559787,
+      "learning_rate": 9.005759205577733e-06,
+      "loss": 0.1654,
+      "step": 4101
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.1162922388206393,
+      "learning_rate": 9.005220022283558e-06,
+      "loss": 0.2149,
+      "step": 4102
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.1572720627350397,
+      "learning_rate": 9.004680708977638e-06,
+      "loss": 0.1854,
+      "step": 4103
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.350867728172819,
+      "learning_rate": 9.004141265677482e-06,
+      "loss": 0.2336,
+      "step": 4104
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.4539227221963542,
+      "learning_rate": 9.003601692400594e-06,
+      "loss": 0.2512,
+      "step": 4105
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.252587667926246,
+      "learning_rate": 9.003061989164495e-06,
+      "loss": 0.2125,
+      "step": 4106
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.434395964527467,
+      "learning_rate": 9.002522155986702e-06,
+      "loss": 0.2332,
+      "step": 4107
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.0697123853417179,
+      "learning_rate": 9.001982192884739e-06,
+      "loss": 0.2486,
+      "step": 4108
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.4272196881790924,
+      "learning_rate": 9.001442099876131e-06,
+      "loss": 0.2439,
+      "step": 4109
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.2229526399709112,
+      "learning_rate": 9.000901876978411e-06,
+      "loss": 0.2271,
+      "step": 4110
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.573706531727278,
+      "learning_rate": 9.000361524209117e-06,
+      "loss": 0.2714,
+      "step": 4111
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.8528543415862091,
+      "learning_rate": 8.999821041585788e-06,
+      "loss": 0.2027,
+      "step": 4112
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.0560111640609617,
+      "learning_rate": 8.999280429125967e-06,
+      "loss": 0.1748,
+      "step": 4113
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.248186856913118,
+      "learning_rate": 8.998739686847202e-06,
+      "loss": 0.1928,
+      "step": 4114
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.9885110550225112,
+      "learning_rate": 8.998198814767049e-06,
+      "loss": 0.1957,
+      "step": 4115
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.9475782104792192,
+      "learning_rate": 8.997657812903063e-06,
+      "loss": 0.1657,
+      "step": 4116
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.4894584608145,
+      "learning_rate": 8.997116681272807e-06,
+      "loss": 0.2246,
+      "step": 4117
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.9648207555073778,
+      "learning_rate": 8.996575419893842e-06,
+      "loss": 0.2784,
+      "step": 4118
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.3267563682154073,
+      "learning_rate": 8.996034028783743e-06,
+      "loss": 0.2307,
+      "step": 4119
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.0074536212232321,
+      "learning_rate": 8.995492507960082e-06,
+      "loss": 0.1779,
+      "step": 4120
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.053205317538902,
+      "learning_rate": 8.994950857440434e-06,
+      "loss": 0.1853,
+      "step": 4121
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.7122956447464797,
+      "learning_rate": 8.994409077242387e-06,
+      "loss": 0.2956,
+      "step": 4122
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.9185005808376376,
+      "learning_rate": 8.993867167383523e-06,
+      "loss": 0.1791,
+      "step": 4123
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.2339909512899174,
+      "learning_rate": 8.993325127881434e-06,
+      "loss": 0.2092,
+      "step": 4124
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.2676403231709357,
+      "learning_rate": 8.992782958753714e-06,
+      "loss": 0.2509,
+      "step": 4125
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.0264891153680922,
+      "learning_rate": 8.992240660017965e-06,
+      "loss": 0.1612,
+      "step": 4126
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.0360654760514612,
+      "learning_rate": 8.991698231691789e-06,
+      "loss": 0.1895,
+      "step": 4127
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 2.8416239997218646,
+      "learning_rate": 8.99115567379279e-06,
+      "loss": 0.1948,
+      "step": 4128
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.2990466959328977,
+      "learning_rate": 8.990612986338584e-06,
+      "loss": 0.191,
+      "step": 4129
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.1759743492322317,
+      "learning_rate": 8.990070169346786e-06,
+      "loss": 0.1924,
+      "step": 4130
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.1578650873805505,
+      "learning_rate": 8.989527222835017e-06,
+      "loss": 0.1809,
+      "step": 4131
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.9212946216544114,
+      "learning_rate": 8.988984146820897e-06,
+      "loss": 0.1933,
+      "step": 4132
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.4891294930629726,
+      "learning_rate": 8.988440941322061e-06,
+      "loss": 0.2362,
+      "step": 4133
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.132407064067968,
+      "learning_rate": 8.987897606356137e-06,
+      "loss": 0.1832,
+      "step": 4134
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.3726216739040702,
+      "learning_rate": 8.987354141940766e-06,
+      "loss": 0.1673,
+      "step": 4135
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.895434469133622,
+      "learning_rate": 8.986810548093583e-06,
+      "loss": 0.169,
+      "step": 4136
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.1027061177037536,
+      "learning_rate": 8.98626682483224e-06,
+      "loss": 0.211,
+      "step": 4137
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.9335209067908207,
+      "learning_rate": 8.98572297217438e-06,
+      "loss": 0.1786,
+      "step": 4138
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.1771923038657055,
+      "learning_rate": 8.985178990137664e-06,
+      "loss": 0.2116,
+      "step": 4139
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.3540906634154162,
+      "learning_rate": 8.984634878739744e-06,
+      "loss": 0.2247,
+      "step": 4140
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.2671268553280426,
+      "learning_rate": 8.984090637998286e-06,
+      "loss": 0.186,
+      "step": 4141
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.9396938854497765,
+      "learning_rate": 8.983546267930953e-06,
+      "loss": 0.2315,
+      "step": 4142
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.2088680272976626,
+      "learning_rate": 8.983001768555419e-06,
+      "loss": 0.2285,
+      "step": 4143
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.2165481104907392,
+      "learning_rate": 8.982457139889358e-06,
+      "loss": 0.2444,
+      "step": 4144
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.1596862068749847,
+      "learning_rate": 8.981912381950447e-06,
+      "loss": 0.2032,
+      "step": 4145
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.4290835454622803,
+      "learning_rate": 8.98136749475637e-06,
+      "loss": 0.1874,
+      "step": 4146
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.4417729253802891,
+      "learning_rate": 8.980822478324815e-06,
+      "loss": 0.2592,
+      "step": 4147
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.982195532341427,
+      "learning_rate": 8.980277332673474e-06,
+      "loss": 0.1997,
+      "step": 4148
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.2319449712626314,
+      "learning_rate": 8.979732057820042e-06,
+      "loss": 0.216,
+      "step": 4149
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.1075698931906703,
+      "learning_rate": 8.979186653782216e-06,
+      "loss": 0.2359,
+      "step": 4150
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.8939238541327754,
+      "learning_rate": 8.978641120577706e-06,
+      "loss": 0.1821,
+      "step": 4151
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.4854715789121167,
+      "learning_rate": 8.978095458224217e-06,
+      "loss": 0.2027,
+      "step": 4152
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.2987225769294264,
+      "learning_rate": 8.977549666739461e-06,
+      "loss": 0.2396,
+      "step": 4153
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 3.794425421423598,
+      "learning_rate": 8.977003746141156e-06,
+      "loss": 0.2069,
+      "step": 4154
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.9294467441987192,
+      "learning_rate": 8.976457696447024e-06,
+      "loss": 0.1402,
+      "step": 4155
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.232865799209807,
+      "learning_rate": 8.975911517674785e-06,
+      "loss": 0.2475,
+      "step": 4156
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.1280423442512608,
+      "learning_rate": 8.975365209842175e-06,
+      "loss": 0.1699,
+      "step": 4157
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.0102348344273901,
+      "learning_rate": 8.974818772966923e-06,
+      "loss": 0.1759,
+      "step": 4158
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.1226994715197114,
+      "learning_rate": 8.974272207066767e-06,
+      "loss": 0.2074,
+      "step": 4159
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.0285743452131137,
+      "learning_rate": 8.973725512159451e-06,
+      "loss": 0.2076,
+      "step": 4160
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.351317502015487,
+      "learning_rate": 8.97317868826272e-06,
+      "loss": 0.2163,
+      "step": 4161
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.0168881035398925,
+      "learning_rate": 8.972631735394324e-06,
+      "loss": 0.2003,
+      "step": 4162
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.7263013514923218,
+      "learning_rate": 8.972084653572016e-06,
+      "loss": 0.2659,
+      "step": 4163
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.9891415788501141,
+      "learning_rate": 8.971537442813558e-06,
+      "loss": 0.1972,
+      "step": 4164
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.5826184807679964,
+      "learning_rate": 8.970990103136709e-06,
+      "loss": 0.2059,
+      "step": 4165
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 8.496267986045657,
+      "learning_rate": 8.97044263455924e-06,
+      "loss": 0.2401,
+      "step": 4166
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.512417912521622,
+      "learning_rate": 8.969895037098916e-06,
+      "loss": 0.2159,
+      "step": 4167
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.3140415819909463,
+      "learning_rate": 8.96934731077352e-06,
+      "loss": 0.2215,
+      "step": 4168
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.7172982673914612,
+      "learning_rate": 8.968799455600827e-06,
+      "loss": 0.246,
+      "step": 4169
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.3920239821327896,
+      "learning_rate": 8.968251471598622e-06,
+      "loss": 0.237,
+      "step": 4170
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.0389365652319726,
+      "learning_rate": 8.96770335878469e-06,
+      "loss": 0.1715,
+      "step": 4171
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.4822808683628241,
+      "learning_rate": 8.967155117176827e-06,
+      "loss": 0.2579,
+      "step": 4172
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.0259310342798171,
+      "learning_rate": 8.966606746792828e-06,
+      "loss": 0.1972,
+      "step": 4173
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.6714019306637364,
+      "learning_rate": 8.966058247650493e-06,
+      "loss": 0.2643,
+      "step": 4174
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.5957615210167895,
+      "learning_rate": 8.965509619767625e-06,
+      "loss": 0.204,
+      "step": 4175
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.8617816939671128,
+      "learning_rate": 8.964960863162035e-06,
+      "loss": 0.1891,
+      "step": 4176
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.468732465350204,
+      "learning_rate": 8.964411977851536e-06,
+      "loss": 0.2007,
+      "step": 4177
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.4207772901348332,
+      "learning_rate": 8.963862963853944e-06,
+      "loss": 0.2273,
+      "step": 4178
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 2.3858374733932806,
+      "learning_rate": 8.963313821187084e-06,
+      "loss": 0.2091,
+      "step": 4179
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.2507784550341914,
+      "learning_rate": 8.962764549868775e-06,
+      "loss": 0.2068,
+      "step": 4180
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.0452296899915356,
+      "learning_rate": 8.962215149916849e-06,
+      "loss": 0.1446,
+      "step": 4181
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.4766659649924592,
+      "learning_rate": 8.961665621349143e-06,
+      "loss": 0.2307,
+      "step": 4182
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.1241979554660875,
+      "learning_rate": 8.961115964183491e-06,
+      "loss": 0.1676,
+      "step": 4183
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.6169964728198982,
+      "learning_rate": 8.960566178437738e-06,
+      "loss": 0.2032,
+      "step": 4184
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.2494679451147417,
+      "learning_rate": 8.96001626412973e-06,
+      "loss": 0.1767,
+      "step": 4185
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.4151400364876454,
+      "learning_rate": 8.959466221277315e-06,
+      "loss": 0.1842,
+      "step": 4186
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.5186245653811778,
+      "learning_rate": 8.958916049898352e-06,
+      "loss": 0.2175,
+      "step": 4187
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.2565147268059946,
+      "learning_rate": 8.958365750010696e-06,
+      "loss": 0.1768,
+      "step": 4188
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.7227010980594,
+      "learning_rate": 8.957815321632212e-06,
+      "loss": 0.215,
+      "step": 4189
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.2495719683463062,
+      "learning_rate": 8.957264764780767e-06,
+      "loss": 0.22,
+      "step": 4190
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.4139786057850101,
+      "learning_rate": 8.956714079474232e-06,
+      "loss": 0.2241,
+      "step": 4191
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.4723317550215913,
+      "learning_rate": 8.956163265730481e-06,
+      "loss": 0.1646,
+      "step": 4192
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.623803739201739,
+      "learning_rate": 8.955612323567398e-06,
+      "loss": 0.2515,
+      "step": 4193
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 2.691763932997076,
+      "learning_rate": 8.955061253002863e-06,
+      "loss": 0.2239,
+      "step": 4194
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.4619560694345073,
+      "learning_rate": 8.954510054054766e-06,
+      "loss": 0.224,
+      "step": 4195
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.9368251477631669,
+      "learning_rate": 8.953958726740999e-06,
+      "loss": 0.1569,
+      "step": 4196
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.253470422340444,
+      "learning_rate": 8.953407271079456e-06,
+      "loss": 0.2632,
+      "step": 4197
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.6165455221187577,
+      "learning_rate": 8.952855687088042e-06,
+      "loss": 0.2136,
+      "step": 4198
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.560952641105342,
+      "learning_rate": 8.952303974784657e-06,
+      "loss": 0.2234,
+      "step": 4199
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.1413909646067562,
+      "learning_rate": 8.951752134187214e-06,
+      "loss": 0.1784,
+      "step": 4200
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.7702786759716713,
+      "learning_rate": 8.951200165313625e-06,
+      "loss": 0.2305,
+      "step": 4201
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.5565619526586238,
+      "learning_rate": 8.950648068181804e-06,
+      "loss": 0.2025,
+      "step": 4202
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.3255835043987552,
+      "learning_rate": 8.950095842809675e-06,
+      "loss": 0.217,
+      "step": 4203
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.8566235451445895,
+      "learning_rate": 8.949543489215167e-06,
+      "loss": 0.1808,
+      "step": 4204
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.2421604999270972,
+      "learning_rate": 8.948991007416203e-06,
+      "loss": 0.1971,
+      "step": 4205
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.2544197118599456,
+      "learning_rate": 8.948438397430721e-06,
+      "loss": 0.2457,
+      "step": 4206
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.159125464182703,
+      "learning_rate": 8.947885659276658e-06,
+      "loss": 0.2298,
+      "step": 4207
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.3559700481379304,
+      "learning_rate": 8.947332792971956e-06,
+      "loss": 0.2556,
+      "step": 4208
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 3.944382903614446,
+      "learning_rate": 8.946779798534563e-06,
+      "loss": 0.2332,
+      "step": 4209
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.1838987385568296,
+      "learning_rate": 8.946226675982428e-06,
+      "loss": 0.2144,
+      "step": 4210
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.4360832081796822,
+      "learning_rate": 8.945673425333504e-06,
+      "loss": 0.2081,
+      "step": 4211
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.8151414190357701,
+      "learning_rate": 8.945120046605753e-06,
+      "loss": 0.1207,
+      "step": 4212
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.2377077174878104,
+      "learning_rate": 8.944566539817136e-06,
+      "loss": 0.224,
+      "step": 4213
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.0995804408520107,
+      "learning_rate": 8.944012904985622e-06,
+      "loss": 0.2618,
+      "step": 4214
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.3518492196094398,
+      "learning_rate": 8.94345914212918e-06,
+      "loss": 0.2373,
+      "step": 4215
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.9800016992965388,
+      "learning_rate": 8.942905251265787e-06,
+      "loss": 0.2305,
+      "step": 4216
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.239608197320148,
+      "learning_rate": 8.942351232413422e-06,
+      "loss": 0.2861,
+      "step": 4217
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.3998050116986656,
+      "learning_rate": 8.941797085590069e-06,
+      "loss": 0.2084,
+      "step": 4218
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.5562940101547762,
+      "learning_rate": 8.941242810813717e-06,
+      "loss": 0.2274,
+      "step": 4219
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.2286735366467563,
+      "learning_rate": 8.940688408102356e-06,
+      "loss": 0.2363,
+      "step": 4220
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.7692398690325375,
+      "learning_rate": 8.940133877473985e-06,
+      "loss": 0.2444,
+      "step": 4221
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.4337613595069358,
+      "learning_rate": 8.9395792189466e-06,
+      "loss": 0.2366,
+      "step": 4222
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.334063157621835,
+      "learning_rate": 8.939024432538207e-06,
+      "loss": 0.2259,
+      "step": 4223
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.3009368757566686,
+      "learning_rate": 8.938469518266819e-06,
+      "loss": 0.2384,
+      "step": 4224
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.158897298820986,
+      "learning_rate": 8.937914476150445e-06,
+      "loss": 0.2598,
+      "step": 4225
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.9568309691144997,
+      "learning_rate": 8.937359306207103e-06,
+      "loss": 0.2177,
+      "step": 4226
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.0138420502367065,
+      "learning_rate": 8.936804008454812e-06,
+      "loss": 0.1869,
+      "step": 4227
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.7206210810785523,
+      "learning_rate": 8.9362485829116e-06,
+      "loss": 0.1908,
+      "step": 4228
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.096322214949548,
+      "learning_rate": 8.935693029595496e-06,
+      "loss": 0.1904,
+      "step": 4229
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.9172589195460586,
+      "learning_rate": 8.935137348524532e-06,
+      "loss": 0.2067,
+      "step": 4230
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.1357987702199552,
+      "learning_rate": 8.934581539716748e-06,
+      "loss": 0.1866,
+      "step": 4231
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.060338434527011,
+      "learning_rate": 8.934025603190184e-06,
+      "loss": 0.1988,
+      "step": 4232
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.5983848635824949,
+      "learning_rate": 8.933469538962887e-06,
+      "loss": 0.2516,
+      "step": 4233
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.413704886186746,
+      "learning_rate": 8.932913347052908e-06,
+      "loss": 0.155,
+      "step": 4234
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.4080196861648016,
+      "learning_rate": 8.932357027478298e-06,
+      "loss": 0.1546,
+      "step": 4235
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.9967581795612049,
+      "learning_rate": 8.931800580257117e-06,
+      "loss": 0.2452,
+      "step": 4236
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.7130677722619752,
+      "learning_rate": 8.93124400540743e-06,
+      "loss": 0.1662,
+      "step": 4237
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.609025442715747,
+      "learning_rate": 8.930687302947302e-06,
+      "loss": 0.241,
+      "step": 4238
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.8022497110871519,
+      "learning_rate": 8.930130472894803e-06,
+      "loss": 0.1606,
+      "step": 4239
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.9139098056183375,
+      "learning_rate": 8.92957351526801e-06,
+      "loss": 0.164,
+      "step": 4240
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.253547198146712,
+      "learning_rate": 8.929016430085e-06,
+      "loss": 0.247,
+      "step": 4241
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.0661913887617487,
+      "learning_rate": 8.928459217363859e-06,
+      "loss": 0.2164,
+      "step": 4242
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.134093330029409,
+      "learning_rate": 8.92790187712267e-06,
+      "loss": 0.1769,
+      "step": 4243
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 2.2374035577394875,
+      "learning_rate": 8.92734440937953e-06,
+      "loss": 0.2169,
+      "step": 4244
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.8490260972094773,
+      "learning_rate": 8.92678681415253e-06,
+      "loss": 0.2153,
+      "step": 4245
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.0477825454920757,
+      "learning_rate": 8.926229091459772e-06,
+      "loss": 0.1975,
+      "step": 4246
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.7673834294146521,
+      "learning_rate": 8.925671241319362e-06,
+      "loss": 0.1438,
+      "step": 4247
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.7090955599992732,
+      "learning_rate": 8.925113263749405e-06,
+      "loss": 0.2493,
+      "step": 4248
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.9920188890075485,
+      "learning_rate": 8.924555158768013e-06,
+      "loss": 0.211,
+      "step": 4249
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.9106020858731595,
+      "learning_rate": 8.923996926393306e-06,
+      "loss": 0.1596,
+      "step": 4250
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.680160158767063,
+      "learning_rate": 8.9234385666434e-06,
+      "loss": 0.2866,
+      "step": 4251
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.9326652214902078,
+      "learning_rate": 8.922880079536425e-06,
+      "loss": 0.2302,
+      "step": 4252
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.4740001733747778,
+      "learning_rate": 8.922321465090504e-06,
+      "loss": 0.2315,
+      "step": 4253
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.1923247689531074,
+      "learning_rate": 8.921762723323774e-06,
+      "loss": 0.2036,
+      "step": 4254
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.092529508281666,
+      "learning_rate": 8.92120385425437e-06,
+      "loss": 0.2345,
+      "step": 4255
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.9865743084352787,
+      "learning_rate": 8.920644857900435e-06,
+      "loss": 0.2328,
+      "step": 4256
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.947539500335367,
+      "learning_rate": 8.920085734280113e-06,
+      "loss": 0.1958,
+      "step": 4257
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.8978038190952445,
+      "learning_rate": 8.919526483411555e-06,
+      "loss": 0.2017,
+      "step": 4258
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.116496852157465,
+      "learning_rate": 8.918967105312913e-06,
+      "loss": 0.1828,
+      "step": 4259
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.373688821099589,
+      "learning_rate": 8.918407600002344e-06,
+      "loss": 0.2145,
+      "step": 4260
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.956060413425276,
+      "learning_rate": 8.917847967498012e-06,
+      "loss": 0.1739,
+      "step": 4261
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.428268559688174,
+      "learning_rate": 8.91728820781808e-06,
+      "loss": 0.2139,
+      "step": 4262
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.2266878729119164,
+      "learning_rate": 8.916728320980723e-06,
+      "loss": 0.2079,
+      "step": 4263
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.1241046079133123,
+      "learning_rate": 8.916168307004112e-06,
+      "loss": 0.2479,
+      "step": 4264
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.612377168584661,
+      "learning_rate": 8.915608165906424e-06,
+      "loss": 0.1878,
+      "step": 4265
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.5765471383963572,
+      "learning_rate": 8.915047897705843e-06,
+      "loss": 0.2155,
+      "step": 4266
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.1512370533479248,
+      "learning_rate": 8.91448750242056e-06,
+      "loss": 0.2432,
+      "step": 4267
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.066719421402153,
+      "learning_rate": 8.913926980068759e-06,
+      "loss": 0.2169,
+      "step": 4268
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.2257715343056266,
+      "learning_rate": 8.913366330668637e-06,
+      "loss": 0.2284,
+      "step": 4269
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.2871332267669846,
+      "learning_rate": 8.912805554238393e-06,
+      "loss": 0.1914,
+      "step": 4270
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.947881549492129,
+      "learning_rate": 8.912244650796233e-06,
+      "loss": 0.1968,
+      "step": 4271
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.8961539895946065,
+      "learning_rate": 8.911683620360361e-06,
+      "loss": 0.2268,
+      "step": 4272
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.3203431300836954,
+      "learning_rate": 8.911122462948988e-06,
+      "loss": 0.2351,
+      "step": 4273
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.0130007990283676,
+      "learning_rate": 8.910561178580333e-06,
+      "loss": 0.1717,
+      "step": 4274
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.3249908477876478,
+      "learning_rate": 8.909999767272612e-06,
+      "loss": 0.2518,
+      "step": 4275
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.008238656113677,
+      "learning_rate": 8.90943822904405e-06,
+      "loss": 0.1937,
+      "step": 4276
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.3762281173010977,
+      "learning_rate": 8.908876563912876e-06,
+      "loss": 0.2218,
+      "step": 4277
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.334658494300119,
+      "learning_rate": 8.908314771897319e-06,
+      "loss": 0.2227,
+      "step": 4278
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.0048716389708643,
+      "learning_rate": 8.907752853015619e-06,
+      "loss": 0.2307,
+      "step": 4279
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.0936867223044233,
+      "learning_rate": 8.907190807286015e-06,
+      "loss": 0.2202,
+      "step": 4280
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.1398856987401422,
+      "learning_rate": 8.906628634726748e-06,
+      "loss": 0.2111,
+      "step": 4281
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.1640185517457198,
+      "learning_rate": 8.906066335356071e-06,
+      "loss": 0.2494,
+      "step": 4282
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.345727597804449,
+      "learning_rate": 8.905503909192234e-06,
+      "loss": 0.2267,
+      "step": 4283
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.9851220594261682,
+      "learning_rate": 8.904941356253493e-06,
+      "loss": 0.2062,
+      "step": 4284
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.8653539161682837,
+      "learning_rate": 8.904378676558113e-06,
+      "loss": 0.1957,
+      "step": 4285
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.0992668424658167,
+      "learning_rate": 8.903815870124353e-06,
+      "loss": 0.1715,
+      "step": 4286
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.8373520816588498,
+      "learning_rate": 8.903252936970487e-06,
+      "loss": 0.1618,
+      "step": 4287
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.2923538497624238,
+      "learning_rate": 8.902689877114786e-06,
+      "loss": 0.2246,
+      "step": 4288
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.9131375017752904,
+      "learning_rate": 8.902126690575525e-06,
+      "loss": 0.2356,
+      "step": 4289
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.2481531439716016,
+      "learning_rate": 8.90156337737099e-06,
+      "loss": 0.1833,
+      "step": 4290
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.365905775355453,
+      "learning_rate": 8.900999937519463e-06,
+      "loss": 0.2117,
+      "step": 4291
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.6295291912436385,
+      "learning_rate": 8.900436371039236e-06,
+      "loss": 0.2082,
+      "step": 4292
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.079056130397898,
+      "learning_rate": 8.899872677948602e-06,
+      "loss": 0.2103,
+      "step": 4293
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.147007385871383,
+      "learning_rate": 8.899308858265858e-06,
+      "loss": 0.1918,
+      "step": 4294
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.219169969214731,
+      "learning_rate": 8.898744912009306e-06,
+      "loss": 0.2182,
+      "step": 4295
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.304552956456611,
+      "learning_rate": 8.89818083919725e-06,
+      "loss": 0.1911,
+      "step": 4296
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.5166581899236633,
+      "learning_rate": 8.897616639848005e-06,
+      "loss": 0.2258,
+      "step": 4297
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.0905416076252017,
+      "learning_rate": 8.89705231397988e-06,
+      "loss": 0.2421,
+      "step": 4298
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.1243437038167055,
+      "learning_rate": 8.896487861611199e-06,
+      "loss": 0.1955,
+      "step": 4299
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.8635618845459881,
+      "learning_rate": 8.89592328276028e-06,
+      "loss": 0.1975,
+      "step": 4300
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.321131142076386,
+      "learning_rate": 8.89535857744545e-06,
+      "loss": 0.2038,
+      "step": 4301
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.8946394958968662,
+      "learning_rate": 8.894793745685043e-06,
+      "loss": 0.2021,
+      "step": 4302
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.3521572414157315,
+      "learning_rate": 8.894228787497389e-06,
+      "loss": 0.2659,
+      "step": 4303
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.8727026096348142,
+      "learning_rate": 8.89366370290083e-06,
+      "loss": 0.1586,
+      "step": 4304
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.0405691309414082,
+      "learning_rate": 8.893098491913707e-06,
+      "loss": 0.16,
+      "step": 4305
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.1285339355303698,
+      "learning_rate": 8.89253315455437e-06,
+      "loss": 0.1928,
+      "step": 4306
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.285467869508033,
+      "learning_rate": 8.891967690841168e-06,
+      "loss": 0.1702,
+      "step": 4307
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.7990669275951089,
+      "learning_rate": 8.891402100792457e-06,
+      "loss": 0.1679,
+      "step": 4308
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.9218937318321224,
+      "learning_rate": 8.890836384426596e-06,
+      "loss": 0.2215,
+      "step": 4309
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.9821420843722041,
+      "learning_rate": 8.890270541761948e-06,
+      "loss": 0.1971,
+      "step": 4310
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.1895339778611815,
+      "learning_rate": 8.889704572816883e-06,
+      "loss": 0.2102,
+      "step": 4311
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.9853957065187221,
+      "learning_rate": 8.889138477609768e-06,
+      "loss": 0.2071,
+      "step": 4312
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.0635002701063578,
+      "learning_rate": 8.888572256158982e-06,
+      "loss": 0.202,
+      "step": 4313
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.1008042370336586,
+      "learning_rate": 8.888005908482906e-06,
+      "loss": 0.2192,
+      "step": 4314
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.460752467815378,
+      "learning_rate": 8.887439434599922e-06,
+      "loss": 0.2218,
+      "step": 4315
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.3200680066195185,
+      "learning_rate": 8.886872834528418e-06,
+      "loss": 0.1887,
+      "step": 4316
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.9616208786672176,
+      "learning_rate": 8.886306108286787e-06,
+      "loss": 0.1866,
+      "step": 4317
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.4094562959590418,
+      "learning_rate": 8.885739255893426e-06,
+      "loss": 0.21,
+      "step": 4318
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.1122848385052118,
+      "learning_rate": 8.885172277366733e-06,
+      "loss": 0.1865,
+      "step": 4319
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.4355669539208662,
+      "learning_rate": 8.884605172725114e-06,
+      "loss": 0.2599,
+      "step": 4320
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.2267237211074946,
+      "learning_rate": 8.884037941986977e-06,
+      "loss": 0.1968,
+      "step": 4321
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.1706176569100404,
+      "learning_rate": 8.883470585170735e-06,
+      "loss": 0.1855,
+      "step": 4322
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.0085786579725609,
+      "learning_rate": 8.882903102294804e-06,
+      "loss": 0.2082,
+      "step": 4323
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.9938504070863514,
+      "learning_rate": 8.882335493377607e-06,
+      "loss": 0.2095,
+      "step": 4324
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.8803637397204285,
+      "learning_rate": 8.881767758437567e-06,
+      "loss": 0.2118,
+      "step": 4325
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.2534792565228814,
+      "learning_rate": 8.881199897493113e-06,
+      "loss": 0.2249,
+      "step": 4326
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.0070009679066354,
+      "learning_rate": 8.880631910562679e-06,
+      "loss": 0.2041,
+      "step": 4327
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.1854221963446199,
+      "learning_rate": 8.8800637976647e-06,
+      "loss": 0.2318,
+      "step": 4328
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.7124599015281865,
+      "learning_rate": 8.879495558817622e-06,
+      "loss": 0.2067,
+      "step": 4329
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.441135916131789,
+      "learning_rate": 8.878927194039884e-06,
+      "loss": 0.2678,
+      "step": 4330
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.1691457963695921,
+      "learning_rate": 8.87835870334994e-06,
+      "loss": 0.1779,
+      "step": 4331
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.6729865256323824,
+      "learning_rate": 8.877790086766243e-06,
+      "loss": 0.2344,
+      "step": 4332
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.1788582855892646,
+      "learning_rate": 8.87722134430725e-06,
+      "loss": 0.2053,
+      "step": 4333
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.1736546942095736,
+      "learning_rate": 8.876652475991421e-06,
+      "loss": 0.1746,
+      "step": 4334
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.2700915523348884,
+      "learning_rate": 8.876083481837224e-06,
+      "loss": 0.1708,
+      "step": 4335
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.3300524161566778,
+      "learning_rate": 8.875514361863129e-06,
+      "loss": 0.1947,
+      "step": 4336
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.4323777321764382,
+      "learning_rate": 8.874945116087609e-06,
+      "loss": 0.2379,
+      "step": 4337
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.2479408589259289,
+      "learning_rate": 8.874375744529143e-06,
+      "loss": 0.2092,
+      "step": 4338
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.0315139714428327,
+      "learning_rate": 8.873806247206211e-06,
+      "loss": 0.185,
+      "step": 4339
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.1460131012898334,
+      "learning_rate": 8.8732366241373e-06,
+      "loss": 0.1923,
+      "step": 4340
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.6808394983859738,
+      "learning_rate": 8.872666875340903e-06,
+      "loss": 0.2884,
+      "step": 4341
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 2.248975096500665,
+      "learning_rate": 8.872097000835511e-06,
+      "loss": 0.2391,
+      "step": 4342
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.4024319182707783,
+      "learning_rate": 8.871527000639624e-06,
+      "loss": 0.2157,
+      "step": 4343
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.3421040187604238,
+      "learning_rate": 8.870956874771745e-06,
+      "loss": 0.2582,
+      "step": 4344
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.879006931048409,
+      "learning_rate": 8.870386623250379e-06,
+      "loss": 0.1538,
+      "step": 4345
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.5155709910847277,
+      "learning_rate": 8.869816246094037e-06,
+      "loss": 0.2605,
+      "step": 4346
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.2978074875029137,
+      "learning_rate": 8.869245743321235e-06,
+      "loss": 0.2202,
+      "step": 4347
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.4995003429342808,
+      "learning_rate": 8.868675114950491e-06,
+      "loss": 0.2521,
+      "step": 4348
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.1073784929016537,
+      "learning_rate": 8.86810436100033e-06,
+      "loss": 0.1917,
+      "step": 4349
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.342272946110347,
+      "learning_rate": 8.867533481489275e-06,
+      "loss": 0.2296,
+      "step": 4350
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.3868181461656137,
+      "learning_rate": 8.866962476435861e-06,
+      "loss": 0.1703,
+      "step": 4351
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.3011921127793131,
+      "learning_rate": 8.866391345858618e-06,
+      "loss": 0.2017,
+      "step": 4352
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.1227210368281981,
+      "learning_rate": 8.86582008977609e-06,
+      "loss": 0.2614,
+      "step": 4353
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.30824444791832,
+      "learning_rate": 8.865248708206821e-06,
+      "loss": 0.1948,
+      "step": 4354
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.2554152639327183,
+      "learning_rate": 8.864677201169356e-06,
+      "loss": 0.2272,
+      "step": 4355
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.4003759300918712,
+      "learning_rate": 8.864105568682245e-06,
+      "loss": 0.2514,
+      "step": 4356
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.437109393974982,
+      "learning_rate": 8.863533810764046e-06,
+      "loss": 0.2192,
+      "step": 4357
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.9402211573059595,
+      "learning_rate": 8.862961927433318e-06,
+      "loss": 0.2085,
+      "step": 4358
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.2367152139898707,
+      "learning_rate": 8.862389918708624e-06,
+      "loss": 0.177,
+      "step": 4359
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.8034450234795716,
+      "learning_rate": 8.861817784608532e-06,
+      "loss": 0.1227,
+      "step": 4360
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.851788165026077,
+      "learning_rate": 8.861245525151614e-06,
+      "loss": 0.2016,
+      "step": 4361
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.1794454866333661,
+      "learning_rate": 8.860673140356447e-06,
+      "loss": 0.2096,
+      "step": 4362
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.080021468706354,
+      "learning_rate": 8.860100630241609e-06,
+      "loss": 0.1789,
+      "step": 4363
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.2942162672689357,
+      "learning_rate": 8.859527994825684e-06,
+      "loss": 0.2448,
+      "step": 4364
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.3585937614616408,
+      "learning_rate": 8.858955234127262e-06,
+      "loss": 0.2155,
+      "step": 4365
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.4951990903700056,
+      "learning_rate": 8.858382348164933e-06,
+      "loss": 0.2005,
+      "step": 4366
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.8487675233950577,
+      "learning_rate": 8.857809336957296e-06,
+      "loss": 0.1612,
+      "step": 4367
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.9217252130211318,
+      "learning_rate": 8.857236200522947e-06,
+      "loss": 0.2227,
+      "step": 4368
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.8273706734358731,
+      "learning_rate": 8.856662938880496e-06,
+      "loss": 0.178,
+      "step": 4369
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.281556325440414,
+      "learning_rate": 8.856089552048546e-06,
+      "loss": 0.2135,
+      "step": 4370
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.5919870766117863,
+      "learning_rate": 8.855516040045713e-06,
+      "loss": 0.2388,
+      "step": 4371
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.2894598781867634,
+      "learning_rate": 8.854942402890612e-06,
+      "loss": 0.2305,
+      "step": 4372
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.7990274949997642,
+      "learning_rate": 8.854368640601862e-06,
+      "loss": 0.1627,
+      "step": 4373
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.196174387203855,
+      "learning_rate": 8.853794753198092e-06,
+      "loss": 0.1935,
+      "step": 4374
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.9901561749774193,
+      "learning_rate": 8.853220740697927e-06,
+      "loss": 0.2218,
+      "step": 4375
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.4817494914345395,
+      "learning_rate": 8.852646603120001e-06,
+      "loss": 0.2187,
+      "step": 4376
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.9687196665291004,
+      "learning_rate": 8.852072340482951e-06,
+      "loss": 0.1725,
+      "step": 4377
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.134367076725257,
+      "learning_rate": 8.851497952805417e-06,
+      "loss": 0.1863,
+      "step": 4378
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.9920398497167429,
+      "learning_rate": 8.850923440106046e-06,
+      "loss": 0.1589,
+      "step": 4379
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.2606460363243293,
+      "learning_rate": 8.850348802403486e-06,
+      "loss": 0.1867,
+      "step": 4380
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.0376992568593206,
+      "learning_rate": 8.849774039716387e-06,
+      "loss": 0.2706,
+      "step": 4381
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.4476384983734518,
+      "learning_rate": 8.849199152063412e-06,
+      "loss": 0.226,
+      "step": 4382
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.169332771494504,
+      "learning_rate": 8.848624139463218e-06,
+      "loss": 0.2233,
+      "step": 4383
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.3930090602619818,
+      "learning_rate": 8.84804900193447e-06,
+      "loss": 0.2397,
+      "step": 4384
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.3350207399815954,
+      "learning_rate": 8.847473739495842e-06,
+      "loss": 0.1968,
+      "step": 4385
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.110576194244664,
+      "learning_rate": 8.846898352166e-06,
+      "loss": 0.1851,
+      "step": 4386
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.2726271719095799,
+      "learning_rate": 8.846322839963628e-06,
+      "loss": 0.2219,
+      "step": 4387
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.5280173357869073,
+      "learning_rate": 8.845747202907404e-06,
+      "loss": 0.286,
+      "step": 4388
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.3360897989630651,
+      "learning_rate": 8.845171441016014e-06,
+      "loss": 0.2085,
+      "step": 4389
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 3.3383997957949436,
+      "learning_rate": 8.844595554308149e-06,
+      "loss": 0.1959,
+      "step": 4390
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.0077436334900505,
+      "learning_rate": 8.8440195428025e-06,
+      "loss": 0.2039,
+      "step": 4391
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.9926797704433552,
+      "learning_rate": 8.843443406517766e-06,
+      "loss": 0.2266,
+      "step": 4392
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.6097936641229331,
+      "learning_rate": 8.842867145472651e-06,
+      "loss": 0.2136,
+      "step": 4393
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.3330958494521614,
+      "learning_rate": 8.842290759685857e-06,
+      "loss": 0.1708,
+      "step": 4394
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.609487619903907,
+      "learning_rate": 8.841714249176097e-06,
+      "loss": 0.2368,
+      "step": 4395
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.8624479827104871,
+      "learning_rate": 8.841137613962081e-06,
+      "loss": 0.1862,
+      "step": 4396
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.09182543367489,
+      "learning_rate": 8.840560854062532e-06,
+      "loss": 0.2024,
+      "step": 4397
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.1749646148527708,
+      "learning_rate": 8.839983969496166e-06,
+      "loss": 0.2322,
+      "step": 4398
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.30908196437541,
+      "learning_rate": 8.839406960281715e-06,
+      "loss": 0.2599,
+      "step": 4399
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.6836234292558578,
+      "learning_rate": 8.838829826437905e-06,
+      "loss": 0.249,
+      "step": 4400
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.095414553770584,
+      "learning_rate": 8.83825256798347e-06,
+      "loss": 0.1931,
+      "step": 4401
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.8100252356798545,
+      "learning_rate": 8.837675184937152e-06,
+      "loss": 0.2004,
+      "step": 4402
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.2804051468968511,
+      "learning_rate": 8.83709767731769e-06,
+      "loss": 0.1941,
+      "step": 4403
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.9373739396219123,
+      "learning_rate": 8.836520045143831e-06,
+      "loss": 0.1556,
+      "step": 4404
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.9100579773562789,
+      "learning_rate": 8.835942288434323e-06,
+      "loss": 0.2042,
+      "step": 4405
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.2402392830338573,
+      "learning_rate": 8.835364407207925e-06,
+      "loss": 0.1826,
+      "step": 4406
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.9349884365445362,
+      "learning_rate": 8.834786401483392e-06,
+      "loss": 0.1802,
+      "step": 4407
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.303647483395873,
+      "learning_rate": 8.834208271279487e-06,
+      "loss": 0.1954,
+      "step": 4408
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.0052355637428072,
+      "learning_rate": 8.833630016614976e-06,
+      "loss": 0.1643,
+      "step": 4409
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.0107555148955278,
+      "learning_rate": 8.833051637508632e-06,
+      "loss": 0.1823,
+      "step": 4410
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.004965066092093,
+      "learning_rate": 8.832473133979226e-06,
+      "loss": 0.2229,
+      "step": 4411
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.049203091729663,
+      "learning_rate": 8.83189450604554e-06,
+      "loss": 0.2105,
+      "step": 4412
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.1504291777311575,
+      "learning_rate": 8.831315753726355e-06,
+      "loss": 0.1944,
+      "step": 4413
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.0741703577412258,
+      "learning_rate": 8.830736877040455e-06,
+      "loss": 0.1486,
+      "step": 4414
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.3832657200617924,
+      "learning_rate": 8.830157876006635e-06,
+      "loss": 0.1669,
+      "step": 4415
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.5919849308069893,
+      "learning_rate": 8.829578750643689e-06,
+      "loss": 0.226,
+      "step": 4416
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.584591651527378,
+      "learning_rate": 8.828999500970412e-06,
+      "loss": 0.2329,
+      "step": 4417
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.1127662839615067,
+      "learning_rate": 8.82842012700561e-06,
+      "loss": 0.2088,
+      "step": 4418
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.5949178979185485,
+      "learning_rate": 8.82784062876809e-06,
+      "loss": 0.1549,
+      "step": 4419
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.6190165574565454,
+      "learning_rate": 8.82726100627666e-06,
+      "loss": 0.2246,
+      "step": 4420
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.5510274273705735,
+      "learning_rate": 8.826681259550139e-06,
+      "loss": 0.1679,
+      "step": 4421
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.8317610565818025,
+      "learning_rate": 8.826101388607343e-06,
+      "loss": 0.2484,
+      "step": 4422
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.0811570643715682,
+      "learning_rate": 8.825521393467095e-06,
+      "loss": 0.1774,
+      "step": 4423
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.3215475865625568,
+      "learning_rate": 8.824941274148224e-06,
+      "loss": 0.2039,
+      "step": 4424
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.4611337805281899,
+      "learning_rate": 8.82436103066956e-06,
+      "loss": 0.2641,
+      "step": 4425
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.104676721134503,
+      "learning_rate": 8.823780663049938e-06,
+      "loss": 0.2078,
+      "step": 4426
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.4169744849331571,
+      "learning_rate": 8.823200171308194e-06,
+      "loss": 0.208,
+      "step": 4427
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.3075142521473497,
+      "learning_rate": 8.822619555463178e-06,
+      "loss": 0.224,
+      "step": 4428
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.0478211191382047,
+      "learning_rate": 8.822038815533732e-06,
+      "loss": 0.165,
+      "step": 4429
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.3253452311802532,
+      "learning_rate": 8.821457951538706e-06,
+      "loss": 0.1936,
+      "step": 4430
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.1484508675815828,
+      "learning_rate": 8.820876963496959e-06,
+      "loss": 0.223,
+      "step": 4431
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.552642085373638,
+      "learning_rate": 8.820295851427348e-06,
+      "loss": 0.2052,
+      "step": 4432
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 13.212202466142715,
+      "learning_rate": 8.819714615348738e-06,
+      "loss": 0.1828,
+      "step": 4433
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.1540565853809188,
+      "learning_rate": 8.819133255279994e-06,
+      "loss": 0.1866,
+      "step": 4434
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.600118507153945,
+      "learning_rate": 8.818551771239988e-06,
+      "loss": 0.2475,
+      "step": 4435
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.2778370645414467,
+      "learning_rate": 8.817970163247597e-06,
+      "loss": 0.1865,
+      "step": 4436
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.3857890222658154,
+      "learning_rate": 8.817388431321698e-06,
+      "loss": 0.2236,
+      "step": 4437
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.4169030085867513,
+      "learning_rate": 8.816806575481176e-06,
+      "loss": 0.2505,
+      "step": 4438
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.4807935524502198,
+      "learning_rate": 8.816224595744917e-06,
+      "loss": 0.2257,
+      "step": 4439
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.197505985918992,
+      "learning_rate": 8.815642492131815e-06,
+      "loss": 0.2415,
+      "step": 4440
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.9785380141995665,
+      "learning_rate": 8.815060264660759e-06,
+      "loss": 0.2255,
+      "step": 4441
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.9648681629814391,
+      "learning_rate": 8.814477913350659e-06,
+      "loss": 0.1991,
+      "step": 4442
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 2.875903837518265,
+      "learning_rate": 8.813895438220408e-06,
+      "loss": 0.238,
+      "step": 4443
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.6140315733288892,
+      "learning_rate": 8.81331283928892e-06,
+      "loss": 0.1443,
+      "step": 4444
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.3226122662373183,
+      "learning_rate": 8.812730116575103e-06,
+      "loss": 0.2047,
+      "step": 4445
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.2589788597144729,
+      "learning_rate": 8.812147270097875e-06,
+      "loss": 0.1621,
+      "step": 4446
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.0215921673327937,
+      "learning_rate": 8.811564299876155e-06,
+      "loss": 0.1611,
+      "step": 4447
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.3797929374113092,
+      "learning_rate": 8.810981205928865e-06,
+      "loss": 0.2258,
+      "step": 4448
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.4130358116997865,
+      "learning_rate": 8.810397988274935e-06,
+      "loss": 0.1612,
+      "step": 4449
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.0113974267164267,
+      "learning_rate": 8.809814646933292e-06,
+      "loss": 0.2147,
+      "step": 4450
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.9423738400126036,
+      "learning_rate": 8.809231181922879e-06,
+      "loss": 0.1804,
+      "step": 4451
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.5219242429697524,
+      "learning_rate": 8.80864759326263e-06,
+      "loss": 0.229,
+      "step": 4452
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.8987879546438536,
+      "learning_rate": 8.808063880971487e-06,
+      "loss": 0.1983,
+      "step": 4453
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.4953476063624258,
+      "learning_rate": 8.807480045068404e-06,
+      "loss": 0.2461,
+      "step": 4454
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 3.0761387026981253,
+      "learning_rate": 8.806896085572328e-06,
+      "loss": 0.1956,
+      "step": 4455
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.4973589276174795,
+      "learning_rate": 8.806312002502215e-06,
+      "loss": 0.2349,
+      "step": 4456
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.8714865880687592,
+      "learning_rate": 8.805727795877028e-06,
+      "loss": 0.2108,
+      "step": 4457
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.6228997959191307,
+      "learning_rate": 8.805143465715727e-06,
+      "loss": 0.2219,
+      "step": 4458
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.1054150666051434,
+      "learning_rate": 8.80455901203728e-06,
+      "loss": 0.2276,
+      "step": 4459
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.460263433344344,
+      "learning_rate": 8.80397443486066e-06,
+      "loss": 0.2501,
+      "step": 4460
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.9395140392199552,
+      "learning_rate": 8.803389734204843e-06,
+      "loss": 0.247,
+      "step": 4461
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.0310348878821212,
+      "learning_rate": 8.80280491008881e-06,
+      "loss": 0.2006,
+      "step": 4462
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.4565265077843046,
+      "learning_rate": 8.80221996253154e-06,
+      "loss": 0.2122,
+      "step": 4463
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.2884407899501376,
+      "learning_rate": 8.801634891552025e-06,
+      "loss": 0.252,
+      "step": 4464
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.2724457810875986,
+      "learning_rate": 8.801049697169254e-06,
+      "loss": 0.2237,
+      "step": 4465
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.9606489963782552,
+      "learning_rate": 8.800464379402227e-06,
+      "loss": 0.2201,
+      "step": 4466
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.2765736504117964,
+      "learning_rate": 8.799878938269937e-06,
+      "loss": 0.192,
+      "step": 4467
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.1289602517540338,
+      "learning_rate": 8.799293373791394e-06,
+      "loss": 0.2703,
+      "step": 4468
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.3681979355577247,
+      "learning_rate": 8.798707685985604e-06,
+      "loss": 0.2249,
+      "step": 4469
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.3553660433551653,
+      "learning_rate": 8.798121874871577e-06,
+      "loss": 0.2049,
+      "step": 4470
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.4624417912634924,
+      "learning_rate": 8.79753594046833e-06,
+      "loss": 0.2301,
+      "step": 4471
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.912091113851784,
+      "learning_rate": 8.796949882794881e-06,
+      "loss": 0.2095,
+      "step": 4472
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.9232877064009657,
+      "learning_rate": 8.796363701870257e-06,
+      "loss": 0.2093,
+      "step": 4473
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.9865135566218108,
+      "learning_rate": 8.795777397713484e-06,
+      "loss": 0.1811,
+      "step": 4474
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 2.9867284392953817,
+      "learning_rate": 8.795190970343593e-06,
+      "loss": 0.2511,
+      "step": 4475
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.0054255837138468,
+      "learning_rate": 8.794604419779621e-06,
+      "loss": 0.2276,
+      "step": 4476
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.1832992077847138,
+      "learning_rate": 8.794017746040608e-06,
+      "loss": 0.2251,
+      "step": 4477
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.1812753081133551,
+      "learning_rate": 8.793430949145598e-06,
+      "loss": 0.249,
+      "step": 4478
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.3264415239296348,
+      "learning_rate": 8.792844029113637e-06,
+      "loss": 0.2237,
+      "step": 4479
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.3161102324324012,
+      "learning_rate": 8.79225698596378e-06,
+      "loss": 0.2719,
+      "step": 4480
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.191403814219502,
+      "learning_rate": 8.791669819715077e-06,
+      "loss": 0.2271,
+      "step": 4481
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.085903803556437,
+      "learning_rate": 8.791082530386594e-06,
+      "loss": 0.1941,
+      "step": 4482
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.9229817455042669,
+      "learning_rate": 8.790495117997391e-06,
+      "loss": 0.1701,
+      "step": 4483
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.5290062288353277,
+      "learning_rate": 8.789907582566536e-06,
+      "loss": 0.1951,
+      "step": 4484
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.654479405175094,
+      "learning_rate": 8.789319924113102e-06,
+      "loss": 0.2204,
+      "step": 4485
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.093459347014425,
+      "learning_rate": 8.788732142656164e-06,
+      "loss": 0.1868,
+      "step": 4486
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.4054830393507447,
+      "learning_rate": 8.788144238214803e-06,
+      "loss": 0.2377,
+      "step": 4487
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.2005850048319102,
+      "learning_rate": 8.787556210808101e-06,
+      "loss": 0.1681,
+      "step": 4488
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.2723612039964693,
+      "learning_rate": 8.786968060455147e-06,
+      "loss": 0.2152,
+      "step": 4489
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.0264758842831903,
+      "learning_rate": 8.786379787175032e-06,
+      "loss": 0.2023,
+      "step": 4490
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.9768487930260423,
+      "learning_rate": 8.785791390986851e-06,
+      "loss": 0.2379,
+      "step": 4491
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.1642923512204857,
+      "learning_rate": 8.785202871909704e-06,
+      "loss": 0.2239,
+      "step": 4492
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.830783053247842,
+      "learning_rate": 8.784614229962697e-06,
+      "loss": 0.2277,
+      "step": 4493
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.1740413637114269,
+      "learning_rate": 8.784025465164936e-06,
+      "loss": 0.2328,
+      "step": 4494
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.174816215566389,
+      "learning_rate": 8.783436577535529e-06,
+      "loss": 0.1875,
+      "step": 4495
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.0264209439240224,
+      "learning_rate": 8.782847567093598e-06,
+      "loss": 0.2041,
+      "step": 4496
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.6287344091782439,
+      "learning_rate": 8.782258433858259e-06,
+      "loss": 0.2148,
+      "step": 4497
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.274197038901747,
+      "learning_rate": 8.781669177848636e-06,
+      "loss": 0.2226,
+      "step": 4498
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.1835827842682838,
+      "learning_rate": 8.781079799083858e-06,
+      "loss": 0.2349,
+      "step": 4499
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.5130580592760174,
+      "learning_rate": 8.780490297583053e-06,
+      "loss": 0.1844,
+      "step": 4500
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.1453456519976721,
+      "learning_rate": 8.779900673365362e-06,
+      "loss": 0.2233,
+      "step": 4501
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.0886748173909688,
+      "learning_rate": 8.77931092644992e-06,
+      "loss": 0.2026,
+      "step": 4502
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.029409471176702,
+      "learning_rate": 8.778721056855872e-06,
+      "loss": 0.1948,
+      "step": 4503
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.8178113408822054,
+      "learning_rate": 8.778131064602366e-06,
+      "loss": 0.187,
+      "step": 4504
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.9598412017037049,
+      "learning_rate": 8.777540949708551e-06,
+      "loss": 0.222,
+      "step": 4505
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.9623104576091149,
+      "learning_rate": 8.776950712193588e-06,
+      "loss": 0.2065,
+      "step": 4506
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.5360619297637297,
+      "learning_rate": 8.776360352076632e-06,
+      "loss": 0.1994,
+      "step": 4507
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.1624136574467085,
+      "learning_rate": 8.775769869376846e-06,
+      "loss": 0.1923,
+      "step": 4508
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.6746423047589587,
+      "learning_rate": 8.775179264113399e-06,
+      "loss": 0.2454,
+      "step": 4509
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.0257687952138226,
+      "learning_rate": 8.774588536305464e-06,
+      "loss": 0.2202,
+      "step": 4510
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.6180185863826613,
+      "learning_rate": 8.773997685972214e-06,
+      "loss": 0.2148,
+      "step": 4511
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.0510500028876357,
+      "learning_rate": 8.773406713132826e-06,
+      "loss": 0.1752,
+      "step": 4512
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.3847977751535978,
+      "learning_rate": 8.77281561780649e-06,
+      "loss": 0.218,
+      "step": 4513
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.2777408630010447,
+      "learning_rate": 8.772224400012388e-06,
+      "loss": 0.2196,
+      "step": 4514
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.165088905106054,
+      "learning_rate": 8.771633059769712e-06,
+      "loss": 0.1943,
+      "step": 4515
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.8892015722978762,
+      "learning_rate": 8.771041597097658e-06,
+      "loss": 0.1685,
+      "step": 4516
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.1760220249437507,
+      "learning_rate": 8.770450012015427e-06,
+      "loss": 0.2223,
+      "step": 4517
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.5032023652840425,
+      "learning_rate": 8.76985830454222e-06,
+      "loss": 0.1853,
+      "step": 4518
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.3488466948140414,
+      "learning_rate": 8.769266474697245e-06,
+      "loss": 0.2149,
+      "step": 4519
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.4282171839421738,
+      "learning_rate": 8.76867452249971e-06,
+      "loss": 0.2653,
+      "step": 4520
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.9468895136000999,
+      "learning_rate": 8.768082447968835e-06,
+      "loss": 0.1466,
+      "step": 4521
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.4972458106366269,
+      "learning_rate": 8.767490251123837e-06,
+      "loss": 0.239,
+      "step": 4522
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 2.1925612617033123,
+      "learning_rate": 8.766897931983939e-06,
+      "loss": 0.1579,
+      "step": 4523
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.0054973191424017,
+      "learning_rate": 8.766305490568367e-06,
+      "loss": 0.2156,
+      "step": 4524
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.2604097420000429,
+      "learning_rate": 8.765712926896354e-06,
+      "loss": 0.2417,
+      "step": 4525
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.7454501954026125,
+      "learning_rate": 8.765120240987133e-06,
+      "loss": 0.1875,
+      "step": 4526
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.328073736211213,
+      "learning_rate": 8.764527432859946e-06,
+      "loss": 0.2021,
+      "step": 4527
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.939144218557902,
+      "learning_rate": 8.763934502534031e-06,
+      "loss": 0.1711,
+      "step": 4528
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.7617404665663385,
+      "learning_rate": 8.763341450028639e-06,
+      "loss": 0.226,
+      "step": 4529
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.8974272936470346,
+      "learning_rate": 8.76274827536302e-06,
+      "loss": 0.2113,
+      "step": 4530
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.0845592100116834,
+      "learning_rate": 8.762154978556428e-06,
+      "loss": 0.1839,
+      "step": 4531
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.2039375471915055,
+      "learning_rate": 8.761561559628122e-06,
+      "loss": 0.2047,
+      "step": 4532
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.124860807920851,
+      "learning_rate": 8.760968018597363e-06,
+      "loss": 0.1452,
+      "step": 4533
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.9033075981276352,
+      "learning_rate": 8.760374355483421e-06,
+      "loss": 0.2364,
+      "step": 4534
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.1682910267241098,
+      "learning_rate": 8.759780570305565e-06,
+      "loss": 0.2203,
+      "step": 4535
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.2075231731774818,
+      "learning_rate": 8.75918666308307e-06,
+      "loss": 0.2409,
+      "step": 4536
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.3185638099685963,
+      "learning_rate": 8.758592633835213e-06,
+      "loss": 0.186,
+      "step": 4537
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.2968831435487904,
+      "learning_rate": 8.757998482581278e-06,
+      "loss": 0.1559,
+      "step": 4538
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.4507705523744094,
+      "learning_rate": 8.757404209340552e-06,
+      "loss": 0.2382,
+      "step": 4539
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.8550416661990157,
+      "learning_rate": 8.756809814132324e-06,
+      "loss": 0.178,
+      "step": 4540
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.101476082557196,
+      "learning_rate": 8.75621529697589e-06,
+      "loss": 0.1974,
+      "step": 4541
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.2734249332311076,
+      "learning_rate": 8.755620657890547e-06,
+      "loss": 0.2477,
+      "step": 4542
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.0945483563738465,
+      "learning_rate": 8.755025896895597e-06,
+      "loss": 0.1695,
+      "step": 4543
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.1100704018313903,
+      "learning_rate": 8.754431014010349e-06,
+      "loss": 0.1706,
+      "step": 4544
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.0335721265602666,
+      "learning_rate": 8.75383600925411e-06,
+      "loss": 0.1881,
+      "step": 4545
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.2210874578078317,
+      "learning_rate": 8.753240882646196e-06,
+      "loss": 0.21,
+      "step": 4546
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.3345510527456095,
+      "learning_rate": 8.752645634205924e-06,
+      "loss": 0.2046,
+      "step": 4547
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.0001342467259746,
+      "learning_rate": 8.752050263952617e-06,
+      "loss": 0.2111,
+      "step": 4548
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.9109952941427139,
+      "learning_rate": 8.7514547719056e-06,
+      "loss": 0.1849,
+      "step": 4549
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.4912063708482128,
+      "learning_rate": 8.750859158084206e-06,
+      "loss": 0.237,
+      "step": 4550
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.3636028107304614,
+      "learning_rate": 8.750263422507766e-06,
+      "loss": 0.2066,
+      "step": 4551
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.9205528877144167,
+      "learning_rate": 8.749667565195619e-06,
+      "loss": 0.1476,
+      "step": 4552
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.0730416316184468,
+      "learning_rate": 8.749071586167106e-06,
+      "loss": 0.2209,
+      "step": 4553
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.1776442619820473,
+      "learning_rate": 8.748475485441573e-06,
+      "loss": 0.1818,
+      "step": 4554
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 2.9888778589613167,
+      "learning_rate": 8.747879263038371e-06,
+      "loss": 0.1917,
+      "step": 4555
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.9048972886374866,
+      "learning_rate": 8.747282918976852e-06,
+      "loss": 0.1857,
+      "step": 4556
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.5115498810715176,
+      "learning_rate": 8.746686453276377e-06,
+      "loss": 0.2296,
+      "step": 4557
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.648599797435606,
+      "learning_rate": 8.746089865956303e-06,
+      "loss": 0.2044,
+      "step": 4558
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.1791796127806085,
+      "learning_rate": 8.745493157035997e-06,
+      "loss": 0.2494,
+      "step": 4559
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 2.1526168123863387,
+      "learning_rate": 8.744896326534832e-06,
+      "loss": 0.3053,
+      "step": 4560
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.3045723144715093,
+      "learning_rate": 8.744299374472178e-06,
+      "loss": 0.1528,
+      "step": 4561
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.498313392694588,
+      "learning_rate": 8.743702300867412e-06,
+      "loss": 0.1895,
+      "step": 4562
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.2964980801383492,
+      "learning_rate": 8.743105105739917e-06,
+      "loss": 0.2229,
+      "step": 4563
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.1910740365730608,
+      "learning_rate": 8.742507789109077e-06,
+      "loss": 0.248,
+      "step": 4564
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.2791132307372295,
+      "learning_rate": 8.741910350994284e-06,
+      "loss": 0.2288,
+      "step": 4565
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.8974198490549334,
+      "learning_rate": 8.74131279141493e-06,
+      "loss": 0.2309,
+      "step": 4566
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.2140714330847795,
+      "learning_rate": 8.740715110390408e-06,
+      "loss": 0.1758,
+      "step": 4567
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.060551977057505,
+      "learning_rate": 8.740117307940123e-06,
+      "loss": 0.2207,
+      "step": 4568
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.3880817659099862,
+      "learning_rate": 8.739519384083482e-06,
+      "loss": 0.2328,
+      "step": 4569
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.4811939612182132,
+      "learning_rate": 8.73892133883989e-06,
+      "loss": 0.269,
+      "step": 4570
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.3778132680785682,
+      "learning_rate": 8.738323172228761e-06,
+      "loss": 0.2513,
+      "step": 4571
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.9889057884992757,
+      "learning_rate": 8.737724884269513e-06,
+      "loss": 0.1752,
+      "step": 4572
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.090363539203332,
+      "learning_rate": 8.737126474981566e-06,
+      "loss": 0.2192,
+      "step": 4573
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.1129666569915162,
+      "learning_rate": 8.736527944384344e-06,
+      "loss": 0.231,
+      "step": 4574
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.0280972172655958,
+      "learning_rate": 8.735929292497275e-06,
+      "loss": 0.2941,
+      "step": 4575
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.132141259707703,
+      "learning_rate": 8.735330519339794e-06,
+      "loss": 0.2117,
+      "step": 4576
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.8170191389905374,
+      "learning_rate": 8.734731624931336e-06,
+      "loss": 0.1662,
+      "step": 4577
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.8949262026145596,
+      "learning_rate": 8.734132609291344e-06,
+      "loss": 0.1727,
+      "step": 4578
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.2099341304109354,
+      "learning_rate": 8.733533472439258e-06,
+      "loss": 0.1883,
+      "step": 4579
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.0053690372573845,
+      "learning_rate": 8.732934214394529e-06,
+      "loss": 0.1463,
+      "step": 4580
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.1764384592194024,
+      "learning_rate": 8.73233483517661e-06,
+      "loss": 0.1555,
+      "step": 4581
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.0088687193183141,
+      "learning_rate": 8.731735334804953e-06,
+      "loss": 0.2113,
+      "step": 4582
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.9706838282541858,
+      "learning_rate": 8.731135713299024e-06,
+      "loss": 0.2176,
+      "step": 4583
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.1676440569071145,
+      "learning_rate": 8.730535970678281e-06,
+      "loss": 0.1781,
+      "step": 4584
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.5421380957871211,
+      "learning_rate": 8.729936106962198e-06,
+      "loss": 0.2577,
+      "step": 4585
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.1124695868333199,
+      "learning_rate": 8.729336122170245e-06,
+      "loss": 0.2416,
+      "step": 4586
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.1607882242638954,
+      "learning_rate": 8.728736016321895e-06,
+      "loss": 0.1768,
+      "step": 4587
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.4806308544356388,
+      "learning_rate": 8.72813578943663e-06,
+      "loss": 0.2341,
+      "step": 4588
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.3439794977899344,
+      "learning_rate": 8.727535441533935e-06,
+      "loss": 0.2205,
+      "step": 4589
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.1405146574838063,
+      "learning_rate": 8.726934972633294e-06,
+      "loss": 0.1893,
+      "step": 4590
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.8629239257665734,
+      "learning_rate": 8.726334382754202e-06,
+      "loss": 0.1777,
+      "step": 4591
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.900777252788889,
+      "learning_rate": 8.725733671916152e-06,
+      "loss": 0.1961,
+      "step": 4592
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.010066631321672,
+      "learning_rate": 8.725132840138644e-06,
+      "loss": 0.1753,
+      "step": 4593
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.421236922070566,
+      "learning_rate": 8.724531887441183e-06,
+      "loss": 0.2086,
+      "step": 4594
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.4430517028753442,
+      "learning_rate": 8.723930813843276e-06,
+      "loss": 0.239,
+      "step": 4595
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.2799717950065006,
+      "learning_rate": 8.723329619364432e-06,
+      "loss": 0.2205,
+      "step": 4596
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.1909030852378988,
+      "learning_rate": 8.722728304024169e-06,
+      "loss": 0.1991,
+      "step": 4597
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.332251311048402,
+      "learning_rate": 8.722126867842002e-06,
+      "loss": 0.2357,
+      "step": 4598
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.9653717171784709,
+      "learning_rate": 8.72152531083746e-06,
+      "loss": 0.1737,
+      "step": 4599
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.5631553515324936,
+      "learning_rate": 8.720923633030062e-06,
+      "loss": 0.23,
+      "step": 4600
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.0645976817115297,
+      "learning_rate": 8.720321834439347e-06,
+      "loss": 0.1783,
+      "step": 4601
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.4936430308215947,
+      "learning_rate": 8.719719915084844e-06,
+      "loss": 0.2541,
+      "step": 4602
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.498751356551389,
+      "learning_rate": 8.719117874986094e-06,
+      "loss": 0.2237,
+      "step": 4603
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.3378654421921408,
+      "learning_rate": 8.71851571416264e-06,
+      "loss": 0.1922,
+      "step": 4604
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.227183338235453,
+      "learning_rate": 8.717913432634027e-06,
+      "loss": 0.2265,
+      "step": 4605
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.3009007716898917,
+      "learning_rate": 8.717311030419806e-06,
+      "loss": 0.2088,
+      "step": 4606
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.4747572028985643,
+      "learning_rate": 8.71670850753953e-06,
+      "loss": 0.2425,
+      "step": 4607
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.9095037171777813,
+      "learning_rate": 8.716105864012761e-06,
+      "loss": 0.1624,
+      "step": 4608
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.2890897690171783,
+      "learning_rate": 8.715503099859057e-06,
+      "loss": 0.1836,
+      "step": 4609
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.183702312816095,
+      "learning_rate": 8.714900215097986e-06,
+      "loss": 0.2222,
+      "step": 4610
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.8994041467175948,
+      "learning_rate": 8.714297209749118e-06,
+      "loss": 0.1464,
+      "step": 4611
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.9493173903165955,
+      "learning_rate": 8.713694083832028e-06,
+      "loss": 0.1832,
+      "step": 4612
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.9137110829350159,
+      "learning_rate": 8.713090837366291e-06,
+      "loss": 0.1614,
+      "step": 4613
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.1162260059531062,
+      "learning_rate": 8.71248747037149e-06,
+      "loss": 0.1761,
+      "step": 4614
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.8007025767268741,
+      "learning_rate": 8.71188398286721e-06,
+      "loss": 0.2265,
+      "step": 4615
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.2248165359982426,
+      "learning_rate": 8.711280374873045e-06,
+      "loss": 0.234,
+      "step": 4616
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.3463334578323978,
+      "learning_rate": 8.710676646408583e-06,
+      "loss": 0.2347,
+      "step": 4617
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.003286633400583,
+      "learning_rate": 8.710072797493422e-06,
+      "loss": 0.2577,
+      "step": 4618
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.9017745096810251,
+      "learning_rate": 8.709468828147165e-06,
+      "loss": 0.1516,
+      "step": 4619
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.9515630673499208,
+      "learning_rate": 8.708864738389417e-06,
+      "loss": 0.2141,
+      "step": 4620
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.108406712169261,
+      "learning_rate": 8.708260528239788e-06,
+      "loss": 0.234,
+      "step": 4621
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.57168840510436,
+      "learning_rate": 8.707656197717889e-06,
+      "loss": 0.1859,
+      "step": 4622
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.9293539576899678,
+      "learning_rate": 8.707051746843338e-06,
+      "loss": 0.1556,
+      "step": 4623
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.291402129866529,
+      "learning_rate": 8.706447175635755e-06,
+      "loss": 0.2464,
+      "step": 4624
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 3.284562936155179,
+      "learning_rate": 8.705842484114766e-06,
+      "loss": 0.2313,
+      "step": 4625
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.3195671073847774,
+      "learning_rate": 8.7052376723e-06,
+      "loss": 0.1338,
+      "step": 4626
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.2329190999830306,
+      "learning_rate": 8.704632740211085e-06,
+      "loss": 0.1641,
+      "step": 4627
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.3914227220158923,
+      "learning_rate": 8.704027687867664e-06,
+      "loss": 0.2707,
+      "step": 4628
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.8921509352772279,
+      "learning_rate": 8.703422515289374e-06,
+      "loss": 0.1959,
+      "step": 4629
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.1038560713495118,
+      "learning_rate": 8.702817222495859e-06,
+      "loss": 0.2031,
+      "step": 4630
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.490603033889611,
+      "learning_rate": 8.702211809506768e-06,
+      "loss": 0.2113,
+      "step": 4631
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.9405958577486082,
+      "learning_rate": 8.701606276341753e-06,
+      "loss": 0.1595,
+      "step": 4632
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.206527222021542,
+      "learning_rate": 8.701000623020468e-06,
+      "loss": 0.2061,
+      "step": 4633
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.8950366649663001,
+      "learning_rate": 8.700394849562575e-06,
+      "loss": 0.1313,
+      "step": 4634
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.3722334089442025,
+      "learning_rate": 8.699788955987739e-06,
+      "loss": 0.2311,
+      "step": 4635
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.3905498050333274,
+      "learning_rate": 8.699182942315623e-06,
+      "loss": 0.1989,
+      "step": 4636
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.116559132011196,
+      "learning_rate": 8.698576808565904e-06,
+      "loss": 0.1781,
+      "step": 4637
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.9645453968614248,
+      "learning_rate": 8.697970554758256e-06,
+      "loss": 0.2128,
+      "step": 4638
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.2727932836312192,
+      "learning_rate": 8.697364180912354e-06,
+      "loss": 0.2221,
+      "step": 4639
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.047470629688314,
+      "learning_rate": 8.696757687047887e-06,
+      "loss": 0.1462,
+      "step": 4640
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.1438768293069932,
+      "learning_rate": 8.696151073184539e-06,
+      "loss": 0.2177,
+      "step": 4641
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.8056459152605402,
+      "learning_rate": 8.695544339342e-06,
+      "loss": 0.2414,
+      "step": 4642
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.4659229893211188,
+      "learning_rate": 8.69493748553997e-06,
+      "loss": 0.2534,
+      "step": 4643
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.42636719904386,
+      "learning_rate": 8.694330511798142e-06,
+      "loss": 0.2462,
+      "step": 4644
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.7793777187837143,
+      "learning_rate": 8.693723418136221e-06,
+      "loss": 0.2769,
+      "step": 4645
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.3366714010709477,
+      "learning_rate": 8.693116204573915e-06,
+      "loss": 0.177,
+      "step": 4646
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.3730766027279273,
+      "learning_rate": 8.692508871130933e-06,
+      "loss": 0.1878,
+      "step": 4647
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.2570257391847357,
+      "learning_rate": 8.69190141782699e-06,
+      "loss": 0.2546,
+      "step": 4648
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.2804305738887503,
+      "learning_rate": 8.691293844681804e-06,
+      "loss": 0.2392,
+      "step": 4649
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.9349479968448662,
+      "learning_rate": 8.690686151715098e-06,
+      "loss": 0.241,
+      "step": 4650
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.1025028526908274,
+      "learning_rate": 8.690078338946595e-06,
+      "loss": 0.2168,
+      "step": 4651
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.9188491089880632,
+      "learning_rate": 8.689470406396029e-06,
+      "loss": 0.2138,
+      "step": 4652
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.1581343908644584,
+      "learning_rate": 8.68886235408313e-06,
+      "loss": 0.1946,
+      "step": 4653
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.9952463023211185,
+      "learning_rate": 8.68825418202764e-06,
+      "loss": 0.2171,
+      "step": 4654
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.2754784381481687,
+      "learning_rate": 8.687645890249296e-06,
+      "loss": 0.1965,
+      "step": 4655
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.1677685266172333,
+      "learning_rate": 8.687037478767848e-06,
+      "loss": 0.2373,
+      "step": 4656
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.1140789052175366,
+      "learning_rate": 8.686428947603043e-06,
+      "loss": 0.184,
+      "step": 4657
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.0049245629196442,
+      "learning_rate": 8.685820296774634e-06,
+      "loss": 0.2077,
+      "step": 4658
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 2.390929377258727,
+      "learning_rate": 8.685211526302378e-06,
+      "loss": 0.2246,
+      "step": 4659
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.9667950586822094,
+      "learning_rate": 8.684602636206037e-06,
+      "loss": 0.1902,
+      "step": 4660
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.9719145639529495,
+      "learning_rate": 8.683993626505378e-06,
+      "loss": 0.2035,
+      "step": 4661
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.2728930344965894,
+      "learning_rate": 8.683384497220165e-06,
+      "loss": 0.1895,
+      "step": 4662
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.1964339304809835,
+      "learning_rate": 8.682775248370173e-06,
+      "loss": 0.2152,
+      "step": 4663
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.1098415391087144,
+      "learning_rate": 8.682165879975179e-06,
+      "loss": 0.2101,
+      "step": 4664
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.1573788136310124,
+      "learning_rate": 8.681556392054964e-06,
+      "loss": 0.2172,
+      "step": 4665
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.5959743277461296,
+      "learning_rate": 8.680946784629309e-06,
+      "loss": 0.2544,
+      "step": 4666
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.0402542393127472,
+      "learning_rate": 8.680337057718008e-06,
+      "loss": 0.2273,
+      "step": 4667
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.1014813104090801,
+      "learning_rate": 8.679727211340848e-06,
+      "loss": 0.2312,
+      "step": 4668
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.2836640595562736,
+      "learning_rate": 8.679117245517627e-06,
+      "loss": 0.2047,
+      "step": 4669
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.9719770120396456,
+      "learning_rate": 8.678507160268145e-06,
+      "loss": 0.1716,
+      "step": 4670
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.4343332584287205,
+      "learning_rate": 8.677896955612205e-06,
+      "loss": 0.2072,
+      "step": 4671
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.0435481444490515,
+      "learning_rate": 8.677286631569613e-06,
+      "loss": 0.1827,
+      "step": 4672
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.3256895110839517,
+      "learning_rate": 8.676676188160184e-06,
+      "loss": 0.1975,
+      "step": 4673
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.9979433203796425,
+      "learning_rate": 8.676065625403733e-06,
+      "loss": 0.1507,
+      "step": 4674
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.7235542916285143,
+      "learning_rate": 8.675454943320076e-06,
+      "loss": 0.1542,
+      "step": 4675
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.12006039546146,
+      "learning_rate": 8.67484414192904e-06,
+      "loss": 0.1736,
+      "step": 4676
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.1295217218945708,
+      "learning_rate": 8.674233221250448e-06,
+      "loss": 0.1647,
+      "step": 4677
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.9296465431936853,
+      "learning_rate": 8.673622181304132e-06,
+      "loss": 0.1953,
+      "step": 4678
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.9228573349868033,
+      "learning_rate": 8.67301102210993e-06,
+      "loss": 0.1291,
+      "step": 4679
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.0465453575452583,
+      "learning_rate": 8.672399743687678e-06,
+      "loss": 0.1915,
+      "step": 4680
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.4953261066881982,
+      "learning_rate": 8.671788346057217e-06,
+      "loss": 0.2212,
+      "step": 4681
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.462471675343796,
+      "learning_rate": 8.671176829238395e-06,
+      "loss": 0.2071,
+      "step": 4682
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.4519100827335467,
+      "learning_rate": 8.670565193251062e-06,
+      "loss": 0.1543,
+      "step": 4683
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.4788293152983258,
+      "learning_rate": 8.669953438115073e-06,
+      "loss": 0.1996,
+      "step": 4684
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.5370398266967844,
+      "learning_rate": 8.669341563850284e-06,
+      "loss": 0.259,
+      "step": 4685
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.1948142103891433,
+      "learning_rate": 8.66872957047656e-06,
+      "loss": 0.2253,
+      "step": 4686
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.3342672497710164,
+      "learning_rate": 8.668117458013763e-06,
+      "loss": 0.2138,
+      "step": 4687
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.1082707503792315,
+      "learning_rate": 8.667505226481763e-06,
+      "loss": 0.229,
+      "step": 4688
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.2004263944848501,
+      "learning_rate": 8.666892875900435e-06,
+      "loss": 0.2109,
+      "step": 4689
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.7474945943432263,
+      "learning_rate": 8.666280406289655e-06,
+      "loss": 0.2231,
+      "step": 4690
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.2981870058446356,
+      "learning_rate": 8.665667817669307e-06,
+      "loss": 0.2328,
+      "step": 4691
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.354228909296925,
+      "learning_rate": 8.665055110059273e-06,
+      "loss": 0.2434,
+      "step": 4692
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.6759931012493317,
+      "learning_rate": 8.664442283479442e-06,
+      "loss": 0.2185,
+      "step": 4693
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.2709938896356465,
+      "learning_rate": 8.663829337949708e-06,
+      "loss": 0.2134,
+      "step": 4694
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.2000964196157093,
+      "learning_rate": 8.663216273489965e-06,
+      "loss": 0.2025,
+      "step": 4695
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.5118264645390407,
+      "learning_rate": 8.662603090120117e-06,
+      "loss": 0.2258,
+      "step": 4696
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.3122387098034543,
+      "learning_rate": 8.661989787860065e-06,
+      "loss": 0.2051,
+      "step": 4697
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.2645121261600503,
+      "learning_rate": 8.661376366729719e-06,
+      "loss": 0.2373,
+      "step": 4698
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.3121464429084255,
+      "learning_rate": 8.660762826748991e-06,
+      "loss": 0.2504,
+      "step": 4699
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.2676279825241876,
+      "learning_rate": 8.660149167937796e-06,
+      "loss": 0.2518,
+      "step": 4700
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.9335211611902908,
+      "learning_rate": 8.659535390316056e-06,
+      "loss": 0.1649,
+      "step": 4701
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.8986016724730912,
+      "learning_rate": 8.658921493903692e-06,
+      "loss": 0.1909,
+      "step": 4702
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.9875405168596175,
+      "learning_rate": 8.658307478720634e-06,
+      "loss": 0.2059,
+      "step": 4703
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.0555359677102802,
+      "learning_rate": 8.657693344786809e-06,
+      "loss": 0.1836,
+      "step": 4704
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.1357555323019315,
+      "learning_rate": 8.657079092122156e-06,
+      "loss": 0.2207,
+      "step": 4705
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.5098903735242934,
+      "learning_rate": 8.656464720746612e-06,
+      "loss": 0.2151,
+      "step": 4706
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.3107547282501626,
+      "learning_rate": 8.65585023068012e-06,
+      "loss": 0.2039,
+      "step": 4707
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.8251987276627335,
+      "learning_rate": 8.655235621942627e-06,
+      "loss": 0.1935,
+      "step": 4708
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.1169070895617097,
+      "learning_rate": 8.654620894554086e-06,
+      "loss": 0.2083,
+      "step": 4709
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.3791221120092378,
+      "learning_rate": 8.654006048534448e-06,
+      "loss": 0.1985,
+      "step": 4710
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.189631969727475,
+      "learning_rate": 8.653391083903673e-06,
+      "loss": 0.1836,
+      "step": 4711
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.2111919174265844,
+      "learning_rate": 8.652776000681722e-06,
+      "loss": 0.2075,
+      "step": 4712
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.8077629273826304,
+      "learning_rate": 8.65216079888856e-06,
+      "loss": 0.2102,
+      "step": 4713
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.803680766647959,
+      "learning_rate": 8.65154547854416e-06,
+      "loss": 0.2386,
+      "step": 4714
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.2738071999093106,
+      "learning_rate": 8.650930039668495e-06,
+      "loss": 0.2151,
+      "step": 4715
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.300478196139016,
+      "learning_rate": 8.65031448228154e-06,
+      "loss": 0.2384,
+      "step": 4716
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.1289847826783732,
+      "learning_rate": 8.649698806403278e-06,
+      "loss": 0.174,
+      "step": 4717
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.5259336269484973,
+      "learning_rate": 8.649083012053694e-06,
+      "loss": 0.2373,
+      "step": 4718
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.4374652384100157,
+      "learning_rate": 8.648467099252778e-06,
+      "loss": 0.2184,
+      "step": 4719
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.287852719518254,
+      "learning_rate": 8.64785106802052e-06,
+      "loss": 0.2946,
+      "step": 4720
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.1441867793337688,
+      "learning_rate": 8.647234918376922e-06,
+      "loss": 0.2026,
+      "step": 4721
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.2919872569295003,
+      "learning_rate": 8.646618650341979e-06,
+      "loss": 0.2012,
+      "step": 4722
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.3703263606400269,
+      "learning_rate": 8.646002263935695e-06,
+      "loss": 0.1872,
+      "step": 4723
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.2470135724416294,
+      "learning_rate": 8.645385759178086e-06,
+      "loss": 0.233,
+      "step": 4724
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.296065339161626,
+      "learning_rate": 8.644769136089156e-06,
+      "loss": 0.2042,
+      "step": 4725
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.2132770119449954,
+      "learning_rate": 8.644152394688923e-06,
+      "loss": 0.2582,
+      "step": 4726
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.9445123172452758,
+      "learning_rate": 8.64353553499741e-06,
+      "loss": 0.1805,
+      "step": 4727
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.8460564205761849,
+      "learning_rate": 8.642918557034639e-06,
+      "loss": 0.2009,
+      "step": 4728
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.2389443161067353,
+      "learning_rate": 8.642301460820634e-06,
+      "loss": 0.2637,
+      "step": 4729
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.8387984135494274,
+      "learning_rate": 8.641684246375432e-06,
+      "loss": 0.1693,
+      "step": 4730
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.2842570633295325,
+      "learning_rate": 8.641066913719063e-06,
+      "loss": 0.1844,
+      "step": 4731
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 2.1634846270380694,
+      "learning_rate": 8.64044946287157e-06,
+      "loss": 0.1642,
+      "step": 4732
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.98997775068773,
+      "learning_rate": 8.639831893852993e-06,
+      "loss": 0.2139,
+      "step": 4733
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.0060482403373279,
+      "learning_rate": 8.63921420668338e-06,
+      "loss": 0.1683,
+      "step": 4734
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.177532991743866,
+      "learning_rate": 8.638596401382779e-06,
+      "loss": 0.2475,
+      "step": 4735
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.9859619423999809,
+      "learning_rate": 8.63797847797125e-06,
+      "loss": 0.1813,
+      "step": 4736
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.025305686339155,
+      "learning_rate": 8.637360436468844e-06,
+      "loss": 0.2304,
+      "step": 4737
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.1953633031146662,
+      "learning_rate": 8.63674227689563e-06,
+      "loss": 0.1721,
+      "step": 4738
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.1759508856514096,
+      "learning_rate": 8.636123999271668e-06,
+      "loss": 0.2163,
+      "step": 4739
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.4648898575202078,
+      "learning_rate": 8.635505603617031e-06,
+      "loss": 0.2625,
+      "step": 4740
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.723820792259735,
+      "learning_rate": 8.634887089951794e-06,
+      "loss": 0.2443,
+      "step": 4741
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.8966673859517071,
+      "learning_rate": 8.63426845829603e-06,
+      "loss": 0.2107,
+      "step": 4742
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.5630775909982706,
+      "learning_rate": 8.63364970866982e-06,
+      "loss": 0.1779,
+      "step": 4743
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.0902398527876365,
+      "learning_rate": 8.633030841093251e-06,
+      "loss": 0.2442,
+      "step": 4744
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.4993596038628163,
+      "learning_rate": 8.632411855586414e-06,
+      "loss": 0.2243,
+      "step": 4745
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.2214983585241628,
+      "learning_rate": 8.631792752169398e-06,
+      "loss": 0.1867,
+      "step": 4746
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 2.4763972380699975,
+      "learning_rate": 8.631173530862302e-06,
+      "loss": 0.2041,
+      "step": 4747
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.1790881876671109,
+      "learning_rate": 8.630554191685225e-06,
+      "loss": 0.2291,
+      "step": 4748
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.7599811633650404,
+      "learning_rate": 8.62993473465827e-06,
+      "loss": 0.2508,
+      "step": 4749
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.024557470896887,
+      "learning_rate": 8.629315159801547e-06,
+      "loss": 0.2217,
+      "step": 4750
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.1842916189110182,
+      "learning_rate": 8.628695467135166e-06,
+      "loss": 0.2477,
+      "step": 4751
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.1592517338693096,
+      "learning_rate": 8.628075656679244e-06,
+      "loss": 0.222,
+      "step": 4752
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.9982503825347447,
+      "learning_rate": 8.627455728453902e-06,
+      "loss": 0.2115,
+      "step": 4753
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.9437424047982352,
+      "learning_rate": 8.62683568247926e-06,
+      "loss": 0.2062,
+      "step": 4754
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.5403901959533914,
+      "learning_rate": 8.626215518775445e-06,
+      "loss": 0.2181,
+      "step": 4755
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.3438141691056382,
+      "learning_rate": 8.62559523736259e-06,
+      "loss": 0.2587,
+      "step": 4756
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.0652671509915654,
+      "learning_rate": 8.624974838260827e-06,
+      "loss": 0.1969,
+      "step": 4757
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.2375302564529758,
+      "learning_rate": 8.624354321490299e-06,
+      "loss": 0.1878,
+      "step": 4758
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.2206665842309226,
+      "learning_rate": 8.623733687071143e-06,
+      "loss": 0.2183,
+      "step": 4759
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.0879462731269482,
+      "learning_rate": 8.623112935023507e-06,
+      "loss": 0.2038,
+      "step": 4760
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.166544569032512,
+      "learning_rate": 8.622492065367544e-06,
+      "loss": 0.2325,
+      "step": 4761
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.9471839800778377,
+      "learning_rate": 8.621871078123404e-06,
+      "loss": 0.1843,
+      "step": 4762
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.8082947535265307,
+      "learning_rate": 8.621249973311247e-06,
+      "loss": 0.1809,
+      "step": 4763
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.1641920580867648,
+      "learning_rate": 8.620628750951234e-06,
+      "loss": 0.1939,
+      "step": 4764
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.1408946619948626,
+      "learning_rate": 8.62000741106353e-06,
+      "loss": 0.2126,
+      "step": 4765
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.1330428366137681,
+      "learning_rate": 8.619385953668303e-06,
+      "loss": 0.2018,
+      "step": 4766
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.190651604587951,
+      "learning_rate": 8.618764378785727e-06,
+      "loss": 0.1941,
+      "step": 4767
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.7936425845988466,
+      "learning_rate": 8.618142686435978e-06,
+      "loss": 0.1605,
+      "step": 4768
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.070251776371876,
+      "learning_rate": 8.617520876639235e-06,
+      "loss": 0.2417,
+      "step": 4769
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.4178175596021203,
+      "learning_rate": 8.616898949415688e-06,
+      "loss": 0.1893,
+      "step": 4770
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.6234100959050979,
+      "learning_rate": 8.616276904785519e-06,
+      "loss": 0.2424,
+      "step": 4771
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.3243394363456567,
+      "learning_rate": 8.615654742768921e-06,
+      "loss": 0.2643,
+      "step": 4772
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.2217317809441708,
+      "learning_rate": 8.615032463386093e-06,
+      "loss": 0.2409,
+      "step": 4773
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.1103973328912426,
+      "learning_rate": 8.614410066657233e-06,
+      "loss": 0.2585,
+      "step": 4774
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.9395613112383444,
+      "learning_rate": 8.613787552602541e-06,
+      "loss": 0.2077,
+      "step": 4775
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.2774565993922617,
+      "learning_rate": 8.61316492124223e-06,
+      "loss": 0.1955,
+      "step": 4776
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.5252477857405298,
+      "learning_rate": 8.612542172596505e-06,
+      "loss": 0.2062,
+      "step": 4777
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.2730936349671944,
+      "learning_rate": 8.611919306685585e-06,
+      "loss": 0.2209,
+      "step": 4778
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.4783399274686195,
+      "learning_rate": 8.611296323529688e-06,
+      "loss": 0.1827,
+      "step": 4779
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.1894867630760617,
+      "learning_rate": 8.610673223149036e-06,
+      "loss": 0.2402,
+      "step": 4780
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.9518077172249931,
+      "learning_rate": 8.610050005563852e-06,
+      "loss": 0.1858,
+      "step": 4781
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.3914060729731899,
+      "learning_rate": 8.609426670794372e-06,
+      "loss": 0.2465,
+      "step": 4782
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.2484209189668598,
+      "learning_rate": 8.608803218860826e-06,
+      "loss": 0.2246,
+      "step": 4783
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.2201639243404947,
+      "learning_rate": 8.60817964978345e-06,
+      "loss": 0.2109,
+      "step": 4784
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.1269013154931604,
+      "learning_rate": 8.60755596358249e-06,
+      "loss": 0.2336,
+      "step": 4785
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.6198912239348462,
+      "learning_rate": 8.60693216027819e-06,
+      "loss": 0.2735,
+      "step": 4786
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.2503036875177063,
+      "learning_rate": 8.606308239890794e-06,
+      "loss": 0.2279,
+      "step": 4787
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.132251237900119,
+      "learning_rate": 8.605684202440564e-06,
+      "loss": 0.2157,
+      "step": 4788
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.983223999068593,
+      "learning_rate": 8.605060047947747e-06,
+      "loss": 0.1626,
+      "step": 4789
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.4220799878669772,
+      "learning_rate": 8.60443577643261e-06,
+      "loss": 0.208,
+      "step": 4790
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.7693184059626684,
+      "learning_rate": 8.603811387915415e-06,
+      "loss": 0.1701,
+      "step": 4791
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.3815525802328659,
+      "learning_rate": 8.603186882416428e-06,
+      "loss": 0.2387,
+      "step": 4792
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.276614217778618,
+      "learning_rate": 8.602562259955925e-06,
+      "loss": 0.2108,
+      "step": 4793
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 2.2992962123253835,
+      "learning_rate": 8.601937520554178e-06,
+      "loss": 0.2161,
+      "step": 4794
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.1686890519603252,
+      "learning_rate": 8.60131266423147e-06,
+      "loss": 0.1347,
+      "step": 4795
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.8938840748850786,
+      "learning_rate": 8.600687691008079e-06,
+      "loss": 0.1629,
+      "step": 4796
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.2083140133951988,
+      "learning_rate": 8.600062600904297e-06,
+      "loss": 0.1647,
+      "step": 4797
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.017716957513384,
+      "learning_rate": 8.599437393940411e-06,
+      "loss": 0.1875,
+      "step": 4798
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.6646231612590273,
+      "learning_rate": 8.59881207013672e-06,
+      "loss": 0.1594,
+      "step": 4799
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.9786605613409622,
+      "learning_rate": 8.598186629513518e-06,
+      "loss": 0.2055,
+      "step": 4800
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.9981809897370817,
+      "learning_rate": 8.59756107209111e-06,
+      "loss": 0.1851,
+      "step": 4801
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.1446091264132507,
+      "learning_rate": 8.5969353978898e-06,
+      "loss": 0.2201,
+      "step": 4802
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.8080420408682425,
+      "learning_rate": 8.596309606929899e-06,
+      "loss": 0.1771,
+      "step": 4803
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.8551533892894022,
+      "learning_rate": 8.59568369923172e-06,
+      "loss": 0.2241,
+      "step": 4804
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.604019536562032,
+      "learning_rate": 8.59505767481558e-06,
+      "loss": 0.2949,
+      "step": 4805
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.4230329316791477,
+      "learning_rate": 8.594431533701801e-06,
+      "loss": 0.2381,
+      "step": 4806
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.2660109709960459,
+      "learning_rate": 8.593805275910707e-06,
+      "loss": 0.2505,
+      "step": 4807
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.1309003361348084,
+      "learning_rate": 8.593178901462628e-06,
+      "loss": 0.2067,
+      "step": 4808
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.515359600908115,
+      "learning_rate": 8.592552410377895e-06,
+      "loss": 0.1656,
+      "step": 4809
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.000150091858213,
+      "learning_rate": 8.591925802676844e-06,
+      "loss": 0.2133,
+      "step": 4810
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.193146347911001,
+      "learning_rate": 8.591299078379819e-06,
+      "loss": 0.2067,
+      "step": 4811
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.5908638261495702,
+      "learning_rate": 8.590672237507156e-06,
+      "loss": 0.2187,
+      "step": 4812
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.3707575100458727,
+      "learning_rate": 8.590045280079212e-06,
+      "loss": 0.1809,
+      "step": 4813
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.4325151179591091,
+      "learning_rate": 8.589418206116332e-06,
+      "loss": 0.2192,
+      "step": 4814
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.9412274932909286,
+      "learning_rate": 8.588791015638873e-06,
+      "loss": 0.2471,
+      "step": 4815
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.0326690474291673,
+      "learning_rate": 8.588163708667192e-06,
+      "loss": 0.2048,
+      "step": 4816
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.3255548333851885,
+      "learning_rate": 8.587536285221656e-06,
+      "loss": 0.2434,
+      "step": 4817
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.98307590063118,
+      "learning_rate": 8.586908745322628e-06,
+      "loss": 0.1742,
+      "step": 4818
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.0421485479673587,
+      "learning_rate": 8.586281088990482e-06,
+      "loss": 0.2205,
+      "step": 4819
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.3826403753192924,
+      "learning_rate": 8.585653316245588e-06,
+      "loss": 0.1781,
+      "step": 4820
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.0170517647400592,
+      "learning_rate": 8.585025427108324e-06,
+      "loss": 0.2053,
+      "step": 4821
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 2.082733737174621,
+      "learning_rate": 8.584397421599076e-06,
+      "loss": 0.1976,
+      "step": 4822
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.4053850480877186,
+      "learning_rate": 8.583769299738224e-06,
+      "loss": 0.2383,
+      "step": 4823
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.0995228768946421,
+      "learning_rate": 8.58314106154616e-06,
+      "loss": 0.2165,
+      "step": 4824
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.003170041271869,
+      "learning_rate": 8.582512707043277e-06,
+      "loss": 0.2092,
+      "step": 4825
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.2730497101807488,
+      "learning_rate": 8.581884236249972e-06,
+      "loss": 0.2347,
+      "step": 4826
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.2369232191851902,
+      "learning_rate": 8.581255649186644e-06,
+      "loss": 0.2172,
+      "step": 4827
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.1968950254991095,
+      "learning_rate": 8.5806269458737e-06,
+      "loss": 0.2029,
+      "step": 4828
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.4348296125233313,
+      "learning_rate": 8.579998126331546e-06,
+      "loss": 0.1763,
+      "step": 4829
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.654030677113438,
+      "learning_rate": 8.579369190580593e-06,
+      "loss": 0.2623,
+      "step": 4830
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.8005228653683798,
+      "learning_rate": 8.578740138641258e-06,
+      "loss": 0.2415,
+      "step": 4831
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.1927296284859568,
+      "learning_rate": 8.57811097053396e-06,
+      "loss": 0.2034,
+      "step": 4832
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.1271646919195586,
+      "learning_rate": 8.577481686279123e-06,
+      "loss": 0.1681,
+      "step": 4833
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.3816502065490806,
+      "learning_rate": 8.576852285897172e-06,
+      "loss": 0.2483,
+      "step": 4834
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.8724490600214686,
+      "learning_rate": 8.57622276940854e-06,
+      "loss": 0.1576,
+      "step": 4835
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.18693470047934,
+      "learning_rate": 8.575593136833662e-06,
+      "loss": 0.2699,
+      "step": 4836
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.9974163568325685,
+      "learning_rate": 8.574963388192971e-06,
+      "loss": 0.2088,
+      "step": 4837
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.8556649819103352,
+      "learning_rate": 8.574333523506916e-06,
+      "loss": 0.1879,
+      "step": 4838
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.1946642425220555,
+      "learning_rate": 8.573703542795938e-06,
+      "loss": 0.2384,
+      "step": 4839
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.8260451235384162,
+      "learning_rate": 8.573073446080488e-06,
+      "loss": 0.2229,
+      "step": 4840
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.2155879124129798,
+      "learning_rate": 8.57244323338102e-06,
+      "loss": 0.2057,
+      "step": 4841
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.0951212920847946,
+      "learning_rate": 8.57181290471799e-06,
+      "loss": 0.2061,
+      "step": 4842
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.5412376036033304,
+      "learning_rate": 8.571182460111857e-06,
+      "loss": 0.2471,
+      "step": 4843
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.0616095638692982,
+      "learning_rate": 8.57055189958309e-06,
+      "loss": 0.2166,
+      "step": 4844
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.2502688796715673,
+      "learning_rate": 8.569921223152156e-06,
+      "loss": 0.2075,
+      "step": 4845
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.540894325907098,
+      "learning_rate": 8.569290430839524e-06,
+      "loss": 0.2125,
+      "step": 4846
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.1633787227039039,
+      "learning_rate": 8.568659522665674e-06,
+      "loss": 0.202,
+      "step": 4847
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.1438084295596544,
+      "learning_rate": 8.568028498651083e-06,
+      "loss": 0.2325,
+      "step": 4848
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.1542583659246672,
+      "learning_rate": 8.567397358816235e-06,
+      "loss": 0.1683,
+      "step": 4849
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.502039686121971,
+      "learning_rate": 8.566766103181618e-06,
+      "loss": 0.2155,
+      "step": 4850
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.1089516372524488,
+      "learning_rate": 8.566134731767722e-06,
+      "loss": 0.2124,
+      "step": 4851
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.091522214663793,
+      "learning_rate": 8.565503244595042e-06,
+      "loss": 0.1913,
+      "step": 4852
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.1297403672749515,
+      "learning_rate": 8.564871641684076e-06,
+      "loss": 0.1931,
+      "step": 4853
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.9293165187996372,
+      "learning_rate": 8.564239923055327e-06,
+      "loss": 0.2054,
+      "step": 4854
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.8929224383158678,
+      "learning_rate": 8.563608088729299e-06,
+      "loss": 0.1506,
+      "step": 4855
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.257241944519261,
+      "learning_rate": 8.562976138726504e-06,
+      "loss": 0.2118,
+      "step": 4856
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.0728918255164743,
+      "learning_rate": 8.562344073067454e-06,
+      "loss": 0.2317,
+      "step": 4857
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.8920141677620694,
+      "learning_rate": 8.561711891772668e-06,
+      "loss": 0.18,
+      "step": 4858
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.9834441484057468,
+      "learning_rate": 8.561079594862666e-06,
+      "loss": 0.2073,
+      "step": 4859
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.3259423173816458,
+      "learning_rate": 8.560447182357974e-06,
+      "loss": 0.1622,
+      "step": 4860
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.027770113495797,
+      "learning_rate": 8.559814654279117e-06,
+      "loss": 0.17,
+      "step": 4861
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.5606407929806496,
+      "learning_rate": 8.55918201064663e-06,
+      "loss": 0.1633,
+      "step": 4862
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.6476777165309142,
+      "learning_rate": 8.558549251481048e-06,
+      "loss": 0.2085,
+      "step": 4863
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.0126937141516201,
+      "learning_rate": 8.55791637680291e-06,
+      "loss": 0.1881,
+      "step": 4864
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.0397543665260418,
+      "learning_rate": 8.557283386632761e-06,
+      "loss": 0.2093,
+      "step": 4865
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.398555706665104,
+      "learning_rate": 8.556650280991148e-06,
+      "loss": 0.2164,
+      "step": 4866
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.3484649857629436,
+      "learning_rate": 8.556017059898623e-06,
+      "loss": 0.2287,
+      "step": 4867
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.9555300031695584,
+      "learning_rate": 8.555383723375737e-06,
+      "loss": 0.2096,
+      "step": 4868
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.1655101111631945,
+      "learning_rate": 8.554750271443051e-06,
+      "loss": 0.2272,
+      "step": 4869
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.3084709992412766,
+      "learning_rate": 8.554116704121129e-06,
+      "loss": 0.208,
+      "step": 4870
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.1174302304279613,
+      "learning_rate": 8.553483021430534e-06,
+      "loss": 0.1749,
+      "step": 4871
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.2506066459574798,
+      "learning_rate": 8.552849223391837e-06,
+      "loss": 0.163,
+      "step": 4872
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.9504686415291429,
+      "learning_rate": 8.55221531002561e-06,
+      "loss": 0.1477,
+      "step": 4873
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.314094100486003,
+      "learning_rate": 8.551581281352433e-06,
+      "loss": 0.239,
+      "step": 4874
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.0499446261087078,
+      "learning_rate": 8.550947137392884e-06,
+      "loss": 0.1878,
+      "step": 4875
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.4984015028869067,
+      "learning_rate": 8.55031287816755e-06,
+      "loss": 0.2269,
+      "step": 4876
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.9326857516664692,
+      "learning_rate": 8.54967850369702e-06,
+      "loss": 0.1823,
+      "step": 4877
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.2458689077214395,
+      "learning_rate": 8.54904401400188e-06,
+      "loss": 0.1998,
+      "step": 4878
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.4114963036939598,
+      "learning_rate": 8.548409409102733e-06,
+      "loss": 0.2304,
+      "step": 4879
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.4827793228168085,
+      "learning_rate": 8.547774689020178e-06,
+      "loss": 0.2111,
+      "step": 4880
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.2048146997085298,
+      "learning_rate": 8.547139853774814e-06,
+      "loss": 0.2029,
+      "step": 4881
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.376019077172159,
+      "learning_rate": 8.546504903387252e-06,
+      "loss": 0.216,
+      "step": 4882
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.1003575227611144,
+      "learning_rate": 8.5458698378781e-06,
+      "loss": 0.177,
+      "step": 4883
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.251858929380012,
+      "learning_rate": 8.545234657267974e-06,
+      "loss": 0.2282,
+      "step": 4884
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.020385234717179,
+      "learning_rate": 8.544599361577493e-06,
+      "loss": 0.1996,
+      "step": 4885
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.398896040941505,
+      "learning_rate": 8.543963950827279e-06,
+      "loss": 0.2332,
+      "step": 4886
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.241522410019394,
+      "learning_rate": 8.543328425037955e-06,
+      "loss": 0.2128,
+      "step": 4887
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.425047521049785,
+      "learning_rate": 8.542692784230155e-06,
+      "loss": 0.237,
+      "step": 4888
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.222622698789704,
+      "learning_rate": 8.542057028424511e-06,
+      "loss": 0.2342,
+      "step": 4889
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.6878832594105139,
+      "learning_rate": 8.541421157641658e-06,
+      "loss": 0.1536,
+      "step": 4890
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.010124862047537,
+      "learning_rate": 8.540785171902237e-06,
+      "loss": 0.194,
+      "step": 4891
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.7604611585617087,
+      "learning_rate": 8.540149071226893e-06,
+      "loss": 0.2426,
+      "step": 4892
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.9020370657872282,
+      "learning_rate": 8.539512855636274e-06,
+      "loss": 0.2007,
+      "step": 4893
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.4339299314508909,
+      "learning_rate": 8.538876525151034e-06,
+      "loss": 0.2195,
+      "step": 4894
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.2040634798956298,
+      "learning_rate": 8.538240079791827e-06,
+      "loss": 0.216,
+      "step": 4895
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.8774154141269477,
+      "learning_rate": 8.537603519579311e-06,
+      "loss": 0.2338,
+      "step": 4896
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.9203007848095982,
+      "learning_rate": 8.536966844534151e-06,
+      "loss": 0.1816,
+      "step": 4897
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.9729013727774264,
+      "learning_rate": 8.536330054677014e-06,
+      "loss": 0.1857,
+      "step": 4898
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.1832263914649943,
+      "learning_rate": 8.535693150028569e-06,
+      "loss": 0.2049,
+      "step": 4899
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.0085793320550898,
+      "learning_rate": 8.53505613060949e-06,
+      "loss": 0.2126,
+      "step": 4900
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.9910318720263521,
+      "learning_rate": 8.534418996440459e-06,
+      "loss": 0.1655,
+      "step": 4901
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.054436918796761,
+      "learning_rate": 8.533781747542154e-06,
+      "loss": 0.184,
+      "step": 4902
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.0565030952437666,
+      "learning_rate": 8.533144383935261e-06,
+      "loss": 0.1833,
+      "step": 4903
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.9290328686258763,
+      "learning_rate": 8.53250690564047e-06,
+      "loss": 0.2004,
+      "step": 4904
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.125326774742863,
+      "learning_rate": 8.531869312678473e-06,
+      "loss": 0.2107,
+      "step": 4905
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.604000535145647,
+      "learning_rate": 8.531231605069967e-06,
+      "loss": 0.2789,
+      "step": 4906
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.1026368490598617,
+      "learning_rate": 8.530593782835653e-06,
+      "loss": 0.1918,
+      "step": 4907
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.1216719362581262,
+      "learning_rate": 8.529955845996235e-06,
+      "loss": 0.1824,
+      "step": 4908
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.1330903214107384,
+      "learning_rate": 8.529317794572419e-06,
+      "loss": 0.2065,
+      "step": 4909
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.1888884000858275,
+      "learning_rate": 8.528679628584918e-06,
+      "loss": 0.2184,
+      "step": 4910
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.9784167310191172,
+      "learning_rate": 8.528041348054448e-06,
+      "loss": 0.2159,
+      "step": 4911
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.168357270984037,
+      "learning_rate": 8.527402953001727e-06,
+      "loss": 0.1969,
+      "step": 4912
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.6957513179754249,
+      "learning_rate": 8.526764443447477e-06,
+      "loss": 0.255,
+      "step": 4913
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.052225260517258,
+      "learning_rate": 8.526125819412426e-06,
+      "loss": 0.2137,
+      "step": 4914
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.7291811464653277,
+      "learning_rate": 8.525487080917302e-06,
+      "loss": 0.2082,
+      "step": 4915
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.1587570982008866,
+      "learning_rate": 8.524848227982843e-06,
+      "loss": 0.2488,
+      "step": 4916
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.3276472031634259,
+      "learning_rate": 8.524209260629781e-06,
+      "loss": 0.1912,
+      "step": 4917
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.7771474205824997,
+      "learning_rate": 8.52357017887886e-06,
+      "loss": 0.1696,
+      "step": 4918
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.0017645827652364,
+      "learning_rate": 8.522930982750825e-06,
+      "loss": 0.2233,
+      "step": 4919
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.954867805929508,
+      "learning_rate": 8.522291672266425e-06,
+      "loss": 0.2229,
+      "step": 4920
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.4127876306950973,
+      "learning_rate": 8.521652247446412e-06,
+      "loss": 0.2515,
+      "step": 4921
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.216810976439642,
+      "learning_rate": 8.52101270831154e-06,
+      "loss": 0.2282,
+      "step": 4922
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.7716794455452329,
+      "learning_rate": 8.520373054882573e-06,
+      "loss": 0.141,
+      "step": 4923
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.2198219842289435,
+      "learning_rate": 8.519733287180272e-06,
+      "loss": 0.2116,
+      "step": 4924
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.1159082512946947,
+      "learning_rate": 8.519093405225404e-06,
+      "loss": 0.2356,
+      "step": 4925
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.3018050732515,
+      "learning_rate": 8.51845340903874e-06,
+      "loss": 0.1898,
+      "step": 4926
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.088384286859492,
+      "learning_rate": 8.517813298641056e-06,
+      "loss": 0.1302,
+      "step": 4927
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.0901246682997527,
+      "learning_rate": 8.51717307405313e-06,
+      "loss": 0.2083,
+      "step": 4928
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.2679326218577813,
+      "learning_rate": 8.516532735295742e-06,
+      "loss": 0.2185,
+      "step": 4929
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.4122028860679332,
+      "learning_rate": 8.515892282389679e-06,
+      "loss": 0.2489,
+      "step": 4930
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.1610111368527793,
+      "learning_rate": 8.515251715355732e-06,
+      "loss": 0.1917,
+      "step": 4931
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.3450171157437438,
+      "learning_rate": 8.514611034214691e-06,
+      "loss": 0.2034,
+      "step": 4932
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.0637606984714427,
+      "learning_rate": 8.513970238987357e-06,
+      "loss": 0.2004,
+      "step": 4933
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.2300515088096722,
+      "learning_rate": 8.513329329694527e-06,
+      "loss": 0.216,
+      "step": 4934
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.9733534017908207,
+      "learning_rate": 8.512688306357007e-06,
+      "loss": 0.1398,
+      "step": 4935
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.16556121767359,
+      "learning_rate": 8.512047168995605e-06,
+      "loss": 0.1728,
+      "step": 4936
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.879215347191256,
+      "learning_rate": 8.511405917631132e-06,
+      "loss": 0.2173,
+      "step": 4937
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.6245089302463294,
+      "learning_rate": 8.510764552284406e-06,
+      "loss": 0.2444,
+      "step": 4938
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.3819669277259456,
+      "learning_rate": 8.51012307297624e-06,
+      "loss": 0.2114,
+      "step": 4939
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.1004657645904719,
+      "learning_rate": 8.509481479727462e-06,
+      "loss": 0.1668,
+      "step": 4940
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.1988295170405197,
+      "learning_rate": 8.508839772558897e-06,
+      "loss": 0.2207,
+      "step": 4941
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.0389599382005208,
+      "learning_rate": 8.508197951491376e-06,
+      "loss": 0.2099,
+      "step": 4942
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.098990583021935,
+      "learning_rate": 8.507556016545733e-06,
+      "loss": 0.2377,
+      "step": 4943
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.0983611834957723,
+      "learning_rate": 8.506913967742803e-06,
+      "loss": 0.2159,
+      "step": 4944
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.29733221772793,
+      "learning_rate": 8.506271805103428e-06,
+      "loss": 0.2326,
+      "step": 4945
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.4658643829003177,
+      "learning_rate": 8.505629528648456e-06,
+      "loss": 0.1996,
+      "step": 4946
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.3911141167608383,
+      "learning_rate": 8.504987138398734e-06,
+      "loss": 0.1952,
+      "step": 4947
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.2423084907868147,
+      "learning_rate": 8.504344634375112e-06,
+      "loss": 0.2297,
+      "step": 4948
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.5713412819691197,
+      "learning_rate": 8.50370201659845e-06,
+      "loss": 0.2075,
+      "step": 4949
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.1317530182818627,
+      "learning_rate": 8.503059285089604e-06,
+      "loss": 0.2161,
+      "step": 4950
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.1781996144943798,
+      "learning_rate": 8.50241643986944e-06,
+      "loss": 0.1722,
+      "step": 4951
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.0800320631004128,
+      "learning_rate": 8.501773480958824e-06,
+      "loss": 0.1443,
+      "step": 4952
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.975722901865251,
+      "learning_rate": 8.501130408378627e-06,
+      "loss": 0.2422,
+      "step": 4953
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.4662797799901315,
+      "learning_rate": 8.500487222149726e-06,
+      "loss": 0.1446,
+      "step": 4954
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.2388716622887772,
+      "learning_rate": 8.499843922292994e-06,
+      "loss": 0.2578,
+      "step": 4955
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.2209665853386422,
+      "learning_rate": 8.499200508829317e-06,
+      "loss": 0.2312,
+      "step": 4956
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.312645932440707,
+      "learning_rate": 8.498556981779577e-06,
+      "loss": 0.2179,
+      "step": 4957
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.9225482048498795,
+      "learning_rate": 8.49791334116467e-06,
+      "loss": 0.1668,
+      "step": 4958
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.3497264524210342,
+      "learning_rate": 8.497269587005482e-06,
+      "loss": 0.2318,
+      "step": 4959
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.1593789103794216,
+      "learning_rate": 8.496625719322911e-06,
+      "loss": 0.1827,
+      "step": 4960
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.2107296980711972,
+      "learning_rate": 8.495981738137859e-06,
+      "loss": 0.2229,
+      "step": 4961
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.141565262745204,
+      "learning_rate": 8.49533764347123e-06,
+      "loss": 0.1925,
+      "step": 4962
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.0153283841477723,
+      "learning_rate": 8.494693435343931e-06,
+      "loss": 0.2155,
+      "step": 4963
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.3318938823524566,
+      "learning_rate": 8.494049113776875e-06,
+      "loss": 0.1818,
+      "step": 4964
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.2994576943240985,
+      "learning_rate": 8.493404678790974e-06,
+      "loss": 0.1755,
+      "step": 4965
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.0373144126699085,
+      "learning_rate": 8.492760130407148e-06,
+      "loss": 0.2074,
+      "step": 4966
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.178462794873808,
+      "learning_rate": 8.49211546864632e-06,
+      "loss": 0.1689,
+      "step": 4967
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.6362963156341424,
+      "learning_rate": 8.491470693529416e-06,
+      "loss": 0.2222,
+      "step": 4968
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.9429326775666719,
+      "learning_rate": 8.490825805077366e-06,
+      "loss": 0.1796,
+      "step": 4969
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.1486154827074835,
+      "learning_rate": 8.490180803311103e-06,
+      "loss": 0.2258,
+      "step": 4970
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.941518270399987,
+      "learning_rate": 8.489535688251562e-06,
+      "loss": 0.214,
+      "step": 4971
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.4376725710031655,
+      "learning_rate": 8.488890459919687e-06,
+      "loss": 0.2579,
+      "step": 4972
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.039882604265742,
+      "learning_rate": 8.488245118336423e-06,
+      "loss": 0.2017,
+      "step": 4973
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.8928706933381235,
+      "learning_rate": 8.487599663522716e-06,
+      "loss": 0.1996,
+      "step": 4974
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.11178862914383,
+      "learning_rate": 8.486954095499518e-06,
+      "loss": 0.2489,
+      "step": 4975
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.1772648490636386,
+      "learning_rate": 8.486308414287784e-06,
+      "loss": 0.2109,
+      "step": 4976
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.3816642250157183,
+      "learning_rate": 8.485662619908474e-06,
+      "loss": 0.2372,
+      "step": 4977
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.252814762615845,
+      "learning_rate": 8.485016712382554e-06,
+      "loss": 0.2362,
+      "step": 4978
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.9379206419982387,
+      "learning_rate": 8.484370691730985e-06,
+      "loss": 0.1866,
+      "step": 4979
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.8859984196552337,
+      "learning_rate": 8.48372455797474e-06,
+      "loss": 0.2121,
+      "step": 4980
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.9974936129430058,
+      "learning_rate": 8.483078311134793e-06,
+      "loss": 0.2066,
+      "step": 4981
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.2557504838096336,
+      "learning_rate": 8.48243195123212e-06,
+      "loss": 0.2352,
+      "step": 4982
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.929496307785681,
+      "learning_rate": 8.481785478287706e-06,
+      "loss": 0.2321,
+      "step": 4983
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.1105304681248303,
+      "learning_rate": 8.48113889232253e-06,
+      "loss": 0.2035,
+      "step": 4984
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.1891445071166615,
+      "learning_rate": 8.480492193357585e-06,
+      "loss": 0.2054,
+      "step": 4985
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.2582417282914178,
+      "learning_rate": 8.479845381413862e-06,
+      "loss": 0.2072,
+      "step": 4986
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.2984509687723575,
+      "learning_rate": 8.479198456512357e-06,
+      "loss": 0.2498,
+      "step": 4987
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.1123268520434375,
+      "learning_rate": 8.478551418674069e-06,
+      "loss": 0.1843,
+      "step": 4988
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.1555056560263208,
+      "learning_rate": 8.477904267920002e-06,
+      "loss": 0.1312,
+      "step": 4989
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.058193117779474,
+      "learning_rate": 8.477257004271162e-06,
+      "loss": 0.1677,
+      "step": 4990
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.2312267027746961,
+      "learning_rate": 8.476609627748563e-06,
+      "loss": 0.2176,
+      "step": 4991
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.0331908263139726,
+      "learning_rate": 8.475962138373212e-06,
+      "loss": 0.1945,
+      "step": 4992
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.6589550339969663,
+      "learning_rate": 8.475314536166133e-06,
+      "loss": 0.2381,
+      "step": 4993
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.2663177511967136,
+      "learning_rate": 8.474666821148346e-06,
+      "loss": 0.171,
+      "step": 4994
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.141146565100039,
+      "learning_rate": 8.474018993340875e-06,
+      "loss": 0.2515,
+      "step": 4995
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.9278245159327012,
+      "learning_rate": 8.47337105276475e-06,
+      "loss": 0.1831,
+      "step": 4996
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.0903263476266765,
+      "learning_rate": 8.472722999441005e-06,
+      "loss": 0.1875,
+      "step": 4997
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.2703622225851994,
+      "learning_rate": 8.472074833390674e-06,
+      "loss": 0.2084,
+      "step": 4998
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.0183186316963766,
+      "learning_rate": 8.471426554634796e-06,
+      "loss": 0.2315,
+      "step": 4999
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.434267737963635,
+      "learning_rate": 8.470778163194416e-06,
+      "loss": 0.1903,
+      "step": 5000
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.3438689988695713,
+      "learning_rate": 8.470129659090581e-06,
+      "loss": 0.2231,
+      "step": 5001
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.7720733808499428,
+      "learning_rate": 8.469481042344341e-06,
+      "loss": 0.1164,
+      "step": 5002
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.9875579248865204,
+      "learning_rate": 8.468832312976753e-06,
+      "loss": 0.1988,
+      "step": 5003
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.2888938114302741,
+      "learning_rate": 8.468183471008872e-06,
+      "loss": 0.2271,
+      "step": 5004
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.0251715671335175,
+      "learning_rate": 8.467534516461763e-06,
+      "loss": 0.228,
+      "step": 5005
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.238345021191209,
+      "learning_rate": 8.466885449356488e-06,
+      "loss": 0.2035,
+      "step": 5006
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.2238876780617032,
+      "learning_rate": 8.466236269714119e-06,
+      "loss": 0.2339,
+      "step": 5007
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.0076031541373949,
+      "learning_rate": 8.465586977555726e-06,
+      "loss": 0.2245,
+      "step": 5008
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.0201055016634366,
+      "learning_rate": 8.464937572902388e-06,
+      "loss": 0.1828,
+      "step": 5009
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.3639406014297801,
+      "learning_rate": 8.464288055775185e-06,
+      "loss": 0.2124,
+      "step": 5010
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.1982003463491038,
+      "learning_rate": 8.463638426195197e-06,
+      "loss": 0.1959,
+      "step": 5011
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.9319373823766639,
+      "learning_rate": 8.462988684183515e-06,
+      "loss": 0.186,
+      "step": 5012
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.8259483355130011,
+      "learning_rate": 8.46233882976123e-06,
+      "loss": 0.2018,
+      "step": 5013
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.2429936090617328,
+      "learning_rate": 8.461688862949433e-06,
+      "loss": 0.2046,
+      "step": 5014
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.1245038011133854,
+      "learning_rate": 8.461038783769227e-06,
+      "loss": 0.1627,
+      "step": 5015
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.114080199278937,
+      "learning_rate": 8.460388592241711e-06,
+      "loss": 0.2203,
+      "step": 5016
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.8028972256351928,
+      "learning_rate": 8.459738288387993e-06,
+      "loss": 0.1976,
+      "step": 5017
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.1954870797740058,
+      "learning_rate": 8.459087872229178e-06,
+      "loss": 0.2423,
+      "step": 5018
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.8444519919752145,
+      "learning_rate": 8.458437343786384e-06,
+      "loss": 0.1858,
+      "step": 5019
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.0927212406950122,
+      "learning_rate": 8.457786703080723e-06,
+      "loss": 0.2007,
+      "step": 5020
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.1514130616371578,
+      "learning_rate": 8.45713595013332e-06,
+      "loss": 0.173,
+      "step": 5021
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.0417707041602482,
+      "learning_rate": 8.456485084965294e-06,
+      "loss": 0.1935,
+      "step": 5022
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.6758960337775326,
+      "learning_rate": 8.455834107597775e-06,
+      "loss": 0.1973,
+      "step": 5023
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.4211737244821354,
+      "learning_rate": 8.455183018051893e-06,
+      "loss": 0.2167,
+      "step": 5024
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.9781213335267108,
+      "learning_rate": 8.454531816348784e-06,
+      "loss": 0.2299,
+      "step": 5025
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.8187413769369815,
+      "learning_rate": 8.453880502509586e-06,
+      "loss": 0.1627,
+      "step": 5026
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.4095782392463156,
+      "learning_rate": 8.45322907655544e-06,
+      "loss": 0.2541,
+      "step": 5027
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.2693708986103864,
+      "learning_rate": 8.452577538507495e-06,
+      "loss": 0.1502,
+      "step": 5028
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.5158712958752005,
+      "learning_rate": 8.451925888386894e-06,
+      "loss": 0.1874,
+      "step": 5029
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.3304397494727664,
+      "learning_rate": 8.451274126214794e-06,
+      "loss": 0.2177,
+      "step": 5030
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.2578783758457999,
+      "learning_rate": 8.450622252012353e-06,
+      "loss": 0.1974,
+      "step": 5031
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.260315229600693,
+      "learning_rate": 8.44997026580073e-06,
+      "loss": 0.1839,
+      "step": 5032
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.2501820461509139,
+      "learning_rate": 8.449318167601086e-06,
+      "loss": 0.2165,
+      "step": 5033
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.3752663094379212,
+      "learning_rate": 8.448665957434592e-06,
+      "loss": 0.1797,
+      "step": 5034
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.7752185193329555,
+      "learning_rate": 8.448013635322419e-06,
+      "loss": 0.1692,
+      "step": 5035
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.0347304314071502,
+      "learning_rate": 8.447361201285738e-06,
+      "loss": 0.2439,
+      "step": 5036
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.2498675633844551,
+      "learning_rate": 8.446708655345732e-06,
+      "loss": 0.2583,
+      "step": 5037
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.9813342387416893,
+      "learning_rate": 8.44605599752358e-06,
+      "loss": 0.184,
+      "step": 5038
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.4223082211251936,
+      "learning_rate": 8.445403227840469e-06,
+      "loss": 0.2363,
+      "step": 5039
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.5005078840480401,
+      "learning_rate": 8.44475034631759e-06,
+      "loss": 0.1969,
+      "step": 5040
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.8617161376374006,
+      "learning_rate": 8.444097352976132e-06,
+      "loss": 0.2552,
+      "step": 5041
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.3027614929545186,
+      "learning_rate": 8.443444247837293e-06,
+      "loss": 0.2142,
+      "step": 5042
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.072944617215986,
+      "learning_rate": 8.442791030922275e-06,
+      "loss": 0.2139,
+      "step": 5043
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.0489802441132459,
+      "learning_rate": 8.44213770225228e-06,
+      "loss": 0.1893,
+      "step": 5044
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.2259087061551466,
+      "learning_rate": 8.441484261848514e-06,
+      "loss": 0.2194,
+      "step": 5045
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.0267972793634903,
+      "learning_rate": 8.440830709732192e-06,
+      "loss": 0.1886,
+      "step": 5046
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.5711281932942067,
+      "learning_rate": 8.440177045924525e-06,
+      "loss": 0.2338,
+      "step": 5047
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.3578442677928841,
+      "learning_rate": 8.439523270446735e-06,
+      "loss": 0.2482,
+      "step": 5048
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.0183379453797459,
+      "learning_rate": 8.438869383320039e-06,
+      "loss": 0.1938,
+      "step": 5049
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.6354858095036113,
+      "learning_rate": 8.438215384565666e-06,
+      "loss": 0.2166,
+      "step": 5050
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.9368239597814129,
+      "learning_rate": 8.437561274204847e-06,
+      "loss": 0.1697,
+      "step": 5051
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.1979258896292972,
+      "learning_rate": 8.436907052258809e-06,
+      "loss": 0.1799,
+      "step": 5052
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.4627450659215064,
+      "learning_rate": 8.436252718748794e-06,
+      "loss": 0.2195,
+      "step": 5053
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.1125874155839417,
+      "learning_rate": 8.43559827369604e-06,
+      "loss": 0.1699,
+      "step": 5054
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.9788727479820374,
+      "learning_rate": 8.43494371712179e-06,
+      "loss": 0.1968,
+      "step": 5055
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.3317712127268277,
+      "learning_rate": 8.43428904904729e-06,
+      "loss": 0.239,
+      "step": 5056
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.1489413479627097,
+      "learning_rate": 8.433634269493793e-06,
+      "loss": 0.2235,
+      "step": 5057
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.5805324000441079,
+      "learning_rate": 8.432979378482555e-06,
+      "loss": 0.2782,
+      "step": 5058
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.2962461082798742,
+      "learning_rate": 8.43232437603483e-06,
+      "loss": 0.2261,
+      "step": 5059
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.112356856196981,
+      "learning_rate": 8.431669262171884e-06,
+      "loss": 0.2127,
+      "step": 5060
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.7614187383381051,
+      "learning_rate": 8.43101403691498e-06,
+      "loss": 0.1785,
+      "step": 5061
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.2763269136433242,
+      "learning_rate": 8.430358700285387e-06,
+      "loss": 0.1975,
+      "step": 5062
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.5039277756376634,
+      "learning_rate": 8.429703252304377e-06,
+      "loss": 0.2122,
+      "step": 5063
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.3793779679375795,
+      "learning_rate": 8.42904769299323e-06,
+      "loss": 0.2321,
+      "step": 5064
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.847372800925885,
+      "learning_rate": 8.42839202237322e-06,
+      "loss": 0.1954,
+      "step": 5065
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.2204775994825752,
+      "learning_rate": 8.427736240465634e-06,
+      "loss": 0.2136,
+      "step": 5066
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.077618175739415,
+      "learning_rate": 8.42708034729176e-06,
+      "loss": 0.1996,
+      "step": 5067
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.3189234311065188,
+      "learning_rate": 8.426424342872886e-06,
+      "loss": 0.181,
+      "step": 5068
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.194030668120899,
+      "learning_rate": 8.425768227230309e-06,
+      "loss": 0.1451,
+      "step": 5069
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.1467909143179793,
+      "learning_rate": 8.425112000385323e-06,
+      "loss": 0.2248,
+      "step": 5070
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.0415964760576564,
+      "learning_rate": 8.424455662359235e-06,
+      "loss": 0.201,
+      "step": 5071
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.4904960212648017,
+      "learning_rate": 8.423799213173345e-06,
+      "loss": 0.2107,
+      "step": 5072
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.0400820870969796,
+      "learning_rate": 8.423142652848964e-06,
+      "loss": 0.1939,
+      "step": 5073
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.9382986124291985,
+      "learning_rate": 8.422485981407403e-06,
+      "loss": 0.2149,
+      "step": 5074
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.009562172121663,
+      "learning_rate": 8.421829198869979e-06,
+      "loss": 0.1668,
+      "step": 5075
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.72253701279603,
+      "learning_rate": 8.421172305258014e-06,
+      "loss": 0.223,
+      "step": 5076
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.0415452125395386,
+      "learning_rate": 8.420515300592826e-06,
+      "loss": 0.1569,
+      "step": 5077
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.521700313655021,
+      "learning_rate": 8.419858184895746e-06,
+      "loss": 0.243,
+      "step": 5078
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.1375578482100326,
+      "learning_rate": 8.419200958188102e-06,
+      "loss": 0.1975,
+      "step": 5079
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.2496615339044075,
+      "learning_rate": 8.418543620491231e-06,
+      "loss": 0.2186,
+      "step": 5080
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.223784088014582,
+      "learning_rate": 8.417886171826466e-06,
+      "loss": 0.1881,
+      "step": 5081
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.2283142187046527,
+      "learning_rate": 8.417228612215151e-06,
+      "loss": 0.2062,
+      "step": 5082
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.7777993147599636,
+      "learning_rate": 8.41657094167863e-06,
+      "loss": 0.2358,
+      "step": 5083
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.7098881800241394,
+      "learning_rate": 8.41591316023825e-06,
+      "loss": 0.1883,
+      "step": 5084
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.4881885930251537,
+      "learning_rate": 8.415255267915368e-06,
+      "loss": 0.195,
+      "step": 5085
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.273275461594516,
+      "learning_rate": 8.414597264731333e-06,
+      "loss": 0.198,
+      "step": 5086
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.3380273512427476,
+      "learning_rate": 8.413939150707508e-06,
+      "loss": 0.1523,
+      "step": 5087
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.4364426834571606,
+      "learning_rate": 8.413280925865255e-06,
+      "loss": 0.2418,
+      "step": 5088
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.0287684798058307,
+      "learning_rate": 8.41262259022594e-06,
+      "loss": 0.1764,
+      "step": 5089
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.5515003290810698,
+      "learning_rate": 8.411964143810936e-06,
+      "loss": 0.3076,
+      "step": 5090
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.3974727747407174,
+      "learning_rate": 8.411305586641611e-06,
+      "loss": 0.1626,
+      "step": 5091
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.7087651553559762,
+      "learning_rate": 8.410646918739348e-06,
+      "loss": 0.2529,
+      "step": 5092
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.349318071275843,
+      "learning_rate": 8.40998814012552e-06,
+      "loss": 0.2401,
+      "step": 5093
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.9627310396229704,
+      "learning_rate": 8.409329250821521e-06,
+      "loss": 0.1922,
+      "step": 5094
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.295046404781302,
+      "learning_rate": 8.40867025084873e-06,
+      "loss": 0.259,
+      "step": 5095
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.9665869600538459,
+      "learning_rate": 8.408011140228544e-06,
+      "loss": 0.1845,
+      "step": 5096
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.5736851225547568,
+      "learning_rate": 8.407351918982358e-06,
+      "loss": 0.2856,
+      "step": 5097
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.9644723739263624,
+      "learning_rate": 8.406692587131569e-06,
+      "loss": 0.2296,
+      "step": 5098
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.221332075214751,
+      "learning_rate": 8.406033144697579e-06,
+      "loss": 0.2409,
+      "step": 5099
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.9742220202485163,
+      "learning_rate": 8.405373591701794e-06,
+      "loss": 0.1825,
+      "step": 5100
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.8186213476899309,
+      "learning_rate": 8.404713928165624e-06,
+      "loss": 0.2129,
+      "step": 5101
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.1807673072319866,
+      "learning_rate": 8.404054154110483e-06,
+      "loss": 0.1577,
+      "step": 5102
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.1128987508512629,
+      "learning_rate": 8.403394269557787e-06,
+      "loss": 0.2089,
+      "step": 5103
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.9579372943896122,
+      "learning_rate": 8.402734274528955e-06,
+      "loss": 0.193,
+      "step": 5104
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.1943078618714904,
+      "learning_rate": 8.40207416904541e-06,
+      "loss": 0.2594,
+      "step": 5105
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.4104282507803398,
+      "learning_rate": 8.401413953128584e-06,
+      "loss": 0.2305,
+      "step": 5106
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.1322237293417878,
+      "learning_rate": 8.400753626799904e-06,
+      "loss": 0.2351,
+      "step": 5107
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.007754529350297,
+      "learning_rate": 8.400093190080807e-06,
+      "loss": 0.1892,
+      "step": 5108
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.7986565489587255,
+      "learning_rate": 8.39943264299273e-06,
+      "loss": 0.1723,
+      "step": 5109
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.4789083380062238,
+      "learning_rate": 8.39877198555711e-06,
+      "loss": 0.2171,
+      "step": 5110
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.8690973764010381,
+      "learning_rate": 8.398111217795402e-06,
+      "loss": 0.1784,
+      "step": 5111
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 2.6797393037554254,
+      "learning_rate": 8.397450339729047e-06,
+      "loss": 0.2106,
+      "step": 5112
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.2354149466677373,
+      "learning_rate": 8.396789351379501e-06,
+      "loss": 0.1989,
+      "step": 5113
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.2981273395441946,
+      "learning_rate": 8.39612825276822e-06,
+      "loss": 0.1712,
+      "step": 5114
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.4446328266178023,
+      "learning_rate": 8.395467043916664e-06,
+      "loss": 0.2218,
+      "step": 5115
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.9903607410105938,
+      "learning_rate": 8.394805724846293e-06,
+      "loss": 0.2147,
+      "step": 5116
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.2950684304825772,
+      "learning_rate": 8.394144295578576e-06,
+      "loss": 0.2451,
+      "step": 5117
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.0516572114299445,
+      "learning_rate": 8.393482756134985e-06,
+      "loss": 0.1718,
+      "step": 5118
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.0801608867047732,
+      "learning_rate": 8.392821106536993e-06,
+      "loss": 0.1966,
+      "step": 5119
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.9858493624979163,
+      "learning_rate": 8.392159346806075e-06,
+      "loss": 0.198,
+      "step": 5120
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.123671372970557,
+      "learning_rate": 8.391497476963715e-06,
+      "loss": 0.2842,
+      "step": 5121
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.3465145914457202,
+      "learning_rate": 8.390835497031395e-06,
+      "loss": 0.1958,
+      "step": 5122
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.214055547731149,
+      "learning_rate": 8.390173407030606e-06,
+      "loss": 0.2304,
+      "step": 5123
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.9936677261399091,
+      "learning_rate": 8.38951120698284e-06,
+      "loss": 0.1969,
+      "step": 5124
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.003847417332076,
+      "learning_rate": 8.38884889690959e-06,
+      "loss": 0.1569,
+      "step": 5125
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.080198351978855,
+      "learning_rate": 8.388186476832355e-06,
+      "loss": 0.2008,
+      "step": 5126
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.9637377129258846,
+      "learning_rate": 8.38752394677264e-06,
+      "loss": 0.1943,
+      "step": 5127
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.225499661513079,
+      "learning_rate": 8.386861306751948e-06,
+      "loss": 0.188,
+      "step": 5128
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.1579608760386393,
+      "learning_rate": 8.386198556791791e-06,
+      "loss": 0.2411,
+      "step": 5129
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.5625178240713105,
+      "learning_rate": 8.385535696913683e-06,
+      "loss": 0.2608,
+      "step": 5130
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.245712371389931,
+      "learning_rate": 8.38487272713914e-06,
+      "loss": 0.2431,
+      "step": 5131
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.5580117972443634,
+      "learning_rate": 8.38420964748968e-06,
+      "loss": 0.1955,
+      "step": 5132
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.9790610641574894,
+      "learning_rate": 8.383546457986828e-06,
+      "loss": 0.2329,
+      "step": 5133
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.8988888850558051,
+      "learning_rate": 8.382883158652114e-06,
+      "loss": 0.2216,
+      "step": 5134
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.1382460738633189,
+      "learning_rate": 8.382219749507067e-06,
+      "loss": 0.1866,
+      "step": 5135
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.0040812729434903,
+      "learning_rate": 8.381556230573223e-06,
+      "loss": 0.1907,
+      "step": 5136
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.0028933190694929,
+      "learning_rate": 8.380892601872118e-06,
+      "loss": 0.142,
+      "step": 5137
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.9452587528160868,
+      "learning_rate": 8.380228863425294e-06,
+      "loss": 0.1703,
+      "step": 5138
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.4419723409655905,
+      "learning_rate": 8.3795650152543e-06,
+      "loss": 0.2198,
+      "step": 5139
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 2.1642014290183496,
+      "learning_rate": 8.378901057380681e-06,
+      "loss": 0.242,
+      "step": 5140
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.5330718366939768,
+      "learning_rate": 8.378236989825991e-06,
+      "loss": 0.2422,
+      "step": 5141
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.1219417981487845,
+      "learning_rate": 8.377572812611786e-06,
+      "loss": 0.1827,
+      "step": 5142
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.932812811022376,
+      "learning_rate": 8.376908525759623e-06,
+      "loss": 0.2221,
+      "step": 5143
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.7747726959774122,
+      "learning_rate": 8.376244129291068e-06,
+      "loss": 0.2346,
+      "step": 5144
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.9018488575476696,
+      "learning_rate": 8.375579623227689e-06,
+      "loss": 0.1688,
+      "step": 5145
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.0309867098324126,
+      "learning_rate": 8.374915007591053e-06,
+      "loss": 0.1914,
+      "step": 5146
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.4512680927996178,
+      "learning_rate": 8.374250282402737e-06,
+      "loss": 0.2694,
+      "step": 5147
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.1906523985163295,
+      "learning_rate": 8.373585447684314e-06,
+      "loss": 0.1656,
+      "step": 5148
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.3576725392149729,
+      "learning_rate": 8.372920503457368e-06,
+      "loss": 0.1876,
+      "step": 5149
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.4004920134026817,
+      "learning_rate": 8.372255449743483e-06,
+      "loss": 0.2122,
+      "step": 5150
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.317579519122393,
+      "learning_rate": 8.371590286564247e-06,
+      "loss": 0.2326,
+      "step": 5151
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.7792708218074846,
+      "learning_rate": 8.370925013941252e-06,
+      "loss": 0.1797,
+      "step": 5152
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.2445916955592669,
+      "learning_rate": 8.370259631896092e-06,
+      "loss": 0.2134,
+      "step": 5153
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.0890018533816999,
+      "learning_rate": 8.369594140450367e-06,
+      "loss": 0.2042,
+      "step": 5154
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.5058042018891162,
+      "learning_rate": 8.368928539625679e-06,
+      "loss": 0.2231,
+      "step": 5155
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.1561195712370915,
+      "learning_rate": 8.368262829443632e-06,
+      "loss": 0.197,
+      "step": 5156
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.3207963502707343,
+      "learning_rate": 8.367597009925836e-06,
+      "loss": 0.198,
+      "step": 5157
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.5319192893101559,
+      "learning_rate": 8.366931081093905e-06,
+      "loss": 0.1738,
+      "step": 5158
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.001995979939299,
+      "learning_rate": 8.366265042969457e-06,
+      "loss": 0.192,
+      "step": 5159
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.2202320868535823,
+      "learning_rate": 8.365598895574108e-06,
+      "loss": 0.1851,
+      "step": 5160
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.4658773844445732,
+      "learning_rate": 8.364932638929484e-06,
+      "loss": 0.1979,
+      "step": 5161
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.2747802931630243,
+      "learning_rate": 8.364266273057212e-06,
+      "loss": 0.2204,
+      "step": 5162
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.4439347105128504,
+      "learning_rate": 8.363599797978923e-06,
+      "loss": 0.1811,
+      "step": 5163
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.3468685667764981,
+      "learning_rate": 8.36293321371625e-06,
+      "loss": 0.1835,
+      "step": 5164
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.7905168067491009,
+      "learning_rate": 8.362266520290828e-06,
+      "loss": 0.1916,
+      "step": 5165
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.2869130927015788,
+      "learning_rate": 8.361599717724305e-06,
+      "loss": 0.1848,
+      "step": 5166
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.4509863754148087,
+      "learning_rate": 8.36093280603832e-06,
+      "loss": 0.2625,
+      "step": 5167
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.3272369739790895,
+      "learning_rate": 8.360265785254524e-06,
+      "loss": 0.2455,
+      "step": 5168
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.1881708960420188,
+      "learning_rate": 8.359598655394569e-06,
+      "loss": 0.2806,
+      "step": 5169
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.1838432810655353,
+      "learning_rate": 8.358931416480109e-06,
+      "loss": 0.1566,
+      "step": 5170
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.7888038236033798,
+      "learning_rate": 8.358264068532803e-06,
+      "loss": 0.1389,
+      "step": 5171
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.4697434231117337,
+      "learning_rate": 8.357596611574315e-06,
+      "loss": 0.2124,
+      "step": 5172
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.2210638916968153,
+      "learning_rate": 8.356929045626311e-06,
+      "loss": 0.207,
+      "step": 5173
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.7885908146368665,
+      "learning_rate": 8.35626137071046e-06,
+      "loss": 0.2172,
+      "step": 5174
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.9829527319353561,
+      "learning_rate": 8.355593586848433e-06,
+      "loss": 0.1495,
+      "step": 5175
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.2424510340377322,
+      "learning_rate": 8.35492569406191e-06,
+      "loss": 0.1791,
+      "step": 5176
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.0439900726771496,
+      "learning_rate": 8.35425769237257e-06,
+      "loss": 0.2212,
+      "step": 5177
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.6711847729842876,
+      "learning_rate": 8.353589581802095e-06,
+      "loss": 0.1869,
+      "step": 5178
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.6336729164589048,
+      "learning_rate": 8.352921362372176e-06,
+      "loss": 0.2396,
+      "step": 5179
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.4080371944109737,
+      "learning_rate": 8.352253034104499e-06,
+      "loss": 0.2848,
+      "step": 5180
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.9386101735775215,
+      "learning_rate": 8.35158459702076e-06,
+      "loss": 0.1918,
+      "step": 5181
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.439269684343868,
+      "learning_rate": 8.35091605114266e-06,
+      "loss": 0.2261,
+      "step": 5182
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.9004219343488684,
+      "learning_rate": 8.350247396491898e-06,
+      "loss": 0.2064,
+      "step": 5183
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.210206845028072,
+      "learning_rate": 8.349578633090179e-06,
+      "loss": 0.2052,
+      "step": 5184
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.1653050920727186,
+      "learning_rate": 8.348909760959211e-06,
+      "loss": 0.1684,
+      "step": 5185
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.0661852156902367,
+      "learning_rate": 8.348240780120707e-06,
+      "loss": 0.2006,
+      "step": 5186
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.3076745049903398,
+      "learning_rate": 8.34757169059638e-06,
+      "loss": 0.2531,
+      "step": 5187
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.2201152789896497,
+      "learning_rate": 8.346902492407954e-06,
+      "loss": 0.1581,
+      "step": 5188
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.7282985512144003,
+      "learning_rate": 8.346233185577146e-06,
+      "loss": 0.202,
+      "step": 5189
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.3248744012407223,
+      "learning_rate": 8.345563770125683e-06,
+      "loss": 0.2288,
+      "step": 5190
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.3103103111507186,
+      "learning_rate": 8.3448942460753e-06,
+      "loss": 0.2105,
+      "step": 5191
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.3879310252018469,
+      "learning_rate": 8.344224613447725e-06,
+      "loss": 0.1701,
+      "step": 5192
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.2978670766576281,
+      "learning_rate": 8.343554872264697e-06,
+      "loss": 0.2118,
+      "step": 5193
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.118409755316829,
+      "learning_rate": 8.342885022547953e-06,
+      "loss": 0.2122,
+      "step": 5194
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.1635599555687337,
+      "learning_rate": 8.34221506431924e-06,
+      "loss": 0.1515,
+      "step": 5195
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.1977215780816546,
+      "learning_rate": 8.341544997600304e-06,
+      "loss": 0.2625,
+      "step": 5196
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.0952061607298957,
+      "learning_rate": 8.340874822412897e-06,
+      "loss": 0.1577,
+      "step": 5197
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.4249747150670826,
+      "learning_rate": 8.340204538778768e-06,
+      "loss": 0.1751,
+      "step": 5198
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.9799319693857397,
+      "learning_rate": 8.339534146719683e-06,
+      "loss": 0.149,
+      "step": 5199
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.2946110865965101,
+      "learning_rate": 8.338863646257398e-06,
+      "loss": 0.2054,
+      "step": 5200
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.4512933523642297,
+      "learning_rate": 8.338193037413679e-06,
+      "loss": 0.2288,
+      "step": 5201
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.3499269491372192,
+      "learning_rate": 8.337522320210297e-06,
+      "loss": 0.2074,
+      "step": 5202
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.0824071140101934,
+      "learning_rate": 8.336851494669016e-06,
+      "loss": 0.1905,
+      "step": 5203
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.4516301952340704,
+      "learning_rate": 8.336180560811619e-06,
+      "loss": 0.1907,
+      "step": 5204
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.0639031030353734,
+      "learning_rate": 8.335509518659884e-06,
+      "loss": 0.2144,
+      "step": 5205
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.178544706745058,
+      "learning_rate": 8.334838368235591e-06,
+      "loss": 0.2533,
+      "step": 5206
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.1183376560640481,
+      "learning_rate": 8.334167109560527e-06,
+      "loss": 0.2193,
+      "step": 5207
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.1186787199643096,
+      "learning_rate": 8.333495742656481e-06,
+      "loss": 0.202,
+      "step": 5208
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.2062960363902755,
+      "learning_rate": 8.332824267545248e-06,
+      "loss": 0.1998,
+      "step": 5209
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.2677643702301336,
+      "learning_rate": 8.33215268424862e-06,
+      "loss": 0.1865,
+      "step": 5210
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.36415865680471,
+      "learning_rate": 8.331480992788402e-06,
+      "loss": 0.2407,
+      "step": 5211
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.1546349490264551,
+      "learning_rate": 8.330809193186394e-06,
+      "loss": 0.1813,
+      "step": 5212
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.8539003087977473,
+      "learning_rate": 8.330137285464404e-06,
+      "loss": 0.2087,
+      "step": 5213
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.8541552157138456,
+      "learning_rate": 8.329465269644243e-06,
+      "loss": 0.2378,
+      "step": 5214
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.1567942073211857,
+      "learning_rate": 8.328793145747727e-06,
+      "loss": 0.146,
+      "step": 5215
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.6772026822664774,
+      "learning_rate": 8.328120913796667e-06,
+      "loss": 0.2509,
+      "step": 5216
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.3058141851750706,
+      "learning_rate": 8.32744857381289e-06,
+      "loss": 0.2283,
+      "step": 5217
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.9628587750396065,
+      "learning_rate": 8.32677612581822e-06,
+      "loss": 0.1781,
+      "step": 5218
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.1178867312667509,
+      "learning_rate": 8.326103569834484e-06,
+      "loss": 0.2093,
+      "step": 5219
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.346973589056966,
+      "learning_rate": 8.325430905883514e-06,
+      "loss": 0.1869,
+      "step": 5220
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.0938325765636452,
+      "learning_rate": 8.324758133987143e-06,
+      "loss": 0.1928,
+      "step": 5221
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.264201550710295,
+      "learning_rate": 8.32408525416721e-06,
+      "loss": 0.2151,
+      "step": 5222
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.9191591153957939,
+      "learning_rate": 8.32341226644556e-06,
+      "loss": 0.187,
+      "step": 5223
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.6106030584135282,
+      "learning_rate": 8.322739170844039e-06,
+      "loss": 0.1523,
+      "step": 5224
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.6836988075454011,
+      "learning_rate": 8.32206596738449e-06,
+      "loss": 0.1118,
+      "step": 5225
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.238317675611323,
+      "learning_rate": 8.32139265608877e-06,
+      "loss": 0.2441,
+      "step": 5226
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.5512557372817823,
+      "learning_rate": 8.320719236978736e-06,
+      "loss": 0.1872,
+      "step": 5227
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.4768066191834874,
+      "learning_rate": 8.320045710076244e-06,
+      "loss": 0.2145,
+      "step": 5228
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.1993644731436337,
+      "learning_rate": 8.31937207540316e-06,
+      "loss": 0.2431,
+      "step": 5229
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.9118751698965081,
+      "learning_rate": 8.31869833298135e-06,
+      "loss": 0.2112,
+      "step": 5230
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.1555946738651708,
+      "learning_rate": 8.318024482832685e-06,
+      "loss": 0.1601,
+      "step": 5231
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.2790434695565611,
+      "learning_rate": 8.317350524979036e-06,
+      "loss": 0.1978,
+      "step": 5232
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.096174787977371,
+      "learning_rate": 8.31667645944228e-06,
+      "loss": 0.232,
+      "step": 5233
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.2147886134941048,
+      "learning_rate": 8.3160022862443e-06,
+      "loss": 0.2038,
+      "step": 5234
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.6495556045068396,
+      "learning_rate": 8.31532800540698e-06,
+      "loss": 0.1936,
+      "step": 5235
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.0807687395005146,
+      "learning_rate": 8.314653616952204e-06,
+      "loss": 0.2301,
+      "step": 5236
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.3900139288474547,
+      "learning_rate": 8.313979120901867e-06,
+      "loss": 0.1773,
+      "step": 5237
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.2548680820416895,
+      "learning_rate": 8.313304517277862e-06,
+      "loss": 0.1576,
+      "step": 5238
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.1630965832111764,
+      "learning_rate": 8.312629806102088e-06,
+      "loss": 0.1723,
+      "step": 5239
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.0515362710784981,
+      "learning_rate": 8.311954987396444e-06,
+      "loss": 0.1504,
+      "step": 5240
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.1891581274445877,
+      "learning_rate": 8.311280061182838e-06,
+      "loss": 0.2074,
+      "step": 5241
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.6276904624314308,
+      "learning_rate": 8.310605027483177e-06,
+      "loss": 0.1884,
+      "step": 5242
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.9949954494699937,
+      "learning_rate": 8.309929886319373e-06,
+      "loss": 0.1978,
+      "step": 5243
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.541976011794168,
+      "learning_rate": 8.30925463771334e-06,
+      "loss": 0.2319,
+      "step": 5244
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.1924898504129173,
+      "learning_rate": 8.308579281687e-06,
+      "loss": 0.1612,
+      "step": 5245
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.1013026573568734,
+      "learning_rate": 8.307903818262274e-06,
+      "loss": 0.2187,
+      "step": 5246
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.3604843994919413,
+      "learning_rate": 8.307228247461088e-06,
+      "loss": 0.2054,
+      "step": 5247
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.4215246872035414,
+      "learning_rate": 8.30655256930537e-06,
+      "loss": 0.2326,
+      "step": 5248
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.1162920327485295,
+      "learning_rate": 8.305876783817054e-06,
+      "loss": 0.1704,
+      "step": 5249
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.9765886894529027,
+      "learning_rate": 8.305200891018079e-06,
+      "loss": 0.1491,
+      "step": 5250
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.9318741959453878,
+      "learning_rate": 8.304524890930381e-06,
+      "loss": 0.2324,
+      "step": 5251
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.319974675495522,
+      "learning_rate": 8.303848783575903e-06,
+      "loss": 0.19,
+      "step": 5252
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.0359041355335532,
+      "learning_rate": 8.303172568976597e-06,
+      "loss": 0.1576,
+      "step": 5253
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.4898269516348421,
+      "learning_rate": 8.302496247154408e-06,
+      "loss": 0.2571,
+      "step": 5254
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.144970436914444,
+      "learning_rate": 8.301819818131291e-06,
+      "loss": 0.1872,
+      "step": 5255
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.1375284326322015,
+      "learning_rate": 8.301143281929208e-06,
+      "loss": 0.1449,
+      "step": 5256
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.2073454833632278,
+      "learning_rate": 8.30046663857011e-06,
+      "loss": 0.2696,
+      "step": 5257
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.42531704335073,
+      "learning_rate": 8.29978988807597e-06,
+      "loss": 0.2426,
+      "step": 5258
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 2.180403438015034,
+      "learning_rate": 8.299113030468755e-06,
+      "loss": 0.2507,
+      "step": 5259
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.1538188141230112,
+      "learning_rate": 8.298436065770431e-06,
+      "loss": 0.2267,
+      "step": 5260
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.2708560486884417,
+      "learning_rate": 8.297758994002976e-06,
+      "loss": 0.2479,
+      "step": 5261
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.035948313574645,
+      "learning_rate": 8.297081815188368e-06,
+      "loss": 0.2065,
+      "step": 5262
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.321297087296098,
+      "learning_rate": 8.296404529348587e-06,
+      "loss": 0.2399,
+      "step": 5263
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.5827613605774833,
+      "learning_rate": 8.29572713650562e-06,
+      "loss": 0.2144,
+      "step": 5264
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.7651078006807908,
+      "learning_rate": 8.295049636681458e-06,
+      "loss": 0.2782,
+      "step": 5265
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.1669261418890509,
+      "learning_rate": 8.294372029898086e-06,
+      "loss": 0.1629,
+      "step": 5266
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.865438472106846,
+      "learning_rate": 8.293694316177506e-06,
+      "loss": 0.2688,
+      "step": 5267
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.5822977500904896,
+      "learning_rate": 8.293016495541712e-06,
+      "loss": 0.2692,
+      "step": 5268
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.0651989301443614,
+      "learning_rate": 8.29233856801271e-06,
+      "loss": 0.2322,
+      "step": 5269
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.444871634181987,
+      "learning_rate": 8.291660533612506e-06,
+      "loss": 0.1706,
+      "step": 5270
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.9902378172504254,
+      "learning_rate": 8.290982392363106e-06,
+      "loss": 0.1839,
+      "step": 5271
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.1935530784719686,
+      "learning_rate": 8.290304144286528e-06,
+      "loss": 0.1915,
+      "step": 5272
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.0566302931839269,
+      "learning_rate": 8.289625789404786e-06,
+      "loss": 0.2294,
+      "step": 5273
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.051310307677124,
+      "learning_rate": 8.288947327739895e-06,
+      "loss": 0.2255,
+      "step": 5274
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.0657249980234182,
+      "learning_rate": 8.288268759313886e-06,
+      "loss": 0.2156,
+      "step": 5275
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.179222404457385,
+      "learning_rate": 8.28759008414878e-06,
+      "loss": 0.1996,
+      "step": 5276
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.1078406648794936,
+      "learning_rate": 8.286911302266611e-06,
+      "loss": 0.1901,
+      "step": 5277
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.2635414949515535,
+      "learning_rate": 8.28623241368941e-06,
+      "loss": 0.2199,
+      "step": 5278
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.8674755826127727,
+      "learning_rate": 8.285553418439216e-06,
+      "loss": 0.2125,
+      "step": 5279
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.3176940955993628,
+      "learning_rate": 8.284874316538067e-06,
+      "loss": 0.2635,
+      "step": 5280
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.2185091688554106,
+      "learning_rate": 8.284195108008011e-06,
+      "loss": 0.2222,
+      "step": 5281
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.2660091708947732,
+      "learning_rate": 8.283515792871091e-06,
+      "loss": 0.2611,
+      "step": 5282
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.3547060576931058,
+      "learning_rate": 8.282836371149361e-06,
+      "loss": 0.182,
+      "step": 5283
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.8330835279517614,
+      "learning_rate": 8.282156842864876e-06,
+      "loss": 0.2271,
+      "step": 5284
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.1811874210184405,
+      "learning_rate": 8.281477208039692e-06,
+      "loss": 0.211,
+      "step": 5285
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.0248888940823728,
+      "learning_rate": 8.28079746669587e-06,
+      "loss": 0.1763,
+      "step": 5286
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.0455889310675295,
+      "learning_rate": 8.280117618855475e-06,
+      "loss": 0.232,
+      "step": 5287
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.0765273040324992,
+      "learning_rate": 8.279437664540577e-06,
+      "loss": 0.2104,
+      "step": 5288
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.360645692332413,
+      "learning_rate": 8.278757603773247e-06,
+      "loss": 0.1999,
+      "step": 5289
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.9341332112004002,
+      "learning_rate": 8.278077436575558e-06,
+      "loss": 0.1871,
+      "step": 5290
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.1576402512246884,
+      "learning_rate": 8.277397162969593e-06,
+      "loss": 0.1675,
+      "step": 5291
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.10781258968439,
+      "learning_rate": 8.27671678297743e-06,
+      "loss": 0.206,
+      "step": 5292
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.3801532704388002,
+      "learning_rate": 8.276036296621155e-06,
+      "loss": 0.1677,
+      "step": 5293
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.0867678508800196,
+      "learning_rate": 8.27535570392286e-06,
+      "loss": 0.2308,
+      "step": 5294
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.3924365520072977,
+      "learning_rate": 8.274675004904635e-06,
+      "loss": 0.2015,
+      "step": 5295
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.0466488079634715,
+      "learning_rate": 8.273994199588575e-06,
+      "loss": 0.2476,
+      "step": 5296
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.3212703145255513,
+      "learning_rate": 8.27331328799678e-06,
+      "loss": 0.2329,
+      "step": 5297
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.9601869569295941,
+      "learning_rate": 8.272632270151355e-06,
+      "loss": 0.1967,
+      "step": 5298
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.985772506091909,
+      "learning_rate": 8.271951146074405e-06,
+      "loss": 0.2097,
+      "step": 5299
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.579767095181114,
+      "learning_rate": 8.271269915788038e-06,
+      "loss": 0.2533,
+      "step": 5300
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.1801778211721987,
+      "learning_rate": 8.27058857931437e-06,
+      "loss": 0.1974,
+      "step": 5301
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.31058952529342,
+      "learning_rate": 8.269907136675515e-06,
+      "loss": 0.2655,
+      "step": 5302
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.3943817081198153,
+      "learning_rate": 8.269225587893594e-06,
+      "loss": 0.1982,
+      "step": 5303
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.9891265351800167,
+      "learning_rate": 8.26854393299073e-06,
+      "loss": 0.2164,
+      "step": 5304
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.0830916431615303,
+      "learning_rate": 8.26786217198905e-06,
+      "loss": 0.1985,
+      "step": 5305
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.1229876079677061,
+      "learning_rate": 8.267180304910687e-06,
+      "loss": 0.1947,
+      "step": 5306
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.1573641956748744,
+      "learning_rate": 8.26649833177777e-06,
+      "loss": 0.217,
+      "step": 5307
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.1516618270751924,
+      "learning_rate": 8.26581625261244e-06,
+      "loss": 0.2153,
+      "step": 5308
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.8743260674830683,
+      "learning_rate": 8.265134067436836e-06,
+      "loss": 0.1638,
+      "step": 5309
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.153127820572337,
+      "learning_rate": 8.264451776273104e-06,
+      "loss": 0.1985,
+      "step": 5310
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.788513376115022,
+      "learning_rate": 8.263769379143388e-06,
+      "loss": 0.1872,
+      "step": 5311
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.2303351980726358,
+      "learning_rate": 8.263086876069842e-06,
+      "loss": 0.2445,
+      "step": 5312
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 2.923951698703847,
+      "learning_rate": 8.26240426707462e-06,
+      "loss": 0.2494,
+      "step": 5313
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.1796244831811684,
+      "learning_rate": 8.261721552179879e-06,
+      "loss": 0.2743,
+      "step": 5314
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.5819326228728647,
+      "learning_rate": 8.261038731407783e-06,
+      "loss": 0.1997,
+      "step": 5315
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.577230809189457,
+      "learning_rate": 8.26035580478049e-06,
+      "loss": 0.2069,
+      "step": 5316
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 2.0357505622493144,
+      "learning_rate": 8.259672772320177e-06,
+      "loss": 0.2385,
+      "step": 5317
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.106898905879881,
+      "learning_rate": 8.258989634049008e-06,
+      "loss": 0.209,
+      "step": 5318
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.0806191729294168,
+      "learning_rate": 8.258306389989165e-06,
+      "loss": 0.2188,
+      "step": 5319
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.0940136813032035,
+      "learning_rate": 8.25762304016282e-06,
+      "loss": 0.2086,
+      "step": 5320
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.8104598883406745,
+      "learning_rate": 8.256939584592159e-06,
+      "loss": 0.1508,
+      "step": 5321
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.3055768208541665,
+      "learning_rate": 8.256256023299368e-06,
+      "loss": 0.2453,
+      "step": 5322
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.1275121370104253,
+      "learning_rate": 8.25557235630663e-06,
+      "loss": 0.1962,
+      "step": 5323
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.2143337430341155,
+      "learning_rate": 8.254888583636143e-06,
+      "loss": 0.1925,
+      "step": 5324
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.0383452829861743,
+      "learning_rate": 8.254204705310102e-06,
+      "loss": 0.189,
+      "step": 5325
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.0543418582819126,
+      "learning_rate": 8.253520721350705e-06,
+      "loss": 0.24,
+      "step": 5326
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.2955712959784207,
+      "learning_rate": 8.252836631780152e-06,
+      "loss": 0.1973,
+      "step": 5327
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.7442113321630643,
+      "learning_rate": 8.252152436620653e-06,
+      "loss": 0.2506,
+      "step": 5328
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.2214966796240607,
+      "learning_rate": 8.251468135894417e-06,
+      "loss": 0.1991,
+      "step": 5329
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.1347713043543504,
+      "learning_rate": 8.250783729623655e-06,
+      "loss": 0.1981,
+      "step": 5330
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.2374824766522288,
+      "learning_rate": 8.250099217830584e-06,
+      "loss": 0.1961,
+      "step": 5331
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.8984420610395011,
+      "learning_rate": 8.249414600537421e-06,
+      "loss": 0.1726,
+      "step": 5332
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.3818659600544914,
+      "learning_rate": 8.248729877766393e-06,
+      "loss": 0.1996,
+      "step": 5333
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.6467108157870116,
+      "learning_rate": 8.248045049539726e-06,
+      "loss": 0.2242,
+      "step": 5334
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.7865894117078478,
+      "learning_rate": 8.247360115879647e-06,
+      "loss": 0.1982,
+      "step": 5335
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.416897163911877,
+      "learning_rate": 8.246675076808392e-06,
+      "loss": 0.1809,
+      "step": 5336
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.422248290914185,
+      "learning_rate": 8.2459899323482e-06,
+      "loss": 0.1821,
+      "step": 5337
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.1573519964685277,
+      "learning_rate": 8.245304682521305e-06,
+      "loss": 0.2025,
+      "step": 5338
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.2757996272826777,
+      "learning_rate": 8.244619327349954e-06,
+      "loss": 0.1846,
+      "step": 5339
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.8009634694515515,
+      "learning_rate": 8.243933866856396e-06,
+      "loss": 0.1472,
+      "step": 5340
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.9388531099156416,
+      "learning_rate": 8.243248301062876e-06,
+      "loss": 0.1537,
+      "step": 5341
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.2830162696135727,
+      "learning_rate": 8.242562629991653e-06,
+      "loss": 0.2298,
+      "step": 5342
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.4386224057705914,
+      "learning_rate": 8.241876853664982e-06,
+      "loss": 0.2147,
+      "step": 5343
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.9203885508462841,
+      "learning_rate": 8.241190972105123e-06,
+      "loss": 0.1649,
+      "step": 5344
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.9528747761693714,
+      "learning_rate": 8.240504985334343e-06,
+      "loss": 0.2042,
+      "step": 5345
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.0676291121385344,
+      "learning_rate": 8.239818893374909e-06,
+      "loss": 0.2475,
+      "step": 5346
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.06184977381909,
+      "learning_rate": 8.239132696249088e-06,
+      "loss": 0.2384,
+      "step": 5347
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.1251608862673435,
+      "learning_rate": 8.238446393979159e-06,
+      "loss": 0.1631,
+      "step": 5348
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.6446158280902232,
+      "learning_rate": 8.237759986587397e-06,
+      "loss": 0.2605,
+      "step": 5349
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.2069473638674217,
+      "learning_rate": 8.237073474096083e-06,
+      "loss": 0.1739,
+      "step": 5350
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.2822789387054472,
+      "learning_rate": 8.236386856527504e-06,
+      "loss": 0.2622,
+      "step": 5351
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.402062033140273,
+      "learning_rate": 8.235700133903945e-06,
+      "loss": 0.1907,
+      "step": 5352
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.2906535271315942,
+      "learning_rate": 8.235013306247702e-06,
+      "loss": 0.2233,
+      "step": 5353
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.4143680800834615,
+      "learning_rate": 8.234326373581064e-06,
+      "loss": 0.2537,
+      "step": 5354
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.0518047602100864,
+      "learning_rate": 8.233639335926333e-06,
+      "loss": 0.2089,
+      "step": 5355
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.8474124105063353,
+      "learning_rate": 8.232952193305808e-06,
+      "loss": 0.1842,
+      "step": 5356
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.9317336205986742,
+      "learning_rate": 8.232264945741799e-06,
+      "loss": 0.1804,
+      "step": 5357
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.085466382177979,
+      "learning_rate": 8.231577593256609e-06,
+      "loss": 0.1843,
+      "step": 5358
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.9524448314923146,
+      "learning_rate": 8.230890135872552e-06,
+      "loss": 0.1901,
+      "step": 5359
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.2755276276346605,
+      "learning_rate": 8.230202573611944e-06,
+      "loss": 0.2376,
+      "step": 5360
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 2.2303553149878925,
+      "learning_rate": 8.229514906497103e-06,
+      "loss": 0.2027,
+      "step": 5361
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.053942246631789,
+      "learning_rate": 8.22882713455035e-06,
+      "loss": 0.1804,
+      "step": 5362
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.1691081603666817,
+      "learning_rate": 8.228139257794012e-06,
+      "loss": 0.2019,
+      "step": 5363
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.6873114268586382,
+      "learning_rate": 8.227451276250419e-06,
+      "loss": 0.2641,
+      "step": 5364
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.0051779874197233,
+      "learning_rate": 8.2267631899419e-06,
+      "loss": 0.1906,
+      "step": 5365
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.7898716932008094,
+      "learning_rate": 8.226074998890792e-06,
+      "loss": 0.1832,
+      "step": 5366
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.3211258916480992,
+      "learning_rate": 8.225386703119435e-06,
+      "loss": 0.2219,
+      "step": 5367
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.0436079445410609,
+      "learning_rate": 8.224698302650172e-06,
+      "loss": 0.2242,
+      "step": 5368
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.9409236759315585,
+      "learning_rate": 8.224009797505346e-06,
+      "loss": 0.1897,
+      "step": 5369
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.5668220705352298,
+      "learning_rate": 8.22332118770731e-06,
+      "loss": 0.2035,
+      "step": 5370
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.0064713366456857,
+      "learning_rate": 8.222632473278413e-06,
+      "loss": 0.1691,
+      "step": 5371
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.0159733486656377,
+      "learning_rate": 8.221943654241012e-06,
+      "loss": 0.1607,
+      "step": 5372
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.2322785464942954,
+      "learning_rate": 8.221254730617469e-06,
+      "loss": 0.2104,
+      "step": 5373
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.317918134247657,
+      "learning_rate": 8.220565702430144e-06,
+      "loss": 0.2035,
+      "step": 5374
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.1267981373075593,
+      "learning_rate": 8.219876569701406e-06,
+      "loss": 0.2245,
+      "step": 5375
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.7916487526539917,
+      "learning_rate": 8.219187332453622e-06,
+      "loss": 0.2167,
+      "step": 5376
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.0793677109685915,
+      "learning_rate": 8.218497990709164e-06,
+      "loss": 0.211,
+      "step": 5377
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.9584324699502451,
+      "learning_rate": 8.217808544490413e-06,
+      "loss": 0.1917,
+      "step": 5378
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.8360641764534672,
+      "learning_rate": 8.217118993819746e-06,
+      "loss": 0.238,
+      "step": 5379
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.3100843579875785,
+      "learning_rate": 8.216429338719547e-06,
+      "loss": 0.2334,
+      "step": 5380
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.5790650004328037,
+      "learning_rate": 8.2157395792122e-06,
+      "loss": 0.2627,
+      "step": 5381
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.2485520201712645,
+      "learning_rate": 8.215049715320099e-06,
+      "loss": 0.2044,
+      "step": 5382
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.42159881634298,
+      "learning_rate": 8.214359747065634e-06,
+      "loss": 0.1799,
+      "step": 5383
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.9765131861633377,
+      "learning_rate": 8.213669674471205e-06,
+      "loss": 0.1941,
+      "step": 5384
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.2327644881507431,
+      "learning_rate": 8.212979497559208e-06,
+      "loss": 0.1829,
+      "step": 5385
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.212152231745307,
+      "learning_rate": 8.212289216352051e-06,
+      "loss": 0.2265,
+      "step": 5386
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.0473769055883095,
+      "learning_rate": 8.211598830872139e-06,
+      "loss": 0.1849,
+      "step": 5387
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.243866763073377,
+      "learning_rate": 8.21090834114188e-06,
+      "loss": 0.2509,
+      "step": 5388
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.9735963139249995,
+      "learning_rate": 8.21021774718369e-06,
+      "loss": 0.2447,
+      "step": 5389
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.375939682086546,
+      "learning_rate": 8.209527049019988e-06,
+      "loss": 0.1987,
+      "step": 5390
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.880538430498774,
+      "learning_rate": 8.208836246673193e-06,
+      "loss": 0.209,
+      "step": 5391
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.1981869617224359,
+      "learning_rate": 8.208145340165724e-06,
+      "loss": 0.1878,
+      "step": 5392
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.1007809690416899,
+      "learning_rate": 8.207454329520016e-06,
+      "loss": 0.2131,
+      "step": 5393
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.2790363410471473,
+      "learning_rate": 8.206763214758495e-06,
+      "loss": 0.1536,
+      "step": 5394
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 2.4114585585350334,
+      "learning_rate": 8.206071995903596e-06,
+      "loss": 0.2133,
+      "step": 5395
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.1030907226429532,
+      "learning_rate": 8.205380672977756e-06,
+      "loss": 0.2332,
+      "step": 5396
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.9175030297981617,
+      "learning_rate": 8.204689246003416e-06,
+      "loss": 0.2296,
+      "step": 5397
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.1757848650854559,
+      "learning_rate": 8.203997715003023e-06,
+      "loss": 0.2397,
+      "step": 5398
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.3402068512207557,
+      "learning_rate": 8.203306079999019e-06,
+      "loss": 0.2334,
+      "step": 5399
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.9062910039849781,
+      "learning_rate": 8.202614341013859e-06,
+      "loss": 0.1973,
+      "step": 5400
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.2473076973069666,
+      "learning_rate": 8.201922498069995e-06,
+      "loss": 0.2355,
+      "step": 5401
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.3974416266279373,
+      "learning_rate": 8.201230551189884e-06,
+      "loss": 0.199,
+      "step": 5402
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.256386133250437,
+      "learning_rate": 8.200538500395992e-06,
+      "loss": 0.1943,
+      "step": 5403
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.8326751474049653,
+      "learning_rate": 8.199846345710776e-06,
+      "loss": 0.1899,
+      "step": 5404
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.2657710774032584,
+      "learning_rate": 8.19915408715671e-06,
+      "loss": 0.21,
+      "step": 5405
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.1000437462978145,
+      "learning_rate": 8.198461724756262e-06,
+      "loss": 0.2526,
+      "step": 5406
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.5577893248721457,
+      "learning_rate": 8.197769258531908e-06,
+      "loss": 0.2224,
+      "step": 5407
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.9803948327048012,
+      "learning_rate": 8.197076688506124e-06,
+      "loss": 0.1947,
+      "step": 5408
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.142879857056832,
+      "learning_rate": 8.196384014701393e-06,
+      "loss": 0.2066,
+      "step": 5409
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.9870822931461053,
+      "learning_rate": 8.195691237140198e-06,
+      "loss": 0.188,
+      "step": 5410
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.0512139730483019,
+      "learning_rate": 8.194998355845028e-06,
+      "loss": 0.1911,
+      "step": 5411
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.07580512386695,
+      "learning_rate": 8.194305370838374e-06,
+      "loss": 0.2091,
+      "step": 5412
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.4699614074240785,
+      "learning_rate": 8.193612282142731e-06,
+      "loss": 0.2565,
+      "step": 5413
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.10635244361905,
+      "learning_rate": 8.192919089780598e-06,
+      "loss": 0.1989,
+      "step": 5414
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.9470102654205845,
+      "learning_rate": 8.192225793774473e-06,
+      "loss": 0.1695,
+      "step": 5415
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.185146271959722,
+      "learning_rate": 8.191532394146865e-06,
+      "loss": 0.1681,
+      "step": 5416
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.0796873634354855,
+      "learning_rate": 8.190838890920279e-06,
+      "loss": 0.2114,
+      "step": 5417
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.9490416601001297,
+      "learning_rate": 8.190145284117229e-06,
+      "loss": 0.1883,
+      "step": 5418
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.25758944482443,
+      "learning_rate": 8.189451573760228e-06,
+      "loss": 0.2055,
+      "step": 5419
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.8481010300675831,
+      "learning_rate": 8.188757759871794e-06,
+      "loss": 0.1934,
+      "step": 5420
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.9902997550222483,
+      "learning_rate": 8.188063842474449e-06,
+      "loss": 0.2524,
+      "step": 5421
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.9416965041120963,
+      "learning_rate": 8.187369821590721e-06,
+      "loss": 0.1934,
+      "step": 5422
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.525738277211034,
+      "learning_rate": 8.186675697243133e-06,
+      "loss": 0.2076,
+      "step": 5423
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.2678371335279341,
+      "learning_rate": 8.185981469454222e-06,
+      "loss": 0.212,
+      "step": 5424
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.5220155290740027,
+      "learning_rate": 8.18528713824652e-06,
+      "loss": 0.1781,
+      "step": 5425
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.9929518271263982,
+      "learning_rate": 8.184592703642564e-06,
+      "loss": 0.1878,
+      "step": 5426
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.3190460306573917,
+      "learning_rate": 8.183898165664899e-06,
+      "loss": 0.159,
+      "step": 5427
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.1751562323466906,
+      "learning_rate": 8.183203524336067e-06,
+      "loss": 0.1839,
+      "step": 5428
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.173679418554201,
+      "learning_rate": 8.182508779678619e-06,
+      "loss": 0.19,
+      "step": 5429
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.9040611775596591,
+      "learning_rate": 8.181813931715108e-06,
+      "loss": 0.2193,
+      "step": 5430
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.4758999356001066,
+      "learning_rate": 8.181118980468083e-06,
+      "loss": 0.2074,
+      "step": 5431
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.2463816725674999,
+      "learning_rate": 8.180423925960111e-06,
+      "loss": 0.2067,
+      "step": 5432
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.1154750657774457,
+      "learning_rate": 8.179728768213747e-06,
+      "loss": 0.1859,
+      "step": 5433
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.9322918773918162,
+      "learning_rate": 8.17903350725156e-06,
+      "loss": 0.174,
+      "step": 5434
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.4108369299277024,
+      "learning_rate": 8.178338143096116e-06,
+      "loss": 0.2235,
+      "step": 5435
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.2391767963834692,
+      "learning_rate": 8.177642675769988e-06,
+      "loss": 0.2089,
+      "step": 5436
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.8745570051266242,
+      "learning_rate": 8.176947105295753e-06,
+      "loss": 0.1478,
+      "step": 5437
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.8004768573572163,
+      "learning_rate": 8.176251431695985e-06,
+      "loss": 0.2539,
+      "step": 5438
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.9481463103926462,
+      "learning_rate": 8.175555654993272e-06,
+      "loss": 0.2478,
+      "step": 5439
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.201788513725113,
+      "learning_rate": 8.174859775210196e-06,
+      "loss": 0.1896,
+      "step": 5440
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.1552110511657807,
+      "learning_rate": 8.174163792369347e-06,
+      "loss": 0.219,
+      "step": 5441
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.8417066350925061,
+      "learning_rate": 8.173467706493314e-06,
+      "loss": 0.153,
+      "step": 5442
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.5169793903080688,
+      "learning_rate": 8.172771517604695e-06,
+      "loss": 0.2033,
+      "step": 5443
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.275264505924178,
+      "learning_rate": 8.172075225726089e-06,
+      "loss": 0.2454,
+      "step": 5444
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.253191610596958,
+      "learning_rate": 8.171378830880096e-06,
+      "loss": 0.2075,
+      "step": 5445
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.884877111556089,
+      "learning_rate": 8.170682333089322e-06,
+      "loss": 0.2242,
+      "step": 5446
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.0943913282530298,
+      "learning_rate": 8.169985732376376e-06,
+      "loss": 0.2075,
+      "step": 5447
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.0715977077229162,
+      "learning_rate": 8.169289028763872e-06,
+      "loss": 0.1889,
+      "step": 5448
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.342081647746723,
+      "learning_rate": 8.168592222274423e-06,
+      "loss": 0.2194,
+      "step": 5449
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.3258168387146245,
+      "learning_rate": 8.167895312930649e-06,
+      "loss": 0.23,
+      "step": 5450
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.182667303559081,
+      "learning_rate": 8.16719830075517e-06,
+      "loss": 0.1946,
+      "step": 5451
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.376216412683566,
+      "learning_rate": 8.166501185770614e-06,
+      "loss": 0.2232,
+      "step": 5452
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.3739451106126663,
+      "learning_rate": 8.165803967999607e-06,
+      "loss": 0.1772,
+      "step": 5453
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.1811937685227467,
+      "learning_rate": 8.165106647464784e-06,
+      "loss": 0.2014,
+      "step": 5454
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.8413515897721566,
+      "learning_rate": 8.16440922418878e-06,
+      "loss": 0.2484,
+      "step": 5455
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.9405002151615256,
+      "learning_rate": 8.163711698194232e-06,
+      "loss": 0.1863,
+      "step": 5456
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.897192137502592,
+      "learning_rate": 8.163014069503783e-06,
+      "loss": 0.2093,
+      "step": 5457
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.2108444193475065,
+      "learning_rate": 8.162316338140078e-06,
+      "loss": 0.222,
+      "step": 5458
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.290380639156941,
+      "learning_rate": 8.161618504125766e-06,
+      "loss": 0.2297,
+      "step": 5459
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.8213519504342248,
+      "learning_rate": 8.1609205674835e-06,
+      "loss": 0.1666,
+      "step": 5460
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.124352952861335,
+      "learning_rate": 8.160222528235935e-06,
+      "loss": 0.2127,
+      "step": 5461
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.424591218410608,
+      "learning_rate": 8.159524386405728e-06,
+      "loss": 0.2188,
+      "step": 5462
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.2789212905847593,
+      "learning_rate": 8.158826142015546e-06,
+      "loss": 0.2189,
+      "step": 5463
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.1551657261299169,
+      "learning_rate": 8.158127795088048e-06,
+      "loss": 0.1816,
+      "step": 5464
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.5316875547862867,
+      "learning_rate": 8.157429345645906e-06,
+      "loss": 0.2179,
+      "step": 5465
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.2787937587301412,
+      "learning_rate": 8.156730793711793e-06,
+      "loss": 0.1764,
+      "step": 5466
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.2982752548283194,
+      "learning_rate": 8.15603213930838e-06,
+      "loss": 0.2342,
+      "step": 5467
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.3746880860272144,
+      "learning_rate": 8.155333382458352e-06,
+      "loss": 0.237,
+      "step": 5468
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.2438376460511547,
+      "learning_rate": 8.154634523184389e-06,
+      "loss": 0.1973,
+      "step": 5469
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.8909588572853484,
+      "learning_rate": 8.153935561509173e-06,
+      "loss": 0.2269,
+      "step": 5470
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.0372626214753569,
+      "learning_rate": 8.153236497455394e-06,
+      "loss": 0.185,
+      "step": 5471
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.7278865811850461,
+      "learning_rate": 8.152537331045748e-06,
+      "loss": 0.1584,
+      "step": 5472
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.0499666293126302,
+      "learning_rate": 8.151838062302925e-06,
+      "loss": 0.1587,
+      "step": 5473
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.3775743692801512,
+      "learning_rate": 8.151138691249627e-06,
+      "loss": 0.2179,
+      "step": 5474
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.3418669925232238,
+      "learning_rate": 8.150439217908557e-06,
+      "loss": 0.1977,
+      "step": 5475
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.9645918148481838,
+      "learning_rate": 8.149739642302415e-06,
+      "loss": 0.1972,
+      "step": 5476
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.177635017872724,
+      "learning_rate": 8.149039964453916e-06,
+      "loss": 0.1624,
+      "step": 5477
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.184163315332817,
+      "learning_rate": 8.148340184385766e-06,
+      "loss": 0.2317,
+      "step": 5478
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.065902874114958,
+      "learning_rate": 8.147640302120685e-06,
+      "loss": 0.2094,
+      "step": 5479
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.0874536431095083,
+      "learning_rate": 8.146940317681388e-06,
+      "loss": 0.1978,
+      "step": 5480
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.0257757535381895,
+      "learning_rate": 8.146240231090601e-06,
+      "loss": 0.1397,
+      "step": 5481
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.319577760367956,
+      "learning_rate": 8.145540042371047e-06,
+      "loss": 0.2373,
+      "step": 5482
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.6392025662260425,
+      "learning_rate": 8.144839751545453e-06,
+      "loss": 0.1936,
+      "step": 5483
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.9568759539198902,
+      "learning_rate": 8.144139358636552e-06,
+      "loss": 0.1925,
+      "step": 5484
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.143772169716477,
+      "learning_rate": 8.14343886366708e-06,
+      "loss": 0.1827,
+      "step": 5485
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.4863811588725977,
+      "learning_rate": 8.142738266659775e-06,
+      "loss": 0.2963,
+      "step": 5486
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.2879514616787389,
+      "learning_rate": 8.14203756763738e-06,
+      "loss": 0.1908,
+      "step": 5487
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.1510591483155124,
+      "learning_rate": 8.141336766622635e-06,
+      "loss": 0.192,
+      "step": 5488
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.2743196174732552,
+      "learning_rate": 8.140635863638295e-06,
+      "loss": 0.2207,
+      "step": 5489
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.8094494364727784,
+      "learning_rate": 8.139934858707109e-06,
+      "loss": 0.2125,
+      "step": 5490
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.1644988506262535,
+      "learning_rate": 8.139233751851831e-06,
+      "loss": 0.2373,
+      "step": 5491
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.3420343355509108,
+      "learning_rate": 8.138532543095221e-06,
+      "loss": 0.2353,
+      "step": 5492
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.111313454408806,
+      "learning_rate": 8.137831232460039e-06,
+      "loss": 0.1818,
+      "step": 5493
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.9133229340796732,
+      "learning_rate": 8.13712981996905e-06,
+      "loss": 0.1694,
+      "step": 5494
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.0128152200602887,
+      "learning_rate": 8.136428305645026e-06,
+      "loss": 0.1394,
+      "step": 5495
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.2521326052459594,
+      "learning_rate": 8.135726689510733e-06,
+      "loss": 0.2394,
+      "step": 5496
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.3565072250792474,
+      "learning_rate": 8.135024971588949e-06,
+      "loss": 0.2125,
+      "step": 5497
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.2879280741207322,
+      "learning_rate": 8.134323151902453e-06,
+      "loss": 0.2336,
+      "step": 5498
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.9534708332721218,
+      "learning_rate": 8.133621230474023e-06,
+      "loss": 0.1955,
+      "step": 5499
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.1100391218434627,
+      "learning_rate": 8.132919207326449e-06,
+      "loss": 0.1935,
+      "step": 5500
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.3964570763060522,
+      "learning_rate": 8.132217082482511e-06,
+      "loss": 0.23,
+      "step": 5501
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.8585048865583026,
+      "learning_rate": 8.13151485596501e-06,
+      "loss": 0.1698,
+      "step": 5502
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.9593063676309916,
+      "learning_rate": 8.130812527796734e-06,
+      "loss": 0.1855,
+      "step": 5503
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.2266608672576147,
+      "learning_rate": 8.130110098000483e-06,
+      "loss": 0.2369,
+      "step": 5504
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.7276585640787555,
+      "learning_rate": 8.129407566599059e-06,
+      "loss": 0.1363,
+      "step": 5505
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.9433318123236341,
+      "learning_rate": 8.128704933615266e-06,
+      "loss": 0.2141,
+      "step": 5506
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.8689763781910005,
+      "learning_rate": 8.12800219907191e-06,
+      "loss": 0.2097,
+      "step": 5507
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.0776437085940016,
+      "learning_rate": 8.127299362991805e-06,
+      "loss": 0.2193,
+      "step": 5508
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.2982117767984611,
+      "learning_rate": 8.126596425397766e-06,
+      "loss": 0.209,
+      "step": 5509
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.359523710250334,
+      "learning_rate": 8.125893386312607e-06,
+      "loss": 0.2256,
+      "step": 5510
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.3685949628589185,
+      "learning_rate": 8.125190245759151e-06,
+      "loss": 0.2032,
+      "step": 5511
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.7685580745940526,
+      "learning_rate": 8.124487003760223e-06,
+      "loss": 0.2763,
+      "step": 5512
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.0559850446491408,
+      "learning_rate": 8.123783660338651e-06,
+      "loss": 0.1809,
+      "step": 5513
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.9582936602220695,
+      "learning_rate": 8.123080215517265e-06,
+      "loss": 0.1988,
+      "step": 5514
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.17750328874559,
+      "learning_rate": 8.122376669318899e-06,
+      "loss": 0.2398,
+      "step": 5515
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.0223497222779585,
+      "learning_rate": 8.12167302176639e-06,
+      "loss": 0.1811,
+      "step": 5516
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.216563500914222,
+      "learning_rate": 8.12096927288258e-06,
+      "loss": 0.1902,
+      "step": 5517
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.9338642620277481,
+      "learning_rate": 8.120265422690312e-06,
+      "loss": 0.1575,
+      "step": 5518
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.9345787002123302,
+      "learning_rate": 8.119561471212436e-06,
+      "loss": 0.1615,
+      "step": 5519
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.2790815375999207,
+      "learning_rate": 8.1188574184718e-06,
+      "loss": 0.2277,
+      "step": 5520
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.3030888312566749,
+      "learning_rate": 8.118153264491258e-06,
+      "loss": 0.2248,
+      "step": 5521
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.5660013997970246,
+      "learning_rate": 8.117449009293668e-06,
+      "loss": 0.2627,
+      "step": 5522
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.9808642247124555,
+      "learning_rate": 8.116744652901891e-06,
+      "loss": 0.1485,
+      "step": 5523
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.2265565926719357,
+      "learning_rate": 8.116040195338792e-06,
+      "loss": 0.1792,
+      "step": 5524
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.295762750641213,
+      "learning_rate": 8.115335636627233e-06,
+      "loss": 0.1824,
+      "step": 5525
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.8260335581001879,
+      "learning_rate": 8.11463097679009e-06,
+      "loss": 0.2472,
+      "step": 5526
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.0383185900194591,
+      "learning_rate": 8.113926215850234e-06,
+      "loss": 0.1709,
+      "step": 5527
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.9380039033080338,
+      "learning_rate": 8.11322135383054e-06,
+      "loss": 0.1896,
+      "step": 5528
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.3542157078690436,
+      "learning_rate": 8.112516390753894e-06,
+      "loss": 0.2071,
+      "step": 5529
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.978692191101818,
+      "learning_rate": 8.111811326643175e-06,
+      "loss": 0.1839,
+      "step": 5530
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.709666348216502,
+      "learning_rate": 8.11110616152127e-06,
+      "loss": 0.2086,
+      "step": 5531
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.075600342509487,
+      "learning_rate": 8.110400895411073e-06,
+      "loss": 0.2144,
+      "step": 5532
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.7873555087278704,
+      "learning_rate": 8.109695528335472e-06,
+      "loss": 0.1976,
+      "step": 5533
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.8034507079876781,
+      "learning_rate": 8.108990060317366e-06,
+      "loss": 0.1704,
+      "step": 5534
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.7609462186094873,
+      "learning_rate": 8.108284491379655e-06,
+      "loss": 0.1755,
+      "step": 5535
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.3639435908017215,
+      "learning_rate": 8.107578821545243e-06,
+      "loss": 0.2233,
+      "step": 5536
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.8034703449645011,
+      "learning_rate": 8.106873050837033e-06,
+      "loss": 0.1588,
+      "step": 5537
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.1115940308864483,
+      "learning_rate": 8.106167179277939e-06,
+      "loss": 0.2696,
+      "step": 5538
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.9396125584811128,
+      "learning_rate": 8.105461206890873e-06,
+      "loss": 0.1939,
+      "step": 5539
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.2526881593934591,
+      "learning_rate": 8.10475513369875e-06,
+      "loss": 0.2165,
+      "step": 5540
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.1874795413137176,
+      "learning_rate": 8.10404895972449e-06,
+      "loss": 0.1809,
+      "step": 5541
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.6443960019018166,
+      "learning_rate": 8.103342684991015e-06,
+      "loss": 0.1502,
+      "step": 5542
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 3.5368726915512734,
+      "learning_rate": 8.102636309521251e-06,
+      "loss": 0.27,
+      "step": 5543
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.5133655115004823,
+      "learning_rate": 8.10192983333813e-06,
+      "loss": 0.1996,
+      "step": 5544
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.8480816665744987,
+      "learning_rate": 8.101223256464582e-06,
+      "loss": 0.2388,
+      "step": 5545
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.9266995330269089,
+      "learning_rate": 8.100516578923543e-06,
+      "loss": 0.1651,
+      "step": 5546
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.9741900159120882,
+      "learning_rate": 8.099809800737954e-06,
+      "loss": 0.1691,
+      "step": 5547
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.001416280044178,
+      "learning_rate": 8.099102921930757e-06,
+      "loss": 0.236,
+      "step": 5548
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.5673393802032387,
+      "learning_rate": 8.098395942524897e-06,
+      "loss": 0.2104,
+      "step": 5549
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.8971160292318429,
+      "learning_rate": 8.09768886254332e-06,
+      "loss": 0.1871,
+      "step": 5550
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.0667738608375006,
+      "learning_rate": 8.096981682008984e-06,
+      "loss": 0.1984,
+      "step": 5551
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.224172824638443,
+      "learning_rate": 8.096274400944843e-06,
+      "loss": 0.236,
+      "step": 5552
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.227300731553365,
+      "learning_rate": 8.095567019373851e-06,
+      "loss": 0.2049,
+      "step": 5553
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.7697755548111624,
+      "learning_rate": 8.094859537318976e-06,
+      "loss": 0.2084,
+      "step": 5554
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.9411770434747077,
+      "learning_rate": 8.09415195480318e-06,
+      "loss": 0.154,
+      "step": 5555
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.3285029883205728,
+      "learning_rate": 8.093444271849434e-06,
+      "loss": 0.191,
+      "step": 5556
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.2425583249924694,
+      "learning_rate": 8.092736488480705e-06,
+      "loss": 0.1604,
+      "step": 5557
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.2457764110856362,
+      "learning_rate": 8.092028604719972e-06,
+      "loss": 0.1786,
+      "step": 5558
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.360584877090192,
+      "learning_rate": 8.091320620590213e-06,
+      "loss": 0.2203,
+      "step": 5559
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.457349455477586,
+      "learning_rate": 8.09061253611441e-06,
+      "loss": 0.1738,
+      "step": 5560
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.774257804192172,
+      "learning_rate": 8.089904351315546e-06,
+      "loss": 0.1911,
+      "step": 5561
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.9905901170725377,
+      "learning_rate": 8.089196066216609e-06,
+      "loss": 0.1663,
+      "step": 5562
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.1396409829622398,
+      "learning_rate": 8.08848768084059e-06,
+      "loss": 0.1757,
+      "step": 5563
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.2172568601402038,
+      "learning_rate": 8.087779195210488e-06,
+      "loss": 0.2201,
+      "step": 5564
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.9757067402230444,
+      "learning_rate": 8.087070609349297e-06,
+      "loss": 0.233,
+      "step": 5565
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.8898875358030532,
+      "learning_rate": 8.086361923280017e-06,
+      "loss": 0.1977,
+      "step": 5566
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.2561190145442764,
+      "learning_rate": 8.085653137025658e-06,
+      "loss": 0.1802,
+      "step": 5567
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.9504805800218922,
+      "learning_rate": 8.084944250609219e-06,
+      "loss": 0.198,
+      "step": 5568
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.8397807918282308,
+      "learning_rate": 8.08423526405372e-06,
+      "loss": 0.1596,
+      "step": 5569
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.607695758981812,
+      "learning_rate": 8.083526177382168e-06,
+      "loss": 0.2507,
+      "step": 5570
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.5601141309329747,
+      "learning_rate": 8.082816990617584e-06,
+      "loss": 0.2724,
+      "step": 5571
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.5818598948797098,
+      "learning_rate": 8.082107703782987e-06,
+      "loss": 0.2841,
+      "step": 5572
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.314589455097089,
+      "learning_rate": 8.081398316901402e-06,
+      "loss": 0.263,
+      "step": 5573
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.310780852532949,
+      "learning_rate": 8.080688829995856e-06,
+      "loss": 0.1782,
+      "step": 5574
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.5134703003579912,
+      "learning_rate": 8.07997924308938e-06,
+      "loss": 0.2128,
+      "step": 5575
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.0773093413791162,
+      "learning_rate": 8.079269556205006e-06,
+      "loss": 0.1617,
+      "step": 5576
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.7481414566878186,
+      "learning_rate": 8.078559769365771e-06,
+      "loss": 0.2326,
+      "step": 5577
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.4794470751843611,
+      "learning_rate": 8.077849882594716e-06,
+      "loss": 0.2173,
+      "step": 5578
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.978454401381363,
+      "learning_rate": 8.077139895914883e-06,
+      "loss": 0.2297,
+      "step": 5579
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.9691982722252324,
+      "learning_rate": 8.07642980934932e-06,
+      "loss": 0.1958,
+      "step": 5580
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.050120896910859,
+      "learning_rate": 8.075719622921077e-06,
+      "loss": 0.2024,
+      "step": 5581
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.4506422388119968,
+      "learning_rate": 8.075009336653205e-06,
+      "loss": 0.2278,
+      "step": 5582
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.8503405811652694,
+      "learning_rate": 8.074298950568762e-06,
+      "loss": 0.2194,
+      "step": 5583
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.0521984762228798,
+      "learning_rate": 8.073588464690808e-06,
+      "loss": 0.2104,
+      "step": 5584
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.1873515955523593,
+      "learning_rate": 8.072877879042406e-06,
+      "loss": 0.2242,
+      "step": 5585
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.8923449180340224,
+      "learning_rate": 8.072167193646618e-06,
+      "loss": 0.1434,
+      "step": 5586
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.2737136728574006,
+      "learning_rate": 8.071456408526518e-06,
+      "loss": 0.2142,
+      "step": 5587
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.837202999310411,
+      "learning_rate": 8.070745523705178e-06,
+      "loss": 0.195,
+      "step": 5588
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.837357462578343,
+      "learning_rate": 8.07003453920567e-06,
+      "loss": 0.1649,
+      "step": 5589
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.2176072191538512,
+      "learning_rate": 8.069323455051079e-06,
+      "loss": 0.1934,
+      "step": 5590
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.1323254977174209,
+      "learning_rate": 8.06861227126448e-06,
+      "loss": 0.1798,
+      "step": 5591
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.9414882772569139,
+      "learning_rate": 8.067900987868964e-06,
+      "loss": 0.1784,
+      "step": 5592
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.6063265921833685,
+      "learning_rate": 8.067189604887618e-06,
+      "loss": 0.2886,
+      "step": 5593
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.9433167242593014,
+      "learning_rate": 8.066478122343533e-06,
+      "loss": 0.2081,
+      "step": 5594
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.155630457990558,
+      "learning_rate": 8.065766540259803e-06,
+      "loss": 0.1937,
+      "step": 5595
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.1771936345605118,
+      "learning_rate": 8.06505485865953e-06,
+      "loss": 0.2033,
+      "step": 5596
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.0609520512286659,
+      "learning_rate": 8.064343077565813e-06,
+      "loss": 0.1967,
+      "step": 5597
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.9240446744198655,
+      "learning_rate": 8.063631197001759e-06,
+      "loss": 0.1733,
+      "step": 5598
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.2868732400759852,
+      "learning_rate": 8.062919216990475e-06,
+      "loss": 0.2087,
+      "step": 5599
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.4175677730889764,
+      "learning_rate": 8.062207137555069e-06,
+      "loss": 0.157,
+      "step": 5600
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.9703563535151343,
+      "learning_rate": 8.06149495871866e-06,
+      "loss": 0.1928,
+      "step": 5601
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.3738302014974868,
+      "learning_rate": 8.060782680504366e-06,
+      "loss": 0.2121,
+      "step": 5602
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.319133113289576,
+      "learning_rate": 8.060070302935305e-06,
+      "loss": 0.1782,
+      "step": 5603
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.0347771040213856,
+      "learning_rate": 8.059357826034601e-06,
+      "loss": 0.1917,
+      "step": 5604
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.1261040382979173,
+      "learning_rate": 8.058645249825382e-06,
+      "loss": 0.2309,
+      "step": 5605
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.224306870778086,
+      "learning_rate": 8.057932574330782e-06,
+      "loss": 0.1667,
+      "step": 5606
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.0556301725525097,
+      "learning_rate": 8.057219799573932e-06,
+      "loss": 0.1363,
+      "step": 5607
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.9291057850403148,
+      "learning_rate": 8.056506925577966e-06,
+      "loss": 0.2115,
+      "step": 5608
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.4477071513302293,
+      "learning_rate": 8.055793952366032e-06,
+      "loss": 0.2395,
+      "step": 5609
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.4738262244241866,
+      "learning_rate": 8.055080879961265e-06,
+      "loss": 0.2367,
+      "step": 5610
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.9808649273361633,
+      "learning_rate": 8.054367708386818e-06,
+      "loss": 0.2159,
+      "step": 5611
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.0361225656187876,
+      "learning_rate": 8.05365443766584e-06,
+      "loss": 0.2412,
+      "step": 5612
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.120157646838677,
+      "learning_rate": 8.05294106782148e-06,
+      "loss": 0.2142,
+      "step": 5613
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.9980636397444888,
+      "learning_rate": 8.052227598876901e-06,
+      "loss": 0.2185,
+      "step": 5614
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.847066757796316,
+      "learning_rate": 8.051514030855256e-06,
+      "loss": 0.1722,
+      "step": 5615
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.173387530465464,
+      "learning_rate": 8.050800363779711e-06,
+      "loss": 0.1609,
+      "step": 5616
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.19888777623286,
+      "learning_rate": 8.050086597673433e-06,
+      "loss": 0.1912,
+      "step": 5617
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.2512810326895174,
+      "learning_rate": 8.04937273255959e-06,
+      "loss": 0.2454,
+      "step": 5618
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.5580208306200873,
+      "learning_rate": 8.048658768461354e-06,
+      "loss": 0.2143,
+      "step": 5619
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.3660308795586629,
+      "learning_rate": 8.047944705401901e-06,
+      "loss": 0.2202,
+      "step": 5620
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.1611092503292946,
+      "learning_rate": 8.047230543404411e-06,
+      "loss": 0.2025,
+      "step": 5621
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.9539176133720344,
+      "learning_rate": 8.046516282492065e-06,
+      "loss": 0.2079,
+      "step": 5622
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.1849309163262027,
+      "learning_rate": 8.045801922688051e-06,
+      "loss": 0.2056,
+      "step": 5623
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.3298494608766216,
+      "learning_rate": 8.045087464015551e-06,
+      "loss": 0.1861,
+      "step": 5624
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.0454059638162474,
+      "learning_rate": 8.044372906497762e-06,
+      "loss": 0.2196,
+      "step": 5625
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.123081236603866,
+      "learning_rate": 8.04365825015788e-06,
+      "loss": 0.1784,
+      "step": 5626
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.1530276289718455,
+      "learning_rate": 8.042943495019098e-06,
+      "loss": 0.1544,
+      "step": 5627
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.3465460239990634,
+      "learning_rate": 8.042228641104622e-06,
+      "loss": 0.2472,
+      "step": 5628
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.160320758394646,
+      "learning_rate": 8.041513688437654e-06,
+      "loss": 0.1741,
+      "step": 5629
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.4430406990350746,
+      "learning_rate": 8.040798637041405e-06,
+      "loss": 0.2232,
+      "step": 5630
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.2474634750146891,
+      "learning_rate": 8.040083486939082e-06,
+      "loss": 0.2317,
+      "step": 5631
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.6497090280880136,
+      "learning_rate": 8.039368238153901e-06,
+      "loss": 0.2198,
+      "step": 5632
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.0284628139922753,
+      "learning_rate": 8.038652890709078e-06,
+      "loss": 0.2055,
+      "step": 5633
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.3031861824314204,
+      "learning_rate": 8.037937444627836e-06,
+      "loss": 0.2306,
+      "step": 5634
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.232880502476384,
+      "learning_rate": 8.037221899933398e-06,
+      "loss": 0.1999,
+      "step": 5635
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.1546279638941421,
+      "learning_rate": 8.036506256648989e-06,
+      "loss": 0.2071,
+      "step": 5636
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.116567112656639,
+      "learning_rate": 8.035790514797843e-06,
+      "loss": 0.2099,
+      "step": 5637
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.9555772781728225,
+      "learning_rate": 8.03507467440319e-06,
+      "loss": 0.1815,
+      "step": 5638
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.521352465962843,
+      "learning_rate": 8.034358735488268e-06,
+      "loss": 0.2617,
+      "step": 5639
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.09669769931313,
+      "learning_rate": 8.033642698076315e-06,
+      "loss": 0.1971,
+      "step": 5640
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.5035297690459906,
+      "learning_rate": 8.032926562190577e-06,
+      "loss": 0.1896,
+      "step": 5641
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.2830911468466075,
+      "learning_rate": 8.0322103278543e-06,
+      "loss": 0.1705,
+      "step": 5642
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.2825649451241834,
+      "learning_rate": 8.03149399509073e-06,
+      "loss": 0.1753,
+      "step": 5643
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.618391625306073,
+      "learning_rate": 8.030777563923123e-06,
+      "loss": 0.1841,
+      "step": 5644
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.334698540350058,
+      "learning_rate": 8.030061034374733e-06,
+      "loss": 0.2487,
+      "step": 5645
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.398603321612057,
+      "learning_rate": 8.029344406468819e-06,
+      "loss": 0.2511,
+      "step": 5646
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.065038546295013,
+      "learning_rate": 8.028627680228645e-06,
+      "loss": 0.2033,
+      "step": 5647
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.1968892596493457,
+      "learning_rate": 8.027910855677473e-06,
+      "loss": 0.2241,
+      "step": 5648
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.735052970739277,
+      "learning_rate": 8.027193932838573e-06,
+      "loss": 0.2457,
+      "step": 5649
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.0890442759796548,
+      "learning_rate": 8.026476911735219e-06,
+      "loss": 0.1743,
+      "step": 5650
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.3807803027486578,
+      "learning_rate": 8.025759792390684e-06,
+      "loss": 0.2357,
+      "step": 5651
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.8830374604099345,
+      "learning_rate": 8.025042574828246e-06,
+      "loss": 0.1841,
+      "step": 5652
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.1277184490880308,
+      "learning_rate": 8.024325259071187e-06,
+      "loss": 0.1827,
+      "step": 5653
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.0220425741789825,
+      "learning_rate": 8.02360784514279e-06,
+      "loss": 0.2535,
+      "step": 5654
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.060186421450639,
+      "learning_rate": 8.022890333066343e-06,
+      "loss": 0.2014,
+      "step": 5655
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.1751561689459222,
+      "learning_rate": 8.022172722865138e-06,
+      "loss": 0.2502,
+      "step": 5656
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.451093479056655,
+      "learning_rate": 8.02145501456247e-06,
+      "loss": 0.1715,
+      "step": 5657
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.2187758161793831,
+      "learning_rate": 8.020737208181632e-06,
+      "loss": 0.2001,
+      "step": 5658
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.2935817394857858,
+      "learning_rate": 8.02001930374593e-06,
+      "loss": 0.1676,
+      "step": 5659
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.090807903564142,
+      "learning_rate": 8.019301301278663e-06,
+      "loss": 0.2218,
+      "step": 5660
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.556286449656624,
+      "learning_rate": 8.018583200803141e-06,
+      "loss": 0.2154,
+      "step": 5661
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.4547055484305627,
+      "learning_rate": 8.017865002342672e-06,
+      "loss": 0.2241,
+      "step": 5662
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.0548358636277897,
+      "learning_rate": 8.01714670592057e-06,
+      "loss": 0.2509,
+      "step": 5663
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.043444678483818,
+      "learning_rate": 8.01642831156015e-06,
+      "loss": 0.187,
+      "step": 5664
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.4789029499708337,
+      "learning_rate": 8.015709819284732e-06,
+      "loss": 0.1855,
+      "step": 5665
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.5877341560514158,
+      "learning_rate": 8.01499122911764e-06,
+      "loss": 0.2133,
+      "step": 5666
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.1541408911069295,
+      "learning_rate": 8.014272541082198e-06,
+      "loss": 0.194,
+      "step": 5667
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.0044908046249315,
+      "learning_rate": 8.013553755201737e-06,
+      "loss": 0.1918,
+      "step": 5668
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.999053712238828,
+      "learning_rate": 8.012834871499589e-06,
+      "loss": 0.2216,
+      "step": 5669
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.5069221439783174,
+      "learning_rate": 8.012115889999087e-06,
+      "loss": 0.2262,
+      "step": 5670
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.1145332958748684,
+      "learning_rate": 8.01139681072357e-06,
+      "loss": 0.2282,
+      "step": 5671
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.078615085847242,
+      "learning_rate": 8.010677633696383e-06,
+      "loss": 0.1999,
+      "step": 5672
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.1795436774617527,
+      "learning_rate": 8.009958358940868e-06,
+      "loss": 0.19,
+      "step": 5673
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.3574677987872217,
+      "learning_rate": 8.009238986480374e-06,
+      "loss": 0.2381,
+      "step": 5674
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.0815007978704017,
+      "learning_rate": 8.008519516338254e-06,
+      "loss": 0.2374,
+      "step": 5675
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.912185689827414,
+      "learning_rate": 8.007799948537857e-06,
+      "loss": 0.2777,
+      "step": 5676
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.1261399077007659,
+      "learning_rate": 8.007080283102547e-06,
+      "loss": 0.2063,
+      "step": 5677
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.4685744322687235,
+      "learning_rate": 8.006360520055679e-06,
+      "loss": 0.1692,
+      "step": 5678
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.9887553583707065,
+      "learning_rate": 8.00564065942062e-06,
+      "loss": 0.1576,
+      "step": 5679
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.1656064956311,
+      "learning_rate": 8.00492070122074e-06,
+      "loss": 0.2188,
+      "step": 5680
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.1397262667612318,
+      "learning_rate": 8.004200645479403e-06,
+      "loss": 0.237,
+      "step": 5681
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.4085737737930701,
+      "learning_rate": 8.003480492219988e-06,
+      "loss": 0.2157,
+      "step": 5682
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.1381196945497258,
+      "learning_rate": 8.002760241465867e-06,
+      "loss": 0.2062,
+      "step": 5683
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.3959697842328656,
+      "learning_rate": 8.002039893240423e-06,
+      "loss": 0.2307,
+      "step": 5684
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.4233738994940863,
+      "learning_rate": 8.001319447567037e-06,
+      "loss": 0.2045,
+      "step": 5685
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.1793734688286477,
+      "learning_rate": 8.000598904469097e-06,
+      "loss": 0.1479,
+      "step": 5686
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.0821699129350462,
+      "learning_rate": 7.999878263969989e-06,
+      "loss": 0.2379,
+      "step": 5687
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.7117163026549316,
+      "learning_rate": 7.999157526093109e-06,
+      "loss": 0.1874,
+      "step": 5688
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.250854969443376,
+      "learning_rate": 7.99843669086185e-06,
+      "loss": 0.2344,
+      "step": 5689
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.533220011624201,
+      "learning_rate": 7.997715758299613e-06,
+      "loss": 0.2284,
+      "step": 5690
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.321024725186281,
+      "learning_rate": 7.996994728429798e-06,
+      "loss": 0.2125,
+      "step": 5691
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.3651491794826598,
+      "learning_rate": 7.996273601275812e-06,
+      "loss": 0.2046,
+      "step": 5692
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.5485890433661686,
+      "learning_rate": 7.995552376861061e-06,
+      "loss": 0.2401,
+      "step": 5693
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.9570238619149418,
+      "learning_rate": 7.994831055208957e-06,
+      "loss": 0.178,
+      "step": 5694
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.243812917610755,
+      "learning_rate": 7.994109636342915e-06,
+      "loss": 0.1799,
+      "step": 5695
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.0363556553030537,
+      "learning_rate": 7.993388120286354e-06,
+      "loss": 0.1925,
+      "step": 5696
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.156902009383616,
+      "learning_rate": 7.99266650706269e-06,
+      "loss": 0.2093,
+      "step": 5697
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.624433558572832,
+      "learning_rate": 7.991944796695352e-06,
+      "loss": 0.2009,
+      "step": 5698
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.401851285996636,
+      "learning_rate": 7.991222989207765e-06,
+      "loss": 0.2111,
+      "step": 5699
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.9476360855095707,
+      "learning_rate": 7.99050108462336e-06,
+      "loss": 0.2091,
+      "step": 5700
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.9510728424184515,
+      "learning_rate": 7.989779082965571e-06,
+      "loss": 0.1606,
+      "step": 5701
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.0597728790515815,
+      "learning_rate": 7.989056984257834e-06,
+      "loss": 0.2227,
+      "step": 5702
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.1237097467335353,
+      "learning_rate": 7.988334788523586e-06,
+      "loss": 0.2108,
+      "step": 5703
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.000105086780349,
+      "learning_rate": 7.987612495786274e-06,
+      "loss": 0.2025,
+      "step": 5704
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.5690249091517157,
+      "learning_rate": 7.986890106069342e-06,
+      "loss": 0.2259,
+      "step": 5705
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.335685760638196,
+      "learning_rate": 7.98616761939624e-06,
+      "loss": 0.2107,
+      "step": 5706
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.051504070874967,
+      "learning_rate": 7.98544503579042e-06,
+      "loss": 0.1921,
+      "step": 5707
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.0172102184366212,
+      "learning_rate": 7.984722355275336e-06,
+      "loss": 0.1842,
+      "step": 5708
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.4881361667874167,
+      "learning_rate": 7.983999577874451e-06,
+      "loss": 0.213,
+      "step": 5709
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.9297986184559828,
+      "learning_rate": 7.98327670361122e-06,
+      "loss": 0.2197,
+      "step": 5710
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.9008981759509662,
+      "learning_rate": 7.982553732509116e-06,
+      "loss": 0.1457,
+      "step": 5711
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.2712298635137047,
+      "learning_rate": 7.981830664591599e-06,
+      "loss": 0.1812,
+      "step": 5712
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.3091724539535172,
+      "learning_rate": 7.981107499882147e-06,
+      "loss": 0.2224,
+      "step": 5713
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.6397753740788104,
+      "learning_rate": 7.980384238404228e-06,
+      "loss": 0.2106,
+      "step": 5714
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.3244357086064396,
+      "learning_rate": 7.979660880181324e-06,
+      "loss": 0.2289,
+      "step": 5715
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.9570183947731314,
+      "learning_rate": 7.978937425236915e-06,
+      "loss": 0.1969,
+      "step": 5716
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.206228380398632,
+      "learning_rate": 7.978213873594482e-06,
+      "loss": 0.179,
+      "step": 5717
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.3349563665528028,
+      "learning_rate": 7.977490225277518e-06,
+      "loss": 0.2456,
+      "step": 5718
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.9519855303438127,
+      "learning_rate": 7.976766480309508e-06,
+      "loss": 0.2539,
+      "step": 5719
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.2767818422406463,
+      "learning_rate": 7.976042638713945e-06,
+      "loss": 0.1905,
+      "step": 5720
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.3228449038567904,
+      "learning_rate": 7.975318700514325e-06,
+      "loss": 0.2625,
+      "step": 5721
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.1541836645955301,
+      "learning_rate": 7.974594665734153e-06,
+      "loss": 0.1529,
+      "step": 5722
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.9822877143882833,
+      "learning_rate": 7.973870534396925e-06,
+      "loss": 0.1934,
+      "step": 5723
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.2975803921486961,
+      "learning_rate": 7.973146306526149e-06,
+      "loss": 0.2213,
+      "step": 5724
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.0335935258901154,
+      "learning_rate": 7.972421982145335e-06,
+      "loss": 0.2213,
+      "step": 5725
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.1456085512007435,
+      "learning_rate": 7.971697561277996e-06,
+      "loss": 0.2021,
+      "step": 5726
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.6864534748986022,
+      "learning_rate": 7.970973043947644e-06,
+      "loss": 0.2135,
+      "step": 5727
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.9321414028467061,
+      "learning_rate": 7.970248430177797e-06,
+      "loss": 0.193,
+      "step": 5728
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.9308373869245171,
+      "learning_rate": 7.969523719991981e-06,
+      "loss": 0.1848,
+      "step": 5729
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.6656239396855912,
+      "learning_rate": 7.968798913413713e-06,
+      "loss": 0.2703,
+      "step": 5730
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.3336625471957055,
+      "learning_rate": 7.968074010466528e-06,
+      "loss": 0.2131,
+      "step": 5731
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.427480385415092,
+      "learning_rate": 7.967349011173955e-06,
+      "loss": 0.1739,
+      "step": 5732
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.0822438365175218,
+      "learning_rate": 7.966623915559524e-06,
+      "loss": 0.2103,
+      "step": 5733
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.099558193004885,
+      "learning_rate": 7.965898723646777e-06,
+      "loss": 0.1964,
+      "step": 5734
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.1809732948772564,
+      "learning_rate": 7.965173435459251e-06,
+      "loss": 0.1734,
+      "step": 5735
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.519330011393755,
+      "learning_rate": 7.96444805102049e-06,
+      "loss": 0.2014,
+      "step": 5736
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.1328356345872728,
+      "learning_rate": 7.963722570354041e-06,
+      "loss": 0.2243,
+      "step": 5737
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.1713913451942588,
+      "learning_rate": 7.962996993483454e-06,
+      "loss": 0.2228,
+      "step": 5738
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.3832115995594034,
+      "learning_rate": 7.962271320432278e-06,
+      "loss": 0.2393,
+      "step": 5739
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.1721998201654344,
+      "learning_rate": 7.961545551224073e-06,
+      "loss": 0.2197,
+      "step": 5740
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.3899445771113035,
+      "learning_rate": 7.960819685882397e-06,
+      "loss": 0.196,
+      "step": 5741
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.834892763176594,
+      "learning_rate": 7.960093724430811e-06,
+      "loss": 0.1817,
+      "step": 5742
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.7666671008140347,
+      "learning_rate": 7.959367666892882e-06,
+      "loss": 0.2043,
+      "step": 5743
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.1303547913630154,
+      "learning_rate": 7.958641513292173e-06,
+      "loss": 0.2028,
+      "step": 5744
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.8365953510714823,
+      "learning_rate": 7.957915263652262e-06,
+      "loss": 0.2164,
+      "step": 5745
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.9901507255766687,
+      "learning_rate": 7.957188917996722e-06,
+      "loss": 0.2305,
+      "step": 5746
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.1488814566282124,
+      "learning_rate": 7.956462476349126e-06,
+      "loss": 0.2097,
+      "step": 5747
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.0055007174774215,
+      "learning_rate": 7.95573593873306e-06,
+      "loss": 0.1917,
+      "step": 5748
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.147501891196683,
+      "learning_rate": 7.955009305172104e-06,
+      "loss": 0.2335,
+      "step": 5749
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.77397041673435,
+      "learning_rate": 7.954282575689848e-06,
+      "loss": 0.1789,
+      "step": 5750
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.9590606717371614,
+      "learning_rate": 7.95355575030988e-06,
+      "loss": 0.2126,
+      "step": 5751
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.2503438938469047,
+      "learning_rate": 7.952828829055794e-06,
+      "loss": 0.2173,
+      "step": 5752
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.0937292003356864,
+      "learning_rate": 7.952101811951188e-06,
+      "loss": 0.1837,
+      "step": 5753
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.2628525977481646,
+      "learning_rate": 7.951374699019658e-06,
+      "loss": 0.1441,
+      "step": 5754
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.1879139142762485,
+      "learning_rate": 7.950647490284808e-06,
+      "loss": 0.2187,
+      "step": 5755
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.2587813413396514,
+      "learning_rate": 7.949920185770244e-06,
+      "loss": 0.2138,
+      "step": 5756
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.6904151410116606,
+      "learning_rate": 7.949192785499573e-06,
+      "loss": 0.1709,
+      "step": 5757
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.0273310276157526,
+      "learning_rate": 7.94846528949641e-06,
+      "loss": 0.1993,
+      "step": 5758
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.2249712538630697,
+      "learning_rate": 7.947737697784368e-06,
+      "loss": 0.1759,
+      "step": 5759
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.9095749984138857,
+      "learning_rate": 7.947010010387064e-06,
+      "loss": 0.1812,
+      "step": 5760
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.0343932110574947,
+      "learning_rate": 7.946282227328123e-06,
+      "loss": 0.2419,
+      "step": 5761
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.2323636171200918,
+      "learning_rate": 7.945554348631165e-06,
+      "loss": 0.2082,
+      "step": 5762
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.151421212812733,
+      "learning_rate": 7.94482637431982e-06,
+      "loss": 0.1692,
+      "step": 5763
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.1333901313820525,
+      "learning_rate": 7.944098304417717e-06,
+      "loss": 0.2083,
+      "step": 5764
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 2.6462875011455536,
+      "learning_rate": 7.943370138948488e-06,
+      "loss": 0.1936,
+      "step": 5765
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.3138973866600503,
+      "learning_rate": 7.942641877935775e-06,
+      "loss": 0.1874,
+      "step": 5766
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.1598100683588466,
+      "learning_rate": 7.941913521403213e-06,
+      "loss": 0.1784,
+      "step": 5767
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.981931180870396,
+      "learning_rate": 7.941185069374446e-06,
+      "loss": 0.2559,
+      "step": 5768
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.180069588368617,
+      "learning_rate": 7.94045652187312e-06,
+      "loss": 0.1955,
+      "step": 5769
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.5405188365284246,
+      "learning_rate": 7.939727878922887e-06,
+      "loss": 0.2548,
+      "step": 5770
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.1244266227695192,
+      "learning_rate": 7.938999140547395e-06,
+      "loss": 0.21,
+      "step": 5771
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.2178626712265874,
+      "learning_rate": 7.9382703067703e-06,
+      "loss": 0.1699,
+      "step": 5772
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.9811772307744785,
+      "learning_rate": 7.93754137761526e-06,
+      "loss": 0.1903,
+      "step": 5773
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.278621096154844,
+      "learning_rate": 7.936812353105941e-06,
+      "loss": 0.2607,
+      "step": 5774
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.2073599050040116,
+      "learning_rate": 7.936083233266003e-06,
+      "loss": 0.2183,
+      "step": 5775
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.9755882610640859,
+      "learning_rate": 7.935354018119115e-06,
+      "loss": 0.192,
+      "step": 5776
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.9839987310064209,
+      "learning_rate": 7.934624707688947e-06,
+      "loss": 0.2181,
+      "step": 5777
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.152006706150701,
+      "learning_rate": 7.933895301999172e-06,
+      "loss": 0.203,
+      "step": 5778
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.9677383789887533,
+      "learning_rate": 7.933165801073473e-06,
+      "loss": 0.2062,
+      "step": 5779
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.309514419307317,
+      "learning_rate": 7.932436204935521e-06,
+      "loss": 0.2143,
+      "step": 5780
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.0721412223470015,
+      "learning_rate": 7.931706513609006e-06,
+      "loss": 0.2195,
+      "step": 5781
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.1710577419743409,
+      "learning_rate": 7.93097672711761e-06,
+      "loss": 0.1664,
+      "step": 5782
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.3468387418544217,
+      "learning_rate": 7.930246845485025e-06,
+      "loss": 0.1988,
+      "step": 5783
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.5709873950684465,
+      "learning_rate": 7.92951686873494e-06,
+      "loss": 0.2712,
+      "step": 5784
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.9928797462126138,
+      "learning_rate": 7.928786796891058e-06,
+      "loss": 0.2007,
+      "step": 5785
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.0422530754854453,
+      "learning_rate": 7.928056629977067e-06,
+      "loss": 0.2087,
+      "step": 5786
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.0010566279326678,
+      "learning_rate": 7.927326368016677e-06,
+      "loss": 0.2044,
+      "step": 5787
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.0876155758971047,
+      "learning_rate": 7.926596011033589e-06,
+      "loss": 0.2358,
+      "step": 5788
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.0438180101816932,
+      "learning_rate": 7.925865559051512e-06,
+      "loss": 0.1776,
+      "step": 5789
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.8610883371423215,
+      "learning_rate": 7.925135012094157e-06,
+      "loss": 0.1773,
+      "step": 5790
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.079574010311049,
+      "learning_rate": 7.924404370185237e-06,
+      "loss": 0.175,
+      "step": 5791
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.0932548594291556,
+      "learning_rate": 7.923673633348467e-06,
+      "loss": 0.183,
+      "step": 5792
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.302320614811216,
+      "learning_rate": 7.922942801607573e-06,
+      "loss": 0.2139,
+      "step": 5793
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.0579966413004283,
+      "learning_rate": 7.922211874986275e-06,
+      "loss": 0.1848,
+      "step": 5794
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.4864189149043732,
+      "learning_rate": 7.921480853508296e-06,
+      "loss": 0.1693,
+      "step": 5795
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.201892261495235,
+      "learning_rate": 7.920749737197371e-06,
+      "loss": 0.2174,
+      "step": 5796
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.5020114207766917,
+      "learning_rate": 7.920018526077228e-06,
+      "loss": 0.2373,
+      "step": 5797
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.1844580629380563,
+      "learning_rate": 7.919287220171608e-06,
+      "loss": 0.1943,
+      "step": 5798
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.3406648372075387,
+      "learning_rate": 7.918555819504244e-06,
+      "loss": 0.2577,
+      "step": 5799
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.7032775285647102,
+      "learning_rate": 7.91782432409888e-06,
+      "loss": 0.2259,
+      "step": 5800
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.1490090563320092,
+      "learning_rate": 7.917092733979259e-06,
+      "loss": 0.1915,
+      "step": 5801
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.2231817883206177,
+      "learning_rate": 7.916361049169132e-06,
+      "loss": 0.2174,
+      "step": 5802
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.3093761297280255,
+      "learning_rate": 7.915629269692248e-06,
+      "loss": 0.241,
+      "step": 5803
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.3281484405468573,
+      "learning_rate": 7.914897395572362e-06,
+      "loss": 0.1657,
+      "step": 5804
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.2769563616830923,
+      "learning_rate": 7.91416542683323e-06,
+      "loss": 0.1644,
+      "step": 5805
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.2061377060460414,
+      "learning_rate": 7.91343336349861e-06,
+      "loss": 0.1955,
+      "step": 5806
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.4363412151513075,
+      "learning_rate": 7.91270120559227e-06,
+      "loss": 0.1901,
+      "step": 5807
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.0406628909315094,
+      "learning_rate": 7.911968953137973e-06,
+      "loss": 0.1907,
+      "step": 5808
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.0399188003497453,
+      "learning_rate": 7.911236606159491e-06,
+      "loss": 0.2108,
+      "step": 5809
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.2253654768031956,
+      "learning_rate": 7.910504164680595e-06,
+      "loss": 0.216,
+      "step": 5810
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.2733379336791213,
+      "learning_rate": 7.909771628725058e-06,
+      "loss": 0.2075,
+      "step": 5811
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.3559187134001536,
+      "learning_rate": 7.909038998316661e-06,
+      "loss": 0.1807,
+      "step": 5812
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.196290986974297,
+      "learning_rate": 7.908306273479186e-06,
+      "loss": 0.2023,
+      "step": 5813
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.0673644229180796,
+      "learning_rate": 7.907573454236417e-06,
+      "loss": 0.1785,
+      "step": 5814
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.1556866729829915,
+      "learning_rate": 7.906840540612143e-06,
+      "loss": 0.1746,
+      "step": 5815
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.0907922654059503,
+      "learning_rate": 7.906107532630151e-06,
+      "loss": 0.1195,
+      "step": 5816
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.4168455204423085,
+      "learning_rate": 7.90537443031424e-06,
+      "loss": 0.2332,
+      "step": 5817
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.5128118323608515,
+      "learning_rate": 7.904641233688202e-06,
+      "loss": 0.1781,
+      "step": 5818
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.423040221725186,
+      "learning_rate": 7.903907942775839e-06,
+      "loss": 0.2173,
+      "step": 5819
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.320580300140151,
+      "learning_rate": 7.903174557600955e-06,
+      "loss": 0.1983,
+      "step": 5820
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.4951111377915092,
+      "learning_rate": 7.902441078187356e-06,
+      "loss": 0.2015,
+      "step": 5821
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 3.15007194330923,
+      "learning_rate": 7.901707504558851e-06,
+      "loss": 0.2351,
+      "step": 5822
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.437261406129813,
+      "learning_rate": 7.90097383673925e-06,
+      "loss": 0.214,
+      "step": 5823
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.3222390542678095,
+      "learning_rate": 7.900240074752372e-06,
+      "loss": 0.1813,
+      "step": 5824
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.2054803816802175,
+      "learning_rate": 7.899506218622032e-06,
+      "loss": 0.2057,
+      "step": 5825
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.364866969943345,
+      "learning_rate": 7.898772268372055e-06,
+      "loss": 0.2147,
+      "step": 5826
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.3725298400494372,
+      "learning_rate": 7.898038224026259e-06,
+      "loss": 0.2397,
+      "step": 5827
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.7382233511206439,
+      "learning_rate": 7.897304085608479e-06,
+      "loss": 0.1689,
+      "step": 5828
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.2162849855830917,
+      "learning_rate": 7.896569853142542e-06,
+      "loss": 0.2273,
+      "step": 5829
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.5707482544687261,
+      "learning_rate": 7.895835526652282e-06,
+      "loss": 0.2262,
+      "step": 5830
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.4275516753188189,
+      "learning_rate": 7.895101106161536e-06,
+      "loss": 0.2473,
+      "step": 5831
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.4485022619371823,
+      "learning_rate": 7.89436659169414e-06,
+      "loss": 0.1819,
+      "step": 5832
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.1219791245819832,
+      "learning_rate": 7.893631983273945e-06,
+      "loss": 0.2272,
+      "step": 5833
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.2578416592908335,
+      "learning_rate": 7.892897280924788e-06,
+      "loss": 0.2698,
+      "step": 5834
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.555581942568174,
+      "learning_rate": 7.892162484670524e-06,
+      "loss": 0.2234,
+      "step": 5835
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.5878724745854444,
+      "learning_rate": 7.891427594535002e-06,
+      "loss": 0.2604,
+      "step": 5836
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.5323998697856243,
+      "learning_rate": 7.890692610542077e-06,
+      "loss": 0.2791,
+      "step": 5837
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.1621269513251569,
+      "learning_rate": 7.889957532715607e-06,
+      "loss": 0.1969,
+      "step": 5838
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.2690546637741464,
+      "learning_rate": 7.889222361079455e-06,
+      "loss": 0.2141,
+      "step": 5839
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.9946803817949741,
+      "learning_rate": 7.888487095657484e-06,
+      "loss": 0.1701,
+      "step": 5840
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.2896607181323128,
+      "learning_rate": 7.887751736473559e-06,
+      "loss": 0.2085,
+      "step": 5841
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.3565721074954595,
+      "learning_rate": 7.887016283551555e-06,
+      "loss": 0.2048,
+      "step": 5842
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.1044163074955642,
+      "learning_rate": 7.88628073691534e-06,
+      "loss": 0.1919,
+      "step": 5843
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.3402922077203756,
+      "learning_rate": 7.885545096588794e-06,
+      "loss": 0.2253,
+      "step": 5844
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.4296343407884033,
+      "learning_rate": 7.884809362595793e-06,
+      "loss": 0.2234,
+      "step": 5845
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.0722246318548074,
+      "learning_rate": 7.884073534960221e-06,
+      "loss": 0.1375,
+      "step": 5846
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.1960139834157062,
+      "learning_rate": 7.883337613705966e-06,
+      "loss": 0.1802,
+      "step": 5847
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.1858360905569427,
+      "learning_rate": 7.88260159885691e-06,
+      "loss": 0.2029,
+      "step": 5848
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.754162923820483,
+      "learning_rate": 7.881865490436954e-06,
+      "loss": 0.1822,
+      "step": 5849
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.5299111955713423,
+      "learning_rate": 7.881129288469985e-06,
+      "loss": 0.1557,
+      "step": 5850
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.2850323185400367,
+      "learning_rate": 7.8803929929799e-06,
+      "loss": 0.1931,
+      "step": 5851
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.056653614105377,
+      "learning_rate": 7.879656603990604e-06,
+      "loss": 0.2151,
+      "step": 5852
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.071471827242674,
+      "learning_rate": 7.878920121525999e-06,
+      "loss": 0.1975,
+      "step": 5853
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.9233877531681168,
+      "learning_rate": 7.87818354560999e-06,
+      "loss": 0.2295,
+      "step": 5854
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.0491640984226291,
+      "learning_rate": 7.87744687626649e-06,
+      "loss": 0.2281,
+      "step": 5855
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.149840276707451,
+      "learning_rate": 7.876710113519407e-06,
+      "loss": 0.1562,
+      "step": 5856
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.2627935924473912,
+      "learning_rate": 7.875973257392664e-06,
+      "loss": 0.2132,
+      "step": 5857
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.3321121001108052,
+      "learning_rate": 7.875236307910172e-06,
+      "loss": 0.2312,
+      "step": 5858
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.0482994626160795,
+      "learning_rate": 7.874499265095857e-06,
+      "loss": 0.1911,
+      "step": 5859
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.2570345617171264,
+      "learning_rate": 7.873762128973641e-06,
+      "loss": 0.2194,
+      "step": 5860
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.2187501020156377,
+      "learning_rate": 7.873024899567457e-06,
+      "loss": 0.1802,
+      "step": 5861
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.1112450483139946,
+      "learning_rate": 7.87228757690123e-06,
+      "loss": 0.2316,
+      "step": 5862
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.382054816390671,
+      "learning_rate": 7.871550160998897e-06,
+      "loss": 0.1997,
+      "step": 5863
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.2306472406658853,
+      "learning_rate": 7.870812651884393e-06,
+      "loss": 0.2236,
+      "step": 5864
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.2014652824261525,
+      "learning_rate": 7.870075049581662e-06,
+      "loss": 0.184,
+      "step": 5865
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.1853865078237555,
+      "learning_rate": 7.869337354114645e-06,
+      "loss": 0.1917,
+      "step": 5866
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.2176965813141751,
+      "learning_rate": 7.868599565507287e-06,
+      "loss": 0.2188,
+      "step": 5867
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.249915767603255,
+      "learning_rate": 7.867861683783537e-06,
+      "loss": 0.1988,
+      "step": 5868
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.3935499318181395,
+      "learning_rate": 7.867123708967349e-06,
+      "loss": 0.2068,
+      "step": 5869
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.14625137372439,
+      "learning_rate": 7.866385641082673e-06,
+      "loss": 0.196,
+      "step": 5870
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.8622068304264079,
+      "learning_rate": 7.865647480153474e-06,
+      "loss": 0.2028,
+      "step": 5871
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.088550357312398,
+      "learning_rate": 7.864909226203709e-06,
+      "loss": 0.2141,
+      "step": 5872
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.3750977352747256,
+      "learning_rate": 7.864170879257343e-06,
+      "loss": 0.201,
+      "step": 5873
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.4011695788949319,
+      "learning_rate": 7.863432439338344e-06,
+      "loss": 0.1911,
+      "step": 5874
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.0094069170260682,
+      "learning_rate": 7.86269390647068e-06,
+      "loss": 0.1712,
+      "step": 5875
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.5542631624510295,
+      "learning_rate": 7.861955280678327e-06,
+      "loss": 0.232,
+      "step": 5876
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.0284186530842205,
+      "learning_rate": 7.861216561985258e-06,
+      "loss": 0.2418,
+      "step": 5877
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.3917892771151354,
+      "learning_rate": 7.860477750415456e-06,
+      "loss": 0.1997,
+      "step": 5878
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.2411738667028207,
+      "learning_rate": 7.8597388459929e-06,
+      "loss": 0.1871,
+      "step": 5879
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.3636268707734283,
+      "learning_rate": 7.858999848741578e-06,
+      "loss": 0.2198,
+      "step": 5880
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.1772563163348295,
+      "learning_rate": 7.858260758685476e-06,
+      "loss": 0.2382,
+      "step": 5881
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.0426109098351433,
+      "learning_rate": 7.857521575848586e-06,
+      "loss": 0.2195,
+      "step": 5882
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.0295844092664819,
+      "learning_rate": 7.856782300254902e-06,
+      "loss": 0.139,
+      "step": 5883
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.1297985760965623,
+      "learning_rate": 7.856042931928422e-06,
+      "loss": 0.222,
+      "step": 5884
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.8915030580655118,
+      "learning_rate": 7.855303470893148e-06,
+      "loss": 0.1868,
+      "step": 5885
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.2600418039442207,
+      "learning_rate": 7.854563917173079e-06,
+      "loss": 0.2449,
+      "step": 5886
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.2560770786420885,
+      "learning_rate": 7.853824270792225e-06,
+      "loss": 0.2508,
+      "step": 5887
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.9008575106098565,
+      "learning_rate": 7.853084531774595e-06,
+      "loss": 0.2269,
+      "step": 5888
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.3440214842077194,
+      "learning_rate": 7.852344700144198e-06,
+      "loss": 0.2027,
+      "step": 5889
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.0513640156201474,
+      "learning_rate": 7.851604775925052e-06,
+      "loss": 0.2141,
+      "step": 5890
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.3826465338717893,
+      "learning_rate": 7.850864759141176e-06,
+      "loss": 0.1925,
+      "step": 5891
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.368732978740005,
+      "learning_rate": 7.850124649816593e-06,
+      "loss": 0.2254,
+      "step": 5892
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.7867576485329646,
+      "learning_rate": 7.849384447975322e-06,
+      "loss": 0.1894,
+      "step": 5893
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.2460589807616334,
+      "learning_rate": 7.848644153641394e-06,
+      "loss": 0.2185,
+      "step": 5894
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.1463612063011357,
+      "learning_rate": 7.847903766838837e-06,
+      "loss": 0.2264,
+      "step": 5895
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.194562838968127,
+      "learning_rate": 7.847163287591691e-06,
+      "loss": 0.2183,
+      "step": 5896
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.7017335019388593,
+      "learning_rate": 7.846422715923982e-06,
+      "loss": 0.2748,
+      "step": 5897
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.9897091627640981,
+      "learning_rate": 7.845682051859757e-06,
+      "loss": 0.2262,
+      "step": 5898
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.3049717724787329,
+      "learning_rate": 7.844941295423058e-06,
+      "loss": 0.2147,
+      "step": 5899
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.8007641111911248,
+      "learning_rate": 7.844200446637925e-06,
+      "loss": 0.2352,
+      "step": 5900
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.9173979161241288,
+      "learning_rate": 7.843459505528413e-06,
+      "loss": 0.1935,
+      "step": 5901
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.6772464045363678,
+      "learning_rate": 7.84271847211857e-06,
+      "loss": 0.1847,
+      "step": 5902
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.1091114120575398,
+      "learning_rate": 7.84197734643245e-06,
+      "loss": 0.2024,
+      "step": 5903
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.8224111844978099,
+      "learning_rate": 7.84123612849411e-06,
+      "loss": 0.1883,
+      "step": 5904
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.0985068066710144,
+      "learning_rate": 7.840494818327613e-06,
+      "loss": 0.2092,
+      "step": 5905
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.9970872337706321,
+      "learning_rate": 7.83975341595702e-06,
+      "loss": 0.1885,
+      "step": 5906
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.5537377240684986,
+      "learning_rate": 7.839011921406398e-06,
+      "loss": 0.2381,
+      "step": 5907
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.0319268812039042,
+      "learning_rate": 7.838270334699818e-06,
+      "loss": 0.2363,
+      "step": 5908
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.2285088567744427,
+      "learning_rate": 7.837528655861349e-06,
+      "loss": 0.2236,
+      "step": 5909
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.0490441268616493,
+      "learning_rate": 7.836786884915069e-06,
+      "loss": 0.2611,
+      "step": 5910
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.0837685619346407,
+      "learning_rate": 7.836045021885055e-06,
+      "loss": 0.2784,
+      "step": 5911
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.9248940284700796,
+      "learning_rate": 7.83530306679539e-06,
+      "loss": 0.2045,
+      "step": 5912
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.9416158292489142,
+      "learning_rate": 7.834561019670154e-06,
+      "loss": 0.1805,
+      "step": 5913
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.8892395260625413,
+      "learning_rate": 7.833818880533441e-06,
+      "loss": 0.1682,
+      "step": 5914
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.961259694862834,
+      "learning_rate": 7.833076649409336e-06,
+      "loss": 0.2258,
+      "step": 5915
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.0237403578036341,
+      "learning_rate": 7.832334326321932e-06,
+      "loss": 0.1964,
+      "step": 5916
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.3013630573952015,
+      "learning_rate": 7.831591911295331e-06,
+      "loss": 0.2439,
+      "step": 5917
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.0008824802006961,
+      "learning_rate": 7.830849404353624e-06,
+      "loss": 0.1771,
+      "step": 5918
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.8256440155992888,
+      "learning_rate": 7.83010680552092e-06,
+      "loss": 0.1984,
+      "step": 5919
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.794034811330162,
+      "learning_rate": 7.829364114821323e-06,
+      "loss": 0.2401,
+      "step": 5920
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.0209291820228141,
+      "learning_rate": 7.828621332278937e-06,
+      "loss": 0.189,
+      "step": 5921
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.2664856074541742,
+      "learning_rate": 7.827878457917876e-06,
+      "loss": 0.2171,
+      "step": 5922
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.8832197705724465,
+      "learning_rate": 7.827135491762256e-06,
+      "loss": 0.2331,
+      "step": 5923
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.1439371129231053,
+      "learning_rate": 7.826392433836191e-06,
+      "loss": 0.2111,
+      "step": 5924
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.5013552047273142,
+      "learning_rate": 7.825649284163803e-06,
+      "loss": 0.266,
+      "step": 5925
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.0123043694367775,
+      "learning_rate": 7.824906042769214e-06,
+      "loss": 0.2248,
+      "step": 5926
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.0098918415336287,
+      "learning_rate": 7.824162709676551e-06,
+      "loss": 0.2015,
+      "step": 5927
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.9560771935593122,
+      "learning_rate": 7.823419284909943e-06,
+      "loss": 0.1438,
+      "step": 5928
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.0908836441984853,
+      "learning_rate": 7.82267576849352e-06,
+      "loss": 0.1577,
+      "step": 5929
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.417540146546797,
+      "learning_rate": 7.821932160451418e-06,
+      "loss": 0.164,
+      "step": 5930
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.8097528314016859,
+      "learning_rate": 7.821188460807776e-06,
+      "loss": 0.149,
+      "step": 5931
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.7593489607718547,
+      "learning_rate": 7.820444669586735e-06,
+      "loss": 0.2329,
+      "step": 5932
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.9808903980552314,
+      "learning_rate": 7.819700786812438e-06,
+      "loss": 0.2088,
+      "step": 5933
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.5087252458328924,
+      "learning_rate": 7.818956812509032e-06,
+      "loss": 0.2059,
+      "step": 5934
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.9329971967288458,
+      "learning_rate": 7.818212746700665e-06,
+      "loss": 0.198,
+      "step": 5935
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.0434772041452205,
+      "learning_rate": 7.817468589411493e-06,
+      "loss": 0.2132,
+      "step": 5936
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.6142194227579645,
+      "learning_rate": 7.816724340665672e-06,
+      "loss": 0.1851,
+      "step": 5937
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.0533185763975181,
+      "learning_rate": 7.815980000487355e-06,
+      "loss": 0.1702,
+      "step": 5938
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.271472692069716,
+      "learning_rate": 7.815235568900712e-06,
+      "loss": 0.2104,
+      "step": 5939
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.3094792828650734,
+      "learning_rate": 7.814491045929902e-06,
+      "loss": 0.222,
+      "step": 5940
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.177506291902643,
+      "learning_rate": 7.813746431599094e-06,
+      "loss": 0.1702,
+      "step": 5941
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.9389341848270831,
+      "learning_rate": 7.81300172593246e-06,
+      "loss": 0.1901,
+      "step": 5942
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.183834657292937,
+      "learning_rate": 7.81225692895417e-06,
+      "loss": 0.1923,
+      "step": 5943
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.047600604114915,
+      "learning_rate": 7.811512040688404e-06,
+      "loss": 0.2161,
+      "step": 5944
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.9646795438117717,
+      "learning_rate": 7.810767061159341e-06,
+      "loss": 0.1846,
+      "step": 5945
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.2218718145163103,
+      "learning_rate": 7.810021990391163e-06,
+      "loss": 0.1665,
+      "step": 5946
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.279317268907903,
+      "learning_rate": 7.809276828408056e-06,
+      "loss": 0.2351,
+      "step": 5947
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.0684966264184708,
+      "learning_rate": 7.808531575234206e-06,
+      "loss": 0.1612,
+      "step": 5948
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.1501938209517544,
+      "learning_rate": 7.807786230893807e-06,
+      "loss": 0.1935,
+      "step": 5949
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.418227461130791,
+      "learning_rate": 7.807040795411052e-06,
+      "loss": 0.2669,
+      "step": 5950
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.5177182092412946,
+      "learning_rate": 7.806295268810139e-06,
+      "loss": 0.2476,
+      "step": 5951
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.8148229980386456,
+      "learning_rate": 7.805549651115267e-06,
+      "loss": 0.2169,
+      "step": 5952
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.089591210683583,
+      "learning_rate": 7.80480394235064e-06,
+      "loss": 0.2011,
+      "step": 5953
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.186374695014241,
+      "learning_rate": 7.804058142540465e-06,
+      "loss": 0.143,
+      "step": 5954
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.7483368489234945,
+      "learning_rate": 7.80331225170895e-06,
+      "loss": 0.2412,
+      "step": 5955
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.0118589773538265,
+      "learning_rate": 7.802566269880307e-06,
+      "loss": 0.1695,
+      "step": 5956
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.1082564124310803,
+      "learning_rate": 7.801820197078753e-06,
+      "loss": 0.1972,
+      "step": 5957
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.9325776855731039,
+      "learning_rate": 7.801074033328501e-06,
+      "loss": 0.2023,
+      "step": 5958
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.3303327834787637,
+      "learning_rate": 7.800327778653778e-06,
+      "loss": 0.2783,
+      "step": 5959
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.2366948796569395,
+      "learning_rate": 7.799581433078805e-06,
+      "loss": 0.1924,
+      "step": 5960
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.0660052375175542,
+      "learning_rate": 7.798834996627807e-06,
+      "loss": 0.1713,
+      "step": 5961
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.1094734272476954,
+      "learning_rate": 7.798088469325016e-06,
+      "loss": 0.2051,
+      "step": 5962
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.2897260634154475,
+      "learning_rate": 7.797341851194664e-06,
+      "loss": 0.2413,
+      "step": 5963
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.189943982772323,
+      "learning_rate": 7.796595142260988e-06,
+      "loss": 0.1543,
+      "step": 5964
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.6380152290341359,
+      "learning_rate": 7.795848342548225e-06,
+      "loss": 0.2194,
+      "step": 5965
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.3354794835317283,
+      "learning_rate": 7.795101452080617e-06,
+      "loss": 0.1989,
+      "step": 5966
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.0169638703596113,
+      "learning_rate": 7.794354470882409e-06,
+      "loss": 0.2042,
+      "step": 5967
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.7098186810641072,
+      "learning_rate": 7.793607398977848e-06,
+      "loss": 0.2296,
+      "step": 5968
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.9512160445655802,
+      "learning_rate": 7.792860236391185e-06,
+      "loss": 0.2356,
+      "step": 5969
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.1899338261390662,
+      "learning_rate": 7.792112983146672e-06,
+      "loss": 0.1669,
+      "step": 5970
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.9645792428668702,
+      "learning_rate": 7.791365639268565e-06,
+      "loss": 0.1452,
+      "step": 5971
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.4315402319707824,
+      "learning_rate": 7.790618204781124e-06,
+      "loss": 0.1936,
+      "step": 5972
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.0070323512009904,
+      "learning_rate": 7.789870679708613e-06,
+      "loss": 0.1947,
+      "step": 5973
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.1140565280834498,
+      "learning_rate": 7.789123064075294e-06,
+      "loss": 0.2348,
+      "step": 5974
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.1839577049189038,
+      "learning_rate": 7.788375357905439e-06,
+      "loss": 0.2081,
+      "step": 5975
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.4078547698945694,
+      "learning_rate": 7.787627561223314e-06,
+      "loss": 0.2472,
+      "step": 5976
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.2459794458442686,
+      "learning_rate": 7.786879674053196e-06,
+      "loss": 0.2345,
+      "step": 5977
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.0988859741138985,
+      "learning_rate": 7.78613169641936e-06,
+      "loss": 0.1947,
+      "step": 5978
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.0392383349578733,
+      "learning_rate": 7.785383628346088e-06,
+      "loss": 0.1874,
+      "step": 5979
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.1964478009842676,
+      "learning_rate": 7.784635469857664e-06,
+      "loss": 0.2344,
+      "step": 5980
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.409031444479326,
+      "learning_rate": 7.78388722097837e-06,
+      "loss": 0.24,
+      "step": 5981
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.4603022760041087,
+      "learning_rate": 7.783138881732495e-06,
+      "loss": 0.2386,
+      "step": 5982
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.9391799041712257,
+      "learning_rate": 7.78239045214433e-06,
+      "loss": 0.1604,
+      "step": 5983
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.4463536546232558,
+      "learning_rate": 7.781641932238172e-06,
+      "loss": 0.1965,
+      "step": 5984
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.9381705925585512,
+      "learning_rate": 7.78089332203832e-06,
+      "loss": 0.11,
+      "step": 5985
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.7443149560690105,
+      "learning_rate": 7.780144621569071e-06,
+      "loss": 0.2243,
+      "step": 5986
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.1219815019034227,
+      "learning_rate": 7.779395830854726e-06,
+      "loss": 0.1811,
+      "step": 5987
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.9242521929716179,
+      "learning_rate": 7.778646949919594e-06,
+      "loss": 0.2176,
+      "step": 5988
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.1466260543051339,
+      "learning_rate": 7.777897978787988e-06,
+      "loss": 0.215,
+      "step": 5989
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.1221301819787755,
+      "learning_rate": 7.777148917484215e-06,
+      "loss": 0.1935,
+      "step": 5990
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.1613614151664833,
+      "learning_rate": 7.77639976603259e-06,
+      "loss": 0.1957,
+      "step": 5991
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.4324673435816606,
+      "learning_rate": 7.775650524457431e-06,
+      "loss": 0.2232,
+      "step": 5992
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.1508526093007856,
+      "learning_rate": 7.774901192783062e-06,
+      "loss": 0.158,
+      "step": 5993
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.137688262655657,
+      "learning_rate": 7.774151771033802e-06,
+      "loss": 0.1437,
+      "step": 5994
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.9209969104855852,
+      "learning_rate": 7.773402259233982e-06,
+      "loss": 0.2028,
+      "step": 5995
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.1842373193592832,
+      "learning_rate": 7.77265265740793e-06,
+      "loss": 0.206,
+      "step": 5996
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.3163703685760437,
+      "learning_rate": 7.771902965579978e-06,
+      "loss": 0.2097,
+      "step": 5997
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.9427791858843201,
+      "learning_rate": 7.771153183774463e-06,
+      "loss": 0.2223,
+      "step": 5998
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.2610375850928541,
+      "learning_rate": 7.77040331201572e-06,
+      "loss": 0.2057,
+      "step": 5999
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.2288222498840904,
+      "learning_rate": 7.769653350328095e-06,
+      "loss": 0.2359,
+      "step": 6000
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.2535686201766956,
+      "learning_rate": 7.768903298735928e-06,
+      "loss": 0.1916,
+      "step": 6001
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.1877618488430541,
+      "learning_rate": 7.768153157263567e-06,
+      "loss": 0.1842,
+      "step": 6002
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.104665403671172,
+      "learning_rate": 7.767402925935364e-06,
+      "loss": 0.1971,
+      "step": 6003
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.2740013360412483,
+      "learning_rate": 7.76665260477567e-06,
+      "loss": 0.168,
+      "step": 6004
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.0797525746488599,
+      "learning_rate": 7.765902193808841e-06,
+      "loss": 0.1778,
+      "step": 6005
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.8930756019381652,
+      "learning_rate": 7.765151693059236e-06,
+      "loss": 0.1444,
+      "step": 6006
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.7505630434635966,
+      "learning_rate": 7.764401102551218e-06,
+      "loss": 0.2744,
+      "step": 6007
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.1836782998079414,
+      "learning_rate": 7.763650422309152e-06,
+      "loss": 0.2351,
+      "step": 6008
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.982574126407658,
+      "learning_rate": 7.762899652357402e-06,
+      "loss": 0.1993,
+      "step": 6009
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.2022387906787637,
+      "learning_rate": 7.76214879272034e-06,
+      "loss": 0.215,
+      "step": 6010
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.0927071061171387,
+      "learning_rate": 7.761397843422342e-06,
+      "loss": 0.1834,
+      "step": 6011
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.330275686158803,
+      "learning_rate": 7.76064680448778e-06,
+      "loss": 0.2519,
+      "step": 6012
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.2527708650156795,
+      "learning_rate": 7.759895675941036e-06,
+      "loss": 0.1684,
+      "step": 6013
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.7550459879091272,
+      "learning_rate": 7.759144457806492e-06,
+      "loss": 0.1542,
+      "step": 6014
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.2155140931596427,
+      "learning_rate": 7.758393150108532e-06,
+      "loss": 0.2284,
+      "step": 6015
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.9831947439592029,
+      "learning_rate": 7.757641752871543e-06,
+      "loss": 0.2219,
+      "step": 6016
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.5330846461301018,
+      "learning_rate": 7.756890266119918e-06,
+      "loss": 0.196,
+      "step": 6017
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.1617256380478553,
+      "learning_rate": 7.75613868987805e-06,
+      "loss": 0.2547,
+      "step": 6018
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.2874043521722203,
+      "learning_rate": 7.755387024170334e-06,
+      "loss": 0.2349,
+      "step": 6019
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.3821392265672496,
+      "learning_rate": 7.754635269021172e-06,
+      "loss": 0.2348,
+      "step": 6020
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.2331125074460474,
+      "learning_rate": 7.753883424454965e-06,
+      "loss": 0.1937,
+      "step": 6021
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.7857567829670097,
+      "learning_rate": 7.753131490496117e-06,
+      "loss": 0.1397,
+      "step": 6022
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.3750933607332485,
+      "learning_rate": 7.75237946716904e-06,
+      "loss": 0.1933,
+      "step": 6023
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.1039547274474797,
+      "learning_rate": 7.751627354498141e-06,
+      "loss": 0.2231,
+      "step": 6024
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.1423126132284307,
+      "learning_rate": 7.750875152507835e-06,
+      "loss": 0.1891,
+      "step": 6025
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.3463352446145653,
+      "learning_rate": 7.750122861222542e-06,
+      "loss": 0.211,
+      "step": 6026
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.9222284544612259,
+      "learning_rate": 7.74937048066668e-06,
+      "loss": 0.1526,
+      "step": 6027
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.9903538385444149,
+      "learning_rate": 7.748618010864668e-06,
+      "loss": 0.1783,
+      "step": 6028
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.3957387654333897,
+      "learning_rate": 7.747865451840937e-06,
+      "loss": 0.2274,
+      "step": 6029
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.314985221418948,
+      "learning_rate": 7.747112803619912e-06,
+      "loss": 0.1579,
+      "step": 6030
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.3701504780102582,
+      "learning_rate": 7.746360066226026e-06,
+      "loss": 0.1849,
+      "step": 6031
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.9979219069132598,
+      "learning_rate": 7.745607239683715e-06,
+      "loss": 0.22,
+      "step": 6032
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.8247756251915629,
+      "learning_rate": 7.74485432401741e-06,
+      "loss": 0.2235,
+      "step": 6033
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.1810068117319967,
+      "learning_rate": 7.74410131925156e-06,
+      "loss": 0.2015,
+      "step": 6034
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.2912513933146548,
+      "learning_rate": 7.743348225410601e-06,
+      "loss": 0.1539,
+      "step": 6035
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.8680648060411716,
+      "learning_rate": 7.742595042518981e-06,
+      "loss": 0.1696,
+      "step": 6036
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.3183694966245574,
+      "learning_rate": 7.741841770601148e-06,
+      "loss": 0.2486,
+      "step": 6037
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.2312282171234703,
+      "learning_rate": 7.741088409681556e-06,
+      "loss": 0.2084,
+      "step": 6038
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.9991808624630522,
+      "learning_rate": 7.740334959784656e-06,
+      "loss": 0.2131,
+      "step": 6039
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.0922895217419468,
+      "learning_rate": 7.73958142093491e-06,
+      "loss": 0.1666,
+      "step": 6040
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.3079466038030854,
+      "learning_rate": 7.738827793156773e-06,
+      "loss": 0.185,
+      "step": 6041
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.6632826280565765,
+      "learning_rate": 7.738074076474714e-06,
+      "loss": 0.2274,
+      "step": 6042
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.9694220765006961,
+      "learning_rate": 7.737320270913193e-06,
+      "loss": 0.249,
+      "step": 6043
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.2064450639919317,
+      "learning_rate": 7.736566376496684e-06,
+      "loss": 0.1701,
+      "step": 6044
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.880133755005519,
+      "learning_rate": 7.735812393249656e-06,
+      "loss": 0.1639,
+      "step": 6045
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.9844799439938398,
+      "learning_rate": 7.735058321196584e-06,
+      "loss": 0.1968,
+      "step": 6046
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.2054183130473355,
+      "learning_rate": 7.734304160361947e-06,
+      "loss": 0.2091,
+      "step": 6047
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.1374105301715751,
+      "learning_rate": 7.733549910770224e-06,
+      "loss": 0.2083,
+      "step": 6048
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.3032894953499023,
+      "learning_rate": 7.732795572445899e-06,
+      "loss": 0.2306,
+      "step": 6049
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.4545313735996284,
+      "learning_rate": 7.73204114541346e-06,
+      "loss": 0.2597,
+      "step": 6050
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.2081334257695426,
+      "learning_rate": 7.731286629697392e-06,
+      "loss": 0.2337,
+      "step": 6051
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.2619856841393586,
+      "learning_rate": 7.73053202532219e-06,
+      "loss": 0.2184,
+      "step": 6052
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.4268692659767455,
+      "learning_rate": 7.729777332312349e-06,
+      "loss": 0.2011,
+      "step": 6053
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.036830828773782,
+      "learning_rate": 7.729022550692365e-06,
+      "loss": 0.1633,
+      "step": 6054
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.936630527312407,
+      "learning_rate": 7.72826768048674e-06,
+      "loss": 0.1832,
+      "step": 6055
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.2691476212928325,
+      "learning_rate": 7.727512721719979e-06,
+      "loss": 0.2091,
+      "step": 6056
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.321370932163559,
+      "learning_rate": 7.726757674416583e-06,
+      "loss": 0.2217,
+      "step": 6057
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.333534248557682,
+      "learning_rate": 7.726002538601069e-06,
+      "loss": 0.2597,
+      "step": 6058
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.1213280801740504,
+      "learning_rate": 7.725247314297943e-06,
+      "loss": 0.1709,
+      "step": 6059
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.465236590797419,
+      "learning_rate": 7.724492001531722e-06,
+      "loss": 0.2013,
+      "step": 6060
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.96884607655865,
+      "learning_rate": 7.723736600326923e-06,
+      "loss": 0.2178,
+      "step": 6061
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.2427494739198695,
+      "learning_rate": 7.72298111070807e-06,
+      "loss": 0.1573,
+      "step": 6062
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.4835032161752355,
+      "learning_rate": 7.722225532699683e-06,
+      "loss": 0.2126,
+      "step": 6063
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.2410632780229875,
+      "learning_rate": 7.72146986632629e-06,
+      "loss": 0.2367,
+      "step": 6064
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.3202775916727152,
+      "learning_rate": 7.720714111612419e-06,
+      "loss": 0.2002,
+      "step": 6065
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.648412225176959,
+      "learning_rate": 7.719958268582605e-06,
+      "loss": 0.2396,
+      "step": 6066
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.4468746944328297,
+      "learning_rate": 7.719202337261382e-06,
+      "loss": 0.1975,
+      "step": 6067
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.187269919467315,
+      "learning_rate": 7.718446317673286e-06,
+      "loss": 0.1337,
+      "step": 6068
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.0402908199580247,
+      "learning_rate": 7.717690209842858e-06,
+      "loss": 0.1673,
+      "step": 6069
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.1581109880334688,
+      "learning_rate": 7.716934013794645e-06,
+      "loss": 0.1916,
+      "step": 6070
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.2417045260131077,
+      "learning_rate": 7.716177729553192e-06,
+      "loss": 0.2214,
+      "step": 6071
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.0829876449812323,
+      "learning_rate": 7.71542135714305e-06,
+      "loss": 0.1722,
+      "step": 6072
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.2712658871677278,
+      "learning_rate": 7.714664896588763e-06,
+      "loss": 0.1834,
+      "step": 6073
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.4708813715061095,
+      "learning_rate": 7.713908347914898e-06,
+      "loss": 0.2541,
+      "step": 6074
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.1656876395067288,
+      "learning_rate": 7.713151711146005e-06,
+      "loss": 0.2006,
+      "step": 6075
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.0326450802300968,
+      "learning_rate": 7.712394986306649e-06,
+      "loss": 0.1969,
+      "step": 6076
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.3973759865947986,
+      "learning_rate": 7.71163817342139e-06,
+      "loss": 0.2112,
+      "step": 6077
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.6196183189066748,
+      "learning_rate": 7.710881272514796e-06,
+      "loss": 0.2353,
+      "step": 6078
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.1844461789546146,
+      "learning_rate": 7.71012428361144e-06,
+      "loss": 0.2105,
+      "step": 6079
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.0983821362506385,
+      "learning_rate": 7.70936720673589e-06,
+      "loss": 0.1613,
+      "step": 6080
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.765360246911758,
+      "learning_rate": 7.70861004191272e-06,
+      "loss": 0.2555,
+      "step": 6081
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.137298950437645,
+      "learning_rate": 7.707852789166514e-06,
+      "loss": 0.2293,
+      "step": 6082
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.9928624849611289,
+      "learning_rate": 7.70709544852185e-06,
+      "loss": 0.1747,
+      "step": 6083
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.2795601115323936,
+      "learning_rate": 7.706338020003308e-06,
+      "loss": 0.2164,
+      "step": 6084
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.9923915915953946,
+      "learning_rate": 7.705580503635477e-06,
+      "loss": 0.2024,
+      "step": 6085
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.9731973096868703,
+      "learning_rate": 7.70482289944295e-06,
+      "loss": 0.204,
+      "step": 6086
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.123226675484634,
+      "learning_rate": 7.704065207450313e-06,
+      "loss": 0.216,
+      "step": 6087
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.9840429044809463,
+      "learning_rate": 7.703307427682167e-06,
+      "loss": 0.2277,
+      "step": 6088
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.0064337932772123,
+      "learning_rate": 7.702549560163107e-06,
+      "loss": 0.2063,
+      "step": 6089
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.2854887204736243,
+      "learning_rate": 7.701791604917734e-06,
+      "loss": 0.1992,
+      "step": 6090
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.5739974820081741,
+      "learning_rate": 7.701033561970653e-06,
+      "loss": 0.203,
+      "step": 6091
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.9584316146991025,
+      "learning_rate": 7.700275431346466e-06,
+      "loss": 0.2272,
+      "step": 6092
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.2799943260350235,
+      "learning_rate": 7.699517213069789e-06,
+      "loss": 0.1796,
+      "step": 6093
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.9778983516255292,
+      "learning_rate": 7.698758907165229e-06,
+      "loss": 0.2141,
+      "step": 6094
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.3663889364756392,
+      "learning_rate": 7.698000513657405e-06,
+      "loss": 0.194,
+      "step": 6095
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.0382232036969836,
+      "learning_rate": 7.697242032570932e-06,
+      "loss": 0.1658,
+      "step": 6096
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.7602103208100675,
+      "learning_rate": 7.696483463930432e-06,
+      "loss": 0.2335,
+      "step": 6097
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.9633232966835029,
+      "learning_rate": 7.695724807760526e-06,
+      "loss": 0.1616,
+      "step": 6098
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.34806480572652,
+      "learning_rate": 7.694966064085844e-06,
+      "loss": 0.1726,
+      "step": 6099
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.0087089580626818,
+      "learning_rate": 7.694207232931015e-06,
+      "loss": 0.1438,
+      "step": 6100
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.9550538508523045,
+      "learning_rate": 7.693448314320668e-06,
+      "loss": 0.2088,
+      "step": 6101
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.3458726939428454,
+      "learning_rate": 7.69268930827944e-06,
+      "loss": 0.2007,
+      "step": 6102
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.2452572807336773,
+      "learning_rate": 7.691930214831969e-06,
+      "loss": 0.1424,
+      "step": 6103
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.9435893246470648,
+      "learning_rate": 7.691171034002897e-06,
+      "loss": 0.2157,
+      "step": 6104
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.1782980308658062,
+      "learning_rate": 7.690411765816864e-06,
+      "loss": 0.2265,
+      "step": 6105
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.1130990130193164,
+      "learning_rate": 7.689652410298518e-06,
+      "loss": 0.2449,
+      "step": 6106
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.5477949481080995,
+      "learning_rate": 7.68889296747251e-06,
+      "loss": 0.196,
+      "step": 6107
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.01166775659609,
+      "learning_rate": 7.688133437363487e-06,
+      "loss": 0.1666,
+      "step": 6108
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.0590268749895921,
+      "learning_rate": 7.68737381999611e-06,
+      "loss": 0.1909,
+      "step": 6109
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.95234168805845,
+      "learning_rate": 7.68661411539503e-06,
+      "loss": 0.1693,
+      "step": 6110
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.421477828255778,
+      "learning_rate": 7.685854323584912e-06,
+      "loss": 0.1835,
+      "step": 6111
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.4850706528303022,
+      "learning_rate": 7.68509444459042e-06,
+      "loss": 0.1976,
+      "step": 6112
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.1125905922908463,
+      "learning_rate": 7.684334478436218e-06,
+      "loss": 0.2128,
+      "step": 6113
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.9207126150827651,
+      "learning_rate": 7.683574425146974e-06,
+      "loss": 0.212,
+      "step": 6114
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.2793382288531494,
+      "learning_rate": 7.682814284747363e-06,
+      "loss": 0.1951,
+      "step": 6115
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.9102131412391055,
+      "learning_rate": 7.682054057262054e-06,
+      "loss": 0.1823,
+      "step": 6116
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.1946492115678453,
+      "learning_rate": 7.68129374271573e-06,
+      "loss": 0.259,
+      "step": 6117
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.064942945420641,
+      "learning_rate": 7.680533341133069e-06,
+      "loss": 0.1516,
+      "step": 6118
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.432989936603774,
+      "learning_rate": 7.679772852538754e-06,
+      "loss": 0.2576,
+      "step": 6119
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.3197401205543247,
+      "learning_rate": 7.679012276957471e-06,
+      "loss": 0.2325,
+      "step": 6120
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.1834071865284084,
+      "learning_rate": 7.67825161441391e-06,
+      "loss": 0.1642,
+      "step": 6121
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.2214790266351139,
+      "learning_rate": 7.677490864932763e-06,
+      "loss": 0.1732,
+      "step": 6122
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.9382109231344442,
+      "learning_rate": 7.67673002853872e-06,
+      "loss": 0.1813,
+      "step": 6123
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.1735852818633674,
+      "learning_rate": 7.675969105256482e-06,
+      "loss": 0.2381,
+      "step": 6124
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.0751779123504235,
+      "learning_rate": 7.675208095110749e-06,
+      "loss": 0.191,
+      "step": 6125
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.9044522281782982,
+      "learning_rate": 7.674446998126222e-06,
+      "loss": 0.1645,
+      "step": 6126
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.826090887622842,
+      "learning_rate": 7.673685814327607e-06,
+      "loss": 0.1991,
+      "step": 6127
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.9460939633363692,
+      "learning_rate": 7.672924543739614e-06,
+      "loss": 0.1987,
+      "step": 6128
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.1965373244853825,
+      "learning_rate": 7.672163186386952e-06,
+      "loss": 0.2595,
+      "step": 6129
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.2252984336633257,
+      "learning_rate": 7.671401742294337e-06,
+      "loss": 0.2362,
+      "step": 6130
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.9234091048385173,
+      "learning_rate": 7.670640211486485e-06,
+      "loss": 0.1823,
+      "step": 6131
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.1839109977605824,
+      "learning_rate": 7.669878593988116e-06,
+      "loss": 0.2174,
+      "step": 6132
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.7049636866829331,
+      "learning_rate": 7.669116889823955e-06,
+      "loss": 0.126,
+      "step": 6133
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.087965835971154,
+      "learning_rate": 7.668355099018722e-06,
+      "loss": 0.2057,
+      "step": 6134
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.2590388441597447,
+      "learning_rate": 7.66759322159715e-06,
+      "loss": 0.2281,
+      "step": 6135
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.2456953194996605,
+      "learning_rate": 7.666831257583967e-06,
+      "loss": 0.2641,
+      "step": 6136
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.4558175531577302,
+      "learning_rate": 7.666069207003905e-06,
+      "loss": 0.2376,
+      "step": 6137
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.002886122116722,
+      "learning_rate": 7.665307069881709e-06,
+      "loss": 0.1875,
+      "step": 6138
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.8523736325680407,
+      "learning_rate": 7.66454484624211e-06,
+      "loss": 0.1654,
+      "step": 6139
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.201593802556529,
+      "learning_rate": 7.663782536109855e-06,
+      "loss": 0.1397,
+      "step": 6140
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.079826395921049,
+      "learning_rate": 7.663020139509683e-06,
+      "loss": 0.1553,
+      "step": 6141
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.9558755465216364,
+      "learning_rate": 7.662257656466348e-06,
+      "loss": 0.1878,
+      "step": 6142
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.0172122271853283,
+      "learning_rate": 7.6614950870046e-06,
+      "loss": 0.2026,
+      "step": 6143
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.1078073580828351,
+      "learning_rate": 7.66073243114919e-06,
+      "loss": 0.2482,
+      "step": 6144
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.0915647457846693,
+      "learning_rate": 7.659969688924872e-06,
+      "loss": 0.1915,
+      "step": 6145
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.329341221461172,
+      "learning_rate": 7.659206860356412e-06,
+      "loss": 0.2446,
+      "step": 6146
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.5085216464241002,
+      "learning_rate": 7.65844394546857e-06,
+      "loss": 0.2191,
+      "step": 6147
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.3595249296137233,
+      "learning_rate": 7.657680944286106e-06,
+      "loss": 0.2228,
+      "step": 6148
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.1228237660249183,
+      "learning_rate": 7.65691785683379e-06,
+      "loss": 0.1946,
+      "step": 6149
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.0751038234359513,
+      "learning_rate": 7.656154683136392e-06,
+      "loss": 0.2173,
+      "step": 6150
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.7985680606489767,
+      "learning_rate": 7.655391423218687e-06,
+      "loss": 0.2283,
+      "step": 6151
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.5905706890067324,
+      "learning_rate": 7.654628077105451e-06,
+      "loss": 0.1992,
+      "step": 6152
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.4339562576205855,
+      "learning_rate": 7.653864644821457e-06,
+      "loss": 0.207,
+      "step": 6153
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.1766429544325367,
+      "learning_rate": 7.653101126391494e-06,
+      "loss": 0.1897,
+      "step": 6154
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.120078624138006,
+      "learning_rate": 7.65233752184034e-06,
+      "loss": 0.1803,
+      "step": 6155
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 3.3186354766453885,
+      "learning_rate": 7.651573831192788e-06,
+      "loss": 0.2263,
+      "step": 6156
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.0652113777366827,
+      "learning_rate": 7.650810054473623e-06,
+      "loss": 0.1586,
+      "step": 6157
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.185947782486981,
+      "learning_rate": 7.650046191707641e-06,
+      "loss": 0.1973,
+      "step": 6158
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.3748624201782698,
+      "learning_rate": 7.649282242919634e-06,
+      "loss": 0.2289,
+      "step": 6159
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.9515215683262584,
+      "learning_rate": 7.648518208134402e-06,
+      "loss": 0.1719,
+      "step": 6160
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.1966655263409962,
+      "learning_rate": 7.647754087376745e-06,
+      "loss": 0.1838,
+      "step": 6161
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 2.5568042932961608,
+      "learning_rate": 7.64698988067147e-06,
+      "loss": 0.1897,
+      "step": 6162
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.9328123777156219,
+      "learning_rate": 7.646225588043379e-06,
+      "loss": 0.2162,
+      "step": 6163
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.4008523799582673,
+      "learning_rate": 7.645461209517284e-06,
+      "loss": 0.2101,
+      "step": 6164
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.4695016276949875,
+      "learning_rate": 7.644696745118e-06,
+      "loss": 0.1901,
+      "step": 6165
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.2743138416389304,
+      "learning_rate": 7.643932194870337e-06,
+      "loss": 0.2497,
+      "step": 6166
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.1013526073991862,
+      "learning_rate": 7.643167558799114e-06,
+      "loss": 0.2165,
+      "step": 6167
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.1616331807774845,
+      "learning_rate": 7.64240283692915e-06,
+      "loss": 0.1872,
+      "step": 6168
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.189985898218828,
+      "learning_rate": 7.641638029285274e-06,
+      "loss": 0.1947,
+      "step": 6169
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.3357489798638504,
+      "learning_rate": 7.640873135892307e-06,
+      "loss": 0.2142,
+      "step": 6170
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.2416435916274495,
+      "learning_rate": 7.640108156775082e-06,
+      "loss": 0.2091,
+      "step": 6171
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.1087704920528125,
+      "learning_rate": 7.639343091958424e-06,
+      "loss": 0.1951,
+      "step": 6172
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.4192209862054421,
+      "learning_rate": 7.638577941467172e-06,
+      "loss": 0.1971,
+      "step": 6173
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.0657688478666274,
+      "learning_rate": 7.637812705326164e-06,
+      "loss": 0.1875,
+      "step": 6174
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.0569223726366521,
+      "learning_rate": 7.63704738356024e-06,
+      "loss": 0.1845,
+      "step": 6175
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.1382804953730379,
+      "learning_rate": 7.636281976194241e-06,
+      "loss": 0.1802,
+      "step": 6176
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.5260924099898427,
+      "learning_rate": 7.635516483253012e-06,
+      "loss": 0.2287,
+      "step": 6177
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.959012395328982,
+      "learning_rate": 7.634750904761404e-06,
+      "loss": 0.1398,
+      "step": 6178
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.6089289797151385,
+      "learning_rate": 7.633985240744264e-06,
+      "loss": 0.2666,
+      "step": 6179
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.1159869968119183,
+      "learning_rate": 7.63321949122645e-06,
+      "loss": 0.1855,
+      "step": 6180
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.0619986925751286,
+      "learning_rate": 7.632453656232816e-06,
+      "loss": 0.1932,
+      "step": 6181
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.1761835746981397,
+      "learning_rate": 7.631687735788222e-06,
+      "loss": 0.165,
+      "step": 6182
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.19911191507917,
+      "learning_rate": 7.630921729917533e-06,
+      "loss": 0.2109,
+      "step": 6183
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.5536145240893335,
+      "learning_rate": 7.63015563864561e-06,
+      "loss": 0.2269,
+      "step": 6184
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.0225836262149497,
+      "learning_rate": 7.629389461997325e-06,
+      "loss": 0.2213,
+      "step": 6185
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.9077958701312716,
+      "learning_rate": 7.628623199997545e-06,
+      "loss": 0.1896,
+      "step": 6186
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.562445163360595,
+      "learning_rate": 7.627856852671142e-06,
+      "loss": 0.2411,
+      "step": 6187
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.186140983072246,
+      "learning_rate": 7.627090420042998e-06,
+      "loss": 0.198,
+      "step": 6188
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.9885432928776738,
+      "learning_rate": 7.6263239021379866e-06,
+      "loss": 0.2402,
+      "step": 6189
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.7428123428824407,
+      "learning_rate": 7.62555729898099e-06,
+      "loss": 0.1781,
+      "step": 6190
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.2060708604724386,
+      "learning_rate": 7.624790610596895e-06,
+      "loss": 0.2424,
+      "step": 6191
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.0514140333198516,
+      "learning_rate": 7.624023837010586e-06,
+      "loss": 0.2623,
+      "step": 6192
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.0593467740290565,
+      "learning_rate": 7.623256978246957e-06,
+      "loss": 0.22,
+      "step": 6193
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.1429146784510038,
+      "learning_rate": 7.622490034330896e-06,
+      "loss": 0.2535,
+      "step": 6194
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.3127814201239183,
+      "learning_rate": 7.621723005287301e-06,
+      "loss": 0.2231,
+      "step": 6195
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.8558221254813159,
+      "learning_rate": 7.6209558911410705e-06,
+      "loss": 0.1658,
+      "step": 6196
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.8560298664551276,
+      "learning_rate": 7.620188691917103e-06,
+      "loss": 0.2315,
+      "step": 6197
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.1264221983643232,
+      "learning_rate": 7.619421407640304e-06,
+      "loss": 0.1437,
+      "step": 6198
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.4533805007272722,
+      "learning_rate": 7.61865403833558e-06,
+      "loss": 0.2449,
+      "step": 6199
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.2668320579183576,
+      "learning_rate": 7.61788658402784e-06,
+      "loss": 0.1965,
+      "step": 6200
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.34520619522887,
+      "learning_rate": 7.617119044741997e-06,
+      "loss": 0.2445,
+      "step": 6201
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.113775683916583,
+      "learning_rate": 7.616351420502966e-06,
+      "loss": 0.2073,
+      "step": 6202
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.1358500422880067,
+      "learning_rate": 7.615583711335661e-06,
+      "loss": 0.1926,
+      "step": 6203
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.966122482596854,
+      "learning_rate": 7.614815917265006e-06,
+      "loss": 0.2223,
+      "step": 6204
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.2919476480512666,
+      "learning_rate": 7.614048038315922e-06,
+      "loss": 0.2444,
+      "step": 6205
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.33728676415265,
+      "learning_rate": 7.613280074513335e-06,
+      "loss": 0.237,
+      "step": 6206
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.3163304512263283,
+      "learning_rate": 7.612512025882175e-06,
+      "loss": 0.2254,
+      "step": 6207
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.2573702209353201,
+      "learning_rate": 7.611743892447372e-06,
+      "loss": 0.2022,
+      "step": 6208
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.153120503248839,
+      "learning_rate": 7.610975674233859e-06,
+      "loss": 0.2516,
+      "step": 6209
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.3044523923869216,
+      "learning_rate": 7.610207371266576e-06,
+      "loss": 0.2158,
+      "step": 6210
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.241159742666667,
+      "learning_rate": 7.609438983570461e-06,
+      "loss": 0.2082,
+      "step": 6211
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.8556820046242344,
+      "learning_rate": 7.608670511170456e-06,
+      "loss": 0.1838,
+      "step": 6212
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.292486280854073,
+      "learning_rate": 7.6079019540915055e-06,
+      "loss": 0.1726,
+      "step": 6213
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.2914531733854968,
+      "learning_rate": 7.607133312358558e-06,
+      "loss": 0.2147,
+      "step": 6214
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.8846155380375277,
+      "learning_rate": 7.6063645859965625e-06,
+      "loss": 0.1716,
+      "step": 6215
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.3923397341680894,
+      "learning_rate": 7.605595775030476e-06,
+      "loss": 0.2237,
+      "step": 6216
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.125845561549959,
+      "learning_rate": 7.604826879485252e-06,
+      "loss": 0.2249,
+      "step": 6217
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.0122216164273732,
+      "learning_rate": 7.6040578993858485e-06,
+      "loss": 0.1751,
+      "step": 6218
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.992340634104162,
+      "learning_rate": 7.6032888347572285e-06,
+      "loss": 0.1897,
+      "step": 6219
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.111949409266117,
+      "learning_rate": 7.602519685624356e-06,
+      "loss": 0.226,
+      "step": 6220
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.1850314677512817,
+      "learning_rate": 7.6017504520122e-06,
+      "loss": 0.2367,
+      "step": 6221
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.0612939430351211,
+      "learning_rate": 7.600981133945725e-06,
+      "loss": 0.2101,
+      "step": 6222
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.0470106567198671,
+      "learning_rate": 7.600211731449909e-06,
+      "loss": 0.2012,
+      "step": 6223
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.95025566160344,
+      "learning_rate": 7.599442244549723e-06,
+      "loss": 0.2444,
+      "step": 6224
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.8422346061243533,
+      "learning_rate": 7.5986726732701485e-06,
+      "loss": 0.164,
+      "step": 6225
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.9200678950438502,
+      "learning_rate": 7.597903017636163e-06,
+      "loss": 0.1886,
+      "step": 6226
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.8641584188344903,
+      "learning_rate": 7.5971332776727525e-06,
+      "loss": 0.2331,
+      "step": 6227
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.870324928838717,
+      "learning_rate": 7.596363453404904e-06,
+      "loss": 0.1999,
+      "step": 6228
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.1620524416529097,
+      "learning_rate": 7.595593544857602e-06,
+      "loss": 0.2179,
+      "step": 6229
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.1301563726367352,
+      "learning_rate": 7.594823552055844e-06,
+      "loss": 0.2128,
+      "step": 6230
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.3048308002558835,
+      "learning_rate": 7.594053475024619e-06,
+      "loss": 0.212,
+      "step": 6231
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.6281408300014615,
+      "learning_rate": 7.593283313788929e-06,
+      "loss": 0.2829,
+      "step": 6232
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.9923039728420462,
+      "learning_rate": 7.592513068373768e-06,
+      "loss": 0.2032,
+      "step": 6233
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.074838809609098,
+      "learning_rate": 7.591742738804145e-06,
+      "loss": 0.1849,
+      "step": 6234
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.2493386694753006,
+      "learning_rate": 7.590972325105059e-06,
+      "loss": 0.2088,
+      "step": 6235
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.0734949222209405,
+      "learning_rate": 7.590201827301524e-06,
+      "loss": 0.1426,
+      "step": 6236
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.8855749222563091,
+      "learning_rate": 7.589431245418548e-06,
+      "loss": 0.1098,
+      "step": 6237
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.0105342568183713,
+      "learning_rate": 7.5886605794811445e-06,
+      "loss": 0.2293,
+      "step": 6238
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.077718782605944,
+      "learning_rate": 7.58788982951433e-06,
+      "loss": 0.2647,
+      "step": 6239
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.3476978461438918,
+      "learning_rate": 7.587118995543124e-06,
+      "loss": 0.231,
+      "step": 6240
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.0541573675628801,
+      "learning_rate": 7.5863480775925465e-06,
+      "loss": 0.212,
+      "step": 6241
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.3316370598926306,
+      "learning_rate": 7.585577075687624e-06,
+      "loss": 0.1868,
+      "step": 6242
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.6051450087388452,
+      "learning_rate": 7.5848059898533825e-06,
+      "loss": 0.2262,
+      "step": 6243
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.38487182446696,
+      "learning_rate": 7.584034820114852e-06,
+      "loss": 0.25,
+      "step": 6244
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.0040489227129072,
+      "learning_rate": 7.583263566497064e-06,
+      "loss": 0.2123,
+      "step": 6245
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.074851537207116,
+      "learning_rate": 7.582492229025058e-06,
+      "loss": 0.1934,
+      "step": 6246
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.534802893625208,
+      "learning_rate": 7.581720807723868e-06,
+      "loss": 0.2522,
+      "step": 6247
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.0228947727399265,
+      "learning_rate": 7.5809493026185365e-06,
+      "loss": 0.1485,
+      "step": 6248
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.140563969743024,
+      "learning_rate": 7.580177713734105e-06,
+      "loss": 0.249,
+      "step": 6249
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.8877469431777978,
+      "learning_rate": 7.579406041095622e-06,
+      "loss": 0.1641,
+      "step": 6250
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.112906369033989,
+      "learning_rate": 7.578634284728136e-06,
+      "loss": 0.177,
+      "step": 6251
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.2379793211174943,
+      "learning_rate": 7.5778624446566975e-06,
+      "loss": 0.1867,
+      "step": 6252
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.9608327568158066,
+      "learning_rate": 7.577090520906363e-06,
+      "loss": 0.2308,
+      "step": 6253
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.7012417507419529,
+      "learning_rate": 7.576318513502187e-06,
+      "loss": 0.1716,
+      "step": 6254
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.0931850660336369,
+      "learning_rate": 7.575546422469231e-06,
+      "loss": 0.1826,
+      "step": 6255
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.2230331278795639,
+      "learning_rate": 7.574774247832557e-06,
+      "loss": 0.204,
+      "step": 6256
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.3823412325184947,
+      "learning_rate": 7.574001989617231e-06,
+      "loss": 0.2279,
+      "step": 6257
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.2195835105993396,
+      "learning_rate": 7.573229647848319e-06,
+      "loss": 0.243,
+      "step": 6258
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.1312033026498955,
+      "learning_rate": 7.572457222550895e-06,
+      "loss": 0.2435,
+      "step": 6259
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.0663040849283079,
+      "learning_rate": 7.571684713750027e-06,
+      "loss": 0.2204,
+      "step": 6260
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.020935233740952,
+      "learning_rate": 7.570912121470797e-06,
+      "loss": 0.2104,
+      "step": 6261
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.9061362260893223,
+      "learning_rate": 7.57013944573828e-06,
+      "loss": 0.1661,
+      "step": 6262
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.9846266545946128,
+      "learning_rate": 7.569366686577559e-06,
+      "loss": 0.1504,
+      "step": 6263
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.9619451953477769,
+      "learning_rate": 7.5685938440137185e-06,
+      "loss": 0.2048,
+      "step": 6264
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.1582911186178657,
+      "learning_rate": 7.567820918071844e-06,
+      "loss": 0.2052,
+      "step": 6265
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.55686653847962,
+      "learning_rate": 7.567047908777027e-06,
+      "loss": 0.2208,
+      "step": 6266
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.0504133296511549,
+      "learning_rate": 7.566274816154357e-06,
+      "loss": 0.2461,
+      "step": 6267
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.292093631974385,
+      "learning_rate": 7.565501640228932e-06,
+      "loss": 0.168,
+      "step": 6268
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.122777773917059,
+      "learning_rate": 7.564728381025847e-06,
+      "loss": 0.1971,
+      "step": 6269
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.1332393105274357,
+      "learning_rate": 7.563955038570205e-06,
+      "loss": 0.1928,
+      "step": 6270
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.0233087695311278,
+      "learning_rate": 7.563181612887109e-06,
+      "loss": 0.1588,
+      "step": 6271
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.392516323392859,
+      "learning_rate": 7.5624081040016614e-06,
+      "loss": 0.2222,
+      "step": 6272
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.0069787250727105,
+      "learning_rate": 7.561634511938976e-06,
+      "loss": 0.1592,
+      "step": 6273
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 3.447181931903469,
+      "learning_rate": 7.5608608367241595e-06,
+      "loss": 0.1453,
+      "step": 6274
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.083788960687657,
+      "learning_rate": 7.56008707838233e-06,
+      "loss": 0.1817,
+      "step": 6275
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.898151752960414,
+      "learning_rate": 7.559313236938599e-06,
+      "loss": 0.1596,
+      "step": 6276
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.0359540528058104,
+      "learning_rate": 7.5585393124180896e-06,
+      "loss": 0.2606,
+      "step": 6277
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.0508084010051806,
+      "learning_rate": 7.5577653048459236e-06,
+      "loss": 0.2027,
+      "step": 6278
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.0009418127959702,
+      "learning_rate": 7.556991214247222e-06,
+      "loss": 0.1819,
+      "step": 6279
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.438499693933825,
+      "learning_rate": 7.556217040647119e-06,
+      "loss": 0.2175,
+      "step": 6280
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.9956295315471142,
+      "learning_rate": 7.555442784070739e-06,
+      "loss": 0.1962,
+      "step": 6281
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.2411029132383242,
+      "learning_rate": 7.5546684445432185e-06,
+      "loss": 0.1758,
+      "step": 6282
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.0439540338107332,
+      "learning_rate": 7.553894022089689e-06,
+      "loss": 0.2157,
+      "step": 6283
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.4571176278331235,
+      "learning_rate": 7.553119516735292e-06,
+      "loss": 0.1604,
+      "step": 6284
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 5.858724898831277,
+      "learning_rate": 7.552344928505168e-06,
+      "loss": 0.2005,
+      "step": 6285
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.2694955703416553,
+      "learning_rate": 7.551570257424459e-06,
+      "loss": 0.2215,
+      "step": 6286
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.0702111150936124,
+      "learning_rate": 7.550795503518311e-06,
+      "loss": 0.1578,
+      "step": 6287
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.042418422237254,
+      "learning_rate": 7.550020666811876e-06,
+      "loss": 0.2267,
+      "step": 6288
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.5142343202183468,
+      "learning_rate": 7.549245747330303e-06,
+      "loss": 0.2497,
+      "step": 6289
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.4356165706817186,
+      "learning_rate": 7.548470745098747e-06,
+      "loss": 0.1958,
+      "step": 6290
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.2271601821334965,
+      "learning_rate": 7.5476956601423664e-06,
+      "loss": 0.1558,
+      "step": 6291
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.7083889866889472,
+      "learning_rate": 7.546920492486318e-06,
+      "loss": 0.2578,
+      "step": 6292
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.5937853720423505,
+      "learning_rate": 7.546145242155768e-06,
+      "loss": 0.2723,
+      "step": 6293
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.205236770722455,
+      "learning_rate": 7.545369909175877e-06,
+      "loss": 0.2112,
+      "step": 6294
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.4932618343060595,
+      "learning_rate": 7.5445944935718175e-06,
+      "loss": 0.1995,
+      "step": 6295
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.2097439990945245,
+      "learning_rate": 7.543818995368755e-06,
+      "loss": 0.2434,
+      "step": 6296
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.2133566457137934,
+      "learning_rate": 7.543043414591865e-06,
+      "loss": 0.2118,
+      "step": 6297
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.1255136648364945,
+      "learning_rate": 7.542267751266326e-06,
+      "loss": 0.1946,
+      "step": 6298
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.401694373238923,
+      "learning_rate": 7.5414920054173105e-06,
+      "loss": 0.2022,
+      "step": 6299
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.1080024616899669,
+      "learning_rate": 7.540716177070006e-06,
+      "loss": 0.2428,
+      "step": 6300
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.0728943897298764,
+      "learning_rate": 7.539940266249592e-06,
+      "loss": 0.2597,
+      "step": 6301
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.7163036775043776,
+      "learning_rate": 7.539164272981258e-06,
+      "loss": 0.2055,
+      "step": 6302
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.0832860398362032,
+      "learning_rate": 7.538388197290189e-06,
+      "loss": 0.1955,
+      "step": 6303
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.1289450100907281,
+      "learning_rate": 7.537612039201581e-06,
+      "loss": 0.1964,
+      "step": 6304
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.448765508384205,
+      "learning_rate": 7.536835798740626e-06,
+      "loss": 0.2428,
+      "step": 6305
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.1232436057559223,
+      "learning_rate": 7.5360594759325225e-06,
+      "loss": 0.1869,
+      "step": 6306
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.175924990983071,
+      "learning_rate": 7.535283070802471e-06,
+      "loss": 0.1931,
+      "step": 6307
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.8534268435605944,
+      "learning_rate": 7.534506583375671e-06,
+      "loss": 0.1998,
+      "step": 6308
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.3731435630474254,
+      "learning_rate": 7.533730013677333e-06,
+      "loss": 0.2385,
+      "step": 6309
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.4104546867802892,
+      "learning_rate": 7.53295336173266e-06,
+      "loss": 0.239,
+      "step": 6310
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.3159891311003185,
+      "learning_rate": 7.5321766275668655e-06,
+      "loss": 0.2034,
+      "step": 6311
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.0853046229977592,
+      "learning_rate": 7.53139981120516e-06,
+      "loss": 0.1734,
+      "step": 6312
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.3736082912574052,
+      "learning_rate": 7.5306229126727615e-06,
+      "loss": 0.2319,
+      "step": 6313
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.3948819864071873,
+      "learning_rate": 7.529845931994888e-06,
+      "loss": 0.2474,
+      "step": 6314
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.2447042072035868,
+      "learning_rate": 7.52906886919676e-06,
+      "loss": 0.2037,
+      "step": 6315
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.8644383328159503,
+      "learning_rate": 7.5282917243036025e-06,
+      "loss": 0.2075,
+      "step": 6316
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.4794690624481923,
+      "learning_rate": 7.527514497340642e-06,
+      "loss": 0.1933,
+      "step": 6317
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.2555559646114114,
+      "learning_rate": 7.5267371883331085e-06,
+      "loss": 0.1485,
+      "step": 6318
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.276559542322176,
+      "learning_rate": 7.525959797306232e-06,
+      "loss": 0.2292,
+      "step": 6319
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.063441014111699,
+      "learning_rate": 7.52518232428525e-06,
+      "loss": 0.218,
+      "step": 6320
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.9904498480939828,
+      "learning_rate": 7.5244047692953944e-06,
+      "loss": 0.2303,
+      "step": 6321
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.206457803515479,
+      "learning_rate": 7.52362713236191e-06,
+      "loss": 0.2267,
+      "step": 6322
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.330752886382217,
+      "learning_rate": 7.522849413510037e-06,
+      "loss": 0.2113,
+      "step": 6323
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.0302280149345242,
+      "learning_rate": 7.522071612765021e-06,
+      "loss": 0.233,
+      "step": 6324
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.180404876013474,
+      "learning_rate": 7.521293730152111e-06,
+      "loss": 0.2047,
+      "step": 6325
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.2564733926394747,
+      "learning_rate": 7.5205157656965564e-06,
+      "loss": 0.2503,
+      "step": 6326
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.523755567310089,
+      "learning_rate": 7.519737719423612e-06,
+      "loss": 0.23,
+      "step": 6327
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.1993455835301206,
+      "learning_rate": 7.51895959135853e-06,
+      "loss": 0.2,
+      "step": 6328
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.3557031110056688,
+      "learning_rate": 7.518181381526574e-06,
+      "loss": 0.2402,
+      "step": 6329
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.949685743235513,
+      "learning_rate": 7.517403089952999e-06,
+      "loss": 0.1898,
+      "step": 6330
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.1670501686600077,
+      "learning_rate": 7.516624716663075e-06,
+      "loss": 0.1988,
+      "step": 6331
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.4847754289704336,
+      "learning_rate": 7.515846261682062e-06,
+      "loss": 0.1965,
+      "step": 6332
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.113439277676414,
+      "learning_rate": 7.515067725035236e-06,
+      "loss": 0.1767,
+      "step": 6333
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.1166617012310152,
+      "learning_rate": 7.514289106747864e-06,
+      "loss": 0.1783,
+      "step": 6334
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.8212989892711211,
+      "learning_rate": 7.513510406845221e-06,
+      "loss": 0.1754,
+      "step": 6335
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.0831344571721528,
+      "learning_rate": 7.5127316253525876e-06,
+      "loss": 0.2127,
+      "step": 6336
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.1566617528962648,
+      "learning_rate": 7.511952762295238e-06,
+      "loss": 0.1429,
+      "step": 6337
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.1418735516251355,
+      "learning_rate": 7.511173817698459e-06,
+      "loss": 0.182,
+      "step": 6338
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.457824889740624,
+      "learning_rate": 7.510394791587532e-06,
+      "loss": 0.2341,
+      "step": 6339
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.880637346016007,
+      "learning_rate": 7.5096156839877475e-06,
+      "loss": 0.2286,
+      "step": 6340
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.0304099801507693,
+      "learning_rate": 7.508836494924394e-06,
+      "loss": 0.1765,
+      "step": 6341
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.0867522702782442,
+      "learning_rate": 7.508057224422767e-06,
+      "loss": 0.1254,
+      "step": 6342
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.129637125219074,
+      "learning_rate": 7.507277872508158e-06,
+      "loss": 0.1777,
+      "step": 6343
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.1436562089950004,
+      "learning_rate": 7.50649843920587e-06,
+      "loss": 0.1544,
+      "step": 6344
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.9651844739654687,
+      "learning_rate": 7.505718924541199e-06,
+      "loss": 0.2077,
+      "step": 6345
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.5834647959845958,
+      "learning_rate": 7.504939328539453e-06,
+      "loss": 0.2288,
+      "step": 6346
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.149583373811058,
+      "learning_rate": 7.504159651225935e-06,
+      "loss": 0.2306,
+      "step": 6347
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.0885836381745106,
+      "learning_rate": 7.503379892625953e-06,
+      "loss": 0.2269,
+      "step": 6348
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.1323737067445654,
+      "learning_rate": 7.502600052764822e-06,
+      "loss": 0.1997,
+      "step": 6349
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.4582002286680427,
+      "learning_rate": 7.501820131667854e-06,
+      "loss": 0.2195,
+      "step": 6350
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.3132557296883012,
+      "learning_rate": 7.501040129360364e-06,
+      "loss": 0.223,
+      "step": 6351
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.0770214694215607,
+      "learning_rate": 7.500260045867676e-06,
+      "loss": 0.19,
+      "step": 6352
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.293153935392636,
+      "learning_rate": 7.499479881215108e-06,
+      "loss": 0.1885,
+      "step": 6353
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.037809110749088,
+      "learning_rate": 7.498699635427986e-06,
+      "loss": 0.2121,
+      "step": 6354
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.1333371295877062,
+      "learning_rate": 7.497919308531637e-06,
+      "loss": 0.1998,
+      "step": 6355
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.8906502929241473,
+      "learning_rate": 7.49713890055139e-06,
+      "loss": 0.1803,
+      "step": 6356
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.8405350623270023,
+      "learning_rate": 7.49635841151258e-06,
+      "loss": 0.1733,
+      "step": 6357
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.8016728161454862,
+      "learning_rate": 7.495577841440539e-06,
+      "loss": 0.1775,
+      "step": 6358
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.0393521561265617,
+      "learning_rate": 7.4947971903606065e-06,
+      "loss": 0.1667,
+      "step": 6359
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.8888012823266481,
+      "learning_rate": 7.4940164582981225e-06,
+      "loss": 0.1903,
+      "step": 6360
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.9655667271978039,
+      "learning_rate": 7.493235645278431e-06,
+      "loss": 0.1702,
+      "step": 6361
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.5906124459241524,
+      "learning_rate": 7.492454751326874e-06,
+      "loss": 0.2461,
+      "step": 6362
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.8607193575281702,
+      "learning_rate": 7.491673776468806e-06,
+      "loss": 0.2007,
+      "step": 6363
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.0343882410780834,
+      "learning_rate": 7.490892720729574e-06,
+      "loss": 0.2071,
+      "step": 6364
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.1253605238421864,
+      "learning_rate": 7.490111584134531e-06,
+      "loss": 0.2192,
+      "step": 6365
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.09364147669682,
+      "learning_rate": 7.489330366709033e-06,
+      "loss": 0.2011,
+      "step": 6366
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.0028761066584357,
+      "learning_rate": 7.4885490684784416e-06,
+      "loss": 0.1834,
+      "step": 6367
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.3170517688609655,
+      "learning_rate": 7.487767689468114e-06,
+      "loss": 0.2096,
+      "step": 6368
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.5834363325606704,
+      "learning_rate": 7.486986229703418e-06,
+      "loss": 0.2187,
+      "step": 6369
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.4016865569064014,
+      "learning_rate": 7.486204689209719e-06,
+      "loss": 0.1864,
+      "step": 6370
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.1107951698123253,
+      "learning_rate": 7.485423068012386e-06,
+      "loss": 0.1806,
+      "step": 6371
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.3822374791642302,
+      "learning_rate": 7.484641366136791e-06,
+      "loss": 0.2318,
+      "step": 6372
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.3167953425567953,
+      "learning_rate": 7.4838595836083095e-06,
+      "loss": 0.218,
+      "step": 6373
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.8948509631957754,
+      "learning_rate": 7.483077720452317e-06,
+      "loss": 0.1436,
+      "step": 6374
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.0447795913653914,
+      "learning_rate": 7.482295776694193e-06,
+      "loss": 0.1983,
+      "step": 6375
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.3780013119191477,
+      "learning_rate": 7.481513752359322e-06,
+      "loss": 0.1837,
+      "step": 6376
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.1991184500211478,
+      "learning_rate": 7.480731647473086e-06,
+      "loss": 0.1988,
+      "step": 6377
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.958407910527363,
+      "learning_rate": 7.479949462060875e-06,
+      "loss": 0.1819,
+      "step": 6378
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.0092579397284214,
+      "learning_rate": 7.479167196148079e-06,
+      "loss": 0.1796,
+      "step": 6379
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.228971571074466,
+      "learning_rate": 7.478384849760088e-06,
+      "loss": 0.2049,
+      "step": 6380
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.3969344474209906,
+      "learning_rate": 7.4776024229223036e-06,
+      "loss": 0.2033,
+      "step": 6381
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.3043812375177708,
+      "learning_rate": 7.476819915660116e-06,
+      "loss": 0.218,
+      "step": 6382
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.1034841617386164,
+      "learning_rate": 7.476037327998933e-06,
+      "loss": 0.1958,
+      "step": 6383
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.7390587600160539,
+      "learning_rate": 7.4752546599641515e-06,
+      "loss": 0.261,
+      "step": 6384
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.9931795349716315,
+      "learning_rate": 7.47447191158118e-06,
+      "loss": 0.1738,
+      "step": 6385
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.0136557267212931,
+      "learning_rate": 7.473689082875431e-06,
+      "loss": 0.1503,
+      "step": 6386
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.1799645098656988,
+      "learning_rate": 7.472906173872309e-06,
+      "loss": 0.2203,
+      "step": 6387
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.600831050852145,
+      "learning_rate": 7.4721231845972305e-06,
+      "loss": 0.1936,
+      "step": 6388
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.3344528470346995,
+      "learning_rate": 7.471340115075613e-06,
+      "loss": 0.2105,
+      "step": 6389
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.3722388330325956,
+      "learning_rate": 7.4705569653328735e-06,
+      "loss": 0.2584,
+      "step": 6390
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.5652326553864597,
+      "learning_rate": 7.4697737353944355e-06,
+      "loss": 0.1733,
+      "step": 6391
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.0153481491261023,
+      "learning_rate": 7.4689904252857214e-06,
+      "loss": 0.2092,
+      "step": 6392
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.5058574651762928,
+      "learning_rate": 7.468207035032158e-06,
+      "loss": 0.1991,
+      "step": 6393
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.032282197115231,
+      "learning_rate": 7.467423564659174e-06,
+      "loss": 0.2172,
+      "step": 6394
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.3474586480161739,
+      "learning_rate": 7.466640014192203e-06,
+      "loss": 0.1939,
+      "step": 6395
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.128564852271032,
+      "learning_rate": 7.465856383656679e-06,
+      "loss": 0.1784,
+      "step": 6396
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.2668403041663536,
+      "learning_rate": 7.465072673078039e-06,
+      "loss": 0.1745,
+      "step": 6397
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.2957898103768628,
+      "learning_rate": 7.464288882481722e-06,
+      "loss": 0.208,
+      "step": 6398
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.9460704066627907,
+      "learning_rate": 7.46350501189317e-06,
+      "loss": 0.1607,
+      "step": 6399
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.2508542747791467,
+      "learning_rate": 7.4627210613378296e-06,
+      "loss": 0.1959,
+      "step": 6400
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.2412633599676863,
+      "learning_rate": 7.461937030841149e-06,
+      "loss": 0.2053,
+      "step": 6401
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.5226109340886438,
+      "learning_rate": 7.461152920428573e-06,
+      "loss": 0.1821,
+      "step": 6402
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.0197564621717328,
+      "learning_rate": 7.46036873012556e-06,
+      "loss": 0.1382,
+      "step": 6403
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.1977529265271054,
+      "learning_rate": 7.459584459957564e-06,
+      "loss": 0.2024,
+      "step": 6404
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.4195454188948218,
+      "learning_rate": 7.4588001099500405e-06,
+      "loss": 0.2159,
+      "step": 6405
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.1155872705234506,
+      "learning_rate": 7.458015680128453e-06,
+      "loss": 0.1705,
+      "step": 6406
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.3333964553310915,
+      "learning_rate": 7.457231170518263e-06,
+      "loss": 0.202,
+      "step": 6407
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.7213350687876825,
+      "learning_rate": 7.456446581144936e-06,
+      "loss": 0.233,
+      "step": 6408
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.442374204223814,
+      "learning_rate": 7.455661912033939e-06,
+      "loss": 0.2385,
+      "step": 6409
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.1542525654080675,
+      "learning_rate": 7.454877163210745e-06,
+      "loss": 0.2297,
+      "step": 6410
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.6250579484364023,
+      "learning_rate": 7.4540923347008264e-06,
+      "loss": 0.2515,
+      "step": 6411
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.2626928077194965,
+      "learning_rate": 7.45330742652966e-06,
+      "loss": 0.1916,
+      "step": 6412
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.2374905926053026,
+      "learning_rate": 7.452522438722725e-06,
+      "loss": 0.2735,
+      "step": 6413
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.2967146636479296,
+      "learning_rate": 7.451737371305499e-06,
+      "loss": 0.1982,
+      "step": 6414
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.030526967967222,
+      "learning_rate": 7.45095222430347e-06,
+      "loss": 0.1851,
+      "step": 6415
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.7117077289046221,
+      "learning_rate": 7.450166997742121e-06,
+      "loss": 0.1891,
+      "step": 6416
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.1230569756095117,
+      "learning_rate": 7.449381691646943e-06,
+      "loss": 0.2063,
+      "step": 6417
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 2.0664281573513454,
+      "learning_rate": 7.448596306043425e-06,
+      "loss": 0.1928,
+      "step": 6418
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.022435323691324,
+      "learning_rate": 7.447810840957066e-06,
+      "loss": 0.2006,
+      "step": 6419
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.4271857255089782,
+      "learning_rate": 7.4470252964133575e-06,
+      "loss": 0.1542,
+      "step": 6420
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.3335559193207018,
+      "learning_rate": 7.446239672437801e-06,
+      "loss": 0.2276,
+      "step": 6421
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.9397404582245014,
+      "learning_rate": 7.445453969055898e-06,
+      "loss": 0.1691,
+      "step": 6422
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.1402731248178826,
+      "learning_rate": 7.444668186293153e-06,
+      "loss": 0.1922,
+      "step": 6423
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.9078930951157003,
+      "learning_rate": 7.443882324175072e-06,
+      "loss": 0.2062,
+      "step": 6424
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.361304968441313,
+      "learning_rate": 7.443096382727167e-06,
+      "loss": 0.1962,
+      "step": 6425
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.1934024362864502,
+      "learning_rate": 7.442310361974949e-06,
+      "loss": 0.1968,
+      "step": 6426
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.8619461134866833,
+      "learning_rate": 7.441524261943931e-06,
+      "loss": 0.2123,
+      "step": 6427
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.1312583753107173,
+      "learning_rate": 7.440738082659631e-06,
+      "loss": 0.2025,
+      "step": 6428
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.188766464508972,
+      "learning_rate": 7.43995182414757e-06,
+      "loss": 0.2276,
+      "step": 6429
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.3097577398274995,
+      "learning_rate": 7.439165486433269e-06,
+      "loss": 0.1938,
+      "step": 6430
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.9664159428905263,
+      "learning_rate": 7.438379069542255e-06,
+      "loss": 0.18,
+      "step": 6431
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.2428347203026266,
+      "learning_rate": 7.437592573500053e-06,
+      "loss": 0.2398,
+      "step": 6432
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.7919017967119454,
+      "learning_rate": 7.436805998332195e-06,
+      "loss": 0.1404,
+      "step": 6433
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.7481623167098159,
+      "learning_rate": 7.436019344064214e-06,
+      "loss": 0.1769,
+      "step": 6434
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.2850516720227432,
+      "learning_rate": 7.435232610721643e-06,
+      "loss": 0.207,
+      "step": 6435
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.1855612617260773,
+      "learning_rate": 7.434445798330021e-06,
+      "loss": 0.2479,
+      "step": 6436
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.0416992556918039,
+      "learning_rate": 7.433658906914889e-06,
+      "loss": 0.188,
+      "step": 6437
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.2257153324956516,
+      "learning_rate": 7.4328719365017885e-06,
+      "loss": 0.2414,
+      "step": 6438
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.2452770160492121,
+      "learning_rate": 7.432084887116267e-06,
+      "loss": 0.2221,
+      "step": 6439
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.4644134601584757,
+      "learning_rate": 7.431297758783872e-06,
+      "loss": 0.2248,
+      "step": 6440
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.8796922814017356,
+      "learning_rate": 7.430510551530152e-06,
+      "loss": 0.2026,
+      "step": 6441
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.228926066986379,
+      "learning_rate": 7.429723265380663e-06,
+      "loss": 0.2014,
+      "step": 6442
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.2240164430161875,
+      "learning_rate": 7.428935900360959e-06,
+      "loss": 0.1919,
+      "step": 6443
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.3557330981341804,
+      "learning_rate": 7.4281484564966e-06,
+      "loss": 0.2197,
+      "step": 6444
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.3422981197748343,
+      "learning_rate": 7.4273609338131456e-06,
+      "loss": 0.2295,
+      "step": 6445
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.9411562077081198,
+      "learning_rate": 7.426573332336159e-06,
+      "loss": 0.1932,
+      "step": 6446
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.0898250728633843,
+      "learning_rate": 7.425785652091206e-06,
+      "loss": 0.2304,
+      "step": 6447
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.3103717290384331,
+      "learning_rate": 7.4249978931038555e-06,
+      "loss": 0.2036,
+      "step": 6448
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.2260869373799728,
+      "learning_rate": 7.424210055399681e-06,
+      "loss": 0.1817,
+      "step": 6449
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.2192889024692974,
+      "learning_rate": 7.423422139004253e-06,
+      "loss": 0.2282,
+      "step": 6450
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.2025586318554615,
+      "learning_rate": 7.422634143943148e-06,
+      "loss": 0.1539,
+      "step": 6451
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.259165048491959,
+      "learning_rate": 7.421846070241947e-06,
+      "loss": 0.1758,
+      "step": 6452
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.019781233661434,
+      "learning_rate": 7.42105791792623e-06,
+      "loss": 0.1706,
+      "step": 6453
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.0855606311929928,
+      "learning_rate": 7.42026968702158e-06,
+      "loss": 0.1833,
+      "step": 6454
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.296805924994685,
+      "learning_rate": 7.419481377553584e-06,
+      "loss": 0.1876,
+      "step": 6455
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.1239165388657508,
+      "learning_rate": 7.418692989547831e-06,
+      "loss": 0.1894,
+      "step": 6456
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.041991161428159,
+      "learning_rate": 7.4179045230299114e-06,
+      "loss": 0.2207,
+      "step": 6457
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.3130548686559709,
+      "learning_rate": 7.4171159780254235e-06,
+      "loss": 0.1915,
+      "step": 6458
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.8749016201962837,
+      "learning_rate": 7.416327354559959e-06,
+      "loss": 0.2102,
+      "step": 6459
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.2788870992374346,
+      "learning_rate": 7.41553865265912e-06,
+      "loss": 0.1638,
+      "step": 6460
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.195034262742573,
+      "learning_rate": 7.4147498723485075e-06,
+      "loss": 0.1073,
+      "step": 6461
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.2702174107714874,
+      "learning_rate": 7.413961013653725e-06,
+      "loss": 0.1852,
+      "step": 6462
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.1714319758638851,
+      "learning_rate": 7.41317207660038e-06,
+      "loss": 0.2428,
+      "step": 6463
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.2485118635538799,
+      "learning_rate": 7.412383061214082e-06,
+      "loss": 0.1936,
+      "step": 6464
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 2.5546395883529556,
+      "learning_rate": 7.411593967520443e-06,
+      "loss": 0.2065,
+      "step": 6465
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.2709212191345343,
+      "learning_rate": 7.410804795545076e-06,
+      "loss": 0.2152,
+      "step": 6466
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.9261190023659244,
+      "learning_rate": 7.410015545313599e-06,
+      "loss": 0.2173,
+      "step": 6467
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.174993176453135,
+      "learning_rate": 7.409226216851632e-06,
+      "loss": 0.2089,
+      "step": 6468
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.2818624904939122,
+      "learning_rate": 7.408436810184797e-06,
+      "loss": 0.2411,
+      "step": 6469
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.1807160015056544,
+      "learning_rate": 7.407647325338718e-06,
+      "loss": 0.2046,
+      "step": 6470
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.0353835594070298,
+      "learning_rate": 7.4068577623390225e-06,
+      "loss": 0.1955,
+      "step": 6471
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.3084477784459403,
+      "learning_rate": 7.40606812121134e-06,
+      "loss": 0.2181,
+      "step": 6472
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.4803035588565197,
+      "learning_rate": 7.405278401981302e-06,
+      "loss": 0.1821,
+      "step": 6473
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.707763869254962,
+      "learning_rate": 7.404488604674544e-06,
+      "loss": 0.1553,
+      "step": 6474
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.1407311011309513,
+      "learning_rate": 7.403698729316703e-06,
+      "loss": 0.1944,
+      "step": 6475
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.1651707180741961,
+      "learning_rate": 7.402908775933419e-06,
+      "loss": 0.2356,
+      "step": 6476
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.1974227329245353,
+      "learning_rate": 7.402118744550335e-06,
+      "loss": 0.2115,
+      "step": 6477
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.029791425429932,
+      "learning_rate": 7.401328635193096e-06,
+      "loss": 0.2046,
+      "step": 6478
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.4360132329227808,
+      "learning_rate": 7.400538447887347e-06,
+      "loss": 0.1901,
+      "step": 6479
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.2342192817201012,
+      "learning_rate": 7.399748182658741e-06,
+      "loss": 0.182,
+      "step": 6480
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.3293102949767066,
+      "learning_rate": 7.398957839532928e-06,
+      "loss": 0.2227,
+      "step": 6481
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.4373534480316263,
+      "learning_rate": 7.3981674185355645e-06,
+      "loss": 0.165,
+      "step": 6482
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.3969139158771997,
+      "learning_rate": 7.3973769196923075e-06,
+      "loss": 0.2175,
+      "step": 6483
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.4638305007989396,
+      "learning_rate": 7.396586343028817e-06,
+      "loss": 0.2012,
+      "step": 6484
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.4028952584655112,
+      "learning_rate": 7.3957956885707545e-06,
+      "loss": 0.2354,
+      "step": 6485
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.1790490570181629,
+      "learning_rate": 7.395004956343788e-06,
+      "loss": 0.1909,
+      "step": 6486
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.7597341672263938,
+      "learning_rate": 7.394214146373583e-06,
+      "loss": 0.2088,
+      "step": 6487
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.2733579651738003,
+      "learning_rate": 7.39342325868581e-06,
+      "loss": 0.2053,
+      "step": 6488
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.0591698447896867,
+      "learning_rate": 7.392632293306143e-06,
+      "loss": 0.175,
+      "step": 6489
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.0314151674996712,
+      "learning_rate": 7.391841250260253e-06,
+      "loss": 0.1941,
+      "step": 6490
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.1558487153923458,
+      "learning_rate": 7.3910501295738226e-06,
+      "loss": 0.1814,
+      "step": 6491
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.643670189664107,
+      "learning_rate": 7.390258931272529e-06,
+      "loss": 0.1807,
+      "step": 6492
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.3710752143585085,
+      "learning_rate": 7.389467655382057e-06,
+      "loss": 0.2233,
+      "step": 6493
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.2552917823382888,
+      "learning_rate": 7.38867630192809e-06,
+      "loss": 0.206,
+      "step": 6494
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.6483462159251552,
+      "learning_rate": 7.387884870936317e-06,
+      "loss": 0.2247,
+      "step": 6495
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.7426131427737876,
+      "learning_rate": 7.387093362432428e-06,
+      "loss": 0.1782,
+      "step": 6496
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.9308151090520451,
+      "learning_rate": 7.386301776442116e-06,
+      "loss": 0.1707,
+      "step": 6497
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.435682045158624,
+      "learning_rate": 7.385510112991077e-06,
+      "loss": 0.219,
+      "step": 6498
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.889528057022971,
+      "learning_rate": 7.384718372105007e-06,
+      "loss": 0.1964,
+      "step": 6499
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.1695172688088706,
+      "learning_rate": 7.383926553809606e-06,
+      "loss": 0.2537,
+      "step": 6500
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.0532292846414517,
+      "learning_rate": 7.383134658130579e-06,
+      "loss": 0.1989,
+      "step": 6501
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.0833601795440304,
+      "learning_rate": 7.382342685093631e-06,
+      "loss": 0.2112,
+      "step": 6502
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.4295239016208208,
+      "learning_rate": 7.381550634724471e-06,
+      "loss": 0.2122,
+      "step": 6503
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.2622987252728668,
+      "learning_rate": 7.380758507048804e-06,
+      "loss": 0.1837,
+      "step": 6504
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.3271835693020926,
+      "learning_rate": 7.379966302092351e-06,
+      "loss": 0.257,
+      "step": 6505
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.9325274639009956,
+      "learning_rate": 7.379174019880821e-06,
+      "loss": 0.1789,
+      "step": 6506
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.400656755981912,
+      "learning_rate": 7.378381660439935e-06,
+      "loss": 0.2321,
+      "step": 6507
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.3777664217616576,
+      "learning_rate": 7.377589223795411e-06,
+      "loss": 0.1897,
+      "step": 6508
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.9600472231996647,
+      "learning_rate": 7.376796709972975e-06,
+      "loss": 0.133,
+      "step": 6509
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.2512893568603123,
+      "learning_rate": 7.376004118998352e-06,
+      "loss": 0.1841,
+      "step": 6510
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.361880645932823,
+      "learning_rate": 7.375211450897267e-06,
+      "loss": 0.1644,
+      "step": 6511
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.9237114225715308,
+      "learning_rate": 7.374418705695454e-06,
+      "loss": 0.2064,
+      "step": 6512
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.111534277193299,
+      "learning_rate": 7.3736258834186435e-06,
+      "loss": 0.2199,
+      "step": 6513
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.3062764745178597,
+      "learning_rate": 7.372832984092572e-06,
+      "loss": 0.265,
+      "step": 6514
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.409955380258786,
+      "learning_rate": 7.372040007742979e-06,
+      "loss": 0.1921,
+      "step": 6515
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.0855596059798724,
+      "learning_rate": 7.371246954395602e-06,
+      "loss": 0.2351,
+      "step": 6516
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.643221372010555,
+      "learning_rate": 7.370453824076185e-06,
+      "loss": 0.1922,
+      "step": 6517
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.3263691192262368,
+      "learning_rate": 7.3696606168104735e-06,
+      "loss": 0.2084,
+      "step": 6518
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.4266113843198753,
+      "learning_rate": 7.368867332624216e-06,
+      "loss": 0.1829,
+      "step": 6519
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.3134859304554913,
+      "learning_rate": 7.368073971543163e-06,
+      "loss": 0.2421,
+      "step": 6520
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.2040895042217854,
+      "learning_rate": 7.367280533593066e-06,
+      "loss": 0.1669,
+      "step": 6521
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.940185032272099,
+      "learning_rate": 7.366487018799683e-06,
+      "loss": 0.1663,
+      "step": 6522
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.4040247158703478,
+      "learning_rate": 7.36569342718877e-06,
+      "loss": 0.2323,
+      "step": 6523
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.6781390427067573,
+      "learning_rate": 7.3648997587860895e-06,
+      "loss": 0.221,
+      "step": 6524
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.0722050050369965,
+      "learning_rate": 7.3641060136174015e-06,
+      "loss": 0.2061,
+      "step": 6525
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.2568103755079507,
+      "learning_rate": 7.363312191708473e-06,
+      "loss": 0.1495,
+      "step": 6526
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.254679869038332,
+      "learning_rate": 7.3625182930850705e-06,
+      "loss": 0.2127,
+      "step": 6527
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.148206131199171,
+      "learning_rate": 7.361724317772967e-06,
+      "loss": 0.2558,
+      "step": 6528
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.05930518951349,
+      "learning_rate": 7.360930265797934e-06,
+      "loss": 0.1651,
+      "step": 6529
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.1323821878337748,
+      "learning_rate": 7.360136137185747e-06,
+      "loss": 0.1757,
+      "step": 6530
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.5903510308353226,
+      "learning_rate": 7.359341931962183e-06,
+      "loss": 0.2277,
+      "step": 6531
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.1139737083795995,
+      "learning_rate": 7.358547650153024e-06,
+      "loss": 0.2288,
+      "step": 6532
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.2534093875093435,
+      "learning_rate": 7.357753291784052e-06,
+      "loss": 0.227,
+      "step": 6533
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.4109442614607741,
+      "learning_rate": 7.356958856881054e-06,
+      "loss": 0.2133,
+      "step": 6534
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.4754035150249183,
+      "learning_rate": 7.3561643454698125e-06,
+      "loss": 0.2119,
+      "step": 6535
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.3942205259214977,
+      "learning_rate": 7.355369757576123e-06,
+      "loss": 0.2156,
+      "step": 6536
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.8502742599228862,
+      "learning_rate": 7.3545750932257775e-06,
+      "loss": 0.1914,
+      "step": 6537
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.2576723701893229,
+      "learning_rate": 7.353780352444569e-06,
+      "loss": 0.1379,
+      "step": 6538
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.3434275528583222,
+      "learning_rate": 7.352985535258299e-06,
+      "loss": 0.2189,
+      "step": 6539
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.0141068850003967,
+      "learning_rate": 7.352190641692764e-06,
+      "loss": 0.1795,
+      "step": 6540
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.1319976580843798,
+      "learning_rate": 7.351395671773768e-06,
+      "loss": 0.2107,
+      "step": 6541
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.0040428981476366,
+      "learning_rate": 7.3506006255271175e-06,
+      "loss": 0.181,
+      "step": 6542
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.1841738572843048,
+      "learning_rate": 7.349805502978619e-06,
+      "loss": 0.2234,
+      "step": 6543
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.3212337440024826,
+      "learning_rate": 7.349010304154081e-06,
+      "loss": 0.2132,
+      "step": 6544
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.3568573031508002,
+      "learning_rate": 7.3482150290793174e-06,
+      "loss": 0.2721,
+      "step": 6545
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.6481128662235573,
+      "learning_rate": 7.347419677780146e-06,
+      "loss": 0.2387,
+      "step": 6546
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.0553145681439469,
+      "learning_rate": 7.346624250282379e-06,
+      "loss": 0.2149,
+      "step": 6547
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.1119140930551017,
+      "learning_rate": 7.345828746611843e-06,
+      "loss": 0.2168,
+      "step": 6548
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.7205106257340852,
+      "learning_rate": 7.3450331667943545e-06,
+      "loss": 0.2006,
+      "step": 6549
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.2614850091256349,
+      "learning_rate": 7.344237510855742e-06,
+      "loss": 0.2347,
+      "step": 6550
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.4261315150676757,
+      "learning_rate": 7.343441778821832e-06,
+      "loss": 0.2302,
+      "step": 6551
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.3383687898183247,
+      "learning_rate": 7.342645970718454e-06,
+      "loss": 0.2356,
+      "step": 6552
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.1266911068176835,
+      "learning_rate": 7.341850086571439e-06,
+      "loss": 0.1923,
+      "step": 6553
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.1080833155154906,
+      "learning_rate": 7.341054126406625e-06,
+      "loss": 0.2142,
+      "step": 6554
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.9921082329812954,
+      "learning_rate": 7.340258090249848e-06,
+      "loss": 0.1661,
+      "step": 6555
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.2533565442289234,
+      "learning_rate": 7.339461978126947e-06,
+      "loss": 0.1975,
+      "step": 6556
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.9891227044332948,
+      "learning_rate": 7.338665790063766e-06,
+      "loss": 0.1763,
+      "step": 6557
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.178686532034179,
+      "learning_rate": 7.337869526086148e-06,
+      "loss": 0.2301,
+      "step": 6558
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.0308934556271825,
+      "learning_rate": 7.337073186219941e-06,
+      "loss": 0.2026,
+      "step": 6559
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.1530258116021732,
+      "learning_rate": 7.336276770490994e-06,
+      "loss": 0.1868,
+      "step": 6560
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.0833579719321451,
+      "learning_rate": 7.33548027892516e-06,
+      "loss": 0.214,
+      "step": 6561
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.2042839193745376,
+      "learning_rate": 7.334683711548293e-06,
+      "loss": 0.2468,
+      "step": 6562
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.9894478945738824,
+      "learning_rate": 7.33388706838625e-06,
+      "loss": 0.1752,
+      "step": 6563
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.9590196039756299,
+      "learning_rate": 7.333090349464891e-06,
+      "loss": 0.1743,
+      "step": 6564
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.1442798022571419,
+      "learning_rate": 7.332293554810077e-06,
+      "loss": 0.1858,
+      "step": 6565
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.1013180786213637,
+      "learning_rate": 7.331496684447674e-06,
+      "loss": 0.2168,
+      "step": 6566
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.9178329610833067,
+      "learning_rate": 7.330699738403547e-06,
+      "loss": 0.1763,
+      "step": 6567
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.1653323703625678,
+      "learning_rate": 7.329902716703568e-06,
+      "loss": 0.1904,
+      "step": 6568
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.2364031797530266,
+      "learning_rate": 7.329105619373605e-06,
+      "loss": 0.2279,
+      "step": 6569
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.3456125270426664,
+      "learning_rate": 7.328308446439536e-06,
+      "loss": 0.2284,
+      "step": 6570
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.3739299294999325,
+      "learning_rate": 7.327511197927234e-06,
+      "loss": 0.2159,
+      "step": 6571
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 2.141175759599005,
+      "learning_rate": 7.3267138738625806e-06,
+      "loss": 0.2458,
+      "step": 6572
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.067215814815237,
+      "learning_rate": 7.325916474271457e-06,
+      "loss": 0.219,
+      "step": 6573
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.944638869368014,
+      "learning_rate": 7.3251189991797475e-06,
+      "loss": 0.1944,
+      "step": 6574
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.121285498566405,
+      "learning_rate": 7.3243214486133385e-06,
+      "loss": 0.1586,
+      "step": 6575
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.951085544953128,
+      "learning_rate": 7.3235238225981176e-06,
+      "loss": 0.1811,
+      "step": 6576
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.5667156665406814,
+      "learning_rate": 7.322726121159979e-06,
+      "loss": 0.2236,
+      "step": 6577
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.4942061465987475,
+      "learning_rate": 7.321928344324815e-06,
+      "loss": 0.2924,
+      "step": 6578
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.9717716691076972,
+      "learning_rate": 7.32113049211852e-06,
+      "loss": 0.2174,
+      "step": 6579
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.8821759091374898,
+      "learning_rate": 7.320332564566995e-06,
+      "loss": 0.1596,
+      "step": 6580
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.1004541579678788,
+      "learning_rate": 7.31953456169614e-06,
+      "loss": 0.1973,
+      "step": 6581
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.0008324161108981,
+      "learning_rate": 7.318736483531861e-06,
+      "loss": 0.2491,
+      "step": 6582
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.9372303336449244,
+      "learning_rate": 7.317938330100061e-06,
+      "loss": 0.2045,
+      "step": 6583
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.9830736838163031,
+      "learning_rate": 7.317140101426651e-06,
+      "loss": 0.209,
+      "step": 6584
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.9779460071369965,
+      "learning_rate": 7.3163417975375394e-06,
+      "loss": 0.201,
+      "step": 6585
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.1679669375389605,
+      "learning_rate": 7.315543418458642e-06,
+      "loss": 0.2595,
+      "step": 6586
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.1917887276990209,
+      "learning_rate": 7.3147449642158745e-06,
+      "loss": 0.204,
+      "step": 6587
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.2908737895345235,
+      "learning_rate": 7.313946434835154e-06,
+      "loss": 0.2123,
+      "step": 6588
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.0602051937974792,
+      "learning_rate": 7.313147830342401e-06,
+      "loss": 0.1701,
+      "step": 6589
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.964624124117465,
+      "learning_rate": 7.3123491507635384e-06,
+      "loss": 0.1943,
+      "step": 6590
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.2907773792124302,
+      "learning_rate": 7.311550396124494e-06,
+      "loss": 0.2459,
+      "step": 6591
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.9458140026186592,
+      "learning_rate": 7.310751566451194e-06,
+      "loss": 0.171,
+      "step": 6592
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.3681006011839267,
+      "learning_rate": 7.3099526617695705e-06,
+      "loss": 0.1908,
+      "step": 6593
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.0513587529316955,
+      "learning_rate": 7.309153682105553e-06,
+      "loss": 0.1864,
+      "step": 6594
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.604788199740871,
+      "learning_rate": 7.308354627485081e-06,
+      "loss": 0.194,
+      "step": 6595
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.5099091233410629,
+      "learning_rate": 7.307555497934091e-06,
+      "loss": 0.191,
+      "step": 6596
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.110026154187179,
+      "learning_rate": 7.306756293478521e-06,
+      "loss": 0.1991,
+      "step": 6597
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.3622676119664372,
+      "learning_rate": 7.305957014144316e-06,
+      "loss": 0.1973,
+      "step": 6598
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.0080723704697352,
+      "learning_rate": 7.305157659957419e-06,
+      "loss": 0.1673,
+      "step": 6599
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.1732269880524133,
+      "learning_rate": 7.3043582309437796e-06,
+      "loss": 0.2521,
+      "step": 6600
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.224239406034099,
+      "learning_rate": 7.303558727129348e-06,
+      "loss": 0.185,
+      "step": 6601
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.158528001743476,
+      "learning_rate": 7.3027591485400735e-06,
+      "loss": 0.1826,
+      "step": 6602
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.9723999813778341,
+      "learning_rate": 7.301959495201915e-06,
+      "loss": 0.2126,
+      "step": 6603
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.8360734432415676,
+      "learning_rate": 7.301159767140826e-06,
+      "loss": 0.1714,
+      "step": 6604
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.6132949512167745,
+      "learning_rate": 7.300359964382768e-06,
+      "loss": 0.2184,
+      "step": 6605
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.309107528717093,
+      "learning_rate": 7.299560086953704e-06,
+      "loss": 0.2032,
+      "step": 6606
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.250527536564769,
+      "learning_rate": 7.298760134879595e-06,
+      "loss": 0.1997,
+      "step": 6607
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.2128194490148638,
+      "learning_rate": 7.297960108186412e-06,
+      "loss": 0.2148,
+      "step": 6608
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.6960546214028283,
+      "learning_rate": 7.297160006900122e-06,
+      "loss": 0.2264,
+      "step": 6609
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.7459915500566915,
+      "learning_rate": 7.296359831046697e-06,
+      "loss": 0.1351,
+      "step": 6610
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.0425393313511497,
+      "learning_rate": 7.295559580652112e-06,
+      "loss": 0.2182,
+      "step": 6611
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.9956335481928061,
+      "learning_rate": 7.294759255742341e-06,
+      "loss": 0.1951,
+      "step": 6612
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.9455890359408989,
+      "learning_rate": 7.2939588563433675e-06,
+      "loss": 0.2063,
+      "step": 6613
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.3537843032965915,
+      "learning_rate": 7.293158382481169e-06,
+      "loss": 0.2171,
+      "step": 6614
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.9342673357681429,
+      "learning_rate": 7.292357834181732e-06,
+      "loss": 0.1958,
+      "step": 6615
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.0193143884354907,
+      "learning_rate": 7.29155721147104e-06,
+      "loss": 0.2039,
+      "step": 6616
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.3516665085149318,
+      "learning_rate": 7.2907565143750824e-06,
+      "loss": 0.2376,
+      "step": 6617
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.297292713892126,
+      "learning_rate": 7.2899557429198535e-06,
+      "loss": 0.238,
+      "step": 6618
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.2368072301597273,
+      "learning_rate": 7.2891548971313416e-06,
+      "loss": 0.2036,
+      "step": 6619
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.3533034330337466,
+      "learning_rate": 7.288353977035548e-06,
+      "loss": 0.2051,
+      "step": 6620
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.2583443038690625,
+      "learning_rate": 7.287552982658467e-06,
+      "loss": 0.2028,
+      "step": 6621
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.9607718836721291,
+      "learning_rate": 7.2867519140261014e-06,
+      "loss": 0.2012,
+      "step": 6622
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.2471460305977877,
+      "learning_rate": 7.285950771164454e-06,
+      "loss": 0.2373,
+      "step": 6623
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.5656168534872121,
+      "learning_rate": 7.285149554099529e-06,
+      "loss": 0.1971,
+      "step": 6624
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.3947650929743078,
+      "learning_rate": 7.284348262857336e-06,
+      "loss": 0.2147,
+      "step": 6625
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.9910497712901272,
+      "learning_rate": 7.283546897463884e-06,
+      "loss": 0.1756,
+      "step": 6626
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.1264814831150067,
+      "learning_rate": 7.282745457945189e-06,
+      "loss": 0.1575,
+      "step": 6627
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.1990420369720356,
+      "learning_rate": 7.281943944327261e-06,
+      "loss": 0.1974,
+      "step": 6628
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.4695501380043214,
+      "learning_rate": 7.281142356636121e-06,
+      "loss": 0.2228,
+      "step": 6629
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.9864968014000383,
+      "learning_rate": 7.280340694897788e-06,
+      "loss": 0.1764,
+      "step": 6630
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.188157022974993,
+      "learning_rate": 7.279538959138286e-06,
+      "loss": 0.1873,
+      "step": 6631
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.248791253264793,
+      "learning_rate": 7.278737149383639e-06,
+      "loss": 0.2673,
+      "step": 6632
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.0820528365643267,
+      "learning_rate": 7.277935265659872e-06,
+      "loss": 0.1595,
+      "step": 6633
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.1324984592017795,
+      "learning_rate": 7.277133307993018e-06,
+      "loss": 0.2444,
+      "step": 6634
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 2.464206619257025,
+      "learning_rate": 7.2763312764091055e-06,
+      "loss": 0.1423,
+      "step": 6635
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.462099881489146,
+      "learning_rate": 7.275529170934172e-06,
+      "loss": 0.1968,
+      "step": 6636
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.6943217164248068,
+      "learning_rate": 7.274726991594253e-06,
+      "loss": 0.2403,
+      "step": 6637
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.9613967152973169,
+      "learning_rate": 7.273924738415388e-06,
+      "loss": 0.2059,
+      "step": 6638
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.4283292655606847,
+      "learning_rate": 7.273122411423617e-06,
+      "loss": 0.2596,
+      "step": 6639
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.3324046003261365,
+      "learning_rate": 7.272320010644988e-06,
+      "loss": 0.1825,
+      "step": 6640
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.4356016842765609,
+      "learning_rate": 7.2715175361055435e-06,
+      "loss": 0.2426,
+      "step": 6641
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.0784009082857222,
+      "learning_rate": 7.270714987831333e-06,
+      "loss": 0.2109,
+      "step": 6642
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.0180557506561672,
+      "learning_rate": 7.269912365848408e-06,
+      "loss": 0.1999,
+      "step": 6643
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.9969555847637441,
+      "learning_rate": 7.269109670182824e-06,
+      "loss": 0.1328,
+      "step": 6644
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.1296053506995984,
+      "learning_rate": 7.268306900860633e-06,
+      "loss": 0.2223,
+      "step": 6645
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.1843010202417452,
+      "learning_rate": 7.2675040579078974e-06,
+      "loss": 0.1998,
+      "step": 6646
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.1771286993300774,
+      "learning_rate": 7.266701141350676e-06,
+      "loss": 0.2196,
+      "step": 6647
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.3975759788094528,
+      "learning_rate": 7.26589815121503e-06,
+      "loss": 0.2342,
+      "step": 6648
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.917065186890614,
+      "learning_rate": 7.26509508752703e-06,
+      "loss": 0.187,
+      "step": 6649
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.341712263235331,
+      "learning_rate": 7.2642919503127405e-06,
+      "loss": 0.2026,
+      "step": 6650
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.2377773697752374,
+      "learning_rate": 7.2634887395982305e-06,
+      "loss": 0.171,
+      "step": 6651
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.1862897446559548,
+      "learning_rate": 7.262685455409576e-06,
+      "loss": 0.2028,
+      "step": 6652
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.701867943313768,
+      "learning_rate": 7.261882097772849e-06,
+      "loss": 0.2341,
+      "step": 6653
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.8868148978449134,
+      "learning_rate": 7.261078666714129e-06,
+      "loss": 0.2318,
+      "step": 6654
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.24003873768891,
+      "learning_rate": 7.260275162259495e-06,
+      "loss": 0.223,
+      "step": 6655
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.1244053394853264,
+      "learning_rate": 7.25947158443503e-06,
+      "loss": 0.2342,
+      "step": 6656
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.617751686533021,
+      "learning_rate": 7.25866793326682e-06,
+      "loss": 0.2508,
+      "step": 6657
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.0017821517430017,
+      "learning_rate": 7.257864208780947e-06,
+      "loss": 0.1647,
+      "step": 6658
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.1258429078233354,
+      "learning_rate": 7.257060411003505e-06,
+      "loss": 0.2142,
+      "step": 6659
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.8244981162351961,
+      "learning_rate": 7.256256539960583e-06,
+      "loss": 0.1624,
+      "step": 6660
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.2711079944264962,
+      "learning_rate": 7.255452595678277e-06,
+      "loss": 0.1844,
+      "step": 6661
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.1732125786743801,
+      "learning_rate": 7.254648578182683e-06,
+      "loss": 0.2192,
+      "step": 6662
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.164243288850292,
+      "learning_rate": 7.253844487499899e-06,
+      "loss": 0.1946,
+      "step": 6663
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.4262648846512394,
+      "learning_rate": 7.253040323656027e-06,
+      "loss": 0.2265,
+      "step": 6664
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.082450841291026,
+      "learning_rate": 7.252236086677171e-06,
+      "loss": 0.2009,
+      "step": 6665
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.9910645475480091,
+      "learning_rate": 7.251431776589437e-06,
+      "loss": 0.1789,
+      "step": 6666
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.1383736286890196,
+      "learning_rate": 7.250627393418932e-06,
+      "loss": 0.2486,
+      "step": 6667
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.447024974043236,
+      "learning_rate": 7.2498229371917675e-06,
+      "loss": 0.2227,
+      "step": 6668
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.8394571978318014,
+      "learning_rate": 7.249018407934056e-06,
+      "loss": 0.1715,
+      "step": 6669
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.9466505314937549,
+      "learning_rate": 7.248213805671915e-06,
+      "loss": 0.1913,
+      "step": 6670
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.9104901476945106,
+      "learning_rate": 7.247409130431461e-06,
+      "loss": 0.1919,
+      "step": 6671
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.0150673839382778,
+      "learning_rate": 7.246604382238814e-06,
+      "loss": 0.2017,
+      "step": 6672
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.8499983867067216,
+      "learning_rate": 7.245799561120096e-06,
+      "loss": 0.1974,
+      "step": 6673
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.2461387512547661,
+      "learning_rate": 7.244994667101435e-06,
+      "loss": 0.2546,
+      "step": 6674
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.8664802167120436,
+      "learning_rate": 7.244189700208955e-06,
+      "loss": 0.2243,
+      "step": 6675
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.5419861187613122,
+      "learning_rate": 7.243384660468786e-06,
+      "loss": 0.1921,
+      "step": 6676
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.15460114333586,
+      "learning_rate": 7.242579547907062e-06,
+      "loss": 0.2308,
+      "step": 6677
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.985647963186518,
+      "learning_rate": 7.241774362549915e-06,
+      "loss": 0.1779,
+      "step": 6678
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.16076763897997,
+      "learning_rate": 7.240969104423484e-06,
+      "loss": 0.1937,
+      "step": 6679
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.173401545609978,
+      "learning_rate": 7.2401637735539065e-06,
+      "loss": 0.1878,
+      "step": 6680
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.054265939386572,
+      "learning_rate": 7.239358369967327e-06,
+      "loss": 0.2089,
+      "step": 6681
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.7823906683495274,
+      "learning_rate": 7.238552893689883e-06,
+      "loss": 0.1366,
+      "step": 6682
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.2404055398438492,
+      "learning_rate": 7.2377473447477285e-06,
+      "loss": 0.205,
+      "step": 6683
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.076668197359605,
+      "learning_rate": 7.236941723167007e-06,
+      "loss": 0.1527,
+      "step": 6684
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.2368442277757288,
+      "learning_rate": 7.236136028973871e-06,
+      "loss": 0.181,
+      "step": 6685
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.3089553532674503,
+      "learning_rate": 7.2353302621944736e-06,
+      "loss": 0.2347,
+      "step": 6686
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.740194519305841,
+      "learning_rate": 7.2345244228549694e-06,
+      "loss": 0.2277,
+      "step": 6687
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.3364060762490693,
+      "learning_rate": 7.23371851098152e-06,
+      "loss": 0.2641,
+      "step": 6688
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.5387420272233243,
+      "learning_rate": 7.232912526600281e-06,
+      "loss": 0.2054,
+      "step": 6689
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.3401723297531458,
+      "learning_rate": 7.232106469737419e-06,
+      "loss": 0.214,
+      "step": 6690
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.5477788529030425,
+      "learning_rate": 7.2313003404190955e-06,
+      "loss": 0.2761,
+      "step": 6691
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.2753224727069634,
+      "learning_rate": 7.230494138671481e-06,
+      "loss": 0.1399,
+      "step": 6692
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.5088556318195863,
+      "learning_rate": 7.2296878645207444e-06,
+      "loss": 0.2288,
+      "step": 6693
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.9579213810155235,
+      "learning_rate": 7.228881517993058e-06,
+      "loss": 0.2046,
+      "step": 6694
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.4301387929252083,
+      "learning_rate": 7.228075099114594e-06,
+      "loss": 0.2434,
+      "step": 6695
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.4350270608226232,
+      "learning_rate": 7.227268607911533e-06,
+      "loss": 0.2155,
+      "step": 6696
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.0769111427257307,
+      "learning_rate": 7.2264620444100505e-06,
+      "loss": 0.2117,
+      "step": 6697
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.7798393412224454,
+      "learning_rate": 7.22565540863633e-06,
+      "loss": 0.1749,
+      "step": 6698
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.7731838212931246,
+      "learning_rate": 7.224848700616557e-06,
+      "loss": 0.2602,
+      "step": 6699
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.8558324754771645,
+      "learning_rate": 7.224041920376914e-06,
+      "loss": 0.1875,
+      "step": 6700
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.0239579246179975,
+      "learning_rate": 7.223235067943591e-06,
+      "loss": 0.1864,
+      "step": 6701
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.6878538577509938,
+      "learning_rate": 7.222428143342781e-06,
+      "loss": 0.2473,
+      "step": 6702
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.5340240734686026,
+      "learning_rate": 7.221621146600675e-06,
+      "loss": 0.2473,
+      "step": 6703
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.8706311229064673,
+      "learning_rate": 7.220814077743468e-06,
+      "loss": 0.1834,
+      "step": 6704
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.5564175138361764,
+      "learning_rate": 7.220006936797359e-06,
+      "loss": 0.2146,
+      "step": 6705
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.547259559898749,
+      "learning_rate": 7.219199723788549e-06,
+      "loss": 0.2277,
+      "step": 6706
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.4582574915465412,
+      "learning_rate": 7.21839243874324e-06,
+      "loss": 0.1736,
+      "step": 6707
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.0758811848860335,
+      "learning_rate": 7.217585081687637e-06,
+      "loss": 0.16,
+      "step": 6708
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.354012625136919,
+      "learning_rate": 7.216777652647946e-06,
+      "loss": 0.1968,
+      "step": 6709
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.894943985401559,
+      "learning_rate": 7.215970151650378e-06,
+      "loss": 0.2322,
+      "step": 6710
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.9911973830174825,
+      "learning_rate": 7.215162578721146e-06,
+      "loss": 0.1909,
+      "step": 6711
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.907384222324215,
+      "learning_rate": 7.2143549338864604e-06,
+      "loss": 0.2295,
+      "step": 6712
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.4555801293699708,
+      "learning_rate": 7.213547217172542e-06,
+      "loss": 0.2224,
+      "step": 6713
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.0786113151853576,
+      "learning_rate": 7.212739428605606e-06,
+      "loss": 0.1997,
+      "step": 6714
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.248941462065853,
+      "learning_rate": 7.211931568211878e-06,
+      "loss": 0.1799,
+      "step": 6715
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.9687379578641795,
+      "learning_rate": 7.2111236360175785e-06,
+      "loss": 0.2057,
+      "step": 6716
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.4725751917763947,
+      "learning_rate": 7.210315632048934e-06,
+      "loss": 0.1892,
+      "step": 6717
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.0836682906980863,
+      "learning_rate": 7.209507556332173e-06,
+      "loss": 0.2336,
+      "step": 6718
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.3542568274585207,
+      "learning_rate": 7.208699408893528e-06,
+      "loss": 0.2438,
+      "step": 6719
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.407100600376805,
+      "learning_rate": 7.207891189759229e-06,
+      "loss": 0.2415,
+      "step": 6720
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.4519414799151165,
+      "learning_rate": 7.207082898955513e-06,
+      "loss": 0.2419,
+      "step": 6721
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.9196952937642014,
+      "learning_rate": 7.206274536508615e-06,
+      "loss": 0.1468,
+      "step": 6722
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.58847317443734,
+      "learning_rate": 7.205466102444777e-06,
+      "loss": 0.2137,
+      "step": 6723
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.917664136464613,
+      "learning_rate": 7.204657596790243e-06,
+      "loss": 0.1685,
+      "step": 6724
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.3410617974875,
+      "learning_rate": 7.203849019571254e-06,
+      "loss": 0.2341,
+      "step": 6725
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.2908144980317107,
+      "learning_rate": 7.20304037081406e-06,
+      "loss": 0.1759,
+      "step": 6726
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.3541846225232768,
+      "learning_rate": 7.202231650544906e-06,
+      "loss": 0.2012,
+      "step": 6727
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.0257952164398627,
+      "learning_rate": 7.2014228587900496e-06,
+      "loss": 0.2203,
+      "step": 6728
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.6681019429933391,
+      "learning_rate": 7.20061399557574e-06,
+      "loss": 0.1352,
+      "step": 6729
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.4984712061904748,
+      "learning_rate": 7.199805060928234e-06,
+      "loss": 0.191,
+      "step": 6730
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.9826764939027913,
+      "learning_rate": 7.198996054873791e-06,
+      "loss": 0.1482,
+      "step": 6731
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.7088976723611857,
+      "learning_rate": 7.198186977438671e-06,
+      "loss": 0.192,
+      "step": 6732
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.9614682979975526,
+      "learning_rate": 7.197377828649139e-06,
+      "loss": 0.1906,
+      "step": 6733
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.135001698032655,
+      "learning_rate": 7.196568608531458e-06,
+      "loss": 0.1674,
+      "step": 6734
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.0984620296403684,
+      "learning_rate": 7.195759317111898e-06,
+      "loss": 0.2292,
+      "step": 6735
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.2863047028004004,
+      "learning_rate": 7.194949954416727e-06,
+      "loss": 0.1889,
+      "step": 6736
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.1616807491767502,
+      "learning_rate": 7.194140520472219e-06,
+      "loss": 0.1867,
+      "step": 6737
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.1967287337766668,
+      "learning_rate": 7.193331015304648e-06,
+      "loss": 0.2487,
+      "step": 6738
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.9079903537945105,
+      "learning_rate": 7.192521438940291e-06,
+      "loss": 0.212,
+      "step": 6739
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.074883928693121,
+      "learning_rate": 7.191711791405426e-06,
+      "loss": 0.2398,
+      "step": 6740
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.0826922284524942,
+      "learning_rate": 7.190902072726336e-06,
+      "loss": 0.2204,
+      "step": 6741
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.9962925151652201,
+      "learning_rate": 7.190092282929307e-06,
+      "loss": 0.2221,
+      "step": 6742
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.0028983011878367,
+      "learning_rate": 7.18928242204062e-06,
+      "loss": 0.1711,
+      "step": 6743
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.1375337215811832,
+      "learning_rate": 7.188472490086569e-06,
+      "loss": 0.2102,
+      "step": 6744
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.9606159421368711,
+      "learning_rate": 7.187662487093441e-06,
+      "loss": 0.1712,
+      "step": 6745
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.439314853051944,
+      "learning_rate": 7.186852413087533e-06,
+      "loss": 0.2123,
+      "step": 6746
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.2679928378921437,
+      "learning_rate": 7.186042268095137e-06,
+      "loss": 0.2182,
+      "step": 6747
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.241084548959192,
+      "learning_rate": 7.18523205214255e-06,
+      "loss": 0.2155,
+      "step": 6748
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.0204919211505554,
+      "learning_rate": 7.184421765256077e-06,
+      "loss": 0.1442,
+      "step": 6749
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.9918332518809805,
+      "learning_rate": 7.183611407462015e-06,
+      "loss": 0.2006,
+      "step": 6750
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.9198105718416592,
+      "learning_rate": 7.1828009787866726e-06,
+      "loss": 0.1895,
+      "step": 6751
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.0142128443542344,
+      "learning_rate": 7.181990479256354e-06,
+      "loss": 0.1968,
+      "step": 6752
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.854112769768019,
+      "learning_rate": 7.181179908897371e-06,
+      "loss": 0.1789,
+      "step": 6753
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.1809298718577816,
+      "learning_rate": 7.180369267736033e-06,
+      "loss": 0.2597,
+      "step": 6754
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.4629583177578571,
+      "learning_rate": 7.179558555798655e-06,
+      "loss": 0.2256,
+      "step": 6755
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.9793338971011948,
+      "learning_rate": 7.178747773111554e-06,
+      "loss": 0.1938,
+      "step": 6756
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.0011703340873515,
+      "learning_rate": 7.177936919701045e-06,
+      "loss": 0.2167,
+      "step": 6757
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.145773977851965,
+      "learning_rate": 7.177125995593452e-06,
+      "loss": 0.1826,
+      "step": 6758
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.2184302020484208,
+      "learning_rate": 7.1763150008150985e-06,
+      "loss": 0.1876,
+      "step": 6759
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.419639176216563,
+      "learning_rate": 7.175503935392308e-06,
+      "loss": 0.1699,
+      "step": 6760
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.7522080064869303,
+      "learning_rate": 7.1746927993514085e-06,
+      "loss": 0.2759,
+      "step": 6761
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.0795182989836873,
+      "learning_rate": 7.17388159271873e-06,
+      "loss": 0.2179,
+      "step": 6762
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.2104558786791433,
+      "learning_rate": 7.173070315520605e-06,
+      "loss": 0.2392,
+      "step": 6763
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.0517945278775949,
+      "learning_rate": 7.172258967783369e-06,
+      "loss": 0.2073,
+      "step": 6764
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.3819761879045835,
+      "learning_rate": 7.171447549533357e-06,
+      "loss": 0.2041,
+      "step": 6765
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.6504638962813133,
+      "learning_rate": 7.170636060796909e-06,
+      "loss": 0.1869,
+      "step": 6766
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.9197579924275575,
+      "learning_rate": 7.169824501600366e-06,
+      "loss": 0.2137,
+      "step": 6767
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.9100699977362863,
+      "learning_rate": 7.169012871970071e-06,
+      "loss": 0.1482,
+      "step": 6768
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.1003882773065896,
+      "learning_rate": 7.168201171932373e-06,
+      "loss": 0.2048,
+      "step": 6769
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.138081143815757,
+      "learning_rate": 7.167389401513617e-06,
+      "loss": 0.1623,
+      "step": 6770
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.2508789578313533,
+      "learning_rate": 7.166577560740155e-06,
+      "loss": 0.2296,
+      "step": 6771
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.9767065666743596,
+      "learning_rate": 7.165765649638339e-06,
+      "loss": 0.1357,
+      "step": 6772
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.9370352586549471,
+      "learning_rate": 7.164953668234526e-06,
+      "loss": 0.1725,
+      "step": 6773
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.8712723018707628,
+      "learning_rate": 7.16414161655507e-06,
+      "loss": 0.2162,
+      "step": 6774
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.1136621593210767,
+      "learning_rate": 7.1633294946263336e-06,
+      "loss": 0.1258,
+      "step": 6775
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.846903125799123,
+      "learning_rate": 7.162517302474679e-06,
+      "loss": 0.2032,
+      "step": 6776
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.2879599711359495,
+      "learning_rate": 7.161705040126466e-06,
+      "loss": 0.202,
+      "step": 6777
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.2076542459472026,
+      "learning_rate": 7.1608927076080666e-06,
+      "loss": 0.2333,
+      "step": 6778
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.1859169073038383,
+      "learning_rate": 7.160080304945847e-06,
+      "loss": 0.1663,
+      "step": 6779
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.1739165563167764,
+      "learning_rate": 7.159267832166179e-06,
+      "loss": 0.2309,
+      "step": 6780
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.419736560183348,
+      "learning_rate": 7.158455289295435e-06,
+      "loss": 0.2258,
+      "step": 6781
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.7147938393017148,
+      "learning_rate": 7.157642676359992e-06,
+      "loss": 0.1829,
+      "step": 6782
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.9679018545550068,
+      "learning_rate": 7.156829993386226e-06,
+      "loss": 0.1899,
+      "step": 6783
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.92783760701094,
+      "learning_rate": 7.156017240400518e-06,
+      "loss": 0.2107,
+      "step": 6784
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.2420995204286769,
+      "learning_rate": 7.155204417429251e-06,
+      "loss": 0.2011,
+      "step": 6785
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.0214208214794032,
+      "learning_rate": 7.154391524498809e-06,
+      "loss": 0.1924,
+      "step": 6786
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.9638107944212705,
+      "learning_rate": 7.1535785616355805e-06,
+      "loss": 0.1802,
+      "step": 6787
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.229921491722576,
+      "learning_rate": 7.152765528865951e-06,
+      "loss": 0.1944,
+      "step": 6788
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.5581880928303284,
+      "learning_rate": 7.1519524262163164e-06,
+      "loss": 0.2562,
+      "step": 6789
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.087592390536974,
+      "learning_rate": 7.151139253713069e-06,
+      "loss": 0.1968,
+      "step": 6790
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 2.659636876719576,
+      "learning_rate": 7.1503260113826035e-06,
+      "loss": 0.1983,
+      "step": 6791
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.2695433542701584,
+      "learning_rate": 7.14951269925132e-06,
+      "loss": 0.1894,
+      "step": 6792
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.1598946135045893,
+      "learning_rate": 7.1486993173456155e-06,
+      "loss": 0.1795,
+      "step": 6793
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.2046983497591186,
+      "learning_rate": 7.147885865691899e-06,
+      "loss": 0.2019,
+      "step": 6794
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.391756366351145,
+      "learning_rate": 7.14707234431657e-06,
+      "loss": 0.1784,
+      "step": 6795
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.1961976931186649,
+      "learning_rate": 7.14625875324604e-06,
+      "loss": 0.1817,
+      "step": 6796
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.3087239955023726,
+      "learning_rate": 7.145445092506713e-06,
+      "loss": 0.1795,
+      "step": 6797
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.0407003028933883,
+      "learning_rate": 7.144631362125009e-06,
+      "loss": 0.2249,
+      "step": 6798
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.2583994858160275,
+      "learning_rate": 7.1438175621273345e-06,
+      "loss": 0.176,
+      "step": 6799
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.4698559523495964,
+      "learning_rate": 7.143003692540112e-06,
+      "loss": 0.2241,
+      "step": 6800
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.0220765718501323,
+      "learning_rate": 7.142189753389754e-06,
+      "loss": 0.1732,
+      "step": 6801
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.2672019453201375,
+      "learning_rate": 7.141375744702687e-06,
+      "loss": 0.26,
+      "step": 6802
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.1780126299884344,
+      "learning_rate": 7.140561666505331e-06,
+      "loss": 0.2219,
+      "step": 6803
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.0351652990701405,
+      "learning_rate": 7.1397475188241125e-06,
+      "loss": 0.2274,
+      "step": 6804
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.5765226203139202,
+      "learning_rate": 7.13893330168546e-06,
+      "loss": 0.1838,
+      "step": 6805
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.3191359401548526,
+      "learning_rate": 7.138119015115802e-06,
+      "loss": 0.182,
+      "step": 6806
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.6917342755493232,
+      "learning_rate": 7.137304659141571e-06,
+      "loss": 0.2305,
+      "step": 6807
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.097639850751884,
+      "learning_rate": 7.136490233789202e-06,
+      "loss": 0.2133,
+      "step": 6808
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.4702020173921186,
+      "learning_rate": 7.135675739085131e-06,
+      "loss": 0.2439,
+      "step": 6809
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.4133250331475615,
+      "learning_rate": 7.134861175055799e-06,
+      "loss": 0.2012,
+      "step": 6810
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.157158109998389,
+      "learning_rate": 7.134046541727644e-06,
+      "loss": 0.1703,
+      "step": 6811
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.4194889274338653,
+      "learning_rate": 7.133231839127111e-06,
+      "loss": 0.2696,
+      "step": 6812
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.4140824901369156,
+      "learning_rate": 7.132417067280646e-06,
+      "loss": 0.2172,
+      "step": 6813
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.880679715415711,
+      "learning_rate": 7.131602226214698e-06,
+      "loss": 0.2014,
+      "step": 6814
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.0148635944543563,
+      "learning_rate": 7.130787315955714e-06,
+      "loss": 0.2437,
+      "step": 6815
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.8883678220049237,
+      "learning_rate": 7.129972336530151e-06,
+      "loss": 0.1895,
+      "step": 6816
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.0169269176715843,
+      "learning_rate": 7.129157287964459e-06,
+      "loss": 0.232,
+      "step": 6817
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.1381065869557039,
+      "learning_rate": 7.128342170285097e-06,
+      "loss": 0.1486,
+      "step": 6818
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.1484005039769627,
+      "learning_rate": 7.127526983518526e-06,
+      "loss": 0.2009,
+      "step": 6819
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.997922755248039,
+      "learning_rate": 7.126711727691205e-06,
+      "loss": 0.1794,
+      "step": 6820
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.7692447119203467,
+      "learning_rate": 7.125896402829598e-06,
+      "loss": 0.1926,
+      "step": 6821
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.9301956475006206,
+      "learning_rate": 7.125081008960171e-06,
+      "loss": 0.1763,
+      "step": 6822
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.2728263347912518,
+      "learning_rate": 7.1242655461093925e-06,
+      "loss": 0.238,
+      "step": 6823
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.0508384609246992,
+      "learning_rate": 7.123450014303732e-06,
+      "loss": 0.1884,
+      "step": 6824
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 2.195471924121211,
+      "learning_rate": 7.122634413569666e-06,
+      "loss": 0.2199,
+      "step": 6825
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.0635132551212367,
+      "learning_rate": 7.1218187439336625e-06,
+      "loss": 0.1839,
+      "step": 6826
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.8492467128147092,
+      "learning_rate": 7.121003005422206e-06,
+      "loss": 0.1916,
+      "step": 6827
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.0522566111098766,
+      "learning_rate": 7.120187198061769e-06,
+      "loss": 0.2093,
+      "step": 6828
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.1836820827545274,
+      "learning_rate": 7.119371321878839e-06,
+      "loss": 0.1841,
+      "step": 6829
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.9208439904922481,
+      "learning_rate": 7.118555376899896e-06,
+      "loss": 0.1133,
+      "step": 6830
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.9666548056207079,
+      "learning_rate": 7.117739363151428e-06,
+      "loss": 0.2125,
+      "step": 6831
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.1846476951790756,
+      "learning_rate": 7.116923280659922e-06,
+      "loss": 0.2104,
+      "step": 6832
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.8036569464368062,
+      "learning_rate": 7.116107129451868e-06,
+      "loss": 0.2434,
+      "step": 6833
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.1754596910504114,
+      "learning_rate": 7.115290909553761e-06,
+      "loss": 0.2559,
+      "step": 6834
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.9447320267205993,
+      "learning_rate": 7.114474620992093e-06,
+      "loss": 0.1506,
+      "step": 6835
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.9775740267572353,
+      "learning_rate": 7.113658263793365e-06,
+      "loss": 0.1738,
+      "step": 6836
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.2664411556462416,
+      "learning_rate": 7.112841837984074e-06,
+      "loss": 0.2101,
+      "step": 6837
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.047775678206662,
+      "learning_rate": 7.112025343590721e-06,
+      "loss": 0.2221,
+      "step": 6838
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.0857724383347003,
+      "learning_rate": 7.111208780639812e-06,
+      "loss": 0.2457,
+      "step": 6839
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.851670807329064,
+      "learning_rate": 7.110392149157851e-06,
+      "loss": 0.2334,
+      "step": 6840
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.0605226763358404,
+      "learning_rate": 7.109575449171348e-06,
+      "loss": 0.2099,
+      "step": 6841
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.4355224996560376,
+      "learning_rate": 7.108758680706811e-06,
+      "loss": 0.2137,
+      "step": 6842
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.8473659470935972,
+      "learning_rate": 7.1079418437907565e-06,
+      "loss": 0.176,
+      "step": 6843
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.3213939477098195,
+      "learning_rate": 7.107124938449696e-06,
+      "loss": 0.236,
+      "step": 6844
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.0273110471531732,
+      "learning_rate": 7.1063079647101504e-06,
+      "loss": 0.2364,
+      "step": 6845
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.1949558910405347,
+      "learning_rate": 7.105490922598637e-06,
+      "loss": 0.1759,
+      "step": 6846
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.1615147391622678,
+      "learning_rate": 7.104673812141676e-06,
+      "loss": 0.2115,
+      "step": 6847
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.2346446310183739,
+      "learning_rate": 7.103856633365793e-06,
+      "loss": 0.1998,
+      "step": 6848
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.3232454025398415,
+      "learning_rate": 7.103039386297514e-06,
+      "loss": 0.2176,
+      "step": 6849
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.8845400196295023,
+      "learning_rate": 7.102222070963367e-06,
+      "loss": 0.1953,
+      "step": 6850
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.7418087959336079,
+      "learning_rate": 7.101404687389883e-06,
+      "loss": 0.2029,
+      "step": 6851
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.1260363123740267,
+      "learning_rate": 7.1005872356035955e-06,
+      "loss": 0.1883,
+      "step": 6852
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 2.4634013361141256,
+      "learning_rate": 7.099769715631037e-06,
+      "loss": 0.2165,
+      "step": 6853
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.1582710284805227,
+      "learning_rate": 7.098952127498746e-06,
+      "loss": 0.1665,
+      "step": 6854
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.291751777933371,
+      "learning_rate": 7.098134471233264e-06,
+      "loss": 0.2177,
+      "step": 6855
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.0341339209266436,
+      "learning_rate": 7.097316746861128e-06,
+      "loss": 0.2297,
+      "step": 6856
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.2257251447671225,
+      "learning_rate": 7.096498954408886e-06,
+      "loss": 0.2399,
+      "step": 6857
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.2156975695687378,
+      "learning_rate": 7.095681093903081e-06,
+      "loss": 0.2188,
+      "step": 6858
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.135273035359985,
+      "learning_rate": 7.094863165370264e-06,
+      "loss": 0.2089,
+      "step": 6859
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.1299246497556352,
+      "learning_rate": 7.094045168836983e-06,
+      "loss": 0.248,
+      "step": 6860
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.8946464754242351,
+      "learning_rate": 7.093227104329792e-06,
+      "loss": 0.2245,
+      "step": 6861
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.604088348767135,
+      "learning_rate": 7.092408971875244e-06,
+      "loss": 0.2034,
+      "step": 6862
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.9436115605059483,
+      "learning_rate": 7.091590771499901e-06,
+      "loss": 0.1735,
+      "step": 6863
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.4859599636237122,
+      "learning_rate": 7.0907725032303166e-06,
+      "loss": 0.1971,
+      "step": 6864
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.2490952868521226,
+      "learning_rate": 7.0899541670930535e-06,
+      "loss": 0.1438,
+      "step": 6865
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.2618907797902794,
+      "learning_rate": 7.089135763114679e-06,
+      "loss": 0.2357,
+      "step": 6866
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.132732368792128,
+      "learning_rate": 7.088317291321754e-06,
+      "loss": 0.2113,
+      "step": 6867
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.2056491679710346,
+      "learning_rate": 7.08749875174085e-06,
+      "loss": 0.1992,
+      "step": 6868
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.8753615743471469,
+      "learning_rate": 7.086680144398535e-06,
+      "loss": 0.1795,
+      "step": 6869
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.9985639124975692,
+      "learning_rate": 7.085861469321385e-06,
+      "loss": 0.2262,
+      "step": 6870
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.9430521115146673,
+      "learning_rate": 7.085042726535969e-06,
+      "loss": 0.1982,
+      "step": 6871
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.2834240216666,
+      "learning_rate": 7.0842239160688695e-06,
+      "loss": 0.2042,
+      "step": 6872
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.4005380497029467,
+      "learning_rate": 7.083405037946664e-06,
+      "loss": 0.2071,
+      "step": 6873
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.9401070109541848,
+      "learning_rate": 7.08258609219593e-06,
+      "loss": 0.2161,
+      "step": 6874
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.3082184637747485,
+      "learning_rate": 7.081767078843256e-06,
+      "loss": 0.2886,
+      "step": 6875
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.9226801738334469,
+      "learning_rate": 7.080947997915224e-06,
+      "loss": 0.1747,
+      "step": 6876
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.3141145142501836,
+      "learning_rate": 7.080128849438424e-06,
+      "loss": 0.2426,
+      "step": 6877
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.9733199168718591,
+      "learning_rate": 7.079309633439445e-06,
+      "loss": 0.2265,
+      "step": 6878
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.0048998728379943,
+      "learning_rate": 7.078490349944881e-06,
+      "loss": 0.1998,
+      "step": 6879
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.1714961265999766,
+      "learning_rate": 7.077670998981323e-06,
+      "loss": 0.1629,
+      "step": 6880
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.3351112482021463,
+      "learning_rate": 7.076851580575373e-06,
+      "loss": 0.2115,
+      "step": 6881
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.5534174778057945,
+      "learning_rate": 7.076032094753623e-06,
+      "loss": 0.1811,
+      "step": 6882
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.343679069986847,
+      "learning_rate": 7.075212541542679e-06,
+      "loss": 0.1987,
+      "step": 6883
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.0224333707487165,
+      "learning_rate": 7.074392920969142e-06,
+      "loss": 0.1956,
+      "step": 6884
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.1357945932776632,
+      "learning_rate": 7.073573233059618e-06,
+      "loss": 0.2092,
+      "step": 6885
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.9592318671933268,
+      "learning_rate": 7.072753477840715e-06,
+      "loss": 0.2382,
+      "step": 6886
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.464154662495526,
+      "learning_rate": 7.071933655339041e-06,
+      "loss": 0.222,
+      "step": 6887
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.186143999692389,
+      "learning_rate": 7.07111376558121e-06,
+      "loss": 0.202,
+      "step": 6888
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.7897974052567036,
+      "learning_rate": 7.070293808593835e-06,
+      "loss": 0.1432,
+      "step": 6889
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.2955042808090051,
+      "learning_rate": 7.069473784403532e-06,
+      "loss": 0.1973,
+      "step": 6890
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.291360987812213,
+      "learning_rate": 7.068653693036921e-06,
+      "loss": 0.1535,
+      "step": 6891
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.1807419646303896,
+      "learning_rate": 7.0678335345206205e-06,
+      "loss": 0.2194,
+      "step": 6892
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.4458833907263091,
+      "learning_rate": 7.0670133088812555e-06,
+      "loss": 0.2221,
+      "step": 6893
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.0721190140417423,
+      "learning_rate": 7.0661930161454485e-06,
+      "loss": 0.1746,
+      "step": 6894
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.3837417885210184,
+      "learning_rate": 7.06537265633983e-06,
+      "loss": 0.2117,
+      "step": 6895
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.169093386355012,
+      "learning_rate": 7.064552229491025e-06,
+      "loss": 0.1717,
+      "step": 6896
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.3705044604876537,
+      "learning_rate": 7.06373173562567e-06,
+      "loss": 0.2207,
+      "step": 6897
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.080465678378226,
+      "learning_rate": 7.062911174770392e-06,
+      "loss": 0.1626,
+      "step": 6898
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.4137274743051285,
+      "learning_rate": 7.062090546951835e-06,
+      "loss": 0.2071,
+      "step": 6899
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.0335196639344015,
+      "learning_rate": 7.061269852196633e-06,
+      "loss": 0.2434,
+      "step": 6900
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.95962381118665,
+      "learning_rate": 7.060449090531425e-06,
+      "loss": 0.209,
+      "step": 6901
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.044668088379447,
+      "learning_rate": 7.059628261982855e-06,
+      "loss": 0.1594,
+      "step": 6902
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.2223537850946316,
+      "learning_rate": 7.058807366577565e-06,
+      "loss": 0.1711,
+      "step": 6903
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.2392546441027914,
+      "learning_rate": 7.0579864043422075e-06,
+      "loss": 0.21,
+      "step": 6904
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.0995932284807712,
+      "learning_rate": 7.057165375303425e-06,
+      "loss": 0.1865,
+      "step": 6905
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.7816997735417293,
+      "learning_rate": 7.056344279487875e-06,
+      "loss": 0.2305,
+      "step": 6906
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.076289639546012,
+      "learning_rate": 7.055523116922204e-06,
+      "loss": 0.2573,
+      "step": 6907
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.5591379365297167,
+      "learning_rate": 7.054701887633072e-06,
+      "loss": 0.2467,
+      "step": 6908
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.017429952395337,
+      "learning_rate": 7.053880591647136e-06,
+      "loss": 0.1466,
+      "step": 6909
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.2540460995795422,
+      "learning_rate": 7.0530592289910525e-06,
+      "loss": 0.14,
+      "step": 6910
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.022254427483016,
+      "learning_rate": 7.0522377996914885e-06,
+      "loss": 0.1699,
+      "step": 6911
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.1991510962373853,
+      "learning_rate": 7.051416303775102e-06,
+      "loss": 0.2184,
+      "step": 6912
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.1207316392001494,
+      "learning_rate": 7.050594741268566e-06,
+      "loss": 0.1952,
+      "step": 6913
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.2678110029255465,
+      "learning_rate": 7.049773112198543e-06,
+      "loss": 0.2004,
+      "step": 6914
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.5367749951982697,
+      "learning_rate": 7.048951416591708e-06,
+      "loss": 0.1779,
+      "step": 6915
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.896532713972683,
+      "learning_rate": 7.04812965447473e-06,
+      "loss": 0.1843,
+      "step": 6916
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.0992468004097702,
+      "learning_rate": 7.047307825874288e-06,
+      "loss": 0.179,
+      "step": 6917
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.0802250929411807,
+      "learning_rate": 7.046485930817055e-06,
+      "loss": 0.1725,
+      "step": 6918
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.0423887601085553,
+      "learning_rate": 7.045663969329712e-06,
+      "loss": 0.1717,
+      "step": 6919
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.0396509637395253,
+      "learning_rate": 7.044841941438941e-06,
+      "loss": 0.181,
+      "step": 6920
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.9476153632602896,
+      "learning_rate": 7.044019847171424e-06,
+      "loss": 0.1428,
+      "step": 6921
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.9170586213573411,
+      "learning_rate": 7.043197686553848e-06,
+      "loss": 0.1705,
+      "step": 6922
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.9954064896728507,
+      "learning_rate": 7.042375459612899e-06,
+      "loss": 0.2027,
+      "step": 6923
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.8952393893903654,
+      "learning_rate": 7.041553166375269e-06,
+      "loss": 0.1488,
+      "step": 6924
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.1697503331919297,
+      "learning_rate": 7.040730806867649e-06,
+      "loss": 0.1692,
+      "step": 6925
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.31236752057181,
+      "learning_rate": 7.039908381116734e-06,
+      "loss": 0.1848,
+      "step": 6926
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.850930327324766,
+      "learning_rate": 7.03908588914922e-06,
+      "loss": 0.2913,
+      "step": 6927
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.2885988872027183,
+      "learning_rate": 7.038263330991803e-06,
+      "loss": 0.1827,
+      "step": 6928
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.224273903388842,
+      "learning_rate": 7.03744070667119e-06,
+      "loss": 0.2642,
+      "step": 6929
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.8139210124534384,
+      "learning_rate": 7.036618016214077e-06,
+      "loss": 0.1561,
+      "step": 6930
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.6201465505154566,
+      "learning_rate": 7.035795259647175e-06,
+      "loss": 0.23,
+      "step": 6931
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.0699426262505045,
+      "learning_rate": 7.034972436997185e-06,
+      "loss": 0.1924,
+      "step": 6932
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.3402990354496247,
+      "learning_rate": 7.034149548290822e-06,
+      "loss": 0.1889,
+      "step": 6933
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.0700321700538724,
+      "learning_rate": 7.033326593554794e-06,
+      "loss": 0.1703,
+      "step": 6934
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.3361085202245293,
+      "learning_rate": 7.032503572815815e-06,
+      "loss": 0.2041,
+      "step": 6935
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.7918901683514541,
+      "learning_rate": 7.031680486100603e-06,
+      "loss": 0.1208,
+      "step": 6936
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.2041115462607803,
+      "learning_rate": 7.030857333435872e-06,
+      "loss": 0.1987,
+      "step": 6937
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.9156503538781151,
+      "learning_rate": 7.030034114848345e-06,
+      "loss": 0.1561,
+      "step": 6938
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.3751412692975318,
+      "learning_rate": 7.0292108303647435e-06,
+      "loss": 0.2145,
+      "step": 6939
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.3582147245046186,
+      "learning_rate": 7.028387480011791e-06,
+      "loss": 0.2327,
+      "step": 6940
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.1912043697478218,
+      "learning_rate": 7.0275640638162145e-06,
+      "loss": 0.2097,
+      "step": 6941
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.8812563904402054,
+      "learning_rate": 7.026740581804743e-06,
+      "loss": 0.1651,
+      "step": 6942
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.1553597595359222,
+      "learning_rate": 7.025917034004107e-06,
+      "loss": 0.1596,
+      "step": 6943
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.2436306790163634,
+      "learning_rate": 7.025093420441038e-06,
+      "loss": 0.1872,
+      "step": 6944
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.078130975996953,
+      "learning_rate": 7.024269741142274e-06,
+      "loss": 0.1964,
+      "step": 6945
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.4693798746205013,
+      "learning_rate": 7.023445996134548e-06,
+      "loss": 0.2328,
+      "step": 6946
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.013682743888901,
+      "learning_rate": 7.022622185444603e-06,
+      "loss": 0.1932,
+      "step": 6947
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.6036420765314787,
+      "learning_rate": 7.021798309099177e-06,
+      "loss": 0.2415,
+      "step": 6948
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.4089317367996632,
+      "learning_rate": 7.020974367125017e-06,
+      "loss": 0.239,
+      "step": 6949
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.1789600611565798,
+      "learning_rate": 7.0201503595488665e-06,
+      "loss": 0.2759,
+      "step": 6950
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.1657834013484587,
+      "learning_rate": 7.019326286397473e-06,
+      "loss": 0.1623,
+      "step": 6951
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.9876649432296858,
+      "learning_rate": 7.018502147697588e-06,
+      "loss": 0.1893,
+      "step": 6952
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.2801426690516262,
+      "learning_rate": 7.017677943475962e-06,
+      "loss": 0.2471,
+      "step": 6953
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.339744202722121,
+      "learning_rate": 7.0168536737593495e-06,
+      "loss": 0.2333,
+      "step": 6954
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.133636116830621,
+      "learning_rate": 7.0160293385745075e-06,
+      "loss": 0.1844,
+      "step": 6955
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.3244524611664539,
+      "learning_rate": 7.015204937948194e-06,
+      "loss": 0.1902,
+      "step": 6956
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.2612730839434405,
+      "learning_rate": 7.014380471907168e-06,
+      "loss": 0.2068,
+      "step": 6957
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.5965808907836745,
+      "learning_rate": 7.013555940478195e-06,
+      "loss": 0.2079,
+      "step": 6958
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.7597785932293806,
+      "learning_rate": 7.012731343688037e-06,
+      "loss": 0.2494,
+      "step": 6959
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.9090024056181167,
+      "learning_rate": 7.011906681563464e-06,
+      "loss": 0.2075,
+      "step": 6960
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.1009046572026433,
+      "learning_rate": 7.011081954131243e-06,
+      "loss": 0.156,
+      "step": 6961
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.3740882993787251,
+      "learning_rate": 7.010257161418143e-06,
+      "loss": 0.208,
+      "step": 6962
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.2171057734579092,
+      "learning_rate": 7.009432303450943e-06,
+      "loss": 0.2287,
+      "step": 6963
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.2278304349803506,
+      "learning_rate": 7.0086073802564115e-06,
+      "loss": 0.2252,
+      "step": 6964
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.735756864343112,
+      "learning_rate": 7.007782391861331e-06,
+      "loss": 0.1533,
+      "step": 6965
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.9149196026684023,
+      "learning_rate": 7.00695733829248e-06,
+      "loss": 0.2458,
+      "step": 6966
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.3044734567964373,
+      "learning_rate": 7.006132219576639e-06,
+      "loss": 0.1955,
+      "step": 6967
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.1152794640106256,
+      "learning_rate": 7.005307035740591e-06,
+      "loss": 0.2159,
+      "step": 6968
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.5986052355196627,
+      "learning_rate": 7.004481786811126e-06,
+      "loss": 0.1768,
+      "step": 6969
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.0927059056412203,
+      "learning_rate": 7.0036564728150295e-06,
+      "loss": 0.2,
+      "step": 6970
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.3709241254057165,
+      "learning_rate": 7.00283109377909e-06,
+      "loss": 0.2308,
+      "step": 6971
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 2.2142513767999197,
+      "learning_rate": 7.002005649730104e-06,
+      "loss": 0.2426,
+      "step": 6972
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.9841920970354686,
+      "learning_rate": 7.001180140694861e-06,
+      "loss": 0.1995,
+      "step": 6973
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.1089727300346275,
+      "learning_rate": 7.0003545667001605e-06,
+      "loss": 0.2113,
+      "step": 6974
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.8979807646428706,
+      "learning_rate": 6.999528927772801e-06,
+      "loss": 0.1901,
+      "step": 6975
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.1278006550028659,
+      "learning_rate": 6.998703223939582e-06,
+      "loss": 0.2096,
+      "step": 6976
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.571978266738517,
+      "learning_rate": 6.997877455227307e-06,
+      "loss": 0.2399,
+      "step": 6977
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.209548325649098,
+      "learning_rate": 6.997051621662781e-06,
+      "loss": 0.2468,
+      "step": 6978
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.321915468877054,
+      "learning_rate": 6.996225723272812e-06,
+      "loss": 0.1924,
+      "step": 6979
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.2444662842404144,
+      "learning_rate": 6.995399760084207e-06,
+      "loss": 0.2368,
+      "step": 6980
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.9628057394081316,
+      "learning_rate": 6.994573732123779e-06,
+      "loss": 0.1863,
+      "step": 6981
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.1577895135180742,
+      "learning_rate": 6.9937476394183404e-06,
+      "loss": 0.2226,
+      "step": 6982
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.1782698291831595,
+      "learning_rate": 6.992921481994706e-06,
+      "loss": 0.1992,
+      "step": 6983
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.903500772956925,
+      "learning_rate": 6.992095259879696e-06,
+      "loss": 0.2021,
+      "step": 6984
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.1243766806284041,
+      "learning_rate": 6.991268973100127e-06,
+      "loss": 0.1837,
+      "step": 6985
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.2017052435447622,
+      "learning_rate": 6.990442621682822e-06,
+      "loss": 0.1915,
+      "step": 6986
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.1824367962069977,
+      "learning_rate": 6.989616205654606e-06,
+      "loss": 0.1678,
+      "step": 6987
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.9711989205416375,
+      "learning_rate": 6.988789725042304e-06,
+      "loss": 0.1782,
+      "step": 6988
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.516247021954435,
+      "learning_rate": 6.987963179872743e-06,
+      "loss": 0.1997,
+      "step": 6989
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.9768696511751251,
+      "learning_rate": 6.987136570172755e-06,
+      "loss": 0.2036,
+      "step": 6990
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.7750918776133445,
+      "learning_rate": 6.98630989596917e-06,
+      "loss": 0.18,
+      "step": 6991
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.043656999869074,
+      "learning_rate": 6.985483157288824e-06,
+      "loss": 0.1877,
+      "step": 6992
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.3853183283598265,
+      "learning_rate": 6.984656354158552e-06,
+      "loss": 0.2351,
+      "step": 6993
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.9811584186185539,
+      "learning_rate": 6.983829486605196e-06,
+      "loss": 0.1453,
+      "step": 6994
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.1224884083441133,
+      "learning_rate": 6.983002554655591e-06,
+      "loss": 0.1904,
+      "step": 6995
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.2087431005953602,
+      "learning_rate": 6.982175558336586e-06,
+      "loss": 0.1844,
+      "step": 6996
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.0465007013212864,
+      "learning_rate": 6.981348497675021e-06,
+      "loss": 0.1888,
+      "step": 6997
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.4755361145919013,
+      "learning_rate": 6.980521372697744e-06,
+      "loss": 0.2072,
+      "step": 6998
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.0298605367203661,
+      "learning_rate": 6.979694183431605e-06,
+      "loss": 0.2332,
+      "step": 6999
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.0007373989304393,
+      "learning_rate": 6.978866929903454e-06,
+      "loss": 0.1481,
+      "step": 7000
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.2235731565481296,
+      "learning_rate": 6.978039612140144e-06,
+      "loss": 0.194,
+      "step": 7001
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.0652870726131534,
+      "learning_rate": 6.977212230168531e-06,
+      "loss": 0.1878,
+      "step": 7002
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.416628101352214,
+      "learning_rate": 6.976384784015473e-06,
+      "loss": 0.203,
+      "step": 7003
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.1780235313550533,
+      "learning_rate": 6.9755572737078266e-06,
+      "loss": 0.1833,
+      "step": 7004
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.5237931557744866,
+      "learning_rate": 6.974729699272458e-06,
+      "loss": 0.2216,
+      "step": 7005
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.431378833727641,
+      "learning_rate": 6.973902060736226e-06,
+      "loss": 0.2912,
+      "step": 7006
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.3581994525946448,
+      "learning_rate": 6.9730743581259976e-06,
+      "loss": 0.1732,
+      "step": 7007
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.1461467440730198,
+      "learning_rate": 6.972246591468642e-06,
+      "loss": 0.1965,
+      "step": 7008
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.229308395646922,
+      "learning_rate": 6.9714187607910264e-06,
+      "loss": 0.1931,
+      "step": 7009
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.2719611901408956,
+      "learning_rate": 6.970590866120026e-06,
+      "loss": 0.2008,
+      "step": 7010
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.1219271194151401,
+      "learning_rate": 6.9697629074825114e-06,
+      "loss": 0.2211,
+      "step": 7011
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.7779791334271036,
+      "learning_rate": 6.9689348849053616e-06,
+      "loss": 0.1602,
+      "step": 7012
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.1607471127046929,
+      "learning_rate": 6.968106798415452e-06,
+      "loss": 0.2139,
+      "step": 7013
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.2449288319254774,
+      "learning_rate": 6.967278648039665e-06,
+      "loss": 0.2301,
+      "step": 7014
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.9279081402723625,
+      "learning_rate": 6.966450433804881e-06,
+      "loss": 0.2023,
+      "step": 7015
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.229581014622536,
+      "learning_rate": 6.965622155737986e-06,
+      "loss": 0.2351,
+      "step": 7016
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.1882644783807734,
+      "learning_rate": 6.964793813865864e-06,
+      "loss": 0.2354,
+      "step": 7017
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.1404525972741493,
+      "learning_rate": 6.963965408215406e-06,
+      "loss": 0.2066,
+      "step": 7018
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.0273826957979018,
+      "learning_rate": 6.963136938813502e-06,
+      "loss": 0.1755,
+      "step": 7019
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.1177146908181383,
+      "learning_rate": 6.962308405687043e-06,
+      "loss": 0.2094,
+      "step": 7020
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.3455104766112758,
+      "learning_rate": 6.961479808862926e-06,
+      "loss": 0.1952,
+      "step": 7021
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.9454175796114898,
+      "learning_rate": 6.960651148368044e-06,
+      "loss": 0.1734,
+      "step": 7022
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.143955255028515,
+      "learning_rate": 6.959822424229301e-06,
+      "loss": 0.2146,
+      "step": 7023
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.7527932935341439,
+      "learning_rate": 6.958993636473595e-06,
+      "loss": 0.2084,
+      "step": 7024
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.5058141335446165,
+      "learning_rate": 6.958164785127827e-06,
+      "loss": 0.2168,
+      "step": 7025
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.2545981117606189,
+      "learning_rate": 6.957335870218906e-06,
+      "loss": 0.2286,
+      "step": 7026
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.2078479377339604,
+      "learning_rate": 6.956506891773734e-06,
+      "loss": 0.2281,
+      "step": 7027
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.9652976906456419,
+      "learning_rate": 6.955677849819227e-06,
+      "loss": 0.2111,
+      "step": 7028
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.2937447273681377,
+      "learning_rate": 6.95484874438229e-06,
+      "loss": 0.2322,
+      "step": 7029
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.0442250504133113,
+      "learning_rate": 6.95401957548984e-06,
+      "loss": 0.2145,
+      "step": 7030
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.1021651520070905,
+      "learning_rate": 6.95319034316879e-06,
+      "loss": 0.2054,
+      "step": 7031
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.9596508472173401,
+      "learning_rate": 6.952361047446059e-06,
+      "loss": 0.2545,
+      "step": 7032
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.8299715744926932,
+      "learning_rate": 6.951531688348566e-06,
+      "loss": 0.2122,
+      "step": 7033
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.1349022728852687,
+      "learning_rate": 6.9507022659032305e-06,
+      "loss": 0.2498,
+      "step": 7034
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.0553088653678082,
+      "learning_rate": 6.949872780136979e-06,
+      "loss": 0.1556,
+      "step": 7035
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.3123161408528796,
+      "learning_rate": 6.949043231076735e-06,
+      "loss": 0.1847,
+      "step": 7036
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.9917103941910845,
+      "learning_rate": 6.948213618749428e-06,
+      "loss": 0.1809,
+      "step": 7037
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 2.1025360366969608,
+      "learning_rate": 6.947383943181986e-06,
+      "loss": 0.1912,
+      "step": 7038
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.0863909717941065,
+      "learning_rate": 6.946554204401342e-06,
+      "loss": 0.2058,
+      "step": 7039
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.551933893160907,
+      "learning_rate": 6.9457244024344284e-06,
+      "loss": 0.2089,
+      "step": 7040
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.6949888944508194,
+      "learning_rate": 6.944894537308183e-06,
+      "loss": 0.223,
+      "step": 7041
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.0405367291876175,
+      "learning_rate": 6.944064609049542e-06,
+      "loss": 0.2393,
+      "step": 7042
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.5668975720683662,
+      "learning_rate": 6.943234617685445e-06,
+      "loss": 0.175,
+      "step": 7043
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.703255525889378,
+      "learning_rate": 6.942404563242836e-06,
+      "loss": 0.1719,
+      "step": 7044
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.3414984514189834,
+      "learning_rate": 6.941574445748657e-06,
+      "loss": 0.2351,
+      "step": 7045
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.0212232044322085,
+      "learning_rate": 6.940744265229855e-06,
+      "loss": 0.252,
+      "step": 7046
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.4189426542862358,
+      "learning_rate": 6.939914021713379e-06,
+      "loss": 0.1885,
+      "step": 7047
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.8840819690130834,
+      "learning_rate": 6.939083715226178e-06,
+      "loss": 0.2052,
+      "step": 7048
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.963008856913813,
+      "learning_rate": 6.938253345795204e-06,
+      "loss": 0.2325,
+      "step": 7049
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.2849216739879132,
+      "learning_rate": 6.937422913447412e-06,
+      "loss": 0.2455,
+      "step": 7050
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.0992643889094265,
+      "learning_rate": 6.93659241820976e-06,
+      "loss": 0.1969,
+      "step": 7051
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.0124623114288345,
+      "learning_rate": 6.935761860109201e-06,
+      "loss": 0.1573,
+      "step": 7052
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.4494419722857146,
+      "learning_rate": 6.934931239172702e-06,
+      "loss": 0.1578,
+      "step": 7053
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.190380054719114,
+      "learning_rate": 6.93410055542722e-06,
+      "loss": 0.2032,
+      "step": 7054
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 7.9359804636345626,
+      "learning_rate": 6.933269808899723e-06,
+      "loss": 0.157,
+      "step": 7055
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.4173582433658258,
+      "learning_rate": 6.932438999617176e-06,
+      "loss": 0.2288,
+      "step": 7056
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.1962102727829307,
+      "learning_rate": 6.931608127606549e-06,
+      "loss": 0.1979,
+      "step": 7057
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.1145296024431732,
+      "learning_rate": 6.930777192894809e-06,
+      "loss": 0.23,
+      "step": 7058
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.0892209449166441,
+      "learning_rate": 6.929946195508933e-06,
+      "loss": 0.2272,
+      "step": 7059
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.2377935104820377,
+      "learning_rate": 6.929115135475893e-06,
+      "loss": 0.1582,
+      "step": 7060
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.138341202986504,
+      "learning_rate": 6.928284012822667e-06,
+      "loss": 0.1585,
+      "step": 7061
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.015063167107443,
+      "learning_rate": 6.927452827576233e-06,
+      "loss": 0.1704,
+      "step": 7062
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.0030980205695736,
+      "learning_rate": 6.926621579763571e-06,
+      "loss": 0.2229,
+      "step": 7063
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.1483613624765767,
+      "learning_rate": 6.925790269411666e-06,
+      "loss": 0.1883,
+      "step": 7064
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.1361321446490737,
+      "learning_rate": 6.924958896547499e-06,
+      "loss": 0.1977,
+      "step": 7065
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.4832997254708131,
+      "learning_rate": 6.924127461198062e-06,
+      "loss": 0.2276,
+      "step": 7066
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.251468549790627,
+      "learning_rate": 6.923295963390338e-06,
+      "loss": 0.2504,
+      "step": 7067
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.1055251097044545,
+      "learning_rate": 6.922464403151324e-06,
+      "loss": 0.2413,
+      "step": 7068
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.6601651516844012,
+      "learning_rate": 6.9216327805080095e-06,
+      "loss": 0.2464,
+      "step": 7069
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.4088249209383474,
+      "learning_rate": 6.920801095487388e-06,
+      "loss": 0.1931,
+      "step": 7070
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.5531415369508141,
+      "learning_rate": 6.919969348116461e-06,
+      "loss": 0.2357,
+      "step": 7071
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.1769804803218489,
+      "learning_rate": 6.919137538422224e-06,
+      "loss": 0.2223,
+      "step": 7072
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.102193631305735,
+      "learning_rate": 6.918305666431678e-06,
+      "loss": 0.133,
+      "step": 7073
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.9942240689375327,
+      "learning_rate": 6.917473732171828e-06,
+      "loss": 0.2059,
+      "step": 7074
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.3085153840925872,
+      "learning_rate": 6.916641735669678e-06,
+      "loss": 0.2086,
+      "step": 7075
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.3626081655049613,
+      "learning_rate": 6.915809676952234e-06,
+      "loss": 0.2626,
+      "step": 7076
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.0044908630359395,
+      "learning_rate": 6.914977556046507e-06,
+      "loss": 0.1795,
+      "step": 7077
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.4253658734736703,
+      "learning_rate": 6.914145372979508e-06,
+      "loss": 0.2073,
+      "step": 7078
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.4288669339356037,
+      "learning_rate": 6.913313127778248e-06,
+      "loss": 0.2339,
+      "step": 7079
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.4209020870201199,
+      "learning_rate": 6.912480820469745e-06,
+      "loss": 0.2076,
+      "step": 7080
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.0784186692336188,
+      "learning_rate": 6.911648451081013e-06,
+      "loss": 0.25,
+      "step": 7081
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.1940418171333045,
+      "learning_rate": 6.910816019639075e-06,
+      "loss": 0.1597,
+      "step": 7082
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 2.3315617841191068,
+      "learning_rate": 6.9099835261709495e-06,
+      "loss": 0.2167,
+      "step": 7083
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.0680483952999003,
+      "learning_rate": 6.90915097070366e-06,
+      "loss": 0.194,
+      "step": 7084
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.3547800386670805,
+      "learning_rate": 6.908318353264233e-06,
+      "loss": 0.2105,
+      "step": 7085
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.033361408389975,
+      "learning_rate": 6.907485673879695e-06,
+      "loss": 0.1891,
+      "step": 7086
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.1811903119144833,
+      "learning_rate": 6.9066529325770745e-06,
+      "loss": 0.2085,
+      "step": 7087
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.4246944387450606,
+      "learning_rate": 6.905820129383402e-06,
+      "loss": 0.214,
+      "step": 7088
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.324592728820979,
+      "learning_rate": 6.904987264325715e-06,
+      "loss": 0.164,
+      "step": 7089
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.0493118051718284,
+      "learning_rate": 6.904154337431044e-06,
+      "loss": 0.1846,
+      "step": 7090
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.1235271107080118,
+      "learning_rate": 6.903321348726429e-06,
+      "loss": 0.2103,
+      "step": 7091
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.023636143458245,
+      "learning_rate": 6.9024882982389074e-06,
+      "loss": 0.199,
+      "step": 7092
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.2358356756097082,
+      "learning_rate": 6.901655185995524e-06,
+      "loss": 0.1984,
+      "step": 7093
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.4145202080369286,
+      "learning_rate": 6.900822012023317e-06,
+      "loss": 0.1817,
+      "step": 7094
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.5949905163048026,
+      "learning_rate": 6.899988776349337e-06,
+      "loss": 0.1923,
+      "step": 7095
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.1443864678735904,
+      "learning_rate": 6.8991554790006276e-06,
+      "loss": 0.184,
+      "step": 7096
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.2828779171376468,
+      "learning_rate": 6.898322120004238e-06,
+      "loss": 0.2538,
+      "step": 7097
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.3672480433546481,
+      "learning_rate": 6.897488699387222e-06,
+      "loss": 0.1852,
+      "step": 7098
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.0123657749828208,
+      "learning_rate": 6.896655217176632e-06,
+      "loss": 0.178,
+      "step": 7099
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.3799064725992345,
+      "learning_rate": 6.895821673399523e-06,
+      "loss": 0.1878,
+      "step": 7100
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.552972283521003,
+      "learning_rate": 6.894988068082952e-06,
+      "loss": 0.2229,
+      "step": 7101
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.280009625895076,
+      "learning_rate": 6.894154401253979e-06,
+      "loss": 0.2435,
+      "step": 7102
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.1769391090376395,
+      "learning_rate": 6.893320672939663e-06,
+      "loss": 0.2059,
+      "step": 7103
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.143248480399778,
+      "learning_rate": 6.892486883167071e-06,
+      "loss": 0.2045,
+      "step": 7104
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.281408491304789,
+      "learning_rate": 6.891653031963267e-06,
+      "loss": 0.2102,
+      "step": 7105
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.9477965705462449,
+      "learning_rate": 6.8908191193553165e-06,
+      "loss": 0.1828,
+      "step": 7106
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.1105744130695,
+      "learning_rate": 6.889985145370291e-06,
+      "loss": 0.1789,
+      "step": 7107
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.304646339309738,
+      "learning_rate": 6.88915111003526e-06,
+      "loss": 0.2307,
+      "step": 7108
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.0154999637663622,
+      "learning_rate": 6.888317013377299e-06,
+      "loss": 0.2069,
+      "step": 7109
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.5727967780782863,
+      "learning_rate": 6.88748285542348e-06,
+      "loss": 0.2249,
+      "step": 7110
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.0458070659884973,
+      "learning_rate": 6.886648636200882e-06,
+      "loss": 0.2049,
+      "step": 7111
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.9621866169247442,
+      "learning_rate": 6.8858143557365865e-06,
+      "loss": 0.1869,
+      "step": 7112
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.9029202209280063,
+      "learning_rate": 6.884980014057672e-06,
+      "loss": 0.1872,
+      "step": 7113
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.3602214184113792,
+      "learning_rate": 6.884145611191222e-06,
+      "loss": 0.2303,
+      "step": 7114
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.9864855574557485,
+      "learning_rate": 6.883311147164322e-06,
+      "loss": 0.2214,
+      "step": 7115
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.1031647841841328,
+      "learning_rate": 6.88247662200406e-06,
+      "loss": 0.1827,
+      "step": 7116
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.9821902498779617,
+      "learning_rate": 6.881642035737521e-06,
+      "loss": 0.1607,
+      "step": 7117
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.334633572895048,
+      "learning_rate": 6.880807388391804e-06,
+      "loss": 0.2344,
+      "step": 7118
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.1809616786543902,
+      "learning_rate": 6.879972679993994e-06,
+      "loss": 0.1725,
+      "step": 7119
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 2.3068247825760455,
+      "learning_rate": 6.879137910571191e-06,
+      "loss": 0.2327,
+      "step": 7120
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.3050185428869527,
+      "learning_rate": 6.8783030801504915e-06,
+      "loss": 0.2033,
+      "step": 7121
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.9831853482287901,
+      "learning_rate": 6.877468188758993e-06,
+      "loss": 0.1821,
+      "step": 7122
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.3990593704962309,
+      "learning_rate": 6.8766332364238e-06,
+      "loss": 0.2122,
+      "step": 7123
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.3277717288790378,
+      "learning_rate": 6.87579822317201e-06,
+      "loss": 0.2181,
+      "step": 7124
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.0301597369762938,
+      "learning_rate": 6.874963149030732e-06,
+      "loss": 0.2061,
+      "step": 7125
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.2370143691290738,
+      "learning_rate": 6.874128014027071e-06,
+      "loss": 0.2034,
+      "step": 7126
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.9485812420775201,
+      "learning_rate": 6.873292818188139e-06,
+      "loss": 0.154,
+      "step": 7127
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.3355804016287742,
+      "learning_rate": 6.8724575615410435e-06,
+      "loss": 0.171,
+      "step": 7128
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.0796649101036098,
+      "learning_rate": 6.871622244112899e-06,
+      "loss": 0.1881,
+      "step": 7129
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.4224363957112973,
+      "learning_rate": 6.870786865930821e-06,
+      "loss": 0.2052,
+      "step": 7130
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.3519743804451119,
+      "learning_rate": 6.869951427021925e-06,
+      "loss": 0.192,
+      "step": 7131
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.253816342918539,
+      "learning_rate": 6.86911592741333e-06,
+      "loss": 0.1672,
+      "step": 7132
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.1691998704122397,
+      "learning_rate": 6.868280367132157e-06,
+      "loss": 0.1646,
+      "step": 7133
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.3119866468492194,
+      "learning_rate": 6.867444746205531e-06,
+      "loss": 0.1903,
+      "step": 7134
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.9806209176363314,
+      "learning_rate": 6.8666090646605725e-06,
+      "loss": 0.1932,
+      "step": 7135
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.3136754449074997,
+      "learning_rate": 6.865773322524411e-06,
+      "loss": 0.2162,
+      "step": 7136
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.1292137466389494,
+      "learning_rate": 6.8649375198241744e-06,
+      "loss": 0.2203,
+      "step": 7137
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.0232217493945033,
+      "learning_rate": 6.864101656586992e-06,
+      "loss": 0.1821,
+      "step": 7138
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.9770285448006083,
+      "learning_rate": 6.863265732840001e-06,
+      "loss": 0.2356,
+      "step": 7139
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.066947906869917,
+      "learning_rate": 6.862429748610331e-06,
+      "loss": 0.2192,
+      "step": 7140
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.8418986000145907,
+      "learning_rate": 6.861593703925121e-06,
+      "loss": 0.1976,
+      "step": 7141
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.2863775554835846,
+      "learning_rate": 6.8607575988115085e-06,
+      "loss": 0.2421,
+      "step": 7142
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.2709460151528744,
+      "learning_rate": 6.859921433296635e-06,
+      "loss": 0.2126,
+      "step": 7143
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.9785957265569274,
+      "learning_rate": 6.859085207407641e-06,
+      "loss": 0.1854,
+      "step": 7144
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.3076697745656771,
+      "learning_rate": 6.858248921171673e-06,
+      "loss": 0.2079,
+      "step": 7145
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.9336809602409832,
+      "learning_rate": 6.857412574615876e-06,
+      "loss": 0.1988,
+      "step": 7146
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.180927537496012,
+      "learning_rate": 6.856576167767399e-06,
+      "loss": 0.1706,
+      "step": 7147
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.1373547547745155,
+      "learning_rate": 6.855739700653392e-06,
+      "loss": 0.211,
+      "step": 7148
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.293804958576749,
+      "learning_rate": 6.854903173301008e-06,
+      "loss": 0.2208,
+      "step": 7149
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.0257249650381348,
+      "learning_rate": 6.8540665857374e-06,
+      "loss": 0.1891,
+      "step": 7150
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.0999642793723912,
+      "learning_rate": 6.853229937989724e-06,
+      "loss": 0.1601,
+      "step": 7151
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.2066222967161806,
+      "learning_rate": 6.852393230085138e-06,
+      "loss": 0.2009,
+      "step": 7152
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.448131942282711,
+      "learning_rate": 6.851556462050804e-06,
+      "loss": 0.2386,
+      "step": 7153
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.0277304882351588,
+      "learning_rate": 6.850719633913881e-06,
+      "loss": 0.2167,
+      "step": 7154
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.294054137637938,
+      "learning_rate": 6.8498827457015345e-06,
+      "loss": 0.2263,
+      "step": 7155
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.8759725861900349,
+      "learning_rate": 6.849045797440931e-06,
+      "loss": 0.1663,
+      "step": 7156
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.5876250047169449,
+      "learning_rate": 6.8482087891592375e-06,
+      "loss": 0.1533,
+      "step": 7157
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.425230498609482,
+      "learning_rate": 6.8473717208836245e-06,
+      "loss": 0.1757,
+      "step": 7158
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.7005910247371304,
+      "learning_rate": 6.846534592641263e-06,
+      "loss": 0.1436,
+      "step": 7159
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.0390105055226502,
+      "learning_rate": 6.845697404459326e-06,
+      "loss": 0.1482,
+      "step": 7160
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.274950190931598,
+      "learning_rate": 6.84486015636499e-06,
+      "loss": 0.2366,
+      "step": 7161
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.267403755185439,
+      "learning_rate": 6.844022848385432e-06,
+      "loss": 0.1952,
+      "step": 7162
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.1140006735714478,
+      "learning_rate": 6.843185480547834e-06,
+      "loss": 0.2495,
+      "step": 7163
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.1358295400521297,
+      "learning_rate": 6.842348052879373e-06,
+      "loss": 0.2003,
+      "step": 7164
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.410285227892467,
+      "learning_rate": 6.841510565407235e-06,
+      "loss": 0.1944,
+      "step": 7165
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.1202284866551766,
+      "learning_rate": 6.840673018158607e-06,
+      "loss": 0.2059,
+      "step": 7166
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.3734395752979,
+      "learning_rate": 6.839835411160674e-06,
+      "loss": 0.188,
+      "step": 7167
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.423148700841765,
+      "learning_rate": 6.838997744440625e-06,
+      "loss": 0.2073,
+      "step": 7168
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.1069384702247795,
+      "learning_rate": 6.83816001802565e-06,
+      "loss": 0.1712,
+      "step": 7169
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.0442231899516987,
+      "learning_rate": 6.837322231942946e-06,
+      "loss": 0.1764,
+      "step": 7170
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.2749267365406007,
+      "learning_rate": 6.836484386219705e-06,
+      "loss": 0.1648,
+      "step": 7171
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.7367313955123767,
+      "learning_rate": 6.835646480883126e-06,
+      "loss": 0.2339,
+      "step": 7172
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.9167009852046818,
+      "learning_rate": 6.834808515960405e-06,
+      "loss": 0.2291,
+      "step": 7173
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.2287379027357606,
+      "learning_rate": 6.8339704914787455e-06,
+      "loss": 0.1781,
+      "step": 7174
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.2447014896364217,
+      "learning_rate": 6.833132407465348e-06,
+      "loss": 0.2175,
+      "step": 7175
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.0160103433685121,
+      "learning_rate": 6.83229426394742e-06,
+      "loss": 0.1924,
+      "step": 7176
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.3285174827708508,
+      "learning_rate": 6.831456060952167e-06,
+      "loss": 0.1854,
+      "step": 7177
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.8741602934348793,
+      "learning_rate": 6.830617798506795e-06,
+      "loss": 0.1539,
+      "step": 7178
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.3443157479952248,
+      "learning_rate": 6.829779476638517e-06,
+      "loss": 0.2127,
+      "step": 7179
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.0007512094781978,
+      "learning_rate": 6.828941095374545e-06,
+      "loss": 0.2449,
+      "step": 7180
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.0331875451973886,
+      "learning_rate": 6.8281026547420945e-06,
+      "loss": 0.2282,
+      "step": 7181
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.4400046726944067,
+      "learning_rate": 6.827264154768379e-06,
+      "loss": 0.2072,
+      "step": 7182
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.107756609508913,
+      "learning_rate": 6.82642559548062e-06,
+      "loss": 0.1867,
+      "step": 7183
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.2423805230798304,
+      "learning_rate": 6.825586976906035e-06,
+      "loss": 0.1969,
+      "step": 7184
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.052023938087176,
+      "learning_rate": 6.824748299071847e-06,
+      "loss": 0.1915,
+      "step": 7185
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.1379165270627296,
+      "learning_rate": 6.823909562005281e-06,
+      "loss": 0.2037,
+      "step": 7186
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.0972525271608025,
+      "learning_rate": 6.823070765733559e-06,
+      "loss": 0.1761,
+      "step": 7187
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.167014928620113,
+      "learning_rate": 6.822231910283914e-06,
+      "loss": 0.2491,
+      "step": 7188
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.6372706773209174,
+      "learning_rate": 6.821392995683572e-06,
+      "loss": 0.2194,
+      "step": 7189
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.179304600779096,
+      "learning_rate": 6.8205540219597675e-06,
+      "loss": 0.2316,
+      "step": 7190
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.2292162227204593,
+      "learning_rate": 6.819714989139732e-06,
+      "loss": 0.2041,
+      "step": 7191
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.8982966457745594,
+      "learning_rate": 6.818875897250701e-06,
+      "loss": 0.224,
+      "step": 7192
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.4017120681278563,
+      "learning_rate": 6.818036746319914e-06,
+      "loss": 0.2222,
+      "step": 7193
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.9899161653291546,
+      "learning_rate": 6.817197536374607e-06,
+      "loss": 0.1879,
+      "step": 7194
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.2469864416036363,
+      "learning_rate": 6.8163582674420224e-06,
+      "loss": 0.2144,
+      "step": 7195
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.2283260338653503,
+      "learning_rate": 6.815518939549406e-06,
+      "loss": 0.2319,
+      "step": 7196
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.9128732709006396,
+      "learning_rate": 6.814679552723999e-06,
+      "loss": 0.2077,
+      "step": 7197
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.0739862433148324,
+      "learning_rate": 6.813840106993051e-06,
+      "loss": 0.2115,
+      "step": 7198
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.5666172332607673,
+      "learning_rate": 6.81300060238381e-06,
+      "loss": 0.2261,
+      "step": 7199
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.8333296439218311,
+      "learning_rate": 6.812161038923527e-06,
+      "loss": 0.1791,
+      "step": 7200
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.1853993291025555,
+      "learning_rate": 6.811321416639456e-06,
+      "loss": 0.1986,
+      "step": 7201
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.0916868335223684,
+      "learning_rate": 6.810481735558849e-06,
+      "loss": 0.218,
+      "step": 7202
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.196646653272429,
+      "learning_rate": 6.809641995708962e-06,
+      "loss": 0.2009,
+      "step": 7203
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.076522811468118,
+      "learning_rate": 6.8088021971170584e-06,
+      "loss": 0.2169,
+      "step": 7204
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.156507878929583,
+      "learning_rate": 6.8079623398103926e-06,
+      "loss": 0.2196,
+      "step": 7205
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.3231678607518913,
+      "learning_rate": 6.807122423816232e-06,
+      "loss": 0.1973,
+      "step": 7206
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.2058856088832597,
+      "learning_rate": 6.806282449161836e-06,
+      "loss": 0.1627,
+      "step": 7207
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.362133931495725,
+      "learning_rate": 6.805442415874475e-06,
+      "loss": 0.2114,
+      "step": 7208
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.2891294635372514,
+      "learning_rate": 6.804602323981413e-06,
+      "loss": 0.2038,
+      "step": 7209
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.3790540751034117,
+      "learning_rate": 6.803762173509924e-06,
+      "loss": 0.1481,
+      "step": 7210
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.3153185581539952,
+      "learning_rate": 6.802921964487277e-06,
+      "loss": 0.2351,
+      "step": 7211
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.3310262001106954,
+      "learning_rate": 6.8020816969407454e-06,
+      "loss": 0.2041,
+      "step": 7212
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.6757591119384339,
+      "learning_rate": 6.801241370897607e-06,
+      "loss": 0.206,
+      "step": 7213
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.9998551539482967,
+      "learning_rate": 6.800400986385136e-06,
+      "loss": 0.216,
+      "step": 7214
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.9914755500909905,
+      "learning_rate": 6.799560543430614e-06,
+      "loss": 0.1622,
+      "step": 7215
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.674437386410269,
+      "learning_rate": 6.798720042061323e-06,
+      "loss": 0.232,
+      "step": 7216
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.1351749286046875,
+      "learning_rate": 6.797879482304546e-06,
+      "loss": 0.1854,
+      "step": 7217
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 2.6815141952284023,
+      "learning_rate": 6.797038864187564e-06,
+      "loss": 0.208,
+      "step": 7218
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.3676481179460072,
+      "learning_rate": 6.7961981877376705e-06,
+      "loss": 0.2379,
+      "step": 7219
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.9690403195634908,
+      "learning_rate": 6.7953574529821495e-06,
+      "loss": 0.1956,
+      "step": 7220
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.9708187698310866,
+      "learning_rate": 6.794516659948293e-06,
+      "loss": 0.2213,
+      "step": 7221
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.0866586945570185,
+      "learning_rate": 6.793675808663394e-06,
+      "loss": 0.2299,
+      "step": 7222
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.1754940971832137,
+      "learning_rate": 6.792834899154745e-06,
+      "loss": 0.1642,
+      "step": 7223
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.3137776933667586,
+      "learning_rate": 6.791993931449647e-06,
+      "loss": 0.2031,
+      "step": 7224
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.1145050015360827,
+      "learning_rate": 6.791152905575395e-06,
+      "loss": 0.187,
+      "step": 7225
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.3299147866907306,
+      "learning_rate": 6.790311821559289e-06,
+      "loss": 0.2475,
+      "step": 7226
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.1454022888204811,
+      "learning_rate": 6.789470679428632e-06,
+      "loss": 0.2124,
+      "step": 7227
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.332789561811058,
+      "learning_rate": 6.788629479210728e-06,
+      "loss": 0.2158,
+      "step": 7228
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.8746366853526278,
+      "learning_rate": 6.787788220932883e-06,
+      "loss": 0.1927,
+      "step": 7229
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.0616432339925848,
+      "learning_rate": 6.7869469046224024e-06,
+      "loss": 0.1915,
+      "step": 7230
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.0203109587851973,
+      "learning_rate": 6.786105530306599e-06,
+      "loss": 0.2359,
+      "step": 7231
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.6628833458077792,
+      "learning_rate": 6.785264098012783e-06,
+      "loss": 0.1511,
+      "step": 7232
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.0195683805054905,
+      "learning_rate": 6.7844226077682675e-06,
+      "loss": 0.1991,
+      "step": 7233
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.3648044580525838,
+      "learning_rate": 6.783581059600367e-06,
+      "loss": 0.1576,
+      "step": 7234
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.001322070248032,
+      "learning_rate": 6.7827394535364e-06,
+      "loss": 0.2135,
+      "step": 7235
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.9063898254142226,
+      "learning_rate": 6.781897789603686e-06,
+      "loss": 0.2093,
+      "step": 7236
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.2926509645285504,
+      "learning_rate": 6.781056067829547e-06,
+      "loss": 0.1702,
+      "step": 7237
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.9373954158304009,
+      "learning_rate": 6.780214288241303e-06,
+      "loss": 0.1864,
+      "step": 7238
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.923656736045116,
+      "learning_rate": 6.779372450866277e-06,
+      "loss": 0.1855,
+      "step": 7239
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.42194372576949,
+      "learning_rate": 6.7785305557318e-06,
+      "loss": 0.2312,
+      "step": 7240
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.2977046510258146,
+      "learning_rate": 6.777688602865198e-06,
+      "loss": 0.1904,
+      "step": 7241
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.2542342131414592,
+      "learning_rate": 6.776846592293801e-06,
+      "loss": 0.2142,
+      "step": 7242
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.9609769758784382,
+      "learning_rate": 6.776004524044943e-06,
+      "loss": 0.2185,
+      "step": 7243
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.173011178138201,
+      "learning_rate": 6.775162398145955e-06,
+      "loss": 0.1498,
+      "step": 7244
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.8845521446176585,
+      "learning_rate": 6.774320214624177e-06,
+      "loss": 0.1314,
+      "step": 7245
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.8249507383609027,
+      "learning_rate": 6.773477973506943e-06,
+      "loss": 0.2239,
+      "step": 7246
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.8509375812146628,
+      "learning_rate": 6.772635674821596e-06,
+      "loss": 0.2414,
+      "step": 7247
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.1807736692201647,
+      "learning_rate": 6.771793318595474e-06,
+      "loss": 0.2035,
+      "step": 7248
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.7424052363272893,
+      "learning_rate": 6.770950904855923e-06,
+      "loss": 0.2169,
+      "step": 7249
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.1223586951068916,
+      "learning_rate": 6.770108433630286e-06,
+      "loss": 0.2265,
+      "step": 7250
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.3736652322168244,
+      "learning_rate": 6.769265904945912e-06,
+      "loss": 0.2076,
+      "step": 7251
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.9365493215935857,
+      "learning_rate": 6.768423318830149e-06,
+      "loss": 0.1505,
+      "step": 7252
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.0028664217033696,
+      "learning_rate": 6.767580675310348e-06,
+      "loss": 0.217,
+      "step": 7253
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.1838345204071512,
+      "learning_rate": 6.766737974413863e-06,
+      "loss": 0.2043,
+      "step": 7254
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.0335032914022877,
+      "learning_rate": 6.765895216168048e-06,
+      "loss": 0.1455,
+      "step": 7255
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.1454338527911117,
+      "learning_rate": 6.765052400600258e-06,
+      "loss": 0.2449,
+      "step": 7256
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.144740371998305,
+      "learning_rate": 6.76420952773785e-06,
+      "loss": 0.2349,
+      "step": 7257
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.893424184055507,
+      "learning_rate": 6.763366597608189e-06,
+      "loss": 0.1782,
+      "step": 7258
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 2.035657085216203,
+      "learning_rate": 6.762523610238633e-06,
+      "loss": 0.2402,
+      "step": 7259
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.2428412936143363,
+      "learning_rate": 6.761680565656548e-06,
+      "loss": 0.1823,
+      "step": 7260
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.9271631710986442,
+      "learning_rate": 6.760837463889297e-06,
+      "loss": 0.1596,
+      "step": 7261
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.9245035553343406,
+      "learning_rate": 6.759994304964251e-06,
+      "loss": 0.1805,
+      "step": 7262
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.153866710321165,
+      "learning_rate": 6.7591510889087775e-06,
+      "loss": 0.2379,
+      "step": 7263
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.2213673790852155,
+      "learning_rate": 6.758307815750248e-06,
+      "loss": 0.2138,
+      "step": 7264
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.9316648764022989,
+      "learning_rate": 6.757464485516035e-06,
+      "loss": 0.2143,
+      "step": 7265
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.1602710994553962,
+      "learning_rate": 6.756621098233514e-06,
+      "loss": 0.2144,
+      "step": 7266
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.3121128930569654,
+      "learning_rate": 6.755777653930065e-06,
+      "loss": 0.222,
+      "step": 7267
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.23828610861514,
+      "learning_rate": 6.75493415263306e-06,
+      "loss": 0.2146,
+      "step": 7268
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.218011381615624,
+      "learning_rate": 6.7540905943698856e-06,
+      "loss": 0.2202,
+      "step": 7269
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.2376101522477794,
+      "learning_rate": 6.75324697916792e-06,
+      "loss": 0.1713,
+      "step": 7270
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.1156896824210891,
+      "learning_rate": 6.752403307054549e-06,
+      "loss": 0.1668,
+      "step": 7271
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.1373817977824758,
+      "learning_rate": 6.751559578057161e-06,
+      "loss": 0.1858,
+      "step": 7272
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.9435761717946491,
+      "learning_rate": 6.7507157922031415e-06,
+      "loss": 0.1621,
+      "step": 7273
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.507728860936834,
+      "learning_rate": 6.74987194951988e-06,
+      "loss": 0.1933,
+      "step": 7274
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.139486276813232,
+      "learning_rate": 6.749028050034769e-06,
+      "loss": 0.1684,
+      "step": 7275
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.1222154259533281,
+      "learning_rate": 6.748184093775202e-06,
+      "loss": 0.2201,
+      "step": 7276
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.5054731059147484,
+      "learning_rate": 6.747340080768574e-06,
+      "loss": 0.2017,
+      "step": 7277
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.966456950761215,
+      "learning_rate": 6.746496011042282e-06,
+      "loss": 0.1812,
+      "step": 7278
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.5615816016037554,
+      "learning_rate": 6.745651884623724e-06,
+      "loss": 0.2132,
+      "step": 7279
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.9004549533098821,
+      "learning_rate": 6.744807701540302e-06,
+      "loss": 0.2068,
+      "step": 7280
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.5353235342885923,
+      "learning_rate": 6.74396346181942e-06,
+      "loss": 0.2212,
+      "step": 7281
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.6164853396087195,
+      "learning_rate": 6.743119165488483e-06,
+      "loss": 0.2023,
+      "step": 7282
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.3865587115284386,
+      "learning_rate": 6.7422748125748935e-06,
+      "loss": 0.2301,
+      "step": 7283
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.9327956982478045,
+      "learning_rate": 6.741430403106061e-06,
+      "loss": 0.1889,
+      "step": 7284
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.369269467004051,
+      "learning_rate": 6.740585937109397e-06,
+      "loss": 0.176,
+      "step": 7285
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.1606343806123418,
+      "learning_rate": 6.739741414612312e-06,
+      "loss": 0.2028,
+      "step": 7286
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.5434541313050014,
+      "learning_rate": 6.738896835642222e-06,
+      "loss": 0.1883,
+      "step": 7287
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.118651382998683,
+      "learning_rate": 6.738052200226538e-06,
+      "loss": 0.1545,
+      "step": 7288
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.236320393132866,
+      "learning_rate": 6.737207508392681e-06,
+      "loss": 0.2382,
+      "step": 7289
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.2254062262464003,
+      "learning_rate": 6.736362760168072e-06,
+      "loss": 0.209,
+      "step": 7290
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.1087042869693926,
+      "learning_rate": 6.735517955580126e-06,
+      "loss": 0.1839,
+      "step": 7291
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.168141224734691,
+      "learning_rate": 6.734673094656271e-06,
+      "loss": 0.1097,
+      "step": 7292
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.1930144196957464,
+      "learning_rate": 6.733828177423929e-06,
+      "loss": 0.2102,
+      "step": 7293
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.9086413607993287,
+      "learning_rate": 6.732983203910528e-06,
+      "loss": 0.196,
+      "step": 7294
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.4673153282275087,
+      "learning_rate": 6.732138174143494e-06,
+      "loss": 0.2224,
+      "step": 7295
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.4984979547212833,
+      "learning_rate": 6.731293088150262e-06,
+      "loss": 0.1827,
+      "step": 7296
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.2232080241339887,
+      "learning_rate": 6.730447945958258e-06,
+      "loss": 0.1726,
+      "step": 7297
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.234738973784294,
+      "learning_rate": 6.7296027475949185e-06,
+      "loss": 0.2027,
+      "step": 7298
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.998093441587907,
+      "learning_rate": 6.728757493087681e-06,
+      "loss": 0.2086,
+      "step": 7299
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.8223793309077831,
+      "learning_rate": 6.72791218246398e-06,
+      "loss": 0.2261,
+      "step": 7300
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.3023130430412744,
+      "learning_rate": 6.727066815751257e-06,
+      "loss": 0.2654,
+      "step": 7301
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.9396605465881231,
+      "learning_rate": 6.7262213929769515e-06,
+      "loss": 0.1406,
+      "step": 7302
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.5656818667475647,
+      "learning_rate": 6.725375914168507e-06,
+      "loss": 0.26,
+      "step": 7303
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.9707217454563495,
+      "learning_rate": 6.724530379353367e-06,
+      "loss": 0.2451,
+      "step": 7304
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 2.011117545354531,
+      "learning_rate": 6.723684788558981e-06,
+      "loss": 0.1902,
+      "step": 7305
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.0390877577214328,
+      "learning_rate": 6.722839141812794e-06,
+      "loss": 0.2093,
+      "step": 7306
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.149164077033988,
+      "learning_rate": 6.721993439142258e-06,
+      "loss": 0.1979,
+      "step": 7307
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.1242788241773474,
+      "learning_rate": 6.7211476805748254e-06,
+      "loss": 0.1916,
+      "step": 7308
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.8813491736739772,
+      "learning_rate": 6.720301866137951e-06,
+      "loss": 0.1797,
+      "step": 7309
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.2210627155911984,
+      "learning_rate": 6.719455995859087e-06,
+      "loss": 0.1882,
+      "step": 7310
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.8784341201018316,
+      "learning_rate": 6.718610069765691e-06,
+      "loss": 0.1728,
+      "step": 7311
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.2544640551136816,
+      "learning_rate": 6.717764087885226e-06,
+      "loss": 0.2169,
+      "step": 7312
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.588024645190128,
+      "learning_rate": 6.716918050245149e-06,
+      "loss": 0.1966,
+      "step": 7313
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.088932232132817,
+      "learning_rate": 6.716071956872927e-06,
+      "loss": 0.278,
+      "step": 7314
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.3159164796725253,
+      "learning_rate": 6.715225807796022e-06,
+      "loss": 0.2328,
+      "step": 7315
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.1947842222190792,
+      "learning_rate": 6.7143796030419e-06,
+      "loss": 0.2021,
+      "step": 7316
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.2409520047972469,
+      "learning_rate": 6.713533342638031e-06,
+      "loss": 0.1898,
+      "step": 7317
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.0335609072558611,
+      "learning_rate": 6.712687026611884e-06,
+      "loss": 0.1798,
+      "step": 7318
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.046778387088222,
+      "learning_rate": 6.711840654990932e-06,
+      "loss": 0.1427,
+      "step": 7319
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.0906449316110838,
+      "learning_rate": 6.710994227802648e-06,
+      "loss": 0.2026,
+      "step": 7320
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.1765484002270294,
+      "learning_rate": 6.710147745074507e-06,
+      "loss": 0.1931,
+      "step": 7321
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.2468818202465601,
+      "learning_rate": 6.709301206833986e-06,
+      "loss": 0.1522,
+      "step": 7322
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.8868284577717673,
+      "learning_rate": 6.708454613108566e-06,
+      "loss": 0.1671,
+      "step": 7323
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.2807883268232714,
+      "learning_rate": 6.707607963925725e-06,
+      "loss": 0.225,
+      "step": 7324
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.4354615244502422,
+      "learning_rate": 6.706761259312948e-06,
+      "loss": 0.1764,
+      "step": 7325
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.9968042980880115,
+      "learning_rate": 6.705914499297721e-06,
+      "loss": 0.1882,
+      "step": 7326
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.354303876332246,
+      "learning_rate": 6.705067683907528e-06,
+      "loss": 0.2291,
+      "step": 7327
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.2389165740654025,
+      "learning_rate": 6.704220813169857e-06,
+      "loss": 0.2177,
+      "step": 7328
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.2022475938475454,
+      "learning_rate": 6.703373887112198e-06,
+      "loss": 0.1869,
+      "step": 7329
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.8646539911540294,
+      "learning_rate": 6.702526905762044e-06,
+      "loss": 0.1907,
+      "step": 7330
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.9666855497518133,
+      "learning_rate": 6.701679869146887e-06,
+      "loss": 0.1906,
+      "step": 7331
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.0653548192503268,
+      "learning_rate": 6.700832777294223e-06,
+      "loss": 0.2191,
+      "step": 7332
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.9328828702833861,
+      "learning_rate": 6.699985630231548e-06,
+      "loss": 0.1852,
+      "step": 7333
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.4651800176574823,
+      "learning_rate": 6.699138427986363e-06,
+      "loss": 0.1996,
+      "step": 7334
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.6144568291194057,
+      "learning_rate": 6.698291170586166e-06,
+      "loss": 0.16,
+      "step": 7335
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.2932356821060695,
+      "learning_rate": 6.697443858058463e-06,
+      "loss": 0.2252,
+      "step": 7336
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.8448682434689774,
+      "learning_rate": 6.696596490430757e-06,
+      "loss": 0.1877,
+      "step": 7337
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.946059072081046,
+      "learning_rate": 6.695749067730552e-06,
+      "loss": 0.2056,
+      "step": 7338
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.0186617384455328,
+      "learning_rate": 6.694901589985357e-06,
+      "loss": 0.1816,
+      "step": 7339
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.0372587390966173,
+      "learning_rate": 6.694054057222681e-06,
+      "loss": 0.2253,
+      "step": 7340
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.5646885957803578,
+      "learning_rate": 6.693206469470037e-06,
+      "loss": 0.2251,
+      "step": 7341
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.3334648236212467,
+      "learning_rate": 6.692358826754937e-06,
+      "loss": 0.192,
+      "step": 7342
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.083334740154869,
+      "learning_rate": 6.691511129104896e-06,
+      "loss": 0.2579,
+      "step": 7343
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.3837275588986446,
+      "learning_rate": 6.690663376547433e-06,
+      "loss": 0.1719,
+      "step": 7344
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.0494257682234052,
+      "learning_rate": 6.689815569110063e-06,
+      "loss": 0.2215,
+      "step": 7345
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.5507441374605361,
+      "learning_rate": 6.688967706820308e-06,
+      "loss": 0.1412,
+      "step": 7346
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.8856260871291639,
+      "learning_rate": 6.68811978970569e-06,
+      "loss": 0.1696,
+      "step": 7347
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.1814419508444909,
+      "learning_rate": 6.687271817793734e-06,
+      "loss": 0.2358,
+      "step": 7348
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.9536411839947718,
+      "learning_rate": 6.686423791111963e-06,
+      "loss": 0.1655,
+      "step": 7349
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.6380889228492497,
+      "learning_rate": 6.685575709687907e-06,
+      "loss": 0.2596,
+      "step": 7350
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.9629514319078729,
+      "learning_rate": 6.684727573549094e-06,
+      "loss": 0.1809,
+      "step": 7351
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.2547470189997512,
+      "learning_rate": 6.683879382723053e-06,
+      "loss": 0.1866,
+      "step": 7352
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.186327365483106,
+      "learning_rate": 6.6830311372373225e-06,
+      "loss": 0.2042,
+      "step": 7353
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.1975571182641305,
+      "learning_rate": 6.682182837119432e-06,
+      "loss": 0.2223,
+      "step": 7354
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.0187594243649287,
+      "learning_rate": 6.68133448239692e-06,
+      "loss": 0.1877,
+      "step": 7355
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.8349357891199445,
+      "learning_rate": 6.680486073097322e-06,
+      "loss": 0.1152,
+      "step": 7356
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.196714022146813,
+      "learning_rate": 6.679637609248181e-06,
+      "loss": 0.214,
+      "step": 7357
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.203091829944978,
+      "learning_rate": 6.678789090877037e-06,
+      "loss": 0.1629,
+      "step": 7358
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.7845378713284687,
+      "learning_rate": 6.677940518011433e-06,
+      "loss": 0.1612,
+      "step": 7359
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.9070274660161328,
+      "learning_rate": 6.677091890678916e-06,
+      "loss": 0.1613,
+      "step": 7360
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.4576056220157931,
+      "learning_rate": 6.676243208907031e-06,
+      "loss": 0.2251,
+      "step": 7361
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.1783526475751385,
+      "learning_rate": 6.675394472723329e-06,
+      "loss": 0.2053,
+      "step": 7362
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 2.6642145242132336,
+      "learning_rate": 6.674545682155357e-06,
+      "loss": 0.2143,
+      "step": 7363
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.3020758851950929,
+      "learning_rate": 6.673696837230672e-06,
+      "loss": 0.2384,
+      "step": 7364
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.7582579568093035,
+      "learning_rate": 6.6728479379768226e-06,
+      "loss": 0.2001,
+      "step": 7365
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.1640314331645454,
+      "learning_rate": 6.671998984421369e-06,
+      "loss": 0.1761,
+      "step": 7366
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.100826274419403,
+      "learning_rate": 6.6711499765918655e-06,
+      "loss": 0.19,
+      "step": 7367
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.337915114714639,
+      "learning_rate": 6.670300914515873e-06,
+      "loss": 0.1946,
+      "step": 7368
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.9928108194785443,
+      "learning_rate": 6.669451798220953e-06,
+      "loss": 0.2276,
+      "step": 7369
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.9272832496434326,
+      "learning_rate": 6.668602627734666e-06,
+      "loss": 0.1659,
+      "step": 7370
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.2987780884017495,
+      "learning_rate": 6.66775340308458e-06,
+      "loss": 0.2193,
+      "step": 7371
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.1696697148044357,
+      "learning_rate": 6.666904124298259e-06,
+      "loss": 0.2161,
+      "step": 7372
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.952886513476514,
+      "learning_rate": 6.666054791403273e-06,
+      "loss": 0.1452,
+      "step": 7373
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.8125449269051339,
+      "learning_rate": 6.665205404427188e-06,
+      "loss": 0.1806,
+      "step": 7374
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.3571009414337898,
+      "learning_rate": 6.66435596339758e-06,
+      "loss": 0.2261,
+      "step": 7375
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.1881444117391589,
+      "learning_rate": 6.663506468342018e-06,
+      "loss": 0.2042,
+      "step": 7376
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.8678578779790317,
+      "learning_rate": 6.66265691928808e-06,
+      "loss": 0.1625,
+      "step": 7377
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.8874453453911639,
+      "learning_rate": 6.661807316263344e-06,
+      "loss": 0.151,
+      "step": 7378
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 2.100972154723861,
+      "learning_rate": 6.660957659295385e-06,
+      "loss": 0.2309,
+      "step": 7379
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.309209509351539,
+      "learning_rate": 6.660107948411787e-06,
+      "loss": 0.1795,
+      "step": 7380
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.2809268919079042,
+      "learning_rate": 6.65925818364013e-06,
+      "loss": 0.167,
+      "step": 7381
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.3658177496461787,
+      "learning_rate": 6.658408365007998e-06,
+      "loss": 0.1802,
+      "step": 7382
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.9696623828006397,
+      "learning_rate": 6.657558492542977e-06,
+      "loss": 0.2195,
+      "step": 7383
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.7836751585610089,
+      "learning_rate": 6.656708566272654e-06,
+      "loss": 0.1523,
+      "step": 7384
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.760238524803821,
+      "learning_rate": 6.6558585862246175e-06,
+      "loss": 0.138,
+      "step": 7385
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.2270085015927326,
+      "learning_rate": 6.6550085524264586e-06,
+      "loss": 0.1931,
+      "step": 7386
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.5046514658807197,
+      "learning_rate": 6.654158464905772e-06,
+      "loss": 0.2127,
+      "step": 7387
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.2193671155759718,
+      "learning_rate": 6.65330832369015e-06,
+      "loss": 0.1929,
+      "step": 7388
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.3708189120978627,
+      "learning_rate": 6.6524581288071885e-06,
+      "loss": 0.2027,
+      "step": 7389
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.1238840962998504,
+      "learning_rate": 6.651607880284488e-06,
+      "loss": 0.2108,
+      "step": 7390
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.2185132395804192,
+      "learning_rate": 6.650757578149644e-06,
+      "loss": 0.2439,
+      "step": 7391
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.1720884383061383,
+      "learning_rate": 6.6499072224302595e-06,
+      "loss": 0.2021,
+      "step": 7392
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.0829616120565113,
+      "learning_rate": 6.649056813153939e-06,
+      "loss": 0.2147,
+      "step": 7393
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.4455848643416729,
+      "learning_rate": 6.648206350348285e-06,
+      "loss": 0.2145,
+      "step": 7394
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.5359507050824575,
+      "learning_rate": 6.647355834040905e-06,
+      "loss": 0.2182,
+      "step": 7395
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.3187613093537554,
+      "learning_rate": 6.646505264259408e-06,
+      "loss": 0.1811,
+      "step": 7396
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.181010183710553,
+      "learning_rate": 6.645654641031403e-06,
+      "loss": 0.2055,
+      "step": 7397
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.1845296300834969,
+      "learning_rate": 6.644803964384502e-06,
+      "loss": 0.2287,
+      "step": 7398
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.099083132535256,
+      "learning_rate": 6.643953234346319e-06,
+      "loss": 0.1657,
+      "step": 7399
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.9208799402800356,
+      "learning_rate": 6.64310245094447e-06,
+      "loss": 0.1763,
+      "step": 7400
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.5380019731158607,
+      "learning_rate": 6.6422516142065676e-06,
+      "loss": 0.2686,
+      "step": 7401
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.8040011775298178,
+      "learning_rate": 6.641400724160235e-06,
+      "loss": 0.2138,
+      "step": 7402
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.2415395282239987,
+      "learning_rate": 6.640549780833091e-06,
+      "loss": 0.2288,
+      "step": 7403
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.0786254315708634,
+      "learning_rate": 6.639698784252757e-06,
+      "loss": 0.2188,
+      "step": 7404
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.2179134689127682,
+      "learning_rate": 6.638847734446859e-06,
+      "loss": 0.2328,
+      "step": 7405
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.5568792095686375,
+      "learning_rate": 6.6379966314430185e-06,
+      "loss": 0.1816,
+      "step": 7406
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.0402024494476254,
+      "learning_rate": 6.637145475268868e-06,
+      "loss": 0.1807,
+      "step": 7407
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.3345448711353556,
+      "learning_rate": 6.636294265952034e-06,
+      "loss": 0.2402,
+      "step": 7408
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.0189502163886561,
+      "learning_rate": 6.635443003520147e-06,
+      "loss": 0.2157,
+      "step": 7409
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.8469615440778637,
+      "learning_rate": 6.634591688000838e-06,
+      "loss": 0.1836,
+      "step": 7410
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.1413349137152295,
+      "learning_rate": 6.633740319421744e-06,
+      "loss": 0.2632,
+      "step": 7411
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.3645110990852876,
+      "learning_rate": 6.632888897810499e-06,
+      "loss": 0.1921,
+      "step": 7412
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.0822427109203874,
+      "learning_rate": 6.6320374231947415e-06,
+      "loss": 0.1561,
+      "step": 7413
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.8320211915854377,
+      "learning_rate": 6.631185895602113e-06,
+      "loss": 0.1635,
+      "step": 7414
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.1398311524090488,
+      "learning_rate": 6.630334315060249e-06,
+      "loss": 0.1756,
+      "step": 7415
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.8905338692470464,
+      "learning_rate": 6.629482681596797e-06,
+      "loss": 0.1427,
+      "step": 7416
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.2772907572728365,
+      "learning_rate": 6.628630995239401e-06,
+      "loss": 0.181,
+      "step": 7417
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.1998096821162647,
+      "learning_rate": 6.627779256015707e-06,
+      "loss": 0.2109,
+      "step": 7418
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.245427588872736,
+      "learning_rate": 6.6269274639533595e-06,
+      "loss": 0.2601,
+      "step": 7419
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.9663877989627389,
+      "learning_rate": 6.626075619080014e-06,
+      "loss": 0.1862,
+      "step": 7420
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.1022776385839639,
+      "learning_rate": 6.625223721423316e-06,
+      "loss": 0.1706,
+      "step": 7421
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 2.2867494946632814,
+      "learning_rate": 6.624371771010922e-06,
+      "loss": 0.1898,
+      "step": 7422
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.435538238707015,
+      "learning_rate": 6.623519767870488e-06,
+      "loss": 0.2102,
+      "step": 7423
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.9687313281843999,
+      "learning_rate": 6.622667712029666e-06,
+      "loss": 0.1818,
+      "step": 7424
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.4707047296543156,
+      "learning_rate": 6.621815603516121e-06,
+      "loss": 0.2147,
+      "step": 7425
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.9823669528208083,
+      "learning_rate": 6.620963442357506e-06,
+      "loss": 0.2423,
+      "step": 7426
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.4997973503870234,
+      "learning_rate": 6.620111228581487e-06,
+      "loss": 0.3096,
+      "step": 7427
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.1860186755125612,
+      "learning_rate": 6.619258962215724e-06,
+      "loss": 0.2428,
+      "step": 7428
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.9796912845739245,
+      "learning_rate": 6.618406643287886e-06,
+      "loss": 0.1833,
+      "step": 7429
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.9084619009863263,
+      "learning_rate": 6.617554271825636e-06,
+      "loss": 0.187,
+      "step": 7430
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.0592784286188206,
+      "learning_rate": 6.616701847856644e-06,
+      "loss": 0.2386,
+      "step": 7431
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.06946815126257,
+      "learning_rate": 6.615849371408582e-06,
+      "loss": 0.1744,
+      "step": 7432
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.150346239428627,
+      "learning_rate": 6.6149968425091205e-06,
+      "loss": 0.2076,
+      "step": 7433
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.1871987630523724,
+      "learning_rate": 6.614144261185931e-06,
+      "loss": 0.2258,
+      "step": 7434
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.8321838552371524,
+      "learning_rate": 6.6132916274666915e-06,
+      "loss": 0.2158,
+      "step": 7435
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.3200607285473915,
+      "learning_rate": 6.612438941379078e-06,
+      "loss": 0.237,
+      "step": 7436
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.9199187153156548,
+      "learning_rate": 6.611586202950768e-06,
+      "loss": 0.1693,
+      "step": 7437
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.5007110787805995,
+      "learning_rate": 6.6107334122094456e-06,
+      "loss": 0.1536,
+      "step": 7438
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.429662714024174,
+      "learning_rate": 6.609880569182788e-06,
+      "loss": 0.2294,
+      "step": 7439
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.6502567187779431,
+      "learning_rate": 6.609027673898483e-06,
+      "loss": 0.1766,
+      "step": 7440
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.1922363828684122,
+      "learning_rate": 6.608174726384213e-06,
+      "loss": 0.2167,
+      "step": 7441
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.2458015266662028,
+      "learning_rate": 6.607321726667668e-06,
+      "loss": 0.1919,
+      "step": 7442
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.0048618889086307,
+      "learning_rate": 6.606468674776536e-06,
+      "loss": 0.1946,
+      "step": 7443
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.073485293983237,
+      "learning_rate": 6.605615570738506e-06,
+      "loss": 0.1635,
+      "step": 7444
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.1273209336893752,
+      "learning_rate": 6.604762414581271e-06,
+      "loss": 0.2292,
+      "step": 7445
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.1496651795686554,
+      "learning_rate": 6.603909206332525e-06,
+      "loss": 0.2299,
+      "step": 7446
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.0306142776549159,
+      "learning_rate": 6.603055946019965e-06,
+      "loss": 0.1926,
+      "step": 7447
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.279085500013146,
+      "learning_rate": 6.602202633671286e-06,
+      "loss": 0.1862,
+      "step": 7448
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.7365671814517671,
+      "learning_rate": 6.601349269314188e-06,
+      "loss": 0.133,
+      "step": 7449
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.9291004162457148,
+      "learning_rate": 6.600495852976373e-06,
+      "loss": 0.2419,
+      "step": 7450
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.9823882303866274,
+      "learning_rate": 6.599642384685543e-06,
+      "loss": 0.1542,
+      "step": 7451
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.1513181519035969,
+      "learning_rate": 6.598788864469401e-06,
+      "loss": 0.2095,
+      "step": 7452
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.1680395398328396,
+      "learning_rate": 6.59793529235565e-06,
+      "loss": 0.1818,
+      "step": 7453
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.0345161906340226,
+      "learning_rate": 6.597081668372004e-06,
+      "loss": 0.1785,
+      "step": 7454
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.388972886909821,
+      "learning_rate": 6.596227992546167e-06,
+      "loss": 0.2219,
+      "step": 7455
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.1692119316637373,
+      "learning_rate": 6.595374264905853e-06,
+      "loss": 0.2234,
+      "step": 7456
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.2608615750755054,
+      "learning_rate": 6.59452048547877e-06,
+      "loss": 0.1603,
+      "step": 7457
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.3178682811321194,
+      "learning_rate": 6.593666654292638e-06,
+      "loss": 0.2032,
+      "step": 7458
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.1282755713616424,
+      "learning_rate": 6.592812771375169e-06,
+      "loss": 0.2004,
+      "step": 7459
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.9986727929799306,
+      "learning_rate": 6.591958836754082e-06,
+      "loss": 0.1884,
+      "step": 7460
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.0806104993205807,
+      "learning_rate": 6.591104850457095e-06,
+      "loss": 0.2734,
+      "step": 7461
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.1261365368260907,
+      "learning_rate": 6.590250812511929e-06,
+      "loss": 0.2101,
+      "step": 7462
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.6865773371845545,
+      "learning_rate": 6.589396722946308e-06,
+      "loss": 0.2096,
+      "step": 7463
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.1109335119977963,
+      "learning_rate": 6.588542581787956e-06,
+      "loss": 0.2,
+      "step": 7464
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.3767268691615266,
+      "learning_rate": 6.587688389064598e-06,
+      "loss": 0.201,
+      "step": 7465
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.426319668628752,
+      "learning_rate": 6.586834144803961e-06,
+      "loss": 0.244,
+      "step": 7466
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.0983035904016032,
+      "learning_rate": 6.585979849033775e-06,
+      "loss": 0.2183,
+      "step": 7467
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.1284083089906107,
+      "learning_rate": 6.585125501781772e-06,
+      "loss": 0.187,
+      "step": 7468
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.2184366410226233,
+      "learning_rate": 6.584271103075685e-06,
+      "loss": 0.1869,
+      "step": 7469
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.3226826093405109,
+      "learning_rate": 6.583416652943245e-06,
+      "loss": 0.2136,
+      "step": 7470
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.9116364096206467,
+      "learning_rate": 6.582562151412191e-06,
+      "loss": 0.187,
+      "step": 7471
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.8815017713388582,
+      "learning_rate": 6.581707598510258e-06,
+      "loss": 0.2005,
+      "step": 7472
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.2298968575572067,
+      "learning_rate": 6.580852994265188e-06,
+      "loss": 0.2068,
+      "step": 7473
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.3524842468060874,
+      "learning_rate": 6.579998338704721e-06,
+      "loss": 0.163,
+      "step": 7474
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.3544754998156163,
+      "learning_rate": 6.579143631856598e-06,
+      "loss": 0.2108,
+      "step": 7475
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.8272731184865447,
+      "learning_rate": 6.578288873748565e-06,
+      "loss": 0.1554,
+      "step": 7476
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.1450688040406611,
+      "learning_rate": 6.577434064408369e-06,
+      "loss": 0.1556,
+      "step": 7477
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.170303086789196,
+      "learning_rate": 6.576579203863756e-06,
+      "loss": 0.1487,
+      "step": 7478
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.3616099086494406,
+      "learning_rate": 6.575724292142474e-06,
+      "loss": 0.2029,
+      "step": 7479
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.3610304468684762,
+      "learning_rate": 6.574869329272277e-06,
+      "loss": 0.2235,
+      "step": 7480
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.024264100101551,
+      "learning_rate": 6.574014315280915e-06,
+      "loss": 0.186,
+      "step": 7481
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.549370857000871,
+      "learning_rate": 6.5731592501961426e-06,
+      "loss": 0.2354,
+      "step": 7482
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.7435305084674915,
+      "learning_rate": 6.5723041340457175e-06,
+      "loss": 0.2115,
+      "step": 7483
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.1799314764501425,
+      "learning_rate": 6.571448966857394e-06,
+      "loss": 0.2223,
+      "step": 7484
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.9625954049683847,
+      "learning_rate": 6.570593748658936e-06,
+      "loss": 0.1961,
+      "step": 7485
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.8454872543010429,
+      "learning_rate": 6.5697384794781004e-06,
+      "loss": 0.1378,
+      "step": 7486
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.0527509571248643,
+      "learning_rate": 6.568883159342651e-06,
+      "loss": 0.1927,
+      "step": 7487
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.1909498403616032,
+      "learning_rate": 6.5680277882803536e-06,
+      "loss": 0.1963,
+      "step": 7488
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.0643902681322979,
+      "learning_rate": 6.567172366318971e-06,
+      "loss": 0.1913,
+      "step": 7489
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.4066068249566168,
+      "learning_rate": 6.5663168934862735e-06,
+      "loss": 0.2475,
+      "step": 7490
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.2596302107547706,
+      "learning_rate": 6.565461369810027e-06,
+      "loss": 0.2067,
+      "step": 7491
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.989674422956714,
+      "learning_rate": 6.564605795318004e-06,
+      "loss": 0.2411,
+      "step": 7492
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.2169500441303132,
+      "learning_rate": 6.563750170037979e-06,
+      "loss": 0.1606,
+      "step": 7493
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.1908402208363094,
+      "learning_rate": 6.562894493997723e-06,
+      "loss": 0.2366,
+      "step": 7494
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.2368207103896822,
+      "learning_rate": 6.562038767225015e-06,
+      "loss": 0.227,
+      "step": 7495
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.215736128868546,
+      "learning_rate": 6.561182989747629e-06,
+      "loss": 0.2098,
+      "step": 7496
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.0272897118690556,
+      "learning_rate": 6.560327161593346e-06,
+      "loss": 0.187,
+      "step": 7497
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.9630546797609961,
+      "learning_rate": 6.559471282789946e-06,
+      "loss": 0.2016,
+      "step": 7498
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.2688881404257555,
+      "learning_rate": 6.55861535336521e-06,
+      "loss": 0.1944,
+      "step": 7499
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.9104090051097662,
+      "learning_rate": 6.557759373346924e-06,
+      "loss": 0.2216,
+      "step": 7500
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.3078359430871325,
+      "learning_rate": 6.556903342762874e-06,
+      "loss": 0.1676,
+      "step": 7501
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.8925735945137957,
+      "learning_rate": 6.5560472616408455e-06,
+      "loss": 0.2218,
+      "step": 7502
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.3778588718138998,
+      "learning_rate": 6.555191130008627e-06,
+      "loss": 0.2083,
+      "step": 7503
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.1398277680974138,
+      "learning_rate": 6.554334947894012e-06,
+      "loss": 0.1838,
+      "step": 7504
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.1865065475329224,
+      "learning_rate": 6.5534787153247905e-06,
+      "loss": 0.1894,
+      "step": 7505
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.3475403887936288,
+      "learning_rate": 6.552622432328757e-06,
+      "loss": 0.1986,
+      "step": 7506
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.0485739497636084,
+      "learning_rate": 6.551766098933705e-06,
+      "loss": 0.1981,
+      "step": 7507
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.332803172518256,
+      "learning_rate": 6.550909715167434e-06,
+      "loss": 0.1806,
+      "step": 7508
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.067963394271604,
+      "learning_rate": 6.55005328105774e-06,
+      "loss": 0.197,
+      "step": 7509
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.77965616143343,
+      "learning_rate": 6.549196796632427e-06,
+      "loss": 0.2306,
+      "step": 7510
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.0568618719062743,
+      "learning_rate": 6.548340261919295e-06,
+      "loss": 0.2035,
+      "step": 7511
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.9130842157847446,
+      "learning_rate": 6.547483676946147e-06,
+      "loss": 0.1218,
+      "step": 7512
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.258577793702149,
+      "learning_rate": 6.54662704174079e-06,
+      "loss": 0.2266,
+      "step": 7513
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.0628143469635445,
+      "learning_rate": 6.5457703563310295e-06,
+      "loss": 0.2057,
+      "step": 7514
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.3857793500566593,
+      "learning_rate": 6.5449136207446754e-06,
+      "loss": 0.2151,
+      "step": 7515
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.3214772041841147,
+      "learning_rate": 6.544056835009535e-06,
+      "loss": 0.2176,
+      "step": 7516
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.3618467183259866,
+      "learning_rate": 6.543199999153422e-06,
+      "loss": 0.2167,
+      "step": 7517
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.47560019647158,
+      "learning_rate": 6.54234311320415e-06,
+      "loss": 0.2074,
+      "step": 7518
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.264316995556445,
+      "learning_rate": 6.541486177189533e-06,
+      "loss": 0.1747,
+      "step": 7519
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.20826978009358,
+      "learning_rate": 6.540629191137389e-06,
+      "loss": 0.149,
+      "step": 7520
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.1706205042952829,
+      "learning_rate": 6.539772155075535e-06,
+      "loss": 0.2189,
+      "step": 7521
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.2656774171523961,
+      "learning_rate": 6.538915069031792e-06,
+      "loss": 0.2038,
+      "step": 7522
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.2411183893739601,
+      "learning_rate": 6.538057933033982e-06,
+      "loss": 0.2008,
+      "step": 7523
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.1347032520006628,
+      "learning_rate": 6.537200747109926e-06,
+      "loss": 0.2052,
+      "step": 7524
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.0413673479465066,
+      "learning_rate": 6.5363435112874495e-06,
+      "loss": 0.2079,
+      "step": 7525
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.1708381070970755,
+      "learning_rate": 6.53548622559438e-06,
+      "loss": 0.1894,
+      "step": 7526
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.1424782440349095,
+      "learning_rate": 6.534628890058543e-06,
+      "loss": 0.2205,
+      "step": 7527
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.3843040611743223,
+      "learning_rate": 6.53377150470777e-06,
+      "loss": 0.2161,
+      "step": 7528
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.9986185855101993,
+      "learning_rate": 6.532914069569894e-06,
+      "loss": 0.1358,
+      "step": 7529
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.303486139583023,
+      "learning_rate": 6.532056584672744e-06,
+      "loss": 0.2231,
+      "step": 7530
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.0166175888659816,
+      "learning_rate": 6.531199050044157e-06,
+      "loss": 0.2246,
+      "step": 7531
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.8733649965025657,
+      "learning_rate": 6.5303414657119675e-06,
+      "loss": 0.1747,
+      "step": 7532
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.0881832129991214,
+      "learning_rate": 6.529483831704015e-06,
+      "loss": 0.1822,
+      "step": 7533
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.727116068823906,
+      "learning_rate": 6.528626148048139e-06,
+      "loss": 0.1996,
+      "step": 7534
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.0996831798007893,
+      "learning_rate": 6.527768414772177e-06,
+      "loss": 0.2,
+      "step": 7535
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.0090819327907132,
+      "learning_rate": 6.526910631903973e-06,
+      "loss": 0.1847,
+      "step": 7536
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.2187087284338207,
+      "learning_rate": 6.526052799471373e-06,
+      "loss": 0.237,
+      "step": 7537
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.2552353523308997,
+      "learning_rate": 6.525194917502222e-06,
+      "loss": 0.1568,
+      "step": 7538
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.3494611912706407,
+      "learning_rate": 6.524336986024366e-06,
+      "loss": 0.2672,
+      "step": 7539
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.0258501222213676,
+      "learning_rate": 6.523479005065657e-06,
+      "loss": 0.1866,
+      "step": 7540
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.4838588582693664,
+      "learning_rate": 6.5226209746539426e-06,
+      "loss": 0.214,
+      "step": 7541
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.2778579132059402,
+      "learning_rate": 6.521762894817075e-06,
+      "loss": 0.2126,
+      "step": 7542
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.0015296691863724,
+      "learning_rate": 6.520904765582909e-06,
+      "loss": 0.1832,
+      "step": 7543
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.266684059627019,
+      "learning_rate": 6.520046586979301e-06,
+      "loss": 0.2073,
+      "step": 7544
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.1858764642783013,
+      "learning_rate": 6.519188359034104e-06,
+      "loss": 0.1831,
+      "step": 7545
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.150600338134889,
+      "learning_rate": 6.51833008177518e-06,
+      "loss": 0.1364,
+      "step": 7546
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.027491714787551,
+      "learning_rate": 6.51747175523039e-06,
+      "loss": 0.1434,
+      "step": 7547
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.398252698739181,
+      "learning_rate": 6.516613379427592e-06,
+      "loss": 0.2291,
+      "step": 7548
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.2146849175787133,
+      "learning_rate": 6.515754954394654e-06,
+      "loss": 0.2321,
+      "step": 7549
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.2768113779657517,
+      "learning_rate": 6.514896480159439e-06,
+      "loss": 0.219,
+      "step": 7550
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.0785134244659422,
+      "learning_rate": 6.514037956749811e-06,
+      "loss": 0.1949,
+      "step": 7551
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.2812862202130273,
+      "learning_rate": 6.5131793841936405e-06,
+      "loss": 0.223,
+      "step": 7552
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.078942196500622,
+      "learning_rate": 6.5123207625187975e-06,
+      "loss": 0.2346,
+      "step": 7553
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.9914781989976144,
+      "learning_rate": 6.511462091753152e-06,
+      "loss": 0.1839,
+      "step": 7554
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.0524197747084478,
+      "learning_rate": 6.510603371924578e-06,
+      "loss": 0.184,
+      "step": 7555
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.216493939305835,
+      "learning_rate": 6.509744603060952e-06,
+      "loss": 0.2049,
+      "step": 7556
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.3471088956440342,
+      "learning_rate": 6.508885785190146e-06,
+      "loss": 0.2399,
+      "step": 7557
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.0475034215400096,
+      "learning_rate": 6.50802691834004e-06,
+      "loss": 0.1617,
+      "step": 7558
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.3327308478840665,
+      "learning_rate": 6.5071680025385155e-06,
+      "loss": 0.1661,
+      "step": 7559
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.1103099204569473,
+      "learning_rate": 6.5063090378134484e-06,
+      "loss": 0.2154,
+      "step": 7560
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.9155277994790533,
+      "learning_rate": 6.5054500241927235e-06,
+      "loss": 0.1678,
+      "step": 7561
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.5817638218280174,
+      "learning_rate": 6.504590961704227e-06,
+      "loss": 0.208,
+      "step": 7562
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.991616813357552,
+      "learning_rate": 6.503731850375841e-06,
+      "loss": 0.176,
+      "step": 7563
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.585169296060556,
+      "learning_rate": 6.502872690235453e-06,
+      "loss": 0.2144,
+      "step": 7564
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.2832474272699577,
+      "learning_rate": 6.502013481310957e-06,
+      "loss": 0.1936,
+      "step": 7565
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.9509489005674993,
+      "learning_rate": 6.501154223630238e-06,
+      "loss": 0.2023,
+      "step": 7566
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.1013977835650854,
+      "learning_rate": 6.5002949172211894e-06,
+      "loss": 0.2367,
+      "step": 7567
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.0905263670340237,
+      "learning_rate": 6.499435562111706e-06,
+      "loss": 0.151,
+      "step": 7568
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.0610709467406347,
+      "learning_rate": 6.498576158329681e-06,
+      "loss": 0.1549,
+      "step": 7569
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.181009795728478,
+      "learning_rate": 6.497716705903012e-06,
+      "loss": 0.163,
+      "step": 7570
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.3704381255543634,
+      "learning_rate": 6.496857204859598e-06,
+      "loss": 0.1878,
+      "step": 7571
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.629054777431496,
+      "learning_rate": 6.4959976552273375e-06,
+      "loss": 0.2382,
+      "step": 7572
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.0981893517560708,
+      "learning_rate": 6.495138057034134e-06,
+      "loss": 0.196,
+      "step": 7573
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.1247631790654928,
+      "learning_rate": 6.494278410307889e-06,
+      "loss": 0.2247,
+      "step": 7574
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.1184649966212843,
+      "learning_rate": 6.493418715076507e-06,
+      "loss": 0.2103,
+      "step": 7575
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.357006163027563,
+      "learning_rate": 6.492558971367895e-06,
+      "loss": 0.2064,
+      "step": 7576
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.1810204636119561,
+      "learning_rate": 6.491699179209961e-06,
+      "loss": 0.226,
+      "step": 7577
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.538312194168128,
+      "learning_rate": 6.490839338630614e-06,
+      "loss": 0.2123,
+      "step": 7578
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.85626092576623,
+      "learning_rate": 6.489979449657765e-06,
+      "loss": 0.1754,
+      "step": 7579
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.397671988149688,
+      "learning_rate": 6.489119512319326e-06,
+      "loss": 0.2291,
+      "step": 7580
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.2529332063525909,
+      "learning_rate": 6.48825952664321e-06,
+      "loss": 0.2182,
+      "step": 7581
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.0685060738612697,
+      "learning_rate": 6.487399492657335e-06,
+      "loss": 0.1276,
+      "step": 7582
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.4942221326349925,
+      "learning_rate": 6.4865394103896175e-06,
+      "loss": 0.2022,
+      "step": 7583
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.8587781605386284,
+      "learning_rate": 6.4856792798679745e-06,
+      "loss": 0.235,
+      "step": 7584
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.1646590704322974,
+      "learning_rate": 6.48481910112033e-06,
+      "loss": 0.166,
+      "step": 7585
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.1229117021896213,
+      "learning_rate": 6.483958874174603e-06,
+      "loss": 0.1976,
+      "step": 7586
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.376106380622552,
+      "learning_rate": 6.483098599058717e-06,
+      "loss": 0.2176,
+      "step": 7587
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.3291137689146255,
+      "learning_rate": 6.4822382758005965e-06,
+      "loss": 0.2074,
+      "step": 7588
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.2917053411703063,
+      "learning_rate": 6.481377904428171e-06,
+      "loss": 0.23,
+      "step": 7589
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.146961067566642,
+      "learning_rate": 6.480517484969366e-06,
+      "loss": 0.1949,
+      "step": 7590
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.5490982786959926,
+      "learning_rate": 6.4796570174521125e-06,
+      "loss": 0.2028,
+      "step": 7591
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.28296671101923,
+      "learning_rate": 6.478796501904341e-06,
+      "loss": 0.2324,
+      "step": 7592
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.459601673844424,
+      "learning_rate": 6.477935938353984e-06,
+      "loss": 0.2211,
+      "step": 7593
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.191695775357881,
+      "learning_rate": 6.4770753268289774e-06,
+      "loss": 0.2374,
+      "step": 7594
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.0785034619398157,
+      "learning_rate": 6.476214667357257e-06,
+      "loss": 0.1822,
+      "step": 7595
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.8348995622782336,
+      "learning_rate": 6.47535395996676e-06,
+      "loss": 0.2192,
+      "step": 7596
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.0623043560371928,
+      "learning_rate": 6.474493204685423e-06,
+      "loss": 0.1908,
+      "step": 7597
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.7880106373564544,
+      "learning_rate": 6.47363240154119e-06,
+      "loss": 0.1848,
+      "step": 7598
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.3472118861363425,
+      "learning_rate": 6.472771550562002e-06,
+      "loss": 0.2824,
+      "step": 7599
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.074532743860371,
+      "learning_rate": 6.471910651775801e-06,
+      "loss": 0.1772,
+      "step": 7600
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.9917893511985212,
+      "learning_rate": 6.471049705210535e-06,
+      "loss": 0.215,
+      "step": 7601
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.6364887481083232,
+      "learning_rate": 6.470188710894149e-06,
+      "loss": 0.2273,
+      "step": 7602
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.2672835729306466,
+      "learning_rate": 6.469327668854594e-06,
+      "loss": 0.179,
+      "step": 7603
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.0853271810454168,
+      "learning_rate": 6.468466579119816e-06,
+      "loss": 0.198,
+      "step": 7604
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.0549177810263493,
+      "learning_rate": 6.46760544171777e-06,
+      "loss": 0.1867,
+      "step": 7605
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 3.3131990234994375,
+      "learning_rate": 6.466744256676406e-06,
+      "loss": 0.2399,
+      "step": 7606
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.6992398304563403,
+      "learning_rate": 6.465883024023679e-06,
+      "loss": 0.2175,
+      "step": 7607
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.855639458089016,
+      "learning_rate": 6.465021743787548e-06,
+      "loss": 0.1271,
+      "step": 7608
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.0759532934952554,
+      "learning_rate": 6.464160415995968e-06,
+      "loss": 0.221,
+      "step": 7609
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.167135427773929,
+      "learning_rate": 6.4632990406769e-06,
+      "loss": 0.225,
+      "step": 7610
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.7661286686402824,
+      "learning_rate": 6.462437617858302e-06,
+      "loss": 0.2366,
+      "step": 7611
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.5613549137370313,
+      "learning_rate": 6.461576147568139e-06,
+      "loss": 0.2313,
+      "step": 7612
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.1554824610987022,
+      "learning_rate": 6.460714629834375e-06,
+      "loss": 0.1896,
+      "step": 7613
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.0696418690372687,
+      "learning_rate": 6.459853064684973e-06,
+      "loss": 0.1774,
+      "step": 7614
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.283823089984368,
+      "learning_rate": 6.458991452147901e-06,
+      "loss": 0.2097,
+      "step": 7615
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.1523866871738628,
+      "learning_rate": 6.4581297922511255e-06,
+      "loss": 0.2262,
+      "step": 7616
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.5106724856768698,
+      "learning_rate": 6.457268085022621e-06,
+      "loss": 0.1882,
+      "step": 7617
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.0588866168300655,
+      "learning_rate": 6.456406330490355e-06,
+      "loss": 0.2002,
+      "step": 7618
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.8180978079328393,
+      "learning_rate": 6.455544528682303e-06,
+      "loss": 0.169,
+      "step": 7619
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.495252624297506,
+      "learning_rate": 6.454682679626438e-06,
+      "loss": 0.2343,
+      "step": 7620
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.9667126378053752,
+      "learning_rate": 6.453820783350737e-06,
+      "loss": 0.2282,
+      "step": 7621
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.234237110910731,
+      "learning_rate": 6.4529588398831786e-06,
+      "loss": 0.1853,
+      "step": 7622
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.4166299894671324,
+      "learning_rate": 6.45209684925174e-06,
+      "loss": 0.2678,
+      "step": 7623
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.030735557654243,
+      "learning_rate": 6.4512348114844015e-06,
+      "loss": 0.1771,
+      "step": 7624
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.143234133076,
+      "learning_rate": 6.450372726609147e-06,
+      "loss": 0.1848,
+      "step": 7625
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.9498111403016825,
+      "learning_rate": 6.44951059465396e-06,
+      "loss": 0.1328,
+      "step": 7626
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.3110966161888198,
+      "learning_rate": 6.448648415646825e-06,
+      "loss": 0.2223,
+      "step": 7627
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.2189642216628174,
+      "learning_rate": 6.44778618961573e-06,
+      "loss": 0.1875,
+      "step": 7628
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.1415818525167767,
+      "learning_rate": 6.446923916588663e-06,
+      "loss": 0.221,
+      "step": 7629
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.0677401964348934,
+      "learning_rate": 6.446061596593613e-06,
+      "loss": 0.2598,
+      "step": 7630
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.1144651158273062,
+      "learning_rate": 6.445199229658573e-06,
+      "loss": 0.2265,
+      "step": 7631
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.0783343353769215,
+      "learning_rate": 6.444336815811535e-06,
+      "loss": 0.21,
+      "step": 7632
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.3198255058392379,
+      "learning_rate": 6.443474355080492e-06,
+      "loss": 0.2173,
+      "step": 7633
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.9549712392367056,
+      "learning_rate": 6.442611847493442e-06,
+      "loss": 0.176,
+      "step": 7634
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.3599999350572318,
+      "learning_rate": 6.441749293078382e-06,
+      "loss": 0.2297,
+      "step": 7635
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.0367346983015653,
+      "learning_rate": 6.440886691863311e-06,
+      "loss": 0.2102,
+      "step": 7636
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.9689741058085333,
+      "learning_rate": 6.44002404387623e-06,
+      "loss": 0.1943,
+      "step": 7637
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.128356128848543,
+      "learning_rate": 6.439161349145139e-06,
+      "loss": 0.2135,
+      "step": 7638
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.0430916417332379,
+      "learning_rate": 6.438298607698044e-06,
+      "loss": 0.2522,
+      "step": 7639
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.8346965393482415,
+      "learning_rate": 6.437435819562949e-06,
+      "loss": 0.1833,
+      "step": 7640
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.3579729621900007,
+      "learning_rate": 6.436572984767861e-06,
+      "loss": 0.2047,
+      "step": 7641
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.1396196242402636,
+      "learning_rate": 6.435710103340787e-06,
+      "loss": 0.1936,
+      "step": 7642
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.0423899644788952,
+      "learning_rate": 6.434847175309737e-06,
+      "loss": 0.2427,
+      "step": 7643
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.4538308756956697,
+      "learning_rate": 6.433984200702724e-06,
+      "loss": 0.1884,
+      "step": 7644
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.0697345461015535,
+      "learning_rate": 6.4331211795477586e-06,
+      "loss": 0.1801,
+      "step": 7645
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.0965775708070375,
+      "learning_rate": 6.432258111872855e-06,
+      "loss": 0.2448,
+      "step": 7646
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.074213525585995,
+      "learning_rate": 6.431394997706031e-06,
+      "loss": 0.1998,
+      "step": 7647
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.3666156451590197,
+      "learning_rate": 6.430531837075301e-06,
+      "loss": 0.1871,
+      "step": 7648
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.6391829056421068,
+      "learning_rate": 6.429668630008687e-06,
+      "loss": 0.2375,
+      "step": 7649
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.4876888953610525,
+      "learning_rate": 6.428805376534205e-06,
+      "loss": 0.2198,
+      "step": 7650
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.2919956547466416,
+      "learning_rate": 6.427942076679879e-06,
+      "loss": 0.1844,
+      "step": 7651
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.0525299114339859,
+      "learning_rate": 6.427078730473732e-06,
+      "loss": 0.1617,
+      "step": 7652
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.4675642792183374,
+      "learning_rate": 6.42621533794379e-06,
+      "loss": 0.183,
+      "step": 7653
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.1452986289637663,
+      "learning_rate": 6.425351899118076e-06,
+      "loss": 0.2298,
+      "step": 7654
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.303507905502256,
+      "learning_rate": 6.424488414024622e-06,
+      "loss": 0.1605,
+      "step": 7655
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.9211970802762438,
+      "learning_rate": 6.4236248826914525e-06,
+      "loss": 0.1693,
+      "step": 7656
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.1118542242173037,
+      "learning_rate": 6.422761305146603e-06,
+      "loss": 0.2223,
+      "step": 7657
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.8198056744312227,
+      "learning_rate": 6.421897681418103e-06,
+      "loss": 0.1817,
+      "step": 7658
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.1219829213212942,
+      "learning_rate": 6.421034011533986e-06,
+      "loss": 0.1568,
+      "step": 7659
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.3448838063751345,
+      "learning_rate": 6.420170295522288e-06,
+      "loss": 0.2652,
+      "step": 7660
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.6040924442654032,
+      "learning_rate": 6.419306533411045e-06,
+      "loss": 0.2113,
+      "step": 7661
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.2410351608510983,
+      "learning_rate": 6.418442725228297e-06,
+      "loss": 0.1936,
+      "step": 7662
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.0191979320142182,
+      "learning_rate": 6.4175788710020815e-06,
+      "loss": 0.2093,
+      "step": 7663
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.4401167604011869,
+      "learning_rate": 6.4167149707604406e-06,
+      "loss": 0.1825,
+      "step": 7664
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.363671760591859,
+      "learning_rate": 6.415851024531417e-06,
+      "loss": 0.2328,
+      "step": 7665
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.0338812947817835,
+      "learning_rate": 6.414987032343056e-06,
+      "loss": 0.1808,
+      "step": 7666
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.11166685402779,
+      "learning_rate": 6.414122994223402e-06,
+      "loss": 0.2013,
+      "step": 7667
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.7119054037898995,
+      "learning_rate": 6.413258910200502e-06,
+      "loss": 0.2478,
+      "step": 7668
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.9221771419949146,
+      "learning_rate": 6.412394780302404e-06,
+      "loss": 0.2104,
+      "step": 7669
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.5313168579219538,
+      "learning_rate": 6.411530604557161e-06,
+      "loss": 0.1921,
+      "step": 7670
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.0369991236704155,
+      "learning_rate": 6.410666382992823e-06,
+      "loss": 0.2113,
+      "step": 7671
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.2474514969396404,
+      "learning_rate": 6.409802115637442e-06,
+      "loss": 0.1884,
+      "step": 7672
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.0654161987359916,
+      "learning_rate": 6.408937802519073e-06,
+      "loss": 0.1732,
+      "step": 7673
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.2983754738821227,
+      "learning_rate": 6.408073443665775e-06,
+      "loss": 0.2159,
+      "step": 7674
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.0472954361289182,
+      "learning_rate": 6.407209039105601e-06,
+      "loss": 0.1811,
+      "step": 7675
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.4145144588611998,
+      "learning_rate": 6.406344588866614e-06,
+      "loss": 0.1921,
+      "step": 7676
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.1620386406664882,
+      "learning_rate": 6.405480092976872e-06,
+      "loss": 0.1584,
+      "step": 7677
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.8899026982656881,
+      "learning_rate": 6.404615551464438e-06,
+      "loss": 0.2116,
+      "step": 7678
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.9196225993594483,
+      "learning_rate": 6.403750964357376e-06,
+      "loss": 0.1377,
+      "step": 7679
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.230765149581572,
+      "learning_rate": 6.402886331683751e-06,
+      "loss": 0.2093,
+      "step": 7680
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.6282568261343022,
+      "learning_rate": 6.402021653471626e-06,
+      "loss": 0.2005,
+      "step": 7681
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.258723641997277,
+      "learning_rate": 6.401156929749075e-06,
+      "loss": 0.2075,
+      "step": 7682
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.0801415968002717,
+      "learning_rate": 6.4002921605441635e-06,
+      "loss": 0.1808,
+      "step": 7683
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.3473238936001475,
+      "learning_rate": 6.399427345884964e-06,
+      "loss": 0.2076,
+      "step": 7684
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.2183167408843094,
+      "learning_rate": 6.398562485799547e-06,
+      "loss": 0.2178,
+      "step": 7685
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.299310686341848,
+      "learning_rate": 6.397697580315989e-06,
+      "loss": 0.206,
+      "step": 7686
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.8774860818620424,
+      "learning_rate": 6.396832629462361e-06,
+      "loss": 0.2061,
+      "step": 7687
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.0267820780997723,
+      "learning_rate": 6.395967633266744e-06,
+      "loss": 0.1293,
+      "step": 7688
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.3570132662844838,
+      "learning_rate": 6.395102591757216e-06,
+      "loss": 0.1859,
+      "step": 7689
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.9576382580438018,
+      "learning_rate": 6.3942375049618536e-06,
+      "loss": 0.1913,
+      "step": 7690
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.1587332514212931,
+      "learning_rate": 6.393372372908742e-06,
+      "loss": 0.2458,
+      "step": 7691
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.0390047150512691,
+      "learning_rate": 6.39250719562596e-06,
+      "loss": 0.1861,
+      "step": 7692
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.9663693571334261,
+      "learning_rate": 6.3916419731415956e-06,
+      "loss": 0.2613,
+      "step": 7693
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.298491980859145,
+      "learning_rate": 6.3907767054837324e-06,
+      "loss": 0.1633,
+      "step": 7694
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.6098850886285523,
+      "learning_rate": 6.3899113926804565e-06,
+      "loss": 0.1929,
+      "step": 7695
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.3521398623110683,
+      "learning_rate": 6.389046034759858e-06,
+      "loss": 0.2272,
+      "step": 7696
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.1053941081428476,
+      "learning_rate": 6.388180631750025e-06,
+      "loss": 0.2119,
+      "step": 7697
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.358498782819607,
+      "learning_rate": 6.387315183679052e-06,
+      "loss": 0.2308,
+      "step": 7698
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.824215079927635,
+      "learning_rate": 6.38644969057503e-06,
+      "loss": 0.2182,
+      "step": 7699
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.7379716275956203,
+      "learning_rate": 6.385584152466054e-06,
+      "loss": 0.1478,
+      "step": 7700
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.9185672797269286,
+      "learning_rate": 6.38471856938022e-06,
+      "loss": 0.199,
+      "step": 7701
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.2438182682681644,
+      "learning_rate": 6.383852941345625e-06,
+      "loss": 0.2227,
+      "step": 7702
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.1786359369646608,
+      "learning_rate": 6.382987268390368e-06,
+      "loss": 0.1821,
+      "step": 7703
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.573079168823242,
+      "learning_rate": 6.382121550542548e-06,
+      "loss": 0.1942,
+      "step": 7704
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.3790566440728511,
+      "learning_rate": 6.381255787830267e-06,
+      "loss": 0.1736,
+      "step": 7705
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.9387010788075417,
+      "learning_rate": 6.3803899802816296e-06,
+      "loss": 0.218,
+      "step": 7706
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.222054821085095,
+      "learning_rate": 6.379524127924741e-06,
+      "loss": 0.1673,
+      "step": 7707
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.213870508092756,
+      "learning_rate": 6.378658230787704e-06,
+      "loss": 0.2676,
+      "step": 7708
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.5554063635040432,
+      "learning_rate": 6.37779228889863e-06,
+      "loss": 0.206,
+      "step": 7709
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.6735627799774149,
+      "learning_rate": 6.376926302285624e-06,
+      "loss": 0.2264,
+      "step": 7710
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.1514341676236544,
+      "learning_rate": 6.376060270976801e-06,
+      "loss": 0.2219,
+      "step": 7711
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.0345425748068384,
+      "learning_rate": 6.3751941950002674e-06,
+      "loss": 0.1789,
+      "step": 7712
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.3353011584453092,
+      "learning_rate": 6.374328074384141e-06,
+      "loss": 0.2065,
+      "step": 7713
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.0673571039927234,
+      "learning_rate": 6.373461909156534e-06,
+      "loss": 0.1797,
+      "step": 7714
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.2313090316301887,
+      "learning_rate": 6.372595699345564e-06,
+      "loss": 0.18,
+      "step": 7715
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.9341612357036625,
+      "learning_rate": 6.371729444979349e-06,
+      "loss": 0.1549,
+      "step": 7716
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.2929206562605606,
+      "learning_rate": 6.370863146086006e-06,
+      "loss": 0.1864,
+      "step": 7717
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.3003448980192094,
+      "learning_rate": 6.369996802693659e-06,
+      "loss": 0.1658,
+      "step": 7718
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.3173110127942658,
+      "learning_rate": 6.369130414830428e-06,
+      "loss": 0.2541,
+      "step": 7719
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.8369864105303144,
+      "learning_rate": 6.368263982524435e-06,
+      "loss": 0.2195,
+      "step": 7720
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.3326470330485731,
+      "learning_rate": 6.367397505803806e-06,
+      "loss": 0.216,
+      "step": 7721
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.1794346986933142,
+      "learning_rate": 6.366530984696669e-06,
+      "loss": 0.2245,
+      "step": 7722
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.1943323255407905,
+      "learning_rate": 6.36566441923115e-06,
+      "loss": 0.225,
+      "step": 7723
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.3950054767502844,
+      "learning_rate": 6.364797809435377e-06,
+      "loss": 0.2478,
+      "step": 7724
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.7293674159047403,
+      "learning_rate": 6.363931155337485e-06,
+      "loss": 0.2017,
+      "step": 7725
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.568764792140813,
+      "learning_rate": 6.3630644569656005e-06,
+      "loss": 0.2214,
+      "step": 7726
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.059911812627467,
+      "learning_rate": 6.362197714347863e-06,
+      "loss": 0.1981,
+      "step": 7727
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.5911008846179564,
+      "learning_rate": 6.361330927512404e-06,
+      "loss": 0.1665,
+      "step": 7728
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.6969705765285503,
+      "learning_rate": 6.360464096487359e-06,
+      "loss": 0.1858,
+      "step": 7729
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.1077733667830245,
+      "learning_rate": 6.359597221300866e-06,
+      "loss": 0.205,
+      "step": 7730
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.0174143910244442,
+      "learning_rate": 6.358730301981066e-06,
+      "loss": 0.1904,
+      "step": 7731
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.055950432963788,
+      "learning_rate": 6.3578633385561e-06,
+      "loss": 0.2011,
+      "step": 7732
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.3412062276432817,
+      "learning_rate": 6.356996331054108e-06,
+      "loss": 0.231,
+      "step": 7733
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.442548624786026,
+      "learning_rate": 6.356129279503237e-06,
+      "loss": 0.1934,
+      "step": 7734
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.529974760401722,
+      "learning_rate": 6.3552621839316275e-06,
+      "loss": 0.2248,
+      "step": 7735
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.8631571436693328,
+      "learning_rate": 6.354395044367429e-06,
+      "loss": 0.1923,
+      "step": 7736
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.1435255385738818,
+      "learning_rate": 6.3535278608387906e-06,
+      "loss": 0.173,
+      "step": 7737
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.1921113054242507,
+      "learning_rate": 6.3526606333738575e-06,
+      "loss": 0.1969,
+      "step": 7738
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.7828416679948215,
+      "learning_rate": 6.351793362000782e-06,
+      "loss": 0.1997,
+      "step": 7739
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.3748674951994129,
+      "learning_rate": 6.350926046747716e-06,
+      "loss": 0.2771,
+      "step": 7740
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.665068707244175,
+      "learning_rate": 6.350058687642816e-06,
+      "loss": 0.2058,
+      "step": 7741
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.7655131345885957,
+      "learning_rate": 6.349191284714232e-06,
+      "loss": 0.1372,
+      "step": 7742
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.9819266320450803,
+      "learning_rate": 6.3483238379901254e-06,
+      "loss": 0.1841,
+      "step": 7743
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.0748578372647146,
+      "learning_rate": 6.347456347498651e-06,
+      "loss": 0.1966,
+      "step": 7744
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.4439936884285767,
+      "learning_rate": 6.3465888132679685e-06,
+      "loss": 0.2171,
+      "step": 7745
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.1467124315725283,
+      "learning_rate": 6.345721235326239e-06,
+      "loss": 0.2177,
+      "step": 7746
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.2199622400710532,
+      "learning_rate": 6.344853613701625e-06,
+      "loss": 0.2586,
+      "step": 7747
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.3481396982782676,
+      "learning_rate": 6.3439859484222874e-06,
+      "loss": 0.2395,
+      "step": 7748
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.9766362315000442,
+      "learning_rate": 6.343118239516394e-06,
+      "loss": 0.161,
+      "step": 7749
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.0551592231097393,
+      "learning_rate": 6.342250487012111e-06,
+      "loss": 0.2118,
+      "step": 7750
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.1852939970753216,
+      "learning_rate": 6.341382690937603e-06,
+      "loss": 0.1726,
+      "step": 7751
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.2495997242496755,
+      "learning_rate": 6.3405148513210444e-06,
+      "loss": 0.187,
+      "step": 7752
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.861984204174815,
+      "learning_rate": 6.339646968190601e-06,
+      "loss": 0.2281,
+      "step": 7753
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.5236059731470528,
+      "learning_rate": 6.3387790415744475e-06,
+      "loss": 0.2507,
+      "step": 7754
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.8025391554768578,
+      "learning_rate": 6.337911071500757e-06,
+      "loss": 0.2007,
+      "step": 7755
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.798711856877519,
+      "learning_rate": 6.3370430579977035e-06,
+      "loss": 0.1933,
+      "step": 7756
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.1395006141143096,
+      "learning_rate": 6.336175001093462e-06,
+      "loss": 0.1787,
+      "step": 7757
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.8671843398741114,
+      "learning_rate": 6.335306900816213e-06,
+      "loss": 0.2067,
+      "step": 7758
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.0328082094305373,
+      "learning_rate": 6.334438757194134e-06,
+      "loss": 0.1736,
+      "step": 7759
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.1224177115742775,
+      "learning_rate": 6.333570570255404e-06,
+      "loss": 0.1842,
+      "step": 7760
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.9488852939524741,
+      "learning_rate": 6.332702340028209e-06,
+      "loss": 0.195,
+      "step": 7761
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.8015339082225182,
+      "learning_rate": 6.331834066540727e-06,
+      "loss": 0.213,
+      "step": 7762
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.8507455956262785,
+      "learning_rate": 6.330965749821148e-06,
+      "loss": 0.151,
+      "step": 7763
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.103489060208765,
+      "learning_rate": 6.3300973898976545e-06,
+      "loss": 0.1394,
+      "step": 7764
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.1466645419277608,
+      "learning_rate": 6.329228986798435e-06,
+      "loss": 0.1806,
+      "step": 7765
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.3929299582890282,
+      "learning_rate": 6.328360540551677e-06,
+      "loss": 0.237,
+      "step": 7766
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.3321681050224985,
+      "learning_rate": 6.327492051185573e-06,
+      "loss": 0.197,
+      "step": 7767
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.2082766262332496,
+      "learning_rate": 6.326623518728313e-06,
+      "loss": 0.2108,
+      "step": 7768
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.0725226901895333,
+      "learning_rate": 6.325754943208091e-06,
+      "loss": 0.143,
+      "step": 7769
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.9332041017072955,
+      "learning_rate": 6.3248863246531015e-06,
+      "loss": 0.1908,
+      "step": 7770
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.3219644860038202,
+      "learning_rate": 6.32401766309154e-06,
+      "loss": 0.1801,
+      "step": 7771
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.1948738072264962,
+      "learning_rate": 6.3231489585516035e-06,
+      "loss": 0.2164,
+      "step": 7772
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.2215866843240237,
+      "learning_rate": 6.322280211061492e-06,
+      "loss": 0.178,
+      "step": 7773
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.14888546115407,
+      "learning_rate": 6.321411420649404e-06,
+      "loss": 0.2618,
+      "step": 7774
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.407350101911903,
+      "learning_rate": 6.32054258734354e-06,
+      "loss": 0.1856,
+      "step": 7775
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.4995713962715076,
+      "learning_rate": 6.319673711172105e-06,
+      "loss": 0.1605,
+      "step": 7776
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 2.197231360274283,
+      "learning_rate": 6.318804792163304e-06,
+      "loss": 0.227,
+      "step": 7777
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.1577563461400788,
+      "learning_rate": 6.3179358303453386e-06,
+      "loss": 0.213,
+      "step": 7778
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.0101771152639687,
+      "learning_rate": 6.317066825746419e-06,
+      "loss": 0.1719,
+      "step": 7779
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.7398126478631248,
+      "learning_rate": 6.316197778394754e-06,
+      "loss": 0.2054,
+      "step": 7780
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.0283719762592136,
+      "learning_rate": 6.3153286883185515e-06,
+      "loss": 0.1646,
+      "step": 7781
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.4353695909190907,
+      "learning_rate": 6.314459555546024e-06,
+      "loss": 0.1861,
+      "step": 7782
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.183037460331001,
+      "learning_rate": 6.313590380105383e-06,
+      "loss": 0.1759,
+      "step": 7783
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.2981078351974698,
+      "learning_rate": 6.312721162024843e-06,
+      "loss": 0.1938,
+      "step": 7784
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.3607285244966065,
+      "learning_rate": 6.311851901332618e-06,
+      "loss": 0.1657,
+      "step": 7785
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.6091285766900225,
+      "learning_rate": 6.310982598056928e-06,
+      "loss": 0.1859,
+      "step": 7786
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.0543400404052004,
+      "learning_rate": 6.310113252225988e-06,
+      "loss": 0.2067,
+      "step": 7787
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.9787904425799815,
+      "learning_rate": 6.309243863868019e-06,
+      "loss": 0.2291,
+      "step": 7788
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.693422837701633,
+      "learning_rate": 6.30837443301124e-06,
+      "loss": 0.2088,
+      "step": 7789
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.6366494102630098,
+      "learning_rate": 6.307504959683876e-06,
+      "loss": 0.2046,
+      "step": 7790
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.8846327937464014,
+      "learning_rate": 6.306635443914148e-06,
+      "loss": 0.2064,
+      "step": 7791
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.9223993717400727,
+      "learning_rate": 6.3057658857302835e-06,
+      "loss": 0.2083,
+      "step": 7792
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.168577917264368,
+      "learning_rate": 6.3048962851605055e-06,
+      "loss": 0.1517,
+      "step": 7793
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.1994399199177863,
+      "learning_rate": 6.304026642233044e-06,
+      "loss": 0.1842,
+      "step": 7794
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.224708484672826,
+      "learning_rate": 6.3031569569761294e-06,
+      "loss": 0.2177,
+      "step": 7795
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.9239539611151004,
+      "learning_rate": 6.3022872294179895e-06,
+      "loss": 0.1549,
+      "step": 7796
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.1557284432162955,
+      "learning_rate": 6.3014174595868586e-06,
+      "loss": 0.2339,
+      "step": 7797
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.2835063967871372,
+      "learning_rate": 6.3005476475109665e-06,
+      "loss": 0.1882,
+      "step": 7798
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.251810086510383,
+      "learning_rate": 6.299677793218551e-06,
+      "loss": 0.2326,
+      "step": 7799
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.335852280068125,
+      "learning_rate": 6.2988078967378474e-06,
+      "loss": 0.189,
+      "step": 7800
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.5488117296522443,
+      "learning_rate": 6.297937958097094e-06,
+      "loss": 0.2232,
+      "step": 7801
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.1629082121975123,
+      "learning_rate": 6.297067977324525e-06,
+      "loss": 0.1713,
+      "step": 7802
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.0638888044351387,
+      "learning_rate": 6.2961979544483845e-06,
+      "loss": 0.1659,
+      "step": 7803
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.1778577641873007,
+      "learning_rate": 6.295327889496913e-06,
+      "loss": 0.1783,
+      "step": 7804
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.0106506085026485,
+      "learning_rate": 6.2944577824983535e-06,
+      "loss": 0.213,
+      "step": 7805
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.2375802318135332,
+      "learning_rate": 6.29358763348095e-06,
+      "loss": 0.1854,
+      "step": 7806
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.131093356776258,
+      "learning_rate": 6.292717442472948e-06,
+      "loss": 0.207,
+      "step": 7807
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.4715332940590158,
+      "learning_rate": 6.291847209502595e-06,
+      "loss": 0.2295,
+      "step": 7808
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.0653479253653475,
+      "learning_rate": 6.290976934598138e-06,
+      "loss": 0.1229,
+      "step": 7809
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.4187161496195153,
+      "learning_rate": 6.290106617787829e-06,
+      "loss": 0.2001,
+      "step": 7810
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.2273480275041329,
+      "learning_rate": 6.289236259099915e-06,
+      "loss": 0.174,
+      "step": 7811
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.4403065360490608,
+      "learning_rate": 6.288365858562652e-06,
+      "loss": 0.1889,
+      "step": 7812
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.064029293375396,
+      "learning_rate": 6.287495416204292e-06,
+      "loss": 0.2258,
+      "step": 7813
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.2492378295433852,
+      "learning_rate": 6.286624932053091e-06,
+      "loss": 0.2037,
+      "step": 7814
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.3932682323029493,
+      "learning_rate": 6.285754406137304e-06,
+      "loss": 0.2227,
+      "step": 7815
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.7682198618801812,
+      "learning_rate": 6.2848838384851896e-06,
+      "loss": 0.164,
+      "step": 7816
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.4279790726006931,
+      "learning_rate": 6.284013229125008e-06,
+      "loss": 0.2284,
+      "step": 7817
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.1113167646528026,
+      "learning_rate": 6.2831425780850185e-06,
+      "loss": 0.1602,
+      "step": 7818
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.2685500094215174,
+      "learning_rate": 6.282271885393483e-06,
+      "loss": 0.1914,
+      "step": 7819
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.8099389645127848,
+      "learning_rate": 6.281401151078663e-06,
+      "loss": 0.1825,
+      "step": 7820
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.3141189365764145,
+      "learning_rate": 6.280530375168826e-06,
+      "loss": 0.1972,
+      "step": 7821
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.9587043015264717,
+      "learning_rate": 6.279659557692237e-06,
+      "loss": 0.2125,
+      "step": 7822
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.6287050877685747,
+      "learning_rate": 6.2787886986771616e-06,
+      "loss": 0.1909,
+      "step": 7823
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.1416633824170108,
+      "learning_rate": 6.277917798151871e-06,
+      "loss": 0.2049,
+      "step": 7824
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.2360819259587235,
+      "learning_rate": 6.277046856144633e-06,
+      "loss": 0.2297,
+      "step": 7825
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.0206728386573742,
+      "learning_rate": 6.276175872683721e-06,
+      "loss": 0.1886,
+      "step": 7826
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.5311656169606789,
+      "learning_rate": 6.275304847797406e-06,
+      "loss": 0.178,
+      "step": 7827
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.305625330764935,
+      "learning_rate": 6.274433781513962e-06,
+      "loss": 0.2005,
+      "step": 7828
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.3731181776001415,
+      "learning_rate": 6.273562673861664e-06,
+      "loss": 0.1899,
+      "step": 7829
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.60445525207542,
+      "learning_rate": 6.272691524868789e-06,
+      "loss": 0.2214,
+      "step": 7830
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.3959614635013577,
+      "learning_rate": 6.271820334563617e-06,
+      "loss": 0.2022,
+      "step": 7831
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.9918810739655017,
+      "learning_rate": 6.270949102974424e-06,
+      "loss": 0.1914,
+      "step": 7832
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.486576689890397,
+      "learning_rate": 6.270077830129493e-06,
+      "loss": 0.2665,
+      "step": 7833
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.2962250057301168,
+      "learning_rate": 6.269206516057105e-06,
+      "loss": 0.2506,
+      "step": 7834
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.2204956366630173,
+      "learning_rate": 6.268335160785544e-06,
+      "loss": 0.1858,
+      "step": 7835
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.8995531494419551,
+      "learning_rate": 6.267463764343093e-06,
+      "loss": 0.1511,
+      "step": 7836
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.6320205609148224,
+      "learning_rate": 6.266592326758042e-06,
+      "loss": 0.2102,
+      "step": 7837
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.2512010168992482,
+      "learning_rate": 6.265720848058674e-06,
+      "loss": 0.1867,
+      "step": 7838
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.4072518554892297,
+      "learning_rate": 6.264849328273279e-06,
+      "loss": 0.2322,
+      "step": 7839
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.5635227384783328,
+      "learning_rate": 6.263977767430148e-06,
+      "loss": 0.2689,
+      "step": 7840
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.1062401078746353,
+      "learning_rate": 6.263106165557571e-06,
+      "loss": 0.1909,
+      "step": 7841
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.2207688916609931,
+      "learning_rate": 6.262234522683843e-06,
+      "loss": 0.1683,
+      "step": 7842
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.0672106765475653,
+      "learning_rate": 6.261362838837255e-06,
+      "loss": 0.1804,
+      "step": 7843
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.2248915677226537,
+      "learning_rate": 6.260491114046106e-06,
+      "loss": 0.1929,
+      "step": 7844
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.9979267710628443,
+      "learning_rate": 6.25961934833869e-06,
+      "loss": 0.1476,
+      "step": 7845
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.292658636122523,
+      "learning_rate": 6.258747541743306e-06,
+      "loss": 0.2028,
+      "step": 7846
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 4.642367320075202,
+      "learning_rate": 6.257875694288251e-06,
+      "loss": 0.257,
+      "step": 7847
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.4823710366278207,
+      "learning_rate": 6.257003806001828e-06,
+      "loss": 0.1824,
+      "step": 7848
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.3974351540672094,
+      "learning_rate": 6.256131876912339e-06,
+      "loss": 0.2224,
+      "step": 7849
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.9440735982162372,
+      "learning_rate": 6.255259907048088e-06,
+      "loss": 0.1631,
+      "step": 7850
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.4567366294496014,
+      "learning_rate": 6.254387896437377e-06,
+      "loss": 0.213,
+      "step": 7851
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.1368950513735545,
+      "learning_rate": 6.253515845108514e-06,
+      "loss": 0.1882,
+      "step": 7852
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.2716028360478742,
+      "learning_rate": 6.252643753089807e-06,
+      "loss": 0.2188,
+      "step": 7853
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.189901627543707,
+      "learning_rate": 6.251771620409563e-06,
+      "loss": 0.1762,
+      "step": 7854
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.0265300305775098,
+      "learning_rate": 6.2508994470960924e-06,
+      "loss": 0.123,
+      "step": 7855
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.47224668985054,
+      "learning_rate": 6.250027233177707e-06,
+      "loss": 0.189,
+      "step": 7856
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.0188379387623838,
+      "learning_rate": 6.249154978682716e-06,
+      "loss": 0.1316,
+      "step": 7857
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.1589828748342939,
+      "learning_rate": 6.2482826836394395e-06,
+      "loss": 0.1706,
+      "step": 7858
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.2310189061727272,
+      "learning_rate": 6.247410348076186e-06,
+      "loss": 0.1881,
+      "step": 7859
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.0776654314399736,
+      "learning_rate": 6.246537972021278e-06,
+      "loss": 0.1983,
+      "step": 7860
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.186504106403379,
+      "learning_rate": 6.245665555503029e-06,
+      "loss": 0.1533,
+      "step": 7861
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.1658604021763803,
+      "learning_rate": 6.244793098549762e-06,
+      "loss": 0.2235,
+      "step": 7862
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.4046293085369457,
+      "learning_rate": 6.243920601189793e-06,
+      "loss": 0.2138,
+      "step": 7863
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.1303927985923223,
+      "learning_rate": 6.243048063451447e-06,
+      "loss": 0.175,
+      "step": 7864
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.6633686356279762,
+      "learning_rate": 6.242175485363046e-06,
+      "loss": 0.2271,
+      "step": 7865
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.9412983401171732,
+      "learning_rate": 6.241302866952913e-06,
+      "loss": 0.1642,
+      "step": 7866
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.1728638037710568,
+      "learning_rate": 6.240430208249377e-06,
+      "loss": 0.2331,
+      "step": 7867
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.2000905092821075,
+      "learning_rate": 6.239557509280761e-06,
+      "loss": 0.1918,
+      "step": 7868
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.2359792027147314,
+      "learning_rate": 6.238684770075398e-06,
+      "loss": 0.2564,
+      "step": 7869
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.1858662076235904,
+      "learning_rate": 6.237811990661613e-06,
+      "loss": 0.1709,
+      "step": 7870
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.1089351379761963,
+      "learning_rate": 6.236939171067741e-06,
+      "loss": 0.1656,
+      "step": 7871
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.9058220629885112,
+      "learning_rate": 6.236066311322111e-06,
+      "loss": 0.2241,
+      "step": 7872
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.1183462369348978,
+      "learning_rate": 6.2351934114530576e-06,
+      "loss": 0.1672,
+      "step": 7873
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.4156078935478007,
+      "learning_rate": 6.234320471488916e-06,
+      "loss": 0.2234,
+      "step": 7874
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.3053322886240313,
+      "learning_rate": 6.233447491458022e-06,
+      "loss": 0.1834,
+      "step": 7875
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.0141695028580544,
+      "learning_rate": 6.232574471388714e-06,
+      "loss": 0.2279,
+      "step": 7876
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.3926562465247008,
+      "learning_rate": 6.231701411309329e-06,
+      "loss": 0.1597,
+      "step": 7877
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.172491223123475,
+      "learning_rate": 6.230828311248208e-06,
+      "loss": 0.1948,
+      "step": 7878
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.5430529269664262,
+      "learning_rate": 6.229955171233691e-06,
+      "loss": 0.2256,
+      "step": 7879
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.1973441970955783,
+      "learning_rate": 6.229081991294124e-06,
+      "loss": 0.2143,
+      "step": 7880
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.0795731027755293,
+      "learning_rate": 6.228208771457849e-06,
+      "loss": 0.1954,
+      "step": 7881
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.8379071396309485,
+      "learning_rate": 6.22733551175321e-06,
+      "loss": 0.2149,
+      "step": 7882
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.1314323609261268,
+      "learning_rate": 6.226462212208554e-06,
+      "loss": 0.2007,
+      "step": 7883
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.0927012004617973,
+      "learning_rate": 6.22558887285223e-06,
+      "loss": 0.2068,
+      "step": 7884
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.453746470486431,
+      "learning_rate": 6.224715493712586e-06,
+      "loss": 0.1984,
+      "step": 7885
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.1039498411671314,
+      "learning_rate": 6.2238420748179725e-06,
+      "loss": 0.1853,
+      "step": 7886
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.337040735656678,
+      "learning_rate": 6.222968616196743e-06,
+      "loss": 0.2614,
+      "step": 7887
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.9789578300915729,
+      "learning_rate": 6.222095117877247e-06,
+      "loss": 0.1335,
+      "step": 7888
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.2078366246966317,
+      "learning_rate": 6.2212215798878424e-06,
+      "loss": 0.1813,
+      "step": 7889
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.276654976444946,
+      "learning_rate": 6.220348002256883e-06,
+      "loss": 0.1492,
+      "step": 7890
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.424844164755713,
+      "learning_rate": 6.219474385012725e-06,
+      "loss": 0.1881,
+      "step": 7891
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.2997609051582426,
+      "learning_rate": 6.218600728183727e-06,
+      "loss": 0.1987,
+      "step": 7892
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.0409901671311665,
+      "learning_rate": 6.217727031798248e-06,
+      "loss": 0.194,
+      "step": 7893
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.4705118470346243,
+      "learning_rate": 6.216853295884652e-06,
+      "loss": 0.2312,
+      "step": 7894
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.1885981626843678,
+      "learning_rate": 6.215979520471296e-06,
+      "loss": 0.1988,
+      "step": 7895
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.123954758422612,
+      "learning_rate": 6.2151057055865464e-06,
+      "loss": 0.2117,
+      "step": 7896
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.8211676794925684,
+      "learning_rate": 6.214231851258765e-06,
+      "loss": 0.2002,
+      "step": 7897
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.2585445015552312,
+      "learning_rate": 6.213357957516322e-06,
+      "loss": 0.1946,
+      "step": 7898
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.4931518274757782,
+      "learning_rate": 6.212484024387582e-06,
+      "loss": 0.2095,
+      "step": 7899
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.4336834341575497,
+      "learning_rate": 6.211610051900911e-06,
+      "loss": 0.2431,
+      "step": 7900
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.069081588874094,
+      "learning_rate": 6.210736040084683e-06,
+      "loss": 0.183,
+      "step": 7901
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.1926981597227728,
+      "learning_rate": 6.209861988967265e-06,
+      "loss": 0.2152,
+      "step": 7902
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.9480090399769585,
+      "learning_rate": 6.208987898577034e-06,
+      "loss": 0.2117,
+      "step": 7903
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.1015864768699197,
+      "learning_rate": 6.208113768942358e-06,
+      "loss": 0.208,
+      "step": 7904
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.141700249249655,
+      "learning_rate": 6.207239600091614e-06,
+      "loss": 0.2345,
+      "step": 7905
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.177603736514902,
+      "learning_rate": 6.2063653920531795e-06,
+      "loss": 0.2527,
+      "step": 7906
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.6954950899236894,
+      "learning_rate": 6.205491144855432e-06,
+      "loss": 0.1404,
+      "step": 7907
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.950063862975474,
+      "learning_rate": 6.204616858526748e-06,
+      "loss": 0.1527,
+      "step": 7908
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.296826534055025,
+      "learning_rate": 6.203742533095507e-06,
+      "loss": 0.1678,
+      "step": 7909
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.258992210550807,
+      "learning_rate": 6.202868168590092e-06,
+      "loss": 0.2197,
+      "step": 7910
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.4410602805668211,
+      "learning_rate": 6.201993765038884e-06,
+      "loss": 0.2149,
+      "step": 7911
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.8121992778878135,
+      "learning_rate": 6.201119322470268e-06,
+      "loss": 0.1693,
+      "step": 7912
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.1092407101629473,
+      "learning_rate": 6.200244840912627e-06,
+      "loss": 0.1866,
+      "step": 7913
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.2973242291885434,
+      "learning_rate": 6.199370320394349e-06,
+      "loss": 0.1924,
+      "step": 7914
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.3911333440831521,
+      "learning_rate": 6.19849576094382e-06,
+      "loss": 0.2096,
+      "step": 7915
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.0152098852728673,
+      "learning_rate": 6.197621162589431e-06,
+      "loss": 0.1884,
+      "step": 7916
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.3431753937172703,
+      "learning_rate": 6.196746525359569e-06,
+      "loss": 0.1932,
+      "step": 7917
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.9649619057157908,
+      "learning_rate": 6.195871849282628e-06,
+      "loss": 0.1723,
+      "step": 7918
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.2946562113331548,
+      "learning_rate": 6.1949971343869975e-06,
+      "loss": 0.1918,
+      "step": 7919
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.02579723470629,
+      "learning_rate": 6.194122380701073e-06,
+      "loss": 0.1914,
+      "step": 7920
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.3276918434183715,
+      "learning_rate": 6.193247588253251e-06,
+      "loss": 0.1961,
+      "step": 7921
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.9448609416439159,
+      "learning_rate": 6.192372757071924e-06,
+      "loss": 0.1787,
+      "step": 7922
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.388954583160774,
+      "learning_rate": 6.191497887185492e-06,
+      "loss": 0.2002,
+      "step": 7923
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.2979686876524064,
+      "learning_rate": 6.190622978622354e-06,
+      "loss": 0.2179,
+      "step": 7924
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.7837226219734138,
+      "learning_rate": 6.1897480314109106e-06,
+      "loss": 0.1695,
+      "step": 7925
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.057741423801113,
+      "learning_rate": 6.188873045579561e-06,
+      "loss": 0.1507,
+      "step": 7926
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.245183749944172,
+      "learning_rate": 6.187998021156707e-06,
+      "loss": 0.1812,
+      "step": 7927
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.3875441525926169,
+      "learning_rate": 6.187122958170757e-06,
+      "loss": 0.2082,
+      "step": 7928
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.0985337903643926,
+      "learning_rate": 6.186247856650111e-06,
+      "loss": 0.1983,
+      "step": 7929
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.365644982604212,
+      "learning_rate": 6.185372716623178e-06,
+      "loss": 0.2469,
+      "step": 7930
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.1531966940178022,
+      "learning_rate": 6.184497538118364e-06,
+      "loss": 0.2,
+      "step": 7931
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.9887669692771592,
+      "learning_rate": 6.183622321164081e-06,
+      "loss": 0.1745,
+      "step": 7932
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.1445576140716542,
+      "learning_rate": 6.182747065788736e-06,
+      "loss": 0.203,
+      "step": 7933
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.2301660225876272,
+      "learning_rate": 6.18187177202074e-06,
+      "loss": 0.1652,
+      "step": 7934
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.147784605437666,
+      "learning_rate": 6.180996439888508e-06,
+      "loss": 0.2476,
+      "step": 7935
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.2683326726604491,
+      "learning_rate": 6.1801210694204515e-06,
+      "loss": 0.1689,
+      "step": 7936
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.9179489884379991,
+      "learning_rate": 6.179245660644988e-06,
+      "loss": 0.2366,
+      "step": 7937
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.1786235091078308,
+      "learning_rate": 6.17837021359053e-06,
+      "loss": 0.2035,
+      "step": 7938
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.0397171704398767,
+      "learning_rate": 6.177494728285499e-06,
+      "loss": 0.2131,
+      "step": 7939
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.5146924214768147,
+      "learning_rate": 6.1766192047583106e-06,
+      "loss": 0.2064,
+      "step": 7940
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.4565285232139644,
+      "learning_rate": 6.175743643037387e-06,
+      "loss": 0.2116,
+      "step": 7941
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.4304903256900197,
+      "learning_rate": 6.174868043151149e-06,
+      "loss": 0.1832,
+      "step": 7942
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.0560977796479665,
+      "learning_rate": 6.173992405128019e-06,
+      "loss": 0.1828,
+      "step": 7943
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.3546203024630417,
+      "learning_rate": 6.17311672899642e-06,
+      "loss": 0.2753,
+      "step": 7944
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.4364664946887413,
+      "learning_rate": 6.172241014784777e-06,
+      "loss": 0.246,
+      "step": 7945
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.3842326269714773,
+      "learning_rate": 6.171365262521517e-06,
+      "loss": 0.2024,
+      "step": 7946
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.1348647830993572,
+      "learning_rate": 6.1704894722350664e-06,
+      "loss": 0.2339,
+      "step": 7947
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.078678878578541,
+      "learning_rate": 6.169613643953855e-06,
+      "loss": 0.1667,
+      "step": 7948
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.3732054064027954,
+      "learning_rate": 6.16873777770631e-06,
+      "loss": 0.2159,
+      "step": 7949
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.5370604956876006,
+      "learning_rate": 6.1678618735208684e-06,
+      "loss": 0.1826,
+      "step": 7950
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.4038691976982844,
+      "learning_rate": 6.166985931425956e-06,
+      "loss": 0.1379,
+      "step": 7951
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.974279215893303,
+      "learning_rate": 6.166109951450011e-06,
+      "loss": 0.2258,
+      "step": 7952
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.7319407166890878,
+      "learning_rate": 6.165233933621466e-06,
+      "loss": 0.1959,
+      "step": 7953
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.5134702142081908,
+      "learning_rate": 6.164357877968756e-06,
+      "loss": 0.2822,
+      "step": 7954
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.4210961063443068,
+      "learning_rate": 6.163481784520322e-06,
+      "loss": 0.1909,
+      "step": 7955
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.2313648230848813,
+      "learning_rate": 6.162605653304598e-06,
+      "loss": 0.2194,
+      "step": 7956
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.097586802664777,
+      "learning_rate": 6.1617294843500275e-06,
+      "loss": 0.1845,
+      "step": 7957
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.6032263207903472,
+      "learning_rate": 6.160853277685049e-06,
+      "loss": 0.2082,
+      "step": 7958
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.1783224475398006,
+      "learning_rate": 6.159977033338107e-06,
+      "loss": 0.2111,
+      "step": 7959
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.2940702284490255,
+      "learning_rate": 6.1591007513376425e-06,
+      "loss": 0.1946,
+      "step": 7960
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.241909939267715,
+      "learning_rate": 6.1582244317121006e-06,
+      "loss": 0.2211,
+      "step": 7961
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.784236205215144,
+      "learning_rate": 6.157348074489928e-06,
+      "loss": 0.1682,
+      "step": 7962
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.0668040745707688,
+      "learning_rate": 6.156471679699572e-06,
+      "loss": 0.1639,
+      "step": 7963
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.9427898091423328,
+      "learning_rate": 6.155595247369479e-06,
+      "loss": 0.1553,
+      "step": 7964
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.2031618769059755,
+      "learning_rate": 6.1547187775281e-06,
+      "loss": 0.1541,
+      "step": 7965
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.116541619098922,
+      "learning_rate": 6.153842270203887e-06,
+      "loss": 0.1982,
+      "step": 7966
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.0635718941323817,
+      "learning_rate": 6.152965725425289e-06,
+      "loss": 0.2401,
+      "step": 7967
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.9213833973091393,
+      "learning_rate": 6.1520891432207605e-06,
+      "loss": 0.2096,
+      "step": 7968
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.053675798567366,
+      "learning_rate": 6.151212523618757e-06,
+      "loss": 0.19,
+      "step": 7969
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.0426052295796149,
+      "learning_rate": 6.150335866647732e-06,
+      "loss": 0.1969,
+      "step": 7970
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.2736497238515276,
+      "learning_rate": 6.149459172336143e-06,
+      "loss": 0.2092,
+      "step": 7971
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.5069420382216308,
+      "learning_rate": 6.1485824407124484e-06,
+      "loss": 0.186,
+      "step": 7972
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.365238675395867,
+      "learning_rate": 6.147705671805109e-06,
+      "loss": 0.1977,
+      "step": 7973
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.496143332772796,
+      "learning_rate": 6.146828865642581e-06,
+      "loss": 0.2024,
+      "step": 7974
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.5087256692944855,
+      "learning_rate": 6.1459520222533305e-06,
+      "loss": 0.2035,
+      "step": 7975
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.4364000030119484,
+      "learning_rate": 6.145075141665817e-06,
+      "loss": 0.2144,
+      "step": 7976
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.040131143066659,
+      "learning_rate": 6.144198223908508e-06,
+      "loss": 0.2211,
+      "step": 7977
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.0049609273460578,
+      "learning_rate": 6.1433212690098655e-06,
+      "loss": 0.1731,
+      "step": 7978
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.718650955901059,
+      "learning_rate": 6.142444276998356e-06,
+      "loss": 0.2042,
+      "step": 7979
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 2.16126259105119,
+      "learning_rate": 6.141567247902449e-06,
+      "loss": 0.2092,
+      "step": 7980
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.18071453280047,
+      "learning_rate": 6.140690181750613e-06,
+      "loss": 0.2161,
+      "step": 7981
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.9667565501116705,
+      "learning_rate": 6.139813078571317e-06,
+      "loss": 0.1894,
+      "step": 7982
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.0873740389103013,
+      "learning_rate": 6.138935938393034e-06,
+      "loss": 0.0929,
+      "step": 7983
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.0589265845587872,
+      "learning_rate": 6.1380587612442355e-06,
+      "loss": 0.148,
+      "step": 7984
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.9632919434683249,
+      "learning_rate": 6.137181547153395e-06,
+      "loss": 0.1671,
+      "step": 7985
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.5319585085518472,
+      "learning_rate": 6.1363042961489875e-06,
+      "loss": 0.1885,
+      "step": 7986
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.164761928394103,
+      "learning_rate": 6.135427008259489e-06,
+      "loss": 0.1735,
+      "step": 7987
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.9452074441778787,
+      "learning_rate": 6.134549683513377e-06,
+      "loss": 0.2135,
+      "step": 7988
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.7763967548277317,
+      "learning_rate": 6.13367232193913e-06,
+      "loss": 0.2323,
+      "step": 7989
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.2482284151194016,
+      "learning_rate": 6.132794923565227e-06,
+      "loss": 0.1687,
+      "step": 7990
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.3436310418835364,
+      "learning_rate": 6.131917488420151e-06,
+      "loss": 0.2293,
+      "step": 7991
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.329161814291636,
+      "learning_rate": 6.131040016532379e-06,
+      "loss": 0.2242,
+      "step": 7992
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.22029364775237,
+      "learning_rate": 6.1301625079304e-06,
+      "loss": 0.2214,
+      "step": 7993
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.3157298758696665,
+      "learning_rate": 6.129284962642697e-06,
+      "loss": 0.2174,
+      "step": 7994
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.9882225746709499,
+      "learning_rate": 6.128407380697754e-06,
+      "loss": 0.2091,
+      "step": 7995
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.0279351144234612,
+      "learning_rate": 6.1275297621240585e-06,
+      "loss": 0.2271,
+      "step": 7996
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.8011853805164717,
+      "learning_rate": 6.126652106950097e-06,
+      "loss": 0.2131,
+      "step": 7997
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.313049085180216,
+      "learning_rate": 6.125774415204362e-06,
+      "loss": 0.1874,
+      "step": 7998
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.597888113605613,
+      "learning_rate": 6.1248966869153416e-06,
+      "loss": 0.2176,
+      "step": 7999
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.2477769335887756,
+      "learning_rate": 6.124018922111528e-06,
+      "loss": 0.198,
+      "step": 8000
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.5740871895085795,
+      "learning_rate": 6.123141120821413e-06,
+      "loss": 0.2177,
+      "step": 8001
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.1745289822295566,
+      "learning_rate": 6.1222632830734925e-06,
+      "loss": 0.184,
+      "step": 8002
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.877444438162986,
+      "learning_rate": 6.12138540889626e-06,
+      "loss": 0.2311,
+      "step": 8003
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.0046905837894178,
+      "learning_rate": 6.1205074983182125e-06,
+      "loss": 0.2315,
+      "step": 8004
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.3024282052769116,
+      "learning_rate": 6.119629551367847e-06,
+      "loss": 0.1814,
+      "step": 8005
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.2296147152882708,
+      "learning_rate": 6.118751568073662e-06,
+      "loss": 0.2325,
+      "step": 8006
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.4451850319403985,
+      "learning_rate": 6.117873548464159e-06,
+      "loss": 0.2076,
+      "step": 8007
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.1444524438436618,
+      "learning_rate": 6.116995492567837e-06,
+      "loss": 0.2529,
+      "step": 8008
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.506467730962317,
+      "learning_rate": 6.1161174004132e-06,
+      "loss": 0.1927,
+      "step": 8009
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.8834812755908494,
+      "learning_rate": 6.115239272028748e-06,
+      "loss": 0.1954,
+      "step": 8010
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.0874260487432852,
+      "learning_rate": 6.11436110744299e-06,
+      "loss": 0.1702,
+      "step": 8011
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.3905099395235019,
+      "learning_rate": 6.113482906684429e-06,
+      "loss": 0.2483,
+      "step": 8012
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.1491497095557965,
+      "learning_rate": 6.112604669781572e-06,
+      "loss": 0.1427,
+      "step": 8013
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.240774781094117,
+      "learning_rate": 6.111726396762929e-06,
+      "loss": 0.1967,
+      "step": 8014
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.023033559231148,
+      "learning_rate": 6.110848087657007e-06,
+      "loss": 0.1718,
+      "step": 8015
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.539630078452226,
+      "learning_rate": 6.109969742492318e-06,
+      "loss": 0.1961,
+      "step": 8016
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.9083261878202101,
+      "learning_rate": 6.109091361297371e-06,
+      "loss": 0.2114,
+      "step": 8017
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.6635367319570917,
+      "learning_rate": 6.108212944100683e-06,
+      "loss": 0.2002,
+      "step": 8018
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.0523726048722648,
+      "learning_rate": 6.107334490930764e-06,
+      "loss": 0.211,
+      "step": 8019
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.1387209153779356,
+      "learning_rate": 6.106456001816131e-06,
+      "loss": 0.1688,
+      "step": 8020
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.8583416035729798,
+      "learning_rate": 6.105577476785301e-06,
+      "loss": 0.205,
+      "step": 8021
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.0263432618826949,
+      "learning_rate": 6.104698915866789e-06,
+      "loss": 0.2187,
+      "step": 8022
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.1197221004331304,
+      "learning_rate": 6.103820319089115e-06,
+      "loss": 0.2201,
+      "step": 8023
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.0579838561834671,
+      "learning_rate": 6.102941686480799e-06,
+      "loss": 0.2316,
+      "step": 8024
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.108018386483737,
+      "learning_rate": 6.102063018070362e-06,
+      "loss": 0.157,
+      "step": 8025
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.285054591254048,
+      "learning_rate": 6.1011843138863246e-06,
+      "loss": 0.185,
+      "step": 8026
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.0889863979617194,
+      "learning_rate": 6.100305573957213e-06,
+      "loss": 0.1579,
+      "step": 8027
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.240882917230203,
+      "learning_rate": 6.099426798311548e-06,
+      "loss": 0.1842,
+      "step": 8028
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.9743624357156077,
+      "learning_rate": 6.0985479869778575e-06,
+      "loss": 0.2052,
+      "step": 8029
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.848674080682624,
+      "learning_rate": 6.097669139984667e-06,
+      "loss": 0.1715,
+      "step": 8030
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.0861862878749324,
+      "learning_rate": 6.096790257360506e-06,
+      "loss": 0.1766,
+      "step": 8031
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.3559235887040233,
+      "learning_rate": 6.095911339133903e-06,
+      "loss": 0.1971,
+      "step": 8032
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.26856170018044,
+      "learning_rate": 6.095032385333387e-06,
+      "loss": 0.1714,
+      "step": 8033
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.026860248447528,
+      "learning_rate": 6.09415339598749e-06,
+      "loss": 0.1491,
+      "step": 8034
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.051362166019802,
+      "learning_rate": 6.093274371124744e-06,
+      "loss": 0.1631,
+      "step": 8035
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.0894109522883075,
+      "learning_rate": 6.092395310773685e-06,
+      "loss": 0.1902,
+      "step": 8036
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.2289473680139704,
+      "learning_rate": 6.091516214962844e-06,
+      "loss": 0.1506,
+      "step": 8037
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.185688442406499,
+      "learning_rate": 6.09063708372076e-06,
+      "loss": 0.2112,
+      "step": 8038
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.2449392692755963,
+      "learning_rate": 6.089757917075969e-06,
+      "loss": 0.1971,
+      "step": 8039
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.1852593033884498,
+      "learning_rate": 6.08887871505701e-06,
+      "loss": 0.1689,
+      "step": 8040
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.0924837991628422,
+      "learning_rate": 6.087999477692421e-06,
+      "loss": 0.1651,
+      "step": 8041
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.1896828114903428,
+      "learning_rate": 6.087120205010744e-06,
+      "loss": 0.2173,
+      "step": 8042
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.2303152503287358,
+      "learning_rate": 6.086240897040519e-06,
+      "loss": 0.2529,
+      "step": 8043
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.312562963701156,
+      "learning_rate": 6.08536155381029e-06,
+      "loss": 0.199,
+      "step": 8044
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.381441125980712,
+      "learning_rate": 6.084482175348602e-06,
+      "loss": 0.1774,
+      "step": 8045
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.336176404728416,
+      "learning_rate": 6.0836027616839976e-06,
+      "loss": 0.2194,
+      "step": 8046
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.2645982356767145,
+      "learning_rate": 6.082723312845025e-06,
+      "loss": 0.2085,
+      "step": 8047
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.653049201165023,
+      "learning_rate": 6.081843828860231e-06,
+      "loss": 0.1737,
+      "step": 8048
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.5925783216763099,
+      "learning_rate": 6.080964309758164e-06,
+      "loss": 0.2003,
+      "step": 8049
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.6965779446685658,
+      "learning_rate": 6.080084755567375e-06,
+      "loss": 0.1791,
+      "step": 8050
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.6911042144192212,
+      "learning_rate": 6.079205166316412e-06,
+      "loss": 0.2127,
+      "step": 8051
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.7869826746151711,
+      "learning_rate": 6.078325542033831e-06,
+      "loss": 0.2237,
+      "step": 8052
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.2009001602881666,
+      "learning_rate": 6.07744588274818e-06,
+      "loss": 0.2043,
+      "step": 8053
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.9557764166612631,
+      "learning_rate": 6.076566188488018e-06,
+      "loss": 0.2046,
+      "step": 8054
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.2332545458537605,
+      "learning_rate": 6.075686459281898e-06,
+      "loss": 0.1959,
+      "step": 8055
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.1232732577828237,
+      "learning_rate": 6.0748066951583775e-06,
+      "loss": 0.2009,
+      "step": 8056
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.6596488446499957,
+      "learning_rate": 6.073926896146012e-06,
+      "loss": 0.1519,
+      "step": 8057
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.3423478440417782,
+      "learning_rate": 6.073047062273363e-06,
+      "loss": 0.2422,
+      "step": 8058
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.1437187337770875,
+      "learning_rate": 6.072167193568989e-06,
+      "loss": 0.2478,
+      "step": 8059
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.3449408167276047,
+      "learning_rate": 6.0712872900614515e-06,
+      "loss": 0.1808,
+      "step": 8060
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.7771213778923266,
+      "learning_rate": 6.070407351779314e-06,
+      "loss": 0.1772,
+      "step": 8061
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.115808436061608,
+      "learning_rate": 6.069527378751135e-06,
+      "loss": 0.1611,
+      "step": 8062
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.1834441128176585,
+      "learning_rate": 6.068647371005485e-06,
+      "loss": 0.196,
+      "step": 8063
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.7277850048986,
+      "learning_rate": 6.067767328570926e-06,
+      "loss": 0.1651,
+      "step": 8064
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.8148912948094513,
+      "learning_rate": 6.066887251476026e-06,
+      "loss": 0.2088,
+      "step": 8065
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.2840203436562139,
+      "learning_rate": 6.066007139749351e-06,
+      "loss": 0.2424,
+      "step": 8066
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.2573807668469288,
+      "learning_rate": 6.0651269934194725e-06,
+      "loss": 0.2462,
+      "step": 8067
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.2213690185118338,
+      "learning_rate": 6.06424681251496e-06,
+      "loss": 0.2062,
+      "step": 8068
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.9088320882004651,
+      "learning_rate": 6.063366597064381e-06,
+      "loss": 0.1476,
+      "step": 8069
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.2045928466413078,
+      "learning_rate": 6.062486347096315e-06,
+      "loss": 0.1975,
+      "step": 8070
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.3047401109241368,
+      "learning_rate": 6.061606062639327e-06,
+      "loss": 0.2329,
+      "step": 8071
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.2629244036748473,
+      "learning_rate": 6.060725743721998e-06,
+      "loss": 0.1967,
+      "step": 8072
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.205115875691531,
+      "learning_rate": 6.0598453903729e-06,
+      "loss": 0.189,
+      "step": 8073
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.9202074146345334,
+      "learning_rate": 6.058965002620614e-06,
+      "loss": 0.157,
+      "step": 8074
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.1183631986178935,
+      "learning_rate": 6.058084580493713e-06,
+      "loss": 0.2265,
+      "step": 8075
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.5420193709042411,
+      "learning_rate": 6.057204124020778e-06,
+      "loss": 0.2164,
+      "step": 8076
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.0882754989110583,
+      "learning_rate": 6.05632363323039e-06,
+      "loss": 0.2296,
+      "step": 8077
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.2769734293495314,
+      "learning_rate": 6.055443108151129e-06,
+      "loss": 0.2113,
+      "step": 8078
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.0752726708075218,
+      "learning_rate": 6.0545625488115775e-06,
+      "loss": 0.2012,
+      "step": 8079
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.258662295117686,
+      "learning_rate": 6.053681955240319e-06,
+      "loss": 0.1783,
+      "step": 8080
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.2518492587145527,
+      "learning_rate": 6.052801327465939e-06,
+      "loss": 0.1986,
+      "step": 8081
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.147451213227319,
+      "learning_rate": 6.051920665517022e-06,
+      "loss": 0.1668,
+      "step": 8082
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.2213892023420134,
+      "learning_rate": 6.051039969422157e-06,
+      "loss": 0.2643,
+      "step": 8083
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.3053751513864056,
+      "learning_rate": 6.050159239209927e-06,
+      "loss": 0.1962,
+      "step": 8084
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.9783881669779624,
+      "learning_rate": 6.049278474908926e-06,
+      "loss": 0.1551,
+      "step": 8085
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.2681188062109547,
+      "learning_rate": 6.048397676547744e-06,
+      "loss": 0.2183,
+      "step": 8086
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.157043542352323,
+      "learning_rate": 6.047516844154968e-06,
+      "loss": 0.1879,
+      "step": 8087
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.5095796803684594,
+      "learning_rate": 6.046635977759194e-06,
+      "loss": 0.2145,
+      "step": 8088
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.2490095446472362,
+      "learning_rate": 6.045755077389014e-06,
+      "loss": 0.2173,
+      "step": 8089
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.2304790435844863,
+      "learning_rate": 6.044874143073023e-06,
+      "loss": 0.2274,
+      "step": 8090
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.468201235460056,
+      "learning_rate": 6.043993174839816e-06,
+      "loss": 0.2017,
+      "step": 8091
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.0796360021529452,
+      "learning_rate": 6.043112172717991e-06,
+      "loss": 0.137,
+      "step": 8092
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.2991112880799833,
+      "learning_rate": 6.042231136736144e-06,
+      "loss": 0.2115,
+      "step": 8093
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.1031107134339673,
+      "learning_rate": 6.0413500669228765e-06,
+      "loss": 0.1836,
+      "step": 8094
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.3697391241268102,
+      "learning_rate": 6.040468963306786e-06,
+      "loss": 0.2066,
+      "step": 8095
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.3900881974194041,
+      "learning_rate": 6.039587825916474e-06,
+      "loss": 0.1866,
+      "step": 8096
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.4629082774178253,
+      "learning_rate": 6.038706654780545e-06,
+      "loss": 0.2291,
+      "step": 8097
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.168658230042251,
+      "learning_rate": 6.037825449927599e-06,
+      "loss": 0.1902,
+      "step": 8098
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.0171096477512573,
+      "learning_rate": 6.036944211386243e-06,
+      "loss": 0.2115,
+      "step": 8099
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.344298512764906,
+      "learning_rate": 6.036062939185081e-06,
+      "loss": 0.2209,
+      "step": 8100
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.1393133592658853,
+      "learning_rate": 6.0351816333527215e-06,
+      "loss": 0.1714,
+      "step": 8101
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.121188683095761,
+      "learning_rate": 6.03430029391777e-06,
+      "loss": 0.1804,
+      "step": 8102
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.9308190171733297,
+      "learning_rate": 6.0334189209088364e-06,
+      "loss": 0.2023,
+      "step": 8103
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.51777273327069,
+      "learning_rate": 6.0325375143545305e-06,
+      "loss": 0.2054,
+      "step": 8104
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.2144902886501634,
+      "learning_rate": 6.031656074283463e-06,
+      "loss": 0.1743,
+      "step": 8105
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.0771192683319415,
+      "learning_rate": 6.030774600724246e-06,
+      "loss": 0.2228,
+      "step": 8106
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.8221627847287114,
+      "learning_rate": 6.029893093705492e-06,
+      "loss": 0.2297,
+      "step": 8107
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.0285780829036566,
+      "learning_rate": 6.029011553255818e-06,
+      "loss": 0.1822,
+      "step": 8108
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.8167466955036353,
+      "learning_rate": 6.028129979403835e-06,
+      "loss": 0.1607,
+      "step": 8109
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.9940333636414097,
+      "learning_rate": 6.027248372178163e-06,
+      "loss": 0.1384,
+      "step": 8110
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.3565610413311715,
+      "learning_rate": 6.026366731607417e-06,
+      "loss": 0.2226,
+      "step": 8111
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.1130195795357687,
+      "learning_rate": 6.025485057720218e-06,
+      "loss": 0.1835,
+      "step": 8112
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.4221699089189508,
+      "learning_rate": 6.024603350545184e-06,
+      "loss": 0.2162,
+      "step": 8113
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.357951471220783,
+      "learning_rate": 6.023721610110935e-06,
+      "loss": 0.2409,
+      "step": 8114
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.31876641665541,
+      "learning_rate": 6.022839836446096e-06,
+      "loss": 0.2302,
+      "step": 8115
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.309979965688301,
+      "learning_rate": 6.021958029579286e-06,
+      "loss": 0.228,
+      "step": 8116
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.895252508588505,
+      "learning_rate": 6.021076189539131e-06,
+      "loss": 0.1742,
+      "step": 8117
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.6827511253532736,
+      "learning_rate": 6.020194316354255e-06,
+      "loss": 0.1332,
+      "step": 8118
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.2246635887398991,
+      "learning_rate": 6.019312410053286e-06,
+      "loss": 0.2173,
+      "step": 8119
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.3620220547433852,
+      "learning_rate": 6.0184304706648485e-06,
+      "loss": 0.1917,
+      "step": 8120
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.292920425756663,
+      "learning_rate": 6.017548498217575e-06,
+      "loss": 0.1806,
+      "step": 8121
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.0772289564207969,
+      "learning_rate": 6.016666492740091e-06,
+      "loss": 0.1831,
+      "step": 8122
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.7398794696034092,
+      "learning_rate": 6.0157844542610265e-06,
+      "loss": 0.1495,
+      "step": 8123
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.7269863320421728,
+      "learning_rate": 6.014902382809017e-06,
+      "loss": 0.1467,
+      "step": 8124
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.4993508021853648,
+      "learning_rate": 6.01402027841269e-06,
+      "loss": 0.2643,
+      "step": 8125
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.7373387234139164,
+      "learning_rate": 6.013138141100683e-06,
+      "loss": 0.1788,
+      "step": 8126
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.8891329334329574,
+      "learning_rate": 6.0122559709016295e-06,
+      "loss": 0.1784,
+      "step": 8127
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.0257066893517461,
+      "learning_rate": 6.011373767844164e-06,
+      "loss": 0.1561,
+      "step": 8128
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.8746139556062871,
+      "learning_rate": 6.0104915319569255e-06,
+      "loss": 0.1836,
+      "step": 8129
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.1838940453501283,
+      "learning_rate": 6.009609263268551e-06,
+      "loss": 0.2299,
+      "step": 8130
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.0024137107564814,
+      "learning_rate": 6.0087269618076786e-06,
+      "loss": 0.1831,
+      "step": 8131
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.2310416598434866,
+      "learning_rate": 6.007844627602949e-06,
+      "loss": 0.2126,
+      "step": 8132
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.0767643365605253,
+      "learning_rate": 6.0069622606830035e-06,
+      "loss": 0.1529,
+      "step": 8133
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.2165667988367803,
+      "learning_rate": 6.0060798610764825e-06,
+      "loss": 0.1784,
+      "step": 8134
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.4677547928266157,
+      "learning_rate": 6.005197428812033e-06,
+      "loss": 0.1971,
+      "step": 8135
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.4109184247477449,
+      "learning_rate": 6.004314963918295e-06,
+      "loss": 0.2349,
+      "step": 8136
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.0350267005323561,
+      "learning_rate": 6.003432466423918e-06,
+      "loss": 0.2155,
+      "step": 8137
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.9955056417799663,
+      "learning_rate": 6.002549936357543e-06,
+      "loss": 0.2108,
+      "step": 8138
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.0546791535506423,
+      "learning_rate": 6.001667373747824e-06,
+      "loss": 0.1566,
+      "step": 8139
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.306906875431877,
+      "learning_rate": 6.000784778623404e-06,
+      "loss": 0.1952,
+      "step": 8140
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.4013517079844175,
+      "learning_rate": 5.999902151012934e-06,
+      "loss": 0.1897,
+      "step": 8141
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.4154948625262487,
+      "learning_rate": 5.9990194909450675e-06,
+      "loss": 0.2107,
+      "step": 8142
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.7640062219765215,
+      "learning_rate": 5.9981367984484515e-06,
+      "loss": 0.1297,
+      "step": 8143
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.078861943563709,
+      "learning_rate": 5.997254073551743e-06,
+      "loss": 0.2006,
+      "step": 8144
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.3592598910211278,
+      "learning_rate": 5.996371316283591e-06,
+      "loss": 0.2425,
+      "step": 8145
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.2136573932447938,
+      "learning_rate": 5.995488526672656e-06,
+      "loss": 0.1392,
+      "step": 8146
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.0113812255474464,
+      "learning_rate": 5.994605704747589e-06,
+      "loss": 0.1915,
+      "step": 8147
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.1970869712025227,
+      "learning_rate": 5.993722850537051e-06,
+      "loss": 0.1827,
+      "step": 8148
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.1795962532072075,
+      "learning_rate": 5.9928399640696966e-06,
+      "loss": 0.1869,
+      "step": 8149
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.9562466343963836,
+      "learning_rate": 5.991957045374186e-06,
+      "loss": 0.1713,
+      "step": 8150
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.3747449063877712,
+      "learning_rate": 5.9910740944791805e-06,
+      "loss": 0.2272,
+      "step": 8151
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.503058623198543,
+      "learning_rate": 5.990191111413337e-06,
+      "loss": 0.197,
+      "step": 8152
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.5485602071356899,
+      "learning_rate": 5.989308096205323e-06,
+      "loss": 0.2424,
+      "step": 8153
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.5299362597461073,
+      "learning_rate": 5.988425048883799e-06,
+      "loss": 0.1853,
+      "step": 8154
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.8893441015748375,
+      "learning_rate": 5.987541969477429e-06,
+      "loss": 0.1663,
+      "step": 8155
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.2593211365710366,
+      "learning_rate": 5.986658858014878e-06,
+      "loss": 0.1818,
+      "step": 8156
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.3606528888419158,
+      "learning_rate": 5.985775714524815e-06,
+      "loss": 0.2046,
+      "step": 8157
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.681873061789619,
+      "learning_rate": 5.984892539035906e-06,
+      "loss": 0.1629,
+      "step": 8158
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.120376114655221,
+      "learning_rate": 5.984009331576817e-06,
+      "loss": 0.2432,
+      "step": 8159
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.8383646026795293,
+      "learning_rate": 5.98312609217622e-06,
+      "loss": 0.1352,
+      "step": 8160
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.9103468185870828,
+      "learning_rate": 5.9822428208627835e-06,
+      "loss": 0.2478,
+      "step": 8161
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.1969973090802049,
+      "learning_rate": 5.981359517665181e-06,
+      "loss": 0.2071,
+      "step": 8162
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.40752497996003,
+      "learning_rate": 5.980476182612085e-06,
+      "loss": 0.1237,
+      "step": 8163
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.9190983660021931,
+      "learning_rate": 5.979592815732168e-06,
+      "loss": 0.2047,
+      "step": 8164
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.1871811037024338,
+      "learning_rate": 5.978709417054104e-06,
+      "loss": 0.2495,
+      "step": 8165
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.9763554277225024,
+      "learning_rate": 5.97782598660657e-06,
+      "loss": 0.1768,
+      "step": 8166
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.0774902482736568,
+      "learning_rate": 5.976942524418244e-06,
+      "loss": 0.1415,
+      "step": 8167
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.3468004467149677,
+      "learning_rate": 5.976059030517799e-06,
+      "loss": 0.1978,
+      "step": 8168
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.9954172165024388,
+      "learning_rate": 5.975175504933919e-06,
+      "loss": 0.1918,
+      "step": 8169
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.8495272302966577,
+      "learning_rate": 5.974291947695279e-06,
+      "loss": 0.2483,
+      "step": 8170
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.48570325802194,
+      "learning_rate": 5.9734083588305635e-06,
+      "loss": 0.1958,
+      "step": 8171
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.1747056312054807,
+      "learning_rate": 5.972524738368452e-06,
+      "loss": 0.2361,
+      "step": 8172
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.9694950061635543,
+      "learning_rate": 5.97164108633763e-06,
+      "loss": 0.1701,
+      "step": 8173
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.2502248114842436,
+      "learning_rate": 5.970757402766777e-06,
+      "loss": 0.1718,
+      "step": 8174
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.9882821075526199,
+      "learning_rate": 5.969873687684583e-06,
+      "loss": 0.1869,
+      "step": 8175
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.8793027174326373,
+      "learning_rate": 5.96898994111973e-06,
+      "loss": 0.2012,
+      "step": 8176
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.302967891324583,
+      "learning_rate": 5.968106163100906e-06,
+      "loss": 0.2003,
+      "step": 8177
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.4062018624649781,
+      "learning_rate": 5.967222353656801e-06,
+      "loss": 0.2207,
+      "step": 8178
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.4365635081186632,
+      "learning_rate": 5.966338512816101e-06,
+      "loss": 0.1868,
+      "step": 8179
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.850198297489142,
+      "learning_rate": 5.965454640607498e-06,
+      "loss": 0.2469,
+      "step": 8180
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.3035006006990841,
+      "learning_rate": 5.964570737059681e-06,
+      "loss": 0.2203,
+      "step": 8181
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.5367360394799263,
+      "learning_rate": 5.9636868022013435e-06,
+      "loss": 0.2047,
+      "step": 8182
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.1126733944143485,
+      "learning_rate": 5.962802836061178e-06,
+      "loss": 0.176,
+      "step": 8183
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.3083326349653923,
+      "learning_rate": 5.961918838667879e-06,
+      "loss": 0.198,
+      "step": 8184
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.9717188688266777,
+      "learning_rate": 5.961034810050143e-06,
+      "loss": 0.2116,
+      "step": 8185
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.4569127645716133,
+      "learning_rate": 5.960150750236661e-06,
+      "loss": 0.2039,
+      "step": 8186
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.369609929189602,
+      "learning_rate": 5.959266659256138e-06,
+      "loss": 0.1802,
+      "step": 8187
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.0787480772173685,
+      "learning_rate": 5.958382537137264e-06,
+      "loss": 0.1504,
+      "step": 8188
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.1502350927111131,
+      "learning_rate": 5.957498383908743e-06,
+      "loss": 0.2288,
+      "step": 8189
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.1447126587357097,
+      "learning_rate": 5.9566141995992735e-06,
+      "loss": 0.1854,
+      "step": 8190
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.7306685768251961,
+      "learning_rate": 5.955729984237558e-06,
+      "loss": 0.1898,
+      "step": 8191
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.1485224176658404,
+      "learning_rate": 5.954845737852296e-06,
+      "loss": 0.2048,
+      "step": 8192
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.6394732860555297,
+      "learning_rate": 5.953961460472194e-06,
+      "loss": 0.2646,
+      "step": 8193
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.3095749094723799,
+      "learning_rate": 5.953077152125954e-06,
+      "loss": 0.2309,
+      "step": 8194
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.0120249030942081,
+      "learning_rate": 5.95219281284228e-06,
+      "loss": 0.1999,
+      "step": 8195
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.119521470702665,
+      "learning_rate": 5.951308442649881e-06,
+      "loss": 0.1697,
+      "step": 8196
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.3689020618361154,
+      "learning_rate": 5.950424041577463e-06,
+      "loss": 0.2194,
+      "step": 8197
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.9111994210986968,
+      "learning_rate": 5.949539609653734e-06,
+      "loss": 0.1668,
+      "step": 8198
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.9742174488110942,
+      "learning_rate": 5.948655146907403e-06,
+      "loss": 0.1805,
+      "step": 8199
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.1054158964765821,
+      "learning_rate": 5.9477706533671805e-06,
+      "loss": 0.2001,
+      "step": 8200
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.1342591067887204,
+      "learning_rate": 5.946886129061777e-06,
+      "loss": 0.1648,
+      "step": 8201
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.5986520466611573,
+      "learning_rate": 5.946001574019907e-06,
+      "loss": 0.2382,
+      "step": 8202
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.9682847509371649,
+      "learning_rate": 5.945116988270281e-06,
+      "loss": 0.1858,
+      "step": 8203
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.1569970301655526,
+      "learning_rate": 5.944232371841614e-06,
+      "loss": 0.2171,
+      "step": 8204
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.364862451387007,
+      "learning_rate": 5.943347724762621e-06,
+      "loss": 0.2505,
+      "step": 8205
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.1724182991422716,
+      "learning_rate": 5.942463047062018e-06,
+      "loss": 0.1883,
+      "step": 8206
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.8406437877738895,
+      "learning_rate": 5.941578338768523e-06,
+      "loss": 0.1999,
+      "step": 8207
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.3221023055826844,
+      "learning_rate": 5.940693599910854e-06,
+      "loss": 0.1918,
+      "step": 8208
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.114325909953396,
+      "learning_rate": 5.939808830517731e-06,
+      "loss": 0.2016,
+      "step": 8209
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.1938077955085875,
+      "learning_rate": 5.93892403061787e-06,
+      "loss": 0.2402,
+      "step": 8210
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.3965669147400763,
+      "learning_rate": 5.938039200239997e-06,
+      "loss": 0.1786,
+      "step": 8211
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.995144829632872,
+      "learning_rate": 5.937154339412831e-06,
+      "loss": 0.1977,
+      "step": 8212
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.1606508189838671,
+      "learning_rate": 5.936269448165097e-06,
+      "loss": 0.2165,
+      "step": 8213
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.1722121580210638,
+      "learning_rate": 5.9353845265255195e-06,
+      "loss": 0.2293,
+      "step": 8214
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.160402355356959,
+      "learning_rate": 5.934499574522819e-06,
+      "loss": 0.1848,
+      "step": 8215
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.305963999636147,
+      "learning_rate": 5.933614592185728e-06,
+      "loss": 0.2195,
+      "step": 8216
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.3442068598601726,
+      "learning_rate": 5.93272957954297e-06,
+      "loss": 0.2161,
+      "step": 8217
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.8854363217229186,
+      "learning_rate": 5.9318445366232726e-06,
+      "loss": 0.1624,
+      "step": 8218
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.1025455988241681,
+      "learning_rate": 5.930959463455367e-06,
+      "loss": 0.228,
+      "step": 8219
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.93332694243403,
+      "learning_rate": 5.93007436006798e-06,
+      "loss": 0.1434,
+      "step": 8220
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.8020511188271813,
+      "learning_rate": 5.929189226489846e-06,
+      "loss": 0.1349,
+      "step": 8221
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.4332860673644432,
+      "learning_rate": 5.928304062749695e-06,
+      "loss": 0.2375,
+      "step": 8222
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.9230461742932812,
+      "learning_rate": 5.92741886887626e-06,
+      "loss": 0.1803,
+      "step": 8223
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.123665009263918,
+      "learning_rate": 5.926533644898275e-06,
+      "loss": 0.2383,
+      "step": 8224
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.9680653275285905,
+      "learning_rate": 5.925648390844476e-06,
+      "loss": 0.2234,
+      "step": 8225
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.1596049145070033,
+      "learning_rate": 5.924763106743596e-06,
+      "loss": 0.2173,
+      "step": 8226
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.0371335883466,
+      "learning_rate": 5.923877792624376e-06,
+      "loss": 0.1738,
+      "step": 8227
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 18.479997698630978,
+      "learning_rate": 5.922992448515551e-06,
+      "loss": 0.185,
+      "step": 8228
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.3439502476758338,
+      "learning_rate": 5.92210707444586e-06,
+      "loss": 0.1835,
+      "step": 8229
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.3169890961376682,
+      "learning_rate": 5.921221670444044e-06,
+      "loss": 0.1851,
+      "step": 8230
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.1350756088854888,
+      "learning_rate": 5.920336236538841e-06,
+      "loss": 0.2248,
+      "step": 8231
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.0596526258991024,
+      "learning_rate": 5.919450772758997e-06,
+      "loss": 0.2048,
+      "step": 8232
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.1037865213392657,
+      "learning_rate": 5.91856527913325e-06,
+      "loss": 0.1896,
+      "step": 8233
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.005715300602532,
+      "learning_rate": 5.917679755690347e-06,
+      "loss": 0.2427,
+      "step": 8234
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.3492583992968583,
+      "learning_rate": 5.91679420245903e-06,
+      "loss": 0.2137,
+      "step": 8235
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.8952248645253664,
+      "learning_rate": 5.915908619468047e-06,
+      "loss": 0.1917,
+      "step": 8236
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.4467080054917758,
+      "learning_rate": 5.9150230067461445e-06,
+      "loss": 0.2288,
+      "step": 8237
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.4612611511574285,
+      "learning_rate": 5.914137364322068e-06,
+      "loss": 0.2318,
+      "step": 8238
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.1467999425306885,
+      "learning_rate": 5.9132516922245665e-06,
+      "loss": 0.2119,
+      "step": 8239
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.5092008306583031,
+      "learning_rate": 5.91236599048239e-06,
+      "loss": 0.2204,
+      "step": 8240
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.2718008705159352,
+      "learning_rate": 5.91148025912429e-06,
+      "loss": 0.1843,
+      "step": 8241
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.8378212106402007,
+      "learning_rate": 5.910594498179015e-06,
+      "loss": 0.1723,
+      "step": 8242
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.0340592095330492,
+      "learning_rate": 5.9097087076753214e-06,
+      "loss": 0.1889,
+      "step": 8243
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.2654577221232712,
+      "learning_rate": 5.9088228876419576e-06,
+      "loss": 0.1722,
+      "step": 8244
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.8129582637251336,
+      "learning_rate": 5.907937038107681e-06,
+      "loss": 0.2116,
+      "step": 8245
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.1505904056602834,
+      "learning_rate": 5.907051159101247e-06,
+      "loss": 0.1748,
+      "step": 8246
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.2258267078759795,
+      "learning_rate": 5.90616525065141e-06,
+      "loss": 0.2116,
+      "step": 8247
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.1969281992157286,
+      "learning_rate": 5.905279312786928e-06,
+      "loss": 0.2151,
+      "step": 8248
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.9195595835758295,
+      "learning_rate": 5.904393345536559e-06,
+      "loss": 0.1824,
+      "step": 8249
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.992077335113965,
+      "learning_rate": 5.903507348929062e-06,
+      "loss": 0.1646,
+      "step": 8250
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.1741127963813212,
+      "learning_rate": 5.902621322993195e-06,
+      "loss": 0.1502,
+      "step": 8251
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.2482064373475956,
+      "learning_rate": 5.901735267757723e-06,
+      "loss": 0.2341,
+      "step": 8252
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.639752601100361,
+      "learning_rate": 5.900849183251405e-06,
+      "loss": 0.284,
+      "step": 8253
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.168868286221025,
+      "learning_rate": 5.8999630695030055e-06,
+      "loss": 0.2293,
+      "step": 8254
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.9023122059481472,
+      "learning_rate": 5.899076926541286e-06,
+      "loss": 0.1744,
+      "step": 8255
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.1062503282632719,
+      "learning_rate": 5.898190754395012e-06,
+      "loss": 0.207,
+      "step": 8256
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.9765242340220145,
+      "learning_rate": 5.897304553092952e-06,
+      "loss": 0.2156,
+      "step": 8257
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.0431416623585825,
+      "learning_rate": 5.896418322663868e-06,
+      "loss": 0.1961,
+      "step": 8258
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.9677093371075385,
+      "learning_rate": 5.895532063136531e-06,
+      "loss": 0.2139,
+      "step": 8259
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.1625085609900154,
+      "learning_rate": 5.894645774539707e-06,
+      "loss": 0.2073,
+      "step": 8260
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.458707249479977,
+      "learning_rate": 5.893759456902167e-06,
+      "loss": 0.2608,
+      "step": 8261
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.4600808027429564,
+      "learning_rate": 5.892873110252682e-06,
+      "loss": 0.1763,
+      "step": 8262
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.0386968593167707,
+      "learning_rate": 5.891986734620022e-06,
+      "loss": 0.1927,
+      "step": 8263
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.0802564404794859,
+      "learning_rate": 5.891100330032959e-06,
+      "loss": 0.1969,
+      "step": 8264
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.455967537864231,
+      "learning_rate": 5.890213896520268e-06,
+      "loss": 0.1624,
+      "step": 8265
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.0962493260975812,
+      "learning_rate": 5.889327434110722e-06,
+      "loss": 0.1619,
+      "step": 8266
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.4232103004197205,
+      "learning_rate": 5.888440942833095e-06,
+      "loss": 0.1526,
+      "step": 8267
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.087909607576121,
+      "learning_rate": 5.887554422716165e-06,
+      "loss": 0.1731,
+      "step": 8268
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.9601226779044019,
+      "learning_rate": 5.886667873788708e-06,
+      "loss": 0.169,
+      "step": 8269
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.1987243869728215,
+      "learning_rate": 5.885781296079501e-06,
+      "loss": 0.2221,
+      "step": 8270
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.3576682601276318,
+      "learning_rate": 5.884894689617324e-06,
+      "loss": 0.2344,
+      "step": 8271
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.3046589922503007,
+      "learning_rate": 5.884008054430957e-06,
+      "loss": 0.1922,
+      "step": 8272
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.1415300061305451,
+      "learning_rate": 5.88312139054918e-06,
+      "loss": 0.2089,
+      "step": 8273
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.9328768842928066,
+      "learning_rate": 5.882234698000775e-06,
+      "loss": 0.2294,
+      "step": 8274
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.1288922384903919,
+      "learning_rate": 5.881347976814524e-06,
+      "loss": 0.1882,
+      "step": 8275
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.4827820123064623,
+      "learning_rate": 5.8804612270192115e-06,
+      "loss": 0.189,
+      "step": 8276
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.138372211709864,
+      "learning_rate": 5.8795744486436214e-06,
+      "loss": 0.1916,
+      "step": 8277
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.0805332749325693,
+      "learning_rate": 5.878687641716539e-06,
+      "loss": 0.2204,
+      "step": 8278
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.0549084537909224,
+      "learning_rate": 5.877800806266749e-06,
+      "loss": 0.1829,
+      "step": 8279
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.1512622089753373,
+      "learning_rate": 5.87691394232304e-06,
+      "loss": 0.1911,
+      "step": 8280
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.9300172205068983,
+      "learning_rate": 5.876027049914202e-06,
+      "loss": 0.2182,
+      "step": 8281
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.7469035018486333,
+      "learning_rate": 5.875140129069022e-06,
+      "loss": 0.1449,
+      "step": 8282
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.1628763857255784,
+      "learning_rate": 5.874253179816289e-06,
+      "loss": 0.2372,
+      "step": 8283
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.0092526429599782,
+      "learning_rate": 5.873366202184796e-06,
+      "loss": 0.1737,
+      "step": 8284
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.2645282165354736,
+      "learning_rate": 5.872479196203333e-06,
+      "loss": 0.1906,
+      "step": 8285
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.1808226441813974,
+      "learning_rate": 5.871592161900695e-06,
+      "loss": 0.2067,
+      "step": 8286
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 2.068910750398681,
+      "learning_rate": 5.8707050993056725e-06,
+      "loss": 0.1814,
+      "step": 8287
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.3401568994946376,
+      "learning_rate": 5.869818008447063e-06,
+      "loss": 0.1924,
+      "step": 8288
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.2083218381460992,
+      "learning_rate": 5.86893088935366e-06,
+      "loss": 0.2341,
+      "step": 8289
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.2526982347124287,
+      "learning_rate": 5.868043742054261e-06,
+      "loss": 0.1977,
+      "step": 8290
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.9492831326066766,
+      "learning_rate": 5.8671565665776635e-06,
+      "loss": 0.1845,
+      "step": 8291
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.029220949228247,
+      "learning_rate": 5.866269362952664e-06,
+      "loss": 0.1865,
+      "step": 8292
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.0916596720543605,
+      "learning_rate": 5.865382131208064e-06,
+      "loss": 0.1434,
+      "step": 8293
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.3359849723389938,
+      "learning_rate": 5.8644948713726615e-06,
+      "loss": 0.1607,
+      "step": 8294
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.2319452827252872,
+      "learning_rate": 5.863607583475259e-06,
+      "loss": 0.205,
+      "step": 8295
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.3470490428367596,
+      "learning_rate": 5.862720267544657e-06,
+      "loss": 0.163,
+      "step": 8296
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.0999918724702638,
+      "learning_rate": 5.861832923609661e-06,
+      "loss": 0.2161,
+      "step": 8297
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.3385535164195994,
+      "learning_rate": 5.86094555169907e-06,
+      "loss": 0.1652,
+      "step": 8298
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.964208780909296,
+      "learning_rate": 5.860058151841692e-06,
+      "loss": 0.1677,
+      "step": 8299
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.1858165817536315,
+      "learning_rate": 5.8591707240663335e-06,
+      "loss": 0.2091,
+      "step": 8300
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 2.55960266954447,
+      "learning_rate": 5.858283268401798e-06,
+      "loss": 0.143,
+      "step": 8301
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.3447503923730086,
+      "learning_rate": 5.857395784876893e-06,
+      "loss": 0.1813,
+      "step": 8302
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.109194020499273,
+      "learning_rate": 5.856508273520429e-06,
+      "loss": 0.1565,
+      "step": 8303
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.4420836596234619,
+      "learning_rate": 5.855620734361215e-06,
+      "loss": 0.1885,
+      "step": 8304
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.9955493898112243,
+      "learning_rate": 5.854733167428057e-06,
+      "loss": 0.1501,
+      "step": 8305
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.157409779011981,
+      "learning_rate": 5.853845572749772e-06,
+      "loss": 0.1787,
+      "step": 8306
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.395236945376065,
+      "learning_rate": 5.852957950355167e-06,
+      "loss": 0.2011,
+      "step": 8307
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.4366605111080244,
+      "learning_rate": 5.852070300273057e-06,
+      "loss": 0.2112,
+      "step": 8308
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.498217845247075,
+      "learning_rate": 5.851182622532256e-06,
+      "loss": 0.2577,
+      "step": 8309
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.9878291068299763,
+      "learning_rate": 5.8502949171615756e-06,
+      "loss": 0.1986,
+      "step": 8310
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.161071371469355,
+      "learning_rate": 5.8494071841898346e-06,
+      "loss": 0.2211,
+      "step": 8311
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.0978495200300533,
+      "learning_rate": 5.848519423645846e-06,
+      "loss": 0.2054,
+      "step": 8312
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.182970833605865,
+      "learning_rate": 5.847631635558432e-06,
+      "loss": 0.1846,
+      "step": 8313
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.9056139514412361,
+      "learning_rate": 5.846743819956405e-06,
+      "loss": 0.1306,
+      "step": 8314
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.1286687964213036,
+      "learning_rate": 5.845855976868589e-06,
+      "loss": 0.1259,
+      "step": 8315
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.458795430563673,
+      "learning_rate": 5.8449681063238e-06,
+      "loss": 0.2028,
+      "step": 8316
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.9200644687543034,
+      "learning_rate": 5.84408020835086e-06,
+      "loss": 0.2071,
+      "step": 8317
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.25650293285401,
+      "learning_rate": 5.843192282978592e-06,
+      "loss": 0.1834,
+      "step": 8318
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.3209388195398553,
+      "learning_rate": 5.8423043302358176e-06,
+      "loss": 0.1716,
+      "step": 8319
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.3899762537011902,
+      "learning_rate": 5.841416350151361e-06,
+      "loss": 0.1995,
+      "step": 8320
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.6841758553683535,
+      "learning_rate": 5.840528342754044e-06,
+      "loss": 0.2455,
+      "step": 8321
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.1577778223509987,
+      "learning_rate": 5.839640308072695e-06,
+      "loss": 0.1996,
+      "step": 8322
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.327179927326027,
+      "learning_rate": 5.838752246136138e-06,
+      "loss": 0.2104,
+      "step": 8323
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.720115266766152,
+      "learning_rate": 5.837864156973202e-06,
+      "loss": 0.1991,
+      "step": 8324
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.2517933200485014,
+      "learning_rate": 5.836976040612712e-06,
+      "loss": 0.2269,
+      "step": 8325
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.456646756020919,
+      "learning_rate": 5.8360878970835e-06,
+      "loss": 0.2208,
+      "step": 8326
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.9966867167100875,
+      "learning_rate": 5.835199726414394e-06,
+      "loss": 0.2231,
+      "step": 8327
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.1986360159214695,
+      "learning_rate": 5.834311528634223e-06,
+      "loss": 0.2105,
+      "step": 8328
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.0449574687278882,
+      "learning_rate": 5.833423303771821e-06,
+      "loss": 0.1893,
+      "step": 8329
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.1784483483583226,
+      "learning_rate": 5.832535051856019e-06,
+      "loss": 0.1645,
+      "step": 8330
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.0723983060673348,
+      "learning_rate": 5.831646772915651e-06,
+      "loss": 0.1754,
+      "step": 8331
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.3969322126734882,
+      "learning_rate": 5.8307584669795504e-06,
+      "loss": 0.2056,
+      "step": 8332
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.4615188196851796,
+      "learning_rate": 5.829870134076552e-06,
+      "loss": 0.1959,
+      "step": 8333
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.1185997909396634,
+      "learning_rate": 5.828981774235492e-06,
+      "loss": 0.1766,
+      "step": 8334
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 2.1594019222927465,
+      "learning_rate": 5.828093387485207e-06,
+      "loss": 0.2306,
+      "step": 8335
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.9285998513645363,
+      "learning_rate": 5.827204973854534e-06,
+      "loss": 0.1911,
+      "step": 8336
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.1783751014834063,
+      "learning_rate": 5.826316533372313e-06,
+      "loss": 0.1902,
+      "step": 8337
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.9047881133470461,
+      "learning_rate": 5.825428066067382e-06,
+      "loss": 0.2291,
+      "step": 8338
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.2103155415450217,
+      "learning_rate": 5.824539571968579e-06,
+      "loss": 0.2323,
+      "step": 8339
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.1410403360957815,
+      "learning_rate": 5.823651051104749e-06,
+      "loss": 0.1702,
+      "step": 8340
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.3263857420442386,
+      "learning_rate": 5.822762503504732e-06,
+      "loss": 0.1989,
+      "step": 8341
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.2129470148661277,
+      "learning_rate": 5.821873929197371e-06,
+      "loss": 0.1914,
+      "step": 8342
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.161603602774359,
+      "learning_rate": 5.82098532821151e-06,
+      "loss": 0.2108,
+      "step": 8343
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.8431524880794254,
+      "learning_rate": 5.820096700575994e-06,
+      "loss": 0.1625,
+      "step": 8344
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.9712210142935954,
+      "learning_rate": 5.819208046319667e-06,
+      "loss": 0.2096,
+      "step": 8345
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.068617094013349,
+      "learning_rate": 5.818319365471374e-06,
+      "loss": 0.2121,
+      "step": 8346
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.1591177251335125,
+      "learning_rate": 5.817430658059966e-06,
+      "loss": 0.1387,
+      "step": 8347
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.9641573275046754,
+      "learning_rate": 5.816541924114287e-06,
+      "loss": 0.1691,
+      "step": 8348
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.0346048415955413,
+      "learning_rate": 5.81565316366319e-06,
+      "loss": 0.2176,
+      "step": 8349
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.8833745100081539,
+      "learning_rate": 5.81476437673552e-06,
+      "loss": 0.1881,
+      "step": 8350
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.3082782920253604,
+      "learning_rate": 5.8138755633601295e-06,
+      "loss": 0.2321,
+      "step": 8351
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.1816151868430766,
+      "learning_rate": 5.8129867235658725e-06,
+      "loss": 0.1893,
+      "step": 8352
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.9649223515698789,
+      "learning_rate": 5.812097857381598e-06,
+      "loss": 0.1911,
+      "step": 8353
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.1441459576842912,
+      "learning_rate": 5.81120896483616e-06,
+      "loss": 0.1632,
+      "step": 8354
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.4026033671216624,
+      "learning_rate": 5.810320045958412e-06,
+      "loss": 0.1848,
+      "step": 8355
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.1714303977299483,
+      "learning_rate": 5.80943110077721e-06,
+      "loss": 0.1891,
+      "step": 8356
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.1394689577144999,
+      "learning_rate": 5.8085421293214085e-06,
+      "loss": 0.2257,
+      "step": 8357
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.2107594836040068,
+      "learning_rate": 5.8076531316198645e-06,
+      "loss": 0.1763,
+      "step": 8358
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.1129534657103766,
+      "learning_rate": 5.806764107701436e-06,
+      "loss": 0.1831,
+      "step": 8359
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.9637409811674034,
+      "learning_rate": 5.80587505759498e-06,
+      "loss": 0.1854,
+      "step": 8360
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.2643829582551285,
+      "learning_rate": 5.804985981329356e-06,
+      "loss": 0.1884,
+      "step": 8361
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.7827166963627927,
+      "learning_rate": 5.804096878933425e-06,
+      "loss": 0.1911,
+      "step": 8362
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.9740607374418269,
+      "learning_rate": 5.803207750436048e-06,
+      "loss": 0.243,
+      "step": 8363
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.9038208456254906,
+      "learning_rate": 5.802318595866083e-06,
+      "loss": 0.2031,
+      "step": 8364
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.0738662470811746,
+      "learning_rate": 5.801429415252396e-06,
+      "loss": 0.1278,
+      "step": 8365
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.1344784900258829,
+      "learning_rate": 5.800540208623849e-06,
+      "loss": 0.2307,
+      "step": 8366
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.0493272201243582,
+      "learning_rate": 5.799650976009308e-06,
+      "loss": 0.2053,
+      "step": 8367
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.9004906456062727,
+      "learning_rate": 5.798761717437635e-06,
+      "loss": 0.1501,
+      "step": 8368
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.8423957336548547,
+      "learning_rate": 5.797872432937697e-06,
+      "loss": 0.2914,
+      "step": 8369
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.0534764860730912,
+      "learning_rate": 5.796983122538362e-06,
+      "loss": 0.1962,
+      "step": 8370
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.3675603174337578,
+      "learning_rate": 5.796093786268497e-06,
+      "loss": 0.2221,
+      "step": 8371
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.198785139651283,
+      "learning_rate": 5.79520442415697e-06,
+      "loss": 0.1749,
+      "step": 8372
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.78720368308451,
+      "learning_rate": 5.794315036232648e-06,
+      "loss": 0.234,
+      "step": 8373
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.8826024071337355,
+      "learning_rate": 5.793425622524406e-06,
+      "loss": 0.1306,
+      "step": 8374
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.922254355013154,
+      "learning_rate": 5.79253618306111e-06,
+      "loss": 0.2045,
+      "step": 8375
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.0459660776244408,
+      "learning_rate": 5.791646717871634e-06,
+      "loss": 0.1476,
+      "step": 8376
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.2051937443496719,
+      "learning_rate": 5.790757226984851e-06,
+      "loss": 0.1732,
+      "step": 8377
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.351655504852443,
+      "learning_rate": 5.789867710429632e-06,
+      "loss": 0.1742,
+      "step": 8378
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.3543065225077673,
+      "learning_rate": 5.788978168234855e-06,
+      "loss": 0.2097,
+      "step": 8379
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.9998735686933486,
+      "learning_rate": 5.788088600429392e-06,
+      "loss": 0.195,
+      "step": 8380
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.0633082173394135,
+      "learning_rate": 5.787199007042121e-06,
+      "loss": 0.1996,
+      "step": 8381
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.8718480342707904,
+      "learning_rate": 5.7863093881019164e-06,
+      "loss": 0.1917,
+      "step": 8382
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.3877680717147847,
+      "learning_rate": 5.785419743637657e-06,
+      "loss": 0.215,
+      "step": 8383
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.4977153973349364,
+      "learning_rate": 5.7845300736782205e-06,
+      "loss": 0.185,
+      "step": 8384
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.2712488747411714,
+      "learning_rate": 5.783640378252488e-06,
+      "loss": 0.2348,
+      "step": 8385
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.8867091224061465,
+      "learning_rate": 5.782750657389338e-06,
+      "loss": 0.2509,
+      "step": 8386
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.078519614190527,
+      "learning_rate": 5.781860911117649e-06,
+      "loss": 0.1925,
+      "step": 8387
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.1793469291089633,
+      "learning_rate": 5.780971139466308e-06,
+      "loss": 0.17,
+      "step": 8388
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.2043820415700757,
+      "learning_rate": 5.780081342464195e-06,
+      "loss": 0.1463,
+      "step": 8389
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.5214525153334275,
+      "learning_rate": 5.779191520140193e-06,
+      "loss": 0.206,
+      "step": 8390
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.2895231348704732,
+      "learning_rate": 5.778301672523185e-06,
+      "loss": 0.2381,
+      "step": 8391
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.0566640382383299,
+      "learning_rate": 5.777411799642058e-06,
+      "loss": 0.1989,
+      "step": 8392
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.085702618575968,
+      "learning_rate": 5.776521901525694e-06,
+      "loss": 0.1858,
+      "step": 8393
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.1261050406586404,
+      "learning_rate": 5.775631978202987e-06,
+      "loss": 0.2071,
+      "step": 8394
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.0451715527579168,
+      "learning_rate": 5.774742029702817e-06,
+      "loss": 0.1913,
+      "step": 8395
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.7174011746616549,
+      "learning_rate": 5.773852056054074e-06,
+      "loss": 0.2374,
+      "step": 8396
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.4819281940896014,
+      "learning_rate": 5.772962057285651e-06,
+      "loss": 0.2129,
+      "step": 8397
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.1426277622624148,
+      "learning_rate": 5.772072033426436e-06,
+      "loss": 0.1538,
+      "step": 8398
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.361936129685688,
+      "learning_rate": 5.771181984505318e-06,
+      "loss": 0.2513,
+      "step": 8399
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.2302488645239438,
+      "learning_rate": 5.770291910551188e-06,
+      "loss": 0.1978,
+      "step": 8400
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.9237736349185874,
+      "learning_rate": 5.769401811592941e-06,
+      "loss": 0.1958,
+      "step": 8401
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.0871560246419851,
+      "learning_rate": 5.768511687659467e-06,
+      "loss": 0.1394,
+      "step": 8402
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.1469979768558942,
+      "learning_rate": 5.767621538779663e-06,
+      "loss": 0.2173,
+      "step": 8403
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.1196871583061325,
+      "learning_rate": 5.766731364982422e-06,
+      "loss": 0.1556,
+      "step": 8404
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.397737730219552,
+      "learning_rate": 5.76584116629664e-06,
+      "loss": 0.2348,
+      "step": 8405
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.5441441030967367,
+      "learning_rate": 5.764950942751213e-06,
+      "loss": 0.2568,
+      "step": 8406
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.8868861901583401,
+      "learning_rate": 5.76406069437504e-06,
+      "loss": 0.1533,
+      "step": 8407
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.3434394932333018,
+      "learning_rate": 5.763170421197016e-06,
+      "loss": 0.2166,
+      "step": 8408
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.1918773782723378,
+      "learning_rate": 5.762280123246043e-06,
+      "loss": 0.2355,
+      "step": 8409
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.7918873257720246,
+      "learning_rate": 5.761389800551017e-06,
+      "loss": 0.1403,
+      "step": 8410
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.1676671777471912,
+      "learning_rate": 5.76049945314084e-06,
+      "loss": 0.2184,
+      "step": 8411
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.1768034202000115,
+      "learning_rate": 5.759609081044415e-06,
+      "loss": 0.1854,
+      "step": 8412
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.9097033139056288,
+      "learning_rate": 5.758718684290641e-06,
+      "loss": 0.1723,
+      "step": 8413
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.169950914738973,
+      "learning_rate": 5.757828262908423e-06,
+      "loss": 0.1927,
+      "step": 8414
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.991431917259602,
+      "learning_rate": 5.756937816926664e-06,
+      "loss": 0.1424,
+      "step": 8415
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.2026214892917313,
+      "learning_rate": 5.756047346374268e-06,
+      "loss": 0.2008,
+      "step": 8416
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.404026103542403,
+      "learning_rate": 5.755156851280142e-06,
+      "loss": 0.2305,
+      "step": 8417
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.2377294114026631,
+      "learning_rate": 5.754266331673187e-06,
+      "loss": 0.1816,
+      "step": 8418
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.8241814711126695,
+      "learning_rate": 5.753375787582316e-06,
+      "loss": 0.1969,
+      "step": 8419
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.257190577246996,
+      "learning_rate": 5.752485219036431e-06,
+      "loss": 0.2327,
+      "step": 8420
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.0096495891875443,
+      "learning_rate": 5.751594626064445e-06,
+      "loss": 0.1709,
+      "step": 8421
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.4776137258117144,
+      "learning_rate": 5.750704008695265e-06,
+      "loss": 0.1671,
+      "step": 8422
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.0494379130325677,
+      "learning_rate": 5.7498133669578e-06,
+      "loss": 0.1758,
+      "step": 8423
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.3434909471047831,
+      "learning_rate": 5.748922700880963e-06,
+      "loss": 0.1953,
+      "step": 8424
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.813355799820931,
+      "learning_rate": 5.748032010493666e-06,
+      "loss": 0.198,
+      "step": 8425
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.277505760826085,
+      "learning_rate": 5.747141295824819e-06,
+      "loss": 0.1735,
+      "step": 8426
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.4261083410608952,
+      "learning_rate": 5.7462505569033345e-06,
+      "loss": 0.2471,
+      "step": 8427
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.3610163513623101,
+      "learning_rate": 5.745359793758129e-06,
+      "loss": 0.2135,
+      "step": 8428
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.6673097380646598,
+      "learning_rate": 5.744469006418114e-06,
+      "loss": 0.2283,
+      "step": 8429
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.7755735407395482,
+      "learning_rate": 5.743578194912209e-06,
+      "loss": 0.1845,
+      "step": 8430
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.1646742877205682,
+      "learning_rate": 5.7426873592693265e-06,
+      "loss": 0.1926,
+      "step": 8431
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.8937785050708216,
+      "learning_rate": 5.741796499518386e-06,
+      "loss": 0.2098,
+      "step": 8432
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.3259535906199937,
+      "learning_rate": 5.740905615688306e-06,
+      "loss": 0.146,
+      "step": 8433
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.3665799534893681,
+      "learning_rate": 5.740014707808003e-06,
+      "loss": 0.187,
+      "step": 8434
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.1667403743513176,
+      "learning_rate": 5.739123775906397e-06,
+      "loss": 0.1944,
+      "step": 8435
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.1628450677550592,
+      "learning_rate": 5.738232820012407e-06,
+      "loss": 0.2084,
+      "step": 8436
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.3509328069876718,
+      "learning_rate": 5.7373418401549565e-06,
+      "loss": 0.2266,
+      "step": 8437
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.3048790488426496,
+      "learning_rate": 5.736450836362965e-06,
+      "loss": 0.1754,
+      "step": 8438
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.8721646688234644,
+      "learning_rate": 5.735559808665357e-06,
+      "loss": 0.2164,
+      "step": 8439
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.1085679382588276,
+      "learning_rate": 5.734668757091052e-06,
+      "loss": 0.2373,
+      "step": 8440
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.9690946977089218,
+      "learning_rate": 5.733777681668978e-06,
+      "loss": 0.2314,
+      "step": 8441
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.2900881155132309,
+      "learning_rate": 5.7328865824280596e-06,
+      "loss": 0.2012,
+      "step": 8442
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.210835018743671,
+      "learning_rate": 5.731995459397221e-06,
+      "loss": 0.1799,
+      "step": 8443
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.1020856267839712,
+      "learning_rate": 5.731104312605389e-06,
+      "loss": 0.1938,
+      "step": 8444
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.9726361115603738,
+      "learning_rate": 5.73021314208149e-06,
+      "loss": 0.2065,
+      "step": 8445
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.0767603163722066,
+      "learning_rate": 5.7293219478544526e-06,
+      "loss": 0.1806,
+      "step": 8446
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.327247907828031,
+      "learning_rate": 5.728430729953205e-06,
+      "loss": 0.1893,
+      "step": 8447
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 2.076624619076507,
+      "learning_rate": 5.727539488406678e-06,
+      "loss": 0.1731,
+      "step": 8448
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.418307975776956,
+      "learning_rate": 5.7266482232438e-06,
+      "loss": 0.2365,
+      "step": 8449
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.6653181776203014,
+      "learning_rate": 5.725756934493503e-06,
+      "loss": 0.1958,
+      "step": 8450
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.8628109626210577,
+      "learning_rate": 5.7248656221847184e-06,
+      "loss": 0.1808,
+      "step": 8451
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.20886489943386,
+      "learning_rate": 5.72397428634638e-06,
+      "loss": 0.1789,
+      "step": 8452
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.4664266604190488,
+      "learning_rate": 5.72308292700742e-06,
+      "loss": 0.1742,
+      "step": 8453
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.9793490174954415,
+      "learning_rate": 5.72219154419677e-06,
+      "loss": 0.2217,
+      "step": 8454
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.2082541807396563,
+      "learning_rate": 5.72130013794337e-06,
+      "loss": 0.1965,
+      "step": 8455
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.4663659324956662,
+      "learning_rate": 5.720408708276151e-06,
+      "loss": 0.1444,
+      "step": 8456
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.2545419103777888,
+      "learning_rate": 5.719517255224052e-06,
+      "loss": 0.2381,
+      "step": 8457
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.369116950272553,
+      "learning_rate": 5.718625778816007e-06,
+      "loss": 0.1929,
+      "step": 8458
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.9276916413754778,
+      "learning_rate": 5.7177342790809576e-06,
+      "loss": 0.2034,
+      "step": 8459
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.3015341140839023,
+      "learning_rate": 5.716842756047841e-06,
+      "loss": 0.252,
+      "step": 8460
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.2144890540333688,
+      "learning_rate": 5.715951209745597e-06,
+      "loss": 0.176,
+      "step": 8461
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.2861867471912682,
+      "learning_rate": 5.715059640203164e-06,
+      "loss": 0.1735,
+      "step": 8462
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.144328577890737,
+      "learning_rate": 5.714168047449484e-06,
+      "loss": 0.1903,
+      "step": 8463
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.8744570610137071,
+      "learning_rate": 5.713276431513499e-06,
+      "loss": 0.227,
+      "step": 8464
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.047732041538181,
+      "learning_rate": 5.71238479242415e-06,
+      "loss": 0.149,
+      "step": 8465
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.2755319858742082,
+      "learning_rate": 5.711493130210381e-06,
+      "loss": 0.2105,
+      "step": 8466
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.2021203785402177,
+      "learning_rate": 5.710601444901136e-06,
+      "loss": 0.2524,
+      "step": 8467
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.4306926372916282,
+      "learning_rate": 5.70970973652536e-06,
+      "loss": 0.2528,
+      "step": 8468
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.1306513461559131,
+      "learning_rate": 5.7088180051119975e-06,
+      "loss": 0.1745,
+      "step": 8469
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.26679149635954,
+      "learning_rate": 5.707926250689996e-06,
+      "loss": 0.1686,
+      "step": 8470
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.6355775459899522,
+      "learning_rate": 5.7070344732883e-06,
+      "loss": 0.1923,
+      "step": 8471
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.2068459312987025,
+      "learning_rate": 5.706142672935858e-06,
+      "loss": 0.2166,
+      "step": 8472
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.584835228485723,
+      "learning_rate": 5.705250849661621e-06,
+      "loss": 0.1782,
+      "step": 8473
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.2876754537607817,
+      "learning_rate": 5.704359003494533e-06,
+      "loss": 0.2137,
+      "step": 8474
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.2023397923999912,
+      "learning_rate": 5.703467134463547e-06,
+      "loss": 0.1686,
+      "step": 8475
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.9934121031273285,
+      "learning_rate": 5.702575242597616e-06,
+      "loss": 0.2205,
+      "step": 8476
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.9611210143852176,
+      "learning_rate": 5.701683327925688e-06,
+      "loss": 0.1901,
+      "step": 8477
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.0572184571005327,
+      "learning_rate": 5.7007913904767135e-06,
+      "loss": 0.2217,
+      "step": 8478
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.9985703926118328,
+      "learning_rate": 5.69989943027965e-06,
+      "loss": 0.1973,
+      "step": 8479
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.3071176487833693,
+      "learning_rate": 5.699007447363447e-06,
+      "loss": 0.1556,
+      "step": 8480
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.1770132935062132,
+      "learning_rate": 5.6981154417570595e-06,
+      "loss": 0.2121,
+      "step": 8481
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.004411827053463,
+      "learning_rate": 5.697223413489446e-06,
+      "loss": 0.1591,
+      "step": 8482
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.2522749405374654,
+      "learning_rate": 5.696331362589557e-06,
+      "loss": 0.2349,
+      "step": 8483
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.9787753287810057,
+      "learning_rate": 5.695439289086351e-06,
+      "loss": 0.2167,
+      "step": 8484
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.6648026618969248,
+      "learning_rate": 5.694547193008788e-06,
+      "loss": 0.1345,
+      "step": 8485
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.7616809639973094,
+      "learning_rate": 5.693655074385823e-06,
+      "loss": 0.131,
+      "step": 8486
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.3366066263619496,
+      "learning_rate": 5.692762933246416e-06,
+      "loss": 0.183,
+      "step": 8487
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.4237034066708847,
+      "learning_rate": 5.691870769619525e-06,
+      "loss": 0.2381,
+      "step": 8488
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.2256256948154938,
+      "learning_rate": 5.690978583534111e-06,
+      "loss": 0.172,
+      "step": 8489
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.2183348272552976,
+      "learning_rate": 5.690086375019135e-06,
+      "loss": 0.1729,
+      "step": 8490
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.3056329425307227,
+      "learning_rate": 5.689194144103558e-06,
+      "loss": 0.1903,
+      "step": 8491
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.5470374075086346,
+      "learning_rate": 5.688301890816343e-06,
+      "loss": 0.2226,
+      "step": 8492
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.3168061778040663,
+      "learning_rate": 5.687409615186452e-06,
+      "loss": 0.1905,
+      "step": 8493
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.481904252341582,
+      "learning_rate": 5.6865173172428525e-06,
+      "loss": 0.2338,
+      "step": 8494
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.3898578199218,
+      "learning_rate": 5.685624997014504e-06,
+      "loss": 0.217,
+      "step": 8495
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.5608695774364516,
+      "learning_rate": 5.684732654530374e-06,
+      "loss": 0.2206,
+      "step": 8496
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.0253740618140041,
+      "learning_rate": 5.683840289819428e-06,
+      "loss": 0.2206,
+      "step": 8497
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.2981922679811453,
+      "learning_rate": 5.682947902910634e-06,
+      "loss": 0.2044,
+      "step": 8498
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.779966661698247,
+      "learning_rate": 5.6820554938329565e-06,
+      "loss": 0.2615,
+      "step": 8499
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.1574536708310026,
+      "learning_rate": 5.681163062615368e-06,
+      "loss": 0.2299,
+      "step": 8500
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.0091109726954028,
+      "learning_rate": 5.680270609286832e-06,
+      "loss": 0.1474,
+      "step": 8501
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.6071522186400227,
+      "learning_rate": 5.679378133876322e-06,
+      "loss": 0.2199,
+      "step": 8502
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.5085531673770942,
+      "learning_rate": 5.678485636412807e-06,
+      "loss": 0.2051,
+      "step": 8503
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.9468470958712256,
+      "learning_rate": 5.6775931169252595e-06,
+      "loss": 0.2375,
+      "step": 8504
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.350913308077239,
+      "learning_rate": 5.6767005754426495e-06,
+      "loss": 0.2828,
+      "step": 8505
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.9541504748798113,
+      "learning_rate": 5.6758080119939486e-06,
+      "loss": 0.1864,
+      "step": 8506
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.2405443780296033,
+      "learning_rate": 5.674915426608132e-06,
+      "loss": 0.1925,
+      "step": 8507
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.2290447078292857,
+      "learning_rate": 5.674022819314171e-06,
+      "loss": 0.1941,
+      "step": 8508
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.972494583022861,
+      "learning_rate": 5.673130190141043e-06,
+      "loss": 0.2247,
+      "step": 8509
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.1243187916509285,
+      "learning_rate": 5.67223753911772e-06,
+      "loss": 0.2172,
+      "step": 8510
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.3090747575511568,
+      "learning_rate": 5.671344866273181e-06,
+      "loss": 0.1584,
+      "step": 8511
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.2960306026902402,
+      "learning_rate": 5.6704521716364025e-06,
+      "loss": 0.1961,
+      "step": 8512
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.9115635353637638,
+      "learning_rate": 5.669559455236361e-06,
+      "loss": 0.1605,
+      "step": 8513
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.2386357058847661,
+      "learning_rate": 5.668666717102035e-06,
+      "loss": 0.2086,
+      "step": 8514
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.272955444870894,
+      "learning_rate": 5.667773957262401e-06,
+      "loss": 0.1989,
+      "step": 8515
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.9157985031384991,
+      "learning_rate": 5.666881175746442e-06,
+      "loss": 0.1849,
+      "step": 8516
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.4247498060485018,
+      "learning_rate": 5.665988372583134e-06,
+      "loss": 0.1946,
+      "step": 8517
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.9965050203584018,
+      "learning_rate": 5.665095547801464e-06,
+      "loss": 0.1701,
+      "step": 8518
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 2.392210621538303,
+      "learning_rate": 5.6642027014304065e-06,
+      "loss": 0.1656,
+      "step": 8519
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.3390703944415314,
+      "learning_rate": 5.663309833498949e-06,
+      "loss": 0.1926,
+      "step": 8520
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.3526532025152518,
+      "learning_rate": 5.662416944036073e-06,
+      "loss": 0.208,
+      "step": 8521
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.3579323117651336,
+      "learning_rate": 5.661524033070763e-06,
+      "loss": 0.1561,
+      "step": 8522
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.4283285170227797,
+      "learning_rate": 5.660631100632003e-06,
+      "loss": 0.1881,
+      "step": 8523
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.1359362678311262,
+      "learning_rate": 5.659738146748776e-06,
+      "loss": 0.182,
+      "step": 8524
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.9935485654431975,
+      "learning_rate": 5.658845171450069e-06,
+      "loss": 0.1813,
+      "step": 8525
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.9861781317847763,
+      "learning_rate": 5.657952174764869e-06,
+      "loss": 0.165,
+      "step": 8526
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.3982086160825542,
+      "learning_rate": 5.657059156722164e-06,
+      "loss": 0.1795,
+      "step": 8527
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.9750240223627307,
+      "learning_rate": 5.656166117350939e-06,
+      "loss": 0.2152,
+      "step": 8528
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.0506142249539134,
+      "learning_rate": 5.6552730566801865e-06,
+      "loss": 0.245,
+      "step": 8529
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.9260090344505943,
+      "learning_rate": 5.654379974738893e-06,
+      "loss": 0.1859,
+      "step": 8530
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.0550295769525184,
+      "learning_rate": 5.65348687155605e-06,
+      "loss": 0.1994,
+      "step": 8531
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.3302794946833112,
+      "learning_rate": 5.652593747160648e-06,
+      "loss": 0.2244,
+      "step": 8532
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.2649927708877222,
+      "learning_rate": 5.651700601581676e-06,
+      "loss": 0.1772,
+      "step": 8533
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.4053503562866296,
+      "learning_rate": 5.650807434848128e-06,
+      "loss": 0.182,
+      "step": 8534
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.283269148917826,
+      "learning_rate": 5.649914246988997e-06,
+      "loss": 0.2209,
+      "step": 8535
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.8918069099533384,
+      "learning_rate": 5.649021038033276e-06,
+      "loss": 0.188,
+      "step": 8536
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.0559186160946772,
+      "learning_rate": 5.648127808009956e-06,
+      "loss": 0.1838,
+      "step": 8537
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.1841745635382195,
+      "learning_rate": 5.647234556948037e-06,
+      "loss": 0.1835,
+      "step": 8538
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.9778014411130219,
+      "learning_rate": 5.646341284876513e-06,
+      "loss": 0.1711,
+      "step": 8539
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.9716590112374497,
+      "learning_rate": 5.6454479918243785e-06,
+      "loss": 0.1883,
+      "step": 8540
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.881652082031607,
+      "learning_rate": 5.644554677820632e-06,
+      "loss": 0.1671,
+      "step": 8541
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.0028102194438286,
+      "learning_rate": 5.643661342894267e-06,
+      "loss": 0.214,
+      "step": 8542
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.36696992232766,
+      "learning_rate": 5.642767987074288e-06,
+      "loss": 0.2486,
+      "step": 8543
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.1288201256341261,
+      "learning_rate": 5.641874610389688e-06,
+      "loss": 0.1658,
+      "step": 8544
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.1865483528629537,
+      "learning_rate": 5.640981212869471e-06,
+      "loss": 0.1665,
+      "step": 8545
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.1900309003471474,
+      "learning_rate": 5.640087794542634e-06,
+      "loss": 0.1979,
+      "step": 8546
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.2890976351949248,
+      "learning_rate": 5.639194355438178e-06,
+      "loss": 0.174,
+      "step": 8547
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.9599381315097114,
+      "learning_rate": 5.638300895585109e-06,
+      "loss": 0.1992,
+      "step": 8548
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.0561206123287095,
+      "learning_rate": 5.637407415012423e-06,
+      "loss": 0.2162,
+      "step": 8549
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.9743499009109431,
+      "learning_rate": 5.636513913749128e-06,
+      "loss": 0.1937,
+      "step": 8550
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.2883228636498614,
+      "learning_rate": 5.635620391824224e-06,
+      "loss": 0.1918,
+      "step": 8551
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.1768061457794734,
+      "learning_rate": 5.634726849266716e-06,
+      "loss": 0.1898,
+      "step": 8552
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.4336568886056407,
+      "learning_rate": 5.63383328610561e-06,
+      "loss": 0.1613,
+      "step": 8553
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.0584068555650248,
+      "learning_rate": 5.632939702369912e-06,
+      "loss": 0.1709,
+      "step": 8554
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.0561346474980182,
+      "learning_rate": 5.632046098088625e-06,
+      "loss": 0.1802,
+      "step": 8555
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.1663903212625413,
+      "learning_rate": 5.6311524732907596e-06,
+      "loss": 0.2408,
+      "step": 8556
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.0875265344309413,
+      "learning_rate": 5.630258828005323e-06,
+      "loss": 0.1931,
+      "step": 8557
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.9733406324853205,
+      "learning_rate": 5.629365162261321e-06,
+      "loss": 0.203,
+      "step": 8558
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.151925240506119,
+      "learning_rate": 5.628471476087766e-06,
+      "loss": 0.1788,
+      "step": 8559
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.42348977112294,
+      "learning_rate": 5.627577769513662e-06,
+      "loss": 0.2045,
+      "step": 8560
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.1766985264956127,
+      "learning_rate": 5.6266840425680256e-06,
+      "loss": 0.2099,
+      "step": 8561
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.3047816077002024,
+      "learning_rate": 5.6257902952798635e-06,
+      "loss": 0.245,
+      "step": 8562
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.8965830634322778,
+      "learning_rate": 5.6248965276781896e-06,
+      "loss": 0.2194,
+      "step": 8563
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.2279073076376374,
+      "learning_rate": 5.624002739792014e-06,
+      "loss": 0.1586,
+      "step": 8564
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.341891216032179,
+      "learning_rate": 5.623108931650349e-06,
+      "loss": 0.1687,
+      "step": 8565
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.1688815699433386,
+      "learning_rate": 5.622215103282212e-06,
+      "loss": 0.2239,
+      "step": 8566
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.068305348071078,
+      "learning_rate": 5.621321254716616e-06,
+      "loss": 0.2745,
+      "step": 8567
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.03831338409343,
+      "learning_rate": 5.6204273859825745e-06,
+      "loss": 0.2333,
+      "step": 8568
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.9016692047177435,
+      "learning_rate": 5.619533497109102e-06,
+      "loss": 0.1714,
+      "step": 8569
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.2809469864906502,
+      "learning_rate": 5.618639588125217e-06,
+      "loss": 0.1921,
+      "step": 8570
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.1142844348237926,
+      "learning_rate": 5.617745659059934e-06,
+      "loss": 0.2287,
+      "step": 8571
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.2814542066849473,
+      "learning_rate": 5.616851709942271e-06,
+      "loss": 0.1887,
+      "step": 8572
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.191992332574747,
+      "learning_rate": 5.615957740801248e-06,
+      "loss": 0.1824,
+      "step": 8573
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.2345763144930335,
+      "learning_rate": 5.6150637516658815e-06,
+      "loss": 0.2372,
+      "step": 8574
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.149115344555347,
+      "learning_rate": 5.614169742565193e-06,
+      "loss": 0.2434,
+      "step": 8575
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.4071898776586422,
+      "learning_rate": 5.6132757135282025e-06,
+      "loss": 0.1869,
+      "step": 8576
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.0934502684798093,
+      "learning_rate": 5.6123816645839295e-06,
+      "loss": 0.2072,
+      "step": 8577
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.297524742585824,
+      "learning_rate": 5.611487595761393e-06,
+      "loss": 0.2325,
+      "step": 8578
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.3573156548525007,
+      "learning_rate": 5.61059350708962e-06,
+      "loss": 0.1955,
+      "step": 8579
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.1230138226496582,
+      "learning_rate": 5.609699398597629e-06,
+      "loss": 0.1216,
+      "step": 8580
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.9949516686575434,
+      "learning_rate": 5.608805270314446e-06,
+      "loss": 0.2069,
+      "step": 8581
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.2692459064538413,
+      "learning_rate": 5.607911122269093e-06,
+      "loss": 0.186,
+      "step": 8582
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.0739263648149238,
+      "learning_rate": 5.607016954490595e-06,
+      "loss": 0.1762,
+      "step": 8583
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.0566977147916825,
+      "learning_rate": 5.606122767007979e-06,
+      "loss": 0.2162,
+      "step": 8584
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.4228338110869605,
+      "learning_rate": 5.6052285598502695e-06,
+      "loss": 0.2124,
+      "step": 8585
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.9688810079190465,
+      "learning_rate": 5.6043343330464926e-06,
+      "loss": 0.1936,
+      "step": 8586
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.209837964992593,
+      "learning_rate": 5.603440086625676e-06,
+      "loss": 0.2232,
+      "step": 8587
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.3561847950163741,
+      "learning_rate": 5.602545820616846e-06,
+      "loss": 0.2081,
+      "step": 8588
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.1529713352221946,
+      "learning_rate": 5.601651535049032e-06,
+      "loss": 0.2265,
+      "step": 8589
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.3885090621168987,
+      "learning_rate": 5.600757229951263e-06,
+      "loss": 0.2088,
+      "step": 8590
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.8000633097399414,
+      "learning_rate": 5.599862905352571e-06,
+      "loss": 0.2323,
+      "step": 8591
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.3019666396938085,
+      "learning_rate": 5.598968561281982e-06,
+      "loss": 0.188,
+      "step": 8592
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.8146043655338062,
+      "learning_rate": 5.598074197768532e-06,
+      "loss": 0.186,
+      "step": 8593
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.3749082003640152,
+      "learning_rate": 5.597179814841247e-06,
+      "loss": 0.2306,
+      "step": 8594
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.2564857338834716,
+      "learning_rate": 5.596285412529164e-06,
+      "loss": 0.1847,
+      "step": 8595
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.1233527156845484,
+      "learning_rate": 5.595390990861311e-06,
+      "loss": 0.1865,
+      "step": 8596
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.1050734352317049,
+      "learning_rate": 5.594496549866726e-06,
+      "loss": 0.1496,
+      "step": 8597
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 2.4718420667901926,
+      "learning_rate": 5.5936020895744395e-06,
+      "loss": 0.1952,
+      "step": 8598
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.508933093162156,
+      "learning_rate": 5.592707610013488e-06,
+      "loss": 0.2201,
+      "step": 8599
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.2735460679749768,
+      "learning_rate": 5.591813111212907e-06,
+      "loss": 0.251,
+      "step": 8600
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.1205871414314803,
+      "learning_rate": 5.590918593201732e-06,
+      "loss": 0.1661,
+      "step": 8601
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.1196605430022875,
+      "learning_rate": 5.5900240560090004e-06,
+      "loss": 0.155,
+      "step": 8602
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.3538892871478643,
+      "learning_rate": 5.589129499663749e-06,
+      "loss": 0.1835,
+      "step": 8603
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.2107772721143892,
+      "learning_rate": 5.588234924195015e-06,
+      "loss": 0.2285,
+      "step": 8604
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.9178109134768164,
+      "learning_rate": 5.5873403296318365e-06,
+      "loss": 0.1942,
+      "step": 8605
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.4443958210734007,
+      "learning_rate": 5.586445716003254e-06,
+      "loss": 0.2274,
+      "step": 8606
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.2084090708216517,
+      "learning_rate": 5.585551083338304e-06,
+      "loss": 0.1788,
+      "step": 8607
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.5965745593817853,
+      "learning_rate": 5.5846564316660314e-06,
+      "loss": 0.1783,
+      "step": 8608
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.3808748441312608,
+      "learning_rate": 5.583761761015474e-06,
+      "loss": 0.1677,
+      "step": 8609
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.794219854172305,
+      "learning_rate": 5.582867071415674e-06,
+      "loss": 0.2594,
+      "step": 8610
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.130355521768373,
+      "learning_rate": 5.581972362895674e-06,
+      "loss": 0.2057,
+      "step": 8611
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.2329749152672707,
+      "learning_rate": 5.581077635484517e-06,
+      "loss": 0.1661,
+      "step": 8612
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.2191042675572685,
+      "learning_rate": 5.580182889211246e-06,
+      "loss": 0.1741,
+      "step": 8613
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.1037414218497967,
+      "learning_rate": 5.579288124104903e-06,
+      "loss": 0.2054,
+      "step": 8614
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.9615811154184886,
+      "learning_rate": 5.578393340194537e-06,
+      "loss": 0.1648,
+      "step": 8615
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.5174755665651876,
+      "learning_rate": 5.577498537509188e-06,
+      "loss": 0.22,
+      "step": 8616
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.015353467296181,
+      "learning_rate": 5.5766037160779055e-06,
+      "loss": 0.1855,
+      "step": 8617
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.9976131556858997,
+      "learning_rate": 5.575708875929734e-06,
+      "loss": 0.1781,
+      "step": 8618
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.4648329772553095,
+      "learning_rate": 5.5748140170937215e-06,
+      "loss": 0.1781,
+      "step": 8619
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.3817554895282638,
+      "learning_rate": 5.573919139598917e-06,
+      "loss": 0.2504,
+      "step": 8620
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.1192031762981949,
+      "learning_rate": 5.573024243474366e-06,
+      "loss": 0.1895,
+      "step": 8621
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.9327425727640551,
+      "learning_rate": 5.572129328749119e-06,
+      "loss": 0.1518,
+      "step": 8622
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.1144816017590757,
+      "learning_rate": 5.5712343954522254e-06,
+      "loss": 0.2713,
+      "step": 8623
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.2518794816208798,
+      "learning_rate": 5.570339443612734e-06,
+      "loss": 0.1643,
+      "step": 8624
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.0300373909816005,
+      "learning_rate": 5.569444473259696e-06,
+      "loss": 0.1715,
+      "step": 8625
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.2869877041397553,
+      "learning_rate": 5.568549484422163e-06,
+      "loss": 0.1737,
+      "step": 8626
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.118052544509132,
+      "learning_rate": 5.567654477129188e-06,
+      "loss": 0.1392,
+      "step": 8627
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.0032849652949767,
+      "learning_rate": 5.56675945140982e-06,
+      "loss": 0.1816,
+      "step": 8628
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.5555856928029999,
+      "learning_rate": 5.565864407293117e-06,
+      "loss": 0.2492,
+      "step": 8629
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.3045151648087745,
+      "learning_rate": 5.5649693448081286e-06,
+      "loss": 0.2422,
+      "step": 8630
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.081146518904182,
+      "learning_rate": 5.564074263983912e-06,
+      "loss": 0.1958,
+      "step": 8631
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.9223506597460107,
+      "learning_rate": 5.563179164849518e-06,
+      "loss": 0.1816,
+      "step": 8632
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.2711569222318757,
+      "learning_rate": 5.562284047434006e-06,
+      "loss": 0.2601,
+      "step": 8633
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.518055755980326,
+      "learning_rate": 5.561388911766429e-06,
+      "loss": 0.1758,
+      "step": 8634
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.902407783242729,
+      "learning_rate": 5.560493757875845e-06,
+      "loss": 0.1494,
+      "step": 8635
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.6064071882646098,
+      "learning_rate": 5.559598585791313e-06,
+      "loss": 0.2437,
+      "step": 8636
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.0680985245623744,
+      "learning_rate": 5.558703395541888e-06,
+      "loss": 0.1723,
+      "step": 8637
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.3587947736222281,
+      "learning_rate": 5.5578081871566305e-06,
+      "loss": 0.2108,
+      "step": 8638
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.6994204258994174,
+      "learning_rate": 5.556912960664598e-06,
+      "loss": 0.198,
+      "step": 8639
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.2816774916992648,
+      "learning_rate": 5.556017716094852e-06,
+      "loss": 0.1918,
+      "step": 8640
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.307825042067426,
+      "learning_rate": 5.555122453476448e-06,
+      "loss": 0.2132,
+      "step": 8641
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.1624688113551054,
+      "learning_rate": 5.5542271728384525e-06,
+      "loss": 0.2229,
+      "step": 8642
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.2483944341777444,
+      "learning_rate": 5.553331874209922e-06,
+      "loss": 0.2168,
+      "step": 8643
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.230771787354503,
+      "learning_rate": 5.552436557619921e-06,
+      "loss": 0.2217,
+      "step": 8644
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.1671648831455093,
+      "learning_rate": 5.551541223097514e-06,
+      "loss": 0.1725,
+      "step": 8645
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.253233113940456,
+      "learning_rate": 5.550645870671759e-06,
+      "loss": 0.2846,
+      "step": 8646
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.9633780226628486,
+      "learning_rate": 5.549750500371725e-06,
+      "loss": 0.1549,
+      "step": 8647
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.9878675571250807,
+      "learning_rate": 5.548855112226473e-06,
+      "loss": 0.1819,
+      "step": 8648
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.1249772440410242,
+      "learning_rate": 5.547959706265068e-06,
+      "loss": 0.1884,
+      "step": 8649
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.869067844815435,
+      "learning_rate": 5.547064282516575e-06,
+      "loss": 0.1147,
+      "step": 8650
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.3353479880507866,
+      "learning_rate": 5.546168841010062e-06,
+      "loss": 0.2045,
+      "step": 8651
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.059273378454977,
+      "learning_rate": 5.545273381774594e-06,
+      "loss": 0.2165,
+      "step": 8652
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.4669488657950318,
+      "learning_rate": 5.544377904839238e-06,
+      "loss": 0.2653,
+      "step": 8653
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.6853445601353767,
+      "learning_rate": 5.543482410233062e-06,
+      "loss": 0.2122,
+      "step": 8654
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.072006566033565,
+      "learning_rate": 5.542586897985135e-06,
+      "loss": 0.1862,
+      "step": 8655
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.3874230329112511,
+      "learning_rate": 5.5416913681245255e-06,
+      "loss": 0.1708,
+      "step": 8656
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.336816810654108,
+      "learning_rate": 5.540795820680302e-06,
+      "loss": 0.2022,
+      "step": 8657
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.299840460900484,
+      "learning_rate": 5.539900255681537e-06,
+      "loss": 0.2264,
+      "step": 8658
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.8463338538324319,
+      "learning_rate": 5.5390046731572976e-06,
+      "loss": 0.1307,
+      "step": 8659
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.8791352143337692,
+      "learning_rate": 5.5381090731366575e-06,
+      "loss": 0.1649,
+      "step": 8660
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.0870870947739246,
+      "learning_rate": 5.537213455648685e-06,
+      "loss": 0.2564,
+      "step": 8661
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.4172609246589594,
+      "learning_rate": 5.536317820722457e-06,
+      "loss": 0.212,
+      "step": 8662
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.5263211468509108,
+      "learning_rate": 5.535422168387044e-06,
+      "loss": 0.1809,
+      "step": 8663
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.9808440724906886,
+      "learning_rate": 5.534526498671519e-06,
+      "loss": 0.1664,
+      "step": 8664
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 2.030986744987564,
+      "learning_rate": 5.533630811604957e-06,
+      "loss": 0.2371,
+      "step": 8665
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.1940123586801803,
+      "learning_rate": 5.532735107216431e-06,
+      "loss": 0.2328,
+      "step": 8666
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.6627467599744579,
+      "learning_rate": 5.531839385535019e-06,
+      "loss": 0.2301,
+      "step": 8667
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.374218707927727,
+      "learning_rate": 5.530943646589792e-06,
+      "loss": 0.1999,
+      "step": 8668
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.1901271379272527,
+      "learning_rate": 5.530047890409832e-06,
+      "loss": 0.2455,
+      "step": 8669
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.0025306212298197,
+      "learning_rate": 5.52915211702421e-06,
+      "loss": 0.2163,
+      "step": 8670
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.2584917562301505,
+      "learning_rate": 5.528256326462007e-06,
+      "loss": 0.2053,
+      "step": 8671
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.1436512903836182,
+      "learning_rate": 5.527360518752301e-06,
+      "loss": 0.1815,
+      "step": 8672
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.1099739137045754,
+      "learning_rate": 5.526464693924167e-06,
+      "loss": 0.2136,
+      "step": 8673
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.1650989673750305,
+      "learning_rate": 5.525568852006688e-06,
+      "loss": 0.2484,
+      "step": 8674
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.4056030228912213,
+      "learning_rate": 5.524672993028942e-06,
+      "loss": 0.2134,
+      "step": 8675
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.344784891197408,
+      "learning_rate": 5.523777117020008e-06,
+      "loss": 0.1968,
+      "step": 8676
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.3540170299555712,
+      "learning_rate": 5.522881224008968e-06,
+      "loss": 0.1969,
+      "step": 8677
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.172730232334628,
+      "learning_rate": 5.521985314024903e-06,
+      "loss": 0.2019,
+      "step": 8678
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.0063477913654204,
+      "learning_rate": 5.521089387096894e-06,
+      "loss": 0.1701,
+      "step": 8679
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.3386258951978143,
+      "learning_rate": 5.520193443254024e-06,
+      "loss": 0.2137,
+      "step": 8680
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.289337961068933,
+      "learning_rate": 5.519297482525375e-06,
+      "loss": 0.1702,
+      "step": 8681
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.8834854289356008,
+      "learning_rate": 5.518401504940032e-06,
+      "loss": 0.1861,
+      "step": 8682
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.9678253502496135,
+      "learning_rate": 5.517505510527079e-06,
+      "loss": 0.1393,
+      "step": 8683
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.3492714835697703,
+      "learning_rate": 5.516609499315598e-06,
+      "loss": 0.1762,
+      "step": 8684
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.9496291611625346,
+      "learning_rate": 5.515713471334677e-06,
+      "loss": 0.1535,
+      "step": 8685
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.3306412086108819,
+      "learning_rate": 5.514817426613399e-06,
+      "loss": 0.2275,
+      "step": 8686
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.9764680187296951,
+      "learning_rate": 5.513921365180852e-06,
+      "loss": 0.1983,
+      "step": 8687
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.7713360535214058,
+      "learning_rate": 5.513025287066121e-06,
+      "loss": 0.1644,
+      "step": 8688
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.0702196706591116,
+      "learning_rate": 5.512129192298293e-06,
+      "loss": 0.1985,
+      "step": 8689
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.0848193896070961,
+      "learning_rate": 5.5112330809064595e-06,
+      "loss": 0.1646,
+      "step": 8690
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.913423804129242,
+      "learning_rate": 5.510336952919705e-06,
+      "loss": 0.191,
+      "step": 8691
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.120531145344605,
+      "learning_rate": 5.50944080836712e-06,
+      "loss": 0.1905,
+      "step": 8692
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.1514571901302737,
+      "learning_rate": 5.5085446472777935e-06,
+      "loss": 0.1861,
+      "step": 8693
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.5073594094466711,
+      "learning_rate": 5.507648469680816e-06,
+      "loss": 0.2294,
+      "step": 8694
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.6564121459368057,
+      "learning_rate": 5.506752275605276e-06,
+      "loss": 0.2323,
+      "step": 8695
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.2833652190741092,
+      "learning_rate": 5.505856065080266e-06,
+      "loss": 0.1834,
+      "step": 8696
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.242176679648051,
+      "learning_rate": 5.5049598381348765e-06,
+      "loss": 0.2155,
+      "step": 8697
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.0768169401372756,
+      "learning_rate": 5.504063594798201e-06,
+      "loss": 0.2225,
+      "step": 8698
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.1971626175196752,
+      "learning_rate": 5.503167335099332e-06,
+      "loss": 0.2028,
+      "step": 8699
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.1075695114355104,
+      "learning_rate": 5.502271059067361e-06,
+      "loss": 0.1985,
+      "step": 8700
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.765940758004841,
+      "learning_rate": 5.501374766731383e-06,
+      "loss": 0.1267,
+      "step": 8701
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.0230108729608065,
+      "learning_rate": 5.500478458120493e-06,
+      "loss": 0.1828,
+      "step": 8702
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.0101953750403232,
+      "learning_rate": 5.499582133263783e-06,
+      "loss": 0.1732,
+      "step": 8703
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.9321845104484179,
+      "learning_rate": 5.498685792190349e-06,
+      "loss": 0.2151,
+      "step": 8704
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.2675032071866799,
+      "learning_rate": 5.49778943492929e-06,
+      "loss": 0.2186,
+      "step": 8705
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.1317906407663605,
+      "learning_rate": 5.496893061509697e-06,
+      "loss": 0.2128,
+      "step": 8706
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.5856691612646772,
+      "learning_rate": 5.495996671960669e-06,
+      "loss": 0.1579,
+      "step": 8707
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.3732624345984181,
+      "learning_rate": 5.495100266311306e-06,
+      "loss": 0.2132,
+      "step": 8708
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.1934677486876444,
+      "learning_rate": 5.494203844590703e-06,
+      "loss": 0.1622,
+      "step": 8709
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.1714983302980428,
+      "learning_rate": 5.493307406827959e-06,
+      "loss": 0.1899,
+      "step": 8710
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.425272500706982,
+      "learning_rate": 5.492410953052174e-06,
+      "loss": 0.2268,
+      "step": 8711
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.0355033868258139,
+      "learning_rate": 5.491514483292445e-06,
+      "loss": 0.1926,
+      "step": 8712
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.9384430016938481,
+      "learning_rate": 5.4906179975778725e-06,
+      "loss": 0.2071,
+      "step": 8713
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.3091179039951408,
+      "learning_rate": 5.489721495937559e-06,
+      "loss": 0.1944,
+      "step": 8714
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.2610480235277943,
+      "learning_rate": 5.488824978400604e-06,
+      "loss": 0.2176,
+      "step": 8715
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.579341246250272,
+      "learning_rate": 5.487928444996108e-06,
+      "loss": 0.2734,
+      "step": 8716
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.3437848890565829,
+      "learning_rate": 5.487031895753175e-06,
+      "loss": 0.1664,
+      "step": 8717
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.0328163259438248,
+      "learning_rate": 5.486135330700906e-06,
+      "loss": 0.2004,
+      "step": 8718
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.3933847346456452,
+      "learning_rate": 5.4852387498684056e-06,
+      "loss": 0.1339,
+      "step": 8719
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.0551677652360796,
+      "learning_rate": 5.484342153284776e-06,
+      "loss": 0.1477,
+      "step": 8720
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.1157647924167031,
+      "learning_rate": 5.483445540979123e-06,
+      "loss": 0.2165,
+      "step": 8721
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.1141352673877831,
+      "learning_rate": 5.482548912980547e-06,
+      "loss": 0.2697,
+      "step": 8722
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.029395321331005,
+      "learning_rate": 5.4816522693181565e-06,
+      "loss": 0.172,
+      "step": 8723
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.2555253168939529,
+      "learning_rate": 5.480755610021058e-06,
+      "loss": 0.2054,
+      "step": 8724
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.0204791434609612,
+      "learning_rate": 5.479858935118354e-06,
+      "loss": 0.2044,
+      "step": 8725
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.349821994501084,
+      "learning_rate": 5.478962244639154e-06,
+      "loss": 0.1755,
+      "step": 8726
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.1999424667275478,
+      "learning_rate": 5.4780655386125635e-06,
+      "loss": 0.1801,
+      "step": 8727
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.4588509566292067,
+      "learning_rate": 5.477168817067692e-06,
+      "loss": 0.2039,
+      "step": 8728
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.1955696340504909,
+      "learning_rate": 5.476272080033646e-06,
+      "loss": 0.2175,
+      "step": 8729
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.698468890202735,
+      "learning_rate": 5.475375327539534e-06,
+      "loss": 0.2017,
+      "step": 8730
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.2167690298406006,
+      "learning_rate": 5.4744785596144645e-06,
+      "loss": 0.2187,
+      "step": 8731
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.0398655545207736,
+      "learning_rate": 5.473581776287549e-06,
+      "loss": 0.1938,
+      "step": 8732
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.250411722803093,
+      "learning_rate": 5.472684977587897e-06,
+      "loss": 0.2397,
+      "step": 8733
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.1204754448548242,
+      "learning_rate": 5.471788163544618e-06,
+      "loss": 0.1598,
+      "step": 8734
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.2313109497775823,
+      "learning_rate": 5.4708913341868255e-06,
+      "loss": 0.2065,
+      "step": 8735
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.9775672093015743,
+      "learning_rate": 5.469994489543628e-06,
+      "loss": 0.2211,
+      "step": 8736
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.2276165785247675,
+      "learning_rate": 5.4690976296441385e-06,
+      "loss": 0.1741,
+      "step": 8737
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.8454908000920245,
+      "learning_rate": 5.468200754517472e-06,
+      "loss": 0.1571,
+      "step": 8738
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.0081109263057957,
+      "learning_rate": 5.46730386419274e-06,
+      "loss": 0.1785,
+      "step": 8739
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.1877256589982605,
+      "learning_rate": 5.4664069586990545e-06,
+      "loss": 0.2271,
+      "step": 8740
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.3348350884166478,
+      "learning_rate": 5.4655100380655305e-06,
+      "loss": 0.1947,
+      "step": 8741
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.0579973367134035,
+      "learning_rate": 5.464613102321284e-06,
+      "loss": 0.1851,
+      "step": 8742
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.0853929732452496,
+      "learning_rate": 5.4637161514954284e-06,
+      "loss": 0.1937,
+      "step": 8743
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.581925960851807,
+      "learning_rate": 5.4628191856170805e-06,
+      "loss": 0.239,
+      "step": 8744
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.322869610152907,
+      "learning_rate": 5.461922204715356e-06,
+      "loss": 0.2323,
+      "step": 8745
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.380989009270311,
+      "learning_rate": 5.461025208819371e-06,
+      "loss": 0.2055,
+      "step": 8746
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.2436471376583436,
+      "learning_rate": 5.460128197958242e-06,
+      "loss": 0.2158,
+      "step": 8747
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.0568932532703048,
+      "learning_rate": 5.459231172161088e-06,
+      "loss": 0.1937,
+      "step": 8748
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.1877941375904815,
+      "learning_rate": 5.4583341314570236e-06,
+      "loss": 0.1607,
+      "step": 8749
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.0551181973442654,
+      "learning_rate": 5.457437075875172e-06,
+      "loss": 0.2293,
+      "step": 8750
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.1380542663650532,
+      "learning_rate": 5.4565400054446495e-06,
+      "loss": 0.206,
+      "step": 8751
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.216554450719619,
+      "learning_rate": 5.455642920194575e-06,
+      "loss": 0.2046,
+      "step": 8752
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.6509581453475548,
+      "learning_rate": 5.4547458201540705e-06,
+      "loss": 0.1862,
+      "step": 8753
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.0325106804505442,
+      "learning_rate": 5.453848705352255e-06,
+      "loss": 0.1527,
+      "step": 8754
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.0390841797309396,
+      "learning_rate": 5.45295157581825e-06,
+      "loss": 0.17,
+      "step": 8755
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.1846031497325573,
+      "learning_rate": 5.452054431581175e-06,
+      "loss": 0.2347,
+      "step": 8756
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.9036709347279174,
+      "learning_rate": 5.451157272670154e-06,
+      "loss": 0.2009,
+      "step": 8757
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.6129572429835928,
+      "learning_rate": 5.450260099114308e-06,
+      "loss": 0.2016,
+      "step": 8758
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.8895807879089161,
+      "learning_rate": 5.449362910942761e-06,
+      "loss": 0.1589,
+      "step": 8759
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.0133910387306753,
+      "learning_rate": 5.448465708184635e-06,
+      "loss": 0.2097,
+      "step": 8760
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.19868946743397,
+      "learning_rate": 5.447568490869054e-06,
+      "loss": 0.2832,
+      "step": 8761
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.0590939088034905,
+      "learning_rate": 5.446671259025142e-06,
+      "loss": 0.2401,
+      "step": 8762
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.2205898415453031,
+      "learning_rate": 5.445774012682026e-06,
+      "loss": 0.2008,
+      "step": 8763
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.8110812447972253,
+      "learning_rate": 5.444876751868828e-06,
+      "loss": 0.2041,
+      "step": 8764
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.1747995885419662,
+      "learning_rate": 5.443979476614674e-06,
+      "loss": 0.1961,
+      "step": 8765
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.070280955190879,
+      "learning_rate": 5.443082186948693e-06,
+      "loss": 0.1804,
+      "step": 8766
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.3241748760348322,
+      "learning_rate": 5.442184882900008e-06,
+      "loss": 0.2143,
+      "step": 8767
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.9150682554605936,
+      "learning_rate": 5.441287564497747e-06,
+      "loss": 0.2254,
+      "step": 8768
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.049808829543348,
+      "learning_rate": 5.440390231771039e-06,
+      "loss": 0.2365,
+      "step": 8769
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.0983738182811909,
+      "learning_rate": 5.439492884749009e-06,
+      "loss": 0.2265,
+      "step": 8770
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.1339253155620805,
+      "learning_rate": 5.438595523460789e-06,
+      "loss": 0.1969,
+      "step": 8771
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.2536565875235453,
+      "learning_rate": 5.437698147935506e-06,
+      "loss": 0.2019,
+      "step": 8772
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.1082577788383086,
+      "learning_rate": 5.43680075820229e-06,
+      "loss": 0.1753,
+      "step": 8773
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.1636942450584356,
+      "learning_rate": 5.435903354290268e-06,
+      "loss": 0.164,
+      "step": 8774
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.9696382057834343,
+      "learning_rate": 5.435005936228575e-06,
+      "loss": 0.1789,
+      "step": 8775
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.2801234516073066,
+      "learning_rate": 5.434108504046337e-06,
+      "loss": 0.1633,
+      "step": 8776
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.5505361507450648,
+      "learning_rate": 5.4332110577726886e-06,
+      "loss": 0.2374,
+      "step": 8777
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.071300732558609,
+      "learning_rate": 5.43231359743676e-06,
+      "loss": 0.2171,
+      "step": 8778
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.4606582953637328,
+      "learning_rate": 5.431416123067682e-06,
+      "loss": 0.2134,
+      "step": 8779
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.1794031124872009,
+      "learning_rate": 5.43051863469459e-06,
+      "loss": 0.1952,
+      "step": 8780
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.6559040410019281,
+      "learning_rate": 5.4296211323466165e-06,
+      "loss": 0.142,
+      "step": 8781
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 2.8186144168158886,
+      "learning_rate": 5.428723616052893e-06,
+      "loss": 0.252,
+      "step": 8782
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.248147872336533,
+      "learning_rate": 5.4278260858425534e-06,
+      "loss": 0.2186,
+      "step": 8783
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.8604326978480055,
+      "learning_rate": 5.426928541744735e-06,
+      "loss": 0.1484,
+      "step": 8784
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.033030639774156,
+      "learning_rate": 5.426030983788569e-06,
+      "loss": 0.2071,
+      "step": 8785
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.158076134614253,
+      "learning_rate": 5.425133412003192e-06,
+      "loss": 0.1943,
+      "step": 8786
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.948515665982623,
+      "learning_rate": 5.424235826417741e-06,
+      "loss": 0.1793,
+      "step": 8787
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.3174759172323207,
+      "learning_rate": 5.42333822706135e-06,
+      "loss": 0.1876,
+      "step": 8788
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.2192519083663214,
+      "learning_rate": 5.422440613963158e-06,
+      "loss": 0.2205,
+      "step": 8789
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.9897796143405764,
+      "learning_rate": 5.4215429871523015e-06,
+      "loss": 0.2143,
+      "step": 8790
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.0533161360595793,
+      "learning_rate": 5.420645346657917e-06,
+      "loss": 0.2177,
+      "step": 8791
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.2971376963055907,
+      "learning_rate": 5.419747692509141e-06,
+      "loss": 0.2133,
+      "step": 8792
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.4644708235010286,
+      "learning_rate": 5.4188500247351135e-06,
+      "loss": 0.1696,
+      "step": 8793
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.8992358061086795,
+      "learning_rate": 5.417952343364974e-06,
+      "loss": 0.1881,
+      "step": 8794
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.9891833327217607,
+      "learning_rate": 5.417054648427861e-06,
+      "loss": 0.1685,
+      "step": 8795
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.0304605359320542,
+      "learning_rate": 5.416156939952914e-06,
+      "loss": 0.1662,
+      "step": 8796
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.027815442166593,
+      "learning_rate": 5.4152592179692744e-06,
+      "loss": 0.1578,
+      "step": 8797
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.9688573934455829,
+      "learning_rate": 5.414361482506082e-06,
+      "loss": 0.1631,
+      "step": 8798
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.345801539483496,
+      "learning_rate": 5.413463733592477e-06,
+      "loss": 0.1987,
+      "step": 8799
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.9578232154744637,
+      "learning_rate": 5.412565971257601e-06,
+      "loss": 0.1714,
+      "step": 8800
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.0453345152471045,
+      "learning_rate": 5.411668195530597e-06,
+      "loss": 0.186,
+      "step": 8801
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.4496773986254787,
+      "learning_rate": 5.4107704064406066e-06,
+      "loss": 0.2091,
+      "step": 8802
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.433525825731134,
+      "learning_rate": 5.409872604016772e-06,
+      "loss": 0.199,
+      "step": 8803
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.3798744703554977,
+      "learning_rate": 5.408974788288236e-06,
+      "loss": 0.1563,
+      "step": 8804
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.246257215199681,
+      "learning_rate": 5.4080769592841456e-06,
+      "loss": 0.1933,
+      "step": 8805
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.0184900769874874,
+      "learning_rate": 5.407179117033641e-06,
+      "loss": 0.158,
+      "step": 8806
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.3540667681953602,
+      "learning_rate": 5.406281261565869e-06,
+      "loss": 0.1829,
+      "step": 8807
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.0373120442156323,
+      "learning_rate": 5.405383392909973e-06,
+      "loss": 0.153,
+      "step": 8808
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.3299608500981983,
+      "learning_rate": 5.4044855110951e-06,
+      "loss": 0.2317,
+      "step": 8809
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.0323330942432234,
+      "learning_rate": 5.4035876161503934e-06,
+      "loss": 0.2209,
+      "step": 8810
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.4991254815276032,
+      "learning_rate": 5.402689708105001e-06,
+      "loss": 0.2445,
+      "step": 8811
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.7350118451614205,
+      "learning_rate": 5.401791786988068e-06,
+      "loss": 0.2506,
+      "step": 8812
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.4578569772401606,
+      "learning_rate": 5.400893852828743e-06,
+      "loss": 0.2118,
+      "step": 8813
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.6861104155756066,
+      "learning_rate": 5.399995905656174e-06,
+      "loss": 0.2405,
+      "step": 8814
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.2595838814866824,
+      "learning_rate": 5.399097945499507e-06,
+      "loss": 0.2091,
+      "step": 8815
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.8249625458939419,
+      "learning_rate": 5.398199972387893e-06,
+      "loss": 0.2139,
+      "step": 8816
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.344329825324646,
+      "learning_rate": 5.397301986350479e-06,
+      "loss": 0.1787,
+      "step": 8817
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.6101762526971133,
+      "learning_rate": 5.396403987416412e-06,
+      "loss": 0.2619,
+      "step": 8818
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.8402047744601973,
+      "learning_rate": 5.395505975614844e-06,
+      "loss": 0.295,
+      "step": 8819
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.217352001131376,
+      "learning_rate": 5.394607950974926e-06,
+      "loss": 0.2483,
+      "step": 8820
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.1776095287772308,
+      "learning_rate": 5.3937099135258054e-06,
+      "loss": 0.2084,
+      "step": 8821
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.466838244325907,
+      "learning_rate": 5.392811863296634e-06,
+      "loss": 0.2717,
+      "step": 8822
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.9354962313329936,
+      "learning_rate": 5.391913800316566e-06,
+      "loss": 0.1497,
+      "step": 8823
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.143823721501819,
+      "learning_rate": 5.391015724614749e-06,
+      "loss": 0.1609,
+      "step": 8824
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.3159576438344023,
+      "learning_rate": 5.3901176362203364e-06,
+      "loss": 0.1769,
+      "step": 8825
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.122414077277136,
+      "learning_rate": 5.389219535162483e-06,
+      "loss": 0.2535,
+      "step": 8826
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.8830775352581488,
+      "learning_rate": 5.38832142147034e-06,
+      "loss": 0.1621,
+      "step": 8827
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.3965512186338667,
+      "learning_rate": 5.387423295173058e-06,
+      "loss": 0.2206,
+      "step": 8828
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.9515837844591997,
+      "learning_rate": 5.386525156299792e-06,
+      "loss": 0.196,
+      "step": 8829
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.0808031018562758,
+      "learning_rate": 5.3856270048797e-06,
+      "loss": 0.2075,
+      "step": 8830
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.9069166259722823,
+      "learning_rate": 5.384728840941932e-06,
+      "loss": 0.2219,
+      "step": 8831
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.528220678622292,
+      "learning_rate": 5.383830664515646e-06,
+      "loss": 0.2368,
+      "step": 8832
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.9097317168033131,
+      "learning_rate": 5.382932475629995e-06,
+      "loss": 0.1692,
+      "step": 8833
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.31946457490567,
+      "learning_rate": 5.3820342743141365e-06,
+      "loss": 0.2241,
+      "step": 8834
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.5915085092536392,
+      "learning_rate": 5.381136060597225e-06,
+      "loss": 0.2562,
+      "step": 8835
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.1623920193123698,
+      "learning_rate": 5.380237834508418e-06,
+      "loss": 0.1893,
+      "step": 8836
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.8097308635841273,
+      "learning_rate": 5.379339596076872e-06,
+      "loss": 0.117,
+      "step": 8837
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.228695663699211,
+      "learning_rate": 5.378441345331744e-06,
+      "loss": 0.1678,
+      "step": 8838
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.26960272367727,
+      "learning_rate": 5.377543082302193e-06,
+      "loss": 0.1809,
+      "step": 8839
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.4265560187603061,
+      "learning_rate": 5.376644807017377e-06,
+      "loss": 0.1804,
+      "step": 8840
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.473480396212672,
+      "learning_rate": 5.375746519506454e-06,
+      "loss": 0.223,
+      "step": 8841
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.4235127090513002,
+      "learning_rate": 5.374848219798582e-06,
+      "loss": 0.2058,
+      "step": 8842
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.333917829772279,
+      "learning_rate": 5.373949907922922e-06,
+      "loss": 0.2103,
+      "step": 8843
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.8376418369023083,
+      "learning_rate": 5.373051583908633e-06,
+      "loss": 0.1586,
+      "step": 8844
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.5962785899477923,
+      "learning_rate": 5.372153247784875e-06,
+      "loss": 0.1858,
+      "step": 8845
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.1019544140991764,
+      "learning_rate": 5.371254899580808e-06,
+      "loss": 0.1603,
+      "step": 8846
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.5599923324243692,
+      "learning_rate": 5.370356539325593e-06,
+      "loss": 0.2271,
+      "step": 8847
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.0116105371582291,
+      "learning_rate": 5.369458167048392e-06,
+      "loss": 0.1903,
+      "step": 8848
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.3580828697105645,
+      "learning_rate": 5.368559782778366e-06,
+      "loss": 0.1998,
+      "step": 8849
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.9219696756013107,
+      "learning_rate": 5.36766138654468e-06,
+      "loss": 0.1889,
+      "step": 8850
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.1383966142661874,
+      "learning_rate": 5.366762978376492e-06,
+      "loss": 0.2082,
+      "step": 8851
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.246795063814652,
+      "learning_rate": 5.365864558302968e-06,
+      "loss": 0.1795,
+      "step": 8852
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.8964931304770039,
+      "learning_rate": 5.364966126353269e-06,
+      "loss": 0.1746,
+      "step": 8853
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.396895994864958,
+      "learning_rate": 5.364067682556561e-06,
+      "loss": 0.2189,
+      "step": 8854
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.1256523906360647,
+      "learning_rate": 5.3631692269420045e-06,
+      "loss": 0.1861,
+      "step": 8855
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.3370166025900767,
+      "learning_rate": 5.362270759538767e-06,
+      "loss": 0.1873,
+      "step": 8856
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.0387534940485255,
+      "learning_rate": 5.361372280376013e-06,
+      "loss": 0.1886,
+      "step": 8857
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.3311975392674493,
+      "learning_rate": 5.360473789482905e-06,
+      "loss": 0.2352,
+      "step": 8858
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.3373461791964776,
+      "learning_rate": 5.359575286888613e-06,
+      "loss": 0.1635,
+      "step": 8859
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.9837074485628777,
+      "learning_rate": 5.358676772622299e-06,
+      "loss": 0.145,
+      "step": 8860
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.0493926903220177,
+      "learning_rate": 5.357778246713131e-06,
+      "loss": 0.1867,
+      "step": 8861
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.2677821654925983,
+      "learning_rate": 5.356879709190276e-06,
+      "loss": 0.1808,
+      "step": 8862
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.5633131009678987,
+      "learning_rate": 5.3559811600829e-06,
+      "loss": 0.1834,
+      "step": 8863
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.9314030818280581,
+      "learning_rate": 5.3550825994201706e-06,
+      "loss": 0.2282,
+      "step": 8864
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.2078470232570986,
+      "learning_rate": 5.354184027231255e-06,
+      "loss": 0.2099,
+      "step": 8865
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.9731642495706657,
+      "learning_rate": 5.353285443545324e-06,
+      "loss": 0.1886,
+      "step": 8866
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.892861039715165,
+      "learning_rate": 5.352386848391543e-06,
+      "loss": 0.1438,
+      "step": 8867
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.491764368111121,
+      "learning_rate": 5.351488241799084e-06,
+      "loss": 0.2286,
+      "step": 8868
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.1017402005085493,
+      "learning_rate": 5.3505896237971124e-06,
+      "loss": 0.1587,
+      "step": 8869
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.4032860196249897,
+      "learning_rate": 5.349690994414801e-06,
+      "loss": 0.1951,
+      "step": 8870
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.3796595307060815,
+      "learning_rate": 5.3487923536813195e-06,
+      "loss": 0.2083,
+      "step": 8871
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.1269967964873822,
+      "learning_rate": 5.347893701625837e-06,
+      "loss": 0.1869,
+      "step": 8872
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 2.7711372184946925,
+      "learning_rate": 5.346995038277523e-06,
+      "loss": 0.212,
+      "step": 8873
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.3585968955860717,
+      "learning_rate": 5.346096363665553e-06,
+      "loss": 0.2096,
+      "step": 8874
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.338100299972496,
+      "learning_rate": 5.345197677819096e-06,
+      "loss": 0.1928,
+      "step": 8875
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.67662554370846,
+      "learning_rate": 5.344298980767322e-06,
+      "loss": 0.2107,
+      "step": 8876
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.1197246805012535,
+      "learning_rate": 5.343400272539407e-06,
+      "loss": 0.1839,
+      "step": 8877
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 2.877471726881006,
+      "learning_rate": 5.342501553164521e-06,
+      "loss": 0.2209,
+      "step": 8878
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.493394363636039,
+      "learning_rate": 5.341602822671838e-06,
+      "loss": 0.2167,
+      "step": 8879
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.578467671799477,
+      "learning_rate": 5.340704081090531e-06,
+      "loss": 0.2358,
+      "step": 8880
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.3645452989732796,
+      "learning_rate": 5.339805328449773e-06,
+      "loss": 0.1872,
+      "step": 8881
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.8000004755539396,
+      "learning_rate": 5.338906564778739e-06,
+      "loss": 0.1995,
+      "step": 8882
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.4068284964109805,
+      "learning_rate": 5.338007790106602e-06,
+      "loss": 0.1927,
+      "step": 8883
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.4145464200843736,
+      "learning_rate": 5.337109004462539e-06,
+      "loss": 0.2054,
+      "step": 8884
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.5078693339650826,
+      "learning_rate": 5.336210207875723e-06,
+      "loss": 0.2312,
+      "step": 8885
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.0561409445688459,
+      "learning_rate": 5.3353114003753316e-06,
+      "loss": 0.158,
+      "step": 8886
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.3580187521276754,
+      "learning_rate": 5.334412581990537e-06,
+      "loss": 0.1884,
+      "step": 8887
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.8512954555676291,
+      "learning_rate": 5.33351375275052e-06,
+      "loss": 0.1858,
+      "step": 8888
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.0210360132018212,
+      "learning_rate": 5.332614912684454e-06,
+      "loss": 0.145,
+      "step": 8889
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.9826774549114597,
+      "learning_rate": 5.331716061821517e-06,
+      "loss": 0.1677,
+      "step": 8890
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.1297287139391974,
+      "learning_rate": 5.330817200190885e-06,
+      "loss": 0.2099,
+      "step": 8891
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.4664438561935553,
+      "learning_rate": 5.329918327821737e-06,
+      "loss": 0.2225,
+      "step": 8892
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.267369758690402,
+      "learning_rate": 5.329019444743251e-06,
+      "loss": 0.2042,
+      "step": 8893
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.1716214287122573,
+      "learning_rate": 5.328120550984603e-06,
+      "loss": 0.1427,
+      "step": 8894
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.2477691592134876,
+      "learning_rate": 5.327221646574975e-06,
+      "loss": 0.2249,
+      "step": 8895
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.9964078071513877,
+      "learning_rate": 5.3263227315435424e-06,
+      "loss": 0.1586,
+      "step": 8896
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.2388557987450433,
+      "learning_rate": 5.325423805919487e-06,
+      "loss": 0.1815,
+      "step": 8897
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.1537967750696296,
+      "learning_rate": 5.32452486973199e-06,
+      "loss": 0.1935,
+      "step": 8898
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.770983281285101,
+      "learning_rate": 5.323625923010227e-06,
+      "loss": 0.2202,
+      "step": 8899
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.226649591063172,
+      "learning_rate": 5.32272696578338e-06,
+      "loss": 0.214,
+      "step": 8900
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.387640987570457,
+      "learning_rate": 5.32182799808063e-06,
+      "loss": 0.2003,
+      "step": 8901
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.1901865985572275,
+      "learning_rate": 5.320929019931159e-06,
+      "loss": 0.2235,
+      "step": 8902
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.0696326902223856,
+      "learning_rate": 5.320030031364147e-06,
+      "loss": 0.1545,
+      "step": 8903
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.4315223332723859,
+      "learning_rate": 5.319131032408777e-06,
+      "loss": 0.185,
+      "step": 8904
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.3330705481218372,
+      "learning_rate": 5.318232023094229e-06,
+      "loss": 0.2123,
+      "step": 8905
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.4374637534402324,
+      "learning_rate": 5.3173330034496875e-06,
+      "loss": 0.2112,
+      "step": 8906
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.3158123504773758,
+      "learning_rate": 5.3164339735043335e-06,
+      "loss": 0.185,
+      "step": 8907
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.6064549222618156,
+      "learning_rate": 5.3155349332873514e-06,
+      "loss": 0.1973,
+      "step": 8908
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.1058075186948675,
+      "learning_rate": 5.314635882827924e-06,
+      "loss": 0.1871,
+      "step": 8909
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.2832531607175741,
+      "learning_rate": 5.313736822155234e-06,
+      "loss": 0.1904,
+      "step": 8910
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.9440466626103811,
+      "learning_rate": 5.312837751298468e-06,
+      "loss": 0.1714,
+      "step": 8911
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.184194122986512,
+      "learning_rate": 5.311938670286806e-06,
+      "loss": 0.1927,
+      "step": 8912
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.2086109453591962,
+      "learning_rate": 5.311039579149437e-06,
+      "loss": 0.1698,
+      "step": 8913
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.2787010406908235,
+      "learning_rate": 5.310140477915544e-06,
+      "loss": 0.2181,
+      "step": 8914
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.3969374075225598,
+      "learning_rate": 5.309241366614314e-06,
+      "loss": 0.2279,
+      "step": 8915
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.701132216577067,
+      "learning_rate": 5.308342245274931e-06,
+      "loss": 0.2185,
+      "step": 8916
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.1679494323536346,
+      "learning_rate": 5.307443113926581e-06,
+      "loss": 0.1915,
+      "step": 8917
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.9947054313394081,
+      "learning_rate": 5.3065439725984506e-06,
+      "loss": 0.229,
+      "step": 8918
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.9245207692129724,
+      "learning_rate": 5.305644821319726e-06,
+      "loss": 0.1813,
+      "step": 8919
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.6746734968134915,
+      "learning_rate": 5.304745660119595e-06,
+      "loss": 0.2114,
+      "step": 8920
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.9878112406603452,
+      "learning_rate": 5.303846489027244e-06,
+      "loss": 0.178,
+      "step": 8921
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.1854911803686123,
+      "learning_rate": 5.302947308071864e-06,
+      "loss": 0.1634,
+      "step": 8922
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.0666254153725272,
+      "learning_rate": 5.302048117282637e-06,
+      "loss": 0.2236,
+      "step": 8923
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.8784089991825639,
+      "learning_rate": 5.301148916688756e-06,
+      "loss": 0.1734,
+      "step": 8924
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.8942484679375616,
+      "learning_rate": 5.3002497063194095e-06,
+      "loss": 0.1606,
+      "step": 8925
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.3293308842220977,
+      "learning_rate": 5.299350486203784e-06,
+      "loss": 0.1561,
+      "step": 8926
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 2.285969496559772,
+      "learning_rate": 5.298451256371068e-06,
+      "loss": 0.2046,
+      "step": 8927
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.241654473629978,
+      "learning_rate": 5.297552016850453e-06,
+      "loss": 0.2008,
+      "step": 8928
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.6278906430217306,
+      "learning_rate": 5.296652767671129e-06,
+      "loss": 0.1948,
+      "step": 8929
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.0867416437015136,
+      "learning_rate": 5.295753508862284e-06,
+      "loss": 0.138,
+      "step": 8930
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.3297160309735099,
+      "learning_rate": 5.294854240453113e-06,
+      "loss": 0.2158,
+      "step": 8931
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.1493888988349414,
+      "learning_rate": 5.293954962472801e-06,
+      "loss": 0.2296,
+      "step": 8932
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.2815771615622897,
+      "learning_rate": 5.293055674950544e-06,
+      "loss": 0.1844,
+      "step": 8933
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.7741192125008194,
+      "learning_rate": 5.29215637791553e-06,
+      "loss": 0.2602,
+      "step": 8934
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.2299183613663305,
+      "learning_rate": 5.291257071396954e-06,
+      "loss": 0.1805,
+      "step": 8935
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.9318669562053383,
+      "learning_rate": 5.290357755424003e-06,
+      "loss": 0.1473,
+      "step": 8936
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.0265555277025873,
+      "learning_rate": 5.289458430025874e-06,
+      "loss": 0.1683,
+      "step": 8937
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.1328970515826042,
+      "learning_rate": 5.288559095231757e-06,
+      "loss": 0.2087,
+      "step": 8938
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.402871709913733,
+      "learning_rate": 5.287659751070846e-06,
+      "loss": 0.2165,
+      "step": 8939
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.0864842568423034,
+      "learning_rate": 5.286760397572334e-06,
+      "loss": 0.1838,
+      "step": 8940
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.0876479366134386,
+      "learning_rate": 5.285861034765415e-06,
+      "loss": 0.2109,
+      "step": 8941
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.17542785897426,
+      "learning_rate": 5.2849616626792834e-06,
+      "loss": 0.1926,
+      "step": 8942
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.277939333872744,
+      "learning_rate": 5.284062281343131e-06,
+      "loss": 0.1744,
+      "step": 8943
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.2270899658989962,
+      "learning_rate": 5.2831628907861545e-06,
+      "loss": 0.2074,
+      "step": 8944
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.9374230989069389,
+      "learning_rate": 5.282263491037547e-06,
+      "loss": 0.1418,
+      "step": 8945
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.2858012415747613,
+      "learning_rate": 5.281364082126503e-06,
+      "loss": 0.2154,
+      "step": 8946
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.826162774782405,
+      "learning_rate": 5.280464664082222e-06,
+      "loss": 0.2476,
+      "step": 8947
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.8178550223583536,
+      "learning_rate": 5.279565236933894e-06,
+      "loss": 0.2038,
+      "step": 8948
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.2446766661216302,
+      "learning_rate": 5.27866580071072e-06,
+      "loss": 0.2575,
+      "step": 8949
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.552751864928583,
+      "learning_rate": 5.2777663554418926e-06,
+      "loss": 0.1869,
+      "step": 8950
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.0921716166365962,
+      "learning_rate": 5.27686690115661e-06,
+      "loss": 0.2103,
+      "step": 8951
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.129876495317292,
+      "learning_rate": 5.275967437884069e-06,
+      "loss": 0.1725,
+      "step": 8952
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.4549676456368978,
+      "learning_rate": 5.275067965653465e-06,
+      "loss": 0.2407,
+      "step": 8953
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.39355410607265,
+      "learning_rate": 5.274168484493998e-06,
+      "loss": 0.1998,
+      "step": 8954
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.2400035042530582,
+      "learning_rate": 5.2732689944348624e-06,
+      "loss": 0.2184,
+      "step": 8955
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.208113119150326,
+      "learning_rate": 5.272369495505261e-06,
+      "loss": 0.1668,
+      "step": 8956
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.7472918861269374,
+      "learning_rate": 5.271469987734387e-06,
+      "loss": 0.1479,
+      "step": 8957
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.159093659188583,
+      "learning_rate": 5.270570471151442e-06,
+      "loss": 0.2103,
+      "step": 8958
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.4688864477166144,
+      "learning_rate": 5.269670945785622e-06,
+      "loss": 0.1959,
+      "step": 8959
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.5382331220654721,
+      "learning_rate": 5.2687714116661304e-06,
+      "loss": 0.2119,
+      "step": 8960
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.9618600975428891,
+      "learning_rate": 5.267871868822164e-06,
+      "loss": 0.1828,
+      "step": 8961
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.1152905434996156,
+      "learning_rate": 5.266972317282921e-06,
+      "loss": 0.1662,
+      "step": 8962
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.0365525059912566,
+      "learning_rate": 5.266072757077605e-06,
+      "loss": 0.2004,
+      "step": 8963
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.0725762450630807,
+      "learning_rate": 5.265173188235413e-06,
+      "loss": 0.1696,
+      "step": 8964
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.8221838909355951,
+      "learning_rate": 5.264273610785547e-06,
+      "loss": 0.128,
+      "step": 8965
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.1907544995516641,
+      "learning_rate": 5.263374024757208e-06,
+      "loss": 0.2172,
+      "step": 8966
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.3890394537164947,
+      "learning_rate": 5.262474430179597e-06,
+      "loss": 0.2291,
+      "step": 8967
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.3859845622220934,
+      "learning_rate": 5.261574827081915e-06,
+      "loss": 0.258,
+      "step": 8968
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.5530182464991376,
+      "learning_rate": 5.260675215493363e-06,
+      "loss": 0.1709,
+      "step": 8969
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.9851977906026356,
+      "learning_rate": 5.259775595443144e-06,
+      "loss": 0.1704,
+      "step": 8970
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.077912883347629,
+      "learning_rate": 5.258875966960459e-06,
+      "loss": 0.1757,
+      "step": 8971
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.7722315948158726,
+      "learning_rate": 5.2579763300745124e-06,
+      "loss": 0.1418,
+      "step": 8972
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.2658703354302914,
+      "learning_rate": 5.257076684814505e-06,
+      "loss": 0.1605,
+      "step": 8973
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.605706858554743,
+      "learning_rate": 5.2561770312096415e-06,
+      "loss": 0.1902,
+      "step": 8974
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.0844912520956507,
+      "learning_rate": 5.2552773692891235e-06,
+      "loss": 0.1634,
+      "step": 8975
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.125630999802777,
+      "learning_rate": 5.254377699082156e-06,
+      "loss": 0.2079,
+      "step": 8976
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.1707285534296241,
+      "learning_rate": 5.25347802061794e-06,
+      "loss": 0.1839,
+      "step": 8977
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.2320440109066182,
+      "learning_rate": 5.252578333925685e-06,
+      "loss": 0.2214,
+      "step": 8978
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.1805234088525716,
+      "learning_rate": 5.2516786390345905e-06,
+      "loss": 0.2114,
+      "step": 8979
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.0032032609281023,
+      "learning_rate": 5.250778935973862e-06,
+      "loss": 0.2008,
+      "step": 8980
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.1814581674994873,
+      "learning_rate": 5.249879224772706e-06,
+      "loss": 0.1967,
+      "step": 8981
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.8684764403054563,
+      "learning_rate": 5.248979505460326e-06,
+      "loss": 0.2043,
+      "step": 8982
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.591949810100069,
+      "learning_rate": 5.248079778065929e-06,
+      "loss": 0.1951,
+      "step": 8983
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.8450188169803684,
+      "learning_rate": 5.247180042618718e-06,
+      "loss": 0.1816,
+      "step": 8984
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.3094339918490445,
+      "learning_rate": 5.246280299147902e-06,
+      "loss": 0.2198,
+      "step": 8985
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.40933314999278,
+      "learning_rate": 5.2453805476826845e-06,
+      "loss": 0.1592,
+      "step": 8986
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.5604861322834553,
+      "learning_rate": 5.244480788252275e-06,
+      "loss": 0.1758,
+      "step": 8987
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.8546584180176532,
+      "learning_rate": 5.243581020885878e-06,
+      "loss": 0.1622,
+      "step": 8988
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.9397661179422246,
+      "learning_rate": 5.242681245612701e-06,
+      "loss": 0.2147,
+      "step": 8989
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.1393371457619177,
+      "learning_rate": 5.241781462461951e-06,
+      "loss": 0.2001,
+      "step": 8990
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.9078441669338014,
+      "learning_rate": 5.240881671462836e-06,
+      "loss": 0.1696,
+      "step": 8991
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.1407482279878298,
+      "learning_rate": 5.239981872644564e-06,
+      "loss": 0.1567,
+      "step": 8992
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.8675999219166698,
+      "learning_rate": 5.239082066036341e-06,
+      "loss": 0.1681,
+      "step": 8993
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.239124020327594,
+      "learning_rate": 5.2381822516673774e-06,
+      "loss": 0.2023,
+      "step": 8994
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.1805154976718177,
+      "learning_rate": 5.23728242956688e-06,
+      "loss": 0.1797,
+      "step": 8995
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.536516142959125,
+      "learning_rate": 5.236382599764061e-06,
+      "loss": 0.2378,
+      "step": 8996
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.417570662511681,
+      "learning_rate": 5.235482762288125e-06,
+      "loss": 0.2254,
+      "step": 8997
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.9203128048032887,
+      "learning_rate": 5.234582917168283e-06,
+      "loss": 0.1712,
+      "step": 8998
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.316549949839707,
+      "learning_rate": 5.233683064433745e-06,
+      "loss": 0.1699,
+      "step": 8999
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.330165109576888,
+      "learning_rate": 5.23278320411372e-06,
+      "loss": 0.1658,
+      "step": 9000
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.4796986678164268,
+      "learning_rate": 5.231883336237419e-06,
+      "loss": 0.1847,
+      "step": 9001
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.3252220994438937,
+      "learning_rate": 5.230983460834051e-06,
+      "loss": 0.2394,
+      "step": 9002
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.9716998624029245,
+      "learning_rate": 5.230083577932827e-06,
+      "loss": 0.2162,
+      "step": 9003
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.9843441564783808,
+      "learning_rate": 5.229183687562959e-06,
+      "loss": 0.2253,
+      "step": 9004
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.1219344260168493,
+      "learning_rate": 5.228283789753655e-06,
+      "loss": 0.1937,
+      "step": 9005
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.0272553445758705,
+      "learning_rate": 5.227383884534129e-06,
+      "loss": 0.1923,
+      "step": 9006
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.2224589514666708,
+      "learning_rate": 5.226483971933591e-06,
+      "loss": 0.2538,
+      "step": 9007
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.3203091931724895,
+      "learning_rate": 5.225584051981253e-06,
+      "loss": 0.1942,
+      "step": 9008
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.111531019545244,
+      "learning_rate": 5.224684124706326e-06,
+      "loss": 0.1587,
+      "step": 9009
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.3904839213742046,
+      "learning_rate": 5.223784190138025e-06,
+      "loss": 0.1936,
+      "step": 9010
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.3305775577182855,
+      "learning_rate": 5.222884248305559e-06,
+      "loss": 0.2169,
+      "step": 9011
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.2645140764302174,
+      "learning_rate": 5.221984299238143e-06,
+      "loss": 0.1904,
+      "step": 9012
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.4114592468310376,
+      "learning_rate": 5.221084342964989e-06,
+      "loss": 0.1995,
+      "step": 9013
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.223232818889236,
+      "learning_rate": 5.220184379515311e-06,
+      "loss": 0.1729,
+      "step": 9014
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.238630868190262,
+      "learning_rate": 5.21928440891832e-06,
+      "loss": 0.1297,
+      "step": 9015
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.4543027540073223,
+      "learning_rate": 5.218384431203231e-06,
+      "loss": 0.162,
+      "step": 9016
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.319491632057646,
+      "learning_rate": 5.217484446399258e-06,
+      "loss": 0.2275,
+      "step": 9017
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.3845670259048988,
+      "learning_rate": 5.2165844545356135e-06,
+      "loss": 0.2455,
+      "step": 9018
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.1443741335779614,
+      "learning_rate": 5.215684455641516e-06,
+      "loss": 0.1959,
+      "step": 9019
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.8759597640622305,
+      "learning_rate": 5.2147844497461745e-06,
+      "loss": 0.1985,
+      "step": 9020
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.855325447342718,
+      "learning_rate": 5.213884436878807e-06,
+      "loss": 0.1706,
+      "step": 9021
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.3580430647171096,
+      "learning_rate": 5.212984417068629e-06,
+      "loss": 0.1884,
+      "step": 9022
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.435686836426357,
+      "learning_rate": 5.212084390344853e-06,
+      "loss": 0.183,
+      "step": 9023
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.111563598947569,
+      "learning_rate": 5.211184356736696e-06,
+      "loss": 0.231,
+      "step": 9024
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.279495992525301,
+      "learning_rate": 5.210284316273372e-06,
+      "loss": 0.1628,
+      "step": 9025
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.284251633866028,
+      "learning_rate": 5.2093842689841e-06,
+      "loss": 0.2423,
+      "step": 9026
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.1821748613503276,
+      "learning_rate": 5.208484214898093e-06,
+      "loss": 0.1942,
+      "step": 9027
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.4819358473733923,
+      "learning_rate": 5.207584154044571e-06,
+      "loss": 0.2173,
+      "step": 9028
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.0172574458706392,
+      "learning_rate": 5.206684086452745e-06,
+      "loss": 0.1167,
+      "step": 9029
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.3535151323369878,
+      "learning_rate": 5.205784012151835e-06,
+      "loss": 0.2392,
+      "step": 9030
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.8813581714372699,
+      "learning_rate": 5.204883931171059e-06,
+      "loss": 0.1511,
+      "step": 9031
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.7179214195241551,
+      "learning_rate": 5.203983843539634e-06,
+      "loss": 0.1303,
+      "step": 9032
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.1328150979374,
+      "learning_rate": 5.203083749286774e-06,
+      "loss": 0.1965,
+      "step": 9033
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.8213862340098216,
+      "learning_rate": 5.202183648441698e-06,
+      "loss": 0.1825,
+      "step": 9034
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.1418837532194455,
+      "learning_rate": 5.201283541033626e-06,
+      "loss": 0.2183,
+      "step": 9035
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.4294359716783185,
+      "learning_rate": 5.200383427091774e-06,
+      "loss": 0.2077,
+      "step": 9036
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.9672971914390904,
+      "learning_rate": 5.199483306645361e-06,
+      "loss": 0.2286,
+      "step": 9037
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.2152437137378587,
+      "learning_rate": 5.198583179723604e-06,
+      "loss": 0.2077,
+      "step": 9038
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.45959880764799,
+      "learning_rate": 5.197683046355725e-06,
+      "loss": 0.2026,
+      "step": 9039
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.0689231711697322,
+      "learning_rate": 5.19678290657094e-06,
+      "loss": 0.1685,
+      "step": 9040
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.4096311234725287,
+      "learning_rate": 5.195882760398469e-06,
+      "loss": 0.1988,
+      "step": 9041
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.6248505831992126,
+      "learning_rate": 5.194982607867529e-06,
+      "loss": 0.2347,
+      "step": 9042
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.010068014381821,
+      "learning_rate": 5.194082449007341e-06,
+      "loss": 0.2114,
+      "step": 9043
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.0206200912073422,
+      "learning_rate": 5.193182283847128e-06,
+      "loss": 0.1395,
+      "step": 9044
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.1919902152067492,
+      "learning_rate": 5.192282112416105e-06,
+      "loss": 0.1943,
+      "step": 9045
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.0991894070762818,
+      "learning_rate": 5.191381934743495e-06,
+      "loss": 0.1542,
+      "step": 9046
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.9589962832300033,
+      "learning_rate": 5.190481750858517e-06,
+      "loss": 0.2379,
+      "step": 9047
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.1285360448672725,
+      "learning_rate": 5.189581560790392e-06,
+      "loss": 0.2112,
+      "step": 9048
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.9524894455828095,
+      "learning_rate": 5.188681364568341e-06,
+      "loss": 0.1987,
+      "step": 9049
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.171614516236263,
+      "learning_rate": 5.1877811622215844e-06,
+      "loss": 0.2404,
+      "step": 9050
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.183764122827849,
+      "learning_rate": 5.1868809537793416e-06,
+      "loss": 0.1635,
+      "step": 9051
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.0278883313777296,
+      "learning_rate": 5.185980739270836e-06,
+      "loss": 0.1951,
+      "step": 9052
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.2295147452129556,
+      "learning_rate": 5.185080518725289e-06,
+      "loss": 0.1923,
+      "step": 9053
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.0411025900238593,
+      "learning_rate": 5.1841802921719235e-06,
+      "loss": 0.14,
+      "step": 9054
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.3658785671344824,
+      "learning_rate": 5.183280059639958e-06,
+      "loss": 0.2271,
+      "step": 9055
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.4567660225834365,
+      "learning_rate": 5.182379821158616e-06,
+      "loss": 0.21,
+      "step": 9056
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.037776078983882,
+      "learning_rate": 5.181479576757123e-06,
+      "loss": 0.1832,
+      "step": 9057
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.9459994968460602,
+      "learning_rate": 5.1805793264646964e-06,
+      "loss": 0.1613,
+      "step": 9058
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.2008291329272442,
+      "learning_rate": 5.1796790703105615e-06,
+      "loss": 0.2104,
+      "step": 9059
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.1234247749414383,
+      "learning_rate": 5.17877880832394e-06,
+      "loss": 0.1798,
+      "step": 9060
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.1938353773862498,
+      "learning_rate": 5.177878540534055e-06,
+      "loss": 0.2144,
+      "step": 9061
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.3072988009986384,
+      "learning_rate": 5.176978266970132e-06,
+      "loss": 0.1996,
+      "step": 9062
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.8143559666994342,
+      "learning_rate": 5.176077987661391e-06,
+      "loss": 0.2012,
+      "step": 9063
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.0767518227236066,
+      "learning_rate": 5.175177702637058e-06,
+      "loss": 0.1711,
+      "step": 9064
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.1407136490744392,
+      "learning_rate": 5.174277411926356e-06,
+      "loss": 0.195,
+      "step": 9065
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.9015894540363959,
+      "learning_rate": 5.17337711555851e-06,
+      "loss": 0.1512,
+      "step": 9066
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.9529164097746962,
+      "learning_rate": 5.172476813562744e-06,
+      "loss": 0.1646,
+      "step": 9067
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.4226553185306456,
+      "learning_rate": 5.17157650596828e-06,
+      "loss": 0.2031,
+      "step": 9068
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.0851783671916164,
+      "learning_rate": 5.170676192804343e-06,
+      "loss": 0.1374,
+      "step": 9069
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.4092943061641523,
+      "learning_rate": 5.1697758741001595e-06,
+      "loss": 0.1629,
+      "step": 9070
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.2316130086051198,
+      "learning_rate": 5.168875549884953e-06,
+      "loss": 0.1972,
+      "step": 9071
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.9078783849323244,
+      "learning_rate": 5.167975220187949e-06,
+      "loss": 0.2196,
+      "step": 9072
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.1775590503569817,
+      "learning_rate": 5.1670748850383734e-06,
+      "loss": 0.1864,
+      "step": 9073
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.1558474858205292,
+      "learning_rate": 5.1661745444654494e-06,
+      "loss": 0.1706,
+      "step": 9074
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.9882142718993266,
+      "learning_rate": 5.165274198498407e-06,
+      "loss": 0.1496,
+      "step": 9075
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.8465135668767528,
+      "learning_rate": 5.164373847166468e-06,
+      "loss": 0.211,
+      "step": 9076
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.8870020972442669,
+      "learning_rate": 5.163473490498858e-06,
+      "loss": 0.1747,
+      "step": 9077
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.19145251324993,
+      "learning_rate": 5.162573128524806e-06,
+      "loss": 0.2033,
+      "step": 9078
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.1086857847570033,
+      "learning_rate": 5.161672761273536e-06,
+      "loss": 0.1734,
+      "step": 9079
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.3723618221472005,
+      "learning_rate": 5.160772388774275e-06,
+      "loss": 0.209,
+      "step": 9080
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.1343708566322004,
+      "learning_rate": 5.15987201105625e-06,
+      "loss": 0.183,
+      "step": 9081
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.2268682511834388,
+      "learning_rate": 5.158971628148689e-06,
+      "loss": 0.2052,
+      "step": 9082
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.341299070165463,
+      "learning_rate": 5.158071240080815e-06,
+      "loss": 0.2204,
+      "step": 9083
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.207888974363254,
+      "learning_rate": 5.15717084688186e-06,
+      "loss": 0.1979,
+      "step": 9084
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.0822208252241656,
+      "learning_rate": 5.156270448581049e-06,
+      "loss": 0.2126,
+      "step": 9085
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.9985228073736904,
+      "learning_rate": 5.155370045207607e-06,
+      "loss": 0.1615,
+      "step": 9086
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.0429627636226915,
+      "learning_rate": 5.154469636790766e-06,
+      "loss": 0.222,
+      "step": 9087
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.0679642453967226,
+      "learning_rate": 5.153569223359751e-06,
+      "loss": 0.1388,
+      "step": 9088
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.638532980487432,
+      "learning_rate": 5.152668804943791e-06,
+      "loss": 0.2406,
+      "step": 9089
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.9643618465164401,
+      "learning_rate": 5.151768381572113e-06,
+      "loss": 0.2064,
+      "step": 9090
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.0384408024874008,
+      "learning_rate": 5.150867953273947e-06,
+      "loss": 0.1823,
+      "step": 9091
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.1326173277142906,
+      "learning_rate": 5.14996752007852e-06,
+      "loss": 0.1759,
+      "step": 9092
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.1006762679906044,
+      "learning_rate": 5.149067082015061e-06,
+      "loss": 0.1823,
+      "step": 9093
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.3097533767268572,
+      "learning_rate": 5.148166639112799e-06,
+      "loss": 0.1921,
+      "step": 9094
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.1540087971354074,
+      "learning_rate": 5.147266191400962e-06,
+      "loss": 0.1703,
+      "step": 9095
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.9352594503505458,
+      "learning_rate": 5.146365738908782e-06,
+      "loss": 0.169,
+      "step": 9096
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.6207101919664177,
+      "learning_rate": 5.145465281665484e-06,
+      "loss": 0.2206,
+      "step": 9097
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.214625317478796,
+      "learning_rate": 5.144564819700299e-06,
+      "loss": 0.2095,
+      "step": 9098
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.3294955089500142,
+      "learning_rate": 5.143664353042458e-06,
+      "loss": 0.2018,
+      "step": 9099
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.176127202285633,
+      "learning_rate": 5.142763881721189e-06,
+      "loss": 0.2325,
+      "step": 9100
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.9649295459856577,
+      "learning_rate": 5.141863405765721e-06,
+      "loss": 0.1911,
+      "step": 9101
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.0123661043236931,
+      "learning_rate": 5.140962925205288e-06,
+      "loss": 0.1926,
+      "step": 9102
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.171176807316075,
+      "learning_rate": 5.1400624400691165e-06,
+      "loss": 0.2159,
+      "step": 9103
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.816933862955751,
+      "learning_rate": 5.139161950386437e-06,
+      "loss": 0.2221,
+      "step": 9104
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.2804336142127002,
+      "learning_rate": 5.138261456186481e-06,
+      "loss": 0.2085,
+      "step": 9105
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.7093776479261207,
+      "learning_rate": 5.137360957498477e-06,
+      "loss": 0.1874,
+      "step": 9106
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.2728404125540653,
+      "learning_rate": 5.136460454351659e-06,
+      "loss": 0.2243,
+      "step": 9107
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.2643358735410304,
+      "learning_rate": 5.135559946775256e-06,
+      "loss": 0.2491,
+      "step": 9108
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.8344542660069358,
+      "learning_rate": 5.134659434798499e-06,
+      "loss": 0.2292,
+      "step": 9109
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.3139891110110393,
+      "learning_rate": 5.133758918450618e-06,
+      "loss": 0.2029,
+      "step": 9110
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.186512852060328,
+      "learning_rate": 5.132858397760848e-06,
+      "loss": 0.2309,
+      "step": 9111
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.9159565644247069,
+      "learning_rate": 5.131957872758417e-06,
+      "loss": 0.1765,
+      "step": 9112
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.9467369708404182,
+      "learning_rate": 5.131057343472558e-06,
+      "loss": 0.2361,
+      "step": 9113
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.1521387160273093,
+      "learning_rate": 5.130156809932502e-06,
+      "loss": 0.1997,
+      "step": 9114
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.2949766691188476,
+      "learning_rate": 5.1292562721674795e-06,
+      "loss": 0.1876,
+      "step": 9115
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.4026794563065885,
+      "learning_rate": 5.128355730206727e-06,
+      "loss": 0.2222,
+      "step": 9116
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.1981586708554695,
+      "learning_rate": 5.127455184079471e-06,
+      "loss": 0.2105,
+      "step": 9117
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.2667482066208602,
+      "learning_rate": 5.126554633814948e-06,
+      "loss": 0.1793,
+      "step": 9118
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.1622829147399563,
+      "learning_rate": 5.125654079442387e-06,
+      "loss": 0.2413,
+      "step": 9119
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.1594072149150414,
+      "learning_rate": 5.124753520991025e-06,
+      "loss": 0.1983,
+      "step": 9120
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.9589293341758575,
+      "learning_rate": 5.123852958490091e-06,
+      "loss": 0.1922,
+      "step": 9121
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.9890152011427289,
+      "learning_rate": 5.122952391968817e-06,
+      "loss": 0.1372,
+      "step": 9122
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.989616302727386,
+      "learning_rate": 5.12205182145644e-06,
+      "loss": 0.1633,
+      "step": 9123
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.2684218356933783,
+      "learning_rate": 5.121151246982189e-06,
+      "loss": 0.1939,
+      "step": 9124
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.7270353530656275,
+      "learning_rate": 5.120250668575301e-06,
+      "loss": 0.203,
+      "step": 9125
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.0365091372104207,
+      "learning_rate": 5.1193500862650045e-06,
+      "loss": 0.2318,
+      "step": 9126
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.5512612522421836,
+      "learning_rate": 5.118449500080538e-06,
+      "loss": 0.2644,
+      "step": 9127
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.0043578560771877,
+      "learning_rate": 5.11754891005113e-06,
+      "loss": 0.1571,
+      "step": 9128
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.8512694987789041,
+      "learning_rate": 5.1166483162060185e-06,
+      "loss": 0.1947,
+      "step": 9129
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.1707102582203157,
+      "learning_rate": 5.115747718574436e-06,
+      "loss": 0.1786,
+      "step": 9130
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.240485581916754,
+      "learning_rate": 5.114847117185615e-06,
+      "loss": 0.1972,
+      "step": 9131
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.191988024391031,
+      "learning_rate": 5.113946512068791e-06,
+      "loss": 0.2451,
+      "step": 9132
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.1252998798425955,
+      "learning_rate": 5.113045903253196e-06,
+      "loss": 0.1693,
+      "step": 9133
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.9173340246600865,
+      "learning_rate": 5.1121452907680695e-06,
+      "loss": 0.17,
+      "step": 9134
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.1441839394228783,
+      "learning_rate": 5.111244674642639e-06,
+      "loss": 0.1644,
+      "step": 9135
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.8002192525562265,
+      "learning_rate": 5.110344054906145e-06,
+      "loss": 0.127,
+      "step": 9136
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.8828247774165022,
+      "learning_rate": 5.109443431587817e-06,
+      "loss": 0.1379,
+      "step": 9137
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.0792280131685739,
+      "learning_rate": 5.108542804716894e-06,
+      "loss": 0.1798,
+      "step": 9138
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.2457191283835949,
+      "learning_rate": 5.107642174322609e-06,
+      "loss": 0.167,
+      "step": 9139
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.3215840257829596,
+      "learning_rate": 5.106741540434197e-06,
+      "loss": 0.2071,
+      "step": 9140
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.1119100692894421,
+      "learning_rate": 5.1058409030808935e-06,
+      "loss": 0.2209,
+      "step": 9141
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.3000336666745396,
+      "learning_rate": 5.104940262291932e-06,
+      "loss": 0.2426,
+      "step": 9142
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.9885669858721221,
+      "learning_rate": 5.104039618096551e-06,
+      "loss": 0.177,
+      "step": 9143
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.3258557371211335,
+      "learning_rate": 5.103138970523982e-06,
+      "loss": 0.218,
+      "step": 9144
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.1094324892890552,
+      "learning_rate": 5.102238319603465e-06,
+      "loss": 0.1839,
+      "step": 9145
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.1103615169028611,
+      "learning_rate": 5.101337665364232e-06,
+      "loss": 0.2253,
+      "step": 9146
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.0614612880505936,
+      "learning_rate": 5.100437007835521e-06,
+      "loss": 0.221,
+      "step": 9147
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.2185884576499872,
+      "learning_rate": 5.099536347046566e-06,
+      "loss": 0.2122,
+      "step": 9148
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.839096066977375,
+      "learning_rate": 5.098635683026605e-06,
+      "loss": 0.1533,
+      "step": 9149
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.33513858536831,
+      "learning_rate": 5.097735015804872e-06,
+      "loss": 0.1712,
+      "step": 9150
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.2985251768924597,
+      "learning_rate": 5.096834345410605e-06,
+      "loss": 0.1938,
+      "step": 9151
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.0429856874040762,
+      "learning_rate": 5.095933671873039e-06,
+      "loss": 0.1857,
+      "step": 9152
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.0784916530709692,
+      "learning_rate": 5.095032995221411e-06,
+      "loss": 0.252,
+      "step": 9153
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.0231532473792289,
+      "learning_rate": 5.094132315484957e-06,
+      "loss": 0.191,
+      "step": 9154
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.4826571512797442,
+      "learning_rate": 5.093231632692915e-06,
+      "loss": 0.1944,
+      "step": 9155
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.036864490656971,
+      "learning_rate": 5.09233094687452e-06,
+      "loss": 0.1833,
+      "step": 9156
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.8457533753158868,
+      "learning_rate": 5.091430258059009e-06,
+      "loss": 0.1408,
+      "step": 9157
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.0347463335980098,
+      "learning_rate": 5.090529566275618e-06,
+      "loss": 0.2349,
+      "step": 9158
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.1870122020599625,
+      "learning_rate": 5.089628871553588e-06,
+      "loss": 0.2095,
+      "step": 9159
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.3370363919318728,
+      "learning_rate": 5.088728173922152e-06,
+      "loss": 0.2382,
+      "step": 9160
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.5641993401332652,
+      "learning_rate": 5.087827473410548e-06,
+      "loss": 0.2192,
+      "step": 9161
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.7346920075973913,
+      "learning_rate": 5.086926770048013e-06,
+      "loss": 0.251,
+      "step": 9162
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.951818581200296,
+      "learning_rate": 5.086026063863787e-06,
+      "loss": 0.2431,
+      "step": 9163
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.9967510410896129,
+      "learning_rate": 5.085125354887104e-06,
+      "loss": 0.2087,
+      "step": 9164
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.2192981141563712,
+      "learning_rate": 5.084224643147204e-06,
+      "loss": 0.165,
+      "step": 9165
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.91206294805416,
+      "learning_rate": 5.083323928673324e-06,
+      "loss": 0.1859,
+      "step": 9166
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.1602799449188883,
+      "learning_rate": 5.0824232114947005e-06,
+      "loss": 0.1721,
+      "step": 9167
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.4021759278126174,
+      "learning_rate": 5.081522491640573e-06,
+      "loss": 0.2181,
+      "step": 9168
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.9048106121678834,
+      "learning_rate": 5.080621769140177e-06,
+      "loss": 0.1806,
+      "step": 9169
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.7796888035130737,
+      "learning_rate": 5.0797210440227525e-06,
+      "loss": 0.2049,
+      "step": 9170
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.2651058533354824,
+      "learning_rate": 5.078820316317537e-06,
+      "loss": 0.2237,
+      "step": 9171
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.1611130169720245,
+      "learning_rate": 5.07791958605377e-06,
+      "loss": 0.1747,
+      "step": 9172
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.363356664697634,
+      "learning_rate": 5.077018853260687e-06,
+      "loss": 0.2472,
+      "step": 9173
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.5754227045464437,
+      "learning_rate": 5.076118117967529e-06,
+      "loss": 0.1888,
+      "step": 9174
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.3342452959993072,
+      "learning_rate": 5.075217380203533e-06,
+      "loss": 0.2528,
+      "step": 9175
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.2135078806805926,
+      "learning_rate": 5.074316639997936e-06,
+      "loss": 0.2015,
+      "step": 9176
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.597832535948791,
+      "learning_rate": 5.07341589737998e-06,
+      "loss": 0.1923,
+      "step": 9177
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.0909381476665072,
+      "learning_rate": 5.0725151523789e-06,
+      "loss": 0.1967,
+      "step": 9178
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.9827468706120943,
+      "learning_rate": 5.071614405023938e-06,
+      "loss": 0.2468,
+      "step": 9179
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.5491504153475604,
+      "learning_rate": 5.07071365534433e-06,
+      "loss": 0.2224,
+      "step": 9180
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.1296741405003146,
+      "learning_rate": 5.069812903369317e-06,
+      "loss": 0.2086,
+      "step": 9181
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.7773475790648346,
+      "learning_rate": 5.0689121491281355e-06,
+      "loss": 0.2086,
+      "step": 9182
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.2009976206354114,
+      "learning_rate": 5.068011392650029e-06,
+      "loss": 0.2005,
+      "step": 9183
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.601134138274503,
+      "learning_rate": 5.067110633964232e-06,
+      "loss": 0.1481,
+      "step": 9184
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.8308942556670207,
+      "learning_rate": 5.066209873099984e-06,
+      "loss": 0.1998,
+      "step": 9185
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.9809246618038999,
+      "learning_rate": 5.065309110086527e-06,
+      "loss": 0.2122,
+      "step": 9186
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.0904150699303268,
+      "learning_rate": 5.0644083449530975e-06,
+      "loss": 0.129,
+      "step": 9187
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.1658182280591012,
+      "learning_rate": 5.063507577728938e-06,
+      "loss": 0.1559,
+      "step": 9188
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.2967950899714091,
+      "learning_rate": 5.062606808443283e-06,
+      "loss": 0.2493,
+      "step": 9189
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.458131165342218,
+      "learning_rate": 5.061706037125378e-06,
+      "loss": 0.2389,
+      "step": 9190
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.1141335454011962,
+      "learning_rate": 5.060805263804458e-06,
+      "loss": 0.157,
+      "step": 9191
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.1204399867655543,
+      "learning_rate": 5.0599044885097646e-06,
+      "loss": 0.1595,
+      "step": 9192
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.2415357835431424,
+      "learning_rate": 5.059003711270537e-06,
+      "loss": 0.2145,
+      "step": 9193
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.0509242364674292,
+      "learning_rate": 5.058102932116015e-06,
+      "loss": 0.1579,
+      "step": 9194
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.5420936589740093,
+      "learning_rate": 5.057202151075439e-06,
+      "loss": 0.1846,
+      "step": 9195
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.8405789993774829,
+      "learning_rate": 5.056301368178047e-06,
+      "loss": 0.2279,
+      "step": 9196
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.8901901749994752,
+      "learning_rate": 5.0554005834530805e-06,
+      "loss": 0.1819,
+      "step": 9197
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.3583425541893392,
+      "learning_rate": 5.054499796929779e-06,
+      "loss": 0.2131,
+      "step": 9198
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.2520852160585099,
+      "learning_rate": 5.053599008637385e-06,
+      "loss": 0.2461,
+      "step": 9199
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.00604203797025,
+      "learning_rate": 5.052698218605133e-06,
+      "loss": 0.1697,
+      "step": 9200
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.42792986615576,
+      "learning_rate": 5.05179742686227e-06,
+      "loss": 0.2092,
+      "step": 9201
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.8571900725665065,
+      "learning_rate": 5.0508966334380306e-06,
+      "loss": 0.1772,
+      "step": 9202
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.0250185954628508,
+      "learning_rate": 5.049995838361656e-06,
+      "loss": 0.1387,
+      "step": 9203
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.7175082554507741,
+      "learning_rate": 5.0490950416623905e-06,
+      "loss": 0.2436,
+      "step": 9204
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.114866569348323,
+      "learning_rate": 5.04819424336947e-06,
+      "loss": 0.1981,
+      "step": 9205
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.4158069285800665,
+      "learning_rate": 5.047293443512138e-06,
+      "loss": 0.2121,
+      "step": 9206
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 2.129182542143261,
+      "learning_rate": 5.046392642119632e-06,
+      "loss": 0.1742,
+      "step": 9207
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.3605153463753559,
+      "learning_rate": 5.045491839221195e-06,
+      "loss": 0.1808,
+      "step": 9208
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.1181292032863277,
+      "learning_rate": 5.044591034846067e-06,
+      "loss": 0.2167,
+      "step": 9209
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.9532557108630738,
+      "learning_rate": 5.043690229023488e-06,
+      "loss": 0.1789,
+      "step": 9210
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.2377154745488677,
+      "learning_rate": 5.0427894217827e-06,
+      "loss": 0.2227,
+      "step": 9211
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.2516965320459574,
+      "learning_rate": 5.041888613152941e-06,
+      "loss": 0.1949,
+      "step": 9212
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.228406582891852,
+      "learning_rate": 5.040987803163455e-06,
+      "loss": 0.2195,
+      "step": 9213
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.119823986148918,
+      "learning_rate": 5.040086991843481e-06,
+      "loss": 0.1803,
+      "step": 9214
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.1365929760123346,
+      "learning_rate": 5.03918617922226e-06,
+      "loss": 0.1962,
+      "step": 9215
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.1128248751637015,
+      "learning_rate": 5.038285365329032e-06,
+      "loss": 0.1778,
+      "step": 9216
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.5144343418533779,
+      "learning_rate": 5.03738455019304e-06,
+      "loss": 0.2022,
+      "step": 9217
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.4353730012084465,
+      "learning_rate": 5.036483733843524e-06,
+      "loss": 0.216,
+      "step": 9218
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.1412598854861016,
+      "learning_rate": 5.035582916309724e-06,
+      "loss": 0.1945,
+      "step": 9219
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.2092321462440014,
+      "learning_rate": 5.0346820976208845e-06,
+      "loss": 0.2411,
+      "step": 9220
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.9391875555867656,
+      "learning_rate": 5.033781277806242e-06,
+      "loss": 0.1467,
+      "step": 9221
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.3517995775261131,
+      "learning_rate": 5.032880456895042e-06,
+      "loss": 0.2093,
+      "step": 9222
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.1485989872285405,
+      "learning_rate": 5.031979634916521e-06,
+      "loss": 0.2168,
+      "step": 9223
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.2214977668684408,
+      "learning_rate": 5.0310788118999245e-06,
+      "loss": 0.1608,
+      "step": 9224
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.1543253799894932,
+      "learning_rate": 5.030177987874491e-06,
+      "loss": 0.1949,
+      "step": 9225
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.1398750445989934,
+      "learning_rate": 5.029277162869463e-06,
+      "loss": 0.1577,
+      "step": 9226
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.2768290121280648,
+      "learning_rate": 5.02837633691408e-06,
+      "loss": 0.1925,
+      "step": 9227
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.02972554467963,
+      "learning_rate": 5.027475510037586e-06,
+      "loss": 0.1477,
+      "step": 9228
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.9358270739209867,
+      "learning_rate": 5.0265746822692215e-06,
+      "loss": 0.1596,
+      "step": 9229
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.471159128467408,
+      "learning_rate": 5.025673853638227e-06,
+      "loss": 0.2588,
+      "step": 9230
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.3608110012122754,
+      "learning_rate": 5.024773024173846e-06,
+      "loss": 0.2102,
+      "step": 9231
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.9277971085871074,
+      "learning_rate": 5.023872193905316e-06,
+      "loss": 0.1455,
+      "step": 9232
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.2066064445073033,
+      "learning_rate": 5.022971362861882e-06,
+      "loss": 0.1705,
+      "step": 9233
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.1851498600505248,
+      "learning_rate": 5.0220705310727835e-06,
+      "loss": 0.1791,
+      "step": 9234
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.3400925983520702,
+      "learning_rate": 5.021169698567264e-06,
+      "loss": 0.2473,
+      "step": 9235
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.997763427378305,
+      "learning_rate": 5.020268865374563e-06,
+      "loss": 0.2098,
+      "step": 9236
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.94563313831463,
+      "learning_rate": 5.019368031523925e-06,
+      "loss": 0.2069,
+      "step": 9237
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.6809572610227215,
+      "learning_rate": 5.018467197044588e-06,
+      "loss": 0.2325,
+      "step": 9238
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.2891474811400998,
+      "learning_rate": 5.017566361965794e-06,
+      "loss": 0.1798,
+      "step": 9239
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.0177604673002434,
+      "learning_rate": 5.016665526316789e-06,
+      "loss": 0.201,
+      "step": 9240
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.1704195059608553,
+      "learning_rate": 5.0157646901268075e-06,
+      "loss": 0.1846,
+      "step": 9241
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.0795383052868686,
+      "learning_rate": 5.014863853425098e-06,
+      "loss": 0.2093,
+      "step": 9242
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.3581628979169371,
+      "learning_rate": 5.013963016240897e-06,
+      "loss": 0.2044,
+      "step": 9243
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.9043549968010642,
+      "learning_rate": 5.013062178603451e-06,
+      "loss": 0.2195,
+      "step": 9244
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.7953105697908194,
+      "learning_rate": 5.012161340541997e-06,
+      "loss": 0.1852,
+      "step": 9245
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.0371241846841155,
+      "learning_rate": 5.011260502085781e-06,
+      "loss": 0.1781,
+      "step": 9246
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.058505394193769,
+      "learning_rate": 5.010359663264042e-06,
+      "loss": 0.2457,
+      "step": 9247
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.9936702267398955,
+      "learning_rate": 5.00945882410602e-06,
+      "loss": 0.1936,
+      "step": 9248
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.1789637581289871,
+      "learning_rate": 5.0085579846409615e-06,
+      "loss": 0.214,
+      "step": 9249
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.1997238921865854,
+      "learning_rate": 5.007657144898105e-06,
+      "loss": 0.1964,
+      "step": 9250
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.1194288272152184,
+      "learning_rate": 5.0067563049066945e-06,
+      "loss": 0.1812,
+      "step": 9251
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.1185981740702864,
+      "learning_rate": 5.005855464695969e-06,
+      "loss": 0.1902,
+      "step": 9252
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.0876571089614175,
+      "learning_rate": 5.004954624295173e-06,
+      "loss": 0.1961,
+      "step": 9253
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.1653894951002897,
+      "learning_rate": 5.004053783733546e-06,
+      "loss": 0.1538,
+      "step": 9254
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.185421278709374,
+      "learning_rate": 5.0031529430403305e-06,
+      "loss": 0.168,
+      "step": 9255
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.0300557968806765,
+      "learning_rate": 5.00225210224477e-06,
+      "loss": 0.2095,
+      "step": 9256
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.1366620147940014,
+      "learning_rate": 5.001351261376104e-06,
+      "loss": 0.2358,
+      "step": 9257
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.3651182973955804,
+      "learning_rate": 5.000450420463576e-06,
+      "loss": 0.2021,
+      "step": 9258
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.9892547697752904,
+      "learning_rate": 4.9995495795364265e-06,
+      "loss": 0.1663,
+      "step": 9259
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.1612048948867773,
+      "learning_rate": 4.998648738623898e-06,
+      "loss": 0.1805,
+      "step": 9260
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.0997343483572257,
+      "learning_rate": 4.9977478977552315e-06,
+      "loss": 0.182,
+      "step": 9261
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.093816022923815,
+      "learning_rate": 4.99684705695967e-06,
+      "loss": 0.1944,
+      "step": 9262
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.34497046907748,
+      "learning_rate": 4.995946216266455e-06,
+      "loss": 0.1891,
+      "step": 9263
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.0444829625057903,
+      "learning_rate": 4.995045375704829e-06,
+      "loss": 0.178,
+      "step": 9264
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.8847898244179921,
+      "learning_rate": 4.994144535304033e-06,
+      "loss": 0.1493,
+      "step": 9265
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.4630195066876732,
+      "learning_rate": 4.993243695093307e-06,
+      "loss": 0.2073,
+      "step": 9266
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.2788542804517036,
+      "learning_rate": 4.992342855101895e-06,
+      "loss": 0.1804,
+      "step": 9267
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.2973468737844307,
+      "learning_rate": 4.99144201535904e-06,
+      "loss": 0.1752,
+      "step": 9268
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.4314427232504188,
+      "learning_rate": 4.990541175893981e-06,
+      "loss": 0.1433,
+      "step": 9269
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.3255118538065835,
+      "learning_rate": 4.98964033673596e-06,
+      "loss": 0.1472,
+      "step": 9270
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.2213016019488865,
+      "learning_rate": 4.988739497914221e-06,
+      "loss": 0.2028,
+      "step": 9271
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.2144348140083363,
+      "learning_rate": 4.987838659458003e-06,
+      "loss": 0.2164,
+      "step": 9272
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.084150648681678,
+      "learning_rate": 4.9869378213965515e-06,
+      "loss": 0.1795,
+      "step": 9273
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.1467394601885588,
+      "learning_rate": 4.9860369837591035e-06,
+      "loss": 0.1956,
+      "step": 9274
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.2437637482894448,
+      "learning_rate": 4.985136146574904e-06,
+      "loss": 0.204,
+      "step": 9275
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.1300321256121437,
+      "learning_rate": 4.9842353098731924e-06,
+      "loss": 0.2343,
+      "step": 9276
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.248222640172032,
+      "learning_rate": 4.9833344736832145e-06,
+      "loss": 0.2152,
+      "step": 9277
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.1105036350410382,
+      "learning_rate": 4.982433638034207e-06,
+      "loss": 0.196,
+      "step": 9278
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.1080220595932764,
+      "learning_rate": 4.981532802955413e-06,
+      "loss": 0.1705,
+      "step": 9279
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.149470844016851,
+      "learning_rate": 4.9806319684760776e-06,
+      "loss": 0.2177,
+      "step": 9280
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.8472663954137287,
+      "learning_rate": 4.979731134625437e-06,
+      "loss": 0.2015,
+      "step": 9281
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.2554182609798454,
+      "learning_rate": 4.978830301432738e-06,
+      "loss": 0.2322,
+      "step": 9282
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.4792096060130457,
+      "learning_rate": 4.977929468927218e-06,
+      "loss": 0.2148,
+      "step": 9283
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.509411322411356,
+      "learning_rate": 4.977028637138119e-06,
+      "loss": 0.2344,
+      "step": 9284
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.084420949638463,
+      "learning_rate": 4.976127806094685e-06,
+      "loss": 0.1932,
+      "step": 9285
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.0990218054564571,
+      "learning_rate": 4.975226975826157e-06,
+      "loss": 0.2018,
+      "step": 9286
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.1165541374059933,
+      "learning_rate": 4.974326146361774e-06,
+      "loss": 0.1872,
+      "step": 9287
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.3100834892827837,
+      "learning_rate": 4.973425317730779e-06,
+      "loss": 0.2089,
+      "step": 9288
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.0403463957258678,
+      "learning_rate": 4.9725244899624145e-06,
+      "loss": 0.1329,
+      "step": 9289
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.0446364485322261,
+      "learning_rate": 4.97162366308592e-06,
+      "loss": 0.2083,
+      "step": 9290
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.8525298212037741,
+      "learning_rate": 4.97072283713054e-06,
+      "loss": 0.1781,
+      "step": 9291
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.9183456945597006,
+      "learning_rate": 4.969822012125512e-06,
+      "loss": 0.1782,
+      "step": 9292
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.4482968320717222,
+      "learning_rate": 4.968921188100077e-06,
+      "loss": 0.2094,
+      "step": 9293
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.4946992515502369,
+      "learning_rate": 4.96802036508348e-06,
+      "loss": 0.2041,
+      "step": 9294
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.0600786255534422,
+      "learning_rate": 4.9671195431049615e-06,
+      "loss": 0.186,
+      "step": 9295
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.2425797071209994,
+      "learning_rate": 4.966218722193759e-06,
+      "loss": 0.1917,
+      "step": 9296
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.252249279492659,
+      "learning_rate": 4.965317902379116e-06,
+      "loss": 0.1595,
+      "step": 9297
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.544669124602791,
+      "learning_rate": 4.9644170836902765e-06,
+      "loss": 0.2244,
+      "step": 9298
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.1478517232144763,
+      "learning_rate": 4.963516266156477e-06,
+      "loss": 0.2159,
+      "step": 9299
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.9510970820604681,
+      "learning_rate": 4.9626154498069625e-06,
+      "loss": 0.185,
+      "step": 9300
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.8363425011022392,
+      "learning_rate": 4.96171463467097e-06,
+      "loss": 0.2092,
+      "step": 9301
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.2497585927788424,
+      "learning_rate": 4.960813820777742e-06,
+      "loss": 0.2127,
+      "step": 9302
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.1538373997628253,
+      "learning_rate": 4.95991300815652e-06,
+      "loss": 0.2011,
+      "step": 9303
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.7799504223795504,
+      "learning_rate": 4.959012196836548e-06,
+      "loss": 0.1552,
+      "step": 9304
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.298024511187678,
+      "learning_rate": 4.9581113868470605e-06,
+      "loss": 0.2263,
+      "step": 9305
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.2750403199178413,
+      "learning_rate": 4.957210578217302e-06,
+      "loss": 0.2108,
+      "step": 9306
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.3395055032514176,
+      "learning_rate": 4.956309770976514e-06,
+      "loss": 0.2001,
+      "step": 9307
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.708117556222496,
+      "learning_rate": 4.955408965153934e-06,
+      "loss": 0.1911,
+      "step": 9308
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.018751924460296,
+      "learning_rate": 4.954508160778807e-06,
+      "loss": 0.1864,
+      "step": 9309
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.237431857612324,
+      "learning_rate": 4.953607357880369e-06,
+      "loss": 0.1795,
+      "step": 9310
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.2868750913239233,
+      "learning_rate": 4.9527065564878635e-06,
+      "loss": 0.1696,
+      "step": 9311
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.165521322042238,
+      "learning_rate": 4.95180575663053e-06,
+      "loss": 0.2297,
+      "step": 9312
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.0417847109507499,
+      "learning_rate": 4.950904958337612e-06,
+      "loss": 0.1695,
+      "step": 9313
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.4314372372279118,
+      "learning_rate": 4.9500041616383446e-06,
+      "loss": 0.2383,
+      "step": 9314
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.16742683201281,
+      "learning_rate": 4.949103366561971e-06,
+      "loss": 0.2055,
+      "step": 9315
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.2526557069777027,
+      "learning_rate": 4.948202573137733e-06,
+      "loss": 0.2182,
+      "step": 9316
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.5177791958116613,
+      "learning_rate": 4.947301781394867e-06,
+      "loss": 0.1485,
+      "step": 9317
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.8387841393391972,
+      "learning_rate": 4.946400991362618e-06,
+      "loss": 0.1539,
+      "step": 9318
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.849085203249134,
+      "learning_rate": 4.945500203070222e-06,
+      "loss": 0.1311,
+      "step": 9319
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.0218147050060382,
+      "learning_rate": 4.94459941654692e-06,
+      "loss": 0.2059,
+      "step": 9320
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.3215226687961625,
+      "learning_rate": 4.9436986318219536e-06,
+      "loss": 0.1619,
+      "step": 9321
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.477692681172344,
+      "learning_rate": 4.942797848924564e-06,
+      "loss": 0.2941,
+      "step": 9322
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.648652532753549,
+      "learning_rate": 4.941897067883987e-06,
+      "loss": 0.2259,
+      "step": 9323
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.195456717399059,
+      "learning_rate": 4.940996288729464e-06,
+      "loss": 0.1688,
+      "step": 9324
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.2187596238196479,
+      "learning_rate": 4.940095511490236e-06,
+      "loss": 0.2273,
+      "step": 9325
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.451727574614885,
+      "learning_rate": 4.939194736195543e-06,
+      "loss": 0.1624,
+      "step": 9326
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.111626311860232,
+      "learning_rate": 4.938293962874625e-06,
+      "loss": 0.201,
+      "step": 9327
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.4357220419584917,
+      "learning_rate": 4.937393191556718e-06,
+      "loss": 0.1785,
+      "step": 9328
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.2639049332803203,
+      "learning_rate": 4.936492422271064e-06,
+      "loss": 0.2021,
+      "step": 9329
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 2.5485213873438983,
+      "learning_rate": 4.935591655046903e-06,
+      "loss": 0.2758,
+      "step": 9330
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 2.12748983317993,
+      "learning_rate": 4.934690889913476e-06,
+      "loss": 0.1851,
+      "step": 9331
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.0982068605149806,
+      "learning_rate": 4.9337901269000175e-06,
+      "loss": 0.1456,
+      "step": 9332
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.3236547917734607,
+      "learning_rate": 4.93288936603577e-06,
+      "loss": 0.2194,
+      "step": 9333
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.012263493819399,
+      "learning_rate": 4.931988607349973e-06,
+      "loss": 0.151,
+      "step": 9334
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.0083373591921307,
+      "learning_rate": 4.9310878508718644e-06,
+      "loss": 0.2203,
+      "step": 9335
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.9595655944695807,
+      "learning_rate": 4.930187096630685e-06,
+      "loss": 0.1797,
+      "step": 9336
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.1170589666520394,
+      "learning_rate": 4.929286344655671e-06,
+      "loss": 0.1248,
+      "step": 9337
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.1410629333175166,
+      "learning_rate": 4.928385594976063e-06,
+      "loss": 0.1915,
+      "step": 9338
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.6944762377956841,
+      "learning_rate": 4.9274848476211e-06,
+      "loss": 0.2058,
+      "step": 9339
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.4039709033842989,
+      "learning_rate": 4.926584102620023e-06,
+      "loss": 0.2005,
+      "step": 9340
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.224605344312444,
+      "learning_rate": 4.925683360002065e-06,
+      "loss": 0.2096,
+      "step": 9341
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.551051387752132,
+      "learning_rate": 4.924782619796469e-06,
+      "loss": 0.1602,
+      "step": 9342
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.1783303355064638,
+      "learning_rate": 4.923881882032472e-06,
+      "loss": 0.1933,
+      "step": 9343
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.4059866420189642,
+      "learning_rate": 4.922981146739313e-06,
+      "loss": 0.1977,
+      "step": 9344
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.1198489525654038,
+      "learning_rate": 4.922080413946232e-06,
+      "loss": 0.2088,
+      "step": 9345
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.1705786178359507,
+      "learning_rate": 4.9211796836824634e-06,
+      "loss": 0.1534,
+      "step": 9346
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.1133345239842543,
+      "learning_rate": 4.920278955977248e-06,
+      "loss": 0.2014,
+      "step": 9347
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.2369061590701467,
+      "learning_rate": 4.919378230859824e-06,
+      "loss": 0.2084,
+      "step": 9348
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.5768178611084218,
+      "learning_rate": 4.91847750835943e-06,
+      "loss": 0.2198,
+      "step": 9349
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.1866564452420025,
+      "learning_rate": 4.917576788505302e-06,
+      "loss": 0.1874,
+      "step": 9350
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.3080561102781645,
+      "learning_rate": 4.916676071326678e-06,
+      "loss": 0.1691,
+      "step": 9351
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.5008808395846869,
+      "learning_rate": 4.915775356852797e-06,
+      "loss": 0.2219,
+      "step": 9352
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.0720390573797471,
+      "learning_rate": 4.914874645112896e-06,
+      "loss": 0.1625,
+      "step": 9353
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.9601107750578328,
+      "learning_rate": 4.913973936136216e-06,
+      "loss": 0.1985,
+      "step": 9354
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.1901942701753636,
+      "learning_rate": 4.913073229951988e-06,
+      "loss": 0.2176,
+      "step": 9355
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.1609570279539105,
+      "learning_rate": 4.9121725265894534e-06,
+      "loss": 0.1969,
+      "step": 9356
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.121143089099271,
+      "learning_rate": 4.911271826077849e-06,
+      "loss": 0.2038,
+      "step": 9357
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.885368392366971,
+      "learning_rate": 4.910371128446414e-06,
+      "loss": 0.1313,
+      "step": 9358
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.963240561820967,
+      "learning_rate": 4.9094704337243825e-06,
+      "loss": 0.1713,
+      "step": 9359
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.1330956154588616,
+      "learning_rate": 4.908569741940992e-06,
+      "loss": 0.198,
+      "step": 9360
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.9907392948834709,
+      "learning_rate": 4.907669053125482e-06,
+      "loss": 0.1941,
+      "step": 9361
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.3076272484048306,
+      "learning_rate": 4.906768367307086e-06,
+      "loss": 0.2475,
+      "step": 9362
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.107917466187855,
+      "learning_rate": 4.905867684515045e-06,
+      "loss": 0.1555,
+      "step": 9363
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.1511530430304808,
+      "learning_rate": 4.904967004778591e-06,
+      "loss": 0.1589,
+      "step": 9364
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.6247053750418294,
+      "learning_rate": 4.904066328126962e-06,
+      "loss": 0.2203,
+      "step": 9365
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.0394261985654367,
+      "learning_rate": 4.903165654589396e-06,
+      "loss": 0.1923,
+      "step": 9366
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.4075291061525632,
+      "learning_rate": 4.90226498419513e-06,
+      "loss": 0.1988,
+      "step": 9367
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.7878143617849356,
+      "learning_rate": 4.901364316973397e-06,
+      "loss": 0.1661,
+      "step": 9368
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.9496201698571544,
+      "learning_rate": 4.9004636529534345e-06,
+      "loss": 0.1948,
+      "step": 9369
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.216814578311777,
+      "learning_rate": 4.89956299216448e-06,
+      "loss": 0.2354,
+      "step": 9370
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.2178054964423597,
+      "learning_rate": 4.898662334635769e-06,
+      "loss": 0.2728,
+      "step": 9371
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.0739670077031258,
+      "learning_rate": 4.8977616803965375e-06,
+      "loss": 0.1817,
+      "step": 9372
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.2940902859068306,
+      "learning_rate": 4.896861029476019e-06,
+      "loss": 0.2182,
+      "step": 9373
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.0532030361082967,
+      "learning_rate": 4.895960381903451e-06,
+      "loss": 0.1537,
+      "step": 9374
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.226116697729359,
+      "learning_rate": 4.895059737708069e-06,
+      "loss": 0.219,
+      "step": 9375
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.3309282605252881,
+      "learning_rate": 4.894159096919109e-06,
+      "loss": 0.1798,
+      "step": 9376
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.0811227158504566,
+      "learning_rate": 4.893258459565805e-06,
+      "loss": 0.1805,
+      "step": 9377
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.1325519253853538,
+      "learning_rate": 4.8923578256773915e-06,
+      "loss": 0.1661,
+      "step": 9378
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.0679603019609045,
+      "learning_rate": 4.891457195283107e-06,
+      "loss": 0.1611,
+      "step": 9379
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.4067967994220896,
+      "learning_rate": 4.890556568412183e-06,
+      "loss": 0.2395,
+      "step": 9380
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.237844203513116,
+      "learning_rate": 4.889655945093858e-06,
+      "loss": 0.2,
+      "step": 9381
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.5073886433714758,
+      "learning_rate": 4.8887553253573625e-06,
+      "loss": 0.2182,
+      "step": 9382
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.3266585444446555,
+      "learning_rate": 4.887854709231932e-06,
+      "loss": 0.2332,
+      "step": 9383
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.4134129644731757,
+      "learning_rate": 4.886954096746803e-06,
+      "loss": 0.1848,
+      "step": 9384
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.0314016140401783,
+      "learning_rate": 4.886053487931212e-06,
+      "loss": 0.1359,
+      "step": 9385
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.0733298698894131,
+      "learning_rate": 4.885152882814387e-06,
+      "loss": 0.1787,
+      "step": 9386
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.2769343131177164,
+      "learning_rate": 4.8842522814255654e-06,
+      "loss": 0.1901,
+      "step": 9387
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.8203641466503633,
+      "learning_rate": 4.883351683793982e-06,
+      "loss": 0.1629,
+      "step": 9388
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.8832665263335026,
+      "learning_rate": 4.88245108994887e-06,
+      "loss": 0.1878,
+      "step": 9389
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.1195846961468259,
+      "learning_rate": 4.881550499919465e-06,
+      "loss": 0.2127,
+      "step": 9390
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.1275017839061345,
+      "learning_rate": 4.880649913734996e-06,
+      "loss": 0.2249,
+      "step": 9391
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.3206484750871892,
+      "learning_rate": 4.879749331424701e-06,
+      "loss": 0.2,
+      "step": 9392
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.9691839399886767,
+      "learning_rate": 4.8788487530178105e-06,
+      "loss": 0.1798,
+      "step": 9393
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.257700352378311,
+      "learning_rate": 4.877948178543563e-06,
+      "loss": 0.1933,
+      "step": 9394
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.0629981359849452,
+      "learning_rate": 4.877047608031184e-06,
+      "loss": 0.2114,
+      "step": 9395
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.34930752471543,
+      "learning_rate": 4.87614704150991e-06,
+      "loss": 0.199,
+      "step": 9396
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.4876296088609107,
+      "learning_rate": 4.875246479008977e-06,
+      "loss": 0.2035,
+      "step": 9397
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.0050574207508483,
+      "learning_rate": 4.874345920557613e-06,
+      "loss": 0.1987,
+      "step": 9398
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.1092544846152834,
+      "learning_rate": 4.873445366185054e-06,
+      "loss": 0.2284,
+      "step": 9399
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.1833452445410713,
+      "learning_rate": 4.872544815920531e-06,
+      "loss": 0.2088,
+      "step": 9400
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.3071779620228212,
+      "learning_rate": 4.871644269793276e-06,
+      "loss": 0.2174,
+      "step": 9401
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.9411194174418948,
+      "learning_rate": 4.8707437278325204e-06,
+      "loss": 0.1363,
+      "step": 9402
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.0082939092295709,
+      "learning_rate": 4.869843190067501e-06,
+      "loss": 0.1698,
+      "step": 9403
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.8413831419857559,
+      "learning_rate": 4.868942656527444e-06,
+      "loss": 0.1865,
+      "step": 9404
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.0903619761269157,
+      "learning_rate": 4.868042127241584e-06,
+      "loss": 0.1946,
+      "step": 9405
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.9630887676902193,
+      "learning_rate": 4.867141602239153e-06,
+      "loss": 0.2036,
+      "step": 9406
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.081337614347852,
+      "learning_rate": 4.866241081549382e-06,
+      "loss": 0.156,
+      "step": 9407
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.1893122923211992,
+      "learning_rate": 4.8653405652015035e-06,
+      "loss": 0.1487,
+      "step": 9408
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.2589207721638191,
+      "learning_rate": 4.864440053224746e-06,
+      "loss": 0.1642,
+      "step": 9409
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.0543412134571986,
+      "learning_rate": 4.863539545648342e-06,
+      "loss": 0.2146,
+      "step": 9410
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.7948755712402873,
+      "learning_rate": 4.862639042501523e-06,
+      "loss": 0.1825,
+      "step": 9411
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.453176815380711,
+      "learning_rate": 4.861738543813522e-06,
+      "loss": 0.1933,
+      "step": 9412
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.9332415303390876,
+      "learning_rate": 4.860838049613565e-06,
+      "loss": 0.2248,
+      "step": 9413
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.0297655526072826,
+      "learning_rate": 4.859937559930885e-06,
+      "loss": 0.2288,
+      "step": 9414
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.0046523907861502,
+      "learning_rate": 4.859037074794713e-06,
+      "loss": 0.1946,
+      "step": 9415
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.1969214609032683,
+      "learning_rate": 4.858136594234279e-06,
+      "loss": 0.2155,
+      "step": 9416
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.8870643034839861,
+      "learning_rate": 4.857236118278813e-06,
+      "loss": 0.1962,
+      "step": 9417
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.2978920998424568,
+      "learning_rate": 4.856335646957544e-06,
+      "loss": 0.1758,
+      "step": 9418
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.4297286976157637,
+      "learning_rate": 4.855435180299702e-06,
+      "loss": 0.2491,
+      "step": 9419
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.2300116749704655,
+      "learning_rate": 4.854534718334517e-06,
+      "loss": 0.1729,
+      "step": 9420
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.0204890281372105,
+      "learning_rate": 4.8536342610912204e-06,
+      "loss": 0.1581,
+      "step": 9421
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.1033996791186411,
+      "learning_rate": 4.852733808599039e-06,
+      "loss": 0.224,
+      "step": 9422
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.105230233237332,
+      "learning_rate": 4.8518333608872015e-06,
+      "loss": 0.1969,
+      "step": 9423
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.2080261919848396,
+      "learning_rate": 4.85093291798494e-06,
+      "loss": 0.2426,
+      "step": 9424
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.8805247197643998,
+      "learning_rate": 4.85003247992148e-06,
+      "loss": 0.2108,
+      "step": 9425
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.113719519295482,
+      "learning_rate": 4.849132046726056e-06,
+      "loss": 0.1764,
+      "step": 9426
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.9458230507847678,
+      "learning_rate": 4.848231618427889e-06,
+      "loss": 0.2023,
+      "step": 9427
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.3451849710848087,
+      "learning_rate": 4.84733119505621e-06,
+      "loss": 0.2179,
+      "step": 9428
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.260236634317522,
+      "learning_rate": 4.84643077664025e-06,
+      "loss": 0.2088,
+      "step": 9429
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.3995174947396791,
+      "learning_rate": 4.845530363209236e-06,
+      "loss": 0.1897,
+      "step": 9430
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.1478079218452477,
+      "learning_rate": 4.8446299547923944e-06,
+      "loss": 0.1976,
+      "step": 9431
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.2285816344704796,
+      "learning_rate": 4.843729551418953e-06,
+      "loss": 0.1948,
+      "step": 9432
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.032738240294867,
+      "learning_rate": 4.8428291531181406e-06,
+      "loss": 0.1608,
+      "step": 9433
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.2991641592277243,
+      "learning_rate": 4.841928759919184e-06,
+      "loss": 0.1907,
+      "step": 9434
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.2980176662932161,
+      "learning_rate": 4.8410283718513134e-06,
+      "loss": 0.2333,
+      "step": 9435
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.0404337281925087,
+      "learning_rate": 4.8401279889437505e-06,
+      "loss": 0.1607,
+      "step": 9436
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.33620109383634,
+      "learning_rate": 4.8392276112257255e-06,
+      "loss": 0.2229,
+      "step": 9437
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.4707878957181055,
+      "learning_rate": 4.838327238726464e-06,
+      "loss": 0.1847,
+      "step": 9438
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.0281232172217174,
+      "learning_rate": 4.837426871475196e-06,
+      "loss": 0.143,
+      "step": 9439
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.1971447122824397,
+      "learning_rate": 4.8365265095011425e-06,
+      "loss": 0.219,
+      "step": 9440
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.2743138357921913,
+      "learning_rate": 4.835626152833534e-06,
+      "loss": 0.2218,
+      "step": 9441
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.2474313002888666,
+      "learning_rate": 4.834725801501595e-06,
+      "loss": 0.1496,
+      "step": 9442
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.7526399349284005,
+      "learning_rate": 4.83382545553455e-06,
+      "loss": 0.2178,
+      "step": 9443
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.4135813843013483,
+      "learning_rate": 4.832925114961629e-06,
+      "loss": 0.1918,
+      "step": 9444
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.9192372445173911,
+      "learning_rate": 4.832024779812052e-06,
+      "loss": 0.1662,
+      "step": 9445
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.3161190976244128,
+      "learning_rate": 4.831124450115048e-06,
+      "loss": 0.2068,
+      "step": 9446
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.224756922495368,
+      "learning_rate": 4.830224125899841e-06,
+      "loss": 0.1762,
+      "step": 9447
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.3356027685733078,
+      "learning_rate": 4.82932380719566e-06,
+      "loss": 0.2664,
+      "step": 9448
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.3590338021171766,
+      "learning_rate": 4.828423494031723e-06,
+      "loss": 0.1858,
+      "step": 9449
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.829246254506679,
+      "learning_rate": 4.827523186437258e-06,
+      "loss": 0.1634,
+      "step": 9450
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.9363715672192625,
+      "learning_rate": 4.826622884441491e-06,
+      "loss": 0.1703,
+      "step": 9451
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.8384930261381355,
+      "learning_rate": 4.825722588073643e-06,
+      "loss": 0.1373,
+      "step": 9452
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.1164138756312054,
+      "learning_rate": 4.824822297362943e-06,
+      "loss": 0.1921,
+      "step": 9453
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.800642489994916,
+      "learning_rate": 4.82392201233861e-06,
+      "loss": 0.1439,
+      "step": 9454
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.313836702798671,
+      "learning_rate": 4.823021733029869e-06,
+      "loss": 0.207,
+      "step": 9455
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.263430137447707,
+      "learning_rate": 4.822121459465945e-06,
+      "loss": 0.1838,
+      "step": 9456
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.828651096917504,
+      "learning_rate": 4.821221191676063e-06,
+      "loss": 0.1708,
+      "step": 9457
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.0959086403909732,
+      "learning_rate": 4.820320929689441e-06,
+      "loss": 0.1618,
+      "step": 9458
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.2541965314988845,
+      "learning_rate": 4.819420673535305e-06,
+      "loss": 0.1563,
+      "step": 9459
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.1951099205388402,
+      "learning_rate": 4.81852042324288e-06,
+      "loss": 0.2232,
+      "step": 9460
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.0689694338278974,
+      "learning_rate": 4.817620178841384e-06,
+      "loss": 0.1623,
+      "step": 9461
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.6812481926266623,
+      "learning_rate": 4.816719940360045e-06,
+      "loss": 0.1645,
+      "step": 9462
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.9222555678213892,
+      "learning_rate": 4.815819707828079e-06,
+      "loss": 0.1596,
+      "step": 9463
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.2297842665070586,
+      "learning_rate": 4.8149194812747115e-06,
+      "loss": 0.2194,
+      "step": 9464
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.410264869585514,
+      "learning_rate": 4.814019260729164e-06,
+      "loss": 0.1946,
+      "step": 9465
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.2033899483068793,
+      "learning_rate": 4.813119046220661e-06,
+      "loss": 0.1481,
+      "step": 9466
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.2879338373496434,
+      "learning_rate": 4.812218837778418e-06,
+      "loss": 0.2141,
+      "step": 9467
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.2140747838030992,
+      "learning_rate": 4.811318635431661e-06,
+      "loss": 0.2257,
+      "step": 9468
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.342875745170349,
+      "learning_rate": 4.8104184392096095e-06,
+      "loss": 0.2421,
+      "step": 9469
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.722377022454115,
+      "learning_rate": 4.809518249141484e-06,
+      "loss": 0.1973,
+      "step": 9470
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.2081308757049036,
+      "learning_rate": 4.808618065256508e-06,
+      "loss": 0.1787,
+      "step": 9471
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.8108589804563466,
+      "learning_rate": 4.807717887583896e-06,
+      "loss": 0.2288,
+      "step": 9472
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.998464810381446,
+      "learning_rate": 4.806817716152873e-06,
+      "loss": 0.1892,
+      "step": 9473
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.033658660961054,
+      "learning_rate": 4.805917550992659e-06,
+      "loss": 0.2178,
+      "step": 9474
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.3392284832374723,
+      "learning_rate": 4.8050173921324736e-06,
+      "loss": 0.2145,
+      "step": 9475
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.0528767087893776,
+      "learning_rate": 4.804117239601534e-06,
+      "loss": 0.18,
+      "step": 9476
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.1440326740682525,
+      "learning_rate": 4.803217093429062e-06,
+      "loss": 0.1998,
+      "step": 9477
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.2582554762784766,
+      "learning_rate": 4.802316953644277e-06,
+      "loss": 0.17,
+      "step": 9478
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.2060759137096573,
+      "learning_rate": 4.801416820276396e-06,
+      "loss": 0.1837,
+      "step": 9479
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.0155227572989303,
+      "learning_rate": 4.800516693354641e-06,
+      "loss": 0.2253,
+      "step": 9480
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.266483202825419,
+      "learning_rate": 4.799616572908228e-06,
+      "loss": 0.2224,
+      "step": 9481
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.0120830193094814,
+      "learning_rate": 4.798716458966375e-06,
+      "loss": 0.182,
+      "step": 9482
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 5.224040108443509,
+      "learning_rate": 4.797816351558303e-06,
+      "loss": 0.1735,
+      "step": 9483
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.421614919183648,
+      "learning_rate": 4.79691625071323e-06,
+      "loss": 0.2414,
+      "step": 9484
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.1346370702992241,
+      "learning_rate": 4.796016156460369e-06,
+      "loss": 0.2021,
+      "step": 9485
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.1321724135614841,
+      "learning_rate": 4.7951160688289415e-06,
+      "loss": 0.1726,
+      "step": 9486
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.144622349937948,
+      "learning_rate": 4.794215987848166e-06,
+      "loss": 0.2191,
+      "step": 9487
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.9539476307571764,
+      "learning_rate": 4.793315913547256e-06,
+      "loss": 0.1742,
+      "step": 9488
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.1548964948793636,
+      "learning_rate": 4.792415845955433e-06,
+      "loss": 0.2444,
+      "step": 9489
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.318900301128665,
+      "learning_rate": 4.791515785101908e-06,
+      "loss": 0.1687,
+      "step": 9490
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.9363167185407693,
+      "learning_rate": 4.790615731015901e-06,
+      "loss": 0.2009,
+      "step": 9491
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.4011346816865877,
+      "learning_rate": 4.7897156837266275e-06,
+      "loss": 0.1651,
+      "step": 9492
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.5534144993308818,
+      "learning_rate": 4.788815643263307e-06,
+      "loss": 0.1894,
+      "step": 9493
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.9558357868783907,
+      "learning_rate": 4.787915609655149e-06,
+      "loss": 0.189,
+      "step": 9494
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.8588524161510468,
+      "learning_rate": 4.787015582931373e-06,
+      "loss": 0.2101,
+      "step": 9495
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.348181401086733,
+      "learning_rate": 4.786115563121194e-06,
+      "loss": 0.2246,
+      "step": 9496
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.1199071848597417,
+      "learning_rate": 4.785215550253826e-06,
+      "loss": 0.1792,
+      "step": 9497
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.4095826012476242,
+      "learning_rate": 4.784315544358488e-06,
+      "loss": 0.2071,
+      "step": 9498
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.0921913417044598,
+      "learning_rate": 4.783415545464387e-06,
+      "loss": 0.2271,
+      "step": 9499
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.1026891871405207,
+      "learning_rate": 4.782515553600744e-06,
+      "loss": 0.1691,
+      "step": 9500
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.8224127517785562,
+      "learning_rate": 4.78161556879677e-06,
+      "loss": 0.1764,
+      "step": 9501
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.0231098413116357,
+      "learning_rate": 4.780715591081683e-06,
+      "loss": 0.1945,
+      "step": 9502
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.1769041421536668,
+      "learning_rate": 4.779815620484692e-06,
+      "loss": 0.1906,
+      "step": 9503
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.1445937741748509,
+      "learning_rate": 4.778915657035012e-06,
+      "loss": 0.1804,
+      "step": 9504
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.5189613362041394,
+      "learning_rate": 4.778015700761858e-06,
+      "loss": 0.173,
+      "step": 9505
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.2093047825194767,
+      "learning_rate": 4.777115751694441e-06,
+      "loss": 0.2055,
+      "step": 9506
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.3356552244085937,
+      "learning_rate": 4.776215809861977e-06,
+      "loss": 0.205,
+      "step": 9507
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.8689793404938584,
+      "learning_rate": 4.775315875293675e-06,
+      "loss": 0.1344,
+      "step": 9508
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.2931920921838171,
+      "learning_rate": 4.7744159480187485e-06,
+      "loss": 0.2233,
+      "step": 9509
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.7770819294961994,
+      "learning_rate": 4.773516028066409e-06,
+      "loss": 0.1769,
+      "step": 9510
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.5091328374325337,
+      "learning_rate": 4.772616115465872e-06,
+      "loss": 0.2366,
+      "step": 9511
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.1249215840455125,
+      "learning_rate": 4.7717162102463465e-06,
+      "loss": 0.198,
+      "step": 9512
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.1138754497036363,
+      "learning_rate": 4.770816312437043e-06,
+      "loss": 0.2187,
+      "step": 9513
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.9617011722337763,
+      "learning_rate": 4.769916422067174e-06,
+      "loss": 0.2113,
+      "step": 9514
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.1355427462749137,
+      "learning_rate": 4.76901653916595e-06,
+      "loss": 0.1891,
+      "step": 9515
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.9776228650629705,
+      "learning_rate": 4.768116663762584e-06,
+      "loss": 0.1889,
+      "step": 9516
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.0073935410448478,
+      "learning_rate": 4.767216795886281e-06,
+      "loss": 0.1662,
+      "step": 9517
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.9464818051531986,
+      "learning_rate": 4.7663169355662565e-06,
+      "loss": 0.1713,
+      "step": 9518
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.1502635850043,
+      "learning_rate": 4.765417082831718e-06,
+      "loss": 0.1507,
+      "step": 9519
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.9178475989890007,
+      "learning_rate": 4.764517237711877e-06,
+      "loss": 0.1624,
+      "step": 9520
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.827633840280214,
+      "learning_rate": 4.763617400235941e-06,
+      "loss": 0.1508,
+      "step": 9521
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.3039112494380272,
+      "learning_rate": 4.762717570433121e-06,
+      "loss": 0.1966,
+      "step": 9522
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.0697624157352015,
+      "learning_rate": 4.761817748332624e-06,
+      "loss": 0.2039,
+      "step": 9523
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.1163341490925858,
+      "learning_rate": 4.760917933963659e-06,
+      "loss": 0.1807,
+      "step": 9524
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.245064851745526,
+      "learning_rate": 4.760018127355439e-06,
+      "loss": 0.1426,
+      "step": 9525
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.1204959425248455,
+      "learning_rate": 4.759118328537166e-06,
+      "loss": 0.2378,
+      "step": 9526
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.0127631250261715,
+      "learning_rate": 4.75821853753805e-06,
+      "loss": 0.1965,
+      "step": 9527
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.19291382914882,
+      "learning_rate": 4.757318754387299e-06,
+      "loss": 0.1978,
+      "step": 9528
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.62872261045925,
+      "learning_rate": 4.756418979114123e-06,
+      "loss": 0.2164,
+      "step": 9529
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.1516022413945406,
+      "learning_rate": 4.755519211747726e-06,
+      "loss": 0.2156,
+      "step": 9530
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.8238926305609188,
+      "learning_rate": 4.754619452317316e-06,
+      "loss": 0.2092,
+      "step": 9531
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.2775080834700343,
+      "learning_rate": 4.7537197008521e-06,
+      "loss": 0.1778,
+      "step": 9532
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.3656297937872088,
+      "learning_rate": 4.7528199573812825e-06,
+      "loss": 0.2403,
+      "step": 9533
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.7939868015273622,
+      "learning_rate": 4.751920221934075e-06,
+      "loss": 0.19,
+      "step": 9534
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.319954444086045,
+      "learning_rate": 4.751020494539676e-06,
+      "loss": 0.2225,
+      "step": 9535
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.0502485029099924,
+      "learning_rate": 4.7501207752272955e-06,
+      "loss": 0.1965,
+      "step": 9536
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.7607381040977483,
+      "learning_rate": 4.749221064026138e-06,
+      "loss": 0.2326,
+      "step": 9537
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.390999170391883,
+      "learning_rate": 4.748321360965411e-06,
+      "loss": 0.1819,
+      "step": 9538
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.3106124082687176,
+      "learning_rate": 4.747421666074316e-06,
+      "loss": 0.1715,
+      "step": 9539
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.0637951628492188,
+      "learning_rate": 4.7465219793820606e-06,
+      "loss": 0.1884,
+      "step": 9540
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.2983712188613863,
+      "learning_rate": 4.745622300917846e-06,
+      "loss": 0.1956,
+      "step": 9541
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.111725667367674,
+      "learning_rate": 4.7447226307108764e-06,
+      "loss": 0.199,
+      "step": 9542
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.1482345456866536,
+      "learning_rate": 4.743822968790361e-06,
+      "loss": 0.1883,
+      "step": 9543
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.515984585962027,
+      "learning_rate": 4.742923315185497e-06,
+      "loss": 0.2074,
+      "step": 9544
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.1604113442842798,
+      "learning_rate": 4.742023669925488e-06,
+      "loss": 0.2048,
+      "step": 9545
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.320926733586764,
+      "learning_rate": 4.741124033039541e-06,
+      "loss": 0.1597,
+      "step": 9546
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.1771474486024538,
+      "learning_rate": 4.740224404556857e-06,
+      "loss": 0.159,
+      "step": 9547
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.205074705987714,
+      "learning_rate": 4.739324784506639e-06,
+      "loss": 0.1859,
+      "step": 9548
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.9725223989193322,
+      "learning_rate": 4.738425172918088e-06,
+      "loss": 0.168,
+      "step": 9549
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.2211196387189176,
+      "learning_rate": 4.737525569820405e-06,
+      "loss": 0.2138,
+      "step": 9550
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.3115328989376456,
+      "learning_rate": 4.736625975242792e-06,
+      "loss": 0.1454,
+      "step": 9551
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.0303889024584014,
+      "learning_rate": 4.735726389214455e-06,
+      "loss": 0.1939,
+      "step": 9552
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.4005452666563611,
+      "learning_rate": 4.734826811764589e-06,
+      "loss": 0.1706,
+      "step": 9553
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.8423674877673517,
+      "learning_rate": 4.733927242922396e-06,
+      "loss": 0.1911,
+      "step": 9554
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.482067534475608,
+      "learning_rate": 4.733027682717079e-06,
+      "loss": 0.2123,
+      "step": 9555
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.3108679012969569,
+      "learning_rate": 4.732128131177838e-06,
+      "loss": 0.2005,
+      "step": 9556
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.0165129484605826,
+      "learning_rate": 4.731228588333871e-06,
+      "loss": 0.211,
+      "step": 9557
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.0069639714675203,
+      "learning_rate": 4.730329054214379e-06,
+      "loss": 0.1572,
+      "step": 9558
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.0597908168530543,
+      "learning_rate": 4.7294295288485605e-06,
+      "loss": 0.1675,
+      "step": 9559
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.8930334718115733,
+      "learning_rate": 4.728530012265614e-06,
+      "loss": 0.1673,
+      "step": 9560
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.1046018470136938,
+      "learning_rate": 4.727630504494743e-06,
+      "loss": 0.187,
+      "step": 9561
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.1418168686412502,
+      "learning_rate": 4.726731005565138e-06,
+      "loss": 0.2341,
+      "step": 9562
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.4940911278598037,
+      "learning_rate": 4.725831515506004e-06,
+      "loss": 0.2052,
+      "step": 9563
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.3447917847617306,
+      "learning_rate": 4.724932034346536e-06,
+      "loss": 0.2278,
+      "step": 9564
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.204723788307355,
+      "learning_rate": 4.724032562115933e-06,
+      "loss": 0.2319,
+      "step": 9565
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.1238777802031696,
+      "learning_rate": 4.723133098843392e-06,
+      "loss": 0.253,
+      "step": 9566
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.5183817553410364,
+      "learning_rate": 4.722233644558109e-06,
+      "loss": 0.2272,
+      "step": 9567
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.0792237501745179,
+      "learning_rate": 4.721334199289281e-06,
+      "loss": 0.2088,
+      "step": 9568
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.2990431041200983,
+      "learning_rate": 4.7204347630661056e-06,
+      "loss": 0.1757,
+      "step": 9569
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.0393964402928741,
+      "learning_rate": 4.719535335917781e-06,
+      "loss": 0.1902,
+      "step": 9570
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.362859012940165,
+      "learning_rate": 4.718635917873497e-06,
+      "loss": 0.2047,
+      "step": 9571
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.1619604832771409,
+      "learning_rate": 4.717736508962454e-06,
+      "loss": 0.2196,
+      "step": 9572
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.202534368065027,
+      "learning_rate": 4.716837109213846e-06,
+      "loss": 0.2208,
+      "step": 9573
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.974240264677956,
+      "learning_rate": 4.71593771865687e-06,
+      "loss": 0.2046,
+      "step": 9574
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.8272976025831353,
+      "learning_rate": 4.715038337320719e-06,
+      "loss": 0.1923,
+      "step": 9575
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.0074317678710545,
+      "learning_rate": 4.714138965234587e-06,
+      "loss": 0.2283,
+      "step": 9576
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.9704652913206516,
+      "learning_rate": 4.713239602427667e-06,
+      "loss": 0.1801,
+      "step": 9577
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.399441863194473,
+      "learning_rate": 4.7123402489291546e-06,
+      "loss": 0.2059,
+      "step": 9578
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.0918035276684557,
+      "learning_rate": 4.711440904768246e-06,
+      "loss": 0.2105,
+      "step": 9579
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.3619929979131395,
+      "learning_rate": 4.710541569974129e-06,
+      "loss": 0.2172,
+      "step": 9580
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.0727239629762342,
+      "learning_rate": 4.709642244575999e-06,
+      "loss": 0.2075,
+      "step": 9581
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.1255474679076458,
+      "learning_rate": 4.708742928603049e-06,
+      "loss": 0.1495,
+      "step": 9582
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.2380403290324773,
+      "learning_rate": 4.707843622084471e-06,
+      "loss": 0.2064,
+      "step": 9583
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.2555917896779627,
+      "learning_rate": 4.706944325049458e-06,
+      "loss": 0.1779,
+      "step": 9584
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.6995162584007589,
+      "learning_rate": 4.7060450375272e-06,
+      "loss": 0.1408,
+      "step": 9585
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.5845555852456947,
+      "learning_rate": 4.705145759546889e-06,
+      "loss": 0.1942,
+      "step": 9586
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.597703545949902,
+      "learning_rate": 4.7042464911377165e-06,
+      "loss": 0.179,
+      "step": 9587
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.2910769385986431,
+      "learning_rate": 4.703347232328874e-06,
+      "loss": 0.1942,
+      "step": 9588
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.404101375399938,
+      "learning_rate": 4.702447983149549e-06,
+      "loss": 0.1821,
+      "step": 9589
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.4698069971829704,
+      "learning_rate": 4.701548743628934e-06,
+      "loss": 0.2205,
+      "step": 9590
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.2911153573023182,
+      "learning_rate": 4.7006495137962185e-06,
+      "loss": 0.1938,
+      "step": 9591
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.082407245654141,
+      "learning_rate": 4.699750293680592e-06,
+      "loss": 0.2167,
+      "step": 9592
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.451341893747394,
+      "learning_rate": 4.6988510833112445e-06,
+      "loss": 0.1814,
+      "step": 9593
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.4388349284800215,
+      "learning_rate": 4.697951882717364e-06,
+      "loss": 0.2244,
+      "step": 9594
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.1094018838875563,
+      "learning_rate": 4.697052691928138e-06,
+      "loss": 0.1595,
+      "step": 9595
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.3002276975412919,
+      "learning_rate": 4.696153510972755e-06,
+      "loss": 0.2237,
+      "step": 9596
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.4910920704319905,
+      "learning_rate": 4.695254339880407e-06,
+      "loss": 0.2328,
+      "step": 9597
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.0044623997837103,
+      "learning_rate": 4.694355178680275e-06,
+      "loss": 0.1902,
+      "step": 9598
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.0543867106146323,
+      "learning_rate": 4.693456027401551e-06,
+      "loss": 0.2445,
+      "step": 9599
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.4026663244509667,
+      "learning_rate": 4.69255688607342e-06,
+      "loss": 0.1869,
+      "step": 9600
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.2549078259776771,
+      "learning_rate": 4.6916577547250705e-06,
+      "loss": 0.2168,
+      "step": 9601
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.2736703829407645,
+      "learning_rate": 4.690758633385687e-06,
+      "loss": 0.247,
+      "step": 9602
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.274529433600462,
+      "learning_rate": 4.689859522084457e-06,
+      "loss": 0.1811,
+      "step": 9603
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.3433979746005884,
+      "learning_rate": 4.688960420850564e-06,
+      "loss": 0.2151,
+      "step": 9604
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.0668287496997524,
+      "learning_rate": 4.688061329713194e-06,
+      "loss": 0.1847,
+      "step": 9605
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.0793368881269445,
+      "learning_rate": 4.687162248701536e-06,
+      "loss": 0.1708,
+      "step": 9606
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.5360401486957078,
+      "learning_rate": 4.686263177844767e-06,
+      "loss": 0.2075,
+      "step": 9607
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.1667967580815026,
+      "learning_rate": 4.6853641171720775e-06,
+      "loss": 0.1796,
+      "step": 9608
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.7268466996052958,
+      "learning_rate": 4.684465066712649e-06,
+      "loss": 0.2341,
+      "step": 9609
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.2470167323277026,
+      "learning_rate": 4.683566026495667e-06,
+      "loss": 0.2274,
+      "step": 9610
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.1836667006896107,
+      "learning_rate": 4.682666996550314e-06,
+      "loss": 0.1671,
+      "step": 9611
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.3046496779847314,
+      "learning_rate": 4.681767976905773e-06,
+      "loss": 0.1798,
+      "step": 9612
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.097026524680963,
+      "learning_rate": 4.680868967591224e-06,
+      "loss": 0.1722,
+      "step": 9613
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.3686843470935974,
+      "learning_rate": 4.679969968635854e-06,
+      "loss": 0.2282,
+      "step": 9614
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.0342109826731796,
+      "learning_rate": 4.6790709800688434e-06,
+      "loss": 0.1948,
+      "step": 9615
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.4084548672269106,
+      "learning_rate": 4.678172001919371e-06,
+      "loss": 0.2461,
+      "step": 9616
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.0364823807829235,
+      "learning_rate": 4.677273034216621e-06,
+      "loss": 0.187,
+      "step": 9617
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.2466684342118035,
+      "learning_rate": 4.676374076989775e-06,
+      "loss": 0.2152,
+      "step": 9618
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.8258303066955477,
+      "learning_rate": 4.675475130268012e-06,
+      "loss": 0.1395,
+      "step": 9619
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.1864207815965633,
+      "learning_rate": 4.674576194080514e-06,
+      "loss": 0.1642,
+      "step": 9620
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.1069582114828709,
+      "learning_rate": 4.673677268456459e-06,
+      "loss": 0.1764,
+      "step": 9621
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.515090117996439,
+      "learning_rate": 4.672778353425027e-06,
+      "loss": 0.2463,
+      "step": 9622
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.1845288202577606,
+      "learning_rate": 4.671879449015398e-06,
+      "loss": 0.2284,
+      "step": 9623
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.102449768275287,
+      "learning_rate": 4.670980555256752e-06,
+      "loss": 0.2064,
+      "step": 9624
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.3263501958948138,
+      "learning_rate": 4.670081672178265e-06,
+      "loss": 0.177,
+      "step": 9625
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.00459766327418,
+      "learning_rate": 4.6691827998091165e-06,
+      "loss": 0.1541,
+      "step": 9626
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.5495368807860475,
+      "learning_rate": 4.668283938178484e-06,
+      "loss": 0.2598,
+      "step": 9627
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.923724336015996,
+      "learning_rate": 4.667385087315547e-06,
+      "loss": 0.2037,
+      "step": 9628
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.0688906086342018,
+      "learning_rate": 4.666486247249482e-06,
+      "loss": 0.2022,
+      "step": 9629
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.2430990521257885,
+      "learning_rate": 4.665587418009464e-06,
+      "loss": 0.1831,
+      "step": 9630
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.7247898670501927,
+      "learning_rate": 4.66468859962467e-06,
+      "loss": 0.2277,
+      "step": 9631
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.3238137471208153,
+      "learning_rate": 4.663789792124278e-06,
+      "loss": 0.1637,
+      "step": 9632
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.134900758021248,
+      "learning_rate": 4.662890995537464e-06,
+      "loss": 0.1706,
+      "step": 9633
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.9417964998985755,
+      "learning_rate": 4.661992209893399e-06,
+      "loss": 0.1948,
+      "step": 9634
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.5846924827683906,
+      "learning_rate": 4.661093435221263e-06,
+      "loss": 0.2273,
+      "step": 9635
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.951368679015134,
+      "learning_rate": 4.660194671550228e-06,
+      "loss": 0.1799,
+      "step": 9636
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.8855843534813419,
+      "learning_rate": 4.659295918909471e-06,
+      "loss": 0.1784,
+      "step": 9637
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.022684850649802,
+      "learning_rate": 4.658397177328164e-06,
+      "loss": 0.1617,
+      "step": 9638
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.3536561518678625,
+      "learning_rate": 4.6574984468354815e-06,
+      "loss": 0.1879,
+      "step": 9639
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.3613631696856023,
+      "learning_rate": 4.6565997274605945e-06,
+      "loss": 0.213,
+      "step": 9640
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.3363744748866224,
+      "learning_rate": 4.655701019232678e-06,
+      "loss": 0.188,
+      "step": 9641
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.3474812214291334,
+      "learning_rate": 4.654802322180907e-06,
+      "loss": 0.2077,
+      "step": 9642
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.071244683124519,
+      "learning_rate": 4.653903636334449e-06,
+      "loss": 0.1545,
+      "step": 9643
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.4807774893686863,
+      "learning_rate": 4.6530049617224775e-06,
+      "loss": 0.2117,
+      "step": 9644
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.108252995565287,
+      "learning_rate": 4.652106298374165e-06,
+      "loss": 0.1839,
+      "step": 9645
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.102090108944677,
+      "learning_rate": 4.651207646318682e-06,
+      "loss": 0.1801,
+      "step": 9646
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.2444398256738713,
+      "learning_rate": 4.6503090055852e-06,
+      "loss": 0.1813,
+      "step": 9647
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.0555586749042605,
+      "learning_rate": 4.64941037620289e-06,
+      "loss": 0.2368,
+      "step": 9648
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.3551362891427647,
+      "learning_rate": 4.648511758200918e-06,
+      "loss": 0.2256,
+      "step": 9649
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.9458123729745805,
+      "learning_rate": 4.647613151608458e-06,
+      "loss": 0.1933,
+      "step": 9650
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.244623762608731,
+      "learning_rate": 4.646714556454679e-06,
+      "loss": 0.175,
+      "step": 9651
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.3805754880976786,
+      "learning_rate": 4.645815972768746e-06,
+      "loss": 0.1958,
+      "step": 9652
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.243305379562208,
+      "learning_rate": 4.644917400579831e-06,
+      "loss": 0.1855,
+      "step": 9653
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.671311097324699,
+      "learning_rate": 4.644018839917101e-06,
+      "loss": 0.1977,
+      "step": 9654
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.0283916155229909,
+      "learning_rate": 4.643120290809725e-06,
+      "loss": 0.2298,
+      "step": 9655
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.1893877221277513,
+      "learning_rate": 4.64222175328687e-06,
+      "loss": 0.1896,
+      "step": 9656
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.0213694480444677,
+      "learning_rate": 4.6413232273777034e-06,
+      "loss": 0.1894,
+      "step": 9657
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.2870769583479478,
+      "learning_rate": 4.640424713111389e-06,
+      "loss": 0.1695,
+      "step": 9658
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.4327823073098216,
+      "learning_rate": 4.639526210517095e-06,
+      "loss": 0.2343,
+      "step": 9659
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.0831859534987016,
+      "learning_rate": 4.63862771962399e-06,
+      "loss": 0.1561,
+      "step": 9660
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.042365773165184,
+      "learning_rate": 4.637729240461235e-06,
+      "loss": 0.1702,
+      "step": 9661
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.0691596804764971,
+      "learning_rate": 4.636830773057997e-06,
+      "loss": 0.2146,
+      "step": 9662
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.0617288462683014,
+      "learning_rate": 4.635932317443441e-06,
+      "loss": 0.1811,
+      "step": 9663
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.7477822483107412,
+      "learning_rate": 4.6350338736467326e-06,
+      "loss": 0.1432,
+      "step": 9664
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.0786985620068155,
+      "learning_rate": 4.634135441697034e-06,
+      "loss": 0.1968,
+      "step": 9665
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.9366811573766799,
+      "learning_rate": 4.63323702162351e-06,
+      "loss": 0.1722,
+      "step": 9666
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.1024047729234012,
+      "learning_rate": 4.632338613455322e-06,
+      "loss": 0.161,
+      "step": 9667
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.2108746712290293,
+      "learning_rate": 4.631440217221634e-06,
+      "loss": 0.208,
+      "step": 9668
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.8052260439275128,
+      "learning_rate": 4.63054183295161e-06,
+      "loss": 0.1623,
+      "step": 9669
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.7125393812931444,
+      "learning_rate": 4.629643460674409e-06,
+      "loss": 0.1932,
+      "step": 9670
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.8014378622833103,
+      "learning_rate": 4.628745100419194e-06,
+      "loss": 0.2038,
+      "step": 9671
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.9506805274886501,
+      "learning_rate": 4.627846752215126e-06,
+      "loss": 0.1877,
+      "step": 9672
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.8142140028689968,
+      "learning_rate": 4.626948416091369e-06,
+      "loss": 0.1902,
+      "step": 9673
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.3697374039104917,
+      "learning_rate": 4.6260500920770794e-06,
+      "loss": 0.2197,
+      "step": 9674
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.8007711639866577,
+      "learning_rate": 4.62515178020142e-06,
+      "loss": 0.1938,
+      "step": 9675
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.476045111679314,
+      "learning_rate": 4.624253480493547e-06,
+      "loss": 0.2446,
+      "step": 9676
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.1709983073603265,
+      "learning_rate": 4.623355192982624e-06,
+      "loss": 0.2154,
+      "step": 9677
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.3607865382315263,
+      "learning_rate": 4.622456917697808e-06,
+      "loss": 0.213,
+      "step": 9678
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.3853423904164792,
+      "learning_rate": 4.621558654668257e-06,
+      "loss": 0.1854,
+      "step": 9679
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.4319290531525863,
+      "learning_rate": 4.620660403923129e-06,
+      "loss": 0.2531,
+      "step": 9680
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.1228466087573095,
+      "learning_rate": 4.6197621654915824e-06,
+      "loss": 0.1803,
+      "step": 9681
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.9833311836406412,
+      "learning_rate": 4.6188639394027766e-06,
+      "loss": 0.1716,
+      "step": 9682
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.422097503104572,
+      "learning_rate": 4.617965725685866e-06,
+      "loss": 0.2175,
+      "step": 9683
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.1804133954503848,
+      "learning_rate": 4.617067524370007e-06,
+      "loss": 0.1974,
+      "step": 9684
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.3461496417928371,
+      "learning_rate": 4.616169335484355e-06,
+      "loss": 0.2322,
+      "step": 9685
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.419627832140847,
+      "learning_rate": 4.615271159058068e-06,
+      "loss": 0.2031,
+      "step": 9686
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.5253706966177827,
+      "learning_rate": 4.614372995120302e-06,
+      "loss": 0.1722,
+      "step": 9687
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.1571921492340782,
+      "learning_rate": 4.6134748437002085e-06,
+      "loss": 0.1832,
+      "step": 9688
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.6819998859151468,
+      "learning_rate": 4.612576704826945e-06,
+      "loss": 0.223,
+      "step": 9689
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.1059788851231414,
+      "learning_rate": 4.611678578529663e-06,
+      "loss": 0.2381,
+      "step": 9690
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.127576579519262,
+      "learning_rate": 4.6107804648375185e-06,
+      "loss": 0.1602,
+      "step": 9691
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.0428132573491662,
+      "learning_rate": 4.609882363779664e-06,
+      "loss": 0.1876,
+      "step": 9692
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.718352228432944,
+      "learning_rate": 4.6089842753852535e-06,
+      "loss": 0.2237,
+      "step": 9693
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.1262483955185583,
+      "learning_rate": 4.608086199683436e-06,
+      "loss": 0.1696,
+      "step": 9694
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.4421372640463055,
+      "learning_rate": 4.607188136703366e-06,
+      "loss": 0.1804,
+      "step": 9695
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.462700960115544,
+      "learning_rate": 4.606290086474198e-06,
+      "loss": 0.2379,
+      "step": 9696
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.970864019741196,
+      "learning_rate": 4.605392049025077e-06,
+      "loss": 0.1914,
+      "step": 9697
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.2861595969282293,
+      "learning_rate": 4.604494024385157e-06,
+      "loss": 0.172,
+      "step": 9698
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.0854806213366717,
+      "learning_rate": 4.6035960125835885e-06,
+      "loss": 0.1371,
+      "step": 9699
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.9235312960845442,
+      "learning_rate": 4.602698013649523e-06,
+      "loss": 0.1412,
+      "step": 9700
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.132483661541123,
+      "learning_rate": 4.601800027612108e-06,
+      "loss": 0.1982,
+      "step": 9701
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.2336924090079775,
+      "learning_rate": 4.6009020545004935e-06,
+      "loss": 0.2118,
+      "step": 9702
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.6877492239373522,
+      "learning_rate": 4.600004094343827e-06,
+      "loss": 0.1921,
+      "step": 9703
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.4018639298851023,
+      "learning_rate": 4.599106147171257e-06,
+      "loss": 0.2515,
+      "step": 9704
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.8420033610997768,
+      "learning_rate": 4.598208213011934e-06,
+      "loss": 0.1344,
+      "step": 9705
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.255955980124129,
+      "learning_rate": 4.597310291895002e-06,
+      "loss": 0.172,
+      "step": 9706
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.1003648828725834,
+      "learning_rate": 4.596412383849609e-06,
+      "loss": 0.2143,
+      "step": 9707
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.2474272268747537,
+      "learning_rate": 4.595514488904902e-06,
+      "loss": 0.2187,
+      "step": 9708
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.8363727902418998,
+      "learning_rate": 4.594616607090028e-06,
+      "loss": 0.1024,
+      "step": 9709
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.4534205564262999,
+      "learning_rate": 4.593718738434133e-06,
+      "loss": 0.2411,
+      "step": 9710
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.8988608534207928,
+      "learning_rate": 4.592820882966361e-06,
+      "loss": 0.1589,
+      "step": 9711
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.0996651576038952,
+      "learning_rate": 4.591923040715856e-06,
+      "loss": 0.2,
+      "step": 9712
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.1616102573295009,
+      "learning_rate": 4.591025211711764e-06,
+      "loss": 0.1539,
+      "step": 9713
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.2727915626351982,
+      "learning_rate": 4.590127395983231e-06,
+      "loss": 0.1732,
+      "step": 9714
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.196965239026385,
+      "learning_rate": 4.589229593559396e-06,
+      "loss": 0.1345,
+      "step": 9715
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.125276577912996,
+      "learning_rate": 4.588331804469405e-06,
+      "loss": 0.1916,
+      "step": 9716
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.8622028035751427,
+      "learning_rate": 4.5874340287423995e-06,
+      "loss": 0.1814,
+      "step": 9717
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.3453715985971189,
+      "learning_rate": 4.586536266407525e-06,
+      "loss": 0.1956,
+      "step": 9718
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.54228921726803,
+      "learning_rate": 4.58563851749392e-06,
+      "loss": 0.2294,
+      "step": 9719
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.8382923656981621,
+      "learning_rate": 4.584740782030727e-06,
+      "loss": 0.1619,
+      "step": 9720
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.1153279146693482,
+      "learning_rate": 4.5838430600470865e-06,
+      "loss": 0.1934,
+      "step": 9721
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.9072704407230653,
+      "learning_rate": 4.58294535157214e-06,
+      "loss": 0.1645,
+      "step": 9722
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.0732063933105642,
+      "learning_rate": 4.582047656635028e-06,
+      "loss": 0.19,
+      "step": 9723
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.358982448790332,
+      "learning_rate": 4.581149975264888e-06,
+      "loss": 0.2074,
+      "step": 9724
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.0091852848347365,
+      "learning_rate": 4.580252307490861e-06,
+      "loss": 0.2066,
+      "step": 9725
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.1043476597052255,
+      "learning_rate": 4.5793546533420856e-06,
+      "loss": 0.1775,
+      "step": 9726
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.4065186138040169,
+      "learning_rate": 4.5784570128477e-06,
+      "loss": 0.2205,
+      "step": 9727
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.6342595113391711,
+      "learning_rate": 4.577559386036842e-06,
+      "loss": 0.2311,
+      "step": 9728
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.8884179475627348,
+      "learning_rate": 4.576661772938651e-06,
+      "loss": 0.1211,
+      "step": 9729
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.1408530360072333,
+      "learning_rate": 4.57576417358226e-06,
+      "loss": 0.1609,
+      "step": 9730
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.0968549474556915,
+      "learning_rate": 4.574866587996809e-06,
+      "loss": 0.1626,
+      "step": 9731
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.1564544258602576,
+      "learning_rate": 4.573969016211434e-06,
+      "loss": 0.1873,
+      "step": 9732
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.174914008529651,
+      "learning_rate": 4.573071458255268e-06,
+      "loss": 0.1645,
+      "step": 9733
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.0723246400084447,
+      "learning_rate": 4.572173914157447e-06,
+      "loss": 0.1552,
+      "step": 9734
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.2769412068452346,
+      "learning_rate": 4.571276383947108e-06,
+      "loss": 0.1723,
+      "step": 9735
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.057028454243303,
+      "learning_rate": 4.570378867653385e-06,
+      "loss": 0.1891,
+      "step": 9736
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.3629115458681582,
+      "learning_rate": 4.569481365305411e-06,
+      "loss": 0.1857,
+      "step": 9737
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.0924916300327514,
+      "learning_rate": 4.568583876932319e-06,
+      "loss": 0.1912,
+      "step": 9738
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.4721362513831044,
+      "learning_rate": 4.567686402563242e-06,
+      "loss": 0.2311,
+      "step": 9739
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.0220139336121403,
+      "learning_rate": 4.566788942227312e-06,
+      "loss": 0.2151,
+      "step": 9740
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.5278265841110525,
+      "learning_rate": 4.565891495953665e-06,
+      "loss": 0.1852,
+      "step": 9741
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.2872953454224008,
+      "learning_rate": 4.564994063771427e-06,
+      "loss": 0.1994,
+      "step": 9742
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.9430716314661851,
+      "learning_rate": 4.564096645709733e-06,
+      "loss": 0.1627,
+      "step": 9743
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.0382802302353376,
+      "learning_rate": 4.563199241797711e-06,
+      "loss": 0.1807,
+      "step": 9744
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.3996750401246567,
+      "learning_rate": 4.562301852064495e-06,
+      "loss": 0.2205,
+      "step": 9745
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.1677980170603615,
+      "learning_rate": 4.561404476539212e-06,
+      "loss": 0.1792,
+      "step": 9746
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.2520691673548012,
+      "learning_rate": 4.560507115250992e-06,
+      "loss": 0.1622,
+      "step": 9747
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.4499108443310011,
+      "learning_rate": 4.5596097682289625e-06,
+      "loss": 0.2442,
+      "step": 9748
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.9379040205132748,
+      "learning_rate": 4.558712435502254e-06,
+      "loss": 0.1861,
+      "step": 9749
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.7550071625549788,
+      "learning_rate": 4.5578151170999926e-06,
+      "loss": 0.2201,
+      "step": 9750
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.1584230040571217,
+      "learning_rate": 4.5569178130513095e-06,
+      "loss": 0.18,
+      "step": 9751
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.022234748662657,
+      "learning_rate": 4.556020523385326e-06,
+      "loss": 0.1981,
+      "step": 9752
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.557126497091569,
+      "learning_rate": 4.555123248131173e-06,
+      "loss": 0.2545,
+      "step": 9753
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.9412768935924704,
+      "learning_rate": 4.554225987317975e-06,
+      "loss": 0.2066,
+      "step": 9754
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.3562341183043474,
+      "learning_rate": 4.553328740974859e-06,
+      "loss": 0.2038,
+      "step": 9755
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.8996238847571005,
+      "learning_rate": 4.5524315091309476e-06,
+      "loss": 0.1923,
+      "step": 9756
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.312316985370045,
+      "learning_rate": 4.551534291815366e-06,
+      "loss": 0.1779,
+      "step": 9757
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.5277365157547158,
+      "learning_rate": 4.55063708905724e-06,
+      "loss": 0.1761,
+      "step": 9758
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.424095887384382,
+      "learning_rate": 4.5497399008856925e-06,
+      "loss": 0.1944,
+      "step": 9759
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.0913211198809005,
+      "learning_rate": 4.548842727329848e-06,
+      "loss": 0.2362,
+      "step": 9760
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.1051425334570544,
+      "learning_rate": 4.547945568418826e-06,
+      "loss": 0.2301,
+      "step": 9761
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.0807675822067833,
+      "learning_rate": 4.547048424181751e-06,
+      "loss": 0.1998,
+      "step": 9762
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.3953866899380516,
+      "learning_rate": 4.546151294647746e-06,
+      "loss": 0.2105,
+      "step": 9763
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.3461065379756896,
+      "learning_rate": 4.545254179845931e-06,
+      "loss": 0.2481,
+      "step": 9764
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.307652175982211,
+      "learning_rate": 4.544357079805427e-06,
+      "loss": 0.1643,
+      "step": 9765
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.5348538739160704,
+      "learning_rate": 4.543459994555352e-06,
+      "loss": 0.2077,
+      "step": 9766
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.497659779934014,
+      "learning_rate": 4.542562924124829e-06,
+      "loss": 0.2375,
+      "step": 9767
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.2340875997234317,
+      "learning_rate": 4.5416658685429756e-06,
+      "loss": 0.154,
+      "step": 9768
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.003675814725018,
+      "learning_rate": 4.540768827838915e-06,
+      "loss": 0.2051,
+      "step": 9769
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.9466958997065058,
+      "learning_rate": 4.53987180204176e-06,
+      "loss": 0.1461,
+      "step": 9770
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.418413360501232,
+      "learning_rate": 4.53897479118063e-06,
+      "loss": 0.2111,
+      "step": 9771
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.1546894203164906,
+      "learning_rate": 4.538077795284645e-06,
+      "loss": 0.1793,
+      "step": 9772
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.3140231034182537,
+      "learning_rate": 4.537180814382921e-06,
+      "loss": 0.1925,
+      "step": 9773
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.0817495609448518,
+      "learning_rate": 4.536283848504573e-06,
+      "loss": 0.1583,
+      "step": 9774
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.1816462697926775,
+      "learning_rate": 4.535386897678717e-06,
+      "loss": 0.1811,
+      "step": 9775
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.1571981627766008,
+      "learning_rate": 4.53448996193447e-06,
+      "loss": 0.1684,
+      "step": 9776
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.2810743810504088,
+      "learning_rate": 4.533593041300946e-06,
+      "loss": 0.2492,
+      "step": 9777
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.9294439044747383,
+      "learning_rate": 4.532696135807262e-06,
+      "loss": 0.157,
+      "step": 9778
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.1503452719538734,
+      "learning_rate": 4.5317992454825285e-06,
+      "loss": 0.2589,
+      "step": 9779
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.0490850199704642,
+      "learning_rate": 4.530902370355862e-06,
+      "loss": 0.1848,
+      "step": 9780
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.3616722594163921,
+      "learning_rate": 4.5300055104563734e-06,
+      "loss": 0.2572,
+      "step": 9781
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.070954815752682,
+      "learning_rate": 4.529108665813177e-06,
+      "loss": 0.2022,
+      "step": 9782
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.3012947051456383,
+      "learning_rate": 4.528211836455384e-06,
+      "loss": 0.1873,
+      "step": 9783
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.0832747499355386,
+      "learning_rate": 4.527315022412104e-06,
+      "loss": 0.1696,
+      "step": 9784
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.8422630461955949,
+      "learning_rate": 4.526418223712451e-06,
+      "loss": 0.1402,
+      "step": 9785
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.9066581793404206,
+      "learning_rate": 4.5255214403855355e-06,
+      "loss": 0.1898,
+      "step": 9786
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.4913017505106299,
+      "learning_rate": 4.524624672460469e-06,
+      "loss": 0.2227,
+      "step": 9787
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.2008116763546974,
+      "learning_rate": 4.523727919966356e-06,
+      "loss": 0.2022,
+      "step": 9788
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.3622263273994204,
+      "learning_rate": 4.52283118293231e-06,
+      "loss": 0.2048,
+      "step": 9789
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.450074971147668,
+      "learning_rate": 4.5219344613874365e-06,
+      "loss": 0.2938,
+      "step": 9790
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.9094185223520639,
+      "learning_rate": 4.521037755360847e-06,
+      "loss": 0.167,
+      "step": 9791
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.7089122244727064,
+      "learning_rate": 4.520141064881648e-06,
+      "loss": 0.1943,
+      "step": 9792
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.7976393213072944,
+      "learning_rate": 4.519244389978944e-06,
+      "loss": 0.1666,
+      "step": 9793
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.0548740310317561,
+      "learning_rate": 4.5183477306818435e-06,
+      "loss": 0.2006,
+      "step": 9794
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.181712506212311,
+      "learning_rate": 4.5174510870194535e-06,
+      "loss": 0.1768,
+      "step": 9795
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.2706467444761211,
+      "learning_rate": 4.5165544590208795e-06,
+      "loss": 0.1818,
+      "step": 9796
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.2082105048688383,
+      "learning_rate": 4.515657846715225e-06,
+      "loss": 0.1916,
+      "step": 9797
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.1109604319420154,
+      "learning_rate": 4.514761250131595e-06,
+      "loss": 0.2517,
+      "step": 9798
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.0231129572159012,
+      "learning_rate": 4.513864669299094e-06,
+      "loss": 0.208,
+      "step": 9799
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.4745473191801606,
+      "learning_rate": 4.512968104246827e-06,
+      "loss": 0.189,
+      "step": 9800
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.2399542424580903,
+      "learning_rate": 4.512071555003894e-06,
+      "loss": 0.2576,
+      "step": 9801
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.810316537967629,
+      "learning_rate": 4.511175021599398e-06,
+      "loss": 0.1843,
+      "step": 9802
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.4512158177988805,
+      "learning_rate": 4.510278504062442e-06,
+      "loss": 0.1827,
+      "step": 9803
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.0818514855337273,
+      "learning_rate": 4.5093820024221275e-06,
+      "loss": 0.2061,
+      "step": 9804
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.9505766566967632,
+      "learning_rate": 4.508485516707557e-06,
+      "loss": 0.1382,
+      "step": 9805
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.3858253247065866,
+      "learning_rate": 4.5075890469478275e-06,
+      "loss": 0.1877,
+      "step": 9806
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.700264376618521,
+      "learning_rate": 4.506692593172043e-06,
+      "loss": 0.184,
+      "step": 9807
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.0918352140816134,
+      "learning_rate": 4.505796155409298e-06,
+      "loss": 0.1877,
+      "step": 9808
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.4557134099140558,
+      "learning_rate": 4.504899733688696e-06,
+      "loss": 0.203,
+      "step": 9809
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.0675275971409988,
+      "learning_rate": 4.504003328039332e-06,
+      "loss": 0.2046,
+      "step": 9810
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.4421154761075545,
+      "learning_rate": 4.503106938490305e-06,
+      "loss": 0.1982,
+      "step": 9811
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.3476328806302167,
+      "learning_rate": 4.502210565070712e-06,
+      "loss": 0.205,
+      "step": 9812
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.089924905004549,
+      "learning_rate": 4.501314207809651e-06,
+      "loss": 0.1625,
+      "step": 9813
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.2554257298275442,
+      "learning_rate": 4.500417866736219e-06,
+      "loss": 0.2599,
+      "step": 9814
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.9281600818682729,
+      "learning_rate": 4.499521541879508e-06,
+      "loss": 0.1723,
+      "step": 9815
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.9636948498112179,
+      "learning_rate": 4.498625233268618e-06,
+      "loss": 0.1436,
+      "step": 9816
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.3011584136069583,
+      "learning_rate": 4.49772894093264e-06,
+      "loss": 0.1889,
+      "step": 9817
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.3208573080718953,
+      "learning_rate": 4.49683266490067e-06,
+      "loss": 0.2173,
+      "step": 9818
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.91682601881576,
+      "learning_rate": 4.4959364052018e-06,
+      "loss": 0.1975,
+      "step": 9819
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.0021474735282883,
+      "learning_rate": 4.495040161865124e-06,
+      "loss": 0.2152,
+      "step": 9820
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.456451094054341,
+      "learning_rate": 4.494143934919735e-06,
+      "loss": 0.2396,
+      "step": 9821
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.0491821323858979,
+      "learning_rate": 4.4932477243947245e-06,
+      "loss": 0.1638,
+      "step": 9822
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.2201134455247766,
+      "learning_rate": 4.492351530319186e-06,
+      "loss": 0.1896,
+      "step": 9823
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.4268711522886353,
+      "learning_rate": 4.491455352722207e-06,
+      "loss": 0.1952,
+      "step": 9824
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.1623792879820543,
+      "learning_rate": 4.490559191632881e-06,
+      "loss": 0.1925,
+      "step": 9825
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.3515295785411434,
+      "learning_rate": 4.489663047080295e-06,
+      "loss": 0.2107,
+      "step": 9826
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.2883891846972484,
+      "learning_rate": 4.488766919093542e-06,
+      "loss": 0.1986,
+      "step": 9827
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.0956422105472166,
+      "learning_rate": 4.487870807701708e-06,
+      "loss": 0.1779,
+      "step": 9828
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.3103635861613745,
+      "learning_rate": 4.486974712933881e-06,
+      "loss": 0.2195,
+      "step": 9829
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.1836890097327042,
+      "learning_rate": 4.48607863481915e-06,
+      "loss": 0.1736,
+      "step": 9830
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.346676680552533,
+      "learning_rate": 4.485182573386602e-06,
+      "loss": 0.1991,
+      "step": 9831
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.0982625006598838,
+      "learning_rate": 4.4842865286653256e-06,
+      "loss": 0.1835,
+      "step": 9832
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.1850148473083115,
+      "learning_rate": 4.483390500684403e-06,
+      "loss": 0.1928,
+      "step": 9833
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.070577515558442,
+      "learning_rate": 4.482494489472923e-06,
+      "loss": 0.1738,
+      "step": 9834
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.9153790936285382,
+      "learning_rate": 4.481598495059968e-06,
+      "loss": 0.1949,
+      "step": 9835
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.9929293314795334,
+      "learning_rate": 4.480702517474626e-06,
+      "loss": 0.1692,
+      "step": 9836
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.8090804765414046,
+      "learning_rate": 4.479806556745978e-06,
+      "loss": 0.2016,
+      "step": 9837
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.161707172195011,
+      "learning_rate": 4.478910612903108e-06,
+      "loss": 0.1756,
+      "step": 9838
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.7237295669329407,
+      "learning_rate": 4.4780146859750975e-06,
+      "loss": 0.2332,
+      "step": 9839
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.3816225244287144,
+      "learning_rate": 4.477118775991033e-06,
+      "loss": 0.1806,
+      "step": 9840
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.1596423513694991,
+      "learning_rate": 4.476222882979994e-06,
+      "loss": 0.171,
+      "step": 9841
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.345533468240246,
+      "learning_rate": 4.47532700697106e-06,
+      "loss": 0.2372,
+      "step": 9842
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.0914240048547426,
+      "learning_rate": 4.474431147993313e-06,
+      "loss": 0.1433,
+      "step": 9843
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.390437483199694,
+      "learning_rate": 4.473535306075834e-06,
+      "loss": 0.1969,
+      "step": 9844
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.0703763733899234,
+      "learning_rate": 4.472639481247702e-06,
+      "loss": 0.1905,
+      "step": 9845
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.2632544593117052,
+      "learning_rate": 4.471743673537994e-06,
+      "loss": 0.1989,
+      "step": 9846
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.3136873891449181,
+      "learning_rate": 4.4708478829757916e-06,
+      "loss": 0.2289,
+      "step": 9847
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.5557944488003828,
+      "learning_rate": 4.4699521095901696e-06,
+      "loss": 0.2091,
+      "step": 9848
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.2621842461600963,
+      "learning_rate": 4.4690563534102076e-06,
+      "loss": 0.1807,
+      "step": 9849
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.206225326760113,
+      "learning_rate": 4.468160614464983e-06,
+      "loss": 0.1947,
+      "step": 9850
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.06613061996859,
+      "learning_rate": 4.4672648927835695e-06,
+      "loss": 0.1813,
+      "step": 9851
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.0140120352662367,
+      "learning_rate": 4.466369188395045e-06,
+      "loss": 0.18,
+      "step": 9852
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.3046834340289142,
+      "learning_rate": 4.465473501328481e-06,
+      "loss": 0.1723,
+      "step": 9853
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.340791847832609,
+      "learning_rate": 4.464577831612958e-06,
+      "loss": 0.1798,
+      "step": 9854
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.1081417038482082,
+      "learning_rate": 4.463682179277545e-06,
+      "loss": 0.1906,
+      "step": 9855
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.4909083790794595,
+      "learning_rate": 4.4627865443513156e-06,
+      "loss": 0.191,
+      "step": 9856
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.9645583081895563,
+      "learning_rate": 4.461890926863344e-06,
+      "loss": 0.1603,
+      "step": 9857
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.5450558899674447,
+      "learning_rate": 4.460995326842703e-06,
+      "loss": 0.1972,
+      "step": 9858
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.17635058879463,
+      "learning_rate": 4.4600997443184655e-06,
+      "loss": 0.2013,
+      "step": 9859
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.3997670704708054,
+      "learning_rate": 4.459204179319699e-06,
+      "loss": 0.1942,
+      "step": 9860
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.4295756186812014,
+      "learning_rate": 4.458308631875476e-06,
+      "loss": 0.1978,
+      "step": 9861
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.3836314206458151,
+      "learning_rate": 4.457413102014866e-06,
+      "loss": 0.1782,
+      "step": 9862
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.3663584443438084,
+      "learning_rate": 4.45651758976694e-06,
+      "loss": 0.2021,
+      "step": 9863
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.3812934160960888,
+      "learning_rate": 4.455622095160765e-06,
+      "loss": 0.2367,
+      "step": 9864
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.2695604350329996,
+      "learning_rate": 4.454726618225408e-06,
+      "loss": 0.1919,
+      "step": 9865
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.4624911375266316,
+      "learning_rate": 4.4538311589899395e-06,
+      "loss": 0.2138,
+      "step": 9866
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.2852733784784478,
+      "learning_rate": 4.452935717483425e-06,
+      "loss": 0.2261,
+      "step": 9867
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.2150489267421913,
+      "learning_rate": 4.452040293734934e-06,
+      "loss": 0.1878,
+      "step": 9868
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.305707315942533,
+      "learning_rate": 4.451144887773529e-06,
+      "loss": 0.1796,
+      "step": 9869
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.9759044623638448,
+      "learning_rate": 4.450249499628276e-06,
+      "loss": 0.179,
+      "step": 9870
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.928750535700401,
+      "learning_rate": 4.449354129328241e-06,
+      "loss": 0.1967,
+      "step": 9871
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.8064762137274554,
+      "learning_rate": 4.448458776902489e-06,
+      "loss": 0.1444,
+      "step": 9872
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.367540017048959,
+      "learning_rate": 4.4475634423800795e-06,
+      "loss": 0.2031,
+      "step": 9873
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.1105769758140318,
+      "learning_rate": 4.446668125790079e-06,
+      "loss": 0.1714,
+      "step": 9874
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.111032097916117,
+      "learning_rate": 4.445772827161549e-06,
+      "loss": 0.224,
+      "step": 9875
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.1463412336838652,
+      "learning_rate": 4.444877546523552e-06,
+      "loss": 0.2131,
+      "step": 9876
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.182042077337019,
+      "learning_rate": 4.443982283905152e-06,
+      "loss": 0.1587,
+      "step": 9877
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.4003404571080933,
+      "learning_rate": 4.443087039335403e-06,
+      "loss": 0.2289,
+      "step": 9878
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.4815236330546435,
+      "learning_rate": 4.442191812843371e-06,
+      "loss": 0.2203,
+      "step": 9879
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.081858362921704,
+      "learning_rate": 4.441296604458113e-06,
+      "loss": 0.2254,
+      "step": 9880
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.0610428733913695,
+      "learning_rate": 4.4404014142086896e-06,
+      "loss": 0.2202,
+      "step": 9881
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.8755804588866353,
+      "learning_rate": 4.439506242124156e-06,
+      "loss": 0.1488,
+      "step": 9882
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.333594738647595,
+      "learning_rate": 4.4386110882335724e-06,
+      "loss": 0.2354,
+      "step": 9883
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.2307684578787115,
+      "learning_rate": 4.437715952565996e-06,
+      "loss": 0.2193,
+      "step": 9884
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.204661787379618,
+      "learning_rate": 4.436820835150483e-06,
+      "loss": 0.2385,
+      "step": 9885
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.309987207345252,
+      "learning_rate": 4.4359257360160914e-06,
+      "loss": 0.1946,
+      "step": 9886
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.4084542906280646,
+      "learning_rate": 4.435030655191872e-06,
+      "loss": 0.2276,
+      "step": 9887
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.1997455409090927,
+      "learning_rate": 4.4341355927068845e-06,
+      "loss": 0.1896,
+      "step": 9888
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.087498842573098,
+      "learning_rate": 4.43324054859018e-06,
+      "loss": 0.2008,
+      "step": 9889
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.1584013950925667,
+      "learning_rate": 4.432345522870815e-06,
+      "loss": 0.1986,
+      "step": 9890
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.4988375818800772,
+      "learning_rate": 4.431450515577838e-06,
+      "loss": 0.2151,
+      "step": 9891
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.093267821567091,
+      "learning_rate": 4.430555526740305e-06,
+      "loss": 0.2035,
+      "step": 9892
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.1181843235229172,
+      "learning_rate": 4.4296605563872675e-06,
+      "loss": 0.1583,
+      "step": 9893
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.33457294566016,
+      "learning_rate": 4.428765604547775e-06,
+      "loss": 0.2403,
+      "step": 9894
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.7145064767396465,
+      "learning_rate": 4.427870671250883e-06,
+      "loss": 0.2141,
+      "step": 9895
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.245022420388905,
+      "learning_rate": 4.426975756525635e-06,
+      "loss": 0.2077,
+      "step": 9896
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.0045780539040985,
+      "learning_rate": 4.426080860401085e-06,
+      "loss": 0.204,
+      "step": 9897
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.041897319847535,
+      "learning_rate": 4.4251859829062784e-06,
+      "loss": 0.1158,
+      "step": 9898
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.3151500874864779,
+      "learning_rate": 4.424291124070268e-06,
+      "loss": 0.2027,
+      "step": 9899
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.2357222487178212,
+      "learning_rate": 4.423396283922097e-06,
+      "loss": 0.2374,
+      "step": 9900
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.2175106477810342,
+      "learning_rate": 4.422501462490813e-06,
+      "loss": 0.2192,
+      "step": 9901
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.3398662654219398,
+      "learning_rate": 4.4216066598054655e-06,
+      "loss": 0.1819,
+      "step": 9902
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.7675768315932853,
+      "learning_rate": 4.420711875895097e-06,
+      "loss": 0.1945,
+      "step": 9903
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.093046583754055,
+      "learning_rate": 4.419817110788756e-06,
+      "loss": 0.1575,
+      "step": 9904
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.2871312398601822,
+      "learning_rate": 4.418922364515484e-06,
+      "loss": 0.1728,
+      "step": 9905
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.1023317214720496,
+      "learning_rate": 4.4180276371043274e-06,
+      "loss": 0.1535,
+      "step": 9906
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.9868870199799017,
+      "learning_rate": 4.417132928584327e-06,
+      "loss": 0.1608,
+      "step": 9907
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.6922409737786246,
+      "learning_rate": 4.416238238984529e-06,
+      "loss": 0.1868,
+      "step": 9908
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.1040901410227222,
+      "learning_rate": 4.415343568333971e-06,
+      "loss": 0.1817,
+      "step": 9909
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.2075410170751082,
+      "learning_rate": 4.414448916661697e-06,
+      "loss": 0.2358,
+      "step": 9910
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.014203769956384,
+      "learning_rate": 4.4135542839967476e-06,
+      "loss": 0.2106,
+      "step": 9911
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.1490380053623659,
+      "learning_rate": 4.412659670368164e-06,
+      "loss": 0.1982,
+      "step": 9912
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.9840175986501952,
+      "learning_rate": 4.411765075804987e-06,
+      "loss": 0.2088,
+      "step": 9913
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.1026612814644814,
+      "learning_rate": 4.410870500336252e-06,
+      "loss": 0.2283,
+      "step": 9914
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.1946541634388015,
+      "learning_rate": 4.409975943991e-06,
+      "loss": 0.2193,
+      "step": 9915
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.987909778697391,
+      "learning_rate": 4.409081406798268e-06,
+      "loss": 0.2062,
+      "step": 9916
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.022546804660881,
+      "learning_rate": 4.408186888787095e-06,
+      "loss": 0.1693,
+      "step": 9917
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.8565305843551796,
+      "learning_rate": 4.407292389986514e-06,
+      "loss": 0.1531,
+      "step": 9918
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.9145668301291805,
+      "learning_rate": 4.406397910425562e-06,
+      "loss": 0.1801,
+      "step": 9919
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.3961204454744067,
+      "learning_rate": 4.405503450133276e-06,
+      "loss": 0.1921,
+      "step": 9920
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.2179651759488541,
+      "learning_rate": 4.40460900913869e-06,
+      "loss": 0.2098,
+      "step": 9921
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.6215702792749478,
+      "learning_rate": 4.403714587470839e-06,
+      "loss": 0.2367,
+      "step": 9922
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.2211654413232067,
+      "learning_rate": 4.402820185158755e-06,
+      "loss": 0.2409,
+      "step": 9923
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.5098812631437204,
+      "learning_rate": 4.401925802231471e-06,
+      "loss": 0.1572,
+      "step": 9924
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.7967808997468441,
+      "learning_rate": 4.401031438718018e-06,
+      "loss": 0.1774,
+      "step": 9925
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.2207251142428655,
+      "learning_rate": 4.400137094647432e-06,
+      "loss": 0.2016,
+      "step": 9926
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.282195550640882,
+      "learning_rate": 4.3992427700487375e-06,
+      "loss": 0.1903,
+      "step": 9927
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.9545943452772848,
+      "learning_rate": 4.3983484649509685e-06,
+      "loss": 0.1775,
+      "step": 9928
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.9763351900672763,
+      "learning_rate": 4.397454179383155e-06,
+      "loss": 0.1878,
+      "step": 9929
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.460504209127626,
+      "learning_rate": 4.396559913374325e-06,
+      "loss": 0.195,
+      "step": 9930
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.2777275302182327,
+      "learning_rate": 4.39566566695351e-06,
+      "loss": 0.1587,
+      "step": 9931
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.1017305063408416,
+      "learning_rate": 4.394771440149732e-06,
+      "loss": 0.165,
+      "step": 9932
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.836389583361739,
+      "learning_rate": 4.393877232992022e-06,
+      "loss": 0.2129,
+      "step": 9933
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.4673395042065538,
+      "learning_rate": 4.392983045509405e-06,
+      "loss": 0.2067,
+      "step": 9934
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.7608445698977895,
+      "learning_rate": 4.39208887773091e-06,
+      "loss": 0.1917,
+      "step": 9935
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.2356942918930487,
+      "learning_rate": 4.3911947296855565e-06,
+      "loss": 0.2215,
+      "step": 9936
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.9696670602327848,
+      "learning_rate": 4.390300601402372e-06,
+      "loss": 0.2179,
+      "step": 9937
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.9948123712115988,
+      "learning_rate": 4.389406492910381e-06,
+      "loss": 0.161,
+      "step": 9938
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.3038606250567737,
+      "learning_rate": 4.388512404238607e-06,
+      "loss": 0.2232,
+      "step": 9939
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.7371014295916964,
+      "learning_rate": 4.387618335416073e-06,
+      "loss": 0.1336,
+      "step": 9940
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.1428926849998577,
+      "learning_rate": 4.386724286471799e-06,
+      "loss": 0.1705,
+      "step": 9941
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.4700222516362562,
+      "learning_rate": 4.385830257434808e-06,
+      "loss": 0.1656,
+      "step": 9942
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.5260658088657897,
+      "learning_rate": 4.384936248334118e-06,
+      "loss": 0.2184,
+      "step": 9943
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.115649667935971,
+      "learning_rate": 4.3840422591987545e-06,
+      "loss": 0.1777,
+      "step": 9944
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.1017628049678179,
+      "learning_rate": 4.383148290057731e-06,
+      "loss": 0.1655,
+      "step": 9945
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.1980860576559307,
+      "learning_rate": 4.382254340940069e-06,
+      "loss": 0.2199,
+      "step": 9946
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.3023058259335931,
+      "learning_rate": 4.381360411874786e-06,
+      "loss": 0.1795,
+      "step": 9947
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.913407327321346,
+      "learning_rate": 4.380466502890899e-06,
+      "loss": 0.1686,
+      "step": 9948
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.85492397951775,
+      "learning_rate": 4.379572614017428e-06,
+      "loss": 0.1482,
+      "step": 9949
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.2671600822135458,
+      "learning_rate": 4.3786787452833854e-06,
+      "loss": 0.226,
+      "step": 9950
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.2441759418882568,
+      "learning_rate": 4.3777848967177885e-06,
+      "loss": 0.1666,
+      "step": 9951
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.1282884085147737,
+      "learning_rate": 4.376891068349651e-06,
+      "loss": 0.1779,
+      "step": 9952
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.0663364356107896,
+      "learning_rate": 4.375997260207989e-06,
+      "loss": 0.1893,
+      "step": 9953
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.4967575941054048,
+      "learning_rate": 4.375103472321813e-06,
+      "loss": 0.2224,
+      "step": 9954
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.0524094350828752,
+      "learning_rate": 4.374209704720138e-06,
+      "loss": 0.183,
+      "step": 9955
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.7761559426319005,
+      "learning_rate": 4.373315957431975e-06,
+      "loss": 0.2196,
+      "step": 9956
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.0537844359154955,
+      "learning_rate": 4.372422230486338e-06,
+      "loss": 0.2246,
+      "step": 9957
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.0527004880529147,
+      "learning_rate": 4.371528523912237e-06,
+      "loss": 0.1692,
+      "step": 9958
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.9171345532193874,
+      "learning_rate": 4.37063483773868e-06,
+      "loss": 0.1603,
+      "step": 9959
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.1740714914212655,
+      "learning_rate": 4.3697411719946795e-06,
+      "loss": 0.1479,
+      "step": 9960
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.2095211991081134,
+      "learning_rate": 4.368847526709241e-06,
+      "loss": 0.19,
+      "step": 9961
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.1299063829289564,
+      "learning_rate": 4.367953901911377e-06,
+      "loss": 0.1803,
+      "step": 9962
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.193312725210471,
+      "learning_rate": 4.367060297630091e-06,
+      "loss": 0.2005,
+      "step": 9963
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.9373410028929963,
+      "learning_rate": 4.366166713894392e-06,
+      "loss": 0.2214,
+      "step": 9964
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.2378645325086879,
+      "learning_rate": 4.3652731507332855e-06,
+      "loss": 0.2126,
+      "step": 9965
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.6306429485233906,
+      "learning_rate": 4.364379608175777e-06,
+      "loss": 0.2204,
+      "step": 9966
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.083467047337678,
+      "learning_rate": 4.363486086250875e-06,
+      "loss": 0.1926,
+      "step": 9967
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.152079032945396,
+      "learning_rate": 4.362592584987578e-06,
+      "loss": 0.1463,
+      "step": 9968
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.095882392606758,
+      "learning_rate": 4.361699104414894e-06,
+      "loss": 0.1727,
+      "step": 9969
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.0407006966492605,
+      "learning_rate": 4.360805644561822e-06,
+      "loss": 0.1286,
+      "step": 9970
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.9931014218451334,
+      "learning_rate": 4.359912205457369e-06,
+      "loss": 0.1973,
+      "step": 9971
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.046712990935866,
+      "learning_rate": 4.359018787130532e-06,
+      "loss": 0.1907,
+      "step": 9972
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.3305318215451356,
+      "learning_rate": 4.358125389610313e-06,
+      "loss": 0.197,
+      "step": 9973
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.3491795154780057,
+      "learning_rate": 4.357232012925714e-06,
+      "loss": 0.2089,
+      "step": 9974
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.1764445920834916,
+      "learning_rate": 4.356338657105733e-06,
+      "loss": 0.2133,
+      "step": 9975
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.345413445352888,
+      "learning_rate": 4.355445322179372e-06,
+      "loss": 0.2159,
+      "step": 9976
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.2679162904225576,
+      "learning_rate": 4.354552008175623e-06,
+      "loss": 0.2473,
+      "step": 9977
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.4028330377030074,
+      "learning_rate": 4.353658715123489e-06,
+      "loss": 0.1909,
+      "step": 9978
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.348506642598944,
+      "learning_rate": 4.352765443051963e-06,
+      "loss": 0.1347,
+      "step": 9979
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.3728006408013316,
+      "learning_rate": 4.351872191990046e-06,
+      "loss": 0.1851,
+      "step": 9980
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.3324970300772851,
+      "learning_rate": 4.350978961966728e-06,
+      "loss": 0.1853,
+      "step": 9981
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.3417188463469796,
+      "learning_rate": 4.350085753011005e-06,
+      "loss": 0.1929,
+      "step": 9982
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.3040772027138243,
+      "learning_rate": 4.349192565151874e-06,
+      "loss": 0.1757,
+      "step": 9983
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.2515098483313094,
+      "learning_rate": 4.348299398418325e-06,
+      "loss": 0.2105,
+      "step": 9984
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.2500036299175963,
+      "learning_rate": 4.347406252839355e-06,
+      "loss": 0.171,
+      "step": 9985
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.178579579123629,
+      "learning_rate": 4.346513128443951e-06,
+      "loss": 0.1994,
+      "step": 9986
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.278603526041566,
+      "learning_rate": 4.345620025261108e-06,
+      "loss": 0.1204,
+      "step": 9987
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.2210477296793356,
+      "learning_rate": 4.344726943319814e-06,
+      "loss": 0.1863,
+      "step": 9988
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.2593334565634744,
+      "learning_rate": 4.343833882649062e-06,
+      "loss": 0.223,
+      "step": 9989
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.3063390694600008,
+      "learning_rate": 4.342940843277838e-06,
+      "loss": 0.1768,
+      "step": 9990
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.2285355837706247,
+      "learning_rate": 4.342047825235132e-06,
+      "loss": 0.2012,
+      "step": 9991
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.4240568105564306,
+      "learning_rate": 4.341154828549932e-06,
+      "loss": 0.2186,
+      "step": 9992
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.9336090086643686,
+      "learning_rate": 4.340261853251225e-06,
+      "loss": 0.1818,
+      "step": 9993
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.2242084147924466,
+      "learning_rate": 4.339368899368e-06,
+      "loss": 0.1546,
+      "step": 9994
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.3582445930994376,
+      "learning_rate": 4.338475966929238e-06,
+      "loss": 0.2088,
+      "step": 9995
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.4794445319731178,
+      "learning_rate": 4.3375830559639275e-06,
+      "loss": 0.1873,
+      "step": 9996
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.4061474365343736,
+      "learning_rate": 4.336690166501051e-06,
+      "loss": 0.2058,
+      "step": 9997
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.2098560549496893,
+      "learning_rate": 4.335797298569593e-06,
+      "loss": 0.1703,
+      "step": 9998
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.019317966382435,
+      "learning_rate": 4.334904452198539e-06,
+      "loss": 0.2324,
+      "step": 9999
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.1752499682266793,
+      "learning_rate": 4.334011627416867e-06,
+      "loss": 0.1734,
+      "step": 10000
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.7853153583564492,
+      "learning_rate": 4.33311882425356e-06,
+      "loss": 0.2193,
+      "step": 10001
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.2178913356116516,
+      "learning_rate": 4.3322260427376e-06,
+      "loss": 0.1751,
+      "step": 10002
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.3614053362236167,
+      "learning_rate": 4.331333282897968e-06,
+      "loss": 0.2288,
+      "step": 10003
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.858456127975246,
+      "learning_rate": 4.33044054476364e-06,
+      "loss": 0.1707,
+      "step": 10004
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.416784133201752,
+      "learning_rate": 4.329547828363599e-06,
+      "loss": 0.2146,
+      "step": 10005
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.4315529272117056,
+      "learning_rate": 4.328655133726819e-06,
+      "loss": 0.1813,
+      "step": 10006
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.9829681917539754,
+      "learning_rate": 4.32776246088228e-06,
+      "loss": 0.1721,
+      "step": 10007
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.9687719265702137,
+      "learning_rate": 4.32686980985896e-06,
+      "loss": 0.2453,
+      "step": 10008
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.457656940775309,
+      "learning_rate": 4.325977180685831e-06,
+      "loss": 0.2363,
+      "step": 10009
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.1951532787665589,
+      "learning_rate": 4.32508457339187e-06,
+      "loss": 0.174,
+      "step": 10010
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.1274300397495731,
+      "learning_rate": 4.324191988006052e-06,
+      "loss": 0.1946,
+      "step": 10011
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.7306777124830238,
+      "learning_rate": 4.323299424557354e-06,
+      "loss": 0.1735,
+      "step": 10012
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.459309573616323,
+      "learning_rate": 4.322406883074742e-06,
+      "loss": 0.222,
+      "step": 10013
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.1302577165796068,
+      "learning_rate": 4.3215143635871935e-06,
+      "loss": 0.1808,
+      "step": 10014
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.1749786571583263,
+      "learning_rate": 4.320621866123679e-06,
+      "loss": 0.1435,
+      "step": 10015
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.0730700576105487,
+      "learning_rate": 4.319729390713168e-06,
+      "loss": 0.2425,
+      "step": 10016
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.1408095096258863,
+      "learning_rate": 4.318836937384635e-06,
+      "loss": 0.1807,
+      "step": 10017
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.0900091332908206,
+      "learning_rate": 4.317944506167044e-06,
+      "loss": 0.1704,
+      "step": 10018
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.9693821505226385,
+      "learning_rate": 4.317052097089368e-06,
+      "loss": 0.2647,
+      "step": 10019
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.453663844009876,
+      "learning_rate": 4.316159710180572e-06,
+      "loss": 0.1847,
+      "step": 10020
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.4276028010201651,
+      "learning_rate": 4.315267345469628e-06,
+      "loss": 0.2377,
+      "step": 10021
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.318598628057142,
+      "learning_rate": 4.314375002985498e-06,
+      "loss": 0.168,
+      "step": 10022
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.3698666738478482,
+      "learning_rate": 4.31348268275715e-06,
+      "loss": 0.235,
+      "step": 10023
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.2264463557998795,
+      "learning_rate": 4.312590384813548e-06,
+      "loss": 0.2384,
+      "step": 10024
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.3208570627007383,
+      "learning_rate": 4.3116981091836575e-06,
+      "loss": 0.1765,
+      "step": 10025
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.297410548465004,
+      "learning_rate": 4.310805855896444e-06,
+      "loss": 0.1871,
+      "step": 10026
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.3306598539522427,
+      "learning_rate": 4.309913624980866e-06,
+      "loss": 0.2128,
+      "step": 10027
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.2803576388629352,
+      "learning_rate": 4.30902141646589e-06,
+      "loss": 0.2228,
+      "step": 10028
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.2146979210712996,
+      "learning_rate": 4.308129230380477e-06,
+      "loss": 0.192,
+      "step": 10029
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.0212337120699637,
+      "learning_rate": 4.307237066753586e-06,
+      "loss": 0.2174,
+      "step": 10030
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.0289504954094326,
+      "learning_rate": 4.3063449256141785e-06,
+      "loss": 0.1722,
+      "step": 10031
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.3381547875899584,
+      "learning_rate": 4.305452806991214e-06,
+      "loss": 0.188,
+      "step": 10032
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.2752660921295218,
+      "learning_rate": 4.304560710913649e-06,
+      "loss": 0.1827,
+      "step": 10033
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.2785507916202707,
+      "learning_rate": 4.303668637410444e-06,
+      "loss": 0.199,
+      "step": 10034
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.2467909584428505,
+      "learning_rate": 4.302776586510558e-06,
+      "loss": 0.2053,
+      "step": 10035
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.3669503578515143,
+      "learning_rate": 4.301884558242941e-06,
+      "loss": 0.1821,
+      "step": 10036
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.1173542905226839,
+      "learning_rate": 4.300992552636555e-06,
+      "loss": 0.2206,
+      "step": 10037
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.9350827126841033,
+      "learning_rate": 4.300100569720353e-06,
+      "loss": 0.1767,
+      "step": 10038
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.6508772831235716,
+      "learning_rate": 4.299208609523287e-06,
+      "loss": 0.1904,
+      "step": 10039
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.1742983049357727,
+      "learning_rate": 4.298316672074314e-06,
+      "loss": 0.1953,
+      "step": 10040
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.8916853314894272,
+      "learning_rate": 4.297424757402385e-06,
+      "loss": 0.1616,
+      "step": 10041
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.422868852896483,
+      "learning_rate": 4.296532865536453e-06,
+      "loss": 0.1779,
+      "step": 10042
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.2449003920962274,
+      "learning_rate": 4.295640996505467e-06,
+      "loss": 0.2354,
+      "step": 10043
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.456516451055543,
+      "learning_rate": 4.294749150338382e-06,
+      "loss": 0.2112,
+      "step": 10044
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.4258027819745134,
+      "learning_rate": 4.293857327064143e-06,
+      "loss": 0.1915,
+      "step": 10045
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.253048496455776,
+      "learning_rate": 4.292965526711701e-06,
+      "loss": 0.2197,
+      "step": 10046
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.0399536384280808,
+      "learning_rate": 4.292073749310006e-06,
+      "loss": 0.1953,
+      "step": 10047
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.1748968613028739,
+      "learning_rate": 4.291181994888004e-06,
+      "loss": 0.1768,
+      "step": 10048
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.2247208844146542,
+      "learning_rate": 4.290290263474642e-06,
+      "loss": 0.2334,
+      "step": 10049
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.0756196926424886,
+      "learning_rate": 4.289398555098865e-06,
+      "loss": 0.1892,
+      "step": 10050
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.0775154159576008,
+      "learning_rate": 4.28850686978962e-06,
+      "loss": 0.1365,
+      "step": 10051
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.4735234689561896,
+      "learning_rate": 4.28761520757585e-06,
+      "loss": 0.1799,
+      "step": 10052
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.154451395397358,
+      "learning_rate": 4.2867235684865035e-06,
+      "loss": 0.1631,
+      "step": 10053
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.2436261063838356,
+      "learning_rate": 4.285831952550518e-06,
+      "loss": 0.1922,
+      "step": 10054
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.0468336780420797,
+      "learning_rate": 4.284940359796837e-06,
+      "loss": 0.1527,
+      "step": 10055
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.2401661561241315,
+      "learning_rate": 4.284048790254405e-06,
+      "loss": 0.1123,
+      "step": 10056
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.1605128950641834,
+      "learning_rate": 4.28315724395216e-06,
+      "loss": 0.2044,
+      "step": 10057
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.5919854690134805,
+      "learning_rate": 4.282265720919044e-06,
+      "loss": 0.2097,
+      "step": 10058
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.1910629121695908,
+      "learning_rate": 4.281374221183994e-06,
+      "loss": 0.2076,
+      "step": 10059
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.292315718667703,
+      "learning_rate": 4.28048274477595e-06,
+      "loss": 0.2001,
+      "step": 10060
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.583959652182121,
+      "learning_rate": 4.279591291723849e-06,
+      "loss": 0.2456,
+      "step": 10061
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.2298232519840346,
+      "learning_rate": 4.2786998620566326e-06,
+      "loss": 0.2072,
+      "step": 10062
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.3823108369266457,
+      "learning_rate": 4.277808455803231e-06,
+      "loss": 0.1493,
+      "step": 10063
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.0317160933022882,
+      "learning_rate": 4.276917072992582e-06,
+      "loss": 0.1834,
+      "step": 10064
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.0785642606717873,
+      "learning_rate": 4.276025713653621e-06,
+      "loss": 0.1698,
+      "step": 10065
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.2933737014073827,
+      "learning_rate": 4.275134377815282e-06,
+      "loss": 0.2158,
+      "step": 10066
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.1398185547442317,
+      "learning_rate": 4.274243065506499e-06,
+      "loss": 0.1706,
+      "step": 10067
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.181891387939812,
+      "learning_rate": 4.273351776756201e-06,
+      "loss": 0.1887,
+      "step": 10068
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.4801024484536054,
+      "learning_rate": 4.2724605115933235e-06,
+      "loss": 0.2241,
+      "step": 10069
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.090047010656954,
+      "learning_rate": 4.271569270046795e-06,
+      "loss": 0.1878,
+      "step": 10070
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.7202906790829742,
+      "learning_rate": 4.270678052145549e-06,
+      "loss": 0.2007,
+      "step": 10071
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.3992825943498888,
+      "learning_rate": 4.269786857918512e-06,
+      "loss": 0.2146,
+      "step": 10072
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.3538098476491884,
+      "learning_rate": 4.2688956873946115e-06,
+      "loss": 0.1894,
+      "step": 10073
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.2465363691000535,
+      "learning_rate": 4.26800454060278e-06,
+      "loss": 0.1504,
+      "step": 10074
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.3998610001770138,
+      "learning_rate": 4.267113417571942e-06,
+      "loss": 0.1538,
+      "step": 10075
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.7708555407160107,
+      "learning_rate": 4.266222318331023e-06,
+      "loss": 0.203,
+      "step": 10076
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.0725052900978593,
+      "learning_rate": 4.265331242908949e-06,
+      "loss": 0.186,
+      "step": 10077
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.5219468820425353,
+      "learning_rate": 4.264440191334645e-06,
+      "loss": 0.2341,
+      "step": 10078
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.4568756189144525,
+      "learning_rate": 4.263549163637035e-06,
+      "loss": 0.2505,
+      "step": 10079
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.231924328370392,
+      "learning_rate": 4.262658159845046e-06,
+      "loss": 0.2187,
+      "step": 10080
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.9990540291883786,
+      "learning_rate": 4.261767179987595e-06,
+      "loss": 0.1741,
+      "step": 10081
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.1971057486396728,
+      "learning_rate": 4.260876224093605e-06,
+      "loss": 0.192,
+      "step": 10082
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.371681115490008,
+      "learning_rate": 4.259985292191999e-06,
+      "loss": 0.2279,
+      "step": 10083
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.524705007428738,
+      "learning_rate": 4.259094384311696e-06,
+      "loss": 0.1857,
+      "step": 10084
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.4865120687464757,
+      "learning_rate": 4.258203500481615e-06,
+      "loss": 0.2061,
+      "step": 10085
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.3333057168494677,
+      "learning_rate": 4.257312640730674e-06,
+      "loss": 0.2414,
+      "step": 10086
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.3397116971350043,
+      "learning_rate": 4.256421805087792e-06,
+      "loss": 0.1581,
+      "step": 10087
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.8920826026860653,
+      "learning_rate": 4.2555309935818865e-06,
+      "loss": 0.1239,
+      "step": 10088
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.1474519120531017,
+      "learning_rate": 4.254640206241875e-06,
+      "loss": 0.1865,
+      "step": 10089
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.101802291709386,
+      "learning_rate": 4.253749443096667e-06,
+      "loss": 0.176,
+      "step": 10090
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.0784918430501385,
+      "learning_rate": 4.252858704175183e-06,
+      "loss": 0.2204,
+      "step": 10091
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.1928349948937622,
+      "learning_rate": 4.251967989506336e-06,
+      "loss": 0.1647,
+      "step": 10092
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.8444878023322833,
+      "learning_rate": 4.251077299119038e-06,
+      "loss": 0.1992,
+      "step": 10093
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.9783699189851386,
+      "learning_rate": 4.250186633042201e-06,
+      "loss": 0.1536,
+      "step": 10094
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.8884104771679364,
+      "learning_rate": 4.249295991304737e-06,
+      "loss": 0.2202,
+      "step": 10095
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.576497765120351,
+      "learning_rate": 4.248405373935556e-06,
+      "loss": 0.2248,
+      "step": 10096
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.8029694636381763,
+      "learning_rate": 4.247514780963569e-06,
+      "loss": 0.244,
+      "step": 10097
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.1939679044298732,
+      "learning_rate": 4.2466242124176875e-06,
+      "loss": 0.1776,
+      "step": 10098
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.9277827260317464,
+      "learning_rate": 4.2457336683268145e-06,
+      "loss": 0.1493,
+      "step": 10099
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.1705033253686055,
+      "learning_rate": 4.24484314871986e-06,
+      "loss": 0.2017,
+      "step": 10100
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.4644978636234078,
+      "learning_rate": 4.243952653625733e-06,
+      "loss": 0.1673,
+      "step": 10101
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.4154890725685205,
+      "learning_rate": 4.243062183073338e-06,
+      "loss": 0.2394,
+      "step": 10102
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.2233438782417214,
+      "learning_rate": 4.242171737091579e-06,
+      "loss": 0.214,
+      "step": 10103
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.209998184874647,
+      "learning_rate": 4.24128131570936e-06,
+      "loss": 0.2198,
+      "step": 10104
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.4596802687297403,
+      "learning_rate": 4.2403909189555865e-06,
+      "loss": 0.2379,
+      "step": 10105
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.3220275878765355,
+      "learning_rate": 4.23950054685916e-06,
+      "loss": 0.1191,
+      "step": 10106
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.4549034372220546,
+      "learning_rate": 4.238610199448986e-06,
+      "loss": 0.1458,
+      "step": 10107
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.0691046564827549,
+      "learning_rate": 4.23771987675396e-06,
+      "loss": 0.1884,
+      "step": 10108
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.3848281047268265,
+      "learning_rate": 4.236829578802985e-06,
+      "loss": 0.1776,
+      "step": 10109
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.193894121322861,
+      "learning_rate": 4.235939305624962e-06,
+      "loss": 0.229,
+      "step": 10110
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.6774112325827044,
+      "learning_rate": 4.2350490572487885e-06,
+      "loss": 0.1862,
+      "step": 10111
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.105314722552342,
+      "learning_rate": 4.234158833703363e-06,
+      "loss": 0.198,
+      "step": 10112
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.1732781719838479,
+      "learning_rate": 4.23326863501758e-06,
+      "loss": 0.1928,
+      "step": 10113
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.1034823792437385,
+      "learning_rate": 4.232378461220339e-06,
+      "loss": 0.1676,
+      "step": 10114
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.185878949108406,
+      "learning_rate": 4.2314883123405335e-06,
+      "loss": 0.2016,
+      "step": 10115
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.5627650870522656,
+      "learning_rate": 4.230598188407062e-06,
+      "loss": 0.2462,
+      "step": 10116
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.1236860413511365,
+      "learning_rate": 4.2297080894488135e-06,
+      "loss": 0.2426,
+      "step": 10117
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.1646637835747846,
+      "learning_rate": 4.228818015494683e-06,
+      "loss": 0.1913,
+      "step": 10118
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.5421771541932527,
+      "learning_rate": 4.227927966573566e-06,
+      "loss": 0.2228,
+      "step": 10119
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.825265044025526,
+      "learning_rate": 4.22703794271435e-06,
+      "loss": 0.1221,
+      "step": 10120
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.164823046832246,
+      "learning_rate": 4.226147943945927e-06,
+      "loss": 0.2187,
+      "step": 10121
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.3450323517623965,
+      "learning_rate": 4.225257970297185e-06,
+      "loss": 0.1982,
+      "step": 10122
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.312909947814666,
+      "learning_rate": 4.224368021797015e-06,
+      "loss": 0.2329,
+      "step": 10123
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.2312610315903378,
+      "learning_rate": 4.223478098474305e-06,
+      "loss": 0.2187,
+      "step": 10124
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.0333618067447956,
+      "learning_rate": 4.2225882003579455e-06,
+      "loss": 0.1895,
+      "step": 10125
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.9806530235959279,
+      "learning_rate": 4.221698327476818e-06,
+      "loss": 0.207,
+      "step": 10126
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.368365426122303,
+      "learning_rate": 4.220808479859809e-06,
+      "loss": 0.2066,
+      "step": 10127
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.9906026693288126,
+      "learning_rate": 4.219918657535806e-06,
+      "loss": 0.1601,
+      "step": 10128
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 3.475849378505034,
+      "learning_rate": 4.2190288605336924e-06,
+      "loss": 0.2148,
+      "step": 10129
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.519301139647838,
+      "learning_rate": 4.2181390888823515e-06,
+      "loss": 0.2255,
+      "step": 10130
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.1157371194515564,
+      "learning_rate": 4.217249342610664e-06,
+      "loss": 0.1891,
+      "step": 10131
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.0914553413464176,
+      "learning_rate": 4.216359621747513e-06,
+      "loss": 0.2022,
+      "step": 10132
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.5530497104809444,
+      "learning_rate": 4.2154699263217794e-06,
+      "loss": 0.1922,
+      "step": 10133
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.7737550180053689,
+      "learning_rate": 4.214580256362345e-06,
+      "loss": 0.1873,
+      "step": 10134
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.0854127562898475,
+      "learning_rate": 4.213690611898085e-06,
+      "loss": 0.1833,
+      "step": 10135
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.1090438274586263,
+      "learning_rate": 4.212800992957881e-06,
+      "loss": 0.1344,
+      "step": 10136
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.309478511906738,
+      "learning_rate": 4.211911399570609e-06,
+      "loss": 0.2355,
+      "step": 10137
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.3963276568420373,
+      "learning_rate": 4.211021831765146e-06,
+      "loss": 0.2483,
+      "step": 10138
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.9714833717816506,
+      "learning_rate": 4.2101322895703685e-06,
+      "loss": 0.1134,
+      "step": 10139
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.1802923060551334,
+      "learning_rate": 4.209242773015151e-06,
+      "loss": 0.1834,
+      "step": 10140
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.3771761774437732,
+      "learning_rate": 4.208353282128367e-06,
+      "loss": 0.1827,
+      "step": 10141
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.8652272730034266,
+      "learning_rate": 4.20746381693889e-06,
+      "loss": 0.203,
+      "step": 10142
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.0712196897928083,
+      "learning_rate": 4.2065743774755965e-06,
+      "loss": 0.1626,
+      "step": 10143
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.1716468271008798,
+      "learning_rate": 4.2056849637673526e-06,
+      "loss": 0.1859,
+      "step": 10144
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.3115903250765055,
+      "learning_rate": 4.204795575843031e-06,
+      "loss": 0.2581,
+      "step": 10145
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.9481029614541384,
+      "learning_rate": 4.2039062137315045e-06,
+      "loss": 0.1475,
+      "step": 10146
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.269173226857978,
+      "learning_rate": 4.203016877461639e-06,
+      "loss": 0.1869,
+      "step": 10147
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.3744847106990425,
+      "learning_rate": 4.202127567062304e-06,
+      "loss": 0.1591,
+      "step": 10148
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.9974192775514118,
+      "learning_rate": 4.201238282562367e-06,
+      "loss": 0.2029,
+      "step": 10149
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.1495858374417776,
+      "learning_rate": 4.200349023990693e-06,
+      "loss": 0.1986,
+      "step": 10150
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.4254721606982212,
+      "learning_rate": 4.19945979137615e-06,
+      "loss": 0.1789,
+      "step": 10151
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.138689567729451,
+      "learning_rate": 4.198570584747606e-06,
+      "loss": 0.1474,
+      "step": 10152
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.9711164941372251,
+      "learning_rate": 4.197681404133918e-06,
+      "loss": 0.1738,
+      "step": 10153
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.6791870081247666,
+      "learning_rate": 4.196792249563953e-06,
+      "loss": 0.172,
+      "step": 10154
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.208733709845889,
+      "learning_rate": 4.195903121066576e-06,
+      "loss": 0.1818,
+      "step": 10155
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.3919015447297627,
+      "learning_rate": 4.195014018670645e-06,
+      "loss": 0.1609,
+      "step": 10156
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.1628181988525987,
+      "learning_rate": 4.194124942405023e-06,
+      "loss": 0.1935,
+      "step": 10157
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.5944940125314697,
+      "learning_rate": 4.193235892298566e-06,
+      "loss": 0.1766,
+      "step": 10158
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.3865662074474732,
+      "learning_rate": 4.192346868380136e-06,
+      "loss": 0.184,
+      "step": 10159
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.9146390017980505,
+      "learning_rate": 4.191457870678592e-06,
+      "loss": 0.2137,
+      "step": 10160
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.231361041411135,
+      "learning_rate": 4.190568899222792e-06,
+      "loss": 0.1946,
+      "step": 10161
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.7355690059414228,
+      "learning_rate": 4.189679954041589e-06,
+      "loss": 0.2008,
+      "step": 10162
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.1596488517390162,
+      "learning_rate": 4.188791035163841e-06,
+      "loss": 0.2173,
+      "step": 10163
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.5298251436902257,
+      "learning_rate": 4.187902142618403e-06,
+      "loss": 0.1947,
+      "step": 10164
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.2111485620344173,
+      "learning_rate": 4.187013276434129e-06,
+      "loss": 0.2289,
+      "step": 10165
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.3353556598479395,
+      "learning_rate": 4.186124436639871e-06,
+      "loss": 0.163,
+      "step": 10166
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.522427561591696,
+      "learning_rate": 4.185235623264481e-06,
+      "loss": 0.1679,
+      "step": 10167
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.0118498404274892,
+      "learning_rate": 4.184346836336812e-06,
+      "loss": 0.1392,
+      "step": 10168
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.1670103686205298,
+      "learning_rate": 4.183458075885712e-06,
+      "loss": 0.1556,
+      "step": 10169
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.1391983418251426,
+      "learning_rate": 4.182569341940036e-06,
+      "loss": 0.2013,
+      "step": 10170
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.5847694718500152,
+      "learning_rate": 4.1816806345286265e-06,
+      "loss": 0.172,
+      "step": 10171
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.6505288715132709,
+      "learning_rate": 4.180791953680334e-06,
+      "loss": 0.2069,
+      "step": 10172
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.341348453758504,
+      "learning_rate": 4.1799032994240075e-06,
+      "loss": 0.1716,
+      "step": 10173
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.432514704475694,
+      "learning_rate": 4.179014671788491e-06,
+      "loss": 0.1742,
+      "step": 10174
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.7424257448182464,
+      "learning_rate": 4.17812607080263e-06,
+      "loss": 0.1447,
+      "step": 10175
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.313651025806853,
+      "learning_rate": 4.177237496495269e-06,
+      "loss": 0.2081,
+      "step": 10176
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.996076299477102,
+      "learning_rate": 4.176348948895252e-06,
+      "loss": 0.186,
+      "step": 10177
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.9555316815428796,
+      "learning_rate": 4.175460428031422e-06,
+      "loss": 0.1967,
+      "step": 10178
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.4578890359255696,
+      "learning_rate": 4.174571933932621e-06,
+      "loss": 0.2384,
+      "step": 10179
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.3616777734505823,
+      "learning_rate": 4.173683466627689e-06,
+      "loss": 0.2079,
+      "step": 10180
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.7093669550321071,
+      "learning_rate": 4.1727950261454665e-06,
+      "loss": 0.187,
+      "step": 10181
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.2575841853442529,
+      "learning_rate": 4.171906612514794e-06,
+      "loss": 0.178,
+      "step": 10182
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.3275750003118607,
+      "learning_rate": 4.17101822576451e-06,
+      "loss": 0.2367,
+      "step": 10183
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.224540212646371,
+      "learning_rate": 4.17012986592345e-06,
+      "loss": 0.221,
+      "step": 10184
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.2272627769192876,
+      "learning_rate": 4.169241533020452e-06,
+      "loss": 0.1753,
+      "step": 10185
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.1168936129472258,
+      "learning_rate": 4.1683532270843505e-06,
+      "loss": 0.15,
+      "step": 10186
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.327839644013892,
+      "learning_rate": 4.167464948143981e-06,
+      "loss": 0.1762,
+      "step": 10187
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.2677367747378085,
+      "learning_rate": 4.166576696228182e-06,
+      "loss": 0.2151,
+      "step": 10188
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.0208643769471344,
+      "learning_rate": 4.1656884713657785e-06,
+      "loss": 0.1637,
+      "step": 10189
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.0961584277385454,
+      "learning_rate": 4.164800273585608e-06,
+      "loss": 0.1493,
+      "step": 10190
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.8334451804005643,
+      "learning_rate": 4.163912102916502e-06,
+      "loss": 0.1347,
+      "step": 10191
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.2374019930783873,
+      "learning_rate": 4.16302395938729e-06,
+      "loss": 0.1926,
+      "step": 10192
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.1123914391746124,
+      "learning_rate": 4.162135843026801e-06,
+      "loss": 0.1901,
+      "step": 10193
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.5134824080482945,
+      "learning_rate": 4.1612477538638635e-06,
+      "loss": 0.2598,
+      "step": 10194
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.0297740286329427,
+      "learning_rate": 4.160359691927306e-06,
+      "loss": 0.2047,
+      "step": 10195
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.222879014729883,
+      "learning_rate": 4.159471657245956e-06,
+      "loss": 0.162,
+      "step": 10196
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.118874380054415,
+      "learning_rate": 4.158583649848642e-06,
+      "loss": 0.1617,
+      "step": 10197
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.2906952661039852,
+      "learning_rate": 4.157695669764183e-06,
+      "loss": 0.2044,
+      "step": 10198
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.2395131604850897,
+      "learning_rate": 4.1568077170214084e-06,
+      "loss": 0.1684,
+      "step": 10199
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.2690309154823152,
+      "learning_rate": 4.155919791649141e-06,
+      "loss": 0.1768,
+      "step": 10200
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.3184951726415872,
+      "learning_rate": 4.155031893676202e-06,
+      "loss": 0.2067,
+      "step": 10201
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.4039504163669514,
+      "learning_rate": 4.1541440231314145e-06,
+      "loss": 0.2115,
+      "step": 10202
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.4166894743172658,
+      "learning_rate": 4.153256180043596e-06,
+      "loss": 0.184,
+      "step": 10203
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.7844753848009063,
+      "learning_rate": 4.15236836444157e-06,
+      "loss": 0.1367,
+      "step": 10204
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.1191443198577442,
+      "learning_rate": 4.151480576354154e-06,
+      "loss": 0.1738,
+      "step": 10205
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.4239158237068037,
+      "learning_rate": 4.150592815810168e-06,
+      "loss": 0.2224,
+      "step": 10206
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.4146985894255815,
+      "learning_rate": 4.149705082838426e-06,
+      "loss": 0.1314,
+      "step": 10207
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.192009716726989,
+      "learning_rate": 4.148817377467746e-06,
+      "loss": 0.1816,
+      "step": 10208
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.1674023438163748,
+      "learning_rate": 4.147929699726945e-06,
+      "loss": 0.2047,
+      "step": 10209
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.9236568832568234,
+      "learning_rate": 4.147042049644835e-06,
+      "loss": 0.1766,
+      "step": 10210
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.0036524217973652,
+      "learning_rate": 4.146154427250231e-06,
+      "loss": 0.2116,
+      "step": 10211
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.4713311190755227,
+      "learning_rate": 4.145266832571944e-06,
+      "loss": 0.1727,
+      "step": 10212
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.6339611115647434,
+      "learning_rate": 4.1443792656387875e-06,
+      "loss": 0.242,
+      "step": 10213
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.1657991352097479,
+      "learning_rate": 4.143491726479571e-06,
+      "loss": 0.194,
+      "step": 10214
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.3805388007015043,
+      "learning_rate": 4.142604215123108e-06,
+      "loss": 0.1794,
+      "step": 10215
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.4231561925063856,
+      "learning_rate": 4.141716731598205e-06,
+      "loss": 0.1706,
+      "step": 10216
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.207401144885806,
+      "learning_rate": 4.140829275933667e-06,
+      "loss": 0.1772,
+      "step": 10217
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.9566014706693643,
+      "learning_rate": 4.139941848158309e-06,
+      "loss": 0.17,
+      "step": 10218
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.0301497217900435,
+      "learning_rate": 4.139054448300932e-06,
+      "loss": 0.2164,
+      "step": 10219
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.0462353766466503,
+      "learning_rate": 4.1381670763903425e-06,
+      "loss": 0.1739,
+      "step": 10220
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.1401172350376592,
+      "learning_rate": 4.1372797324553445e-06,
+      "loss": 0.1665,
+      "step": 10221
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.2991306755643366,
+      "learning_rate": 4.136392416524742e-06,
+      "loss": 0.1798,
+      "step": 10222
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.176694657779773,
+      "learning_rate": 4.135505128627339e-06,
+      "loss": 0.2102,
+      "step": 10223
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.3912145966697684,
+      "learning_rate": 4.134617868791938e-06,
+      "loss": 0.2078,
+      "step": 10224
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.0663066929327656,
+      "learning_rate": 4.133730637047337e-06,
+      "loss": 0.175,
+      "step": 10225
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.4935176647731276,
+      "learning_rate": 4.132843433422337e-06,
+      "loss": 0.2285,
+      "step": 10226
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.5134467074366005,
+      "learning_rate": 4.13195625794574e-06,
+      "loss": 0.2025,
+      "step": 10227
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.5431311509823642,
+      "learning_rate": 4.131069110646342e-06,
+      "loss": 0.2266,
+      "step": 10228
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.1992299921210872,
+      "learning_rate": 4.130181991552939e-06,
+      "loss": 0.168,
+      "step": 10229
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.9967178371812175,
+      "learning_rate": 4.129294900694329e-06,
+      "loss": 0.2324,
+      "step": 10230
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.131325459334565,
+      "learning_rate": 4.128407838099306e-06,
+      "loss": 0.2045,
+      "step": 10231
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.8039709295189978,
+      "learning_rate": 4.127520803796667e-06,
+      "loss": 0.1714,
+      "step": 10232
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.3184777948388846,
+      "learning_rate": 4.126633797815206e-06,
+      "loss": 0.1703,
+      "step": 10233
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.1046646364670625,
+      "learning_rate": 4.125746820183712e-06,
+      "loss": 0.1663,
+      "step": 10234
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.3198727067119476,
+      "learning_rate": 4.1248598709309786e-06,
+      "loss": 0.2215,
+      "step": 10235
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.3023449202671298,
+      "learning_rate": 4.123972950085799e-06,
+      "loss": 0.2048,
+      "step": 10236
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.8894641697951335,
+      "learning_rate": 4.123086057676961e-06,
+      "loss": 0.1621,
+      "step": 10237
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.9898193211841632,
+      "learning_rate": 4.122199193733253e-06,
+      "loss": 0.1635,
+      "step": 10238
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.5146908388284002,
+      "learning_rate": 4.121312358283464e-06,
+      "loss": 0.1799,
+      "step": 10239
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.6473382747008822,
+      "learning_rate": 4.12042555135638e-06,
+      "loss": 0.24,
+      "step": 10240
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.2933373444460068,
+      "learning_rate": 4.119538772980789e-06,
+      "loss": 0.2042,
+      "step": 10241
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.4383572634492445,
+      "learning_rate": 4.1186520231854775e-06,
+      "loss": 0.2379,
+      "step": 10242
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.0474672243696754,
+      "learning_rate": 4.117765301999226e-06,
+      "loss": 0.1784,
+      "step": 10243
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.2354691038641392,
+      "learning_rate": 4.116878609450821e-06,
+      "loss": 0.2251,
+      "step": 10244
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.266903981595735,
+      "learning_rate": 4.115991945569044e-06,
+      "loss": 0.2368,
+      "step": 10245
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.4753479276482693,
+      "learning_rate": 4.1151053103826766e-06,
+      "loss": 0.2584,
+      "step": 10246
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.8387137529163259,
+      "learning_rate": 4.114218703920501e-06,
+      "loss": 0.1537,
+      "step": 10247
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.4732732972591807,
+      "learning_rate": 4.113332126211295e-06,
+      "loss": 0.1873,
+      "step": 10248
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.3842730648237027,
+      "learning_rate": 4.112445577283837e-06,
+      "loss": 0.1718,
+      "step": 10249
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.4101692579520098,
+      "learning_rate": 4.111559057166906e-06,
+      "loss": 0.1957,
+      "step": 10250
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.9917419026221909,
+      "learning_rate": 4.110672565889281e-06,
+      "loss": 0.147,
+      "step": 10251
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.3425240024893048,
+      "learning_rate": 4.109786103479733e-06,
+      "loss": 0.1884,
+      "step": 10252
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.1452178457992865,
+      "learning_rate": 4.1088996699670416e-06,
+      "loss": 0.2126,
+      "step": 10253
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.514384696227239,
+      "learning_rate": 4.1080132653799794e-06,
+      "loss": 0.1947,
+      "step": 10254
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.1337521273488922,
+      "learning_rate": 4.107126889747319e-06,
+      "loss": 0.1831,
+      "step": 10255
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.282846982006945,
+      "learning_rate": 4.106240543097835e-06,
+      "loss": 0.1885,
+      "step": 10256
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.0796008115959235,
+      "learning_rate": 4.105354225460295e-06,
+      "loss": 0.1694,
+      "step": 10257
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.3156462125132768,
+      "learning_rate": 4.104467936863471e-06,
+      "loss": 0.1964,
+      "step": 10258
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.5105573594129242,
+      "learning_rate": 4.103581677336133e-06,
+      "loss": 0.2098,
+      "step": 10259
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.1193749214163817,
+      "learning_rate": 4.102695446907051e-06,
+      "loss": 0.195,
+      "step": 10260
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.146559335899718,
+      "learning_rate": 4.101809245604989e-06,
+      "loss": 0.1854,
+      "step": 10261
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.1198973302742772,
+      "learning_rate": 4.100923073458715e-06,
+      "loss": 0.163,
+      "step": 10262
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.2020858962201388,
+      "learning_rate": 4.100036930496996e-06,
+      "loss": 0.1814,
+      "step": 10263
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.070956836745155,
+      "learning_rate": 4.099150816748596e-06,
+      "loss": 0.1753,
+      "step": 10264
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.034574839303231,
+      "learning_rate": 4.098264732242279e-06,
+      "loss": 0.225,
+      "step": 10265
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.4676675409325168,
+      "learning_rate": 4.0973786770068055e-06,
+      "loss": 0.1905,
+      "step": 10266
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.9958206176278507,
+      "learning_rate": 4.09649265107094e-06,
+      "loss": 0.1816,
+      "step": 10267
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.249227032982883,
+      "learning_rate": 4.095606654463442e-06,
+      "loss": 0.1825,
+      "step": 10268
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.0603546895774871,
+      "learning_rate": 4.094720687213075e-06,
+      "loss": 0.195,
+      "step": 10269
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.0184438138545435,
+      "learning_rate": 4.093834749348592e-06,
+      "loss": 0.2002,
+      "step": 10270
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.4862383891823965,
+      "learning_rate": 4.092948840898755e-06,
+      "loss": 0.2324,
+      "step": 10271
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.1326446474617466,
+      "learning_rate": 4.09206296189232e-06,
+      "loss": 0.1673,
+      "step": 10272
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.3761229670699555,
+      "learning_rate": 4.091177112358043e-06,
+      "loss": 0.1739,
+      "step": 10273
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.4375898758116874,
+      "learning_rate": 4.090291292324682e-06,
+      "loss": 0.1871,
+      "step": 10274
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.2098132328893016,
+      "learning_rate": 4.089405501820986e-06,
+      "loss": 0.1502,
+      "step": 10275
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.9274158232045343,
+      "learning_rate": 4.088519740875712e-06,
+      "loss": 0.1411,
+      "step": 10276
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.1984300753675545,
+      "learning_rate": 4.08763400951761e-06,
+      "loss": 0.221,
+      "step": 10277
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.0308409876771816,
+      "learning_rate": 4.086748307775435e-06,
+      "loss": 0.1862,
+      "step": 10278
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.5111656655759529,
+      "learning_rate": 4.085862635677934e-06,
+      "loss": 0.2322,
+      "step": 10279
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.0305681430871059,
+      "learning_rate": 4.084976993253857e-06,
+      "loss": 0.1477,
+      "step": 10280
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.9628058574187983,
+      "learning_rate": 4.084091380531954e-06,
+      "loss": 0.1713,
+      "step": 10281
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.9065150292679275,
+      "learning_rate": 4.08320579754097e-06,
+      "loss": 0.1312,
+      "step": 10282
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.9526386544478248,
+      "learning_rate": 4.082320244309656e-06,
+      "loss": 0.1958,
+      "step": 10283
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.9413069775580644,
+      "learning_rate": 4.081434720866752e-06,
+      "loss": 0.228,
+      "step": 10284
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.7293421757246843,
+      "learning_rate": 4.0805492272410055e-06,
+      "loss": 0.2159,
+      "step": 10285
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.83468158051074,
+      "learning_rate": 4.0796637634611595e-06,
+      "loss": 0.1944,
+      "step": 10286
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.0723827416224694,
+      "learning_rate": 4.078778329555959e-06,
+      "loss": 0.1743,
+      "step": 10287
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.4583253508303928,
+      "learning_rate": 4.077892925554142e-06,
+      "loss": 0.2143,
+      "step": 10288
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.766531403573393,
+      "learning_rate": 4.0770075514844495e-06,
+      "loss": 0.2272,
+      "step": 10289
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.8682918873760301,
+      "learning_rate": 4.076122207375625e-06,
+      "loss": 0.2074,
+      "step": 10290
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.2644009086473627,
+      "learning_rate": 4.075236893256403e-06,
+      "loss": 0.2279,
+      "step": 10291
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.9819327932512918,
+      "learning_rate": 4.074351609155527e-06,
+      "loss": 0.1444,
+      "step": 10292
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.5252699150943376,
+      "learning_rate": 4.073466355101727e-06,
+      "loss": 0.2007,
+      "step": 10293
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.3199018906443367,
+      "learning_rate": 4.072581131123741e-06,
+      "loss": 0.2297,
+      "step": 10294
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.1487581274206355,
+      "learning_rate": 4.071695937250305e-06,
+      "loss": 0.1817,
+      "step": 10295
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.2543595822559983,
+      "learning_rate": 4.070810773510156e-06,
+      "loss": 0.2212,
+      "step": 10296
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.1259468584203773,
+      "learning_rate": 4.069925639932021e-06,
+      "loss": 0.1991,
+      "step": 10297
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.8952862539322309,
+      "learning_rate": 4.0690405365446346e-06,
+      "loss": 0.1473,
+      "step": 10298
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.9675508699363132,
+      "learning_rate": 4.068155463376728e-06,
+      "loss": 0.1557,
+      "step": 10299
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.3470996992864017,
+      "learning_rate": 4.067270420457031e-06,
+      "loss": 0.1829,
+      "step": 10300
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.2847779885494846,
+      "learning_rate": 4.066385407814274e-06,
+      "loss": 0.1641,
+      "step": 10301
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.552294233307052,
+      "learning_rate": 4.0655004254771816e-06,
+      "loss": 0.2284,
+      "step": 10302
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.449273244650648,
+      "learning_rate": 4.064615473474482e-06,
+      "loss": 0.2011,
+      "step": 10303
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.0717601175162133,
+      "learning_rate": 4.063730551834903e-06,
+      "loss": 0.1578,
+      "step": 10304
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.6824260902602135,
+      "learning_rate": 4.062845660587171e-06,
+      "loss": 0.1597,
+      "step": 10305
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.4200558028623342,
+      "learning_rate": 4.0619607997600046e-06,
+      "loss": 0.1523,
+      "step": 10306
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.074786587534497,
+      "learning_rate": 4.061075969382132e-06,
+      "loss": 0.1728,
+      "step": 10307
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.5216664510157842,
+      "learning_rate": 4.060191169482272e-06,
+      "loss": 0.1952,
+      "step": 10308
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.2788992110835244,
+      "learning_rate": 4.059306400089147e-06,
+      "loss": 0.2409,
+      "step": 10309
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.0011317118917336,
+      "learning_rate": 4.0584216612314785e-06,
+      "loss": 0.2228,
+      "step": 10310
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.190475951149326,
+      "learning_rate": 4.057536952937983e-06,
+      "loss": 0.2114,
+      "step": 10311
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.905448485140276,
+      "learning_rate": 4.05665227523738e-06,
+      "loss": 0.1948,
+      "step": 10312
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.4984759751841037,
+      "learning_rate": 4.055767628158387e-06,
+      "loss": 0.2054,
+      "step": 10313
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.004734395358024,
+      "learning_rate": 4.054883011729721e-06,
+      "loss": 0.1924,
+      "step": 10314
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.302452131177008,
+      "learning_rate": 4.053998425980094e-06,
+      "loss": 0.1692,
+      "step": 10315
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.25127400205649,
+      "learning_rate": 4.053113870938224e-06,
+      "loss": 0.1634,
+      "step": 10316
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.1129196717531018,
+      "learning_rate": 4.05222934663282e-06,
+      "loss": 0.1962,
+      "step": 10317
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.2811067547622537,
+      "learning_rate": 4.051344853092597e-06,
+      "loss": 0.2319,
+      "step": 10318
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.9653298135117048,
+      "learning_rate": 4.050460390346268e-06,
+      "loss": 0.1885,
+      "step": 10319
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.22735209016012,
+      "learning_rate": 4.0495759584225386e-06,
+      "loss": 0.1802,
+      "step": 10320
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.2660268039792435,
+      "learning_rate": 4.04869155735012e-06,
+      "loss": 0.1621,
+      "step": 10321
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.2159407529196737,
+      "learning_rate": 4.0478071871577204e-06,
+      "loss": 0.2046,
+      "step": 10322
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.5238117027424574,
+      "learning_rate": 4.046922847874049e-06,
+      "loss": 0.1845,
+      "step": 10323
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.395694082227101,
+      "learning_rate": 4.046038539527808e-06,
+      "loss": 0.2132,
+      "step": 10324
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.3841098690819924,
+      "learning_rate": 4.045154262147705e-06,
+      "loss": 0.2208,
+      "step": 10325
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.2724626970022805,
+      "learning_rate": 4.044270015762444e-06,
+      "loss": 0.1847,
+      "step": 10326
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.070130353590054,
+      "learning_rate": 4.0433858004007265e-06,
+      "loss": 0.175,
+      "step": 10327
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.0731501656982405,
+      "learning_rate": 4.042501616091259e-06,
+      "loss": 0.1918,
+      "step": 10328
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.4187037872148551,
+      "learning_rate": 4.041617462862737e-06,
+      "loss": 0.2142,
+      "step": 10329
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.1834942426493131,
+      "learning_rate": 4.040733340743864e-06,
+      "loss": 0.1974,
+      "step": 10330
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.315007451407095,
+      "learning_rate": 4.039849249763338e-06,
+      "loss": 0.1956,
+      "step": 10331
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.2787401021507216,
+      "learning_rate": 4.03896518994986e-06,
+      "loss": 0.2027,
+      "step": 10332
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.2872434974565625,
+      "learning_rate": 4.038081161332122e-06,
+      "loss": 0.1535,
+      "step": 10333
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.9394355145744888,
+      "learning_rate": 4.037197163938823e-06,
+      "loss": 0.1636,
+      "step": 10334
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.6177162835473768,
+      "learning_rate": 4.036313197798658e-06,
+      "loss": 0.2338,
+      "step": 10335
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.890953593120451,
+      "learning_rate": 4.03542926294032e-06,
+      "loss": 0.1974,
+      "step": 10336
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.4977166957129169,
+      "learning_rate": 4.034545359392505e-06,
+      "loss": 0.2054,
+      "step": 10337
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.3377677499076175,
+      "learning_rate": 4.033661487183901e-06,
+      "loss": 0.1912,
+      "step": 10338
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.0724508389602088,
+      "learning_rate": 4.0327776463432e-06,
+      "loss": 0.2061,
+      "step": 10339
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.574698746533742,
+      "learning_rate": 4.031893836899094e-06,
+      "loss": 0.2108,
+      "step": 10340
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.4016936822354242,
+      "learning_rate": 4.031010058880272e-06,
+      "loss": 0.2525,
+      "step": 10341
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.5435420697447204,
+      "learning_rate": 4.0301263123154185e-06,
+      "loss": 0.1967,
+      "step": 10342
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.2848904328979491,
+      "learning_rate": 4.0292425972332235e-06,
+      "loss": 0.1683,
+      "step": 10343
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.3264879270032313,
+      "learning_rate": 4.028358913662372e-06,
+      "loss": 0.1494,
+      "step": 10344
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.05166161184877,
+      "learning_rate": 4.0274752616315485e-06,
+      "loss": 0.2161,
+      "step": 10345
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.40830843792257,
+      "learning_rate": 4.026591641169439e-06,
+      "loss": 0.219,
+      "step": 10346
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.4835813383212997,
+      "learning_rate": 4.025708052304723e-06,
+      "loss": 0.1672,
+      "step": 10347
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.2971116964513238,
+      "learning_rate": 4.024824495066083e-06,
+      "loss": 0.2424,
+      "step": 10348
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.0733466266774898,
+      "learning_rate": 4.023940969482201e-06,
+      "loss": 0.1894,
+      "step": 10349
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.4826623874810243,
+      "learning_rate": 4.023057475581759e-06,
+      "loss": 0.2094,
+      "step": 10350
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.1886483807205388,
+      "learning_rate": 4.0221740133934306e-06,
+      "loss": 0.1879,
+      "step": 10351
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.0407443006107686,
+      "learning_rate": 4.021290582945898e-06,
+      "loss": 0.1701,
+      "step": 10352
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.9688657068555694,
+      "learning_rate": 4.020407184267833e-06,
+      "loss": 0.1972,
+      "step": 10353
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.3753426325408806,
+      "learning_rate": 4.019523817387915e-06,
+      "loss": 0.185,
+      "step": 10354
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.8272624742248175,
+      "learning_rate": 4.01864048233482e-06,
+      "loss": 0.1561,
+      "step": 10355
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.3409124038533031,
+      "learning_rate": 4.017757179137218e-06,
+      "loss": 0.1917,
+      "step": 10356
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.156607198956893,
+      "learning_rate": 4.016873907823782e-06,
+      "loss": 0.1609,
+      "step": 10357
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.2505089885839717,
+      "learning_rate": 4.015990668423183e-06,
+      "loss": 0.2082,
+      "step": 10358
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.0834609683971685,
+      "learning_rate": 4.0151074609640966e-06,
+      "loss": 0.1704,
+      "step": 10359
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.5335345523542068,
+      "learning_rate": 4.014224285475186e-06,
+      "loss": 0.2403,
+      "step": 10360
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.2775038901687858,
+      "learning_rate": 4.0133411419851224e-06,
+      "loss": 0.1937,
+      "step": 10361
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.9765827062423704,
+      "learning_rate": 4.012458030522572e-06,
+      "loss": 0.23,
+      "step": 10362
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.0007315898816422,
+      "learning_rate": 4.011574951116202e-06,
+      "loss": 0.1439,
+      "step": 10363
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.3093984435629842,
+      "learning_rate": 4.010691903794679e-06,
+      "loss": 0.184,
+      "step": 10364
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.3906881875244788,
+      "learning_rate": 4.009808888586664e-06,
+      "loss": 0.1706,
+      "step": 10365
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.1883841347781177,
+      "learning_rate": 4.008925905520822e-06,
+      "loss": 0.1943,
+      "step": 10366
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.997736653080482,
+      "learning_rate": 4.008042954625815e-06,
+      "loss": 0.1918,
+      "step": 10367
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.189002617269255,
+      "learning_rate": 4.007160035930305e-06,
+      "loss": 0.2056,
+      "step": 10368
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.0388195356100127,
+      "learning_rate": 4.00627714946295e-06,
+      "loss": 0.2097,
+      "step": 10369
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.639477326107933,
+      "learning_rate": 4.005394295252412e-06,
+      "loss": 0.2073,
+      "step": 10370
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.1270416942283323,
+      "learning_rate": 4.004511473327345e-06,
+      "loss": 0.2218,
+      "step": 10371
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.5013415625598572,
+      "learning_rate": 4.003628683716409e-06,
+      "loss": 0.1755,
+      "step": 10372
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.643599017174973,
+      "learning_rate": 4.002745926448261e-06,
+      "loss": 0.226,
+      "step": 10373
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.2662413003138757,
+      "learning_rate": 4.00186320155155e-06,
+      "loss": 0.2066,
+      "step": 10374
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.369097299475295,
+      "learning_rate": 4.000980509054935e-06,
+      "loss": 0.2092,
+      "step": 10375
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.0932656203315767,
+      "learning_rate": 4.000097848987066e-06,
+      "loss": 0.2096,
+      "step": 10376
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.563325609000064,
+      "learning_rate": 3.999215221376598e-06,
+      "loss": 0.2173,
+      "step": 10377
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.9455063282262116,
+      "learning_rate": 3.9983326262521785e-06,
+      "loss": 0.1781,
+      "step": 10378
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.9951920504575876,
+      "learning_rate": 3.997450063642458e-06,
+      "loss": 0.1921,
+      "step": 10379
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.2763375714462624,
+      "learning_rate": 3.996567533576084e-06,
+      "loss": 0.1954,
+      "step": 10380
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.8547574966542071,
+      "learning_rate": 3.995685036081705e-06,
+      "loss": 0.1301,
+      "step": 10381
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.4309152662528888,
+      "learning_rate": 3.99480257118797e-06,
+      "loss": 0.2196,
+      "step": 10382
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.0144683395349423,
+      "learning_rate": 3.993920138923518e-06,
+      "loss": 0.1586,
+      "step": 10383
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.3119168689147456,
+      "learning_rate": 3.993037739316998e-06,
+      "loss": 0.1737,
+      "step": 10384
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.2969145567732672,
+      "learning_rate": 3.992155372397052e-06,
+      "loss": 0.1914,
+      "step": 10385
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.1388567206841054,
+      "learning_rate": 3.991273038192324e-06,
+      "loss": 0.1866,
+      "step": 10386
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.0322990879214864,
+      "learning_rate": 3.990390736731451e-06,
+      "loss": 0.195,
+      "step": 10387
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.1529282913832861,
+      "learning_rate": 3.989508468043076e-06,
+      "loss": 0.2148,
+      "step": 10388
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.394906647961482,
+      "learning_rate": 3.988626232155837e-06,
+      "loss": 0.1927,
+      "step": 10389
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.0515959552272345,
+      "learning_rate": 3.987744029098371e-06,
+      "loss": 0.1999,
+      "step": 10390
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.2140929395033702,
+      "learning_rate": 3.986861858899319e-06,
+      "loss": 0.1839,
+      "step": 10391
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.7861828116566933,
+      "learning_rate": 3.9859797215873115e-06,
+      "loss": 0.2212,
+      "step": 10392
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.2004194601278144,
+      "learning_rate": 3.985097617190985e-06,
+      "loss": 0.2028,
+      "step": 10393
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.8349691758669605,
+      "learning_rate": 3.9842155457389734e-06,
+      "loss": 0.1611,
+      "step": 10394
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.2834498576137527,
+      "learning_rate": 3.983333507259911e-06,
+      "loss": 0.1726,
+      "step": 10395
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.1329350306572978,
+      "learning_rate": 3.982451501782427e-06,
+      "loss": 0.1818,
+      "step": 10396
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.0593352975691837,
+      "learning_rate": 3.981569529335152e-06,
+      "loss": 0.2032,
+      "step": 10397
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.370939459365634,
+      "learning_rate": 3.980687589946715e-06,
+      "loss": 0.1673,
+      "step": 10398
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.1742328368209352,
+      "learning_rate": 3.979805683645745e-06,
+      "loss": 0.2124,
+      "step": 10399
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.1589984061096066,
+      "learning_rate": 3.978923810460871e-06,
+      "loss": 0.1566,
+      "step": 10400
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.121302854329886,
+      "learning_rate": 3.9780419704207165e-06,
+      "loss": 0.1947,
+      "step": 10401
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.9901221416757447,
+      "learning_rate": 3.977160163553906e-06,
+      "loss": 0.1359,
+      "step": 10402
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.1406581821579438,
+      "learning_rate": 3.976278389889065e-06,
+      "loss": 0.1913,
+      "step": 10403
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.1132610988466307,
+      "learning_rate": 3.975396649454818e-06,
+      "loss": 0.2299,
+      "step": 10404
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.4125984091679256,
+      "learning_rate": 3.9745149422797835e-06,
+      "loss": 0.245,
+      "step": 10405
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.0170692830015164,
+      "learning_rate": 3.973633268392584e-06,
+      "loss": 0.1283,
+      "step": 10406
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.9905984348501752,
+      "learning_rate": 3.972751627821838e-06,
+      "loss": 0.1596,
+      "step": 10407
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.8385139449890077,
+      "learning_rate": 3.971870020596166e-06,
+      "loss": 0.1538,
+      "step": 10408
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.9101568639532155,
+      "learning_rate": 3.970988446744185e-06,
+      "loss": 0.147,
+      "step": 10409
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.9818503896098474,
+      "learning_rate": 3.970106906294509e-06,
+      "loss": 0.1759,
+      "step": 10410
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.9942492798520147,
+      "learning_rate": 3.969225399275755e-06,
+      "loss": 0.1762,
+      "step": 10411
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.539036164808792,
+      "learning_rate": 3.968343925716538e-06,
+      "loss": 0.2082,
+      "step": 10412
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.8565150920840444,
+      "learning_rate": 3.967462485645471e-06,
+      "loss": 0.1392,
+      "step": 10413
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.281577112146665,
+      "learning_rate": 3.966581079091165e-06,
+      "loss": 0.1832,
+      "step": 10414
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.0374491811997055,
+      "learning_rate": 3.965699706082232e-06,
+      "loss": 0.1982,
+      "step": 10415
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.5642363723160044,
+      "learning_rate": 3.964818366647279e-06,
+      "loss": 0.2451,
+      "step": 10416
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.1573814642524096,
+      "learning_rate": 3.963937060814919e-06,
+      "loss": 0.2007,
+      "step": 10417
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.1993106802500004,
+      "learning_rate": 3.963055788613759e-06,
+      "loss": 0.1973,
+      "step": 10418
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.3216916223130608,
+      "learning_rate": 3.962174550072402e-06,
+      "loss": 0.1976,
+      "step": 10419
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.6026009054298513,
+      "learning_rate": 3.9612933452194565e-06,
+      "loss": 0.2632,
+      "step": 10420
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.9258677506576979,
+      "learning_rate": 3.9604121740835265e-06,
+      "loss": 0.2173,
+      "step": 10421
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.250330529620457,
+      "learning_rate": 3.959531036693216e-06,
+      "loss": 0.1552,
+      "step": 10422
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.4925859029288204,
+      "learning_rate": 3.958649933077125e-06,
+      "loss": 0.2278,
+      "step": 10423
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.969493377454576,
+      "learning_rate": 3.957768863263857e-06,
+      "loss": 0.2086,
+      "step": 10424
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.1904831609067756,
+      "learning_rate": 3.9568878272820105e-06,
+      "loss": 0.1837,
+      "step": 10425
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.133511508479588,
+      "learning_rate": 3.9560068251601845e-06,
+      "loss": 0.1733,
+      "step": 10426
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.7007561842578722,
+      "learning_rate": 3.9551258569269795e-06,
+      "loss": 0.2369,
+      "step": 10427
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.1263874822739237,
+      "learning_rate": 3.954244922610988e-06,
+      "loss": 0.1953,
+      "step": 10428
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.5929796937844405,
+      "learning_rate": 3.953364022240807e-06,
+      "loss": 0.1672,
+      "step": 10429
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.4796813466227114,
+      "learning_rate": 3.952483155845033e-06,
+      "loss": 0.226,
+      "step": 10430
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.8125392372487303,
+      "learning_rate": 3.9516023234522585e-06,
+      "loss": 0.1764,
+      "step": 10431
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.9500727353908889,
+      "learning_rate": 3.950721525091075e-06,
+      "loss": 0.1425,
+      "step": 10432
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.7998584780461394,
+      "learning_rate": 3.949840760790074e-06,
+      "loss": 0.1594,
+      "step": 10433
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.1208083573720402,
+      "learning_rate": 3.948960030577845e-06,
+      "loss": 0.2239,
+      "step": 10434
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.2789244597483762,
+      "learning_rate": 3.948079334482978e-06,
+      "loss": 0.2158,
+      "step": 10435
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.0745510835690666,
+      "learning_rate": 3.947198672534063e-06,
+      "loss": 0.1881,
+      "step": 10436
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.333811428984658,
+      "learning_rate": 3.946318044759683e-06,
+      "loss": 0.2142,
+      "step": 10437
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.9722943092151847,
+      "learning_rate": 3.945437451188424e-06,
+      "loss": 0.1846,
+      "step": 10438
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.1082153000967898,
+      "learning_rate": 3.944556891848872e-06,
+      "loss": 0.1546,
+      "step": 10439
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.7114445357444674,
+      "learning_rate": 3.9436763667696125e-06,
+      "loss": 0.2161,
+      "step": 10440
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.172577361389314,
+      "learning_rate": 3.942795875979224e-06,
+      "loss": 0.178,
+      "step": 10441
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.5831209268227644,
+      "learning_rate": 3.941915419506289e-06,
+      "loss": 0.1911,
+      "step": 10442
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.3824401019970547,
+      "learning_rate": 3.9410349973793884e-06,
+      "loss": 0.2009,
+      "step": 10443
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.5403876118834376,
+      "learning_rate": 3.940154609627099e-06,
+      "loss": 0.2374,
+      "step": 10444
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.5524222486103034,
+      "learning_rate": 3.939274256278004e-06,
+      "loss": 0.1986,
+      "step": 10445
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.2792084583275245,
+      "learning_rate": 3.938393937360674e-06,
+      "loss": 0.2013,
+      "step": 10446
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.4237879052780578,
+      "learning_rate": 3.9375136529036876e-06,
+      "loss": 0.1365,
+      "step": 10447
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.5619142531155978,
+      "learning_rate": 3.936633402935619e-06,
+      "loss": 0.1933,
+      "step": 10448
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.066839842627678,
+      "learning_rate": 3.935753187485043e-06,
+      "loss": 0.1858,
+      "step": 10449
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.1554188479621776,
+      "learning_rate": 3.934873006580528e-06,
+      "loss": 0.1707,
+      "step": 10450
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.3162563957473767,
+      "learning_rate": 3.9339928602506505e-06,
+      "loss": 0.2222,
+      "step": 10451
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.422407255006979,
+      "learning_rate": 3.933112748523976e-06,
+      "loss": 0.147,
+      "step": 10452
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.260963724825589,
+      "learning_rate": 3.9322326714290745e-06,
+      "loss": 0.1812,
+      "step": 10453
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.1047158794723502,
+      "learning_rate": 3.931352628994517e-06,
+      "loss": 0.1759,
+      "step": 10454
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.0493642157965528,
+      "learning_rate": 3.9304726212488655e-06,
+      "loss": 0.162,
+      "step": 10455
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.1922611452556797,
+      "learning_rate": 3.929592648220688e-06,
+      "loss": 0.1662,
+      "step": 10456
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.050328671297693,
+      "learning_rate": 3.9287127099385485e-06,
+      "loss": 0.1533,
+      "step": 10457
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.42558434054213,
+      "learning_rate": 3.927832806431013e-06,
+      "loss": 0.2357,
+      "step": 10458
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.2083163688416991,
+      "learning_rate": 3.926952937726638e-06,
+      "loss": 0.2114,
+      "step": 10459
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.2111593720000455,
+      "learning_rate": 3.9260731038539896e-06,
+      "loss": 0.1962,
+      "step": 10460
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.9701023506948478,
+      "learning_rate": 3.925193304841625e-06,
+      "loss": 0.2081,
+      "step": 10461
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.273913490525622,
+      "learning_rate": 3.924313540718103e-06,
+      "loss": 0.1839,
+      "step": 10462
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.9284777784839099,
+      "learning_rate": 3.923433811511984e-06,
+      "loss": 0.2097,
+      "step": 10463
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.3779007863366892,
+      "learning_rate": 3.922554117251821e-06,
+      "loss": 0.292,
+      "step": 10464
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.0958768553951703,
+      "learning_rate": 3.921674457966172e-06,
+      "loss": 0.2338,
+      "step": 10465
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.3523650405151069,
+      "learning_rate": 3.920794833683589e-06,
+      "loss": 0.2741,
+      "step": 10466
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.7879097318888924,
+      "learning_rate": 3.919915244432627e-06,
+      "loss": 0.1965,
+      "step": 10467
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.0058008788052648,
+      "learning_rate": 3.919035690241837e-06,
+      "loss": 0.1541,
+      "step": 10468
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.5332094131632348,
+      "learning_rate": 3.918156171139771e-06,
+      "loss": 0.1908,
+      "step": 10469
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.1287187697809988,
+      "learning_rate": 3.917276687154976e-06,
+      "loss": 0.1985,
+      "step": 10470
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.0256544573009738,
+      "learning_rate": 3.916397238316003e-06,
+      "loss": 0.2038,
+      "step": 10471
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.8228821244213163,
+      "learning_rate": 3.915517824651401e-06,
+      "loss": 0.1504,
+      "step": 10472
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.5186361652852105,
+      "learning_rate": 3.914638446189711e-06,
+      "loss": 0.1557,
+      "step": 10473
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.0100280634656686,
+      "learning_rate": 3.9137591029594814e-06,
+      "loss": 0.1727,
+      "step": 10474
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.1521614706411634,
+      "learning_rate": 3.912879794989257e-06,
+      "loss": 0.1726,
+      "step": 10475
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.1347591084803663,
+      "learning_rate": 3.9120005223075805e-06,
+      "loss": 0.1887,
+      "step": 10476
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.0390578033226392,
+      "learning_rate": 3.911121284942991e-06,
+      "loss": 0.222,
+      "step": 10477
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.0824462390912293,
+      "learning_rate": 3.910242082924032e-06,
+      "loss": 0.168,
+      "step": 10478
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.6408208276043805,
+      "learning_rate": 3.9093629162792404e-06,
+      "loss": 0.2264,
+      "step": 10479
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.9925937234012445,
+      "learning_rate": 3.908483785037156e-06,
+      "loss": 0.2154,
+      "step": 10480
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.456533455633798,
+      "learning_rate": 3.907604689226318e-06,
+      "loss": 0.2396,
+      "step": 10481
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.4092632911292893,
+      "learning_rate": 3.906725628875257e-06,
+      "loss": 0.2098,
+      "step": 10482
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.048914635853607,
+      "learning_rate": 3.905846604012511e-06,
+      "loss": 0.2063,
+      "step": 10483
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.148592523245428,
+      "learning_rate": 3.904967614666614e-06,
+      "loss": 0.1701,
+      "step": 10484
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.2135725177750822,
+      "learning_rate": 3.904088660866099e-06,
+      "loss": 0.2139,
+      "step": 10485
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.127935497054294,
+      "learning_rate": 3.903209742639495e-06,
+      "loss": 0.1976,
+      "step": 10486
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.5671815261650899,
+      "learning_rate": 3.902330860015334e-06,
+      "loss": 0.2023,
+      "step": 10487
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.1886848386320248,
+      "learning_rate": 3.901452013022143e-06,
+      "loss": 0.1517,
+      "step": 10488
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.3394478582822105,
+      "learning_rate": 3.900573201688452e-06,
+      "loss": 0.2284,
+      "step": 10489
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.1600230313889068,
+      "learning_rate": 3.89969442604279e-06,
+      "loss": 0.1775,
+      "step": 10490
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.167599533563745,
+      "learning_rate": 3.898815686113676e-06,
+      "loss": 0.2047,
+      "step": 10491
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.088343466125674,
+      "learning_rate": 3.89793698192964e-06,
+      "loss": 0.1768,
+      "step": 10492
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.4648901105538068,
+      "learning_rate": 3.897058313519202e-06,
+      "loss": 0.1866,
+      "step": 10493
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.0874870551643334,
+      "learning_rate": 3.896179680910886e-06,
+      "loss": 0.1728,
+      "step": 10494
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.2928915922763042,
+      "learning_rate": 3.895301084133213e-06,
+      "loss": 0.2106,
+      "step": 10495
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.9487399475283813,
+      "learning_rate": 3.894422523214702e-06,
+      "loss": 0.1472,
+      "step": 10496
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.2627000447219245,
+      "learning_rate": 3.89354399818387e-06,
+      "loss": 0.1875,
+      "step": 10497
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.1551331268002378,
+      "learning_rate": 3.892665509069237e-06,
+      "loss": 0.1548,
+      "step": 10498
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.1582087923525608,
+      "learning_rate": 3.89178705589932e-06,
+      "loss": 0.1408,
+      "step": 10499
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.2102576962549179,
+      "learning_rate": 3.89090863870263e-06,
+      "loss": 0.2004,
+      "step": 10500
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.619512832709858,
+      "learning_rate": 3.890030257507684e-06,
+      "loss": 0.1803,
+      "step": 10501
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.1203216509132727,
+      "learning_rate": 3.889151912342993e-06,
+      "loss": 0.1892,
+      "step": 10502
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.1458006579624112,
+      "learning_rate": 3.8882736032370716e-06,
+      "loss": 0.1841,
+      "step": 10503
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 3.5821005794758856,
+      "learning_rate": 3.887395330218429e-06,
+      "loss": 0.2275,
+      "step": 10504
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.1602488586762925,
+      "learning_rate": 3.886517093315572e-06,
+      "loss": 0.1669,
+      "step": 10505
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.2268145326578788,
+      "learning_rate": 3.885638892557011e-06,
+      "loss": 0.173,
+      "step": 10506
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.1272456088738818,
+      "learning_rate": 3.884760727971252e-06,
+      "loss": 0.1514,
+      "step": 10507
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.2966429002878288,
+      "learning_rate": 3.883882599586803e-06,
+      "loss": 0.1898,
+      "step": 10508
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.8160845656647778,
+      "learning_rate": 3.883004507432164e-06,
+      "loss": 0.13,
+      "step": 10509
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.2108968850984831,
+      "learning_rate": 3.882126451535842e-06,
+      "loss": 0.1895,
+      "step": 10510
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.8852853681008475,
+      "learning_rate": 3.8812484319263375e-06,
+      "loss": 0.2138,
+      "step": 10511
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.382977378843293,
+      "learning_rate": 3.880370448632154e-06,
+      "loss": 0.1737,
+      "step": 10512
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.079399252518572,
+      "learning_rate": 3.879492501681788e-06,
+      "loss": 0.1729,
+      "step": 10513
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.3041646646473095,
+      "learning_rate": 3.878614591103742e-06,
+      "loss": 0.1484,
+      "step": 10514
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.815043654087621,
+      "learning_rate": 3.877736716926509e-06,
+      "loss": 0.2641,
+      "step": 10515
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.0861427367041157,
+      "learning_rate": 3.876858879178587e-06,
+      "loss": 0.1351,
+      "step": 10516
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.339658735530174,
+      "learning_rate": 3.8759810778884746e-06,
+      "loss": 0.2372,
+      "step": 10517
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.3454069633858565,
+      "learning_rate": 3.87510331308466e-06,
+      "loss": 0.1466,
+      "step": 10518
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 4.8118286159909305,
+      "learning_rate": 3.874225584795639e-06,
+      "loss": 0.1985,
+      "step": 10519
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.4930644749476256,
+      "learning_rate": 3.873347893049903e-06,
+      "loss": 0.2151,
+      "step": 10520
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.2434100067347298,
+      "learning_rate": 3.872470237875943e-06,
+      "loss": 0.1647,
+      "step": 10521
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.240193683227057,
+      "learning_rate": 3.8715926193022475e-06,
+      "loss": 0.2195,
+      "step": 10522
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.1958921716112547,
+      "learning_rate": 3.870715037357305e-06,
+      "loss": 0.1896,
+      "step": 10523
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.2779025275929679,
+      "learning_rate": 3.8698374920696005e-06,
+      "loss": 0.1728,
+      "step": 10524
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.9742219036207266,
+      "learning_rate": 3.8689599834676206e-06,
+      "loss": 0.1452,
+      "step": 10525
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.6371501294949866,
+      "learning_rate": 3.8680825115798526e-06,
+      "loss": 0.247,
+      "step": 10526
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.1252784914123117,
+      "learning_rate": 3.867205076434775e-06,
+      "loss": 0.1745,
+      "step": 10527
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.1936673174429726,
+      "learning_rate": 3.866327678060871e-06,
+      "loss": 0.2065,
+      "step": 10528
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.953874232952228,
+      "learning_rate": 3.865450316486623e-06,
+      "loss": 0.2087,
+      "step": 10529
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.2966219816081734,
+      "learning_rate": 3.864572991740512e-06,
+      "loss": 0.2072,
+      "step": 10530
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.015317501688056,
+      "learning_rate": 3.863695703851013e-06,
+      "loss": 0.1512,
+      "step": 10531
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.1012076996885622,
+      "learning_rate": 3.862818452846607e-06,
+      "loss": 0.2043,
+      "step": 10532
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.1512513025542563,
+      "learning_rate": 3.861941238755766e-06,
+      "loss": 0.2095,
+      "step": 10533
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.5198587564790529,
+      "learning_rate": 3.861064061606966e-06,
+      "loss": 0.1901,
+      "step": 10534
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.4366161490267042,
+      "learning_rate": 3.8601869214286844e-06,
+      "loss": 0.2317,
+      "step": 10535
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.1385800175724041,
+      "learning_rate": 3.859309818249389e-06,
+      "loss": 0.1976,
+      "step": 10536
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.4793767071969128,
+      "learning_rate": 3.8584327520975515e-06,
+      "loss": 0.2034,
+      "step": 10537
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.3053242505845817,
+      "learning_rate": 3.857555723001645e-06,
+      "loss": 0.237,
+      "step": 10538
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.3383686033267765,
+      "learning_rate": 3.856678730990136e-06,
+      "loss": 0.2567,
+      "step": 10539
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.1811347698340913,
+      "learning_rate": 3.855801776091495e-06,
+      "loss": 0.1595,
+      "step": 10540
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.522388238743273,
+      "learning_rate": 3.8549248583341845e-06,
+      "loss": 0.1644,
+      "step": 10541
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.197748765028546,
+      "learning_rate": 3.854047977746671e-06,
+      "loss": 0.1479,
+      "step": 10542
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.1344509180795022,
+      "learning_rate": 3.8531711343574194e-06,
+      "loss": 0.1641,
+      "step": 10543
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.2517343231725717,
+      "learning_rate": 3.852294328194894e-06,
+      "loss": 0.181,
+      "step": 10544
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.1229195829482184,
+      "learning_rate": 3.851417559287552e-06,
+      "loss": 0.1976,
+      "step": 10545
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.3026163885627928,
+      "learning_rate": 3.850540827663858e-06,
+      "loss": 0.2194,
+      "step": 10546
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.2537868270571233,
+      "learning_rate": 3.849664133352269e-06,
+      "loss": 0.2214,
+      "step": 10547
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.060014698344453,
+      "learning_rate": 3.848787476381245e-06,
+      "loss": 0.189,
+      "step": 10548
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.213840989177699,
+      "learning_rate": 3.84791085677924e-06,
+      "loss": 0.1652,
+      "step": 10549
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.2063530433205314,
+      "learning_rate": 3.847034274574713e-06,
+      "loss": 0.1957,
+      "step": 10550
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.0722690071511214,
+      "learning_rate": 3.846157729796115e-06,
+      "loss": 0.1731,
+      "step": 10551
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.143011542384206,
+      "learning_rate": 3.8452812224719e-06,
+      "loss": 0.2625,
+      "step": 10552
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.2027554997382042,
+      "learning_rate": 3.844404752630523e-06,
+      "loss": 0.2128,
+      "step": 10553
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.6291795588208453,
+      "learning_rate": 3.843528320300431e-06,
+      "loss": 0.1422,
+      "step": 10554
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.1510860028581622,
+      "learning_rate": 3.842651925510073e-06,
+      "loss": 0.1605,
+      "step": 10555
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.9940932286923599,
+      "learning_rate": 3.841775568287901e-06,
+      "loss": 0.1699,
+      "step": 10556
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.3460482969929142,
+      "learning_rate": 3.840899248662358e-06,
+      "loss": 0.2213,
+      "step": 10557
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.2865647348876066,
+      "learning_rate": 3.840022966661895e-06,
+      "loss": 0.1731,
+      "step": 10558
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.2268425248245196,
+      "learning_rate": 3.839146722314952e-06,
+      "loss": 0.1359,
+      "step": 10559
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.5157307263494915,
+      "learning_rate": 3.838270515649973e-06,
+      "loss": 0.1864,
+      "step": 10560
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.2014484180219673,
+      "learning_rate": 3.8373943466954025e-06,
+      "loss": 0.2,
+      "step": 10561
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.0100592439043243,
+      "learning_rate": 3.83651821547968e-06,
+      "loss": 0.1567,
+      "step": 10562
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.515441116504269,
+      "learning_rate": 3.835642122031245e-06,
+      "loss": 0.2133,
+      "step": 10563
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.084720979897562,
+      "learning_rate": 3.834766066378535e-06,
+      "loss": 0.1304,
+      "step": 10564
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.0512883405690387,
+      "learning_rate": 3.8338900485499905e-06,
+      "loss": 0.1803,
+      "step": 10565
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.4390512627526333,
+      "learning_rate": 3.833014068574044e-06,
+      "loss": 0.2047,
+      "step": 10566
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.4195928126723014,
+      "learning_rate": 3.832138126479133e-06,
+      "loss": 0.2374,
+      "step": 10567
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.3741993849074656,
+      "learning_rate": 3.8312622222936905e-06,
+      "loss": 0.1712,
+      "step": 10568
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.219008076070706,
+      "learning_rate": 3.830386356046146e-06,
+      "loss": 0.1837,
+      "step": 10569
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.4748099386480158,
+      "learning_rate": 3.829510527764934e-06,
+      "loss": 0.1959,
+      "step": 10570
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.0495886104950345,
+      "learning_rate": 3.828634737478486e-06,
+      "loss": 0.1931,
+      "step": 10571
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.325853632635276,
+      "learning_rate": 3.827758985215225e-06,
+      "loss": 0.2578,
+      "step": 10572
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.0433270485460693,
+      "learning_rate": 3.826883271003581e-06,
+      "loss": 0.1782,
+      "step": 10573
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.0093110406027546,
+      "learning_rate": 3.826007594871983e-06,
+      "loss": 0.1484,
+      "step": 10574
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.1408133334962098,
+      "learning_rate": 3.825131956848852e-06,
+      "loss": 0.1857,
+      "step": 10575
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.1055137453256667,
+      "learning_rate": 3.8242563569626144e-06,
+      "loss": 0.2289,
+      "step": 10576
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.3094571780862436,
+      "learning_rate": 3.823380795241691e-06,
+      "loss": 0.1949,
+      "step": 10577
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.1183052023885802,
+      "learning_rate": 3.822505271714502e-06,
+      "loss": 0.1729,
+      "step": 10578
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.2698338074715594,
+      "learning_rate": 3.8216297864094695e-06,
+      "loss": 0.1972,
+      "step": 10579
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.8719821577807311,
+      "learning_rate": 3.820754339355015e-06,
+      "loss": 0.1472,
+      "step": 10580
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.0798829766375313,
+      "learning_rate": 3.819878930579549e-06,
+      "loss": 0.1524,
+      "step": 10581
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.312744928111088,
+      "learning_rate": 3.819003560111493e-06,
+      "loss": 0.2385,
+      "step": 10582
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.355077665124916,
+      "learning_rate": 3.818128227979261e-06,
+      "loss": 0.173,
+      "step": 10583
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.5088115999087883,
+      "learning_rate": 3.817252934211265e-06,
+      "loss": 0.2039,
+      "step": 10584
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.1301502283896556,
+      "learning_rate": 3.816377678835922e-06,
+      "loss": 0.2242,
+      "step": 10585
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.9775343408003727,
+      "learning_rate": 3.8155024618816364e-06,
+      "loss": 0.1232,
+      "step": 10586
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.9473603270079319,
+      "learning_rate": 3.814627283376823e-06,
+      "loss": 0.1832,
+      "step": 10587
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.425833936266527,
+      "learning_rate": 3.8137521433498898e-06,
+      "loss": 0.1658,
+      "step": 10588
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.25336055378917,
+      "learning_rate": 3.812877041829246e-06,
+      "loss": 0.1821,
+      "step": 10589
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.9868183485160185,
+      "learning_rate": 3.812001978843294e-06,
+      "loss": 0.1527,
+      "step": 10590
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.4299163101101342,
+      "learning_rate": 3.811126954420441e-06,
+      "loss": 0.2314,
+      "step": 10591
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.2472797358354135,
+      "learning_rate": 3.8102519685890915e-06,
+      "loss": 0.1781,
+      "step": 10592
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.1392793981324296,
+      "learning_rate": 3.809377021377646e-06,
+      "loss": 0.1621,
+      "step": 10593
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.617018114884963,
+      "learning_rate": 3.8085021128145095e-06,
+      "loss": 0.196,
+      "step": 10594
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.9270643846668464,
+      "learning_rate": 3.807627242928078e-06,
+      "loss": 0.1674,
+      "step": 10595
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.365498005703867,
+      "learning_rate": 3.8067524117467514e-06,
+      "loss": 0.1801,
+      "step": 10596
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.541122007384571,
+      "learning_rate": 3.8058776192989273e-06,
+      "loss": 0.2158,
+      "step": 10597
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.3752814189152596,
+      "learning_rate": 3.8050028656130046e-06,
+      "loss": 0.1976,
+      "step": 10598
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.2401795142347816,
+      "learning_rate": 3.804128150717374e-06,
+      "loss": 0.1858,
+      "step": 10599
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.9844248736326642,
+      "learning_rate": 3.8032534746404315e-06,
+      "loss": 0.198,
+      "step": 10600
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.2842111010752406,
+      "learning_rate": 3.802378837410571e-06,
+      "loss": 0.1899,
+      "step": 10601
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.8655549439337245,
+      "learning_rate": 3.8015042390561796e-06,
+      "loss": 0.1961,
+      "step": 10602
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.2399898077657425,
+      "learning_rate": 3.800629679605653e-06,
+      "loss": 0.1843,
+      "step": 10603
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.1903514007036473,
+      "learning_rate": 3.7997551590873745e-06,
+      "loss": 0.1507,
+      "step": 10604
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.0396486579397435,
+      "learning_rate": 3.7988806775297334e-06,
+      "loss": 0.2068,
+      "step": 10605
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.1246893771813857,
+      "learning_rate": 3.7980062349611163e-06,
+      "loss": 0.1964,
+      "step": 10606
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.0913221866179297,
+      "learning_rate": 3.7971318314099106e-06,
+      "loss": 0.2252,
+      "step": 10607
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.1861440169660895,
+      "learning_rate": 3.7962574669044944e-06,
+      "loss": 0.1824,
+      "step": 10608
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.8740307209932883,
+      "learning_rate": 3.7953831414732532e-06,
+      "loss": 0.1946,
+      "step": 10609
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.1955755573853706,
+      "learning_rate": 3.7945088551445698e-06,
+      "loss": 0.1804,
+      "step": 10610
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.290745595835092,
+      "learning_rate": 3.79363460794682e-06,
+      "loss": 0.2113,
+      "step": 10611
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.173090476065114,
+      "learning_rate": 3.7927603999083873e-06,
+      "loss": 0.1799,
+      "step": 10612
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.0675906496398517,
+      "learning_rate": 3.7918862310576438e-06,
+      "loss": 0.1501,
+      "step": 10613
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.3413727685596781,
+      "learning_rate": 3.7910121014229684e-06,
+      "loss": 0.1699,
+      "step": 10614
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.8472217333879953,
+      "learning_rate": 3.7901380110327347e-06,
+      "loss": 0.1322,
+      "step": 10615
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.4503623069918563,
+      "learning_rate": 3.7892639599153196e-06,
+      "loss": 0.1959,
+      "step": 10616
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.237875068545135,
+      "learning_rate": 3.7883899480990905e-06,
+      "loss": 0.1634,
+      "step": 10617
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.0968022027542834,
+      "learning_rate": 3.7875159756124203e-06,
+      "loss": 0.1968,
+      "step": 10618
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.33684413105627,
+      "learning_rate": 3.7866420424836797e-06,
+      "loss": 0.2095,
+      "step": 10619
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.6717963200175203,
+      "learning_rate": 3.785768148741235e-06,
+      "loss": 0.2059,
+      "step": 10620
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.512139520563913,
+      "learning_rate": 3.7848942944134565e-06,
+      "loss": 0.1325,
+      "step": 10621
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.6072737809811786,
+      "learning_rate": 3.784020479528706e-06,
+      "loss": 0.1752,
+      "step": 10622
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.2861887978292759,
+      "learning_rate": 3.78314670411535e-06,
+      "loss": 0.2565,
+      "step": 10623
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.2808192796821365,
+      "learning_rate": 3.7822729682017518e-06,
+      "loss": 0.1826,
+      "step": 10624
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.5362320193141692,
+      "learning_rate": 3.781399271816275e-06,
+      "loss": 0.1693,
+      "step": 10625
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.4086322051356628,
+      "learning_rate": 3.780525614987277e-06,
+      "loss": 0.2116,
+      "step": 10626
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.2010149364758549,
+      "learning_rate": 3.7796519977431183e-06,
+      "loss": 0.1384,
+      "step": 10627
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.913900344680853,
+      "learning_rate": 3.7787784201121592e-06,
+      "loss": 0.1927,
+      "step": 10628
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.2689612267981603,
+      "learning_rate": 3.7779048821227533e-06,
+      "loss": 0.1737,
+      "step": 10629
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.2162009811422658,
+      "learning_rate": 3.7770313838032596e-06,
+      "loss": 0.2116,
+      "step": 10630
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.1298918200292898,
+      "learning_rate": 3.7761579251820283e-06,
+      "loss": 0.1822,
+      "step": 10631
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.5816447033008727,
+      "learning_rate": 3.775284506287415e-06,
+      "loss": 0.2204,
+      "step": 10632
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.1575332074365965,
+      "learning_rate": 3.774411127147771e-06,
+      "loss": 0.1672,
+      "step": 10633
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.2074460486587226,
+      "learning_rate": 3.773537787791448e-06,
+      "loss": 0.1807,
+      "step": 10634
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.1773819103397611,
+      "learning_rate": 3.772664488246792e-06,
+      "loss": 0.1682,
+      "step": 10635
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.7994678845474145,
+      "learning_rate": 3.7717912285421527e-06,
+      "loss": 0.1721,
+      "step": 10636
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.9408344257445254,
+      "learning_rate": 3.7709180087058773e-06,
+      "loss": 0.1746,
+      "step": 10637
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.1868381208537762,
+      "learning_rate": 3.7700448287663085e-06,
+      "loss": 0.1556,
+      "step": 10638
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.107906481088231,
+      "learning_rate": 3.7691716887517947e-06,
+      "loss": 0.1706,
+      "step": 10639
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.451693306399091,
+      "learning_rate": 3.768298588690673e-06,
+      "loss": 0.1844,
+      "step": 10640
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.0979783800489142,
+      "learning_rate": 3.7674255286112876e-06,
+      "loss": 0.1539,
+      "step": 10641
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.2755375612448896,
+      "learning_rate": 3.7665525085419784e-06,
+      "loss": 0.268,
+      "step": 10642
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.2756636825277339,
+      "learning_rate": 3.7656795285110858e-06,
+      "loss": 0.2154,
+      "step": 10643
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.0305844854357569,
+      "learning_rate": 3.7648065885469437e-06,
+      "loss": 0.1417,
+      "step": 10644
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.1158578436021793,
+      "learning_rate": 3.76393368867789e-06,
+      "loss": 0.1734,
+      "step": 10645
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.0644597867004801,
+      "learning_rate": 3.7630608289322608e-06,
+      "loss": 0.1992,
+      "step": 10646
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.179179151534091,
+      "learning_rate": 3.7621880093383873e-06,
+      "loss": 0.217,
+      "step": 10647
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.0678307726677865,
+      "learning_rate": 3.7613152299246046e-06,
+      "loss": 0.1616,
+      "step": 10648
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.8837080444189684,
+      "learning_rate": 3.76044249071924e-06,
+      "loss": 0.2421,
+      "step": 10649
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.9390343408758108,
+      "learning_rate": 3.7595697917506247e-06,
+      "loss": 0.1933,
+      "step": 10650
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.3693751502961913,
+      "learning_rate": 3.758697133047087e-06,
+      "loss": 0.2113,
+      "step": 10651
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.2141644611771703,
+      "learning_rate": 3.757824514636957e-06,
+      "loss": 0.167,
+      "step": 10652
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.2122060837369728,
+      "learning_rate": 3.756951936548555e-06,
+      "loss": 0.155,
+      "step": 10653
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.0912162656365503,
+      "learning_rate": 3.7560793988102085e-06,
+      "loss": 0.1735,
+      "step": 10654
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.9825959932956615,
+      "learning_rate": 3.75520690145024e-06,
+      "loss": 0.1882,
+      "step": 10655
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.2539823099045317,
+      "learning_rate": 3.754334444496971e-06,
+      "loss": 0.1587,
+      "step": 10656
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.1559054467036192,
+      "learning_rate": 3.7534620279787243e-06,
+      "loss": 0.1443,
+      "step": 10657
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.4224444391398225,
+      "learning_rate": 3.752589651923815e-06,
+      "loss": 0.1806,
+      "step": 10658
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.5274635716728657,
+      "learning_rate": 3.7517173163605626e-06,
+      "loss": 0.2511,
+      "step": 10659
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.1341380683695101,
+      "learning_rate": 3.7508450213172836e-06,
+      "loss": 0.1639,
+      "step": 10660
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.4592397024674584,
+      "learning_rate": 3.7499727668222964e-06,
+      "loss": 0.2574,
+      "step": 10661
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.29572744614066,
+      "learning_rate": 3.7491005529039092e-06,
+      "loss": 0.2305,
+      "step": 10662
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.2178161173104314,
+      "learning_rate": 3.748228379590438e-06,
+      "loss": 0.2252,
+      "step": 10663
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.444662605263414,
+      "learning_rate": 3.7473562469101943e-06,
+      "loss": 0.2228,
+      "step": 10664
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.255631518725803,
+      "learning_rate": 3.7464841548914856e-06,
+      "loss": 0.1996,
+      "step": 10665
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.3237816615321172,
+      "learning_rate": 3.745612103562625e-06,
+      "loss": 0.2044,
+      "step": 10666
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.191195619443673,
+      "learning_rate": 3.744740092951914e-06,
+      "loss": 0.1806,
+      "step": 10667
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.1249821052187694,
+      "learning_rate": 3.7438681230876615e-06,
+      "loss": 0.213,
+      "step": 10668
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.0820703155130889,
+      "learning_rate": 3.7429961939981725e-06,
+      "loss": 0.2164,
+      "step": 10669
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.0329193021057954,
+      "learning_rate": 3.7421243057117517e-06,
+      "loss": 0.2053,
+      "step": 10670
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.0692036655782944,
+      "learning_rate": 3.7412524582566965e-06,
+      "loss": 0.1531,
+      "step": 10671
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.7290010578829065,
+      "learning_rate": 3.7403806516613117e-06,
+      "loss": 0.2566,
+      "step": 10672
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.2250235828980938,
+      "learning_rate": 3.7395088859538953e-06,
+      "loss": 0.1958,
+      "step": 10673
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.482414652689955,
+      "learning_rate": 3.7386371611627448e-06,
+      "loss": 0.2002,
+      "step": 10674
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.174104772202224,
+      "learning_rate": 3.7377654773161593e-06,
+      "loss": 0.196,
+      "step": 10675
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.119192505011794,
+      "learning_rate": 3.73689383444243e-06,
+      "loss": 0.2374,
+      "step": 10676
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.7943723084241239,
+      "learning_rate": 3.7360222325698538e-06,
+      "loss": 0.1723,
+      "step": 10677
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.4991129596830721,
+      "learning_rate": 3.735150671726722e-06,
+      "loss": 0.1765,
+      "step": 10678
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.181764870409032,
+      "learning_rate": 3.734279151941329e-06,
+      "loss": 0.1864,
+      "step": 10679
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.6443953288034927,
+      "learning_rate": 3.7334076732419604e-06,
+      "loss": 0.2625,
+      "step": 10680
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.102025502886404,
+      "learning_rate": 3.732536235656907e-06,
+      "loss": 0.2286,
+      "step": 10681
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.2861800130815932,
+      "learning_rate": 3.731664839214458e-06,
+      "loss": 0.2139,
+      "step": 10682
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.1878759893176198,
+      "learning_rate": 3.730793483942896e-06,
+      "loss": 0.1931,
+      "step": 10683
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.1228286000304366,
+      "learning_rate": 3.7299221698705096e-06,
+      "loss": 0.1725,
+      "step": 10684
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.3949706376439686,
+      "learning_rate": 3.729050897025578e-06,
+      "loss": 0.1658,
+      "step": 10685
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.260418540845481,
+      "learning_rate": 3.7281796654363846e-06,
+      "loss": 0.1585,
+      "step": 10686
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.1440764815875804,
+      "learning_rate": 3.7273084751312107e-06,
+      "loss": 0.157,
+      "step": 10687
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.4460606108031262,
+      "learning_rate": 3.726437326138338e-06,
+      "loss": 0.1981,
+      "step": 10688
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.0282482962464377,
+      "learning_rate": 3.72556621848604e-06,
+      "loss": 0.181,
+      "step": 10689
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.136202911449327,
+      "learning_rate": 3.7246951522025953e-06,
+      "loss": 0.1421,
+      "step": 10690
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.0719966342576501,
+      "learning_rate": 3.72382412731628e-06,
+      "loss": 0.1907,
+      "step": 10691
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.2911047854200626,
+      "learning_rate": 3.7229531438553666e-06,
+      "loss": 0.1989,
+      "step": 10692
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.0720345225504533,
+      "learning_rate": 3.7220822018481307e-06,
+      "loss": 0.1931,
+      "step": 10693
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.1312068922433605,
+      "learning_rate": 3.721211301322839e-06,
+      "loss": 0.204,
+      "step": 10694
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.4219940209092015,
+      "learning_rate": 3.7203404423077645e-06,
+      "loss": 0.2155,
+      "step": 10695
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.2766876139548082,
+      "learning_rate": 3.7194696248311744e-06,
+      "loss": 0.2152,
+      "step": 10696
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.2408372960428726,
+      "learning_rate": 3.718598848921339e-06,
+      "loss": 0.1561,
+      "step": 10697
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.7674076293154017,
+      "learning_rate": 3.7177281146065198e-06,
+      "loss": 0.2182,
+      "step": 10698
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.4035223202926574,
+      "learning_rate": 3.716857421914983e-06,
+      "loss": 0.2215,
+      "step": 10699
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.227426321941714,
+      "learning_rate": 3.7159867708749937e-06,
+      "loss": 0.1617,
+      "step": 10700
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.9958922032671543,
+      "learning_rate": 3.715116161514811e-06,
+      "loss": 0.2063,
+      "step": 10701
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.4895236569123733,
+      "learning_rate": 3.714245593862699e-06,
+      "loss": 0.1647,
+      "step": 10702
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.0428534276119492,
+      "learning_rate": 3.713375067946911e-06,
+      "loss": 0.1448,
+      "step": 10703
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.0694658801413117,
+      "learning_rate": 3.712504583795709e-06,
+      "loss": 0.1602,
+      "step": 10704
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.0845692955217037,
+      "learning_rate": 3.7116341414373488e-06,
+      "loss": 0.1566,
+      "step": 10705
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.1062710347842823,
+      "learning_rate": 3.7107637409000867e-06,
+      "loss": 0.1601,
+      "step": 10706
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.2499118177780775,
+      "learning_rate": 3.709893382212173e-06,
+      "loss": 0.1881,
+      "step": 10707
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.4150639144157813,
+      "learning_rate": 3.709023065401862e-06,
+      "loss": 0.2339,
+      "step": 10708
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.3860494099692224,
+      "learning_rate": 3.708152790497406e-06,
+      "loss": 0.2049,
+      "step": 10709
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.0680445289097737,
+      "learning_rate": 3.707282557527052e-06,
+      "loss": 0.1394,
+      "step": 10710
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.6274215349093841,
+      "learning_rate": 3.706412366519052e-06,
+      "loss": 0.2457,
+      "step": 10711
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.2414329894434801,
+      "learning_rate": 3.7055422175016486e-06,
+      "loss": 0.1947,
+      "step": 10712
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.0592321533061881,
+      "learning_rate": 3.7046721105030886e-06,
+      "loss": 0.2047,
+      "step": 10713
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.259627821134872,
+      "learning_rate": 3.7038020455516168e-06,
+      "loss": 0.2103,
+      "step": 10714
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.2583912738061507,
+      "learning_rate": 3.7029320226754783e-06,
+      "loss": 0.1491,
+      "step": 10715
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.0455658790886764,
+      "learning_rate": 3.7020620419029095e-06,
+      "loss": 0.198,
+      "step": 10716
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.3030615800295056,
+      "learning_rate": 3.701192103262154e-06,
+      "loss": 0.1917,
+      "step": 10717
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.616596680675628,
+      "learning_rate": 3.7003222067814504e-06,
+      "loss": 0.2293,
+      "step": 10718
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.502276263057738,
+      "learning_rate": 3.6994523524890343e-06,
+      "loss": 0.2208,
+      "step": 10719
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.4617115611843061,
+      "learning_rate": 3.698582540413145e-06,
+      "loss": 0.1771,
+      "step": 10720
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.9984089592091516,
+      "learning_rate": 3.697712770582012e-06,
+      "loss": 0.1605,
+      "step": 10721
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.5022039117267758,
+      "learning_rate": 3.6968430430238722e-06,
+      "loss": 0.1737,
+      "step": 10722
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.5098691364448917,
+      "learning_rate": 3.695973357766956e-06,
+      "loss": 0.2199,
+      "step": 10723
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.2753977724257681,
+      "learning_rate": 3.695103714839496e-06,
+      "loss": 0.1835,
+      "step": 10724
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.1527015033715844,
+      "learning_rate": 3.694234114269719e-06,
+      "loss": 0.1983,
+      "step": 10725
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.0623894591890968,
+      "learning_rate": 3.6933645560858534e-06,
+      "loss": 0.1935,
+      "step": 10726
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.0693227549397915,
+      "learning_rate": 3.6924950403161263e-06,
+      "loss": 0.2123,
+      "step": 10727
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.4902096367578803,
+      "learning_rate": 3.6916255669887608e-06,
+      "loss": 0.1837,
+      "step": 10728
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.9756334883181939,
+      "learning_rate": 3.6907561361319843e-06,
+      "loss": 0.1813,
+      "step": 10729
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.1004922652640023,
+      "learning_rate": 3.689886747774014e-06,
+      "loss": 0.1701,
+      "step": 10730
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.552357561945604,
+      "learning_rate": 3.6890174019430734e-06,
+      "loss": 0.2166,
+      "step": 10731
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.9808968392148292,
+      "learning_rate": 3.6881480986673816e-06,
+      "loss": 0.1383,
+      "step": 10732
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.058460051198053,
+      "learning_rate": 3.6872788379751573e-06,
+      "loss": 0.2251,
+      "step": 10733
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.1607832225957897,
+      "learning_rate": 3.6864096198946187e-06,
+      "loss": 0.1625,
+      "step": 10734
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.20623620400915,
+      "learning_rate": 3.6855404444539773e-06,
+      "loss": 0.1263,
+      "step": 10735
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.5144986362029238,
+      "learning_rate": 3.6846713116814497e-06,
+      "loss": 0.1957,
+      "step": 10736
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.388085876095557,
+      "learning_rate": 3.6838022216052467e-06,
+      "loss": 0.1786,
+      "step": 10737
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.2583472080865763,
+      "learning_rate": 3.6829331742535824e-06,
+      "loss": 0.1924,
+      "step": 10738
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.3556608014547733,
+      "learning_rate": 3.682064169654663e-06,
+      "loss": 0.212,
+      "step": 10739
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.2076177915707034,
+      "learning_rate": 3.681195207836698e-06,
+      "loss": 0.1951,
+      "step": 10740
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.1462809140072894,
+      "learning_rate": 3.6803262888278957e-06,
+      "loss": 0.2104,
+      "step": 10741
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.147942634281306,
+      "learning_rate": 3.6794574126564596e-06,
+      "loss": 0.1368,
+      "step": 10742
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.40400536791748,
+      "learning_rate": 3.6785885793505984e-06,
+      "loss": 0.1657,
+      "step": 10743
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.3340193435308108,
+      "learning_rate": 3.6777197889385097e-06,
+      "loss": 0.1665,
+      "step": 10744
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.3802592040229922,
+      "learning_rate": 3.6768510414483977e-06,
+      "loss": 0.225,
+      "step": 10745
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.0093617323248507,
+      "learning_rate": 3.6759823369084605e-06,
+      "loss": 0.1543,
+      "step": 10746
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.3758403583416965,
+      "learning_rate": 3.6751136753469006e-06,
+      "loss": 0.162,
+      "step": 10747
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.9941368402735047,
+      "learning_rate": 3.674245056791911e-06,
+      "loss": 0.1633,
+      "step": 10748
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.3388257121110956,
+      "learning_rate": 3.6733764812716875e-06,
+      "loss": 0.2128,
+      "step": 10749
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.0713612680782445,
+      "learning_rate": 3.6725079488144277e-06,
+      "loss": 0.1823,
+      "step": 10750
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.2302657580586411,
+      "learning_rate": 3.671639459448323e-06,
+      "loss": 0.2209,
+      "step": 10751
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.5628703965811326,
+      "learning_rate": 3.6707710132015674e-06,
+      "loss": 0.2132,
+      "step": 10752
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.5010168731577842,
+      "learning_rate": 3.669902610102347e-06,
+      "loss": 0.2064,
+      "step": 10753
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.5049450147919525,
+      "learning_rate": 3.6690342501788534e-06,
+      "loss": 0.198,
+      "step": 10754
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.5030436382757524,
+      "learning_rate": 3.668165933459273e-06,
+      "loss": 0.192,
+      "step": 10755
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.3486933656615598,
+      "learning_rate": 3.6672976599717934e-06,
+      "loss": 0.2075,
+      "step": 10756
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.0008518048206276,
+      "learning_rate": 3.6664294297445967e-06,
+      "loss": 0.1995,
+      "step": 10757
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.0739230286425128,
+      "learning_rate": 3.665561242805868e-06,
+      "loss": 0.1506,
+      "step": 10758
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.149545187264818,
+      "learning_rate": 3.6646930991837882e-06,
+      "loss": 0.239,
+      "step": 10759
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.3133392262185402,
+      "learning_rate": 3.6638249989065377e-06,
+      "loss": 0.1843,
+      "step": 10760
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.6175030386658888,
+      "learning_rate": 3.662956942002299e-06,
+      "loss": 0.2189,
+      "step": 10761
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.9672584452977689,
+      "learning_rate": 3.6620889284992443e-06,
+      "loss": 0.1643,
+      "step": 10762
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.0123128656244298,
+      "learning_rate": 3.661220958425554e-06,
+      "loss": 0.159,
+      "step": 10763
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.414879215286774,
+      "learning_rate": 3.6603530318093994e-06,
+      "loss": 0.1922,
+      "step": 10764
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.9689623346200326,
+      "learning_rate": 3.6594851486789585e-06,
+      "loss": 0.1798,
+      "step": 10765
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.4557745169557275,
+      "learning_rate": 3.658617309062398e-06,
+      "loss": 0.1984,
+      "step": 10766
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.6915422917865621,
+      "learning_rate": 3.657749512987891e-06,
+      "loss": 0.2057,
+      "step": 10767
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.9419434862481326,
+      "learning_rate": 3.656881760483607e-06,
+      "loss": 0.1764,
+      "step": 10768
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.3275042794359775,
+      "learning_rate": 3.656014051577713e-06,
+      "loss": 0.2048,
+      "step": 10769
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.4939495687565,
+      "learning_rate": 3.6551463862983782e-06,
+      "loss": 0.1949,
+      "step": 10770
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.2485569342965457,
+      "learning_rate": 3.654278764673762e-06,
+      "loss": 0.1768,
+      "step": 10771
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.4475423056756613,
+      "learning_rate": 3.653411186732033e-06,
+      "loss": 0.2187,
+      "step": 10772
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.350699302069546,
+      "learning_rate": 3.65254365250135e-06,
+      "loss": 0.1637,
+      "step": 10773
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.2826488377694656,
+      "learning_rate": 3.651676162009877e-06,
+      "loss": 0.1724,
+      "step": 10774
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.6133167595453728,
+      "learning_rate": 3.6508087152857684e-06,
+      "loss": 0.2171,
+      "step": 10775
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.9658591664736846,
+      "learning_rate": 3.6499413123571857e-06,
+      "loss": 0.1421,
+      "step": 10776
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.0255164913856003,
+      "learning_rate": 3.649073953252284e-06,
+      "loss": 0.2046,
+      "step": 10777
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.5474322596469385,
+      "learning_rate": 3.648206637999219e-06,
+      "loss": 0.1961,
+      "step": 10778
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.8899189509498715,
+      "learning_rate": 3.6473393666261446e-06,
+      "loss": 0.1323,
+      "step": 10779
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.1527882122936142,
+      "learning_rate": 3.646472139161211e-06,
+      "loss": 0.208,
+      "step": 10780
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.1147090322555318,
+      "learning_rate": 3.645604955632571e-06,
+      "loss": 0.2272,
+      "step": 10781
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.255782973426072,
+      "learning_rate": 3.644737816068372e-06,
+      "loss": 0.1995,
+      "step": 10782
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.894962962475748,
+      "learning_rate": 3.6438707204967654e-06,
+      "loss": 0.1192,
+      "step": 10783
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.1650196483110424,
+      "learning_rate": 3.6430036689458926e-06,
+      "loss": 0.2109,
+      "step": 10784
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.2529326131880176,
+      "learning_rate": 3.642136661443901e-06,
+      "loss": 0.166,
+      "step": 10785
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.2693206377658415,
+      "learning_rate": 3.641269698018934e-06,
+      "loss": 0.2038,
+      "step": 10786
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.4789638570537373,
+      "learning_rate": 3.6404027786991334e-06,
+      "loss": 0.2006,
+      "step": 10787
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.477551738182907,
+      "learning_rate": 3.6395359035126438e-06,
+      "loss": 0.199,
+      "step": 10788
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.0528952107687346,
+      "learning_rate": 3.6386690724875984e-06,
+      "loss": 0.1696,
+      "step": 10789
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.1457343011744439,
+      "learning_rate": 3.637802285652139e-06,
+      "loss": 0.167,
+      "step": 10790
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.9774938910200737,
+      "learning_rate": 3.636935543034399e-06,
+      "loss": 0.1856,
+      "step": 10791
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.2031319853344546,
+      "learning_rate": 3.636068844662518e-06,
+      "loss": 0.1735,
+      "step": 10792
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.9144199357167428,
+      "learning_rate": 3.635202190564624e-06,
+      "loss": 0.1609,
+      "step": 10793
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.6529682323059431,
+      "learning_rate": 3.6343355807688518e-06,
+      "loss": 0.2294,
+      "step": 10794
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.0270487412536446,
+      "learning_rate": 3.6334690153033318e-06,
+      "loss": 0.193,
+      "step": 10795
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.401213886768019,
+      "learning_rate": 3.6326024941961937e-06,
+      "loss": 0.1881,
+      "step": 10796
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.756117109110594,
+      "learning_rate": 3.6317360174755666e-06,
+      "loss": 0.2139,
+      "step": 10797
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.3719214633555101,
+      "learning_rate": 3.6308695851695742e-06,
+      "loss": 0.208,
+      "step": 10798
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.4370556797739578,
+      "learning_rate": 3.630003197306342e-06,
+      "loss": 0.1676,
+      "step": 10799
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.201203483847953,
+      "learning_rate": 3.6291368539139937e-06,
+      "loss": 0.2336,
+      "step": 10800
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.8635838450803677,
+      "learning_rate": 3.6282705550206533e-06,
+      "loss": 0.1756,
+      "step": 10801
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.4010315469875052,
+      "learning_rate": 3.6274043006544373e-06,
+      "loss": 0.2324,
+      "step": 10802
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.350982320987506,
+      "learning_rate": 3.6265380908434668e-06,
+      "loss": 0.1881,
+      "step": 10803
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.3381099966589827,
+      "learning_rate": 3.6256719256158603e-06,
+      "loss": 0.239,
+      "step": 10804
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.9729148681788734,
+      "learning_rate": 3.624805804999734e-06,
+      "loss": 0.1242,
+      "step": 10805
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.0578601378472237,
+      "learning_rate": 3.623939729023202e-06,
+      "loss": 0.2484,
+      "step": 10806
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.0484378095532572,
+      "learning_rate": 3.6230736977143764e-06,
+      "loss": 0.1988,
+      "step": 10807
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.4350831833452833,
+      "learning_rate": 3.622207711101372e-06,
+      "loss": 0.2186,
+      "step": 10808
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.0830336929682503,
+      "learning_rate": 3.6213417692122967e-06,
+      "loss": 0.1613,
+      "step": 10809
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.8810243156101217,
+      "learning_rate": 3.620475872075262e-06,
+      "loss": 0.1769,
+      "step": 10810
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.1799598373179128,
+      "learning_rate": 3.6196100197183713e-06,
+      "loss": 0.2267,
+      "step": 10811
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.600892594422696,
+      "learning_rate": 3.6187442121697336e-06,
+      "loss": 0.2246,
+      "step": 10812
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.2210264189810105,
+      "learning_rate": 3.617878449457453e-06,
+      "loss": 0.2229,
+      "step": 10813
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.0484879046687507,
+      "learning_rate": 3.6170127316096337e-06,
+      "loss": 0.1697,
+      "step": 10814
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.8532743724070122,
+      "learning_rate": 3.616147058654377e-06,
+      "loss": 0.1721,
+      "step": 10815
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.5960328381305877,
+      "learning_rate": 3.6152814306197805e-06,
+      "loss": 0.2277,
+      "step": 10816
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.1170952231971865,
+      "learning_rate": 3.6144158475339463e-06,
+      "loss": 0.2011,
+      "step": 10817
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.4085380245913572,
+      "learning_rate": 3.61355030942497e-06,
+      "loss": 0.165,
+      "step": 10818
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.248573499559983,
+      "learning_rate": 3.6126848163209494e-06,
+      "loss": 0.1964,
+      "step": 10819
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.3047890566514964,
+      "learning_rate": 3.6118193682499758e-06,
+      "loss": 0.2113,
+      "step": 10820
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.2524861913336711,
+      "learning_rate": 3.6109539652401436e-06,
+      "loss": 0.1872,
+      "step": 10821
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.1331979227456177,
+      "learning_rate": 3.610088607319544e-06,
+      "loss": 0.1905,
+      "step": 10822
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.997332134665158,
+      "learning_rate": 3.6092232945162697e-06,
+      "loss": 0.198,
+      "step": 10823
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.3881213173866802,
+      "learning_rate": 3.608358026858406e-06,
+      "loss": 0.2487,
+      "step": 10824
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.0609966715237795,
+      "learning_rate": 3.6074928043740408e-06,
+      "loss": 0.2174,
+      "step": 10825
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 9.762851168559989,
+      "learning_rate": 3.606627627091259e-06,
+      "loss": 0.1698,
+      "step": 10826
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.7931466107879477,
+      "learning_rate": 3.6057624950381464e-06,
+      "loss": 0.1732,
+      "step": 10827
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.421336858170369,
+      "learning_rate": 3.6048974082427867e-06,
+      "loss": 0.1952,
+      "step": 10828
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.3126493351993398,
+      "learning_rate": 3.604032366733258e-06,
+      "loss": 0.2112,
+      "step": 10829
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.4854793205218173,
+      "learning_rate": 3.6031673705376403e-06,
+      "loss": 0.2325,
+      "step": 10830
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.2173938343118942,
+      "learning_rate": 3.6023024196840128e-06,
+      "loss": 0.2023,
+      "step": 10831
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.2478336516244417,
+      "learning_rate": 3.6014375142004544e-06,
+      "loss": 0.1477,
+      "step": 10832
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.0293887213204735,
+      "learning_rate": 3.6005726541150377e-06,
+      "loss": 0.1141,
+      "step": 10833
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.6609499705681732,
+      "learning_rate": 3.599707839455838e-06,
+      "loss": 0.1911,
+      "step": 10834
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.0262053676987122,
+      "learning_rate": 3.598843070250926e-06,
+      "loss": 0.1715,
+      "step": 10835
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.3143413972345823,
+      "learning_rate": 3.5979783465283733e-06,
+      "loss": 0.2185,
+      "step": 10836
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.4002341764305521,
+      "learning_rate": 3.5971136683162523e-06,
+      "loss": 0.1915,
+      "step": 10837
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.6041976068555601,
+      "learning_rate": 3.596249035642626e-06,
+      "loss": 0.193,
+      "step": 10838
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.2319121780378677,
+      "learning_rate": 3.5953844485355627e-06,
+      "loss": 0.2245,
+      "step": 10839
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.197094007330067,
+      "learning_rate": 3.5945199070231286e-06,
+      "loss": 0.1902,
+      "step": 10840
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.314826337026979,
+      "learning_rate": 3.5936554111333875e-06,
+      "loss": 0.2087,
+      "step": 10841
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.1605663923216458,
+      "learning_rate": 3.5927909608944e-06,
+      "loss": 0.2148,
+      "step": 10842
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.243472253635409,
+      "learning_rate": 3.5919265563342274e-06,
+      "loss": 0.2048,
+      "step": 10843
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.0004014498338352,
+      "learning_rate": 3.5910621974809277e-06,
+      "loss": 0.2083,
+      "step": 10844
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.0702685743910811,
+      "learning_rate": 3.590197884362559e-06,
+      "loss": 0.1683,
+      "step": 10845
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.8321526667158804,
+      "learning_rate": 3.5893336170071803e-06,
+      "loss": 0.1514,
+      "step": 10846
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.3903354493345865,
+      "learning_rate": 3.5884693954428408e-06,
+      "loss": 0.2317,
+      "step": 10847
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.2227259088971045,
+      "learning_rate": 3.587605219697596e-06,
+      "loss": 0.175,
+      "step": 10848
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.6131665579273131,
+      "learning_rate": 3.5867410897994993e-06,
+      "loss": 0.1763,
+      "step": 10849
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.7607578312860686,
+      "learning_rate": 3.5858770057766e-06,
+      "loss": 0.2235,
+      "step": 10850
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.8894389211338842,
+      "learning_rate": 3.5850129676569457e-06,
+      "loss": 0.196,
+      "step": 10851
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.8551083104241568,
+      "learning_rate": 3.584148975468584e-06,
+      "loss": 0.2127,
+      "step": 10852
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.3779483553286276,
+      "learning_rate": 3.5832850292395603e-06,
+      "loss": 0.197,
+      "step": 10853
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.26649272865824,
+      "learning_rate": 3.5824211289979194e-06,
+      "loss": 0.2159,
+      "step": 10854
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.483777206913829,
+      "learning_rate": 3.5815572747717054e-06,
+      "loss": 0.2643,
+      "step": 10855
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.9672761326484016,
+      "learning_rate": 3.580693466588956e-06,
+      "loss": 0.1676,
+      "step": 10856
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.2809571652602467,
+      "learning_rate": 3.5798297044777132e-06,
+      "loss": 0.2173,
+      "step": 10857
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.0491541989623183,
+      "learning_rate": 3.5789659884660145e-06,
+      "loss": 0.1767,
+      "step": 10858
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.4734411525782176,
+      "learning_rate": 3.578102318581898e-06,
+      "loss": 0.1537,
+      "step": 10859
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.1913759728008377,
+      "learning_rate": 3.5772386948533978e-06,
+      "loss": 0.1924,
+      "step": 10860
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.9393318518324774,
+      "learning_rate": 3.5763751173085483e-06,
+      "loss": 0.1608,
+      "step": 10861
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.1217207434658176,
+      "learning_rate": 3.57551158597538e-06,
+      "loss": 0.169,
+      "step": 10862
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.2941010953586911,
+      "learning_rate": 3.574648100881924e-06,
+      "loss": 0.2211,
+      "step": 10863
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.1471534365675842,
+      "learning_rate": 3.573784662056212e-06,
+      "loss": 0.1474,
+      "step": 10864
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.3062246655558127,
+      "learning_rate": 3.5729212695262693e-06,
+      "loss": 0.1923,
+      "step": 10865
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.7416669983232294,
+      "learning_rate": 3.5720579233201226e-06,
+      "loss": 0.2018,
+      "step": 10866
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.2649862861785193,
+      "learning_rate": 3.5711946234657956e-06,
+      "loss": 0.1499,
+      "step": 10867
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.3780195097436685,
+      "learning_rate": 3.570331369991315e-06,
+      "loss": 0.2286,
+      "step": 10868
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.9283195336689923,
+      "learning_rate": 3.5694681629246997e-06,
+      "loss": 0.1826,
+      "step": 10869
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.3735985753145277,
+      "learning_rate": 3.568605002293971e-06,
+      "loss": 0.1851,
+      "step": 10870
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.158974961260913,
+      "learning_rate": 3.5677418881271452e-06,
+      "loss": 0.1649,
+      "step": 10871
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.3126823945694717,
+      "learning_rate": 3.566878820452242e-06,
+      "loss": 0.1805,
+      "step": 10872
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.20537012881015,
+      "learning_rate": 3.566015799297278e-06,
+      "loss": 0.1724,
+      "step": 10873
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.5482382065813385,
+      "learning_rate": 3.5651528246902647e-06,
+      "loss": 0.2071,
+      "step": 10874
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.4539060595725959,
+      "learning_rate": 3.5642898966592145e-06,
+      "loss": 0.2586,
+      "step": 10875
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.3009194481096857,
+      "learning_rate": 3.5634270152321403e-06,
+      "loss": 0.1757,
+      "step": 10876
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.242136378405928,
+      "learning_rate": 3.5625641804370526e-06,
+      "loss": 0.2232,
+      "step": 10877
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.2792551922013453,
+      "learning_rate": 3.5617013923019576e-06,
+      "loss": 0.2547,
+      "step": 10878
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.952558624726865,
+      "learning_rate": 3.560838650854863e-06,
+      "loss": 0.2584,
+      "step": 10879
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.2494925959242937,
+      "learning_rate": 3.5599759561237725e-06,
+      "loss": 0.1912,
+      "step": 10880
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.3222910626442463,
+      "learning_rate": 3.5591133081366895e-06,
+      "loss": 0.2148,
+      "step": 10881
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.5617135500570865,
+      "learning_rate": 3.5582507069216197e-06,
+      "loss": 0.2037,
+      "step": 10882
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.127358989916579,
+      "learning_rate": 3.5573881525065594e-06,
+      "loss": 0.1891,
+      "step": 10883
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.2288995819007895,
+      "learning_rate": 3.556525644919509e-06,
+      "loss": 0.2362,
+      "step": 10884
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.2124015617598531,
+      "learning_rate": 3.555663184188466e-06,
+      "loss": 0.185,
+      "step": 10885
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.9352354375476137,
+      "learning_rate": 3.554800770341428e-06,
+      "loss": 0.1968,
+      "step": 10886
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.317107776353738,
+      "learning_rate": 3.553938403406388e-06,
+      "loss": 0.2138,
+      "step": 10887
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.7314004671729881,
+      "learning_rate": 3.5530760834113388e-06,
+      "loss": 0.192,
+      "step": 10888
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.0062685180671993,
+      "learning_rate": 3.552213810384271e-06,
+      "loss": 0.1577,
+      "step": 10889
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.9345762378181915,
+      "learning_rate": 3.5513515843531756e-06,
+      "loss": 0.1927,
+      "step": 10890
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.9140537742459095,
+      "learning_rate": 3.5504894053460427e-06,
+      "loss": 0.1311,
+      "step": 10891
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.3377147322258942,
+      "learning_rate": 3.5496272733908545e-06,
+      "loss": 0.1991,
+      "step": 10892
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.4357703093430425,
+      "learning_rate": 3.5487651885156e-06,
+      "loss": 0.1993,
+      "step": 10893
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.334483565709186,
+      "learning_rate": 3.5479031507482614e-06,
+      "loss": 0.1837,
+      "step": 10894
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.272366026166051,
+      "learning_rate": 3.547041160116823e-06,
+      "loss": 0.163,
+      "step": 10895
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.1950843209904483,
+      "learning_rate": 3.5461792166492635e-06,
+      "loss": 0.1986,
+      "step": 10896
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.2907535096445344,
+      "learning_rate": 3.5453173203735635e-06,
+      "loss": 0.1218,
+      "step": 10897
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.7417071561255513,
+      "learning_rate": 3.5444554713176984e-06,
+      "loss": 0.2488,
+      "step": 10898
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.284955417744705,
+      "learning_rate": 3.543593669509645e-06,
+      "loss": 0.1818,
+      "step": 10899
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.3229263407310947,
+      "learning_rate": 3.5427319149773815e-06,
+      "loss": 0.2082,
+      "step": 10900
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.1330233098235927,
+      "learning_rate": 3.5418702077488753e-06,
+      "loss": 0.2112,
+      "step": 10901
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.1213221699422249,
+      "learning_rate": 3.541008547852102e-06,
+      "loss": 0.2488,
+      "step": 10902
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.168868898142452,
+      "learning_rate": 3.540146935315029e-06,
+      "loss": 0.15,
+      "step": 10903
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.2129407187420276,
+      "learning_rate": 3.539285370165627e-06,
+      "loss": 0.1745,
+      "step": 10904
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.0963328773609198,
+      "learning_rate": 3.538423852431862e-06,
+      "loss": 0.1959,
+      "step": 10905
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.3170948915744756,
+      "learning_rate": 3.5375623821416992e-06,
+      "loss": 0.1686,
+      "step": 10906
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.332214868592359,
+      "learning_rate": 3.5367009593231017e-06,
+      "loss": 0.2302,
+      "step": 10907
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.3882111403408677,
+      "learning_rate": 3.535839584004032e-06,
+      "loss": 0.1817,
+      "step": 10908
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.2018754991688043,
+      "learning_rate": 3.534978256212454e-06,
+      "loss": 0.1632,
+      "step": 10909
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.0931347990114815,
+      "learning_rate": 3.5341169759763215e-06,
+      "loss": 0.1551,
+      "step": 10910
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.4730936824935574,
+      "learning_rate": 3.533255743323596e-06,
+      "loss": 0.1948,
+      "step": 10911
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.1481686686173358,
+      "learning_rate": 3.5323945582822316e-06,
+      "loss": 0.1602,
+      "step": 10912
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.0147096617275435,
+      "learning_rate": 3.531533420880185e-06,
+      "loss": 0.1668,
+      "step": 10913
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.1112842484110814,
+      "learning_rate": 3.5306723311454085e-06,
+      "loss": 0.1589,
+      "step": 10914
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.1927326198531736,
+      "learning_rate": 3.5298112891058523e-06,
+      "loss": 0.2086,
+      "step": 10915
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.1892255305386494,
+      "learning_rate": 3.528950294789466e-06,
+      "loss": 0.2301,
+      "step": 10916
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.9433294492915119,
+      "learning_rate": 3.5280893482241995e-06,
+      "loss": 0.1605,
+      "step": 10917
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.0779428777425082,
+      "learning_rate": 3.5272284494380012e-06,
+      "loss": 0.1731,
+      "step": 10918
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.1470927568990943,
+      "learning_rate": 3.526367598458812e-06,
+      "loss": 0.2575,
+      "step": 10919
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.409130279937423,
+      "learning_rate": 3.5255067953145784e-06,
+      "loss": 0.2399,
+      "step": 10920
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.7255862785990992,
+      "learning_rate": 3.5246460400332416e-06,
+      "loss": 0.2227,
+      "step": 10921
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.2307344331465873,
+      "learning_rate": 3.523785332642744e-06,
+      "loss": 0.1829,
+      "step": 10922
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.23757619370355,
+      "learning_rate": 3.522924673171023e-06,
+      "loss": 0.2096,
+      "step": 10923
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.7022868990322686,
+      "learning_rate": 3.5220640616460173e-06,
+      "loss": 0.2159,
+      "step": 10924
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.29981107301813,
+      "learning_rate": 3.5212034980956606e-06,
+      "loss": 0.1945,
+      "step": 10925
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.2506570326069033,
+      "learning_rate": 3.5203429825478887e-06,
+      "loss": 0.1694,
+      "step": 10926
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.4830762426614237,
+      "learning_rate": 3.5194825150306365e-06,
+      "loss": 0.1852,
+      "step": 10927
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.3070605391591326,
+      "learning_rate": 3.518622095571831e-06,
+      "loss": 0.1809,
+      "step": 10928
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.8042655674525885,
+      "learning_rate": 3.5177617241994043e-06,
+      "loss": 0.1963,
+      "step": 10929
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.3672056632879008,
+      "learning_rate": 3.516901400941285e-06,
+      "loss": 0.1921,
+      "step": 10930
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.1347025970315088,
+      "learning_rate": 3.516041125825399e-06,
+      "loss": 0.2358,
+      "step": 10931
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.2765331817071075,
+      "learning_rate": 3.515180898879672e-06,
+      "loss": 0.1991,
+      "step": 10932
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.3400357361866073,
+      "learning_rate": 3.5143207201320268e-06,
+      "loss": 0.1732,
+      "step": 10933
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.4338888701306047,
+      "learning_rate": 3.513460589610384e-06,
+      "loss": 0.2236,
+      "step": 10934
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.120585685339267,
+      "learning_rate": 3.5126005073426654e-06,
+      "loss": 0.181,
+      "step": 10935
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.157364373511958,
+      "learning_rate": 3.5117404733567922e-06,
+      "loss": 0.1599,
+      "step": 10936
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.2435535261347366,
+      "learning_rate": 3.5108804876806767e-06,
+      "loss": 0.1616,
+      "step": 10937
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.6074139072225373,
+      "learning_rate": 3.510020550342237e-06,
+      "loss": 0.2152,
+      "step": 10938
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.9958289972647166,
+      "learning_rate": 3.509160661369386e-06,
+      "loss": 0.2052,
+      "step": 10939
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.9355014320119952,
+      "learning_rate": 3.5083008207900394e-06,
+      "loss": 0.1508,
+      "step": 10940
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.9547106478611616,
+      "learning_rate": 3.507441028632106e-06,
+      "loss": 0.1578,
+      "step": 10941
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.173631232659427,
+      "learning_rate": 3.506581284923495e-06,
+      "loss": 0.1668,
+      "step": 10942
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.2264240089517788,
+      "learning_rate": 3.5057215896921128e-06,
+      "loss": 0.2016,
+      "step": 10943
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.3687886208698756,
+      "learning_rate": 3.504861942965867e-06,
+      "loss": 0.2409,
+      "step": 10944
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.1573846700960428,
+      "learning_rate": 3.5040023447726646e-06,
+      "loss": 0.154,
+      "step": 10945
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.1907872952962386,
+      "learning_rate": 3.503142795140404e-06,
+      "loss": 0.1894,
+      "step": 10946
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.2715525982392235,
+      "learning_rate": 3.5022832940969896e-06,
+      "loss": 0.1581,
+      "step": 10947
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.8971349167603243,
+      "learning_rate": 3.5014238416703207e-06,
+      "loss": 0.1395,
+      "step": 10948
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.3966542040655607,
+      "learning_rate": 3.5005644378882962e-06,
+      "loss": 0.2485,
+      "step": 10949
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.1897907873978812,
+      "learning_rate": 3.4997050827788126e-06,
+      "loss": 0.1911,
+      "step": 10950
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.333056170246657,
+      "learning_rate": 3.498845776369765e-06,
+      "loss": 0.1781,
+      "step": 10951
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.0233943071059595,
+      "learning_rate": 3.497986518689045e-06,
+      "loss": 0.1277,
+      "step": 10952
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.1769669437958241,
+      "learning_rate": 3.4971273097645466e-06,
+      "loss": 0.2439,
+      "step": 10953
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.8743808953063982,
+      "learning_rate": 3.496268149624162e-06,
+      "loss": 0.1188,
+      "step": 10954
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.0437905542124517,
+      "learning_rate": 3.4954090382957763e-06,
+      "loss": 0.1736,
+      "step": 10955
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.0903611510266582,
+      "learning_rate": 3.4945499758072778e-06,
+      "loss": 0.205,
+      "step": 10956
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.2724566338818393,
+      "learning_rate": 3.493690962186553e-06,
+      "loss": 0.158,
+      "step": 10957
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.715322158398491,
+      "learning_rate": 3.4928319974614866e-06,
+      "loss": 0.1677,
+      "step": 10958
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.4201011374023944,
+      "learning_rate": 3.4919730816599605e-06,
+      "loss": 0.2118,
+      "step": 10959
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.2304120580540836,
+      "learning_rate": 3.4911142148098554e-06,
+      "loss": 0.1625,
+      "step": 10960
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.1740827839571193,
+      "learning_rate": 3.4902553969390498e-06,
+      "loss": 0.2549,
+      "step": 10961
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.9641330781262156,
+      "learning_rate": 3.4893966280754216e-06,
+      "loss": 0.135,
+      "step": 10962
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.3265487292757807,
+      "learning_rate": 3.4885379082468496e-06,
+      "loss": 0.1991,
+      "step": 10963
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.9133044873781982,
+      "learning_rate": 3.4876792374812045e-06,
+      "loss": 0.1151,
+      "step": 10964
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.5759157116798488,
+      "learning_rate": 3.486820615806361e-06,
+      "loss": 0.1665,
+      "step": 10965
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.83907796226101,
+      "learning_rate": 3.4859620432501905e-06,
+      "loss": 0.2272,
+      "step": 10966
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.1292605570069971,
+      "learning_rate": 3.4851035198405638e-06,
+      "loss": 0.1962,
+      "step": 10967
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.292698033737764,
+      "learning_rate": 3.4842450456053476e-06,
+      "loss": 0.1993,
+      "step": 10968
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.285622764255358,
+      "learning_rate": 3.483386620572409e-06,
+      "loss": 0.1698,
+      "step": 10969
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.3812859293204525,
+      "learning_rate": 3.4825282447696117e-06,
+      "loss": 0.2064,
+      "step": 10970
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.2399991270416018,
+      "learning_rate": 3.4816699182248203e-06,
+      "loss": 0.2028,
+      "step": 10971
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.2560611165326594,
+      "learning_rate": 3.480811640965899e-06,
+      "loss": 0.1629,
+      "step": 10972
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.6400580925363815,
+      "learning_rate": 3.4799534130207024e-06,
+      "loss": 0.1887,
+      "step": 10973
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.3745449345450336,
+      "learning_rate": 3.4790952344170927e-06,
+      "loss": 0.1867,
+      "step": 10974
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.0872934644067458,
+      "learning_rate": 3.4782371051829255e-06,
+      "loss": 0.2432,
+      "step": 10975
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.1986325785335357,
+      "learning_rate": 3.477379025346059e-06,
+      "loss": 0.1634,
+      "step": 10976
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.2195794633013948,
+      "learning_rate": 3.476520994934345e-06,
+      "loss": 0.1785,
+      "step": 10977
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.9746674843323263,
+      "learning_rate": 3.475663013975634e-06,
+      "loss": 0.1474,
+      "step": 10978
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.411345935026039,
+      "learning_rate": 3.4748050824977786e-06,
+      "loss": 0.2062,
+      "step": 10979
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.1288783027067122,
+      "learning_rate": 3.473947200528627e-06,
+      "loss": 0.1602,
+      "step": 10980
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.6560597400428494,
+      "learning_rate": 3.4730893680960267e-06,
+      "loss": 0.2407,
+      "step": 10981
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.3631772682080892,
+      "learning_rate": 3.4722315852278253e-06,
+      "loss": 0.1986,
+      "step": 10982
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.9101820152139461,
+      "learning_rate": 3.4713738519518635e-06,
+      "loss": 0.1884,
+      "step": 10983
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.1759645249492412,
+      "learning_rate": 3.4705161682959853e-06,
+      "loss": 0.1792,
+      "step": 10984
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.3566748505198278,
+      "learning_rate": 3.469658534288033e-06,
+      "loss": 0.2062,
+      "step": 10985
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.2420016455616394,
+      "learning_rate": 3.468800949955845e-06,
+      "loss": 0.2107,
+      "step": 10986
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.98638009038324,
+      "learning_rate": 3.467943415327258e-06,
+      "loss": 0.13,
+      "step": 10987
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.0532773004941935,
+      "learning_rate": 3.467085930430108e-06,
+      "loss": 0.2598,
+      "step": 10988
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.8974215925047382,
+      "learning_rate": 3.46622849529223e-06,
+      "loss": 0.2223,
+      "step": 10989
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.196306507216956,
+      "learning_rate": 3.4653711099414572e-06,
+      "loss": 0.1607,
+      "step": 10990
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.1018523923747616,
+      "learning_rate": 3.4645137744056227e-06,
+      "loss": 0.2537,
+      "step": 10991
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.9649428382371065,
+      "learning_rate": 3.463656488712552e-06,
+      "loss": 0.2139,
+      "step": 10992
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.130198748572824,
+      "learning_rate": 3.462799252890075e-06,
+      "loss": 0.1938,
+      "step": 10993
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.2445047645121308,
+      "learning_rate": 3.4619420669660197e-06,
+      "loss": 0.2211,
+      "step": 10994
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.9772881977689536,
+      "learning_rate": 3.4610849309682086e-06,
+      "loss": 0.1793,
+      "step": 10995
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.049560532677476,
+      "learning_rate": 3.4602278449244663e-06,
+      "loss": 0.196,
+      "step": 10996
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.0790159096534822,
+      "learning_rate": 3.4593708088626123e-06,
+      "loss": 0.226,
+      "step": 10997
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.0972754676624559,
+      "learning_rate": 3.4585138228104673e-06,
+      "loss": 0.1837,
+      "step": 10998
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.2005237381404175,
+      "learning_rate": 3.4576568867958506e-06,
+      "loss": 0.196,
+      "step": 10999
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.442220375248412,
+      "learning_rate": 3.4568000008465803e-06,
+      "loss": 0.1865,
+      "step": 11000
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.4731413854174078,
+      "learning_rate": 3.4559431649904675e-06,
+      "loss": 0.2045,
+      "step": 11001
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.1878538169288737,
+      "learning_rate": 3.455086379255327e-06,
+      "loss": 0.2271,
+      "step": 11002
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.9109769321890892,
+      "learning_rate": 3.4542296436689713e-06,
+      "loss": 0.1602,
+      "step": 11003
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.2379702589687138,
+      "learning_rate": 3.4533729582592113e-06,
+      "loss": 0.175,
+      "step": 11004
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.753653794876605,
+      "learning_rate": 3.4525163230538544e-06,
+      "loss": 0.2182,
+      "step": 11005
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.3013731093989733,
+      "learning_rate": 3.4516597380807063e-06,
+      "loss": 0.2388,
+      "step": 11006
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.112059096367871,
+      "learning_rate": 3.4508032033675737e-06,
+      "loss": 0.1167,
+      "step": 11007
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.4334888263941512,
+      "learning_rate": 3.44994671894226e-06,
+      "loss": 0.1703,
+      "step": 11008
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.208915853330333,
+      "learning_rate": 3.449090284832569e-06,
+      "loss": 0.174,
+      "step": 11009
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.1373286708377728,
+      "learning_rate": 3.4482339010662968e-06,
+      "loss": 0.166,
+      "step": 11010
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.1061652722213169,
+      "learning_rate": 3.447377567671245e-06,
+      "loss": 0.1531,
+      "step": 11011
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.1225058471446687,
+      "learning_rate": 3.446521284675211e-06,
+      "loss": 0.23,
+      "step": 11012
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.8195096583108546,
+      "learning_rate": 3.44566505210599e-06,
+      "loss": 0.1672,
+      "step": 11013
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.2554441584331097,
+      "learning_rate": 3.444808869991374e-06,
+      "loss": 0.1662,
+      "step": 11014
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.571650938157704,
+      "learning_rate": 3.4439527383591566e-06,
+      "loss": 0.197,
+      "step": 11015
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.1679557812331147,
+      "learning_rate": 3.443096657237127e-06,
+      "loss": 0.2162,
+      "step": 11016
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.033161546487098,
+      "learning_rate": 3.442240626653076e-06,
+      "loss": 0.1668,
+      "step": 11017
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.2492999001280831,
+      "learning_rate": 3.4413846466347917e-06,
+      "loss": 0.2145,
+      "step": 11018
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.9300479370530771,
+      "learning_rate": 3.440528717210056e-06,
+      "loss": 0.1699,
+      "step": 11019
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.2705971571334025,
+      "learning_rate": 3.439672838406655e-06,
+      "loss": 0.1933,
+      "step": 11020
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.331682450389343,
+      "learning_rate": 3.4388170102523723e-06,
+      "loss": 0.1746,
+      "step": 11021
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.6367658708476482,
+      "learning_rate": 3.437961232774987e-06,
+      "loss": 0.2207,
+      "step": 11022
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.0884554097194101,
+      "learning_rate": 3.437105506002277e-06,
+      "loss": 0.2003,
+      "step": 11023
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.8701230893432624,
+      "learning_rate": 3.4362498299620215e-06,
+      "loss": 0.1851,
+      "step": 11024
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.053390754793062,
+      "learning_rate": 3.4353942046819957e-06,
+      "loss": 0.1685,
+      "step": 11025
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.471619035854387,
+      "learning_rate": 3.434538630189973e-06,
+      "loss": 0.2524,
+      "step": 11026
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.5103338980742445,
+      "learning_rate": 3.4336831065137286e-06,
+      "loss": 0.1776,
+      "step": 11027
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.1240925572915452,
+      "learning_rate": 3.4328276336810303e-06,
+      "loss": 0.1915,
+      "step": 11028
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.1641551899330043,
+      "learning_rate": 3.4319722117196473e-06,
+      "loss": 0.1605,
+      "step": 11029
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.3430449554587636,
+      "learning_rate": 3.431116840657349e-06,
+      "loss": 0.2037,
+      "step": 11030
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.7022356721166094,
+      "learning_rate": 3.430261520521901e-06,
+      "loss": 0.2262,
+      "step": 11031
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.615313709068178,
+      "learning_rate": 3.429406251341066e-06,
+      "loss": 0.2468,
+      "step": 11032
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.0505735926018949,
+      "learning_rate": 3.4285510331426063e-06,
+      "loss": 0.2055,
+      "step": 11033
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.1679958066070983,
+      "learning_rate": 3.4276958659542838e-06,
+      "loss": 0.1951,
+      "step": 11034
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.0093651241079178,
+      "learning_rate": 3.4268407498038574e-06,
+      "loss": 0.1555,
+      "step": 11035
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.3462216273898038,
+      "learning_rate": 3.4259856847190874e-06,
+      "loss": 0.2323,
+      "step": 11036
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.2701849354909198,
+      "learning_rate": 3.4251306707277247e-06,
+      "loss": 0.1681,
+      "step": 11037
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.861287810232022,
+      "learning_rate": 3.424275707857526e-06,
+      "loss": 0.1938,
+      "step": 11038
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.5964734518798918,
+      "learning_rate": 3.4234207961362454e-06,
+      "loss": 0.2078,
+      "step": 11039
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 5.421367409881396,
+      "learning_rate": 3.4225659355916324e-06,
+      "loss": 0.1753,
+      "step": 11040
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.1937978254344686,
+      "learning_rate": 3.421711126251436e-06,
+      "loss": 0.2195,
+      "step": 11041
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.4597551181610062,
+      "learning_rate": 3.4208563681434025e-06,
+      "loss": 0.1832,
+      "step": 11042
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.3049289228733978,
+      "learning_rate": 3.42000166129528e-06,
+      "loss": 0.1638,
+      "step": 11043
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.251796162089707,
+      "learning_rate": 3.419147005734812e-06,
+      "loss": 0.1754,
+      "step": 11044
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.2201077161201186,
+      "learning_rate": 3.4182924014897432e-06,
+      "loss": 0.2212,
+      "step": 11045
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.5250000618520319,
+      "learning_rate": 3.4174378485878114e-06,
+      "loss": 0.182,
+      "step": 11046
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.9034098441629013,
+      "learning_rate": 3.4165833470567557e-06,
+      "loss": 0.1599,
+      "step": 11047
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.6158404291506039,
+      "learning_rate": 3.415728896924317e-06,
+      "loss": 0.1655,
+      "step": 11048
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.289260927013651,
+      "learning_rate": 3.4148744982182293e-06,
+      "loss": 0.2087,
+      "step": 11049
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.2102977077733563,
+      "learning_rate": 3.4140201509662265e-06,
+      "loss": 0.1519,
+      "step": 11050
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.4294293566636085,
+      "learning_rate": 3.41316585519604e-06,
+      "loss": 0.1949,
+      "step": 11051
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.0258676854227604,
+      "learning_rate": 3.412311610935404e-06,
+      "loss": 0.1838,
+      "step": 11052
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.394144218228563,
+      "learning_rate": 3.4114574182120443e-06,
+      "loss": 0.199,
+      "step": 11053
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.1554838103472906,
+      "learning_rate": 3.4106032770536933e-06,
+      "loss": 0.2021,
+      "step": 11054
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.9947285515634067,
+      "learning_rate": 3.4097491874880718e-06,
+      "loss": 0.2124,
+      "step": 11055
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.8624155841597607,
+      "learning_rate": 3.4088951495429055e-06,
+      "loss": 0.1638,
+      "step": 11056
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.2223574773009598,
+      "learning_rate": 3.4080411632459197e-06,
+      "loss": 0.1643,
+      "step": 11057
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.0711927101112713,
+      "learning_rate": 3.407187228624833e-06,
+      "loss": 0.2033,
+      "step": 11058
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.098209326610489,
+      "learning_rate": 3.4063333457073645e-06,
+      "loss": 0.187,
+      "step": 11059
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.3061797237200357,
+      "learning_rate": 3.4054795145212305e-06,
+      "loss": 0.1877,
+      "step": 11060
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.0639549559527208,
+      "learning_rate": 3.4046257350941492e-06,
+      "loss": 0.1896,
+      "step": 11061
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.1327629473141394,
+      "learning_rate": 3.403772007453833e-06,
+      "loss": 0.2016,
+      "step": 11062
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.1512611929238588,
+      "learning_rate": 3.402918331627998e-06,
+      "loss": 0.1741,
+      "step": 11063
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.4859489779425163,
+      "learning_rate": 3.4020647076443502e-06,
+      "loss": 0.2066,
+      "step": 11064
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.7705020683519423,
+      "learning_rate": 3.401211135530601e-06,
+      "loss": 0.223,
+      "step": 11065
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.4724178778994426,
+      "learning_rate": 3.400357615314459e-06,
+      "loss": 0.2137,
+      "step": 11066
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.301078304246276,
+      "learning_rate": 3.3995041470236277e-06,
+      "loss": 0.1488,
+      "step": 11067
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.0897895327792495,
+      "learning_rate": 3.398650730685813e-06,
+      "loss": 0.1776,
+      "step": 11068
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.6849367180000387,
+      "learning_rate": 3.3977973663287156e-06,
+      "loss": 0.1936,
+      "step": 11069
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.3196582301708373,
+      "learning_rate": 3.3969440539800362e-06,
+      "loss": 0.1733,
+      "step": 11070
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.3795991350866792,
+      "learning_rate": 3.3960907936674746e-06,
+      "loss": 0.229,
+      "step": 11071
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.1180201192099655,
+      "learning_rate": 3.3952375854187316e-06,
+      "loss": 0.1974,
+      "step": 11072
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.2892653623538155,
+      "learning_rate": 3.394384429261496e-06,
+      "loss": 0.2056,
+      "step": 11073
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.9638580942159799,
+      "learning_rate": 3.3935313252234653e-06,
+      "loss": 0.1626,
+      "step": 11074
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.3446706346061015,
+      "learning_rate": 3.3926782733323325e-06,
+      "loss": 0.1785,
+      "step": 11075
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.322309986311349,
+      "learning_rate": 3.3918252736157874e-06,
+      "loss": 0.2101,
+      "step": 11076
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.1309151808651623,
+      "learning_rate": 3.3909723261015194e-06,
+      "loss": 0.1919,
+      "step": 11077
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.0734913565383655,
+      "learning_rate": 3.3901194308172126e-06,
+      "loss": 0.1277,
+      "step": 11078
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.3615985475889762,
+      "learning_rate": 3.3892665877905557e-06,
+      "loss": 0.1751,
+      "step": 11079
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.5621342564723943,
+      "learning_rate": 3.3884137970492313e-06,
+      "loss": 0.2254,
+      "step": 11080
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.928960383596822,
+      "learning_rate": 3.387561058620924e-06,
+      "loss": 0.1934,
+      "step": 11081
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.1470441410268912,
+      "learning_rate": 3.38670837253331e-06,
+      "loss": 0.1803,
+      "step": 11082
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.223999802583557,
+      "learning_rate": 3.3858557388140705e-06,
+      "loss": 0.1972,
+      "step": 11083
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.3645304937785379,
+      "learning_rate": 3.3850031574908815e-06,
+      "loss": 0.184,
+      "step": 11084
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.4477132910552393,
+      "learning_rate": 3.3841506285914193e-06,
+      "loss": 0.2,
+      "step": 11085
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.8791541661184357,
+      "learning_rate": 3.3832981521433573e-06,
+      "loss": 0.151,
+      "step": 11086
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.8639876588056774,
+      "learning_rate": 3.382445728174365e-06,
+      "loss": 0.2103,
+      "step": 11087
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.1429399031716414,
+      "learning_rate": 3.381593356712115e-06,
+      "loss": 0.224,
+      "step": 11088
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.1397126350528517,
+      "learning_rate": 3.380741037784276e-06,
+      "loss": 0.1676,
+      "step": 11089
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.2308992510758374,
+      "learning_rate": 3.3798887714185157e-06,
+      "loss": 0.2081,
+      "step": 11090
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.3435524972774966,
+      "learning_rate": 3.379036557642495e-06,
+      "loss": 0.156,
+      "step": 11091
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.1923175640634427,
+      "learning_rate": 3.3781843964838814e-06,
+      "loss": 0.2186,
+      "step": 11092
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.0093626550090768,
+      "learning_rate": 3.377332287970333e-06,
+      "loss": 0.155,
+      "step": 11093
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.3420655427176185,
+      "learning_rate": 3.376480232129513e-06,
+      "loss": 0.2039,
+      "step": 11094
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.5477791176578537,
+      "learning_rate": 3.3756282289890785e-06,
+      "loss": 0.1763,
+      "step": 11095
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.3746284096547112,
+      "learning_rate": 3.3747762785766846e-06,
+      "loss": 0.1505,
+      "step": 11096
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.1042509616608391,
+      "learning_rate": 3.373924380919988e-06,
+      "loss": 0.2391,
+      "step": 11097
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.2781362029426397,
+      "learning_rate": 3.37307253604664e-06,
+      "loss": 0.1713,
+      "step": 11098
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.0867216791679786,
+      "learning_rate": 3.372220743984296e-06,
+      "loss": 0.1881,
+      "step": 11099
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.0334880017876913,
+      "learning_rate": 3.3713690047605994e-06,
+      "loss": 0.1493,
+      "step": 11100
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.523626081092145,
+      "learning_rate": 3.3705173184032035e-06,
+      "loss": 0.2186,
+      "step": 11101
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.318002408761823,
+      "learning_rate": 3.3696656849397514e-06,
+      "loss": 0.2181,
+      "step": 11102
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.129551538091805,
+      "learning_rate": 3.3688141043978905e-06,
+      "loss": 0.2117,
+      "step": 11103
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.2631618113016543,
+      "learning_rate": 3.3679625768052594e-06,
+      "loss": 0.168,
+      "step": 11104
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.05862767175751,
+      "learning_rate": 3.367111102189502e-06,
+      "loss": 0.2066,
+      "step": 11105
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.0153220513482486,
+      "learning_rate": 3.3662596805782572e-06,
+      "loss": 0.155,
+      "step": 11106
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.1604063891467067,
+      "learning_rate": 3.365408311999162e-06,
+      "loss": 0.1893,
+      "step": 11107
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.3898031117832859,
+      "learning_rate": 3.364556996479856e-06,
+      "loss": 0.1861,
+      "step": 11108
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.2430955219178015,
+      "learning_rate": 3.3637057340479674e-06,
+      "loss": 0.1977,
+      "step": 11109
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.8466201462098927,
+      "learning_rate": 3.3628545247311332e-06,
+      "loss": 0.1606,
+      "step": 11110
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.1381616447096528,
+      "learning_rate": 3.3620033685569814e-06,
+      "loss": 0.2044,
+      "step": 11111
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.4338741703857436,
+      "learning_rate": 3.3611522655531443e-06,
+      "loss": 0.2161,
+      "step": 11112
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.3467314656948917,
+      "learning_rate": 3.360301215747245e-06,
+      "loss": 0.185,
+      "step": 11113
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.0809893157014248,
+      "learning_rate": 3.359450219166911e-06,
+      "loss": 0.2166,
+      "step": 11114
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.4413867744949562,
+      "learning_rate": 3.3585992758397658e-06,
+      "loss": 0.2147,
+      "step": 11115
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.508764860582529,
+      "learning_rate": 3.3577483857934324e-06,
+      "loss": 0.1811,
+      "step": 11116
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.690599368001571,
+      "learning_rate": 3.3568975490555335e-06,
+      "loss": 0.2129,
+      "step": 11117
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.8550636255421247,
+      "learning_rate": 3.3560467656536823e-06,
+      "loss": 0.1895,
+      "step": 11118
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.3872164088479069,
+      "learning_rate": 3.355196035615499e-06,
+      "loss": 0.2348,
+      "step": 11119
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.3374823984654636,
+      "learning_rate": 3.3543453589685976e-06,
+      "loss": 0.1821,
+      "step": 11120
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.633076765039365,
+      "learning_rate": 3.3534947357405946e-06,
+      "loss": 0.1696,
+      "step": 11121
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.1225065274834674,
+      "learning_rate": 3.352644165959097e-06,
+      "loss": 0.1719,
+      "step": 11122
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.9144499316316598,
+      "learning_rate": 3.351793649651717e-06,
+      "loss": 0.1401,
+      "step": 11123
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.8845685187734524,
+      "learning_rate": 3.350943186846063e-06,
+      "loss": 0.1251,
+      "step": 11124
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.2718701048529524,
+      "learning_rate": 3.3500927775697413e-06,
+      "loss": 0.1743,
+      "step": 11125
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.9608657042501703,
+      "learning_rate": 3.3492424218503585e-06,
+      "loss": 0.1642,
+      "step": 11126
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.0207376668423458,
+      "learning_rate": 3.3483921197155145e-06,
+      "loss": 0.1738,
+      "step": 11127
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.1474411674581988,
+      "learning_rate": 3.3475418711928127e-06,
+      "loss": 0.1859,
+      "step": 11128
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.3010017276901982,
+      "learning_rate": 3.346691676309851e-06,
+      "loss": 0.2076,
+      "step": 11129
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.6967627150090223,
+      "learning_rate": 3.345841535094231e-06,
+      "loss": 0.2502,
+      "step": 11130
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.6493471514883002,
+      "learning_rate": 3.3449914475735423e-06,
+      "loss": 0.211,
+      "step": 11131
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.363636112754972,
+      "learning_rate": 3.344141413775384e-06,
+      "loss": 0.2072,
+      "step": 11132
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.8827161314362272,
+      "learning_rate": 3.343291433727348e-06,
+      "loss": 0.1451,
+      "step": 11133
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.8197246784173101,
+      "learning_rate": 3.3424415074570237e-06,
+      "loss": 0.1706,
+      "step": 11134
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.6252600525966292,
+      "learning_rate": 3.341591634992004e-06,
+      "loss": 0.2194,
+      "step": 11135
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.0632776612074784,
+      "learning_rate": 3.3407418163598714e-06,
+      "loss": 0.1884,
+      "step": 11136
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.4192417265739243,
+      "learning_rate": 3.3398920515882145e-06,
+      "loss": 0.1822,
+      "step": 11137
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.221247450965966,
+      "learning_rate": 3.3390423407046154e-06,
+      "loss": 0.1773,
+      "step": 11138
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.1874210714160172,
+      "learning_rate": 3.3381926837366585e-06,
+      "loss": 0.1666,
+      "step": 11139
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.110239925370867,
+      "learning_rate": 3.3373430807119212e-06,
+      "loss": 0.2373,
+      "step": 11140
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.3509711587115885,
+      "learning_rate": 3.3364935316579833e-06,
+      "loss": 0.1692,
+      "step": 11141
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.6667609098788296,
+      "learning_rate": 3.3356440366024225e-06,
+      "loss": 0.2333,
+      "step": 11142
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.5087404014863606,
+      "learning_rate": 3.334794595572812e-06,
+      "loss": 0.2054,
+      "step": 11143
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.9187613920394313,
+      "learning_rate": 3.3339452085967293e-06,
+      "loss": 0.1964,
+      "step": 11144
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.1150747275115303,
+      "learning_rate": 3.3330958757017417e-06,
+      "loss": 0.194,
+      "step": 11145
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.9383416629657142,
+      "learning_rate": 3.332246596915421e-06,
+      "loss": 0.2173,
+      "step": 11146
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.3758442573481444,
+      "learning_rate": 3.3313973722653337e-06,
+      "loss": 0.1909,
+      "step": 11147
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.9362646188791511,
+      "learning_rate": 3.3305482017790496e-06,
+      "loss": 0.1526,
+      "step": 11148
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.3009701797816715,
+      "learning_rate": 3.329699085484129e-06,
+      "loss": 0.213,
+      "step": 11149
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.5098613496465425,
+      "learning_rate": 3.3288500234081357e-06,
+      "loss": 0.2384,
+      "step": 11150
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.222450142998796,
+      "learning_rate": 3.3280010155786325e-06,
+      "loss": 0.1769,
+      "step": 11151
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.70049671041097,
+      "learning_rate": 3.3271520620231774e-06,
+      "loss": 0.2079,
+      "step": 11152
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.006624033002556,
+      "learning_rate": 3.3263031627693304e-06,
+      "loss": 0.1857,
+      "step": 11153
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.401417783827311,
+      "learning_rate": 3.325454317844643e-06,
+      "loss": 0.2187,
+      "step": 11154
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.4542575954857306,
+      "learning_rate": 3.324605527276673e-06,
+      "loss": 0.1622,
+      "step": 11155
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.2629752236091463,
+      "learning_rate": 3.323756791092969e-06,
+      "loss": 0.1815,
+      "step": 11156
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.5676299925226471,
+      "learning_rate": 3.322908109321086e-06,
+      "loss": 0.2206,
+      "step": 11157
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.3419456732590183,
+      "learning_rate": 3.3220594819885686e-06,
+      "loss": 0.2225,
+      "step": 11158
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.161495658961849,
+      "learning_rate": 3.3212109091229648e-06,
+      "loss": 0.1821,
+      "step": 11159
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.1492623396748942,
+      "learning_rate": 3.3203623907518203e-06,
+      "loss": 0.2373,
+      "step": 11160
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.1347722498163608,
+      "learning_rate": 3.3195139269026784e-06,
+      "loss": 0.204,
+      "step": 11161
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.3281480099984817,
+      "learning_rate": 3.318665517603083e-06,
+      "loss": 0.1552,
+      "step": 11162
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.954951130708656,
+      "learning_rate": 3.3178171628805695e-06,
+      "loss": 0.2067,
+      "step": 11163
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.1786252000879145,
+      "learning_rate": 3.3169688627626796e-06,
+      "loss": 0.1867,
+      "step": 11164
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.8500218080955699,
+      "learning_rate": 3.3161206172769468e-06,
+      "loss": 0.1642,
+      "step": 11165
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.4952715514345123,
+      "learning_rate": 3.3152724264509095e-06,
+      "loss": 0.1956,
+      "step": 11166
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.162153042846368,
+      "learning_rate": 3.3144242903120954e-06,
+      "loss": 0.1631,
+      "step": 11167
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.193378996180481,
+      "learning_rate": 3.3135762088880387e-06,
+      "loss": 0.2035,
+      "step": 11168
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.2736129464359618,
+      "learning_rate": 3.3127281822062674e-06,
+      "loss": 0.2103,
+      "step": 11169
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.1820944300858927,
+      "learning_rate": 3.3118802102943096e-06,
+      "loss": 0.1904,
+      "step": 11170
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.2392094078861382,
+      "learning_rate": 3.3110322931796933e-06,
+      "loss": 0.2332,
+      "step": 11171
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.9101610347310966,
+      "learning_rate": 3.3101844308899383e-06,
+      "loss": 0.1421,
+      "step": 11172
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.027801984406701,
+      "learning_rate": 3.309336623452569e-06,
+      "loss": 0.1909,
+      "step": 11173
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.2683380623293592,
+      "learning_rate": 3.308488870895104e-06,
+      "loss": 0.19,
+      "step": 11174
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.4228885386629768,
+      "learning_rate": 3.3076411732450653e-06,
+      "loss": 0.2121,
+      "step": 11175
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.1324060138277168,
+      "learning_rate": 3.3067935305299647e-06,
+      "loss": 0.1562,
+      "step": 11176
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.02436009893117,
+      "learning_rate": 3.3059459427773204e-06,
+      "loss": 0.2082,
+      "step": 11177
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.199540421278292,
+      "learning_rate": 3.3050984100146442e-06,
+      "loss": 0.1844,
+      "step": 11178
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.473192443675419,
+      "learning_rate": 3.3042509322694494e-06,
+      "loss": 0.2519,
+      "step": 11179
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.3790944878688864,
+      "learning_rate": 3.303403509569246e-06,
+      "loss": 0.1965,
+      "step": 11180
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.1385190936494922,
+      "learning_rate": 3.3025561419415377e-06,
+      "loss": 0.1521,
+      "step": 11181
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.243682543444446,
+      "learning_rate": 3.3017088294138343e-06,
+      "loss": 0.1692,
+      "step": 11182
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.1540071507879859,
+      "learning_rate": 3.3008615720136385e-06,
+      "loss": 0.2196,
+      "step": 11183
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.4661989885612905,
+      "learning_rate": 3.3000143697684546e-06,
+      "loss": 0.2243,
+      "step": 11184
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.8490300582160744,
+      "learning_rate": 3.2991672227057805e-06,
+      "loss": 0.1935,
+      "step": 11185
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.6491516518869105,
+      "learning_rate": 3.2983201308531153e-06,
+      "loss": 0.2442,
+      "step": 11186
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.0492178407780166,
+      "learning_rate": 3.297473094237957e-06,
+      "loss": 0.1986,
+      "step": 11187
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.2275527249276206,
+      "learning_rate": 3.2966261128878025e-06,
+      "loss": 0.1916,
+      "step": 11188
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.1656743896953448,
+      "learning_rate": 3.295779186830145e-06,
+      "loss": 0.1627,
+      "step": 11189
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.2759857045442755,
+      "learning_rate": 3.2949323160924735e-06,
+      "loss": 0.1934,
+      "step": 11190
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.1074370193118717,
+      "learning_rate": 3.2940855007022804e-06,
+      "loss": 0.1674,
+      "step": 11191
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.442432477608671,
+      "learning_rate": 3.2932387406870513e-06,
+      "loss": 0.1664,
+      "step": 11192
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.1623974963855217,
+      "learning_rate": 3.292392036074277e-06,
+      "loss": 0.2117,
+      "step": 11193
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.1433212341682322,
+      "learning_rate": 3.2915453868914372e-06,
+      "loss": 0.2077,
+      "step": 11194
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.5857926335435026,
+      "learning_rate": 3.290698793166016e-06,
+      "loss": 0.1869,
+      "step": 11195
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.236617288024054,
+      "learning_rate": 3.289852254925495e-06,
+      "loss": 0.2118,
+      "step": 11196
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.2147990016116665,
+      "learning_rate": 3.2890057721973535e-06,
+      "loss": 0.2413,
+      "step": 11197
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.120589046316382,
+      "learning_rate": 3.28815934500907e-06,
+      "loss": 0.169,
+      "step": 11198
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.3971051070334723,
+      "learning_rate": 3.287312973388117e-06,
+      "loss": 0.1752,
+      "step": 11199
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.350249755459237,
+      "learning_rate": 3.286466657361971e-06,
+      "loss": 0.1382,
+      "step": 11200
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.3245331382580505,
+      "learning_rate": 3.285620396958101e-06,
+      "loss": 0.1815,
+      "step": 11201
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.3246871828644435,
+      "learning_rate": 3.284774192203981e-06,
+      "loss": 0.19,
+      "step": 11202
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.469624938493075,
+      "learning_rate": 3.2839280431270746e-06,
+      "loss": 0.2364,
+      "step": 11203
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.9860468623934311,
+      "learning_rate": 3.2830819497548516e-06,
+      "loss": 0.1731,
+      "step": 11204
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.9721359446916118,
+      "learning_rate": 3.282235912114775e-06,
+      "loss": 0.1864,
+      "step": 11205
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.0183019644559177,
+      "learning_rate": 3.281389930234309e-06,
+      "loss": 0.1936,
+      "step": 11206
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.2642789381912072,
+      "learning_rate": 3.2805440041409164e-06,
+      "loss": 0.148,
+      "step": 11207
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.492938362063842,
+      "learning_rate": 3.2796981338620513e-06,
+      "loss": 0.2333,
+      "step": 11208
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.7184792454385883,
+      "learning_rate": 3.278852319425176e-06,
+      "loss": 0.1419,
+      "step": 11209
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.4587131743534865,
+      "learning_rate": 3.2780065608577427e-06,
+      "loss": 0.1381,
+      "step": 11210
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.2450492870291696,
+      "learning_rate": 3.2771608581872085e-06,
+      "loss": 0.1822,
+      "step": 11211
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.2184917423911008,
+      "learning_rate": 3.2763152114410214e-06,
+      "loss": 0.1862,
+      "step": 11212
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.2690802829565477,
+      "learning_rate": 3.275469620646634e-06,
+      "loss": 0.2083,
+      "step": 11213
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.4024026507703264,
+      "learning_rate": 3.274624085831495e-06,
+      "loss": 0.2069,
+      "step": 11214
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.9255892598788218,
+      "learning_rate": 3.2737786070230494e-06,
+      "loss": 0.1604,
+      "step": 11215
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.4431150614196744,
+      "learning_rate": 3.272933184248745e-06,
+      "loss": 0.1965,
+      "step": 11216
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.8408046049717712,
+      "learning_rate": 3.2720878175360204e-06,
+      "loss": 0.1861,
+      "step": 11217
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.9391935092889645,
+      "learning_rate": 3.2712425069123205e-06,
+      "loss": 0.206,
+      "step": 11218
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.9542971969772589,
+      "learning_rate": 3.270397252405082e-06,
+      "loss": 0.1982,
+      "step": 11219
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.573480243537923,
+      "learning_rate": 3.2695520540417447e-06,
+      "loss": 0.2237,
+      "step": 11220
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.3776304621671691,
+      "learning_rate": 3.2687069118497412e-06,
+      "loss": 0.1895,
+      "step": 11221
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.8441567699628216,
+      "learning_rate": 3.2678618258565066e-06,
+      "loss": 0.2441,
+      "step": 11222
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.23750204847147,
+      "learning_rate": 3.267016796089473e-06,
+      "loss": 0.2335,
+      "step": 11223
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.164621625300881,
+      "learning_rate": 3.266171822576071e-06,
+      "loss": 0.1982,
+      "step": 11224
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.3105269244362536,
+      "learning_rate": 3.265326905343731e-06,
+      "loss": 0.1931,
+      "step": 11225
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.9128076220979205,
+      "learning_rate": 3.264482044419875e-06,
+      "loss": 0.1777,
+      "step": 11226
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.163690648434893,
+      "learning_rate": 3.2636372398319304e-06,
+      "loss": 0.1723,
+      "step": 11227
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.493929249640604,
+      "learning_rate": 3.262792491607319e-06,
+      "loss": 0.2258,
+      "step": 11228
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.1020364620983858,
+      "learning_rate": 3.2619477997734617e-06,
+      "loss": 0.1485,
+      "step": 11229
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.2096831962776935,
+      "learning_rate": 3.2611031643577813e-06,
+      "loss": 0.0963,
+      "step": 11230
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.9301979723108063,
+      "learning_rate": 3.260258585387689e-06,
+      "loss": 0.1507,
+      "step": 11231
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.3473172715160047,
+      "learning_rate": 3.259414062890604e-06,
+      "loss": 0.2091,
+      "step": 11232
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.2982239658941552,
+      "learning_rate": 3.258569596893939e-06,
+      "loss": 0.1503,
+      "step": 11233
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.8225285829976625,
+      "learning_rate": 3.257725187425109e-06,
+      "loss": 0.2001,
+      "step": 11234
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.6696260156721914,
+      "learning_rate": 3.256880834511519e-06,
+      "loss": 0.1992,
+      "step": 11235
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.1973412444735494,
+      "learning_rate": 3.2560365381805802e-06,
+      "loss": 0.1627,
+      "step": 11236
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.4642957348929269,
+      "learning_rate": 3.2551922984596975e-06,
+      "loss": 0.1914,
+      "step": 11237
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.1805247783934378,
+      "learning_rate": 3.254348115376276e-06,
+      "loss": 0.2207,
+      "step": 11238
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.5339294996905883,
+      "learning_rate": 3.25350398895772e-06,
+      "loss": 0.2136,
+      "step": 11239
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.6038474751316552,
+      "learning_rate": 3.252659919231428e-06,
+      "loss": 0.2273,
+      "step": 11240
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.922715781582224,
+      "learning_rate": 3.251815906224799e-06,
+      "loss": 0.1714,
+      "step": 11241
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.0135232342353746,
+      "learning_rate": 3.2509719499652314e-06,
+      "loss": 0.1596,
+      "step": 11242
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.4133745701452587,
+      "learning_rate": 3.2501280504801215e-06,
+      "loss": 0.1981,
+      "step": 11243
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.1910683637423534,
+      "learning_rate": 3.24928420779686e-06,
+      "loss": 0.1698,
+      "step": 11244
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.29437675912499,
+      "learning_rate": 3.24844042194284e-06,
+      "loss": 0.2374,
+      "step": 11245
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.875032754453031,
+      "learning_rate": 3.2475966929454505e-06,
+      "loss": 0.208,
+      "step": 11246
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 4.529279050014701,
+      "learning_rate": 3.246753020832081e-06,
+      "loss": 0.1674,
+      "step": 11247
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.9799265483169047,
+      "learning_rate": 3.245909405630118e-06,
+      "loss": 0.1775,
+      "step": 11248
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.1470711910556863,
+      "learning_rate": 3.245065847366942e-06,
+      "loss": 0.1989,
+      "step": 11249
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.0306770691115106,
+      "learning_rate": 3.2442223460699376e-06,
+      "loss": 0.2045,
+      "step": 11250
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.9881476511762842,
+      "learning_rate": 3.2433789017664854e-06,
+      "loss": 0.1983,
+      "step": 11251
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.8625319711661366,
+      "learning_rate": 3.2425355144839664e-06,
+      "loss": 0.0852,
+      "step": 11252
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.2181158494483317,
+      "learning_rate": 3.241692184249754e-06,
+      "loss": 0.1773,
+      "step": 11253
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.1748467134234637,
+      "learning_rate": 3.2408489110912246e-06,
+      "loss": 0.221,
+      "step": 11254
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.0734265510133951,
+      "learning_rate": 3.2400056950357506e-06,
+      "loss": 0.1885,
+      "step": 11255
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.0212109657523687,
+      "learning_rate": 3.2391625361107033e-06,
+      "loss": 0.1523,
+      "step": 11256
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.0652744187066598,
+      "learning_rate": 3.238319434343455e-06,
+      "loss": 0.1987,
+      "step": 11257
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.1188165435101023,
+      "learning_rate": 3.237476389761368e-06,
+      "loss": 0.1448,
+      "step": 11258
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.114283055747543,
+      "learning_rate": 3.2366334023918123e-06,
+      "loss": 0.1834,
+      "step": 11259
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.6935026730774516,
+      "learning_rate": 3.2357904722621492e-06,
+      "loss": 0.1967,
+      "step": 11260
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.1715324441601034,
+      "learning_rate": 3.2349475993997446e-06,
+      "loss": 0.193,
+      "step": 11261
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.3271079544955349,
+      "learning_rate": 3.234104783831954e-06,
+      "loss": 0.2225,
+      "step": 11262
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.9916566307130372,
+      "learning_rate": 3.233262025586138e-06,
+      "loss": 0.2114,
+      "step": 11263
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.372171918154612,
+      "learning_rate": 3.2324193246896517e-06,
+      "loss": 0.2227,
+      "step": 11264
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.2326878443549545,
+      "learning_rate": 3.2315766811698514e-06,
+      "loss": 0.1971,
+      "step": 11265
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.6524579795509922,
+      "learning_rate": 3.2307340950540898e-06,
+      "loss": 0.258,
+      "step": 11266
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.2650534198624148,
+      "learning_rate": 3.2298915663697154e-06,
+      "loss": 0.2088,
+      "step": 11267
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.270185052805809,
+      "learning_rate": 3.229049095144079e-06,
+      "loss": 0.2228,
+      "step": 11268
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.2139997709115393,
+      "learning_rate": 3.2282066814045267e-06,
+      "loss": 0.1758,
+      "step": 11269
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.5128413475278997,
+      "learning_rate": 3.2273643251784066e-06,
+      "loss": 0.2058,
+      "step": 11270
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.1484835771153323,
+      "learning_rate": 3.2265220264930575e-06,
+      "loss": 0.1354,
+      "step": 11271
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.98205310354142,
+      "learning_rate": 3.225679785375825e-06,
+      "loss": 0.176,
+      "step": 11272
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.5332540382144388,
+      "learning_rate": 3.224837601854045e-06,
+      "loss": 0.1948,
+      "step": 11273
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.097808757721853,
+      "learning_rate": 3.2239954759550586e-06,
+      "loss": 0.2129,
+      "step": 11274
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.1285133040930153,
+      "learning_rate": 3.223153407706201e-06,
+      "loss": 0.193,
+      "step": 11275
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.864969355034838,
+      "learning_rate": 3.2223113971348042e-06,
+      "loss": 0.2095,
+      "step": 11276
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.9933867359630962,
+      "learning_rate": 3.2214694442682017e-06,
+      "loss": 0.1755,
+      "step": 11277
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.5322343961030407,
+      "learning_rate": 3.2206275491337236e-06,
+      "loss": 0.1826,
+      "step": 11278
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.154038398706899,
+      "learning_rate": 3.2197857117587006e-06,
+      "loss": 0.1816,
+      "step": 11279
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.110040716780891,
+      "learning_rate": 3.218943932170455e-06,
+      "loss": 0.2402,
+      "step": 11280
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.1938424763635391,
+      "learning_rate": 3.2181022103963143e-06,
+      "loss": 0.1979,
+      "step": 11281
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.377514252736042,
+      "learning_rate": 3.2172605464635997e-06,
+      "loss": 0.2076,
+      "step": 11282
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.1409330278725325,
+      "learning_rate": 3.2164189403996326e-06,
+      "loss": 0.1912,
+      "step": 11283
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.0962998215322752,
+      "learning_rate": 3.215577392231735e-06,
+      "loss": 0.1875,
+      "step": 11284
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.214044906238344,
+      "learning_rate": 3.2147359019872194e-06,
+      "loss": 0.1381,
+      "step": 11285
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.3875573148037306,
+      "learning_rate": 3.2138944696934028e-06,
+      "loss": 0.2088,
+      "step": 11286
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.0661513299836043,
+      "learning_rate": 3.2130530953775984e-06,
+      "loss": 0.1992,
+      "step": 11287
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.184279504413699,
+      "learning_rate": 3.21221177906712e-06,
+      "loss": 0.2153,
+      "step": 11288
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.0560517512302345,
+      "learning_rate": 3.2113705207892743e-06,
+      "loss": 0.1826,
+      "step": 11289
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.219059564021706,
+      "learning_rate": 3.21052932057137e-06,
+      "loss": 0.2428,
+      "step": 11290
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.9876534160216688,
+      "learning_rate": 3.2096881784407126e-06,
+      "loss": 0.1937,
+      "step": 11291
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.210146328668085,
+      "learning_rate": 3.2088470944246062e-06,
+      "loss": 0.1901,
+      "step": 11292
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.2032112517960794,
+      "learning_rate": 3.208006068550355e-06,
+      "loss": 0.216,
+      "step": 11293
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.2361099267565783,
+      "learning_rate": 3.2071651008452555e-06,
+      "loss": 0.1908,
+      "step": 11294
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.3482627899682464,
+      "learning_rate": 3.2063241913366074e-06,
+      "loss": 0.2499,
+      "step": 11295
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.6095403475217767,
+      "learning_rate": 3.2054833400517077e-06,
+      "loss": 0.1883,
+      "step": 11296
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.4257632659404975,
+      "learning_rate": 3.2046425470178534e-06,
+      "loss": 0.1943,
+      "step": 11297
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.352906248193321,
+      "learning_rate": 3.2038018122623315e-06,
+      "loss": 0.2603,
+      "step": 11298
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.240981713354868,
+      "learning_rate": 3.202961135812437e-06,
+      "loss": 0.2021,
+      "step": 11299
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.8679041049551537,
+      "learning_rate": 3.202120517695456e-06,
+      "loss": 0.1833,
+      "step": 11300
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.5006374181982731,
+      "learning_rate": 3.201279957938677e-06,
+      "loss": 0.2506,
+      "step": 11301
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.1557961552674638,
+      "learning_rate": 3.2004394565693877e-06,
+      "loss": 0.1059,
+      "step": 11302
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.6407078676639741,
+      "learning_rate": 3.199599013614866e-06,
+      "loss": 0.2073,
+      "step": 11303
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.2763724673569858,
+      "learning_rate": 3.198758629102395e-06,
+      "loss": 0.2079,
+      "step": 11304
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.182039133765434,
+      "learning_rate": 3.1979183030592554e-06,
+      "loss": 0.2146,
+      "step": 11305
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.9088983233563445,
+      "learning_rate": 3.1970780355127255e-06,
+      "loss": 0.1549,
+      "step": 11306
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.2395664308768677,
+      "learning_rate": 3.1962378264900774e-06,
+      "loss": 0.1691,
+      "step": 11307
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.3480197775545557,
+      "learning_rate": 3.1953976760185877e-06,
+      "loss": 0.2367,
+      "step": 11308
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.0313408118057805,
+      "learning_rate": 3.1945575841255266e-06,
+      "loss": 0.2012,
+      "step": 11309
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.2687996003128588,
+      "learning_rate": 3.1937175508381645e-06,
+      "loss": 0.1844,
+      "step": 11310
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.2796618836322737,
+      "learning_rate": 3.1928775761837705e-06,
+      "loss": 0.193,
+      "step": 11311
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.4099231987058825,
+      "learning_rate": 3.1920376601896087e-06,
+      "loss": 0.2075,
+      "step": 11312
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.04374204678275,
+      "learning_rate": 3.1911978028829432e-06,
+      "loss": 0.1992,
+      "step": 11313
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.8320375088582768,
+      "learning_rate": 3.1903580042910377e-06,
+      "loss": 0.2906,
+      "step": 11314
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.2036864314570468,
+      "learning_rate": 3.1895182644411537e-06,
+      "loss": 0.2066,
+      "step": 11315
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.3542642545993857,
+      "learning_rate": 3.188678583360546e-06,
+      "loss": 0.235,
+      "step": 11316
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.744581194656819,
+      "learning_rate": 3.1878389610764743e-06,
+      "loss": 0.2168,
+      "step": 11317
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.1573259528249062,
+      "learning_rate": 3.1869993976161906e-06,
+      "loss": 0.2558,
+      "step": 11318
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.2106684462762378,
+      "learning_rate": 3.1861598930069492e-06,
+      "loss": 0.193,
+      "step": 11319
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.0128176513171292,
+      "learning_rate": 3.185320447276003e-06,
+      "loss": 0.2055,
+      "step": 11320
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.4530972739620167,
+      "learning_rate": 3.184481060450596e-06,
+      "loss": 0.1925,
+      "step": 11321
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.0825528392218633,
+      "learning_rate": 3.1836417325579784e-06,
+      "loss": 0.1848,
+      "step": 11322
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.3981529777534434,
+      "learning_rate": 3.182802463625394e-06,
+      "loss": 0.1951,
+      "step": 11323
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.1704680691468694,
+      "learning_rate": 3.1819632536800893e-06,
+      "loss": 0.1827,
+      "step": 11324
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.2186592355578056,
+      "learning_rate": 3.1811241027493e-06,
+      "loss": 0.2127,
+      "step": 11325
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.038962664146084,
+      "learning_rate": 3.1802850108602703e-06,
+      "loss": 0.1932,
+      "step": 11326
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.2538232003395202,
+      "learning_rate": 3.1794459780402338e-06,
+      "loss": 0.2546,
+      "step": 11327
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.4070824278611214,
+      "learning_rate": 3.1786070043164273e-06,
+      "loss": 0.2401,
+      "step": 11328
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.5657167147757625,
+      "learning_rate": 3.1777680897160876e-06,
+      "loss": 0.1677,
+      "step": 11329
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.9073572434255814,
+      "learning_rate": 3.1769292342664414e-06,
+      "loss": 0.1886,
+      "step": 11330
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.108709666379727,
+      "learning_rate": 3.176090437994721e-06,
+      "loss": 0.2142,
+      "step": 11331
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.9453239046149724,
+      "learning_rate": 3.175251700928154e-06,
+      "loss": 0.2142,
+      "step": 11332
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.138701337102746,
+      "learning_rate": 3.174413023093967e-06,
+      "loss": 0.1894,
+      "step": 11333
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.4648044021408002,
+      "learning_rate": 3.173574404519381e-06,
+      "loss": 0.2245,
+      "step": 11334
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.2862577906243824,
+      "learning_rate": 3.172735845231622e-06,
+      "loss": 0.2158,
+      "step": 11335
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.1280867694283803,
+      "learning_rate": 3.171897345257907e-06,
+      "loss": 0.1699,
+      "step": 11336
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.182937737704471,
+      "learning_rate": 3.1710589046254546e-06,
+      "loss": 0.1874,
+      "step": 11337
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.0217863843565838,
+      "learning_rate": 3.1702205233614846e-06,
+      "loss": 0.1836,
+      "step": 11338
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.1598108239776168,
+      "learning_rate": 3.1693822014932068e-06,
+      "loss": 0.1521,
+      "step": 11339
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.2617659456620332,
+      "learning_rate": 3.168543939047836e-06,
+      "loss": 0.1771,
+      "step": 11340
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.073996864291183,
+      "learning_rate": 3.1677057360525815e-06,
+      "loss": 0.1275,
+      "step": 11341
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.486918388589899,
+      "learning_rate": 3.166867592534653e-06,
+      "loss": 0.2217,
+      "step": 11342
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.3048550104206489,
+      "learning_rate": 3.166029508521256e-06,
+      "loss": 0.2501,
+      "step": 11343
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.3090023370806445,
+      "learning_rate": 3.165191484039597e-06,
+      "loss": 0.1601,
+      "step": 11344
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.9219245816491304,
+      "learning_rate": 3.1643535191168755e-06,
+      "loss": 0.1474,
+      "step": 11345
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.444456269550011,
+      "learning_rate": 3.163515613780295e-06,
+      "loss": 0.1812,
+      "step": 11346
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.0977034982850025,
+      "learning_rate": 3.1626777680570558e-06,
+      "loss": 0.1773,
+      "step": 11347
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.6037232035793956,
+      "learning_rate": 3.1618399819743505e-06,
+      "loss": 0.1916,
+      "step": 11348
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.3683598315003789,
+      "learning_rate": 3.161002255559377e-06,
+      "loss": 0.1673,
+      "step": 11349
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.0667471498000278,
+      "learning_rate": 3.160164588839328e-06,
+      "loss": 0.1917,
+      "step": 11350
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.9411753414821996,
+      "learning_rate": 3.1593269818413946e-06,
+      "loss": 0.1474,
+      "step": 11351
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.8855499513199945,
+      "learning_rate": 3.1584894345927663e-06,
+      "loss": 0.1614,
+      "step": 11352
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.0410804835686143,
+      "learning_rate": 3.157651947120628e-06,
+      "loss": 0.1528,
+      "step": 11353
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.961991205365592,
+      "learning_rate": 3.156814519452168e-06,
+      "loss": 0.1302,
+      "step": 11354
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.3181693992783672,
+      "learning_rate": 3.1559771516145675e-06,
+      "loss": 0.1926,
+      "step": 11355
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.1852121304770655,
+      "learning_rate": 3.1551398436350124e-06,
+      "loss": 0.1628,
+      "step": 11356
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.4857431932518401,
+      "learning_rate": 3.1543025955406758e-06,
+      "loss": 0.1928,
+      "step": 11357
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.225535242012468,
+      "learning_rate": 3.153465407358739e-06,
+      "loss": 0.238,
+      "step": 11358
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.976797549780528,
+      "learning_rate": 3.152628279116377e-06,
+      "loss": 0.1974,
+      "step": 11359
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.525388348879319,
+      "learning_rate": 3.1517912108407638e-06,
+      "loss": 0.1975,
+      "step": 11360
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.9361108362271443,
+      "learning_rate": 3.150954202559071e-06,
+      "loss": 0.1678,
+      "step": 11361
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.4087844090693602,
+      "learning_rate": 3.150117254298467e-06,
+      "loss": 0.2028,
+      "step": 11362
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.1201895413564513,
+      "learning_rate": 3.1492803660861206e-06,
+      "loss": 0.1788,
+      "step": 11363
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.0816395253637034,
+      "learning_rate": 3.1484435379491975e-06,
+      "loss": 0.1911,
+      "step": 11364
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.9919645629689655,
+      "learning_rate": 3.147606769914864e-06,
+      "loss": 0.2009,
+      "step": 11365
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.123844695938663,
+      "learning_rate": 3.146770062010278e-06,
+      "loss": 0.132,
+      "step": 11366
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.672786049971332,
+      "learning_rate": 3.1459334142626015e-06,
+      "loss": 0.2311,
+      "step": 11367
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.432667768400679,
+      "learning_rate": 3.1450968266989935e-06,
+      "loss": 0.1544,
+      "step": 11368
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.2699543614663276,
+      "learning_rate": 3.144260299346609e-06,
+      "loss": 0.2008,
+      "step": 11369
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.8351002732140138,
+      "learning_rate": 3.1434238322326028e-06,
+      "loss": 0.2254,
+      "step": 11370
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.2138594383919223,
+      "learning_rate": 3.142587425384125e-06,
+      "loss": 0.2034,
+      "step": 11371
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.366290490299967,
+      "learning_rate": 3.141751078828328e-06,
+      "loss": 0.2126,
+      "step": 11372
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.212857668270719,
+      "learning_rate": 3.1409147925923587e-06,
+      "loss": 0.1387,
+      "step": 11373
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.1909384168532764,
+      "learning_rate": 3.1400785667033675e-06,
+      "loss": 0.171,
+      "step": 11374
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.0001181555026155,
+      "learning_rate": 3.1392424011884927e-06,
+      "loss": 0.1718,
+      "step": 11375
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.6155741632863942,
+      "learning_rate": 3.1384062960748797e-06,
+      "loss": 0.2424,
+      "step": 11376
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.5200074756314568,
+      "learning_rate": 3.13757025138967e-06,
+      "loss": 0.1814,
+      "step": 11377
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.4518344336995164,
+      "learning_rate": 3.136734267160001e-06,
+      "loss": 0.2299,
+      "step": 11378
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.7050273271577312,
+      "learning_rate": 3.1358983434130085e-06,
+      "loss": 0.1997,
+      "step": 11379
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.2034953959150727,
+      "learning_rate": 3.1350624801758276e-06,
+      "loss": 0.2033,
+      "step": 11380
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.4559373405343463,
+      "learning_rate": 3.1342266774755905e-06,
+      "loss": 0.2404,
+      "step": 11381
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.336515969246225,
+      "learning_rate": 3.1333909353394287e-06,
+      "loss": 0.1797,
+      "step": 11382
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.4011566243167257,
+      "learning_rate": 3.1325552537944725e-06,
+      "loss": 0.1731,
+      "step": 11383
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.2064547257552536,
+      "learning_rate": 3.1317196328678438e-06,
+      "loss": 0.1837,
+      "step": 11384
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.562715553674197,
+      "learning_rate": 3.130884072586671e-06,
+      "loss": 0.2384,
+      "step": 11385
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.4360902445583001,
+      "learning_rate": 3.1300485729780767e-06,
+      "loss": 0.174,
+      "step": 11386
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.845391338999564,
+      "learning_rate": 3.129213134069181e-06,
+      "loss": 0.2298,
+      "step": 11387
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.0577309284003311,
+      "learning_rate": 3.128377755887103e-06,
+      "loss": 0.1598,
+      "step": 11388
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.9670347017555436,
+      "learning_rate": 3.127542438458958e-06,
+      "loss": 0.1958,
+      "step": 11389
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.2690820544844457,
+      "learning_rate": 3.126707181811862e-06,
+      "loss": 0.1953,
+      "step": 11390
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.3008153383983476,
+      "learning_rate": 3.125871985972929e-06,
+      "loss": 0.1679,
+      "step": 11391
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.4733113266207916,
+      "learning_rate": 3.12503685096927e-06,
+      "loss": 0.1791,
+      "step": 11392
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.6791253102757415,
+      "learning_rate": 3.124201776827992e-06,
+      "loss": 0.1805,
+      "step": 11393
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.9887371067334976,
+      "learning_rate": 3.123366763576202e-06,
+      "loss": 0.198,
+      "step": 11394
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.6551156883127622,
+      "learning_rate": 3.1225318112410076e-06,
+      "loss": 0.225,
+      "step": 11395
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.315853648126555,
+      "learning_rate": 3.12169691984951e-06,
+      "loss": 0.1877,
+      "step": 11396
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.118281998724032,
+      "learning_rate": 3.1208620894288105e-06,
+      "loss": 0.1986,
+      "step": 11397
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.9866208598763968,
+      "learning_rate": 3.1200273200060074e-06,
+      "loss": 0.2453,
+      "step": 11398
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.0159232600289825,
+      "learning_rate": 3.119192611608198e-06,
+      "loss": 0.2284,
+      "step": 11399
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.3448013085429136,
+      "learning_rate": 3.118357964262478e-06,
+      "loss": 0.214,
+      "step": 11400
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.5835029538382708,
+      "learning_rate": 3.1175233779959437e-06,
+      "loss": 0.1971,
+      "step": 11401
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.2884190600766734,
+      "learning_rate": 3.1166888528356797e-06,
+      "loss": 0.2223,
+      "step": 11402
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.83167866158687,
+      "learning_rate": 3.115854388808779e-06,
+      "loss": 0.1965,
+      "step": 11403
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.4111389344818477,
+      "learning_rate": 3.115019985942329e-06,
+      "loss": 0.2156,
+      "step": 11404
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.2687771949296436,
+      "learning_rate": 3.114185644263415e-06,
+      "loss": 0.1985,
+      "step": 11405
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.4344938633065514,
+      "learning_rate": 3.113351363799119e-06,
+      "loss": 0.1897,
+      "step": 11406
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.3838402597975032,
+      "learning_rate": 3.1125171445765213e-06,
+      "loss": 0.1869,
+      "step": 11407
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.094772644162024,
+      "learning_rate": 3.1116829866227027e-06,
+      "loss": 0.2136,
+      "step": 11408
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.6849093043591161,
+      "learning_rate": 3.1108488899647406e-06,
+      "loss": 0.2131,
+      "step": 11409
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.2399069286226994,
+      "learning_rate": 3.1100148546297116e-06,
+      "loss": 0.1895,
+      "step": 11410
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.437793836416532,
+      "learning_rate": 3.1091808806446847e-06,
+      "loss": 0.2699,
+      "step": 11411
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.7457087474227986,
+      "learning_rate": 3.1083469680367338e-06,
+      "loss": 0.222,
+      "step": 11412
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.4986659651984138,
+      "learning_rate": 3.10751311683293e-06,
+      "loss": 0.2019,
+      "step": 11413
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.2832141161582917,
+      "learning_rate": 3.106679327060338e-06,
+      "loss": 0.1899,
+      "step": 11414
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.2090987365569348,
+      "learning_rate": 3.105845598746024e-06,
+      "loss": 0.144,
+      "step": 11415
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.2756160198093691,
+      "learning_rate": 3.10501193191705e-06,
+      "loss": 0.176,
+      "step": 11416
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.136211314775556,
+      "learning_rate": 3.1041783266004784e-06,
+      "loss": 0.1741,
+      "step": 11417
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.2308448035207844,
+      "learning_rate": 3.1033447828233686e-06,
+      "loss": 0.1867,
+      "step": 11418
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.386221974001847,
+      "learning_rate": 3.1025113006127795e-06,
+      "loss": 0.2226,
+      "step": 11419
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.4324964980844896,
+      "learning_rate": 3.1016778799957624e-06,
+      "loss": 0.1575,
+      "step": 11420
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.2892694480543154,
+      "learning_rate": 3.1008445209993737e-06,
+      "loss": 0.2302,
+      "step": 11421
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.2014455685065053,
+      "learning_rate": 3.1000112236506646e-06,
+      "loss": 0.1606,
+      "step": 11422
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.0790424745012286,
+      "learning_rate": 3.099177987976684e-06,
+      "loss": 0.1845,
+      "step": 11423
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.4541157298425882,
+      "learning_rate": 3.0983448140044785e-06,
+      "loss": 0.1811,
+      "step": 11424
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.006471534928242,
+      "learning_rate": 3.0975117017610934e-06,
+      "loss": 0.2233,
+      "step": 11425
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.2609565723966856,
+      "learning_rate": 3.0966786512735727e-06,
+      "loss": 0.2127,
+      "step": 11426
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.0115454315428387,
+      "learning_rate": 3.0958456625689566e-06,
+      "loss": 0.1801,
+      "step": 11427
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.1251139020214591,
+      "learning_rate": 3.0950127356742875e-06,
+      "loss": 0.1783,
+      "step": 11428
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.2529155436049741,
+      "learning_rate": 3.094179870616598e-06,
+      "loss": 0.1466,
+      "step": 11429
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.752436168340772,
+      "learning_rate": 3.093347067422927e-06,
+      "loss": 0.2266,
+      "step": 11430
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.2783265663734202,
+      "learning_rate": 3.092514326120307e-06,
+      "loss": 0.1964,
+      "step": 11431
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.975997605019654,
+      "learning_rate": 3.091681646735769e-06,
+      "loss": 0.1252,
+      "step": 11432
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.079007287028355,
+      "learning_rate": 3.0908490292963417e-06,
+      "loss": 0.1683,
+      "step": 11433
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.2903714296895181,
+      "learning_rate": 3.0900164738290517e-06,
+      "loss": 0.2131,
+      "step": 11434
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.1756521356859002,
+      "learning_rate": 3.0891839803609257e-06,
+      "loss": 0.2136,
+      "step": 11435
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 6.841340777510882,
+      "learning_rate": 3.0883515489189862e-06,
+      "loss": 0.1433,
+      "step": 11436
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.0754181841371913,
+      "learning_rate": 3.0875191795302573e-06,
+      "loss": 0.1711,
+      "step": 11437
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.3298078365987305,
+      "learning_rate": 3.0866868722217535e-06,
+      "loss": 0.1996,
+      "step": 11438
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.5159578744356221,
+      "learning_rate": 3.0858546270204938e-06,
+      "loss": 0.1985,
+      "step": 11439
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.2934896854528342,
+      "learning_rate": 3.0850224439534943e-06,
+      "loss": 0.2292,
+      "step": 11440
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.1918159093246417,
+      "learning_rate": 3.0841903230477678e-06,
+      "loss": 0.2035,
+      "step": 11441
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.2703494736380467,
+      "learning_rate": 3.083358264330325e-06,
+      "loss": 0.2167,
+      "step": 11442
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.8393508041101319,
+      "learning_rate": 3.0825262678281744e-06,
+      "loss": 0.144,
+      "step": 11443
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.1183111901917042,
+      "learning_rate": 3.081694333568323e-06,
+      "loss": 0.1958,
+      "step": 11444
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.481679728097848,
+      "learning_rate": 3.080862461577777e-06,
+      "loss": 0.1976,
+      "step": 11445
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.2853887747650008,
+      "learning_rate": 3.080030651883541e-06,
+      "loss": 0.191,
+      "step": 11446
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.547805961153762,
+      "learning_rate": 3.0791989045126124e-06,
+      "loss": 0.2013,
+      "step": 11447
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.3465916837910252,
+      "learning_rate": 3.0783672194919917e-06,
+      "loss": 0.1709,
+      "step": 11448
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.3232756944756636,
+      "learning_rate": 3.0775355968486776e-06,
+      "loss": 0.1787,
+      "step": 11449
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.319863401022581,
+      "learning_rate": 3.076704036609663e-06,
+      "loss": 0.1926,
+      "step": 11450
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.81263487870086,
+      "learning_rate": 3.0758725388019413e-06,
+      "loss": 0.1515,
+      "step": 11451
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.513377745569388,
+      "learning_rate": 3.0750411034525023e-06,
+      "loss": 0.1981,
+      "step": 11452
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.0735861007691083,
+      "learning_rate": 3.074209730588337e-06,
+      "loss": 0.1447,
+      "step": 11453
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.5492995892708163,
+      "learning_rate": 3.0733784202364303e-06,
+      "loss": 0.1896,
+      "step": 11454
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.2660679130502508,
+      "learning_rate": 3.07254717242377e-06,
+      "loss": 0.1881,
+      "step": 11455
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.123308652139499,
+      "learning_rate": 3.071715987177335e-06,
+      "loss": 0.2359,
+      "step": 11456
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.497646666263806,
+      "learning_rate": 3.0708848645241074e-06,
+      "loss": 0.2123,
+      "step": 11457
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.210607742202318,
+      "learning_rate": 3.0700538044910684e-06,
+      "loss": 0.2329,
+      "step": 11458
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.1032864746236926,
+      "learning_rate": 3.069222807105192e-06,
+      "loss": 0.1596,
+      "step": 11459
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.367608732999466,
+      "learning_rate": 3.0683918723934535e-06,
+      "loss": 0.1318,
+      "step": 11460
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.9053738815072132,
+      "learning_rate": 3.0675610003828255e-06,
+      "loss": 0.1605,
+      "step": 11461
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.273604633620354,
+      "learning_rate": 3.066730191100278e-06,
+      "loss": 0.212,
+      "step": 11462
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.4221772538128048,
+      "learning_rate": 3.06589944457278e-06,
+      "loss": 0.1606,
+      "step": 11463
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.9487736820229333,
+      "learning_rate": 3.0650687608273007e-06,
+      "loss": 0.1648,
+      "step": 11464
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.998384489523724,
+      "learning_rate": 3.0642381398907993e-06,
+      "loss": 0.209,
+      "step": 11465
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.3661826391881848,
+      "learning_rate": 3.063407581790242e-06,
+      "loss": 0.1306,
+      "step": 11466
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.0047927979245357,
+      "learning_rate": 3.062577086552589e-06,
+      "loss": 0.1581,
+      "step": 11467
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.5024340818150372,
+      "learning_rate": 3.0617466542047976e-06,
+      "loss": 0.2067,
+      "step": 11468
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.3076007804952368,
+      "learning_rate": 3.0609162847738243e-06,
+      "loss": 0.216,
+      "step": 11469
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.3603481886932096,
+      "learning_rate": 3.0600859782866227e-06,
+      "loss": 0.2044,
+      "step": 11470
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.0709204374255923,
+      "learning_rate": 3.059255734770146e-06,
+      "loss": 0.1592,
+      "step": 11471
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.1726892948259902,
+      "learning_rate": 3.0584255542513434e-06,
+      "loss": 0.1705,
+      "step": 11472
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.0904732171576745,
+      "learning_rate": 3.0575954367571663e-06,
+      "loss": 0.1854,
+      "step": 11473
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.0900610361924379,
+      "learning_rate": 3.056765382314556e-06,
+      "loss": 0.2349,
+      "step": 11474
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.1474861564299832,
+      "learning_rate": 3.0559353909504596e-06,
+      "loss": 0.1522,
+      "step": 11475
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.1400331993403199,
+      "learning_rate": 3.0551054626918187e-06,
+      "loss": 0.1515,
+      "step": 11476
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.191806480534047,
+      "learning_rate": 3.0542755975655715e-06,
+      "loss": 0.1825,
+      "step": 11477
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.0510145676395646,
+      "learning_rate": 3.0534457955986597e-06,
+      "loss": 0.1491,
+      "step": 11478
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.849650568840731,
+      "learning_rate": 3.0526160568180153e-06,
+      "loss": 0.1858,
+      "step": 11479
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.3438742602359333,
+      "learning_rate": 3.051786381250573e-06,
+      "loss": 0.1402,
+      "step": 11480
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.3809499589487935,
+      "learning_rate": 3.0509567689232646e-06,
+      "loss": 0.2001,
+      "step": 11481
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.3094936408143032,
+      "learning_rate": 3.050127219863023e-06,
+      "loss": 0.1922,
+      "step": 11482
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.1215634043454201,
+      "learning_rate": 3.0492977340967703e-06,
+      "loss": 0.1852,
+      "step": 11483
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.289794823673439,
+      "learning_rate": 3.0484683116514357e-06,
+      "loss": 0.1973,
+      "step": 11484
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.1741099281164036,
+      "learning_rate": 3.0476389525539423e-06,
+      "loss": 0.2023,
+      "step": 11485
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.1966888164647105,
+      "learning_rate": 3.0468096568312105e-06,
+      "loss": 0.1661,
+      "step": 11486
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.302023228847563,
+      "learning_rate": 3.0459804245101627e-06,
+      "loss": 0.1429,
+      "step": 11487
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.820016771451472,
+      "learning_rate": 3.0451512556177115e-06,
+      "loss": 0.2227,
+      "step": 11488
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.0860733592172178,
+      "learning_rate": 3.044322150180775e-06,
+      "loss": 0.1698,
+      "step": 11489
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.2063735447751212,
+      "learning_rate": 3.0434931082262655e-06,
+      "loss": 0.157,
+      "step": 11490
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.3839374481522972,
+      "learning_rate": 3.0426641297810977e-06,
+      "loss": 0.199,
+      "step": 11491
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.696366887566579,
+      "learning_rate": 3.0418352148721753e-06,
+      "loss": 0.2593,
+      "step": 11492
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.2913911925313235,
+      "learning_rate": 3.0410063635264073e-06,
+      "loss": 0.1763,
+      "step": 11493
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.692202691091894,
+      "learning_rate": 3.0401775757707007e-06,
+      "loss": 0.2127,
+      "step": 11494
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.1190323236333481,
+      "learning_rate": 3.039348851631956e-06,
+      "loss": 0.2056,
+      "step": 11495
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.9039114013103067,
+      "learning_rate": 3.0385201911370778e-06,
+      "loss": 0.1741,
+      "step": 11496
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.184222664887626,
+      "learning_rate": 3.037691594312959e-06,
+      "loss": 0.1886,
+      "step": 11497
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.3491573984657552,
+      "learning_rate": 3.0368630611864995e-06,
+      "loss": 0.2115,
+      "step": 11498
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.4315309625652095,
+      "learning_rate": 3.0360345917845944e-06,
+      "loss": 0.2578,
+      "step": 11499
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.3763379718163886,
+      "learning_rate": 3.0352061861341374e-06,
+      "loss": 0.2361,
+      "step": 11500
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.015689785431597,
+      "learning_rate": 3.0343778442620163e-06,
+      "loss": 0.1995,
+      "step": 11501
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.4326832761908201,
+      "learning_rate": 3.03354956619512e-06,
+      "loss": 0.1757,
+      "step": 11502
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.9305900962633221,
+      "learning_rate": 3.0327213519603366e-06,
+      "loss": 0.1312,
+      "step": 11503
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.488046973104229,
+      "learning_rate": 3.031893201584548e-06,
+      "loss": 0.1713,
+      "step": 11504
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.2740613398103389,
+      "learning_rate": 3.0310651150946405e-06,
+      "loss": 0.2342,
+      "step": 11505
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.9345708426510894,
+      "learning_rate": 3.0302370925174894e-06,
+      "loss": 0.1442,
+      "step": 11506
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.162208652262819,
+      "learning_rate": 3.029409133879975e-06,
+      "loss": 0.1503,
+      "step": 11507
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.0105303318787584,
+      "learning_rate": 3.028581239208973e-06,
+      "loss": 0.2902,
+      "step": 11508
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.9303392702139406,
+      "learning_rate": 3.0277534085313603e-06,
+      "loss": 0.1727,
+      "step": 11509
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.1453472067139718,
+      "learning_rate": 3.0269256418740032e-06,
+      "loss": 0.1372,
+      "step": 11510
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.3916811829928672,
+      "learning_rate": 3.0260979392637753e-06,
+      "loss": 0.2026,
+      "step": 11511
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.1312692966580473,
+      "learning_rate": 3.0252703007275436e-06,
+      "loss": 0.2005,
+      "step": 11512
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.1897225724630032,
+      "learning_rate": 3.0244427262921726e-06,
+      "loss": 0.2029,
+      "step": 11513
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.2389110443757598,
+      "learning_rate": 3.023615215984529e-06,
+      "loss": 0.217,
+      "step": 11514
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.2749540376736244,
+      "learning_rate": 3.02278776983147e-06,
+      "loss": 0.1966,
+      "step": 11515
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.4053243584118766,
+      "learning_rate": 3.021960387859857e-06,
+      "loss": 0.1741,
+      "step": 11516
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.4901374795392885,
+      "learning_rate": 3.021133070096547e-06,
+      "loss": 0.2388,
+      "step": 11517
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.5217758800492451,
+      "learning_rate": 3.0203058165683973e-06,
+      "loss": 0.1809,
+      "step": 11518
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.0180661940174203,
+      "learning_rate": 3.0194786273022575e-06,
+      "loss": 0.1639,
+      "step": 11519
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.083029549864692,
+      "learning_rate": 3.0186515023249808e-06,
+      "loss": 0.2261,
+      "step": 11520
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.3597240246570972,
+      "learning_rate": 3.0178244416634163e-06,
+      "loss": 0.1745,
+      "step": 11521
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.0816415401687691,
+      "learning_rate": 3.0169974453444083e-06,
+      "loss": 0.1628,
+      "step": 11522
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.2272429310106856,
+      "learning_rate": 3.0161705133948072e-06,
+      "loss": 0.2038,
+      "step": 11523
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.0987571137363996,
+      "learning_rate": 3.015343645841449e-06,
+      "loss": 0.2064,
+      "step": 11524
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.0657529506608947,
+      "learning_rate": 3.014516842711177e-06,
+      "loss": 0.1571,
+      "step": 11525
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.2055204745040822,
+      "learning_rate": 3.013690104030831e-06,
+      "loss": 0.2429,
+      "step": 11526
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.2774004179957203,
+      "learning_rate": 3.0128634298272473e-06,
+      "loss": 0.1866,
+      "step": 11527
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.558018611733285,
+      "learning_rate": 3.012036820127259e-06,
+      "loss": 0.1992,
+      "step": 11528
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.0066989413279293,
+      "learning_rate": 3.0112102749576972e-06,
+      "loss": 0.1937,
+      "step": 11529
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.1112344162623875,
+      "learning_rate": 3.010383794345395e-06,
+      "loss": 0.1722,
+      "step": 11530
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.9398409940912267,
+      "learning_rate": 3.0095573783171784e-06,
+      "loss": 0.1838,
+      "step": 11531
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.0940265510251357,
+      "learning_rate": 3.008731026899875e-06,
+      "loss": 0.1714,
+      "step": 11532
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.8543464013122053,
+      "learning_rate": 3.007904740120306e-06,
+      "loss": 0.1433,
+      "step": 11533
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.9286598749092266,
+      "learning_rate": 3.0070785180052946e-06,
+      "loss": 0.1539,
+      "step": 11534
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.1444736375076523,
+      "learning_rate": 3.006252360581661e-06,
+      "loss": 0.1801,
+      "step": 11535
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.1534632465834505,
+      "learning_rate": 3.0054262678762236e-06,
+      "loss": 0.1624,
+      "step": 11536
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.9218855348082082,
+      "learning_rate": 3.0046002399157948e-06,
+      "loss": 0.1652,
+      "step": 11537
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.8160224324453877,
+      "learning_rate": 3.0037742767271894e-06,
+      "loss": 0.1778,
+      "step": 11538
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.2559323356400747,
+      "learning_rate": 3.0029483783372204e-06,
+      "loss": 0.1707,
+      "step": 11539
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.0761642282702382,
+      "learning_rate": 3.0021225447726936e-06,
+      "loss": 0.1849,
+      "step": 11540
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.343808691273359,
+      "learning_rate": 3.0012967760604207e-06,
+      "loss": 0.2381,
+      "step": 11541
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.4965708771083048,
+      "learning_rate": 3.0004710722272013e-06,
+      "loss": 0.1692,
+      "step": 11542
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.2405637198670934,
+      "learning_rate": 2.9996454332998408e-06,
+      "loss": 0.2015,
+      "step": 11543
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.0454769870397649,
+      "learning_rate": 2.99881985930514e-06,
+      "loss": 0.1602,
+      "step": 11544
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.320103177288203,
+      "learning_rate": 2.9979943502698994e-06,
+      "loss": 0.2022,
+      "step": 11545
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.10752357540699,
+      "learning_rate": 2.997168906220911e-06,
+      "loss": 0.1971,
+      "step": 11546
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.1850504204932275,
+      "learning_rate": 2.9963435271849717e-06,
+      "loss": 0.2191,
+      "step": 11547
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.073405325433772,
+      "learning_rate": 2.995518213188875e-06,
+      "loss": 0.2359,
+      "step": 11548
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.8317889559661615,
+      "learning_rate": 2.9946929642594087e-06,
+      "loss": 0.2358,
+      "step": 11549
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.3665188707513585,
+      "learning_rate": 2.993867780423364e-06,
+      "loss": 0.2415,
+      "step": 11550
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.0170840603375106,
+      "learning_rate": 2.9930426617075224e-06,
+      "loss": 0.1412,
+      "step": 11551
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.4261608337020912,
+      "learning_rate": 2.99221760813867e-06,
+      "loss": 0.2093,
+      "step": 11552
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.1614047855669218,
+      "learning_rate": 2.991392619743589e-06,
+      "loss": 0.1839,
+      "step": 11553
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.6367005367676843,
+      "learning_rate": 2.9905676965490602e-06,
+      "loss": 0.2217,
+      "step": 11554
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.472349182919633,
+      "learning_rate": 2.9897428385818584e-06,
+      "loss": 0.2141,
+      "step": 11555
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.834683162256544,
+      "learning_rate": 2.988918045868759e-06,
+      "loss": 0.2315,
+      "step": 11556
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.642877150089448,
+      "learning_rate": 2.9880933184365374e-06,
+      "loss": 0.2048,
+      "step": 11557
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.4940970520447723,
+      "learning_rate": 2.987268656311963e-06,
+      "loss": 0.2206,
+      "step": 11558
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.055172562001889,
+      "learning_rate": 2.9864440595218068e-06,
+      "loss": 0.2806,
+      "step": 11559
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.6713176493480688,
+      "learning_rate": 2.9856195280928335e-06,
+      "loss": 0.1508,
+      "step": 11560
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.1459778409202366,
+      "learning_rate": 2.9847950620518074e-06,
+      "loss": 0.2114,
+      "step": 11561
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.4098608375903874,
+      "learning_rate": 2.9839706614254934e-06,
+      "loss": 0.2141,
+      "step": 11562
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.0832498639378703,
+      "learning_rate": 2.9831463262406522e-06,
+      "loss": 0.1808,
+      "step": 11563
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.3306315610839432,
+      "learning_rate": 2.9823220565240396e-06,
+      "loss": 0.1766,
+      "step": 11564
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.3282529124203428,
+      "learning_rate": 2.9814978523024128e-06,
+      "loss": 0.1828,
+      "step": 11565
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.0412861756959018,
+      "learning_rate": 2.9806737136025284e-06,
+      "loss": 0.1949,
+      "step": 11566
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.7377795706326529,
+      "learning_rate": 2.9798496404511347e-06,
+      "loss": 0.1801,
+      "step": 11567
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.2531642173994586,
+      "learning_rate": 2.9790256328749855e-06,
+      "loss": 0.1926,
+      "step": 11568
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.7592164854120997,
+      "learning_rate": 2.978201690900824e-06,
+      "loss": 0.2181,
+      "step": 11569
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.4642889167338584,
+      "learning_rate": 2.977377814555399e-06,
+      "loss": 0.217,
+      "step": 11570
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.650446180954751,
+      "learning_rate": 2.9765540038654527e-06,
+      "loss": 0.1892,
+      "step": 11571
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.1354634967239912,
+      "learning_rate": 2.9757302588577284e-06,
+      "loss": 0.2381,
+      "step": 11572
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.941480086398115,
+      "learning_rate": 2.9749065795589625e-06,
+      "loss": 0.2367,
+      "step": 11573
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.0398569905837431,
+      "learning_rate": 2.9740829659958943e-06,
+      "loss": 0.1703,
+      "step": 11574
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.441978529651397,
+      "learning_rate": 2.9732594181952584e-06,
+      "loss": 0.1959,
+      "step": 11575
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.4538962486328957,
+      "learning_rate": 2.9724359361837863e-06,
+      "loss": 0.1968,
+      "step": 11576
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.8777326207807639,
+      "learning_rate": 2.9716125199882116e-06,
+      "loss": 0.1819,
+      "step": 11577
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.5279667308188218,
+      "learning_rate": 2.9707891696352586e-06,
+      "loss": 0.209,
+      "step": 11578
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.0384993210057043,
+      "learning_rate": 2.9699658851516568e-06,
+      "loss": 0.1658,
+      "step": 11579
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.171578865781506,
+      "learning_rate": 2.9691426665641287e-06,
+      "loss": 0.2024,
+      "step": 11580
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.3753346285245693,
+      "learning_rate": 2.9683195138994003e-06,
+      "loss": 0.1961,
+      "step": 11581
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.2391219248718597,
+      "learning_rate": 2.9674964271841867e-06,
+      "loss": 0.1775,
+      "step": 11582
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.2311278212991665,
+      "learning_rate": 2.9666734064452075e-06,
+      "loss": 0.2145,
+      "step": 11583
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.2418289626812606,
+      "learning_rate": 2.96585045170918e-06,
+      "loss": 0.1966,
+      "step": 11584
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.6684180509966626,
+      "learning_rate": 2.965027563002815e-06,
+      "loss": 0.1976,
+      "step": 11585
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.5516566152926567,
+      "learning_rate": 2.964204740352828e-06,
+      "loss": 0.18,
+      "step": 11586
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.5364323522235326,
+      "learning_rate": 2.9633819837859236e-06,
+      "loss": 0.2486,
+      "step": 11587
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.1505955091061155,
+      "learning_rate": 2.9625592933288117e-06,
+      "loss": 0.1693,
+      "step": 11588
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.2634503142324818,
+      "learning_rate": 2.961736669008196e-06,
+      "loss": 0.1663,
+      "step": 11589
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.209514780431378,
+      "learning_rate": 2.9609141108507824e-06,
+      "loss": 0.1747,
+      "step": 11590
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.403374950271749,
+      "learning_rate": 2.9600916188832675e-06,
+      "loss": 0.1795,
+      "step": 11591
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.011423520789807,
+      "learning_rate": 2.9592691931323516e-06,
+      "loss": 0.2008,
+      "step": 11592
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.17968144952094,
+      "learning_rate": 2.958446833624732e-06,
+      "loss": 0.1672,
+      "step": 11593
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.3006869459477777,
+      "learning_rate": 2.9576245403871016e-06,
+      "loss": 0.1389,
+      "step": 11594
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.3763745871375153,
+      "learning_rate": 2.9568023134461543e-06,
+      "loss": 0.1979,
+      "step": 11595
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.1879162811649748,
+      "learning_rate": 2.9559801528285775e-06,
+      "loss": 0.1674,
+      "step": 11596
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.44020700576142,
+      "learning_rate": 2.9551580585610604e-06,
+      "loss": 0.1748,
+      "step": 11597
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.6802899943485703,
+      "learning_rate": 2.9543360306702884e-06,
+      "loss": 0.1826,
+      "step": 11598
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.2970220442218776,
+      "learning_rate": 2.9535140691829467e-06,
+      "loss": 0.1866,
+      "step": 11599
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.991328468307663,
+      "learning_rate": 2.9526921741257136e-06,
+      "loss": 0.1422,
+      "step": 11600
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.8800469434880465,
+      "learning_rate": 2.9518703455252705e-06,
+      "loss": 0.177,
+      "step": 11601
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.2884057655099475,
+      "learning_rate": 2.9510485834082935e-06,
+      "loss": 0.2313,
+      "step": 11602
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.4450483854073155,
+      "learning_rate": 2.9502268878014574e-06,
+      "loss": 0.201,
+      "step": 11603
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.3067281662335894,
+      "learning_rate": 2.949405258731437e-06,
+      "loss": 0.1804,
+      "step": 11604
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.381732802030034,
+      "learning_rate": 2.9485836962248993e-06,
+      "loss": 0.2106,
+      "step": 11605
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.8423629347991528,
+      "learning_rate": 2.9477622003085136e-06,
+      "loss": 0.2197,
+      "step": 11606
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.0734298444805321,
+      "learning_rate": 2.9469407710089484e-06,
+      "loss": 0.175,
+      "step": 11607
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.2084609332894134,
+      "learning_rate": 2.9461194083528673e-06,
+      "loss": 0.1663,
+      "step": 11608
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.9922873304110671,
+      "learning_rate": 2.9452981123669295e-06,
+      "loss": 0.161,
+      "step": 11609
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.3167110848579997,
+      "learning_rate": 2.9444768830777975e-06,
+      "loss": 0.2089,
+      "step": 11610
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.4099450256997037,
+      "learning_rate": 2.9436557205121273e-06,
+      "loss": 0.1676,
+      "step": 11611
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.9219035006821213,
+      "learning_rate": 2.9428346246965743e-06,
+      "loss": 0.2137,
+      "step": 11612
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.937378025067346,
+      "learning_rate": 2.9420135956577946e-06,
+      "loss": 0.1539,
+      "step": 11613
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.3108566077353374,
+      "learning_rate": 2.941192633422435e-06,
+      "loss": 0.2324,
+      "step": 11614
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.188759054747151,
+      "learning_rate": 2.940371738017147e-06,
+      "loss": 0.1711,
+      "step": 11615
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.2645395092097593,
+      "learning_rate": 2.939550909468576e-06,
+      "loss": 0.1853,
+      "step": 11616
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.1282653078073681,
+      "learning_rate": 2.9387301478033694e-06,
+      "loss": 0.2173,
+      "step": 11617
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.6255844502578012,
+      "learning_rate": 2.9379094530481654e-06,
+      "loss": 0.1675,
+      "step": 11618
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.6174234273450074,
+      "learning_rate": 2.937088825229608e-06,
+      "loss": 0.245,
+      "step": 11619
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.0472048908680416,
+      "learning_rate": 2.9362682643743325e-06,
+      "loss": 0.2101,
+      "step": 11620
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.4157929501893631,
+      "learning_rate": 2.9354477705089755e-06,
+      "loss": 0.206,
+      "step": 11621
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.6526960781874374,
+      "learning_rate": 2.934627343660173e-06,
+      "loss": 0.2147,
+      "step": 11622
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.1305393695193338,
+      "learning_rate": 2.9338069838545523e-06,
+      "loss": 0.2314,
+      "step": 11623
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.057828690082386,
+      "learning_rate": 2.9329866911187454e-06,
+      "loss": 0.2283,
+      "step": 11624
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.328307229050397,
+      "learning_rate": 2.93216646547938e-06,
+      "loss": 0.2129,
+      "step": 11625
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.638872313943209,
+      "learning_rate": 2.931346306963081e-06,
+      "loss": 0.2715,
+      "step": 11626
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.0463795076478668,
+      "learning_rate": 2.930526215596469e-06,
+      "loss": 0.2056,
+      "step": 11627
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.174138261443786,
+      "learning_rate": 2.929706191406167e-06,
+      "loss": 0.1925,
+      "step": 11628
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.8674736452977678,
+      "learning_rate": 2.928886234418792e-06,
+      "loss": 0.1826,
+      "step": 11629
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.551494439198692,
+      "learning_rate": 2.9280663446609596e-06,
+      "loss": 0.2285,
+      "step": 11630
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.1658641518715642,
+      "learning_rate": 2.9272465221592885e-06,
+      "loss": 0.1921,
+      "step": 11631
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.547892197517802,
+      "learning_rate": 2.926426766940384e-06,
+      "loss": 0.1694,
+      "step": 11632
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.0726856736094994,
+      "learning_rate": 2.9256070790308595e-06,
+      "loss": 0.1521,
+      "step": 11633
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.2802733443876302,
+      "learning_rate": 2.924787458457322e-06,
+      "loss": 0.2079,
+      "step": 11634
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.044329994345984,
+      "learning_rate": 2.923967905246379e-06,
+      "loss": 0.1369,
+      "step": 11635
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.057199989378132,
+      "learning_rate": 2.9231484194246295e-06,
+      "loss": 0.1958,
+      "step": 11636
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.4749246298194794,
+      "learning_rate": 2.922329001018678e-06,
+      "loss": 0.2353,
+      "step": 11637
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.0245918359684199,
+      "learning_rate": 2.9215096500551206e-06,
+      "loss": 0.172,
+      "step": 11638
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.9023242902928075,
+      "learning_rate": 2.9206903665605552e-06,
+      "loss": 0.1911,
+      "step": 11639
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.1593303453966142,
+      "learning_rate": 2.9198711505615777e-06,
+      "loss": 0.2217,
+      "step": 11640
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.1687411972214183,
+      "learning_rate": 2.9190520020847777e-06,
+      "loss": 0.2361,
+      "step": 11641
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.2233794879435391,
+      "learning_rate": 2.918232921156746e-06,
+      "loss": 0.1585,
+      "step": 11642
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.3123784051496348,
+      "learning_rate": 2.917413907804071e-06,
+      "loss": 0.1664,
+      "step": 11643
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.0103867020866724,
+      "learning_rate": 2.9165949620533397e-06,
+      "loss": 0.1786,
+      "step": 11644
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.216882411834292,
+      "learning_rate": 2.9157760839311317e-06,
+      "loss": 0.1559,
+      "step": 11645
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.2081175856799389,
+      "learning_rate": 2.914957273464031e-06,
+      "loss": 0.2095,
+      "step": 11646
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.2024524840327397,
+      "learning_rate": 2.914138530678616e-06,
+      "loss": 0.1828,
+      "step": 11647
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.326072335194481,
+      "learning_rate": 2.913319855601464e-06,
+      "loss": 0.1987,
+      "step": 11648
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 5.270116636367191,
+      "learning_rate": 2.912501248259151e-06,
+      "loss": 0.1795,
+      "step": 11649
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.6287161032295947,
+      "learning_rate": 2.9116827086782463e-06,
+      "loss": 0.1549,
+      "step": 11650
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.4054509144277405,
+      "learning_rate": 2.9108642368853233e-06,
+      "loss": 0.1955,
+      "step": 11651
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.4487385121203558,
+      "learning_rate": 2.910045832906947e-06,
+      "loss": 0.161,
+      "step": 11652
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.2319368792538068,
+      "learning_rate": 2.9092274967696864e-06,
+      "loss": 0.1564,
+      "step": 11653
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.3222036530299826,
+      "learning_rate": 2.908409228500102e-06,
+      "loss": 0.1804,
+      "step": 11654
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.405997913336155,
+      "learning_rate": 2.9075910281247565e-06,
+      "loss": 0.2204,
+      "step": 11655
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.4707406863701153,
+      "learning_rate": 2.9067728956702096e-06,
+      "loss": 0.2204,
+      "step": 11656
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.0503962513442064,
+      "learning_rate": 2.9059548311630182e-06,
+      "loss": 0.1598,
+      "step": 11657
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.0699275326866484,
+      "learning_rate": 2.905136834629739e-06,
+      "loss": 0.1891,
+      "step": 11658
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.9994082320421399,
+      "learning_rate": 2.90431890609692e-06,
+      "loss": 0.2258,
+      "step": 11659
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.7183400532199182,
+      "learning_rate": 2.9035010455911156e-06,
+      "loss": 0.1626,
+      "step": 11660
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.369218337047967,
+      "learning_rate": 2.9026832531388726e-06,
+      "loss": 0.2327,
+      "step": 11661
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.3124876986789982,
+      "learning_rate": 2.9018655287667393e-06,
+      "loss": 0.2105,
+      "step": 11662
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.3765838885512969,
+      "learning_rate": 2.901047872501255e-06,
+      "loss": 0.2084,
+      "step": 11663
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.331340717081294,
+      "learning_rate": 2.900230284368964e-06,
+      "loss": 0.1781,
+      "step": 11664
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.2745245406982146,
+      "learning_rate": 2.8994127643964053e-06,
+      "loss": 0.2411,
+      "step": 11665
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.5104493274665252,
+      "learning_rate": 2.898595312610116e-06,
+      "loss": 0.1895,
+      "step": 11666
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.0436837267393704,
+      "learning_rate": 2.8977779290366336e-06,
+      "loss": 0.1847,
+      "step": 11667
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.1788212361912818,
+      "learning_rate": 2.8969606137024865e-06,
+      "loss": 0.2202,
+      "step": 11668
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.9116639893689595,
+      "learning_rate": 2.8961433666342092e-06,
+      "loss": 0.2154,
+      "step": 11669
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.1693708207837141,
+      "learning_rate": 2.8953261878583263e-06,
+      "loss": 0.1577,
+      "step": 11670
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.1952500108827784,
+      "learning_rate": 2.8945090774013668e-06,
+      "loss": 0.2075,
+      "step": 11671
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.071957987027902,
+      "learning_rate": 2.8936920352898525e-06,
+      "loss": 0.2181,
+      "step": 11672
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.0982779035572883,
+      "learning_rate": 2.892875061550305e-06,
+      "loss": 0.1629,
+      "step": 11673
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.4255338584835395,
+      "learning_rate": 2.8920581562092447e-06,
+      "loss": 0.201,
+      "step": 11674
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.3232166076665157,
+      "learning_rate": 2.8912413192931886e-06,
+      "loss": 0.1524,
+      "step": 11675
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.3943873045330122,
+      "learning_rate": 2.890424550828655e-06,
+      "loss": 0.1973,
+      "step": 11676
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.5072250358294024,
+      "learning_rate": 2.8896078508421512e-06,
+      "loss": 0.1907,
+      "step": 11677
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.5296205200088557,
+      "learning_rate": 2.8887912193601895e-06,
+      "loss": 0.2309,
+      "step": 11678
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.0100083197827003,
+      "learning_rate": 2.8879746564092792e-06,
+      "loss": 0.1774,
+      "step": 11679
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.4134512180714387,
+      "learning_rate": 2.887158162015928e-06,
+      "loss": 0.2281,
+      "step": 11680
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.2712098189983387,
+      "learning_rate": 2.886341736206636e-06,
+      "loss": 0.19,
+      "step": 11681
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.1872835024691148,
+      "learning_rate": 2.885525379007906e-06,
+      "loss": 0.1875,
+      "step": 11682
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.6543119850954873,
+      "learning_rate": 2.8847090904462392e-06,
+      "loss": 0.2373,
+      "step": 11683
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.3939375161071457,
+      "learning_rate": 2.8838928705481318e-06,
+      "loss": 0.228,
+      "step": 11684
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.2141053694197765,
+      "learning_rate": 2.88307671934008e-06,
+      "loss": 0.1958,
+      "step": 11685
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.4758745791002592,
+      "learning_rate": 2.8822606368485727e-06,
+      "loss": 0.216,
+      "step": 11686
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.1054966724225763,
+      "learning_rate": 2.881444623100106e-06,
+      "loss": 0.1666,
+      "step": 11687
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.9965233493482054,
+      "learning_rate": 2.880628678121162e-06,
+      "loss": 0.1707,
+      "step": 11688
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.3746599080238424,
+      "learning_rate": 2.879812801938233e-06,
+      "loss": 0.1792,
+      "step": 11689
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.537493205830829,
+      "learning_rate": 2.8789969945777974e-06,
+      "loss": 0.2462,
+      "step": 11690
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.0830592615343417,
+      "learning_rate": 2.878181256066338e-06,
+      "loss": 0.1913,
+      "step": 11691
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.5234584904716288,
+      "learning_rate": 2.8773655864303364e-06,
+      "loss": 0.2177,
+      "step": 11692
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.4171636255560198,
+      "learning_rate": 2.8765499856962677e-06,
+      "loss": 0.2191,
+      "step": 11693
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.106614772415211,
+      "learning_rate": 2.8757344538906096e-06,
+      "loss": 0.2232,
+      "step": 11694
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.573243539573993,
+      "learning_rate": 2.874918991039831e-06,
+      "loss": 0.2137,
+      "step": 11695
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.2682593238921336,
+      "learning_rate": 2.8741035971704035e-06,
+      "loss": 0.2305,
+      "step": 11696
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.2600997097388236,
+      "learning_rate": 2.8732882723087963e-06,
+      "loss": 0.1982,
+      "step": 11697
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.477092801591884,
+      "learning_rate": 2.8724730164814764e-06,
+      "loss": 0.2138,
+      "step": 11698
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.2112315497700732,
+      "learning_rate": 2.871657829714903e-06,
+      "loss": 0.1593,
+      "step": 11699
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.3706958633582726,
+      "learning_rate": 2.870842712035542e-06,
+      "loss": 0.1758,
+      "step": 11700
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.43060386192551,
+      "learning_rate": 2.8700276634698497e-06,
+      "loss": 0.2098,
+      "step": 11701
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.8205944258012563,
+      "learning_rate": 2.869212684044285e-06,
+      "loss": 0.141,
+      "step": 11702
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.9165728120909808,
+      "learning_rate": 2.868397773785303e-06,
+      "loss": 0.1396,
+      "step": 11703
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.9477311868121063,
+      "learning_rate": 2.867582932719354e-06,
+      "loss": 0.1352,
+      "step": 11704
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.5404806429257536,
+      "learning_rate": 2.866768160872891e-06,
+      "loss": 0.2239,
+      "step": 11705
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.4543314099469968,
+      "learning_rate": 2.865953458272358e-06,
+      "loss": 0.2727,
+      "step": 11706
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.2711416271831708,
+      "learning_rate": 2.8651388249442047e-06,
+      "loss": 0.221,
+      "step": 11707
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.460132252009693,
+      "learning_rate": 2.8643242609148707e-06,
+      "loss": 0.2138,
+      "step": 11708
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.3608962021002389,
+      "learning_rate": 2.8635097662107997e-06,
+      "loss": 0.19,
+      "step": 11709
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.2642417135853434,
+      "learning_rate": 2.86269534085843e-06,
+      "loss": 0.1844,
+      "step": 11710
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.1347702735372327,
+      "learning_rate": 2.8618809848841988e-06,
+      "loss": 0.1812,
+      "step": 11711
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.1832461554887763,
+      "learning_rate": 2.861066698314542e-06,
+      "loss": 0.1859,
+      "step": 11712
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.8766295318166808,
+      "learning_rate": 2.8602524811758888e-06,
+      "loss": 0.1572,
+      "step": 11713
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.6809209885453809,
+      "learning_rate": 2.8594383334946695e-06,
+      "loss": 0.1663,
+      "step": 11714
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.133118002902024,
+      "learning_rate": 2.8586242552973137e-06,
+      "loss": 0.2063,
+      "step": 11715
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.2419235740841879,
+      "learning_rate": 2.8578102466102454e-06,
+      "loss": 0.1713,
+      "step": 11716
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.5289237564753206,
+      "learning_rate": 2.8569963074598904e-06,
+      "loss": 0.2101,
+      "step": 11717
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.0176873889770288,
+      "learning_rate": 2.856182437872666e-06,
+      "loss": 0.1555,
+      "step": 11718
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.397037913915921,
+      "learning_rate": 2.8553686378749925e-06,
+      "loss": 0.1859,
+      "step": 11719
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.3562175622699866,
+      "learning_rate": 2.8545549074932876e-06,
+      "loss": 0.1647,
+      "step": 11720
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.2329450878067039,
+      "learning_rate": 2.853741246753963e-06,
+      "loss": 0.1729,
+      "step": 11721
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.2563541262445597,
+      "learning_rate": 2.8529276556834305e-06,
+      "loss": 0.1786,
+      "step": 11722
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.3801415451252828,
+      "learning_rate": 2.852114134308104e-06,
+      "loss": 0.1968,
+      "step": 11723
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.1295505882599286,
+      "learning_rate": 2.8513006826543853e-06,
+      "loss": 0.194,
+      "step": 11724
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.3025445871354677,
+      "learning_rate": 2.850487300748682e-06,
+      "loss": 0.1815,
+      "step": 11725
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.9762160258786183,
+      "learning_rate": 2.8496739886173994e-06,
+      "loss": 0.1956,
+      "step": 11726
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.1144679169821408,
+      "learning_rate": 2.8488607462869333e-06,
+      "loss": 0.176,
+      "step": 11727
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.3542468653980284,
+      "learning_rate": 2.8480475737836844e-06,
+      "loss": 0.2019,
+      "step": 11728
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.2011786980163814,
+      "learning_rate": 2.8472344711340487e-06,
+      "loss": 0.1443,
+      "step": 11729
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.0256283865886686,
+      "learning_rate": 2.8464214383644224e-06,
+      "loss": 0.1592,
+      "step": 11730
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.3277223144863373,
+      "learning_rate": 2.8456084755011926e-06,
+      "loss": 0.1763,
+      "step": 11731
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.881039188750156,
+      "learning_rate": 2.8447955825707496e-06,
+      "loss": 0.1668,
+      "step": 11732
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.29238382806114,
+      "learning_rate": 2.8439827595994825e-06,
+      "loss": 0.1914,
+      "step": 11733
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.2011408266846397,
+      "learning_rate": 2.843170006613774e-06,
+      "loss": 0.1627,
+      "step": 11734
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.5095553924143046,
+      "learning_rate": 2.84235732364001e-06,
+      "loss": 0.1738,
+      "step": 11735
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.2510507756657812,
+      "learning_rate": 2.841544710704566e-06,
+      "loss": 0.1754,
+      "step": 11736
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.468251047200278,
+      "learning_rate": 2.8407321678338206e-06,
+      "loss": 0.2216,
+      "step": 11737
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.8311353335104932,
+      "learning_rate": 2.8399196950541543e-06,
+      "loss": 0.1738,
+      "step": 11738
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.5347538971640966,
+      "learning_rate": 2.839107292391934e-06,
+      "loss": 0.2152,
+      "step": 11739
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.3724627005002679,
+      "learning_rate": 2.8382949598735354e-06,
+      "loss": 0.2271,
+      "step": 11740
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.8106946724266021,
+      "learning_rate": 2.8374826975253245e-06,
+      "loss": 0.114,
+      "step": 11741
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.2730141796163645,
+      "learning_rate": 2.8366705053736677e-06,
+      "loss": 0.1719,
+      "step": 11742
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.0702426219321421,
+      "learning_rate": 2.8358583834449305e-06,
+      "loss": 0.1581,
+      "step": 11743
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.113937893676346,
+      "learning_rate": 2.8350463317654776e-06,
+      "loss": 0.1637,
+      "step": 11744
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.467013661787729,
+      "learning_rate": 2.8342343503616625e-06,
+      "loss": 0.1706,
+      "step": 11745
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.070109678807088,
+      "learning_rate": 2.8334224392598464e-06,
+      "loss": 0.153,
+      "step": 11746
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.2398841693878322,
+      "learning_rate": 2.832610598486384e-06,
+      "loss": 0.188,
+      "step": 11747
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.359172301691015,
+      "learning_rate": 2.831798828067629e-06,
+      "loss": 0.1834,
+      "step": 11748
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.2934011531523786,
+      "learning_rate": 2.8309871280299295e-06,
+      "loss": 0.1861,
+      "step": 11749
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.018234965323338,
+      "learning_rate": 2.8301754983996355e-06,
+      "loss": 0.1206,
+      "step": 11750
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.7825957628311139,
+      "learning_rate": 2.8293639392030915e-06,
+      "loss": 0.1205,
+      "step": 11751
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.1791565818265504,
+      "learning_rate": 2.8285524504666427e-06,
+      "loss": 0.1649,
+      "step": 11752
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.143889699804985,
+      "learning_rate": 2.8277410322166323e-06,
+      "loss": 0.1507,
+      "step": 11753
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.1348800322462784,
+      "learning_rate": 2.826929684479395e-06,
+      "loss": 0.162,
+      "step": 11754
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.473022392654196,
+      "learning_rate": 2.82611840728127e-06,
+      "loss": 0.2285,
+      "step": 11755
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 4.1905768179061536,
+      "learning_rate": 2.8253072006485928e-06,
+      "loss": 0.2337,
+      "step": 11756
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.104461794299276,
+      "learning_rate": 2.8244960646076936e-06,
+      "loss": 0.1749,
+      "step": 11757
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.4213740860309838,
+      "learning_rate": 2.8236849991849036e-06,
+      "loss": 0.2244,
+      "step": 11758
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.3771679460406732,
+      "learning_rate": 2.8228740044065494e-06,
+      "loss": 0.1411,
+      "step": 11759
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.6190701804119716,
+      "learning_rate": 2.8220630802989564e-06,
+      "loss": 0.2021,
+      "step": 11760
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.61220992957623,
+      "learning_rate": 2.821252226888448e-06,
+      "loss": 0.2165,
+      "step": 11761
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.311048420247585,
+      "learning_rate": 2.820441444201347e-06,
+      "loss": 0.1832,
+      "step": 11762
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.2858568212570824,
+      "learning_rate": 2.8196307322639687e-06,
+      "loss": 0.1645,
+      "step": 11763
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.5341275677973019,
+      "learning_rate": 2.818820091102631e-06,
+      "loss": 0.1861,
+      "step": 11764
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.0171017511030644,
+      "learning_rate": 2.8180095207436464e-06,
+      "loss": 0.1298,
+      "step": 11765
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.9783028658733272,
+      "learning_rate": 2.81719902121333e-06,
+      "loss": 0.1617,
+      "step": 11766
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.3323805901878691,
+      "learning_rate": 2.816388592537986e-06,
+      "loss": 0.1853,
+      "step": 11767
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.4104130978611922,
+      "learning_rate": 2.8155782347439247e-06,
+      "loss": 0.1859,
+      "step": 11768
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.2082388263019384,
+      "learning_rate": 2.8147679478574498e-06,
+      "loss": 0.1636,
+      "step": 11769
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.2800821311243096,
+      "learning_rate": 2.813957731904864e-06,
+      "loss": 0.1783,
+      "step": 11770
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.659754455413906,
+      "learning_rate": 2.813147586912469e-06,
+      "loss": 0.2483,
+      "step": 11771
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.8952215979067374,
+      "learning_rate": 2.812337512906559e-06,
+      "loss": 0.1803,
+      "step": 11772
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.9889012679892052,
+      "learning_rate": 2.811527509913431e-06,
+      "loss": 0.165,
+      "step": 11773
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.0688303235471053,
+      "learning_rate": 2.810717577959381e-06,
+      "loss": 0.1971,
+      "step": 11774
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.1424700514715211,
+      "learning_rate": 2.809907717070695e-06,
+      "loss": 0.1634,
+      "step": 11775
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.2595742871373647,
+      "learning_rate": 2.8090979272736663e-06,
+      "loss": 0.1662,
+      "step": 11776
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.389790335385194,
+      "learning_rate": 2.8082882085945763e-06,
+      "loss": 0.2239,
+      "step": 11777
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.2749151888857342,
+      "learning_rate": 2.8074785610597114e-06,
+      "loss": 0.2157,
+      "step": 11778
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.4921836952840024,
+      "learning_rate": 2.806668984695354e-06,
+      "loss": 0.2083,
+      "step": 11779
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.3936274268260673,
+      "learning_rate": 2.805859479527784e-06,
+      "loss": 0.1981,
+      "step": 11780
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.4766714419184743,
+      "learning_rate": 2.8050500455832745e-06,
+      "loss": 0.2272,
+      "step": 11781
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.2583219077157881,
+      "learning_rate": 2.804240682888103e-06,
+      "loss": 0.1626,
+      "step": 11782
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.1014767978688174,
+      "learning_rate": 2.803431391468542e-06,
+      "loss": 0.1781,
+      "step": 11783
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.1292855836507665,
+      "learning_rate": 2.8026221713508628e-06,
+      "loss": 0.2023,
+      "step": 11784
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.6576189578026546,
+      "learning_rate": 2.80181302256133e-06,
+      "loss": 0.1642,
+      "step": 11785
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.1577277727824988,
+      "learning_rate": 2.8010039451262098e-06,
+      "loss": 0.1667,
+      "step": 11786
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.4278477964215128,
+      "learning_rate": 2.8001949390717664e-06,
+      "loss": 0.1966,
+      "step": 11787
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.1605176715906487,
+      "learning_rate": 2.7993860044242605e-06,
+      "loss": 0.2363,
+      "step": 11788
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.1164065812833173,
+      "learning_rate": 2.798577141209952e-06,
+      "loss": 0.1597,
+      "step": 11789
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.1918646117182576,
+      "learning_rate": 2.797768349455094e-06,
+      "loss": 0.2257,
+      "step": 11790
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.836369574373572,
+      "learning_rate": 2.796959629185941e-06,
+      "loss": 0.2105,
+      "step": 11791
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.831305476090657,
+      "learning_rate": 2.796150980428748e-06,
+      "loss": 0.1659,
+      "step": 11792
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.0832405872105486,
+      "learning_rate": 2.7953424032097587e-06,
+      "loss": 0.1628,
+      "step": 11793
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.2952700444615972,
+      "learning_rate": 2.794533897555225e-06,
+      "loss": 0.1965,
+      "step": 11794
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.5307436611929592,
+      "learning_rate": 2.7937254634913873e-06,
+      "loss": 0.1759,
+      "step": 11795
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.246399444345346,
+      "learning_rate": 2.79291710104449e-06,
+      "loss": 0.1653,
+      "step": 11796
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.0560802518002022,
+      "learning_rate": 2.7921088102407724e-06,
+      "loss": 0.1635,
+      "step": 11797
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.281630340390737,
+      "learning_rate": 2.7913005911064746e-06,
+      "loss": 0.1847,
+      "step": 11798
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.4454228255976524,
+      "learning_rate": 2.7904924436678275e-06,
+      "loss": 0.2307,
+      "step": 11799
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.7583404206418097,
+      "learning_rate": 2.7896843679510667e-06,
+      "loss": 0.1716,
+      "step": 11800
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.9834791413634856,
+      "learning_rate": 2.788876363982422e-06,
+      "loss": 0.1923,
+      "step": 11801
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.4048330837821108,
+      "learning_rate": 2.788068431788124e-06,
+      "loss": 0.1938,
+      "step": 11802
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.1888618816309349,
+      "learning_rate": 2.7872605713943944e-06,
+      "loss": 0.186,
+      "step": 11803
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.4804247155095076,
+      "learning_rate": 2.7864527828274597e-06,
+      "loss": 0.2075,
+      "step": 11804
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.2607218639529258,
+      "learning_rate": 2.7856450661135404e-06,
+      "loss": 0.1785,
+      "step": 11805
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.2513374013676597,
+      "learning_rate": 2.7848374212788552e-06,
+      "loss": 0.1521,
+      "step": 11806
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.964333016620233,
+      "learning_rate": 2.7840298483496232e-06,
+      "loss": 0.2173,
+      "step": 11807
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.2288433632399067,
+      "learning_rate": 2.7832223473520546e-06,
+      "loss": 0.1916,
+      "step": 11808
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.564458177917931,
+      "learning_rate": 2.7824149183123637e-06,
+      "loss": 0.2041,
+      "step": 11809
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.0770291481124843,
+      "learning_rate": 2.7816075612567615e-06,
+      "loss": 0.1399,
+      "step": 11810
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.0708117991105444,
+      "learning_rate": 2.7808002762114515e-06,
+      "loss": 0.2285,
+      "step": 11811
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.1991274516240547,
+      "learning_rate": 2.779993063202643e-06,
+      "loss": 0.1861,
+      "step": 11812
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.2817798716745579,
+      "learning_rate": 2.779185922256534e-06,
+      "loss": 0.1867,
+      "step": 11813
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.7461493267798183,
+      "learning_rate": 2.778378853399327e-06,
+      "loss": 0.2424,
+      "step": 11814
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.6241668298376102,
+      "learning_rate": 2.7775718566572203e-06,
+      "loss": 0.1998,
+      "step": 11815
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.3314269756382346,
+      "learning_rate": 2.7767649320564104e-06,
+      "loss": 0.215,
+      "step": 11816
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.4976512257148946,
+      "learning_rate": 2.775958079623088e-06,
+      "loss": 0.1781,
+      "step": 11817
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.1982777970299467,
+      "learning_rate": 2.7751512993834452e-06,
+      "loss": 0.2134,
+      "step": 11818
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.1199306941408653,
+      "learning_rate": 2.77434459136367e-06,
+      "loss": 0.2226,
+      "step": 11819
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.2719087214885116,
+      "learning_rate": 2.773537955589951e-06,
+      "loss": 0.2086,
+      "step": 11820
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.237137359408623,
+      "learning_rate": 2.772731392088469e-06,
+      "loss": 0.2085,
+      "step": 11821
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.5027386331361554,
+      "learning_rate": 2.7719249008854067e-06,
+      "loss": 0.1903,
+      "step": 11822
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.4134846449754395,
+      "learning_rate": 2.771118482006943e-06,
+      "loss": 0.1909,
+      "step": 11823
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.1562770086433198,
+      "learning_rate": 2.7703121354792555e-06,
+      "loss": 0.1373,
+      "step": 11824
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.3014194767142766,
+      "learning_rate": 2.76950586132852e-06,
+      "loss": 0.1584,
+      "step": 11825
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.4125909735913509,
+      "learning_rate": 2.7686996595809045e-06,
+      "loss": 0.1739,
+      "step": 11826
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.4213981012054377,
+      "learning_rate": 2.767893530262582e-06,
+      "loss": 0.1953,
+      "step": 11827
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.5396760937341984,
+      "learning_rate": 2.7670874733997207e-06,
+      "loss": 0.211,
+      "step": 11828
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.5544875989286162,
+      "learning_rate": 2.766281489018482e-06,
+      "loss": 0.2055,
+      "step": 11829
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.2858570791013446,
+      "learning_rate": 2.7654755771450326e-06,
+      "loss": 0.174,
+      "step": 11830
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.1314078336803919,
+      "learning_rate": 2.7646697378055285e-06,
+      "loss": 0.181,
+      "step": 11831
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.0690117540556348,
+      "learning_rate": 2.7638639710261307e-06,
+      "loss": 0.1993,
+      "step": 11832
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.0049579136313238,
+      "learning_rate": 2.763058276832994e-06,
+      "loss": 0.1533,
+      "step": 11833
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.1714840300363285,
+      "learning_rate": 2.7622526552522745e-06,
+      "loss": 0.1342,
+      "step": 11834
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.3064068241526468,
+      "learning_rate": 2.7614471063101177e-06,
+      "loss": 0.1841,
+      "step": 11835
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.949482125967886,
+      "learning_rate": 2.7606416300326755e-06,
+      "loss": 0.241,
+      "step": 11836
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.1994193589603406,
+      "learning_rate": 2.759836226446094e-06,
+      "loss": 0.2129,
+      "step": 11837
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.1472696315291482,
+      "learning_rate": 2.7590308955765183e-06,
+      "loss": 0.1833,
+      "step": 11838
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.395309861320572,
+      "learning_rate": 2.7582256374500865e-06,
+      "loss": 0.2405,
+      "step": 11839
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.3668117170244838,
+      "learning_rate": 2.75742045209294e-06,
+      "loss": 0.2057,
+      "step": 11840
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.8104249281488056,
+      "learning_rate": 2.7566153395312144e-06,
+      "loss": 0.1943,
+      "step": 11841
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.0359123419674274,
+      "learning_rate": 2.755810299791046e-06,
+      "loss": 0.1118,
+      "step": 11842
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.0542800363480012,
+      "learning_rate": 2.755005332898567e-06,
+      "loss": 0.1671,
+      "step": 11843
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.7745348228464974,
+      "learning_rate": 2.7542004388799037e-06,
+      "loss": 0.1548,
+      "step": 11844
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.2978229318270698,
+      "learning_rate": 2.7533956177611864e-06,
+      "loss": 0.227,
+      "step": 11845
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.268693154334684,
+      "learning_rate": 2.752590869568541e-06,
+      "loss": 0.1869,
+      "step": 11846
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.3712991386752962,
+      "learning_rate": 2.751786194328086e-06,
+      "loss": 0.1965,
+      "step": 11847
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.6290719670679772,
+      "learning_rate": 2.750981592065946e-06,
+      "loss": 0.2081,
+      "step": 11848
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.137786085576002,
+      "learning_rate": 2.7501770628082346e-06,
+      "loss": 0.1879,
+      "step": 11849
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.2751251195541624,
+      "learning_rate": 2.74937260658107e-06,
+      "loss": 0.241,
+      "step": 11850
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.3714910159905,
+      "learning_rate": 2.7485682234105648e-06,
+      "loss": 0.1696,
+      "step": 11851
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.0998330439314616,
+      "learning_rate": 2.7477639133228314e-06,
+      "loss": 0.2143,
+      "step": 11852
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.4844300209689936,
+      "learning_rate": 2.746959676343974e-06,
+      "loss": 0.1946,
+      "step": 11853
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.559343726431273,
+      "learning_rate": 2.7461555125001023e-06,
+      "loss": 0.1991,
+      "step": 11854
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.3416852390986214,
+      "learning_rate": 2.7453514218173176e-06,
+      "loss": 0.2599,
+      "step": 11855
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.1809867973112522,
+      "learning_rate": 2.7445474043217244e-06,
+      "loss": 0.1937,
+      "step": 11856
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.2460613366099724,
+      "learning_rate": 2.7437434600394186e-06,
+      "loss": 0.1669,
+      "step": 11857
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.3122049664231799,
+      "learning_rate": 2.7429395889964966e-06,
+      "loss": 0.1619,
+      "step": 11858
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.9524904067374271,
+      "learning_rate": 2.7421357912190536e-06,
+      "loss": 0.1609,
+      "step": 11859
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.5444913956435513,
+      "learning_rate": 2.741332066733181e-06,
+      "loss": 0.2297,
+      "step": 11860
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.977636439135051,
+      "learning_rate": 2.7405284155649702e-06,
+      "loss": 0.2066,
+      "step": 11861
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.1281526634147885,
+      "learning_rate": 2.739724837740505e-06,
+      "loss": 0.1589,
+      "step": 11862
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.802469365260861,
+      "learning_rate": 2.738921333285871e-06,
+      "loss": 0.139,
+      "step": 11863
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.3479069131127848,
+      "learning_rate": 2.738117902227152e-06,
+      "loss": 0.1882,
+      "step": 11864
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.278411168424286,
+      "learning_rate": 2.737314544590426e-06,
+      "loss": 0.1873,
+      "step": 11865
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.302817820974836,
+      "learning_rate": 2.736511260401772e-06,
+      "loss": 0.2519,
+      "step": 11866
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.160191769810761,
+      "learning_rate": 2.7357080496872624e-06,
+      "loss": 0.1601,
+      "step": 11867
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.9831855035778202,
+      "learning_rate": 2.734904912472971e-06,
+      "loss": 0.1412,
+      "step": 11868
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.2366496627701968,
+      "learning_rate": 2.7341018487849693e-06,
+      "loss": 0.192,
+      "step": 11869
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.364705799306357,
+      "learning_rate": 2.733298858649327e-06,
+      "loss": 0.1506,
+      "step": 11870
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.3453710115758706,
+      "learning_rate": 2.7324959420921042e-06,
+      "loss": 0.1477,
+      "step": 11871
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.1849457789643485,
+      "learning_rate": 2.7316930991393675e-06,
+      "loss": 0.1776,
+      "step": 11872
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.1267761408835513,
+      "learning_rate": 2.7308903298171775e-06,
+      "loss": 0.131,
+      "step": 11873
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.1056125769717804,
+      "learning_rate": 2.7300876341515936e-06,
+      "loss": 0.2359,
+      "step": 11874
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.597825491412398,
+      "learning_rate": 2.7292850121686685e-06,
+      "loss": 0.1764,
+      "step": 11875
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.6067231909017112,
+      "learning_rate": 2.7284824638944573e-06,
+      "loss": 0.168,
+      "step": 11876
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.151932219389669,
+      "learning_rate": 2.7276799893550126e-06,
+      "loss": 0.2078,
+      "step": 11877
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.2868777545688268,
+      "learning_rate": 2.7268775885763815e-06,
+      "loss": 0.2115,
+      "step": 11878
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.5351473259605795,
+      "learning_rate": 2.7260752615846132e-06,
+      "loss": 0.189,
+      "step": 11879
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.1860313619629124,
+      "learning_rate": 2.7252730084057473e-06,
+      "loss": 0.2122,
+      "step": 11880
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.316021081851415,
+      "learning_rate": 2.7244708290658275e-06,
+      "loss": 0.179,
+      "step": 11881
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.471976223125551,
+      "learning_rate": 2.7236687235908953e-06,
+      "loss": 0.1909,
+      "step": 11882
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.6949840526336983,
+      "learning_rate": 2.7228666920069835e-06,
+      "loss": 0.1916,
+      "step": 11883
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.9635998182104134,
+      "learning_rate": 2.7220647343401302e-06,
+      "loss": 0.2008,
+      "step": 11884
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.1598842635974447,
+      "learning_rate": 2.721262850616363e-06,
+      "loss": 0.1703,
+      "step": 11885
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.3665221393617097,
+      "learning_rate": 2.7204610408617147e-06,
+      "loss": 0.1837,
+      "step": 11886
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.9475125276151296,
+      "learning_rate": 2.719659305102212e-06,
+      "loss": 0.1783,
+      "step": 11887
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.1344727024847736,
+      "learning_rate": 2.7188576433638812e-06,
+      "loss": 0.1996,
+      "step": 11888
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.2044055087932355,
+      "learning_rate": 2.718056055672741e-06,
+      "loss": 0.1588,
+      "step": 11889
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.5363647470151525,
+      "learning_rate": 2.7172545420548135e-06,
+      "loss": 0.2189,
+      "step": 11890
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.9837291278951533,
+      "learning_rate": 2.7164531025361165e-06,
+      "loss": 0.1696,
+      "step": 11891
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.4603955135591171,
+      "learning_rate": 2.715651737142666e-06,
+      "loss": 0.1835,
+      "step": 11892
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.2536314547318668,
+      "learning_rate": 2.714850445900473e-06,
+      "loss": 0.2018,
+      "step": 11893
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.1480303159274292,
+      "learning_rate": 2.7140492288355476e-06,
+      "loss": 0.1902,
+      "step": 11894
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.6423541811079294,
+      "learning_rate": 2.713248085973899e-06,
+      "loss": 0.1499,
+      "step": 11895
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.207440856147587,
+      "learning_rate": 2.712447017341533e-06,
+      "loss": 0.1692,
+      "step": 11896
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.5447662265463686,
+      "learning_rate": 2.7116460229644536e-06,
+      "loss": 0.2088,
+      "step": 11897
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.100161748571969,
+      "learning_rate": 2.7108451028686584e-06,
+      "loss": 0.2306,
+      "step": 11898
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.2691337373921754,
+      "learning_rate": 2.7100442570801478e-06,
+      "loss": 0.2041,
+      "step": 11899
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.5039503512170906,
+      "learning_rate": 2.709243485624919e-06,
+      "loss": 0.2124,
+      "step": 11900
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.0001030101470783,
+      "learning_rate": 2.7084427885289623e-06,
+      "loss": 0.1517,
+      "step": 11901
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.7658118427631706,
+      "learning_rate": 2.7076421658182716e-06,
+      "loss": 0.2255,
+      "step": 11902
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.6041935778090408,
+      "learning_rate": 2.7068416175188332e-06,
+      "loss": 0.1772,
+      "step": 11903
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.5545018962312962,
+      "learning_rate": 2.7060411436566346e-06,
+      "loss": 0.1999,
+      "step": 11904
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.0792328215295155,
+      "learning_rate": 2.705240744257659e-06,
+      "loss": 0.1546,
+      "step": 11905
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.104260252489354,
+      "learning_rate": 2.7044404193478913e-06,
+      "loss": 0.198,
+      "step": 11906
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.9288158016588945,
+      "learning_rate": 2.7036401689533054e-06,
+      "loss": 0.1806,
+      "step": 11907
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.9426731986493867,
+      "learning_rate": 2.7028399930998795e-06,
+      "loss": 0.159,
+      "step": 11908
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.4133223867676605,
+      "learning_rate": 2.702039891813589e-06,
+      "loss": 0.1729,
+      "step": 11909
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.3573251621180757,
+      "learning_rate": 2.7012398651204063e-06,
+      "loss": 0.2023,
+      "step": 11910
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.0845242307268463,
+      "learning_rate": 2.7004399130462982e-06,
+      "loss": 0.1719,
+      "step": 11911
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.427784333092127,
+      "learning_rate": 2.6996400356172325e-06,
+      "loss": 0.211,
+      "step": 11912
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.3072031859994067,
+      "learning_rate": 2.6988402328591745e-06,
+      "loss": 0.1935,
+      "step": 11913
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.83172228023648,
+      "learning_rate": 2.6980405047980853e-06,
+      "loss": 0.2746,
+      "step": 11914
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.8721083058624532,
+      "learning_rate": 2.697240851459927e-06,
+      "loss": 0.2599,
+      "step": 11915
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.8096510369400767,
+      "learning_rate": 2.6964412728706537e-06,
+      "loss": 0.2116,
+      "step": 11916
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.6602670251128195,
+      "learning_rate": 2.6956417690562196e-06,
+      "loss": 0.2387,
+      "step": 11917
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.3933653859959831,
+      "learning_rate": 2.6948423400425828e-06,
+      "loss": 0.1562,
+      "step": 11918
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.1811008513108832,
+      "learning_rate": 2.694042985855686e-06,
+      "loss": 0.1581,
+      "step": 11919
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.3527899388956421,
+      "learning_rate": 2.6932437065214824e-06,
+      "loss": 0.2217,
+      "step": 11920
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.9656396586962875,
+      "learning_rate": 2.692444502065912e-06,
+      "loss": 0.1952,
+      "step": 11921
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.7474010329351557,
+      "learning_rate": 2.69164537251492e-06,
+      "loss": 0.1935,
+      "step": 11922
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.5497648189538926,
+      "learning_rate": 2.690846317894447e-06,
+      "loss": 0.2061,
+      "step": 11923
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.5251329475438378,
+      "learning_rate": 2.690047338230433e-06,
+      "loss": 0.1849,
+      "step": 11924
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.1024968499886887,
+      "learning_rate": 2.6892484335488073e-06,
+      "loss": 0.2045,
+      "step": 11925
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.3768016055455037,
+      "learning_rate": 2.688449603875507e-06,
+      "loss": 0.2229,
+      "step": 11926
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.413470892679361,
+      "learning_rate": 2.6876508492364615e-06,
+      "loss": 0.1957,
+      "step": 11927
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.5013489171377061,
+      "learning_rate": 2.6868521696576017e-06,
+      "loss": 0.2267,
+      "step": 11928
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.4999775388148588,
+      "learning_rate": 2.686053565164848e-06,
+      "loss": 0.1841,
+      "step": 11929
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.0151510966990556,
+      "learning_rate": 2.6852550357841268e-06,
+      "loss": 0.1558,
+      "step": 11930
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.8646170348490658,
+      "learning_rate": 2.684456581541358e-06,
+      "loss": 0.2307,
+      "step": 11931
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.0945571271463574,
+      "learning_rate": 2.6836582024624597e-06,
+      "loss": 0.1354,
+      "step": 11932
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.522292004111019,
+      "learning_rate": 2.68285989857335e-06,
+      "loss": 0.1913,
+      "step": 11933
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.19557728982285,
+      "learning_rate": 2.682061669899939e-06,
+      "loss": 0.1543,
+      "step": 11934
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.1402833701333586,
+      "learning_rate": 2.681263516468139e-06,
+      "loss": 0.1591,
+      "step": 11935
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.515206790524491,
+      "learning_rate": 2.680465438303861e-06,
+      "loss": 0.1862,
+      "step": 11936
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.4352595554273595,
+      "learning_rate": 2.6796674354330067e-06,
+      "loss": 0.1931,
+      "step": 11937
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.5225420716733553,
+      "learning_rate": 2.678869507881483e-06,
+      "loss": 0.2215,
+      "step": 11938
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.6748677945272783,
+      "learning_rate": 2.6780716556751878e-06,
+      "loss": 0.2142,
+      "step": 11939
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.275196419924423,
+      "learning_rate": 2.677273878840022e-06,
+      "loss": 0.1676,
+      "step": 11940
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.250063362601817,
+      "learning_rate": 2.676476177401882e-06,
+      "loss": 0.1666,
+      "step": 11941
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.2018750931258222,
+      "learning_rate": 2.6756785513866636e-06,
+      "loss": 0.1706,
+      "step": 11942
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.6083439514900038,
+      "learning_rate": 2.674881000820254e-06,
+      "loss": 0.2029,
+      "step": 11943
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.25140322810025,
+      "learning_rate": 2.6740835257285434e-06,
+      "loss": 0.2171,
+      "step": 11944
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.3822101854553746,
+      "learning_rate": 2.67328612613742e-06,
+      "loss": 0.1874,
+      "step": 11945
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.3498228100330245,
+      "learning_rate": 2.6724888020727687e-06,
+      "loss": 0.1809,
+      "step": 11946
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.373496946365467,
+      "learning_rate": 2.6716915535604672e-06,
+      "loss": 0.1932,
+      "step": 11947
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.6510259139174908,
+      "learning_rate": 2.6708943806263966e-06,
+      "loss": 0.2173,
+      "step": 11948
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.690464713594861,
+      "learning_rate": 2.6700972832964332e-06,
+      "loss": 0.2241,
+      "step": 11949
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.1944857119006693,
+      "learning_rate": 2.6693002615964526e-06,
+      "loss": 0.151,
+      "step": 11950
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.5110781751569036,
+      "learning_rate": 2.6685033155523275e-06,
+      "loss": 0.1743,
+      "step": 11951
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.3636657287450404,
+      "learning_rate": 2.6677064451899237e-06,
+      "loss": 0.1593,
+      "step": 11952
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.9480112192253838,
+      "learning_rate": 2.6669096505351096e-06,
+      "loss": 0.1756,
+      "step": 11953
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.6623873629452712,
+      "learning_rate": 2.6661129316137514e-06,
+      "loss": 0.2679,
+      "step": 11954
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.5331092732855702,
+      "learning_rate": 2.665316288451708e-06,
+      "loss": 0.212,
+      "step": 11955
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.3687110504504825,
+      "learning_rate": 2.664519721074843e-06,
+      "loss": 0.2142,
+      "step": 11956
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.606503573117763,
+      "learning_rate": 2.663723229509008e-06,
+      "loss": 0.248,
+      "step": 11957
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.2738166724342055,
+      "learning_rate": 2.6629268137800607e-06,
+      "loss": 0.1783,
+      "step": 11958
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.7687878678787194,
+      "learning_rate": 2.662130473913853e-06,
+      "loss": 0.185,
+      "step": 11959
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.8877347773310873,
+      "learning_rate": 2.661334209936236e-06,
+      "loss": 0.1494,
+      "step": 11960
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.9903802677847389,
+      "learning_rate": 2.660538021873054e-06,
+      "loss": 0.1827,
+      "step": 11961
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.916770384169705,
+      "learning_rate": 2.659741909750153e-06,
+      "loss": 0.1552,
+      "step": 11962
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.3161062640225036,
+      "learning_rate": 2.6589458735933753e-06,
+      "loss": 0.1982,
+      "step": 11963
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.4216099638781452,
+      "learning_rate": 2.6581499134285603e-06,
+      "loss": 0.1778,
+      "step": 11964
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.2392704138629902,
+      "learning_rate": 2.6573540292815484e-06,
+      "loss": 0.2233,
+      "step": 11965
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.199429689177277,
+      "learning_rate": 2.65655822117817e-06,
+      "loss": 0.1406,
+      "step": 11966
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.8140515477443755,
+      "learning_rate": 2.655762489144259e-06,
+      "loss": 0.1854,
+      "step": 11967
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.9195878218083381,
+      "learning_rate": 2.6549668332056455e-06,
+      "loss": 0.1688,
+      "step": 11968
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.3369096932851814,
+      "learning_rate": 2.6541712533881594e-06,
+      "loss": 0.2293,
+      "step": 11969
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.6823597742668994,
+      "learning_rate": 2.6533757497176206e-06,
+      "loss": 0.2186,
+      "step": 11970
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.3632714454276451,
+      "learning_rate": 2.652580322219855e-06,
+      "loss": 0.2083,
+      "step": 11971
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.5336469360239122,
+      "learning_rate": 2.6517849709206834e-06,
+      "loss": 0.196,
+      "step": 11972
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.4522261608388058,
+      "learning_rate": 2.650989695845921e-06,
+      "loss": 0.1299,
+      "step": 11973
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.865073323911365,
+      "learning_rate": 2.650194497021385e-06,
+      "loss": 0.1912,
+      "step": 11974
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.076404788641589,
+      "learning_rate": 2.649399374472885e-06,
+      "loss": 0.2143,
+      "step": 11975
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.3215084330832387,
+      "learning_rate": 2.648604328226233e-06,
+      "loss": 0.212,
+      "step": 11976
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.327913576015175,
+      "learning_rate": 2.647809358307237e-06,
+      "loss": 0.1843,
+      "step": 11977
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.3579199720986046,
+      "learning_rate": 2.647014464741704e-06,
+      "loss": 0.2242,
+      "step": 11978
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.6603609923725533,
+      "learning_rate": 2.646219647555432e-06,
+      "loss": 0.256,
+      "step": 11979
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.8152411795468708,
+      "learning_rate": 2.6454249067742237e-06,
+      "loss": 0.1692,
+      "step": 11980
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.36043650190119,
+      "learning_rate": 2.644630242423877e-06,
+      "loss": 0.1957,
+      "step": 11981
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.244942674561835,
+      "learning_rate": 2.6438356545301875e-06,
+      "loss": 0.153,
+      "step": 11982
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.089171171307947,
+      "learning_rate": 2.6430411431189495e-06,
+      "loss": 0.1722,
+      "step": 11983
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.2019814097114145,
+      "learning_rate": 2.642246708215949e-06,
+      "loss": 0.1557,
+      "step": 11984
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.2095560607982192,
+      "learning_rate": 2.6414523498469756e-06,
+      "loss": 0.1609,
+      "step": 11985
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.0282803878206708,
+      "learning_rate": 2.640658068037816e-06,
+      "loss": 0.1543,
+      "step": 11986
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.6700462228087372,
+      "learning_rate": 2.639863862814255e-06,
+      "loss": 0.2459,
+      "step": 11987
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.2844692933490924,
+      "learning_rate": 2.6390697342020665e-06,
+      "loss": 0.2001,
+      "step": 11988
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.0576406072980553,
+      "learning_rate": 2.6382756822270327e-06,
+      "loss": 0.1354,
+      "step": 11989
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.1502264323391402,
+      "learning_rate": 2.6374817069149308e-06,
+      "loss": 0.1319,
+      "step": 11990
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.4585527013796171,
+      "learning_rate": 2.6366878082915288e-06,
+      "loss": 0.156,
+      "step": 11991
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.1589406834856772,
+      "learning_rate": 2.6358939863826014e-06,
+      "loss": 0.2135,
+      "step": 11992
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.9895231356213549,
+      "learning_rate": 2.635100241213913e-06,
+      "loss": 0.1483,
+      "step": 11993
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 3.1644351315949177,
+      "learning_rate": 2.6343065728112306e-06,
+      "loss": 0.2229,
+      "step": 11994
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.410684676769322,
+      "learning_rate": 2.633512981200317e-06,
+      "loss": 0.2099,
+      "step": 11995
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.5892460142262363,
+      "learning_rate": 2.6327194664069357e-06,
+      "loss": 0.2331,
+      "step": 11996
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.4069076166053747,
+      "learning_rate": 2.631926028456839e-06,
+      "loss": 0.2076,
+      "step": 11997
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.4287583870612384,
+      "learning_rate": 2.6311326673757854e-06,
+      "loss": 0.2102,
+      "step": 11998
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.5115702270273246,
+      "learning_rate": 2.6303393831895273e-06,
+      "loss": 0.2403,
+      "step": 11999
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.533072860918216,
+      "learning_rate": 2.6295461759238155e-06,
+      "loss": 0.1807,
+      "step": 12000
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.7130301015302951,
+      "learning_rate": 2.6287530456044003e-06,
+      "loss": 0.2556,
+      "step": 12001
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.8282649117099923,
+      "learning_rate": 2.6279599922570226e-06,
+      "loss": 0.1582,
+      "step": 12002
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.6502605680425797,
+      "learning_rate": 2.627167015907428e-06,
+      "loss": 0.238,
+      "step": 12003
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.3013347101259634,
+      "learning_rate": 2.626374116581356e-06,
+      "loss": 0.209,
+      "step": 12004
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.6816865779610342,
+      "learning_rate": 2.6255812943045465e-06,
+      "loss": 0.2297,
+      "step": 12005
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.9966264085921177,
+      "learning_rate": 2.624788549102733e-06,
+      "loss": 0.1541,
+      "step": 12006
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.2498032146766498,
+      "learning_rate": 2.6239958810016504e-06,
+      "loss": 0.1764,
+      "step": 12007
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.472310424036058,
+      "learning_rate": 2.6232032900270254e-06,
+      "loss": 0.208,
+      "step": 12008
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.2789322049414829,
+      "learning_rate": 2.622410776204589e-06,
+      "loss": 0.2187,
+      "step": 12009
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.360101861759202,
+      "learning_rate": 2.621618339560068e-06,
+      "loss": 0.1465,
+      "step": 12010
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.3756803514566,
+      "learning_rate": 2.620825980119181e-06,
+      "loss": 0.2056,
+      "step": 12011
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.405351383467828,
+      "learning_rate": 2.620033697907651e-06,
+      "loss": 0.1968,
+      "step": 12012
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.0206698640352403,
+      "learning_rate": 2.6192414929511956e-06,
+      "loss": 0.1473,
+      "step": 12013
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.3223797961576362,
+      "learning_rate": 2.6184493652755326e-06,
+      "loss": 0.1792,
+      "step": 12014
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.134114786258275,
+      "learning_rate": 2.61765731490637e-06,
+      "loss": 0.1903,
+      "step": 12015
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.1015203751970886,
+      "learning_rate": 2.6168653418694216e-06,
+      "loss": 0.1459,
+      "step": 12016
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.5685628804090375,
+      "learning_rate": 2.6160734461903946e-06,
+      "loss": 0.1764,
+      "step": 12017
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.9539106484514035,
+      "learning_rate": 2.615281627894994e-06,
+      "loss": 0.172,
+      "step": 12018
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.5818774172083556,
+      "learning_rate": 2.614489887008925e-06,
+      "loss": 0.2215,
+      "step": 12019
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.206071601779957,
+      "learning_rate": 2.613698223557885e-06,
+      "loss": 0.2042,
+      "step": 12020
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.2584278070698427,
+      "learning_rate": 2.612906637567572e-06,
+      "loss": 0.1847,
+      "step": 12021
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.5480208126880373,
+      "learning_rate": 2.6121151290636826e-06,
+      "loss": 0.1763,
+      "step": 12022
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.2556921579146312,
+      "learning_rate": 2.6113236980719115e-06,
+      "loss": 0.1924,
+      "step": 12023
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.349275829320465,
+      "learning_rate": 2.6105323446179443e-06,
+      "loss": 0.1894,
+      "step": 12024
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.444107692756778,
+      "learning_rate": 2.6097410687274726e-06,
+      "loss": 0.181,
+      "step": 12025
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.3573299294487533,
+      "learning_rate": 2.6089498704261795e-06,
+      "loss": 0.2267,
+      "step": 12026
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.3899629870896708,
+      "learning_rate": 2.6081587497397474e-06,
+      "loss": 0.1923,
+      "step": 12027
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.4330316137879924,
+      "learning_rate": 2.6073677066938608e-06,
+      "loss": 0.1971,
+      "step": 12028
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.9058453770584505,
+      "learning_rate": 2.6065767413141916e-06,
+      "loss": 0.1401,
+      "step": 12029
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.9469083230131984,
+      "learning_rate": 2.605785853626418e-06,
+      "loss": 0.177,
+      "step": 12030
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.9697088987157341,
+      "learning_rate": 2.604995043656212e-06,
+      "loss": 0.1551,
+      "step": 12031
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.6108354821078157,
+      "learning_rate": 2.604204311429247e-06,
+      "loss": 0.195,
+      "step": 12032
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.2010857093733451,
+      "learning_rate": 2.603413656971185e-06,
+      "loss": 0.1462,
+      "step": 12033
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.8995184635865742,
+      "learning_rate": 2.602623080307694e-06,
+      "loss": 0.2047,
+      "step": 12034
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.669587421944958,
+      "learning_rate": 2.6018325814644364e-06,
+      "loss": 0.2067,
+      "step": 12035
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.381786752640976,
+      "learning_rate": 2.601042160467072e-06,
+      "loss": 0.192,
+      "step": 12036
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.4069892370802617,
+      "learning_rate": 2.6002518173412606e-06,
+      "loss": 0.2296,
+      "step": 12037
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.3190880125317326,
+      "learning_rate": 2.599461552112654e-06,
+      "loss": 0.171,
+      "step": 12038
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.2679582446139674,
+      "learning_rate": 2.5986713648069053e-06,
+      "loss": 0.1906,
+      "step": 12039
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.0043505051641364,
+      "learning_rate": 2.597881255449665e-06,
+      "loss": 0.2317,
+      "step": 12040
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.5032461179048424,
+      "learning_rate": 2.5970912240665815e-06,
+      "loss": 0.2566,
+      "step": 12041
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.6928907447732862,
+      "learning_rate": 2.596301270683298e-06,
+      "loss": 0.2247,
+      "step": 12042
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.3412419050182147,
+      "learning_rate": 2.5955113953254583e-06,
+      "loss": 0.2145,
+      "step": 12043
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.7921683589421095,
+      "learning_rate": 2.5947215980187e-06,
+      "loss": 0.1722,
+      "step": 12044
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.0969437835732478,
+      "learning_rate": 2.593931878788662e-06,
+      "loss": 0.2049,
+      "step": 12045
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.4444269302520856,
+      "learning_rate": 2.5931422376609805e-06,
+      "loss": 0.1888,
+      "step": 12046
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.9294631370722783,
+      "learning_rate": 2.592352674661284e-06,
+      "loss": 0.147,
+      "step": 12047
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.2189177373240407,
+      "learning_rate": 2.5915631898152045e-06,
+      "loss": 0.195,
+      "step": 12048
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.285849645053896,
+      "learning_rate": 2.590773783148368e-06,
+      "loss": 0.1794,
+      "step": 12049
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.2283610700049592,
+      "learning_rate": 2.589984454686403e-06,
+      "loss": 0.1452,
+      "step": 12050
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.5695879104283554,
+      "learning_rate": 2.5891952044549262e-06,
+      "loss": 0.2085,
+      "step": 12051
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.4131943635396391,
+      "learning_rate": 2.588406032479559e-06,
+      "loss": 0.2366,
+      "step": 12052
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.3478723933168082,
+      "learning_rate": 2.5876169387859185e-06,
+      "loss": 0.1918,
+      "step": 12053
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.0245289915361597,
+      "learning_rate": 2.5868279233996197e-06,
+      "loss": 0.1743,
+      "step": 12054
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.2912567075165562,
+      "learning_rate": 2.5860389863462765e-06,
+      "loss": 0.1792,
+      "step": 12055
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.4476343295295173,
+      "learning_rate": 2.5852501276514934e-06,
+      "loss": 0.1915,
+      "step": 12056
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.1315415026568896,
+      "learning_rate": 2.5844613473408804e-06,
+      "loss": 0.2355,
+      "step": 12057
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.2301405150992688,
+      "learning_rate": 2.5836726454400405e-06,
+      "loss": 0.2193,
+      "step": 12058
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.1820883572606917,
+      "learning_rate": 2.582884021974578e-06,
+      "loss": 0.1992,
+      "step": 12059
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.2305311520208768,
+      "learning_rate": 2.5820954769700885e-06,
+      "loss": 0.174,
+      "step": 12060
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.5509322450861243,
+      "learning_rate": 2.581307010452172e-06,
+      "loss": 0.2437,
+      "step": 12061
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.6483606980498062,
+      "learning_rate": 2.5805186224464187e-06,
+      "loss": 0.1968,
+      "step": 12062
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.4952010187906755,
+      "learning_rate": 2.5797303129784225e-06,
+      "loss": 0.1909,
+      "step": 12063
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.0168127780451774,
+      "learning_rate": 2.5789420820737736e-06,
+      "loss": 0.1631,
+      "step": 12064
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.160340670175095,
+      "learning_rate": 2.5781539297580553e-06,
+      "loss": 0.1772,
+      "step": 12065
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.8401303024250948,
+      "learning_rate": 2.577365856056853e-06,
+      "loss": 0.1649,
+      "step": 12066
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.4714553221529947,
+      "learning_rate": 2.576577860995748e-06,
+      "loss": 0.243,
+      "step": 12067
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.354794601201827,
+      "learning_rate": 2.5757899446003214e-06,
+      "loss": 0.2037,
+      "step": 12068
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.1175893997668993,
+      "learning_rate": 2.575002106896145e-06,
+      "loss": 0.1574,
+      "step": 12069
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.1674681391238448,
+      "learning_rate": 2.5742143479087955e-06,
+      "loss": 0.1775,
+      "step": 12070
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.2139077337096122,
+      "learning_rate": 2.573426667663842e-06,
+      "loss": 0.192,
+      "step": 12071
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.7941192683820242,
+      "learning_rate": 2.572639066186855e-06,
+      "loss": 0.2274,
+      "step": 12072
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.22571207286509,
+      "learning_rate": 2.571851543503402e-06,
+      "loss": 0.1989,
+      "step": 12073
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.4110689722951764,
+      "learning_rate": 2.571064099639041e-06,
+      "loss": 0.1697,
+      "step": 12074
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.269733671043177,
+      "learning_rate": 2.570276734619337e-06,
+      "loss": 0.1818,
+      "step": 12075
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.1943714420409335,
+      "learning_rate": 2.5694894484698475e-06,
+      "loss": 0.1818,
+      "step": 12076
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.2224231168641049,
+      "learning_rate": 2.56870224121613e-06,
+      "loss": 0.1715,
+      "step": 12077
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.2538903434433757,
+      "learning_rate": 2.5679151128837334e-06,
+      "loss": 0.1609,
+      "step": 12078
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.503066244483867,
+      "learning_rate": 2.567128063498213e-06,
+      "loss": 0.1728,
+      "step": 12079
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.2291325322627136,
+      "learning_rate": 2.566341093085113e-06,
+      "loss": 0.2215,
+      "step": 12080
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.1552063671774107,
+      "learning_rate": 2.5655542016699796e-06,
+      "loss": 0.1972,
+      "step": 12081
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.257265884379476,
+      "learning_rate": 2.5647673892783598e-06,
+      "loss": 0.1937,
+      "step": 12082
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.172192508282052,
+      "learning_rate": 2.563980655935788e-06,
+      "loss": 0.1978,
+      "step": 12083
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.3801335540942934,
+      "learning_rate": 2.563194001667806e-06,
+      "loss": 0.1315,
+      "step": 12084
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.5192268503073771,
+      "learning_rate": 2.5624074264999475e-06,
+      "loss": 0.2072,
+      "step": 12085
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.1842573873707434,
+      "learning_rate": 2.5616209304577477e-06,
+      "loss": 0.1529,
+      "step": 12086
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.3351155144831575,
+      "learning_rate": 2.5608345135667322e-06,
+      "loss": 0.195,
+      "step": 12087
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.3973243268068691,
+      "learning_rate": 2.560048175852431e-06,
+      "loss": 0.2152,
+      "step": 12088
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.5517686971927265,
+      "learning_rate": 2.55926191734037e-06,
+      "loss": 0.1439,
+      "step": 12089
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.1446990550845926,
+      "learning_rate": 2.55847573805607e-06,
+      "loss": 0.1857,
+      "step": 12090
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.3762335689553016,
+      "learning_rate": 2.557689638025054e-06,
+      "loss": 0.1962,
+      "step": 12091
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.2123970710679686,
+      "learning_rate": 2.5569036172728335e-06,
+      "loss": 0.1478,
+      "step": 12092
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.8700630372362744,
+      "learning_rate": 2.5561176758249273e-06,
+      "loss": 0.1549,
+      "step": 12093
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.3333771805209438,
+      "learning_rate": 2.5553318137068473e-06,
+      "loss": 0.1875,
+      "step": 12094
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.267734529694228,
+      "learning_rate": 2.554546030944104e-06,
+      "loss": 0.1818,
+      "step": 12095
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.452602837431901,
+      "learning_rate": 2.5537603275622e-06,
+      "loss": 0.1849,
+      "step": 12096
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.3882411498040634,
+      "learning_rate": 2.5529747035866455e-06,
+      "loss": 0.209,
+      "step": 12097
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.0603298681811484,
+      "learning_rate": 2.5521891590429366e-06,
+      "loss": 0.1889,
+      "step": 12098
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.8538808848530488,
+      "learning_rate": 2.5514036939565755e-06,
+      "loss": 0.1756,
+      "step": 12099
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.0585989459286809,
+      "learning_rate": 2.55061830835306e-06,
+      "loss": 0.1398,
+      "step": 12100
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.322822602812688,
+      "learning_rate": 2.5498330022578815e-06,
+      "loss": 0.1787,
+      "step": 12101
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.31879138226236,
+      "learning_rate": 2.549047775696533e-06,
+      "loss": 0.1799,
+      "step": 12102
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.0230153865855007,
+      "learning_rate": 2.5482626286945024e-06,
+      "loss": 0.179,
+      "step": 12103
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.3343862037277683,
+      "learning_rate": 2.5474775612772783e-06,
+      "loss": 0.155,
+      "step": 12104
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.192478397921469,
+      "learning_rate": 2.5466925734703415e-06,
+      "loss": 0.1641,
+      "step": 12105
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.0983214992681354,
+      "learning_rate": 2.5459076652991744e-06,
+      "loss": 0.1763,
+      "step": 12106
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.0629791308435412,
+      "learning_rate": 2.5451228367892554e-06,
+      "loss": 0.1831,
+      "step": 12107
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.1115837138470563,
+      "learning_rate": 2.544338087966061e-06,
+      "loss": 0.1956,
+      "step": 12108
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.200983491717226,
+      "learning_rate": 2.5435534188550665e-06,
+      "loss": 0.1758,
+      "step": 12109
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.3488311290293233,
+      "learning_rate": 2.542768829481739e-06,
+      "loss": 0.1675,
+      "step": 12110
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.6435135616930188,
+      "learning_rate": 2.5419843198715473e-06,
+      "loss": 0.2422,
+      "step": 12111
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.5595226722468394,
+      "learning_rate": 2.5411998900499586e-06,
+      "loss": 0.1652,
+      "step": 12112
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.3298045982054003,
+      "learning_rate": 2.5404155400424367e-06,
+      "loss": 0.1741,
+      "step": 12113
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.0761991348552629,
+      "learning_rate": 2.5396312698744396e-06,
+      "loss": 0.1658,
+      "step": 12114
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.305554490210164,
+      "learning_rate": 2.5388470795714283e-06,
+      "loss": 0.1394,
+      "step": 12115
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.7051983907786845,
+      "learning_rate": 2.538062969158854e-06,
+      "loss": 0.1624,
+      "step": 12116
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 6.365293349782651,
+      "learning_rate": 2.537278938662171e-06,
+      "loss": 0.2217,
+      "step": 12117
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.490109449677456,
+      "learning_rate": 2.536494988106832e-06,
+      "loss": 0.1565,
+      "step": 12118
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.311095777988968,
+      "learning_rate": 2.53571111751828e-06,
+      "loss": 0.1933,
+      "step": 12119
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.3402090797939126,
+      "learning_rate": 2.534927326921963e-06,
+      "loss": 0.1838,
+      "step": 12120
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.4374764896626395,
+      "learning_rate": 2.5341436163433218e-06,
+      "loss": 0.1597,
+      "step": 12121
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.4149544179174491,
+      "learning_rate": 2.533359985807799e-06,
+      "loss": 0.2055,
+      "step": 12122
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.3946119814032012,
+      "learning_rate": 2.5325764353408277e-06,
+      "loss": 0.2111,
+      "step": 12123
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.2648595272915686,
+      "learning_rate": 2.531792964967844e-06,
+      "loss": 0.1925,
+      "step": 12124
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.9966207344430863,
+      "learning_rate": 2.5310095747142794e-06,
+      "loss": 0.1661,
+      "step": 12125
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.39325712538382,
+      "learning_rate": 2.530226264605565e-06,
+      "loss": 0.1991,
+      "step": 12126
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.3843437757479793,
+      "learning_rate": 2.5294430346671273e-06,
+      "loss": 0.1796,
+      "step": 12127
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.5128083356151065,
+      "learning_rate": 2.5286598849243876e-06,
+      "loss": 0.1738,
+      "step": 12128
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.3521271997789897,
+      "learning_rate": 2.5278768154027687e-06,
+      "loss": 0.221,
+      "step": 12129
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.4529424906526915,
+      "learning_rate": 2.5270938261276912e-06,
+      "loss": 0.1723,
+      "step": 12130
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.4632965767338286,
+      "learning_rate": 2.5263109171245714e-06,
+      "loss": 0.1958,
+      "step": 12131
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.5111122146707403,
+      "learning_rate": 2.5255280884188193e-06,
+      "loss": 0.1958,
+      "step": 12132
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.6566561254969343,
+      "learning_rate": 2.5247453400358505e-06,
+      "loss": 0.1501,
+      "step": 12133
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.2787365800259634,
+      "learning_rate": 2.5239626720010702e-06,
+      "loss": 0.2216,
+      "step": 12134
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.3878506352734692,
+      "learning_rate": 2.5231800843398847e-06,
+      "loss": 0.2142,
+      "step": 12135
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.5314375363553738,
+      "learning_rate": 2.5223975770777e-06,
+      "loss": 0.1585,
+      "step": 12136
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.250187174373219,
+      "learning_rate": 2.521615150239912e-06,
+      "loss": 0.2024,
+      "step": 12137
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.3038111073162075,
+      "learning_rate": 2.5208328038519226e-06,
+      "loss": 0.1446,
+      "step": 12138
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.3731731123432473,
+      "learning_rate": 2.5200505379391256e-06,
+      "loss": 0.1732,
+      "step": 12139
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.398779448999478,
+      "learning_rate": 2.5192683525269156e-06,
+      "loss": 0.1545,
+      "step": 12140
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.1898038657072956,
+      "learning_rate": 2.5184862476406804e-06,
+      "loss": 0.1778,
+      "step": 12141
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.3705641438254583,
+      "learning_rate": 2.517704223305808e-06,
+      "loss": 0.2104,
+      "step": 12142
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.6458564683237829,
+      "learning_rate": 2.516922279547684e-06,
+      "loss": 0.1929,
+      "step": 12143
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.8089594296238949,
+      "learning_rate": 2.51614041639169e-06,
+      "loss": 0.2353,
+      "step": 12144
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.2261633086718,
+      "learning_rate": 2.51535863386321e-06,
+      "loss": 0.1913,
+      "step": 12145
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.398858928580941,
+      "learning_rate": 2.5145769319876145e-06,
+      "loss": 0.1836,
+      "step": 12146
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.4744543410901403,
+      "learning_rate": 2.5137953107902814e-06,
+      "loss": 0.1857,
+      "step": 12147
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.267877464040434,
+      "learning_rate": 2.513013770296582e-06,
+      "loss": 0.2017,
+      "step": 12148
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.5311217351637754,
+      "learning_rate": 2.5122323105318867e-06,
+      "loss": 0.1996,
+      "step": 12149
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.21127941165268,
+      "learning_rate": 2.5114509315215605e-06,
+      "loss": 0.2037,
+      "step": 12150
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.7431966634314413,
+      "learning_rate": 2.5106696332909696e-06,
+      "loss": 0.2596,
+      "step": 12151
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.085447495871415,
+      "learning_rate": 2.5098884158654723e-06,
+      "loss": 0.1685,
+      "step": 12152
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.6833608537039757,
+      "learning_rate": 2.5091072792704285e-06,
+      "loss": 0.1945,
+      "step": 12153
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.4020081307075778,
+      "learning_rate": 2.5083262235311965e-06,
+      "loss": 0.1569,
+      "step": 12154
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.3493906939779017,
+      "learning_rate": 2.507545248673127e-06,
+      "loss": 0.1767,
+      "step": 12155
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.9516910935145677,
+      "learning_rate": 2.5067643547215712e-06,
+      "loss": 0.2245,
+      "step": 12156
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.1598447998146961,
+      "learning_rate": 2.5059835417018783e-06,
+      "loss": 0.1844,
+      "step": 12157
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.1095228499629215,
+      "learning_rate": 2.505202809639396e-06,
+      "loss": 0.1679,
+      "step": 12158
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.5860924739003885,
+      "learning_rate": 2.504422158559463e-06,
+      "loss": 0.2554,
+      "step": 12159
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.0831670411871797,
+      "learning_rate": 2.5036415884874223e-06,
+      "loss": 0.208,
+      "step": 12160
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.0302846935073162,
+      "learning_rate": 2.50286109944861e-06,
+      "loss": 0.1376,
+      "step": 12161
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.4376006324415416,
+      "learning_rate": 2.502080691468363e-06,
+      "loss": 0.2187,
+      "step": 12162
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.006135306838201,
+      "learning_rate": 2.501300364572016e-06,
+      "loss": 0.2049,
+      "step": 12163
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.4719454089411417,
+      "learning_rate": 2.5005201187848924e-06,
+      "loss": 0.221,
+      "step": 12164
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.3440376184622125,
+      "learning_rate": 2.499739954132324e-06,
+      "loss": 0.2199,
+      "step": 12165
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.5670292535758303,
+      "learning_rate": 2.4989598706396344e-06,
+      "loss": 0.2472,
+      "step": 12166
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.1282912562430703,
+      "learning_rate": 2.4981798683321473e-06,
+      "loss": 0.1631,
+      "step": 12167
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.229883387763725,
+      "learning_rate": 2.497399947235179e-06,
+      "loss": 0.2231,
+      "step": 12168
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.2077669537285485,
+      "learning_rate": 2.4966201073740486e-06,
+      "loss": 0.2028,
+      "step": 12169
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.3167339016027904,
+      "learning_rate": 2.495840348774068e-06,
+      "loss": 0.1992,
+      "step": 12170
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.9060063440911785,
+      "learning_rate": 2.495060671460549e-06,
+      "loss": 0.2424,
+      "step": 12171
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.3542052654731884,
+      "learning_rate": 2.494281075458803e-06,
+      "loss": 0.2293,
+      "step": 12172
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.4759927185151651,
+      "learning_rate": 2.4935015607941324e-06,
+      "loss": 0.198,
+      "step": 12173
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.7665811232253575,
+      "learning_rate": 2.492722127491843e-06,
+      "loss": 0.104,
+      "step": 12174
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.2158714590076956,
+      "learning_rate": 2.491942775577234e-06,
+      "loss": 0.1311,
+      "step": 12175
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.5527539521908522,
+      "learning_rate": 2.491163505075608e-06,
+      "loss": 0.1907,
+      "step": 12176
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.7983693170423884,
+      "learning_rate": 2.4903843160122542e-06,
+      "loss": 0.2163,
+      "step": 12177
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 3.5590205685095406,
+      "learning_rate": 2.4896052084124696e-06,
+      "loss": 0.2303,
+      "step": 12178
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.4414857164801416,
+      "learning_rate": 2.4888261823015426e-06,
+      "loss": 0.2052,
+      "step": 12179
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.3898086830762986,
+      "learning_rate": 2.488047237704762e-06,
+      "loss": 0.1307,
+      "step": 12180
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.491641269188805,
+      "learning_rate": 2.4872683746474145e-06,
+      "loss": 0.1862,
+      "step": 12181
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.5926603023313204,
+      "learning_rate": 2.486489593154779e-06,
+      "loss": 0.1897,
+      "step": 12182
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.384345288097759,
+      "learning_rate": 2.485710893252136e-06,
+      "loss": 0.2178,
+      "step": 12183
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.3391083375659192,
+      "learning_rate": 2.4849322749647643e-06,
+      "loss": 0.1875,
+      "step": 12184
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.7856554464455665,
+      "learning_rate": 2.484153738317938e-06,
+      "loss": 0.2153,
+      "step": 12185
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.0363374163607553,
+      "learning_rate": 2.483375283336927e-06,
+      "loss": 0.2135,
+      "step": 12186
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.243815518814839,
+      "learning_rate": 2.4825969100470025e-06,
+      "loss": 0.1543,
+      "step": 12187
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.4026472937319339,
+      "learning_rate": 2.481818618473429e-06,
+      "loss": 0.195,
+      "step": 12188
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.3941165991476756,
+      "learning_rate": 2.4810404086414704e-06,
+      "loss": 0.2621,
+      "step": 12189
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.5674252798441513,
+      "learning_rate": 2.480262280576391e-06,
+      "loss": 0.2297,
+      "step": 12190
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.3135347601607905,
+      "learning_rate": 2.479484234303445e-06,
+      "loss": 0.1605,
+      "step": 12191
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.2180029707059665,
+      "learning_rate": 2.4787062698478896e-06,
+      "loss": 0.1822,
+      "step": 12192
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.0408605366996497,
+      "learning_rate": 2.477928387234979e-06,
+      "loss": 0.19,
+      "step": 12193
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.0122670819677675,
+      "learning_rate": 2.4771505864899652e-06,
+      "loss": 0.1684,
+      "step": 12194
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.657759019490608,
+      "learning_rate": 2.476372867638092e-06,
+      "loss": 0.2378,
+      "step": 12195
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.5502719571349244,
+      "learning_rate": 2.475595230704607e-06,
+      "loss": 0.2087,
+      "step": 12196
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.8790084351096,
+      "learning_rate": 2.4748176757147517e-06,
+      "loss": 0.1774,
+      "step": 12197
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.2551591917630567,
+      "learning_rate": 2.4740402026937677e-06,
+      "loss": 0.2031,
+      "step": 12198
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.27627783523136,
+      "learning_rate": 2.4732628116668927e-06,
+      "loss": 0.1428,
+      "step": 12199
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.5860848992538292,
+      "learning_rate": 2.472485502659358e-06,
+      "loss": 0.1669,
+      "step": 12200
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.5179887017368638,
+      "learning_rate": 2.471708275696397e-06,
+      "loss": 0.2024,
+      "step": 12201
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.678819933554486,
+      "learning_rate": 2.4709311308032396e-06,
+      "loss": 0.2078,
+      "step": 12202
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.371311181371827,
+      "learning_rate": 2.470154068005114e-06,
+      "loss": 0.162,
+      "step": 12203
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.6166217339513,
+      "learning_rate": 2.4693770873272398e-06,
+      "loss": 0.1937,
+      "step": 12204
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.2317809419707786,
+      "learning_rate": 2.468600188794843e-06,
+      "loss": 0.1377,
+      "step": 12205
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.1016257788697483,
+      "learning_rate": 2.4678233724331375e-06,
+      "loss": 0.1934,
+      "step": 12206
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.6350569197753182,
+      "learning_rate": 2.4670466382673412e-06,
+      "loss": 0.1847,
+      "step": 12207
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.8535669719630394,
+      "learning_rate": 2.4662699863226698e-06,
+      "loss": 0.2425,
+      "step": 12208
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.3476419862738702,
+      "learning_rate": 2.4654934166243295e-06,
+      "loss": 0.1895,
+      "step": 12209
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.5214283630481606,
+      "learning_rate": 2.4647169291975304e-06,
+      "loss": 0.2283,
+      "step": 12210
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.2505565804892333,
+      "learning_rate": 2.463940524067478e-06,
+      "loss": 0.1539,
+      "step": 12211
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.5544921857745433,
+      "learning_rate": 2.4631642012593742e-06,
+      "loss": 0.2094,
+      "step": 12212
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.9997787519542447,
+      "learning_rate": 2.462387960798421e-06,
+      "loss": 0.1649,
+      "step": 12213
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.216172672059989,
+      "learning_rate": 2.4616118027098124e-06,
+      "loss": 0.1608,
+      "step": 12214
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.1323341889002103,
+      "learning_rate": 2.460835727018744e-06,
+      "loss": 0.1781,
+      "step": 12215
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.502718647298671,
+      "learning_rate": 2.4600597337504084e-06,
+      "loss": 0.189,
+      "step": 12216
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.291415558012403,
+      "learning_rate": 2.459283822929996e-06,
+      "loss": 0.1795,
+      "step": 12217
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.9586632142988691,
+      "learning_rate": 2.4585079945826895e-06,
+      "loss": 0.2678,
+      "step": 12218
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.30187640898322,
+      "learning_rate": 2.4577322487336752e-06,
+      "loss": 0.2166,
+      "step": 12219
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.1384694163172937,
+      "learning_rate": 2.4569565854081347e-06,
+      "loss": 0.2025,
+      "step": 12220
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.6099419474797811,
+      "learning_rate": 2.4561810046312466e-06,
+      "loss": 0.182,
+      "step": 12221
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.1126325390387495,
+      "learning_rate": 2.4554055064281846e-06,
+      "loss": 0.1693,
+      "step": 12222
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.0982271013304088,
+      "learning_rate": 2.454630090824125e-06,
+      "loss": 0.1645,
+      "step": 12223
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.028839943674749,
+      "learning_rate": 2.4538547578442344e-06,
+      "loss": 0.1644,
+      "step": 12224
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.5243481090874893,
+      "learning_rate": 2.4530795075136824e-06,
+      "loss": 0.1996,
+      "step": 12225
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.2156555733232524,
+      "learning_rate": 2.452304339857636e-06,
+      "loss": 0.2678,
+      "step": 12226
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.0598785514578584,
+      "learning_rate": 2.4515292549012543e-06,
+      "loss": 0.163,
+      "step": 12227
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.9911254244594312,
+      "learning_rate": 2.450754252669698e-06,
+      "loss": 0.1991,
+      "step": 12228
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.5916260539881024,
+      "learning_rate": 2.449979333188125e-06,
+      "loss": 0.2115,
+      "step": 12229
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.633737322711428,
+      "learning_rate": 2.4492044964816886e-06,
+      "loss": 0.2292,
+      "step": 12230
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.6462041270179864,
+      "learning_rate": 2.448429742575543e-06,
+      "loss": 0.2035,
+      "step": 12231
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.2799433348823674,
+      "learning_rate": 2.4476550714948337e-06,
+      "loss": 0.1839,
+      "step": 12232
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.2328232173671392,
+      "learning_rate": 2.4468804832647087e-06,
+      "loss": 0.1632,
+      "step": 12233
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.9487914234160116,
+      "learning_rate": 2.4461059779103107e-06,
+      "loss": 0.108,
+      "step": 12234
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.2544535277418791,
+      "learning_rate": 2.4453315554567836e-06,
+      "loss": 0.186,
+      "step": 12235
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.1358342645898576,
+      "learning_rate": 2.4445572159292614e-06,
+      "loss": 0.1761,
+      "step": 12236
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.2678882531018076,
+      "learning_rate": 2.443782959352881e-06,
+      "loss": 0.1977,
+      "step": 12237
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.5615068187867056,
+      "learning_rate": 2.443008785752776e-06,
+      "loss": 0.2086,
+      "step": 12238
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.239302391651262,
+      "learning_rate": 2.442234695154078e-06,
+      "loss": 0.1306,
+      "step": 12239
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.4916526874583367,
+      "learning_rate": 2.4414606875819113e-06,
+      "loss": 0.2243,
+      "step": 12240
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.0785031312068942,
+      "learning_rate": 2.4406867630614033e-06,
+      "loss": 0.1465,
+      "step": 12241
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.8660751180384119,
+      "learning_rate": 2.4399129216176724e-06,
+      "loss": 0.1617,
+      "step": 12242
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.4422183656540937,
+      "learning_rate": 2.439139163275841e-06,
+      "loss": 0.1934,
+      "step": 12243
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.0105499158414633,
+      "learning_rate": 2.438365488061026e-06,
+      "loss": 0.1389,
+      "step": 12244
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.259341137362164,
+      "learning_rate": 2.437591895998339e-06,
+      "loss": 0.1691,
+      "step": 12245
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.2904883908163824,
+      "learning_rate": 2.4368183871128926e-06,
+      "loss": 0.1744,
+      "step": 12246
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.1571040285803549,
+      "learning_rate": 2.4360449614297955e-06,
+      "loss": 0.2077,
+      "step": 12247
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.7120453196124394,
+      "learning_rate": 2.4352716189741526e-06,
+      "loss": 0.1312,
+      "step": 12248
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.6418185025482657,
+      "learning_rate": 2.4344983597710704e-06,
+      "loss": 0.1827,
+      "step": 12249
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.1392438704907768,
+      "learning_rate": 2.4337251838456443e-06,
+      "loss": 0.1615,
+      "step": 12250
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.0736670655339422,
+      "learning_rate": 2.4329520912229744e-06,
+      "loss": 0.1284,
+      "step": 12251
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.1029835675837485,
+      "learning_rate": 2.432179081928156e-06,
+      "loss": 0.169,
+      "step": 12252
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.2390917763732907,
+      "learning_rate": 2.4314061559862836e-06,
+      "loss": 0.2022,
+      "step": 12253
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.7717115515214406,
+      "learning_rate": 2.4306333134224413e-06,
+      "loss": 0.1768,
+      "step": 12254
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 4.316553277320871,
+      "learning_rate": 2.4298605542617197e-06,
+      "loss": 0.1715,
+      "step": 12255
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.4814539225761247,
+      "learning_rate": 2.429087878529205e-06,
+      "loss": 0.1718,
+      "step": 12256
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.2315042528511921,
+      "learning_rate": 2.4283152862499737e-06,
+      "loss": 0.1814,
+      "step": 12257
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.754810227732942,
+      "learning_rate": 2.4275427774491067e-06,
+      "loss": 0.2093,
+      "step": 12258
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.2696615651078176,
+      "learning_rate": 2.4267703521516827e-06,
+      "loss": 0.1777,
+      "step": 12259
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.9864772759397603,
+      "learning_rate": 2.425998010382771e-06,
+      "loss": 0.1461,
+      "step": 12260
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.2885838458014909,
+      "learning_rate": 2.425225752167444e-06,
+      "loss": 0.1854,
+      "step": 12261
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.4098279787534562,
+      "learning_rate": 2.4244535775307715e-06,
+      "loss": 0.2161,
+      "step": 12262
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.036367733108971,
+      "learning_rate": 2.423681486497815e-06,
+      "loss": 0.1383,
+      "step": 12263
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.5537645317535123,
+      "learning_rate": 2.4229094790936385e-06,
+      "loss": 0.1649,
+      "step": 12264
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.9837177860445907,
+      "learning_rate": 2.422137555343303e-06,
+      "loss": 0.1955,
+      "step": 12265
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.8996769006408851,
+      "learning_rate": 2.4213657152718646e-06,
+      "loss": 0.1696,
+      "step": 12266
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.144723023371879,
+      "learning_rate": 2.4205939589043797e-06,
+      "loss": 0.1714,
+      "step": 12267
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.2220260257265412,
+      "learning_rate": 2.419822286265896e-06,
+      "loss": 0.1927,
+      "step": 12268
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.5885544124988282,
+      "learning_rate": 2.4190506973814647e-06,
+      "loss": 0.2123,
+      "step": 12269
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.2092756927979254,
+      "learning_rate": 2.4182791922761324e-06,
+      "loss": 0.1915,
+      "step": 12270
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.1758860655415306,
+      "learning_rate": 2.4175077709749436e-06,
+      "loss": 0.1638,
+      "step": 12271
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.1548052505777218,
+      "learning_rate": 2.416736433502936e-06,
+      "loss": 0.1762,
+      "step": 12272
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.2927149396756246,
+      "learning_rate": 2.4159651798851484e-06,
+      "loss": 0.1935,
+      "step": 12273
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.3098975931109014,
+      "learning_rate": 2.41519401014662e-06,
+      "loss": 0.2115,
+      "step": 12274
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.2288397314471933,
+      "learning_rate": 2.414422924312378e-06,
+      "loss": 0.1905,
+      "step": 12275
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.9559074459919579,
+      "learning_rate": 2.413651922407457e-06,
+      "loss": 0.1356,
+      "step": 12276
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.8936488741448032,
+      "learning_rate": 2.412881004456879e-06,
+      "loss": 0.2226,
+      "step": 12277
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.2301906210808353,
+      "learning_rate": 2.412110170485672e-06,
+      "loss": 0.2207,
+      "step": 12278
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.0077991632569736,
+      "learning_rate": 2.4113394205188568e-06,
+      "loss": 0.1581,
+      "step": 12279
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.1625776100913774,
+      "learning_rate": 2.4105687545814544e-06,
+      "loss": 0.1861,
+      "step": 12280
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.3163955274875332,
+      "learning_rate": 2.4097981726984777e-06,
+      "loss": 0.1377,
+      "step": 12281
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.2260820957479235,
+      "learning_rate": 2.4090276748949415e-06,
+      "loss": 0.233,
+      "step": 12282
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.3965109122362485,
+      "learning_rate": 2.408257261195857e-06,
+      "loss": 0.1997,
+      "step": 12283
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.097230034342987,
+      "learning_rate": 2.407486931626232e-06,
+      "loss": 0.1606,
+      "step": 12284
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.825855964222305,
+      "learning_rate": 2.406716686211074e-06,
+      "loss": 0.1866,
+      "step": 12285
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.2549767603755226,
+      "learning_rate": 2.405946524975382e-06,
+      "loss": 0.1733,
+      "step": 12286
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.8741439216053972,
+      "learning_rate": 2.4051764479441576e-06,
+      "loss": 0.2214,
+      "step": 12287
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.3336664158801723,
+      "learning_rate": 2.4044064551423973e-06,
+      "loss": 0.1855,
+      "step": 12288
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.9583202245410041,
+      "learning_rate": 2.4036365465950978e-06,
+      "loss": 0.192,
+      "step": 12289
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.5066004808342846,
+      "learning_rate": 2.402866722327248e-06,
+      "loss": 0.2344,
+      "step": 12290
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.1616624341345558,
+      "learning_rate": 2.4020969823638368e-06,
+      "loss": 0.1862,
+      "step": 12291
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.075583364533145,
+      "learning_rate": 2.4013273267298544e-06,
+      "loss": 0.1575,
+      "step": 12292
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.88957849023178,
+      "learning_rate": 2.4005577554502786e-06,
+      "loss": 0.2327,
+      "step": 12293
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.3504432029077589,
+      "learning_rate": 2.399788268550095e-06,
+      "loss": 0.172,
+      "step": 12294
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.3724235622951324,
+      "learning_rate": 2.3990188660542773e-06,
+      "loss": 0.1716,
+      "step": 12295
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.0358906462014699,
+      "learning_rate": 2.398249547987803e-06,
+      "loss": 0.176,
+      "step": 12296
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.7232181731151268,
+      "learning_rate": 2.3974803143756446e-06,
+      "loss": 0.2118,
+      "step": 12297
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.1486369433407384,
+      "learning_rate": 2.3967111652427736e-06,
+      "loss": 0.1798,
+      "step": 12298
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.3778479110491595,
+      "learning_rate": 2.3959421006141536e-06,
+      "loss": 0.1863,
+      "step": 12299
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.0668305087575591,
+      "learning_rate": 2.39517312051475e-06,
+      "loss": 0.2081,
+      "step": 12300
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.4156100409182637,
+      "learning_rate": 2.394404224969525e-06,
+      "loss": 0.174,
+      "step": 12301
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.3358662179144405,
+      "learning_rate": 2.3936354140034374e-06,
+      "loss": 0.2047,
+      "step": 12302
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.5521012460925185,
+      "learning_rate": 2.392866687641445e-06,
+      "loss": 0.1884,
+      "step": 12303
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.7251485044541988,
+      "learning_rate": 2.3920980459084966e-06,
+      "loss": 0.192,
+      "step": 12304
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.2297080882868232,
+      "learning_rate": 2.3913294888295452e-06,
+      "loss": 0.224,
+      "step": 12305
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.551677351270579,
+      "learning_rate": 2.3905610164295394e-06,
+      "loss": 0.1519,
+      "step": 12306
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.228133647102024,
+      "learning_rate": 2.389792628733425e-06,
+      "loss": 0.2033,
+      "step": 12307
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.6141094086469419,
+      "learning_rate": 2.3890243257661413e-06,
+      "loss": 0.2117,
+      "step": 12308
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.565421144268214,
+      "learning_rate": 2.388256107552629e-06,
+      "loss": 0.2243,
+      "step": 12309
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.4099818991030164,
+      "learning_rate": 2.3874879741178276e-06,
+      "loss": 0.2052,
+      "step": 12310
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.1203460226802817,
+      "learning_rate": 2.3867199254866663e-06,
+      "loss": 0.1582,
+      "step": 12311
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.179642903830248,
+      "learning_rate": 2.3859519616840814e-06,
+      "loss": 0.1786,
+      "step": 12312
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.042858607301995,
+      "learning_rate": 2.385184082734997e-06,
+      "loss": 0.1341,
+      "step": 12313
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.7824247401588782,
+      "learning_rate": 2.384416288664341e-06,
+      "loss": 0.2143,
+      "step": 12314
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.5981572585119217,
+      "learning_rate": 2.383648579497036e-06,
+      "loss": 0.1624,
+      "step": 12315
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.8701700650469786,
+      "learning_rate": 2.382880955258005e-06,
+      "loss": 0.1339,
+      "step": 12316
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.1015141050381099,
+      "learning_rate": 2.382113415972161e-06,
+      "loss": 0.1594,
+      "step": 12317
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.1512954113616014,
+      "learning_rate": 2.381345961664421e-06,
+      "loss": 0.1927,
+      "step": 12318
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.3900651501320342,
+      "learning_rate": 2.3805785923596968e-06,
+      "loss": 0.1902,
+      "step": 12319
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.266023784965435,
+      "learning_rate": 2.379811308082898e-06,
+      "loss": 0.1821,
+      "step": 12320
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.2208411939740047,
+      "learning_rate": 2.3790441088589324e-06,
+      "loss": 0.1959,
+      "step": 12321
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.2227654271237687,
+      "learning_rate": 2.3782769947127007e-06,
+      "loss": 0.1922,
+      "step": 12322
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.2669641833020646,
+      "learning_rate": 2.377509965669105e-06,
+      "loss": 0.2026,
+      "step": 12323
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.0465338848686496,
+      "learning_rate": 2.3767430217530434e-06,
+      "loss": 0.177,
+      "step": 12324
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.0529824108013701,
+      "learning_rate": 2.3759761629894144e-06,
+      "loss": 0.1955,
+      "step": 12325
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.6749310415221645,
+      "learning_rate": 2.3752093894031057e-06,
+      "loss": 0.1537,
+      "step": 12326
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.5431278098487533,
+      "learning_rate": 2.3744427010190103e-06,
+      "loss": 0.1998,
+      "step": 12327
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.4054934267821981,
+      "learning_rate": 2.373676097862016e-06,
+      "loss": 0.1878,
+      "step": 12328
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.5285036897181137,
+      "learning_rate": 2.372909579957004e-06,
+      "loss": 0.1486,
+      "step": 12329
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.509552761730503,
+      "learning_rate": 2.3721431473288597e-06,
+      "loss": 0.2001,
+      "step": 12330
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.202064808741901,
+      "learning_rate": 2.3713768000024576e-06,
+      "loss": 0.1741,
+      "step": 12331
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.123283056219763,
+      "learning_rate": 2.370610538002677e-06,
+      "loss": 0.2137,
+      "step": 12332
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.7041633682191852,
+      "learning_rate": 2.3698443613543897e-06,
+      "loss": 0.1924,
+      "step": 12333
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.1830115822235552,
+      "learning_rate": 2.3690782700824693e-06,
+      "loss": 0.1759,
+      "step": 12334
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.3533253611700178,
+      "learning_rate": 2.368312264211779e-06,
+      "loss": 0.158,
+      "step": 12335
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.2624626122264118,
+      "learning_rate": 2.367546343767186e-06,
+      "loss": 0.1915,
+      "step": 12336
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.2194861945627562,
+      "learning_rate": 2.366780508773552e-06,
+      "loss": 0.2243,
+      "step": 12337
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.3588036838373754,
+      "learning_rate": 2.3660147592557368e-06,
+      "loss": 0.2165,
+      "step": 12338
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.22336950612552,
+      "learning_rate": 2.3652490952386e-06,
+      "loss": 0.2003,
+      "step": 12339
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.3382654208056075,
+      "learning_rate": 2.3644835167469898e-06,
+      "loss": 0.1779,
+      "step": 12340
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.4740225628539012,
+      "learning_rate": 2.3637180238057606e-06,
+      "loss": 0.1711,
+      "step": 12341
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.3075543788101398,
+      "learning_rate": 2.36295261643976e-06,
+      "loss": 0.2126,
+      "step": 12342
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.5066637520551798,
+      "learning_rate": 2.3621872946738366e-06,
+      "loss": 0.2165,
+      "step": 12343
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.751308905341228,
+      "learning_rate": 2.361422058532828e-06,
+      "loss": 0.1235,
+      "step": 12344
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.108540165448273,
+      "learning_rate": 2.360656908041577e-06,
+      "loss": 0.1606,
+      "step": 12345
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.192686057411782,
+      "learning_rate": 2.3598918432249218e-06,
+      "loss": 0.1969,
+      "step": 12346
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.5644157776678755,
+      "learning_rate": 2.3591268641076943e-06,
+      "loss": 0.2005,
+      "step": 12347
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.8749692749371927,
+      "learning_rate": 2.3583619707147286e-06,
+      "loss": 0.1519,
+      "step": 12348
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.3511205126927595,
+      "learning_rate": 2.357597163070851e-06,
+      "loss": 0.193,
+      "step": 12349
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.4340483648850864,
+      "learning_rate": 2.3568324412008886e-06,
+      "loss": 0.1631,
+      "step": 12350
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.5515111675587394,
+      "learning_rate": 2.3560678051296653e-06,
+      "loss": 0.2124,
+      "step": 12351
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.3981242868031987,
+      "learning_rate": 2.355303254882004e-06,
+      "loss": 0.1786,
+      "step": 12352
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.523497682996565,
+      "learning_rate": 2.3545387904827164e-06,
+      "loss": 0.2448,
+      "step": 12353
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.4417633244970494,
+      "learning_rate": 2.353774411956622e-06,
+      "loss": 0.1817,
+      "step": 12354
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.2375607244446958,
+      "learning_rate": 2.353010119328532e-06,
+      "loss": 0.1962,
+      "step": 12355
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.4033312442064403,
+      "learning_rate": 2.352245912623255e-06,
+      "loss": 0.1924,
+      "step": 12356
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.1573052552584555,
+      "learning_rate": 2.3514817918656007e-06,
+      "loss": 0.1968,
+      "step": 12357
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.4097089348585987,
+      "learning_rate": 2.350717757080368e-06,
+      "loss": 0.1907,
+      "step": 12358
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.3044804220272999,
+      "learning_rate": 2.3499538082923607e-06,
+      "loss": 0.1868,
+      "step": 12359
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.3662930205570543,
+      "learning_rate": 2.3491899455263765e-06,
+      "loss": 0.197,
+      "step": 12360
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.5198484888843138,
+      "learning_rate": 2.3484261688072134e-06,
+      "loss": 0.2365,
+      "step": 12361
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.0541364996567932,
+      "learning_rate": 2.347662478159659e-06,
+      "loss": 0.1831,
+      "step": 12362
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.2679329656133551,
+      "learning_rate": 2.3468988736085063e-06,
+      "loss": 0.1562,
+      "step": 12363
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.5479874319618592,
+      "learning_rate": 2.346135355178544e-06,
+      "loss": 0.2157,
+      "step": 12364
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.3139683763448744,
+      "learning_rate": 2.3453719228945515e-06,
+      "loss": 0.1689,
+      "step": 12365
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.1286723445665068,
+      "learning_rate": 2.3446085767813148e-06,
+      "loss": 0.1388,
+      "step": 12366
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.468474610869018,
+      "learning_rate": 2.343845316863609e-06,
+      "loss": 0.1977,
+      "step": 12367
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.086912492749228,
+      "learning_rate": 2.3430821431662113e-06,
+      "loss": 0.1309,
+      "step": 12368
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.3917904320759642,
+      "learning_rate": 2.3423190557138957e-06,
+      "loss": 0.2168,
+      "step": 12369
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.4868862066812694,
+      "learning_rate": 2.3415560545314338e-06,
+      "loss": 0.2373,
+      "step": 12370
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.3371999419619274,
+      "learning_rate": 2.3407931396435886e-06,
+      "loss": 0.1845,
+      "step": 12371
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.3292218755772964,
+      "learning_rate": 2.3400303110751276e-06,
+      "loss": 0.1455,
+      "step": 12372
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.2598001785862374,
+      "learning_rate": 2.3392675688508123e-06,
+      "loss": 0.1614,
+      "step": 12373
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.1075723972994485,
+      "learning_rate": 2.338504912995401e-06,
+      "loss": 0.2267,
+      "step": 12374
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.3070507232802353,
+      "learning_rate": 2.3377423435336533e-06,
+      "loss": 0.1918,
+      "step": 12375
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.3073980305881965,
+      "learning_rate": 2.336979860490318e-06,
+      "loss": 0.2127,
+      "step": 12376
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.4041783411458513,
+      "learning_rate": 2.336217463890148e-06,
+      "loss": 0.2293,
+      "step": 12377
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.0419417729288127,
+      "learning_rate": 2.3354551537578905e-06,
+      "loss": 0.1604,
+      "step": 12378
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.2731372034520796,
+      "learning_rate": 2.334692930118293e-06,
+      "loss": 0.209,
+      "step": 12379
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.147498645568705,
+      "learning_rate": 2.3339307929960943e-06,
+      "loss": 0.1512,
+      "step": 12380
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.1018645636702062,
+      "learning_rate": 2.3331687424160345e-06,
+      "loss": 0.2157,
+      "step": 12381
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.2200570281127694,
+      "learning_rate": 2.3324067784028524e-06,
+      "loss": 0.1279,
+      "step": 12382
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.1423182428314456,
+      "learning_rate": 2.3316449009812793e-06,
+      "loss": 0.1735,
+      "step": 12383
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.5194067511757874,
+      "learning_rate": 2.330883110176049e-06,
+      "loss": 0.1968,
+      "step": 12384
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.9712748695439626,
+      "learning_rate": 2.330121406011885e-06,
+      "loss": 0.0966,
+      "step": 12385
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.1396698185631995,
+      "learning_rate": 2.329359788513516e-06,
+      "loss": 0.1785,
+      "step": 12386
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.6241426315196135,
+      "learning_rate": 2.3285982577056635e-06,
+      "loss": 0.1805,
+      "step": 12387
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.3296650764735225,
+      "learning_rate": 2.3278368136130503e-06,
+      "loss": 0.219,
+      "step": 12388
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.4086366163470958,
+      "learning_rate": 2.3270754562603886e-06,
+      "loss": 0.1957,
+      "step": 12389
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.986285289929575,
+      "learning_rate": 2.326314185672395e-06,
+      "loss": 0.1522,
+      "step": 12390
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.1148218136709693,
+      "learning_rate": 2.32555300187378e-06,
+      "loss": 0.1893,
+      "step": 12391
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.3259713157497188,
+      "learning_rate": 2.324791904889252e-06,
+      "loss": 0.2021,
+      "step": 12392
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.1855198167355725,
+      "learning_rate": 2.3240308947435198e-06,
+      "loss": 0.157,
+      "step": 12393
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.1276281969477375,
+      "learning_rate": 2.323269971461281e-06,
+      "loss": 0.1776,
+      "step": 12394
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.7454288110508223,
+      "learning_rate": 2.322509135067239e-06,
+      "loss": 0.2337,
+      "step": 12395
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.1353515058012469,
+      "learning_rate": 2.3217483855860894e-06,
+      "loss": 0.1194,
+      "step": 12396
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.6406839087673184,
+      "learning_rate": 2.3209877230425295e-06,
+      "loss": 0.1664,
+      "step": 12397
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.3488944598826802,
+      "learning_rate": 2.320227147461247e-06,
+      "loss": 0.1561,
+      "step": 12398
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.1098144695055907,
+      "learning_rate": 2.319466658866932e-06,
+      "loss": 0.1754,
+      "step": 12399
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.6382078927934127,
+      "learning_rate": 2.3187062572842724e-06,
+      "loss": 0.1756,
+      "step": 12400
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.6222215245528806,
+      "learning_rate": 2.3179459427379477e-06,
+      "loss": 0.2074,
+      "step": 12401
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.719486033457373,
+      "learning_rate": 2.3171857152526424e-06,
+      "loss": 0.26,
+      "step": 12402
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.4733102292447229,
+      "learning_rate": 2.316425574853029e-06,
+      "loss": 0.2144,
+      "step": 12403
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.2939927981318475,
+      "learning_rate": 2.3156655215637846e-06,
+      "loss": 0.2086,
+      "step": 12404
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.238133183962075,
+      "learning_rate": 2.314905555409581e-06,
+      "loss": 0.1523,
+      "step": 12405
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.2713604645479963,
+      "learning_rate": 2.3141456764150894e-06,
+      "loss": 0.1529,
+      "step": 12406
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.4773064661591497,
+      "learning_rate": 2.313385884604972e-06,
+      "loss": 0.1528,
+      "step": 12407
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.3883048053947764,
+      "learning_rate": 2.3126261800038927e-06,
+      "loss": 0.212,
+      "step": 12408
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.2751001170297447,
+      "learning_rate": 2.3118665626365134e-06,
+      "loss": 0.1739,
+      "step": 12409
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.4796805636395431,
+      "learning_rate": 2.3111070325274913e-06,
+      "loss": 0.2572,
+      "step": 12410
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.1471138027318488,
+      "learning_rate": 2.3103475897014833e-06,
+      "loss": 0.1726,
+      "step": 12411
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.389655578980958,
+      "learning_rate": 2.309588234183137e-06,
+      "loss": 0.2182,
+      "step": 12412
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.2122247499022094,
+      "learning_rate": 2.308828965997104e-06,
+      "loss": 0.1646,
+      "step": 12413
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.505231971511114,
+      "learning_rate": 2.30806978516803e-06,
+      "loss": 0.219,
+      "step": 12414
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.4743009257295052,
+      "learning_rate": 2.3073106917205606e-06,
+      "loss": 0.224,
+      "step": 12415
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.0360416548427416,
+      "learning_rate": 2.306551685679333e-06,
+      "loss": 0.1483,
+      "step": 12416
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.392353194851884,
+      "learning_rate": 2.3057927670689863e-06,
+      "loss": 0.1908,
+      "step": 12417
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.1146197487625173,
+      "learning_rate": 2.3050339359141577e-06,
+      "loss": 0.1936,
+      "step": 12418
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.2632510915917974,
+      "learning_rate": 2.304275192239475e-06,
+      "loss": 0.1791,
+      "step": 12419
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.1177028987689592,
+      "learning_rate": 2.303516536069572e-06,
+      "loss": 0.1593,
+      "step": 12420
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.1069329846098193,
+      "learning_rate": 2.3027579674290708e-06,
+      "loss": 0.1295,
+      "step": 12421
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.3828229149463924,
+      "learning_rate": 2.301999486342597e-06,
+      "loss": 0.1658,
+      "step": 12422
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.308100000863432,
+      "learning_rate": 2.301241092834771e-06,
+      "loss": 0.2261,
+      "step": 12423
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.2197658052995894,
+      "learning_rate": 2.300482786930213e-06,
+      "loss": 0.1868,
+      "step": 12424
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.2235941534488635,
+      "learning_rate": 2.2997245686535346e-06,
+      "loss": 0.2204,
+      "step": 12425
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.428396517612808,
+      "learning_rate": 2.2989664380293496e-06,
+      "loss": 0.2382,
+      "step": 12426
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.0499994425545076,
+      "learning_rate": 2.298208395082267e-06,
+      "loss": 0.1634,
+      "step": 12427
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.0757053769573106,
+      "learning_rate": 2.297450439836893e-06,
+      "loss": 0.1538,
+      "step": 12428
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.2292673573497206,
+      "learning_rate": 2.2966925723178344e-06,
+      "loss": 0.2027,
+      "step": 12429
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.3549280572325788,
+      "learning_rate": 2.295934792549687e-06,
+      "loss": 0.1375,
+      "step": 12430
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.1908502500830505,
+      "learning_rate": 2.2951771005570516e-06,
+      "loss": 0.1875,
+      "step": 12431
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.2509933994658051,
+      "learning_rate": 2.2944194963645227e-06,
+      "loss": 0.1844,
+      "step": 12432
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.3349634099350018,
+      "learning_rate": 2.2936619799966948e-06,
+      "loss": 0.167,
+      "step": 12433
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.235403043254768,
+      "learning_rate": 2.2929045514781533e-06,
+      "loss": 0.1825,
+      "step": 12434
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.138312706810485,
+      "learning_rate": 2.2921472108334868e-06,
+      "loss": 0.1769,
+      "step": 12435
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.4718945503961245,
+      "learning_rate": 2.291389958087281e-06,
+      "loss": 0.1639,
+      "step": 12436
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.4645841546506284,
+      "learning_rate": 2.2906327932641125e-06,
+      "loss": 0.1709,
+      "step": 12437
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.4448000695420378,
+      "learning_rate": 2.289875716388563e-06,
+      "loss": 0.1741,
+      "step": 12438
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.5595110438293078,
+      "learning_rate": 2.289118727485205e-06,
+      "loss": 0.1789,
+      "step": 12439
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.5584451609366419,
+      "learning_rate": 2.2883618265786122e-06,
+      "loss": 0.1901,
+      "step": 12440
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.24985685870038,
+      "learning_rate": 2.287605013693353e-06,
+      "loss": 0.1995,
+      "step": 12441
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.2986079552839334,
+      "learning_rate": 2.2868482888539974e-06,
+      "loss": 0.2216,
+      "step": 12442
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.3886935370535631,
+      "learning_rate": 2.2860916520851046e-06,
+      "loss": 0.2073,
+      "step": 12443
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.8923714939066724,
+      "learning_rate": 2.285335103411237e-06,
+      "loss": 0.18,
+      "step": 12444
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.3783320833611423,
+      "learning_rate": 2.284578642856953e-06,
+      "loss": 0.2133,
+      "step": 12445
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.2644623137085098,
+      "learning_rate": 2.2838222704468075e-06,
+      "loss": 0.1864,
+      "step": 12446
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.8833629072731173,
+      "learning_rate": 2.2830659862053554e-06,
+      "loss": 0.2224,
+      "step": 12447
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.9372362997801847,
+      "learning_rate": 2.282309790157142e-06,
+      "loss": 0.1948,
+      "step": 12448
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.1125195892623816,
+      "learning_rate": 2.2815536823267152e-06,
+      "loss": 0.1485,
+      "step": 12449
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.773274040251208,
+      "learning_rate": 2.280797662738619e-06,
+      "loss": 0.2037,
+      "step": 12450
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.6224293917088126,
+      "learning_rate": 2.2800417314173967e-06,
+      "loss": 0.21,
+      "step": 12451
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.2508172909835442,
+      "learning_rate": 2.279285888387581e-06,
+      "loss": 0.1895,
+      "step": 12452
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.1011684267464377,
+      "learning_rate": 2.2785301336737108e-06,
+      "loss": 0.0882,
+      "step": 12453
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.3913884933383898,
+      "learning_rate": 2.277774467300319e-06,
+      "loss": 0.1749,
+      "step": 12454
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.0706362530500508,
+      "learning_rate": 2.2770188892919322e-06,
+      "loss": 0.1681,
+      "step": 12455
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.6682540506616432,
+      "learning_rate": 2.2762633996730787e-06,
+      "loss": 0.1672,
+      "step": 12456
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.3419729725252254,
+      "learning_rate": 2.2755079984682805e-06,
+      "loss": 0.1644,
+      "step": 12457
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.3029710191550705,
+      "learning_rate": 2.274752685702059e-06,
+      "loss": 0.1834,
+      "step": 12458
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.1323974375145693,
+      "learning_rate": 2.2739974613989328e-06,
+      "loss": 0.1801,
+      "step": 12459
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.3746536533935017,
+      "learning_rate": 2.273242325583416e-06,
+      "loss": 0.1646,
+      "step": 12460
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.2594717644085434,
+      "learning_rate": 2.272487278280024e-06,
+      "loss": 0.2074,
+      "step": 12461
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.7685559850442594,
+      "learning_rate": 2.2717323195132607e-06,
+      "loss": 0.2045,
+      "step": 12462
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.4164575295021247,
+      "learning_rate": 2.2709774493076358e-06,
+      "loss": 0.1675,
+      "step": 12463
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.5421547807007283,
+      "learning_rate": 2.2702226676876515e-06,
+      "loss": 0.1857,
+      "step": 12464
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.5088345665844316,
+      "learning_rate": 2.2694679746778116e-06,
+      "loss": 0.1616,
+      "step": 12465
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.3309117980073242,
+      "learning_rate": 2.2687133703026093e-06,
+      "loss": 0.1719,
+      "step": 12466
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.3602451904508412,
+      "learning_rate": 2.2679588545865417e-06,
+      "loss": 0.1951,
+      "step": 12467
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.110269702856377,
+      "learning_rate": 2.2672044275541007e-06,
+      "loss": 0.1781,
+      "step": 12468
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.1525118899119964,
+      "learning_rate": 2.2664500892297754e-06,
+      "loss": 0.1718,
+      "step": 12469
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 3.1638793162312338,
+      "learning_rate": 2.2656958396380543e-06,
+      "loss": 0.1959,
+      "step": 12470
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.9949591037674368,
+      "learning_rate": 2.264941678803416e-06,
+      "loss": 0.1949,
+      "step": 12471
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.4500156278425764,
+      "learning_rate": 2.264187606750346e-06,
+      "loss": 0.2445,
+      "step": 12472
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.815163123396983,
+      "learning_rate": 2.263433623503317e-06,
+      "loss": 0.1957,
+      "step": 12473
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.2175803471172617,
+      "learning_rate": 2.262679729086809e-06,
+      "loss": 0.1967,
+      "step": 12474
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.1420414617955845,
+      "learning_rate": 2.2619259235252888e-06,
+      "loss": 0.1914,
+      "step": 12475
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.1701829520260767,
+      "learning_rate": 2.2611722068432275e-06,
+      "loss": 0.153,
+      "step": 12476
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.8546677860692187,
+      "learning_rate": 2.2604185790650912e-06,
+      "loss": 0.1633,
+      "step": 12477
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.4370268841331124,
+      "learning_rate": 2.2596650402153437e-06,
+      "loss": 0.1521,
+      "step": 12478
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.5130010484287137,
+      "learning_rate": 2.2589115903184463e-06,
+      "loss": 0.2003,
+      "step": 12479
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.3659937197134546,
+      "learning_rate": 2.258158229398853e-06,
+      "loss": 0.1686,
+      "step": 12480
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.1772618619036075,
+      "learning_rate": 2.25740495748102e-06,
+      "loss": 0.1334,
+      "step": 12481
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.2679860879508302,
+      "learning_rate": 2.2566517745894e-06,
+      "loss": 0.1864,
+      "step": 12482
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.0368762391837767,
+      "learning_rate": 2.2558986807484423e-06,
+      "loss": 0.2159,
+      "step": 12483
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.2851138209379998,
+      "learning_rate": 2.2551456759825896e-06,
+      "loss": 0.147,
+      "step": 12484
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.0286235742534753,
+      "learning_rate": 2.2543927603162867e-06,
+      "loss": 0.1431,
+      "step": 12485
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.0804685504386315,
+      "learning_rate": 2.253639933773973e-06,
+      "loss": 0.1825,
+      "step": 12486
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.392365823321826,
+      "learning_rate": 2.2528871963800876e-06,
+      "loss": 0.17,
+      "step": 12487
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.162544744227129,
+      "learning_rate": 2.2521345481590647e-06,
+      "loss": 0.2407,
+      "step": 12488
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.1130947893831247,
+      "learning_rate": 2.251381989135332e-06,
+      "loss": 0.1734,
+      "step": 12489
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.6310978149807769,
+      "learning_rate": 2.2506295193333233e-06,
+      "loss": 0.1668,
+      "step": 12490
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.9452056765296536,
+      "learning_rate": 2.2498771387774595e-06,
+      "loss": 0.1243,
+      "step": 12491
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.8772514320167797,
+      "learning_rate": 2.249124847492166e-06,
+      "loss": 0.1407,
+      "step": 12492
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.8255735100416723,
+      "learning_rate": 2.2483726455018612e-06,
+      "loss": 0.2081,
+      "step": 12493
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 14.13088621863272,
+      "learning_rate": 2.247620532830962e-06,
+      "loss": 0.1695,
+      "step": 12494
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.4591995601452223,
+      "learning_rate": 2.2468685095038834e-06,
+      "loss": 0.2435,
+      "step": 12495
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.5392371916021954,
+      "learning_rate": 2.246116575545036e-06,
+      "loss": 0.1891,
+      "step": 12496
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.0254011862074697,
+      "learning_rate": 2.2453647309788295e-06,
+      "loss": 0.1543,
+      "step": 12497
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.4489485470774741,
+      "learning_rate": 2.2446129758296665e-06,
+      "loss": 0.1575,
+      "step": 12498
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.7687154367190385,
+      "learning_rate": 2.243861310121951e-06,
+      "loss": 0.1814,
+      "step": 12499
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.1449883609262412,
+      "learning_rate": 2.243109733880082e-06,
+      "loss": 0.2024,
+      "step": 12500
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.1697188791942126,
+      "learning_rate": 2.2423582471284582e-06,
+      "loss": 0.1302,
+      "step": 12501
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.5640979368404515,
+      "learning_rate": 2.2416068498914696e-06,
+      "loss": 0.1968,
+      "step": 12502
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.3276809875807305,
+      "learning_rate": 2.2408555421935084e-06,
+      "loss": 0.1846,
+      "step": 12503
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.3269356001852515,
+      "learning_rate": 2.2401043240589636e-06,
+      "loss": 0.2024,
+      "step": 12504
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.9909990732557233,
+      "learning_rate": 2.239353195512221e-06,
+      "loss": 0.1671,
+      "step": 12505
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.3078370368870729,
+      "learning_rate": 2.2386021565776594e-06,
+      "loss": 0.1781,
+      "step": 12506
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.3186846974552495,
+      "learning_rate": 2.2378512072796597e-06,
+      "loss": 0.2307,
+      "step": 12507
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.5669658427120108,
+      "learning_rate": 2.2371003476426e-06,
+      "loss": 0.1655,
+      "step": 12508
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.113204799545523,
+      "learning_rate": 2.23634957769085e-06,
+      "loss": 0.2555,
+      "step": 12509
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.5290406305966036,
+      "learning_rate": 2.2355988974487843e-06,
+      "loss": 0.2474,
+      "step": 12510
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.1681748147816282,
+      "learning_rate": 2.2348483069407655e-06,
+      "loss": 0.2,
+      "step": 12511
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.9949061513436347,
+      "learning_rate": 2.2340978061911607e-06,
+      "loss": 0.1709,
+      "step": 12512
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.4260470416922337,
+      "learning_rate": 2.2333473952243315e-06,
+      "loss": 0.2023,
+      "step": 12513
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.7537105510595261,
+      "learning_rate": 2.2325970740646373e-06,
+      "loss": 0.26,
+      "step": 12514
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.4055588454270633,
+      "learning_rate": 2.2318468427364353e-06,
+      "loss": 0.1917,
+      "step": 12515
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.6772594956663551,
+      "learning_rate": 2.2310967012640743e-06,
+      "loss": 0.2195,
+      "step": 12516
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.2986724874993796,
+      "learning_rate": 2.2303466496719065e-06,
+      "loss": 0.149,
+      "step": 12517
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.1659888227695465,
+      "learning_rate": 2.22959668798428e-06,
+      "loss": 0.1622,
+      "step": 12518
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.389049505509827,
+      "learning_rate": 2.2288468162255396e-06,
+      "loss": 0.1808,
+      "step": 12519
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.1434107912465405,
+      "learning_rate": 2.2280970344200224e-06,
+      "loss": 0.1745,
+      "step": 12520
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.1138054266216388,
+      "learning_rate": 2.2273473425920704e-06,
+      "loss": 0.1799,
+      "step": 12521
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.2976667302700013,
+      "learning_rate": 2.226597740766018e-06,
+      "loss": 0.1606,
+      "step": 12522
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.5469370959562938,
+      "learning_rate": 2.2258482289661986e-06,
+      "loss": 0.235,
+      "step": 12523
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.4315369322566445,
+      "learning_rate": 2.22509880721694e-06,
+      "loss": 0.1641,
+      "step": 12524
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.7923073265439198,
+      "learning_rate": 2.2243494755425717e-06,
+      "loss": 0.2572,
+      "step": 12525
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.3873151544355713,
+      "learning_rate": 2.2236002339674134e-06,
+      "loss": 0.208,
+      "step": 12526
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.0654712987058974,
+      "learning_rate": 2.2228510825157877e-06,
+      "loss": 0.1821,
+      "step": 12527
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.0310971515262395,
+      "learning_rate": 2.222102021212015e-06,
+      "loss": 0.1594,
+      "step": 12528
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.9174225800637306,
+      "learning_rate": 2.2213530500804065e-06,
+      "loss": 0.2567,
+      "step": 12529
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.4029381111420014,
+      "learning_rate": 2.2206041691452756e-06,
+      "loss": 0.1744,
+      "step": 12530
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.3102509438784802,
+      "learning_rate": 2.2198553784309317e-06,
+      "loss": 0.1249,
+      "step": 12531
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.4118814614221142,
+      "learning_rate": 2.219106677961681e-06,
+      "loss": 0.193,
+      "step": 12532
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.0923493817904089,
+      "learning_rate": 2.2183580677618285e-06,
+      "loss": 0.1999,
+      "step": 12533
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.57945816943048,
+      "learning_rate": 2.2176095478556707e-06,
+      "loss": 0.2365,
+      "step": 12534
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.5749435446392854,
+      "learning_rate": 2.216861118267507e-06,
+      "loss": 0.1786,
+      "step": 12535
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.1459086335611053,
+      "learning_rate": 2.216112779021632e-06,
+      "loss": 0.173,
+      "step": 12536
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.9090006868786824,
+      "learning_rate": 2.2153645301423382e-06,
+      "loss": 0.2623,
+      "step": 12537
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.776947268604046,
+      "learning_rate": 2.214616371653912e-06,
+      "loss": 0.2205,
+      "step": 12538
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.3914584919515856,
+      "learning_rate": 2.21386830358064e-06,
+      "loss": 0.2048,
+      "step": 12539
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.0069941474570725,
+      "learning_rate": 2.213120325946804e-06,
+      "loss": 0.2485,
+      "step": 12540
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.084308967448692,
+      "learning_rate": 2.2123724387766877e-06,
+      "loss": 0.192,
+      "step": 12541
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.4236692175081271,
+      "learning_rate": 2.211624642094563e-06,
+      "loss": 0.193,
+      "step": 12542
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.2554772004636547,
+      "learning_rate": 2.2108769359247077e-06,
+      "loss": 0.1693,
+      "step": 12543
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.1643256075708597,
+      "learning_rate": 2.2101293202913885e-06,
+      "loss": 0.1452,
+      "step": 12544
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.1894035203734568,
+      "learning_rate": 2.209381795218877e-06,
+      "loss": 0.1739,
+      "step": 12545
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.210597836633311,
+      "learning_rate": 2.208634360731438e-06,
+      "loss": 0.186,
+      "step": 12546
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.3488497217804132,
+      "learning_rate": 2.207887016853331e-06,
+      "loss": 0.1763,
+      "step": 12547
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.384345296170796,
+      "learning_rate": 2.2071397636088176e-06,
+      "loss": 0.2213,
+      "step": 12548
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.9751625555195521,
+      "learning_rate": 2.206392601022153e-06,
+      "loss": 0.211,
+      "step": 12549
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.3157626709749224,
+      "learning_rate": 2.2056455291175915e-06,
+      "loss": 0.2307,
+      "step": 12550
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.6701591100105282,
+      "learning_rate": 2.2048985479193846e-06,
+      "loss": 0.1884,
+      "step": 12551
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.1616382894249426,
+      "learning_rate": 2.204151657451776e-06,
+      "loss": 0.164,
+      "step": 12552
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.1452220583096673,
+      "learning_rate": 2.2034048577390126e-06,
+      "loss": 0.1853,
+      "step": 12553
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.337127061125668,
+      "learning_rate": 2.2026581488053355e-06,
+      "loss": 0.1937,
+      "step": 12554
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.4584483089990548,
+      "learning_rate": 2.2019115306749855e-06,
+      "loss": 0.1801,
+      "step": 12555
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.507863024127012,
+      "learning_rate": 2.2011650033721943e-06,
+      "loss": 0.1874,
+      "step": 12556
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.745980066412544,
+      "learning_rate": 2.2004185669211966e-06,
+      "loss": 0.1837,
+      "step": 12557
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.9292569265321688,
+      "learning_rate": 2.199672221346222e-06,
+      "loss": 0.162,
+      "step": 12558
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.9191163741492916,
+      "learning_rate": 2.198925966671499e-06,
+      "loss": 0.2812,
+      "step": 12559
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.5725835443206517,
+      "learning_rate": 2.1981798029212485e-06,
+      "loss": 0.2038,
+      "step": 12560
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.1384134048322525,
+      "learning_rate": 2.1974337301196947e-06,
+      "loss": 0.2101,
+      "step": 12561
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.1258539826235734,
+      "learning_rate": 2.1966877482910508e-06,
+      "loss": 0.1438,
+      "step": 12562
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.8174284325090149,
+      "learning_rate": 2.195941857459536e-06,
+      "loss": 0.1892,
+      "step": 12563
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.3866398157569484,
+      "learning_rate": 2.195196057649362e-06,
+      "loss": 0.1908,
+      "step": 12564
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.5163839838750641,
+      "learning_rate": 2.1944503488847346e-06,
+      "loss": 0.2243,
+      "step": 12565
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.0718086012190395,
+      "learning_rate": 2.193704731189863e-06,
+      "loss": 0.1545,
+      "step": 12566
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.478352564156119,
+      "learning_rate": 2.192959204588949e-06,
+      "loss": 0.2325,
+      "step": 12567
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.114454547922621,
+      "learning_rate": 2.1922137691061936e-06,
+      "loss": 0.1574,
+      "step": 12568
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.4266248846044838,
+      "learning_rate": 2.1914684247657956e-06,
+      "loss": 0.197,
+      "step": 12569
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.2225320893154088,
+      "learning_rate": 2.190723171591946e-06,
+      "loss": 0.1484,
+      "step": 12570
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.232563474654766,
+      "learning_rate": 2.1899780096088375e-06,
+      "loss": 0.1944,
+      "step": 12571
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.200808490859479,
+      "learning_rate": 2.1892329388406586e-06,
+      "loss": 0.1725,
+      "step": 12572
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.0300644649081587,
+      "learning_rate": 2.1884879593115967e-06,
+      "loss": 0.1721,
+      "step": 12573
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.3195348599544312,
+      "learning_rate": 2.1877430710458307e-06,
+      "loss": 0.2158,
+      "step": 12574
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.2714378096437569,
+      "learning_rate": 2.1869982740675414e-06,
+      "loss": 0.2676,
+      "step": 12575
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.3646933393569662,
+      "learning_rate": 2.186253568400906e-06,
+      "loss": 0.1927,
+      "step": 12576
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.4421378504264977,
+      "learning_rate": 2.1855089540700998e-06,
+      "loss": 0.1626,
+      "step": 12577
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.5288770618293825,
+      "learning_rate": 2.1847644310992893e-06,
+      "loss": 0.1814,
+      "step": 12578
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.292205879375929,
+      "learning_rate": 2.184019999512646e-06,
+      "loss": 0.1658,
+      "step": 12579
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.201159956337948,
+      "learning_rate": 2.1832756593343313e-06,
+      "loss": 0.1583,
+      "step": 12580
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.4411505004477287,
+      "learning_rate": 2.182531410588508e-06,
+      "loss": 0.1636,
+      "step": 12581
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.3278322121514217,
+      "learning_rate": 2.181787253299338e-06,
+      "loss": 0.1386,
+      "step": 12582
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.2563109587118153,
+      "learning_rate": 2.1810431874909717e-06,
+      "loss": 0.179,
+      "step": 12583
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.2630896044881859,
+      "learning_rate": 2.1802992131875646e-06,
+      "loss": 0.2049,
+      "step": 12584
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.3554850640538192,
+      "learning_rate": 2.1795553304132668e-06,
+      "loss": 0.2029,
+      "step": 12585
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.446044550903211,
+      "learning_rate": 2.1788115391922242e-06,
+      "loss": 0.1937,
+      "step": 12586
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.1287854007859968,
+      "learning_rate": 2.1780678395485837e-06,
+      "loss": 0.1881,
+      "step": 12587
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.4000398177845237,
+      "learning_rate": 2.177324231506482e-06,
+      "loss": 0.2028,
+      "step": 12588
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.0068911621458863,
+      "learning_rate": 2.1765807150900587e-06,
+      "loss": 0.1717,
+      "step": 12589
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.3425427575068727,
+      "learning_rate": 2.1758372903234492e-06,
+      "loss": 0.1661,
+      "step": 12590
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.6624807532350836,
+      "learning_rate": 2.1750939572307874e-06,
+      "loss": 0.18,
+      "step": 12591
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.477352702912023,
+      "learning_rate": 2.1743507158361976e-06,
+      "loss": 0.189,
+      "step": 12592
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.4570291636761412,
+      "learning_rate": 2.173607566163809e-06,
+      "loss": 0.1488,
+      "step": 12593
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.3439153580611416,
+      "learning_rate": 2.1728645082377435e-06,
+      "loss": 0.184,
+      "step": 12594
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.3967640309985234,
+      "learning_rate": 2.172121542082124e-06,
+      "loss": 0.1432,
+      "step": 12595
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.2007607508530913,
+      "learning_rate": 2.1713786677210634e-06,
+      "loss": 0.1714,
+      "step": 12596
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.140050390597997,
+      "learning_rate": 2.17063588517868e-06,
+      "loss": 0.1742,
+      "step": 12597
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.3475974526225425,
+      "learning_rate": 2.169893194479081e-06,
+      "loss": 0.209,
+      "step": 12598
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.7758401610055086,
+      "learning_rate": 2.169150595646376e-06,
+      "loss": 0.1846,
+      "step": 12599
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.14118897199669,
+      "learning_rate": 2.1684080887046725e-06,
+      "loss": 0.2058,
+      "step": 12600
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.9153302334614764,
+      "learning_rate": 2.167665673678069e-06,
+      "loss": 0.2266,
+      "step": 12601
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.5892716814917185,
+      "learning_rate": 2.1669233505906663e-06,
+      "loss": 0.176,
+      "step": 12602
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.6692553277468418,
+      "learning_rate": 2.1661811194665605e-06,
+      "loss": 0.2225,
+      "step": 12603
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.3459681760650302,
+      "learning_rate": 2.1654389803298452e-06,
+      "loss": 0.2209,
+      "step": 12604
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.8014635017488831,
+      "learning_rate": 2.1646969332046127e-06,
+      "loss": 0.1579,
+      "step": 12605
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.58796408554889,
+      "learning_rate": 2.163954978114946e-06,
+      "loss": 0.2519,
+      "step": 12606
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.5313751257162376,
+      "learning_rate": 2.163213115084932e-06,
+      "loss": 0.2223,
+      "step": 12607
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.1513910008007069,
+      "learning_rate": 2.1624713441386515e-06,
+      "loss": 0.1736,
+      "step": 12608
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.1393886031334295,
+      "learning_rate": 2.1617296653001846e-06,
+      "loss": 0.1485,
+      "step": 12609
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.5873417110826173,
+      "learning_rate": 2.1609880785936027e-06,
+      "loss": 0.1267,
+      "step": 12610
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.0822366354294035,
+      "learning_rate": 2.1602465840429802e-06,
+      "loss": 0.1574,
+      "step": 12611
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.3609020984043856,
+      "learning_rate": 2.1595051816723873e-06,
+      "loss": 0.1858,
+      "step": 12612
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.475482757818579,
+      "learning_rate": 2.158763871505891e-06,
+      "loss": 0.1918,
+      "step": 12613
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.2076251349757563,
+      "learning_rate": 2.1580226535675517e-06,
+      "loss": 0.1721,
+      "step": 12614
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.1642968354759167,
+      "learning_rate": 2.157281527881433e-06,
+      "loss": 0.1567,
+      "step": 12615
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.4164203144433563,
+      "learning_rate": 2.1565404944715883e-06,
+      "loss": 0.1789,
+      "step": 12616
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.7900206159364348,
+      "learning_rate": 2.155799553362075e-06,
+      "loss": 0.1869,
+      "step": 12617
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.5897012407696234,
+      "learning_rate": 2.155058704576945e-06,
+      "loss": 0.1934,
+      "step": 12618
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.1912795108860426,
+      "learning_rate": 2.1543179481402443e-06,
+      "loss": 0.1767,
+      "step": 12619
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.177502800741732,
+      "learning_rate": 2.153577284076019e-06,
+      "loss": 0.1651,
+      "step": 12620
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.24469463583385,
+      "learning_rate": 2.1528367124083115e-06,
+      "loss": 0.2101,
+      "step": 12621
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.4220861969762597,
+      "learning_rate": 2.1520962331611615e-06,
+      "loss": 0.1998,
+      "step": 12622
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.1811302470000742,
+      "learning_rate": 2.1513558463586075e-06,
+      "loss": 0.1948,
+      "step": 12623
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.407512373692004,
+      "learning_rate": 2.1506155520246795e-06,
+      "loss": 0.1901,
+      "step": 12624
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.292730640833169,
+      "learning_rate": 2.1498753501834086e-06,
+      "loss": 0.1692,
+      "step": 12625
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.5493670195912201,
+      "learning_rate": 2.149135240858823e-06,
+      "loss": 0.2166,
+      "step": 12626
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.4015495500636115,
+      "learning_rate": 2.1483952240749485e-06,
+      "loss": 0.1913,
+      "step": 12627
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.0047318407243908,
+      "learning_rate": 2.147655299855803e-06,
+      "loss": 0.1733,
+      "step": 12628
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.5925868595754866,
+      "learning_rate": 2.146915468225407e-06,
+      "loss": 0.2101,
+      "step": 12629
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.2101024270940672,
+      "learning_rate": 2.146175729207775e-06,
+      "loss": 0.1864,
+      "step": 12630
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.2289990003303757,
+      "learning_rate": 2.1454360828269223e-06,
+      "loss": 0.1987,
+      "step": 12631
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.9133818957637666,
+      "learning_rate": 2.144696529106854e-06,
+      "loss": 0.155,
+      "step": 12632
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.344791283362423,
+      "learning_rate": 2.14395706807158e-06,
+      "loss": 0.1945,
+      "step": 12633
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.5958511092484113,
+      "learning_rate": 2.1432176997450993e-06,
+      "loss": 0.1756,
+      "step": 12634
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.5687834046519382,
+      "learning_rate": 2.1424784241514155e-06,
+      "loss": 0.177,
+      "step": 12635
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.4451016452704708,
+      "learning_rate": 2.141739241314527e-06,
+      "loss": 0.1856,
+      "step": 12636
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.5852584888620198,
+      "learning_rate": 2.141000151258425e-06,
+      "loss": 0.1874,
+      "step": 12637
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.3159827278353091,
+      "learning_rate": 2.1402611540071015e-06,
+      "loss": 0.1406,
+      "step": 12638
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.2799412451331225,
+      "learning_rate": 2.1395222495845456e-06,
+      "loss": 0.1821,
+      "step": 12639
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.3707437448905018,
+      "learning_rate": 2.1387834380147426e-06,
+      "loss": 0.1714,
+      "step": 12640
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.222315365982187,
+      "learning_rate": 2.138044719321676e-06,
+      "loss": 0.2506,
+      "step": 12641
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.2244601424943522,
+      "learning_rate": 2.1373060935293217e-06,
+      "loss": 0.197,
+      "step": 12642
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.5531815442781614,
+      "learning_rate": 2.136567560661658e-06,
+      "loss": 0.1591,
+      "step": 12643
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.186217456186314,
+      "learning_rate": 2.135829120742658e-06,
+      "loss": 0.1727,
+      "step": 12644
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.1127424475536334,
+      "learning_rate": 2.1350907737962935e-06,
+      "loss": 0.1357,
+      "step": 12645
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.251162015701593,
+      "learning_rate": 2.1343525198465278e-06,
+      "loss": 0.1705,
+      "step": 12646
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.1947027097088572,
+      "learning_rate": 2.1336143589173273e-06,
+      "loss": 0.1857,
+      "step": 12647
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.0578619408710255,
+      "learning_rate": 2.1328762910326527e-06,
+      "loss": 0.1323,
+      "step": 12648
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.245292190520164,
+      "learning_rate": 2.1321383162164647e-06,
+      "loss": 0.2289,
+      "step": 12649
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.0705643935232738,
+      "learning_rate": 2.131400434492714e-06,
+      "loss": 0.1923,
+      "step": 12650
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.1697176435012397,
+      "learning_rate": 2.130662645885357e-06,
+      "loss": 0.1259,
+      "step": 12651
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.9837066358346835,
+      "learning_rate": 2.1299249504183384e-06,
+      "loss": 0.2389,
+      "step": 12652
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.570988618661361,
+      "learning_rate": 2.1291873481156066e-06,
+      "loss": 0.1918,
+      "step": 12653
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.525827840942395,
+      "learning_rate": 2.1284498390011057e-06,
+      "loss": 0.1817,
+      "step": 12654
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.2782543506316992,
+      "learning_rate": 2.1277124230987728e-06,
+      "loss": 0.2086,
+      "step": 12655
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.3760134500468915,
+      "learning_rate": 2.1269751004325463e-06,
+      "loss": 0.1832,
+      "step": 12656
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.5673145124819003,
+      "learning_rate": 2.12623787102636e-06,
+      "loss": 0.1857,
+      "step": 12657
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.3845675613304955,
+      "learning_rate": 2.1255007349041447e-06,
+      "loss": 0.1833,
+      "step": 12658
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.899221933253142,
+      "learning_rate": 2.1247636920898308e-06,
+      "loss": 0.1225,
+      "step": 12659
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.7238138033585098,
+      "learning_rate": 2.1240267426073384e-06,
+      "loss": 0.22,
+      "step": 12660
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.318729102954449,
+      "learning_rate": 2.123289886480592e-06,
+      "loss": 0.2314,
+      "step": 12661
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.265659726219841,
+      "learning_rate": 2.1225531237335107e-06,
+      "loss": 0.167,
+      "step": 12662
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.5393743634735662,
+      "learning_rate": 2.121816454390011e-06,
+      "loss": 0.2006,
+      "step": 12663
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.4022167727435269,
+      "learning_rate": 2.121079878474002e-06,
+      "loss": 0.2298,
+      "step": 12664
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.1564715372960583,
+      "learning_rate": 2.1203433960093963e-06,
+      "loss": 0.1568,
+      "step": 12665
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.678421794496139,
+      "learning_rate": 2.1196070070200998e-06,
+      "loss": 0.199,
+      "step": 12666
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.1156531881997367,
+      "learning_rate": 2.1188707115300176e-06,
+      "loss": 0.1789,
+      "step": 12667
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.2705803709844434,
+      "learning_rate": 2.118134509563047e-06,
+      "loss": 0.193,
+      "step": 12668
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.0837707171500406,
+      "learning_rate": 2.1173984011430904e-06,
+      "loss": 0.1871,
+      "step": 12669
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.7129823405680003,
+      "learning_rate": 2.1166623862940366e-06,
+      "loss": 0.193,
+      "step": 12670
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.2773525576039944,
+      "learning_rate": 2.1159264650397797e-06,
+      "loss": 0.1251,
+      "step": 12671
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.3742670359756992,
+      "learning_rate": 2.1151906374042104e-06,
+      "loss": 0.1804,
+      "step": 12672
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.269374573649753,
+      "learning_rate": 2.1144549034112093e-06,
+      "loss": 0.1466,
+      "step": 12673
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.969688487770052,
+      "learning_rate": 2.113719263084662e-06,
+      "loss": 0.1858,
+      "step": 12674
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.7781669806165696,
+      "learning_rate": 2.1129837164484473e-06,
+      "loss": 0.2283,
+      "step": 12675
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.1448752306231873,
+      "learning_rate": 2.112248263526441e-06,
+      "loss": 0.183,
+      "step": 12676
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.2529889789208488,
+      "learning_rate": 2.1115129043425188e-06,
+      "loss": 0.1601,
+      "step": 12677
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.2929729622948019,
+      "learning_rate": 2.1107776389205463e-06,
+      "loss": 0.1974,
+      "step": 12678
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.9621280490241618,
+      "learning_rate": 2.110042467284393e-06,
+      "loss": 0.1463,
+      "step": 12679
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.2388661263846383,
+      "learning_rate": 2.1093073894579237e-06,
+      "loss": 0.1722,
+      "step": 12680
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.3506232509575233,
+      "learning_rate": 2.108572405465e-06,
+      "loss": 0.1785,
+      "step": 12681
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.5143038949949843,
+      "learning_rate": 2.107837515329478e-06,
+      "loss": 0.21,
+      "step": 12682
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.1982295370638867,
+      "learning_rate": 2.1071027190752126e-06,
+      "loss": 0.1744,
+      "step": 12683
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.5336827348028572,
+      "learning_rate": 2.106368016726056e-06,
+      "loss": 0.1689,
+      "step": 12684
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.2017563320642433,
+      "learning_rate": 2.10563340830586e-06,
+      "loss": 0.1659,
+      "step": 12685
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.1391443937399013,
+      "learning_rate": 2.104898893838466e-06,
+      "loss": 0.1538,
+      "step": 12686
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.759185064406472,
+      "learning_rate": 2.1041644733477205e-06,
+      "loss": 0.2272,
+      "step": 12687
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.4634514053404188,
+      "learning_rate": 2.10343014685746e-06,
+      "loss": 0.1844,
+      "step": 12688
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.101239606865629,
+      "learning_rate": 2.102695914391522e-06,
+      "loss": 0.1727,
+      "step": 12689
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.0418459745893847,
+      "learning_rate": 2.1019617759737426e-06,
+      "loss": 0.1525,
+      "step": 12690
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.1471167027754394,
+      "learning_rate": 2.101227731627949e-06,
+      "loss": 0.1993,
+      "step": 12691
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.0285438519236285,
+      "learning_rate": 2.1004937813779696e-06,
+      "loss": 0.1478,
+      "step": 12692
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.2210187946336775,
+      "learning_rate": 2.0997599252476297e-06,
+      "loss": 0.1792,
+      "step": 12693
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.988871138261727,
+      "learning_rate": 2.09902616326075e-06,
+      "loss": 0.1422,
+      "step": 12694
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.4878630914241189,
+      "learning_rate": 2.0982924954411516e-06,
+      "loss": 0.2121,
+      "step": 12695
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.4811450953295846,
+      "learning_rate": 2.097558921812645e-06,
+      "loss": 0.1967,
+      "step": 12696
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.3640987601163626,
+      "learning_rate": 2.0968254423990454e-06,
+      "loss": 0.1914,
+      "step": 12697
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 3.187903098255048,
+      "learning_rate": 2.096092057224161e-06,
+      "loss": 0.1948,
+      "step": 12698
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.425249575417975,
+      "learning_rate": 2.0953587663117988e-06,
+      "loss": 0.1951,
+      "step": 12699
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.443508754738031,
+      "learning_rate": 2.0946255696857627e-06,
+      "loss": 0.1906,
+      "step": 12700
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.170594596732747,
+      "learning_rate": 2.0938924673698496e-06,
+      "loss": 0.1323,
+      "step": 12701
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.2428242353039225,
+      "learning_rate": 2.093159459387858e-06,
+      "loss": 0.1892,
+      "step": 12702
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.033312999024718,
+      "learning_rate": 2.092426545763584e-06,
+      "loss": 0.1668,
+      "step": 12703
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.4473515921323323,
+      "learning_rate": 2.0916937265208144e-06,
+      "loss": 0.195,
+      "step": 12704
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.3733043402131357,
+      "learning_rate": 2.0909610016833403e-06,
+      "loss": 0.1992,
+      "step": 12705
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.1158039188633029,
+      "learning_rate": 2.090228371274944e-06,
+      "loss": 0.2124,
+      "step": 12706
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.162511233562624,
+      "learning_rate": 2.0894958353194067e-06,
+      "loss": 0.1548,
+      "step": 12707
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.318229776299661,
+      "learning_rate": 2.0887633938405094e-06,
+      "loss": 0.1864,
+      "step": 12708
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.234162660937996,
+      "learning_rate": 2.0880310468620275e-06,
+      "loss": 0.1943,
+      "step": 12709
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.2418398280603413,
+      "learning_rate": 2.087298794407731e-06,
+      "loss": 0.1763,
+      "step": 12710
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.1091760100961843,
+      "learning_rate": 2.0865666365013903e-06,
+      "loss": 0.1584,
+      "step": 12711
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.5691400392629542,
+      "learning_rate": 2.085834573166772e-06,
+      "loss": 0.2185,
+      "step": 12712
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.2410289171735966,
+      "learning_rate": 2.0851026044276405e-06,
+      "loss": 0.181,
+      "step": 12713
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.3322573131574234,
+      "learning_rate": 2.0843707303077536e-06,
+      "loss": 0.1957,
+      "step": 12714
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.488661574781264,
+      "learning_rate": 2.083638950830869e-06,
+      "loss": 0.2107,
+      "step": 12715
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.2841592939374777,
+      "learning_rate": 2.082907266020741e-06,
+      "loss": 0.1734,
+      "step": 12716
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.3723570065974833,
+      "learning_rate": 2.0821756759011208e-06,
+      "loss": 0.2053,
+      "step": 12717
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.3234147469592499,
+      "learning_rate": 2.081444180495758e-06,
+      "loss": 0.2015,
+      "step": 12718
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.1436355979174875,
+      "learning_rate": 2.0807127798283933e-06,
+      "loss": 0.1728,
+      "step": 12719
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.2489812782694458,
+      "learning_rate": 2.079981473922771e-06,
+      "loss": 0.1759,
+      "step": 12720
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.3043576754473698,
+      "learning_rate": 2.07925026280263e-06,
+      "loss": 0.1712,
+      "step": 12721
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.6098090762314374,
+      "learning_rate": 2.078519146491704e-06,
+      "loss": 0.2002,
+      "step": 12722
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.2727711636794259,
+      "learning_rate": 2.077788125013728e-06,
+      "loss": 0.1865,
+      "step": 12723
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.5370469862262393,
+      "learning_rate": 2.077057198392428e-06,
+      "loss": 0.2232,
+      "step": 12724
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.331903121468771,
+      "learning_rate": 2.0763263666515327e-06,
+      "loss": 0.182,
+      "step": 12725
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.984707399664778,
+      "learning_rate": 2.0755956298147646e-06,
+      "loss": 0.122,
+      "step": 12726
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.2383842906740088,
+      "learning_rate": 2.0748649879058454e-06,
+      "loss": 0.2401,
+      "step": 12727
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 3.777675144684605,
+      "learning_rate": 2.0741344409484886e-06,
+      "loss": 0.1839,
+      "step": 12728
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.33694916338432,
+      "learning_rate": 2.073403988966411e-06,
+      "loss": 0.1672,
+      "step": 12729
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.1044793908198336,
+      "learning_rate": 2.072673631983323e-06,
+      "loss": 0.1454,
+      "step": 12730
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.9510021998345417,
+      "learning_rate": 2.0719433700229334e-06,
+      "loss": 0.1103,
+      "step": 12731
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.3140861486422963,
+      "learning_rate": 2.0712132031089444e-06,
+      "loss": 0.1623,
+      "step": 12732
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.3918748660830507,
+      "learning_rate": 2.0704831312650593e-06,
+      "loss": 0.1512,
+      "step": 12733
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.7874782827698639,
+      "learning_rate": 2.069753154514976e-06,
+      "loss": 0.2306,
+      "step": 12734
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.1173230802522975,
+      "learning_rate": 2.06902327288239e-06,
+      "loss": 0.1829,
+      "step": 12735
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.635198679344823,
+      "learning_rate": 2.068293486390996e-06,
+      "loss": 0.1864,
+      "step": 12736
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.2566031957049675,
+      "learning_rate": 2.0675637950644795e-06,
+      "loss": 0.1868,
+      "step": 12737
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.215392329251651,
+      "learning_rate": 2.0668341989265283e-06,
+      "loss": 0.1854,
+      "step": 12738
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.9765729522145724,
+      "learning_rate": 2.0661046980008283e-06,
+      "loss": 0.1708,
+      "step": 12739
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.0489992431907451,
+      "learning_rate": 2.065375292311054e-06,
+      "loss": 0.1046,
+      "step": 12740
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.4419547874811884,
+      "learning_rate": 2.064645981880888e-06,
+      "loss": 0.1791,
+      "step": 12741
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.2344109795553815,
+      "learning_rate": 2.063916766733999e-06,
+      "loss": 0.1556,
+      "step": 12742
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.7624813256390426,
+      "learning_rate": 2.0631876468940605e-06,
+      "loss": 0.2159,
+      "step": 12743
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.5000358120299022,
+      "learning_rate": 2.062458622384739e-06,
+      "loss": 0.1506,
+      "step": 12744
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.2903843604560095,
+      "learning_rate": 2.0617296932297028e-06,
+      "loss": 0.1721,
+      "step": 12745
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.474670641302161,
+      "learning_rate": 2.0610008594526075e-06,
+      "loss": 0.2094,
+      "step": 12746
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.671673343653987,
+      "learning_rate": 2.0602721210771147e-06,
+      "loss": 0.2229,
+      "step": 12747
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.3524948718842826,
+      "learning_rate": 2.05954347812688e-06,
+      "loss": 0.1803,
+      "step": 12748
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.4274837780162672,
+      "learning_rate": 2.0588149306255557e-06,
+      "loss": 0.179,
+      "step": 12749
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.3303315513596197,
+      "learning_rate": 2.0580864785967886e-06,
+      "loss": 0.1678,
+      "step": 12750
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.1126946694183233,
+      "learning_rate": 2.057358122064226e-06,
+      "loss": 0.1808,
+      "step": 12751
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.3723316378180008,
+      "learning_rate": 2.056629861051511e-06,
+      "loss": 0.2089,
+      "step": 12752
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.2872345795464377,
+      "learning_rate": 2.0559016955822835e-06,
+      "loss": 0.1778,
+      "step": 12753
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.1695512493413738,
+      "learning_rate": 2.055173625680182e-06,
+      "loss": 0.178,
+      "step": 12754
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.35331816285665,
+      "learning_rate": 2.0544456513688353e-06,
+      "loss": 0.1709,
+      "step": 12755
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.2318996329508327,
+      "learning_rate": 2.053717772671877e-06,
+      "loss": 0.1826,
+      "step": 12756
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.1998270865126643,
+      "learning_rate": 2.0529899896129364e-06,
+      "loss": 0.1929,
+      "step": 12757
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.0563940088301715,
+      "learning_rate": 2.052262302215633e-06,
+      "loss": 0.1832,
+      "step": 12758
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.7712753197779059,
+      "learning_rate": 2.051534710503592e-06,
+      "loss": 0.1986,
+      "step": 12759
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.0411360125651385,
+      "learning_rate": 2.0508072145004284e-06,
+      "loss": 0.1906,
+      "step": 12760
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.2411436757517043,
+      "learning_rate": 2.050079814229758e-06,
+      "loss": 0.2323,
+      "step": 12761
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.3882274479871013,
+      "learning_rate": 2.0493525097151933e-06,
+      "loss": 0.2555,
+      "step": 12762
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.4609222411312515,
+      "learning_rate": 2.0486253009803454e-06,
+      "loss": 0.1662,
+      "step": 12763
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.3523639606924729,
+      "learning_rate": 2.0478981880488146e-06,
+      "loss": 0.1867,
+      "step": 12764
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.110616020038604,
+      "learning_rate": 2.0471711709442066e-06,
+      "loss": 0.1544,
+      "step": 12765
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.446536293509357,
+      "learning_rate": 2.0464442496901204e-06,
+      "loss": 0.246,
+      "step": 12766
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.3225121120455305,
+      "learning_rate": 2.045717424310154e-06,
+      "loss": 0.1987,
+      "step": 12767
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.9461400925542172,
+      "learning_rate": 2.0449906948278974e-06,
+      "loss": 0.1404,
+      "step": 12768
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.5371238969555536,
+      "learning_rate": 2.0442640612669416e-06,
+      "loss": 0.2017,
+      "step": 12769
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.1956942671678585,
+      "learning_rate": 2.043537523650875e-06,
+      "loss": 0.2004,
+      "step": 12770
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.3707182878213835,
+      "learning_rate": 2.042811082003279e-06,
+      "loss": 0.1643,
+      "step": 12771
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.4981269118750782,
+      "learning_rate": 2.042084736347739e-06,
+      "loss": 0.1936,
+      "step": 12772
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.2067214173116203,
+      "learning_rate": 2.041358486707827e-06,
+      "loss": 0.2109,
+      "step": 12773
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.0642284506172388,
+      "learning_rate": 2.0406323331071216e-06,
+      "loss": 0.1419,
+      "step": 12774
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.1192809449600967,
+      "learning_rate": 2.03990627556919e-06,
+      "loss": 0.1457,
+      "step": 12775
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.443307034236467,
+      "learning_rate": 2.0391803141176036e-06,
+      "loss": 0.1705,
+      "step": 12776
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.9690970944980428,
+      "learning_rate": 2.0384544487759283e-06,
+      "loss": 0.2014,
+      "step": 12777
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.9286224422191869,
+      "learning_rate": 2.037728679567723e-06,
+      "loss": 0.1729,
+      "step": 12778
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.486922407208278,
+      "learning_rate": 2.0370030065165485e-06,
+      "loss": 0.14,
+      "step": 12779
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.0563395065835393,
+      "learning_rate": 2.0362774296459597e-06,
+      "loss": 0.1505,
+      "step": 12780
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.247049533130186,
+      "learning_rate": 2.035551948979512e-06,
+      "loss": 0.2094,
+      "step": 12781
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.7955300443911159,
+      "learning_rate": 2.0348265645407507e-06,
+      "loss": 0.0976,
+      "step": 12782
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.6245820626753422,
+      "learning_rate": 2.0341012763532243e-06,
+      "loss": 0.1936,
+      "step": 12783
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.413044241479202,
+      "learning_rate": 2.033376084440476e-06,
+      "loss": 0.2033,
+      "step": 12784
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.5737116259501114,
+      "learning_rate": 2.0326509888260474e-06,
+      "loss": 0.2331,
+      "step": 12785
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.1779066216068947,
+      "learning_rate": 2.0319259895334725e-06,
+      "loss": 0.1963,
+      "step": 12786
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.4892260088416989,
+      "learning_rate": 2.031201086586287e-06,
+      "loss": 0.2428,
+      "step": 12787
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.468739631291578,
+      "learning_rate": 2.030476280008021e-06,
+      "loss": 0.1242,
+      "step": 12788
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.2742378259162956,
+      "learning_rate": 2.0297515698222026e-06,
+      "loss": 0.1617,
+      "step": 12789
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.2661005998558206,
+      "learning_rate": 2.0290269560523584e-06,
+      "loss": 0.1977,
+      "step": 12790
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.8220476237560458,
+      "learning_rate": 2.0283024387220057e-06,
+      "loss": 0.1531,
+      "step": 12791
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.0719278095265081,
+      "learning_rate": 2.027578017854666e-06,
+      "loss": 0.1668,
+      "step": 12792
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.5159308773769773,
+      "learning_rate": 2.026853693473852e-06,
+      "loss": 0.179,
+      "step": 12793
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.7328435305129368,
+      "learning_rate": 2.026129465603078e-06,
+      "loss": 0.1745,
+      "step": 12794
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.4923111929231336,
+      "learning_rate": 2.0254053342658504e-06,
+      "loss": 0.2209,
+      "step": 12795
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.609630038309768,
+      "learning_rate": 2.024681299485676e-06,
+      "loss": 0.2077,
+      "step": 12796
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.8877624168250333,
+      "learning_rate": 2.0239573612860577e-06,
+      "loss": 0.229,
+      "step": 12797
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.113088867229161,
+      "learning_rate": 2.023233519690494e-06,
+      "loss": 0.1672,
+      "step": 12798
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.2924657978627447,
+      "learning_rate": 2.0225097747224846e-06,
+      "loss": 0.1911,
+      "step": 12799
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.5076994977123488,
+      "learning_rate": 2.0217861264055176e-06,
+      "loss": 0.166,
+      "step": 12800
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.480412172630454,
+      "learning_rate": 2.0210625747630865e-06,
+      "loss": 0.2033,
+      "step": 12801
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.260691472576657,
+      "learning_rate": 2.0203391198186765e-06,
+      "loss": 0.1328,
+      "step": 12802
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.5012337160934568,
+      "learning_rate": 2.019615761595774e-06,
+      "loss": 0.1959,
+      "step": 12803
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.401249700532621,
+      "learning_rate": 2.018892500117856e-06,
+      "loss": 0.1799,
+      "step": 12804
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.1696424039004707,
+      "learning_rate": 2.018169335408402e-06,
+      "loss": 0.1723,
+      "step": 12805
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.3450637206330873,
+      "learning_rate": 2.017446267490886e-06,
+      "loss": 0.2017,
+      "step": 12806
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.5030137423383894,
+      "learning_rate": 2.016723296388779e-06,
+      "loss": 0.1893,
+      "step": 12807
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.5658886118397948,
+      "learning_rate": 2.016000422125551e-06,
+      "loss": 0.163,
+      "step": 12808
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.3258820443173494,
+      "learning_rate": 2.0152776447246636e-06,
+      "loss": 0.1688,
+      "step": 12809
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.4878667419391585,
+      "learning_rate": 2.014554964209582e-06,
+      "loss": 0.1784,
+      "step": 12810
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.4742109808947053,
+      "learning_rate": 2.013832380603761e-06,
+      "loss": 0.2042,
+      "step": 12811
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.3378026334687894,
+      "learning_rate": 2.0131098939306606e-06,
+      "loss": 0.174,
+      "step": 12812
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.0700498940965408,
+      "learning_rate": 2.012387504213728e-06,
+      "loss": 0.179,
+      "step": 12813
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.276676003442309,
+      "learning_rate": 2.0116652114764152e-06,
+      "loss": 0.2135,
+      "step": 12814
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.8312208095216003,
+      "learning_rate": 2.0109430157421685e-06,
+      "loss": 0.2011,
+      "step": 12815
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.0897876937013253,
+      "learning_rate": 2.0102209170344296e-06,
+      "loss": 0.1475,
+      "step": 12816
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.1810776685721838,
+      "learning_rate": 2.0094989153766416e-06,
+      "loss": 0.235,
+      "step": 12817
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.7076575092859378,
+      "learning_rate": 2.0087770107922362e-06,
+      "loss": 0.166,
+      "step": 12818
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.1678818156827313,
+      "learning_rate": 2.008055203304649e-06,
+      "loss": 0.1385,
+      "step": 12819
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.474429199124882,
+      "learning_rate": 2.0073334929373108e-06,
+      "loss": 0.2223,
+      "step": 12820
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.3164975844939515,
+      "learning_rate": 2.00661187971365e-06,
+      "loss": 0.1853,
+      "step": 12821
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.3893440379640953,
+      "learning_rate": 2.0058903636570862e-06,
+      "loss": 0.2228,
+      "step": 12822
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.6183323148232478,
+      "learning_rate": 2.005168944791044e-06,
+      "loss": 0.1768,
+      "step": 12823
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.2183792516243408,
+      "learning_rate": 2.00444762313894e-06,
+      "loss": 0.2138,
+      "step": 12824
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.7094414671924065,
+      "learning_rate": 2.0037263987241884e-06,
+      "loss": 0.1669,
+      "step": 12825
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.4325652031805383,
+      "learning_rate": 2.0030052715702026e-06,
+      "loss": 0.156,
+      "step": 12826
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.565175388596564,
+      "learning_rate": 2.002284241700387e-06,
+      "loss": 0.1792,
+      "step": 12827
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.3561292380304437,
+      "learning_rate": 2.001563309138151e-06,
+      "loss": 0.2043,
+      "step": 12828
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.3424819796434302,
+      "learning_rate": 2.0008424739068925e-06,
+      "loss": 0.1703,
+      "step": 12829
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.4806275093891461,
+      "learning_rate": 2.0001217360300135e-06,
+      "loss": 0.1919,
+      "step": 12830
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.2805255318713749,
+      "learning_rate": 1.9994010955309064e-06,
+      "loss": 0.2107,
+      "step": 12831
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.3009247371360002,
+      "learning_rate": 1.998680552432965e-06,
+      "loss": 0.1707,
+      "step": 12832
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.2969573724393482,
+      "learning_rate": 1.997960106759579e-06,
+      "loss": 0.2191,
+      "step": 12833
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.078361533658275,
+      "learning_rate": 1.9972397585341336e-06,
+      "loss": 0.1399,
+      "step": 12834
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.4096166550506917,
+      "learning_rate": 1.996519507780015e-06,
+      "loss": 0.2259,
+      "step": 12835
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.4508372777207443,
+      "learning_rate": 1.995799354520598e-06,
+      "loss": 0.1717,
+      "step": 12836
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.2427197101975027,
+      "learning_rate": 1.9950792987792617e-06,
+      "loss": 0.1883,
+      "step": 12837
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.6951003579162343,
+      "learning_rate": 1.994359340579379e-06,
+      "loss": 0.2132,
+      "step": 12838
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.712208758948213,
+      "learning_rate": 1.9936394799443227e-06,
+      "loss": 0.2222,
+      "step": 12839
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.5102558277807336,
+      "learning_rate": 1.9929197168974555e-06,
+      "loss": 0.1397,
+      "step": 12840
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.4899631318389046,
+      "learning_rate": 1.9922000514621438e-06,
+      "loss": 0.1836,
+      "step": 12841
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.1626036129942667,
+      "learning_rate": 1.9914804836617474e-06,
+      "loss": 0.1909,
+      "step": 12842
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.8646041284676667,
+      "learning_rate": 1.9907610135196255e-06,
+      "loss": 0.1902,
+      "step": 12843
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.0844536993919596,
+      "learning_rate": 1.9900416410591324e-06,
+      "loss": 0.191,
+      "step": 12844
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.9185209034071047,
+      "learning_rate": 1.989322366303617e-06,
+      "loss": 0.1715,
+      "step": 12845
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.436178880520987,
+      "learning_rate": 1.988603189276431e-06,
+      "loss": 0.1948,
+      "step": 12846
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.8015576590004003,
+      "learning_rate": 1.987884110000915e-06,
+      "loss": 0.2211,
+      "step": 12847
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.7351667341167187,
+      "learning_rate": 1.9871651285004147e-06,
+      "loss": 0.1601,
+      "step": 12848
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.4027652532057209,
+      "learning_rate": 1.9864462447982652e-06,
+      "loss": 0.1428,
+      "step": 12849
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.3481917657974352,
+      "learning_rate": 1.985727458917803e-06,
+      "loss": 0.134,
+      "step": 12850
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.4809495565042434,
+      "learning_rate": 1.9850087708823617e-06,
+      "loss": 0.2332,
+      "step": 12851
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.331424916332069,
+      "learning_rate": 1.984290180715269e-06,
+      "loss": 0.1707,
+      "step": 12852
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.5829472750855476,
+      "learning_rate": 1.9835716884398534e-06,
+      "loss": 0.1805,
+      "step": 12853
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.2831194427934758,
+      "learning_rate": 1.9828532940794325e-06,
+      "loss": 0.1919,
+      "step": 12854
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.298490801725943,
+      "learning_rate": 1.9821349976573296e-06,
+      "loss": 0.1528,
+      "step": 12855
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.4037089702252188,
+      "learning_rate": 1.9814167991968596e-06,
+      "loss": 0.1859,
+      "step": 12856
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.3772337553143363,
+      "learning_rate": 1.9806986987213384e-06,
+      "loss": 0.1604,
+      "step": 12857
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.7160611187214192,
+      "learning_rate": 1.9799806962540713e-06,
+      "loss": 0.2094,
+      "step": 12858
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.2532003754019099,
+      "learning_rate": 1.9792627918183677e-06,
+      "loss": 0.157,
+      "step": 12859
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.3682620076770082,
+      "learning_rate": 1.978544985437531e-06,
+      "loss": 0.1924,
+      "step": 12860
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.1020505818107809,
+      "learning_rate": 1.9778272771348613e-06,
+      "loss": 0.1443,
+      "step": 12861
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.631275713195425,
+      "learning_rate": 1.977109666933658e-06,
+      "loss": 0.165,
+      "step": 12862
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.1158442191045708,
+      "learning_rate": 1.9763921548572113e-06,
+      "loss": 0.1652,
+      "step": 12863
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.8588142711569016,
+      "learning_rate": 1.975674740928816e-06,
+      "loss": 0.2023,
+      "step": 12864
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.2540948197178046,
+      "learning_rate": 1.9749574251717556e-06,
+      "loss": 0.1875,
+      "step": 12865
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.3991233326695716,
+      "learning_rate": 1.9742402076093186e-06,
+      "loss": 0.2032,
+      "step": 12866
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.246654856519848,
+      "learning_rate": 1.9735230882647826e-06,
+      "loss": 0.1482,
+      "step": 12867
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.259855293808643,
+      "learning_rate": 1.972806067161428e-06,
+      "loss": 0.1566,
+      "step": 12868
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.0936924578652791,
+      "learning_rate": 1.9720891443225287e-06,
+      "loss": 0.168,
+      "step": 12869
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.604564147375715,
+      "learning_rate": 1.971372319771357e-06,
+      "loss": 0.2067,
+      "step": 12870
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.686842366397943,
+      "learning_rate": 1.9706555935311835e-06,
+      "loss": 0.1876,
+      "step": 12871
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.9711460239092834,
+      "learning_rate": 1.969938965625269e-06,
+      "loss": 0.1383,
+      "step": 12872
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.52756576680579,
+      "learning_rate": 1.9692224360768786e-06,
+      "loss": 0.199,
+      "step": 12873
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.4275385334840107,
+      "learning_rate": 1.9685060049092704e-06,
+      "loss": 0.1861,
+      "step": 12874
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.791973028678237,
+      "learning_rate": 1.967789672145702e-06,
+      "loss": 0.2335,
+      "step": 12875
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.9801770950525973,
+      "learning_rate": 1.9670734378094237e-06,
+      "loss": 0.148,
+      "step": 12876
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.2572106437866477,
+      "learning_rate": 1.9663573019236856e-06,
+      "loss": 0.1856,
+      "step": 12877
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.2338485319374708,
+      "learning_rate": 1.965641264511733e-06,
+      "loss": 0.2201,
+      "step": 12878
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.0378140212716678,
+      "learning_rate": 1.9649253255968104e-06,
+      "loss": 0.1862,
+      "step": 12879
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.6000429597503874,
+      "learning_rate": 1.9642094852021586e-06,
+      "loss": 0.1526,
+      "step": 12880
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.1904589060326196,
+      "learning_rate": 1.963493743351011e-06,
+      "loss": 0.1692,
+      "step": 12881
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.2892385362589243,
+      "learning_rate": 1.9627781000666046e-06,
+      "loss": 0.1552,
+      "step": 12882
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.6167073343223637,
+      "learning_rate": 1.9620625553721656e-06,
+      "loss": 0.2265,
+      "step": 12883
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.7043155524878297,
+      "learning_rate": 1.9613471092909248e-06,
+      "loss": 0.2359,
+      "step": 12884
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.2469321811442702,
+      "learning_rate": 1.9606317618461023e-06,
+      "loss": 0.1924,
+      "step": 12885
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.3120216536775746,
+      "learning_rate": 1.9599165130609205e-06,
+      "loss": 0.1555,
+      "step": 12886
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.3537517358182307,
+      "learning_rate": 1.9592013629585972e-06,
+      "loss": 0.1953,
+      "step": 12887
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.2542033809006774,
+      "learning_rate": 1.958486311562346e-06,
+      "loss": 0.1964,
+      "step": 12888
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.4969402811671937,
+      "learning_rate": 1.9577713588953797e-06,
+      "loss": 0.2074,
+      "step": 12889
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.828259207538908,
+      "learning_rate": 1.957056504980903e-06,
+      "loss": 0.1602,
+      "step": 12890
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.370905034481034,
+      "learning_rate": 1.956341749842122e-06,
+      "loss": 0.1989,
+      "step": 12891
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.6998245211801506,
+      "learning_rate": 1.9556270935022383e-06,
+      "loss": 0.1716,
+      "step": 12892
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.40536505087059,
+      "learning_rate": 1.9549125359844504e-06,
+      "loss": 0.1829,
+      "step": 12893
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.4813959416293705,
+      "learning_rate": 1.9541980773119516e-06,
+      "loss": 0.203,
+      "step": 12894
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.8906920181959107,
+      "learning_rate": 1.9534837175079353e-06,
+      "loss": 0.1754,
+      "step": 12895
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.7242919007561819,
+      "learning_rate": 1.9527694565955886e-06,
+      "loss": 0.2098,
+      "step": 12896
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.5522982554687768,
+      "learning_rate": 1.952055294598098e-06,
+      "loss": 0.1868,
+      "step": 12897
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.4442843569197952,
+      "learning_rate": 1.951341231538647e-06,
+      "loss": 0.2336,
+      "step": 12898
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.050994633063823,
+      "learning_rate": 1.950627267440411e-06,
+      "loss": 0.2012,
+      "step": 12899
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.940179896628678,
+      "learning_rate": 1.9499134023265688e-06,
+      "loss": 0.0839,
+      "step": 12900
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.5735460694708236,
+      "learning_rate": 1.94919963622029e-06,
+      "loss": 0.1991,
+      "step": 12901
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.4063494117354483,
+      "learning_rate": 1.948485969144747e-06,
+      "loss": 0.2292,
+      "step": 12902
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.042423861250601,
+      "learning_rate": 1.9477724011231026e-06,
+      "loss": 0.2116,
+      "step": 12903
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.91359278029522,
+      "learning_rate": 1.9470589321785203e-06,
+      "loss": 0.1837,
+      "step": 12904
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.3684245208121233,
+      "learning_rate": 1.946345562334162e-06,
+      "loss": 0.1881,
+      "step": 12905
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.3359779970823842,
+      "learning_rate": 1.945632291613182e-06,
+      "loss": 0.1891,
+      "step": 12906
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.3687423039572708,
+      "learning_rate": 1.944919120038736e-06,
+      "loss": 0.2011,
+      "step": 12907
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.38058725559746,
+      "learning_rate": 1.9442060476339703e-06,
+      "loss": 0.2087,
+      "step": 12908
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.4374236210973603,
+      "learning_rate": 1.943493074422034e-06,
+      "loss": 0.2156,
+      "step": 12909
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.397380818563472,
+      "learning_rate": 1.94278020042607e-06,
+      "loss": 0.1738,
+      "step": 12910
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.2276786536674968,
+      "learning_rate": 1.9420674256692203e-06,
+      "loss": 0.1836,
+      "step": 12911
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.2847062456437777,
+      "learning_rate": 1.941354750174618e-06,
+      "loss": 0.1996,
+      "step": 12912
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.5966340294705599,
+      "learning_rate": 1.9406421739654004e-06,
+      "loss": 0.1832,
+      "step": 12913
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.1657279313025555,
+      "learning_rate": 1.9399296970646963e-06,
+      "loss": 0.2112,
+      "step": 12914
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.4453698894344733,
+      "learning_rate": 1.9392173194956344e-06,
+      "loss": 0.2403,
+      "step": 12915
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.2619336125392038,
+      "learning_rate": 1.93850504128134e-06,
+      "loss": 0.125,
+      "step": 12916
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.358881268779916,
+      "learning_rate": 1.937792862444931e-06,
+      "loss": 0.1802,
+      "step": 12917
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.487084455709692,
+      "learning_rate": 1.9370807830095278e-06,
+      "loss": 0.1647,
+      "step": 12918
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.3442804875175587,
+      "learning_rate": 1.9363688029982426e-06,
+      "loss": 0.202,
+      "step": 12919
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.2183228318781167,
+      "learning_rate": 1.935656922434189e-06,
+      "loss": 0.1253,
+      "step": 12920
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.0885599631810567,
+      "learning_rate": 1.9349451413404717e-06,
+      "loss": 0.2414,
+      "step": 12921
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.3145590976094672,
+      "learning_rate": 1.9342334597401984e-06,
+      "loss": 0.1736,
+      "step": 12922
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.354756863676101,
+      "learning_rate": 1.9335218776564694e-06,
+      "loss": 0.2056,
+      "step": 12923
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.3196424626540715,
+      "learning_rate": 1.932810395112384e-06,
+      "loss": 0.1817,
+      "step": 12924
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.1394259931167336,
+      "learning_rate": 1.9320990121310384e-06,
+      "loss": 0.1467,
+      "step": 12925
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.0645766650262902,
+      "learning_rate": 1.9313877287355217e-06,
+      "loss": 0.1473,
+      "step": 12926
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.3584854583510393,
+      "learning_rate": 1.9306765449489234e-06,
+      "loss": 0.178,
+      "step": 12927
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.0386880866912758,
+      "learning_rate": 1.9299654607943295e-06,
+      "loss": 0.14,
+      "step": 12928
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.0523775943448184,
+      "learning_rate": 1.9292544762948246e-06,
+      "loss": 0.1632,
+      "step": 12929
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.020232390695318,
+      "learning_rate": 1.928543591473482e-06,
+      "loss": 0.1696,
+      "step": 12930
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.5333846015194206,
+      "learning_rate": 1.927832806353382e-06,
+      "loss": 0.2446,
+      "step": 12931
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.2423957494197988,
+      "learning_rate": 1.9271221209575952e-06,
+      "loss": 0.2117,
+      "step": 12932
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.3553115907524846,
+      "learning_rate": 1.9264115353091915e-06,
+      "loss": 0.1634,
+      "step": 12933
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.0790444889769435,
+      "learning_rate": 1.9257010494312384e-06,
+      "loss": 0.1848,
+      "step": 12934
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.36255718885494,
+      "learning_rate": 1.924990663346796e-06,
+      "loss": 0.1818,
+      "step": 12935
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.371836463847524,
+      "learning_rate": 1.924280377078926e-06,
+      "loss": 0.1979,
+      "step": 12936
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.171519320870818,
+      "learning_rate": 1.9235701906506816e-06,
+      "loss": 0.1227,
+      "step": 12937
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.0209236251840816,
+      "learning_rate": 1.9228601040851204e-06,
+      "loss": 0.1161,
+      "step": 12938
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.188617813881736,
+      "learning_rate": 1.9221501174052877e-06,
+      "loss": 0.2051,
+      "step": 12939
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.1766411307961766,
+      "learning_rate": 1.9214402306342316e-06,
+      "loss": 0.1742,
+      "step": 12940
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.1141818335770963,
+      "learning_rate": 1.9207304437949964e-06,
+      "loss": 0.117,
+      "step": 12941
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.217638523731999,
+      "learning_rate": 1.9200207569106216e-06,
+      "loss": 0.1428,
+      "step": 12942
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.1049025234590377,
+      "learning_rate": 1.9193111700041454e-06,
+      "loss": 0.2305,
+      "step": 12943
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.105602168736439,
+      "learning_rate": 1.9186016830985993e-06,
+      "loss": 0.1844,
+      "step": 12944
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.4512616587982932,
+      "learning_rate": 1.917892296217014e-06,
+      "loss": 0.1793,
+      "step": 12945
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.1987293624095898,
+      "learning_rate": 1.9171830093824173e-06,
+      "loss": 0.2208,
+      "step": 12946
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.0140836684099765,
+      "learning_rate": 1.9164738226178325e-06,
+      "loss": 0.1334,
+      "step": 12947
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.1082260636385022,
+      "learning_rate": 1.9157647359462828e-06,
+      "loss": 0.1586,
+      "step": 12948
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.3628024050017886,
+      "learning_rate": 1.915055749390781e-06,
+      "loss": 0.166,
+      "step": 12949
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.5178910967391068,
+      "learning_rate": 1.914346862974344e-06,
+      "loss": 0.2001,
+      "step": 12950
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.4228763617433897,
+      "learning_rate": 1.913638076719982e-06,
+      "loss": 0.2093,
+      "step": 12951
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.3575338852589132,
+      "learning_rate": 1.9129293906507047e-06,
+      "loss": 0.1512,
+      "step": 12952
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.4188788018387832,
+      "learning_rate": 1.9122208047895124e-06,
+      "loss": 0.2134,
+      "step": 12953
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.1635241479701732,
+      "learning_rate": 1.911512319159411e-06,
+      "loss": 0.1887,
+      "step": 12954
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.8680315849915233,
+      "learning_rate": 1.9108039337833934e-06,
+      "loss": 0.1805,
+      "step": 12955
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.2326528881308576,
+      "learning_rate": 1.9100956486844564e-06,
+      "loss": 0.2154,
+      "step": 12956
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.2141513967432014,
+      "learning_rate": 1.9093874638855935e-06,
+      "loss": 0.1533,
+      "step": 12957
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.1365958897896495,
+      "learning_rate": 1.9086793794097885e-06,
+      "loss": 0.2157,
+      "step": 12958
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.4267750829738874,
+      "learning_rate": 1.9079713952800294e-06,
+      "loss": 0.1937,
+      "step": 12959
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.379857362052188,
+      "learning_rate": 1.907263511519296e-06,
+      "loss": 0.1959,
+      "step": 12960
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.415744029339853,
+      "learning_rate": 1.9065557281505699e-06,
+      "loss": 0.1981,
+      "step": 12961
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.0700705735159544,
+      "learning_rate": 1.9058480451968214e-06,
+      "loss": 0.1346,
+      "step": 12962
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.6806432976512216,
+      "learning_rate": 1.9051404626810249e-06,
+      "loss": 0.2377,
+      "step": 12963
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.9358026477420189,
+      "learning_rate": 1.9044329806261485e-06,
+      "loss": 0.1365,
+      "step": 12964
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.202072100845805,
+      "learning_rate": 1.9037255990551578e-06,
+      "loss": 0.1829,
+      "step": 12965
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.1003193502800659,
+      "learning_rate": 1.9030183179910161e-06,
+      "loss": 0.1572,
+      "step": 12966
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.0954043836187313,
+      "learning_rate": 1.9023111374566793e-06,
+      "loss": 0.184,
+      "step": 12967
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.502180164177197,
+      "learning_rate": 1.9016040574751043e-06,
+      "loss": 0.2229,
+      "step": 12968
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.479122002245179,
+      "learning_rate": 1.9008970780692431e-06,
+      "loss": 0.1782,
+      "step": 12969
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.0774894928364573,
+      "learning_rate": 1.9001901992620463e-06,
+      "loss": 0.1966,
+      "step": 12970
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.6209289191220968,
+      "learning_rate": 1.8994834210764569e-06,
+      "loss": 0.1679,
+      "step": 12971
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.4110003953731411,
+      "learning_rate": 1.8987767435354205e-06,
+      "loss": 0.219,
+      "step": 12972
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.4643216919445832,
+      "learning_rate": 1.8980701666618716e-06,
+      "loss": 0.155,
+      "step": 12973
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.4971601057402872,
+      "learning_rate": 1.89736369047875e-06,
+      "loss": 0.1797,
+      "step": 12974
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.4257125837957225,
+      "learning_rate": 1.8966573150089883e-06,
+      "loss": 0.1989,
+      "step": 12975
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.3678363241413438,
+      "learning_rate": 1.8959510402755133e-06,
+      "loss": 0.1455,
+      "step": 12976
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.8584231480501678,
+      "learning_rate": 1.895244866301252e-06,
+      "loss": 0.1325,
+      "step": 12977
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.5289985614251218,
+      "learning_rate": 1.894538793109128e-06,
+      "loss": 0.1908,
+      "step": 12978
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.2345337463863517,
+      "learning_rate": 1.893832820722062e-06,
+      "loss": 0.1993,
+      "step": 12979
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.4897332455041328,
+      "learning_rate": 1.8931269491629672e-06,
+      "loss": 0.1886,
+      "step": 12980
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.1354481126939506,
+      "learning_rate": 1.8924211784547587e-06,
+      "loss": 0.1644,
+      "step": 12981
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.460653200893403,
+      "learning_rate": 1.8917155086203449e-06,
+      "loss": 0.176,
+      "step": 12982
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.0406741177935213,
+      "learning_rate": 1.8910099396826337e-06,
+      "loss": 0.1581,
+      "step": 12983
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.3225681122346695,
+      "learning_rate": 1.890304471664529e-06,
+      "loss": 0.178,
+      "step": 12984
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.121540307094525,
+      "learning_rate": 1.8895991045889278e-06,
+      "loss": 0.2372,
+      "step": 12985
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.1466347857302752,
+      "learning_rate": 1.8888938384787287e-06,
+      "loss": 0.1991,
+      "step": 12986
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.1980806131404118,
+      "learning_rate": 1.888188673356824e-06,
+      "loss": 0.1972,
+      "step": 12987
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.304334806980028,
+      "learning_rate": 1.8874836092461063e-06,
+      "loss": 0.1442,
+      "step": 12988
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.8742481949700716,
+      "learning_rate": 1.8867786461694588e-06,
+      "loss": 0.1859,
+      "step": 12989
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.4682649714513485,
+      "learning_rate": 1.8860737841497684e-06,
+      "loss": 0.1968,
+      "step": 12990
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.1404743160368591,
+      "learning_rate": 1.8853690232099114e-06,
+      "loss": 0.1862,
+      "step": 12991
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.419361260942678,
+      "learning_rate": 1.884664363372768e-06,
+      "loss": 0.1931,
+      "step": 12992
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.0990710863543218,
+      "learning_rate": 1.8839598046612117e-06,
+      "loss": 0.1826,
+      "step": 12993
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.3161568730893538,
+      "learning_rate": 1.8832553470981103e-06,
+      "loss": 0.2086,
+      "step": 12994
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.5128458143985573,
+      "learning_rate": 1.8825509907063328e-06,
+      "loss": 0.1767,
+      "step": 12995
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.451885527955638,
+      "learning_rate": 1.8818467355087428e-06,
+      "loss": 0.1963,
+      "step": 12996
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.0741816080781543,
+      "learning_rate": 1.8811425815282024e-06,
+      "loss": 0.1385,
+      "step": 12997
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.2281477321214744,
+      "learning_rate": 1.880438528787566e-06,
+      "loss": 0.1861,
+      "step": 12998
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.3197834941467466,
+      "learning_rate": 1.879734577309688e-06,
+      "loss": 0.1756,
+      "step": 12999
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.1284817620224665,
+      "learning_rate": 1.8790307271174207e-06,
+      "loss": 0.1812,
+      "step": 13000
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.0316667780220674,
+      "learning_rate": 1.8783269782336105e-06,
+      "loss": 0.1394,
+      "step": 13001
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.4462083463019924,
+      "learning_rate": 1.8776233306811032e-06,
+      "loss": 0.2246,
+      "step": 13002
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.1371946780498963,
+      "learning_rate": 1.8769197844827363e-06,
+      "loss": 0.1873,
+      "step": 13003
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.4014925432129313,
+      "learning_rate": 1.8762163396613498e-06,
+      "loss": 0.1809,
+      "step": 13004
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.3688856015249877,
+      "learning_rate": 1.8755129962397762e-06,
+      "loss": 0.1915,
+      "step": 13005
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.5514889651025703,
+      "learning_rate": 1.87480975424085e-06,
+      "loss": 0.2157,
+      "step": 13006
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.2202690512411254,
+      "learning_rate": 1.8741066136873943e-06,
+      "loss": 0.1924,
+      "step": 13007
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.3543457909247856,
+      "learning_rate": 1.8734035746022366e-06,
+      "loss": 0.1686,
+      "step": 13008
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.3343736854028592,
+      "learning_rate": 1.8727006370081956e-06,
+      "loss": 0.1807,
+      "step": 13009
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.6020065436378403,
+      "learning_rate": 1.8719978009280903e-06,
+      "loss": 0.2059,
+      "step": 13010
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.1038194322564394,
+      "learning_rate": 1.8712950663847368e-06,
+      "loss": 0.1838,
+      "step": 13011
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.4460042998987321,
+      "learning_rate": 1.870592433400943e-06,
+      "loss": 0.1867,
+      "step": 13012
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.9368260305469592,
+      "learning_rate": 1.8698899019995175e-06,
+      "loss": 0.1641,
+      "step": 13013
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.7746582744964154,
+      "learning_rate": 1.8691874722032666e-06,
+      "loss": 0.1139,
+      "step": 13014
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.1187456575767591,
+      "learning_rate": 1.868485144034992e-06,
+      "loss": 0.1675,
+      "step": 13015
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.4601800199220005,
+      "learning_rate": 1.867782917517489e-06,
+      "loss": 0.1942,
+      "step": 13016
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.2834801106407236,
+      "learning_rate": 1.8670807926735535e-06,
+      "loss": 0.2237,
+      "step": 13017
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.423498340504057,
+      "learning_rate": 1.8663787695259772e-06,
+      "loss": 0.1618,
+      "step": 13018
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.509127675795902,
+      "learning_rate": 1.865676848097548e-06,
+      "loss": 0.2313,
+      "step": 13019
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.2616915271762978,
+      "learning_rate": 1.8649750284110517e-06,
+      "loss": 0.1921,
+      "step": 13020
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.4381300851432754,
+      "learning_rate": 1.8642733104892679e-06,
+      "loss": 0.2117,
+      "step": 13021
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.6281344096814283,
+      "learning_rate": 1.8635716943549748e-06,
+      "loss": 0.2063,
+      "step": 13022
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.5407268437310944,
+      "learning_rate": 1.8628701800309506e-06,
+      "loss": 0.2326,
+      "step": 13023
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.04702712846795,
+      "learning_rate": 1.862168767539962e-06,
+      "loss": 0.1655,
+      "step": 13024
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.0962650197891217,
+      "learning_rate": 1.8614674569047796e-06,
+      "loss": 0.1302,
+      "step": 13025
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.691691341556874,
+      "learning_rate": 1.8607662481481708e-06,
+      "loss": 0.2387,
+      "step": 13026
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.637349041503312,
+      "learning_rate": 1.8600651412928922e-06,
+      "loss": 0.1929,
+      "step": 13027
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.896621263083629,
+      "learning_rate": 1.8593641363617054e-06,
+      "loss": 0.2151,
+      "step": 13028
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.9099420454684648,
+      "learning_rate": 1.8586632333773657e-06,
+      "loss": 0.1882,
+      "step": 13029
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.305887561234874,
+      "learning_rate": 1.8579624323626228e-06,
+      "loss": 0.1666,
+      "step": 13030
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.669634271137726,
+      "learning_rate": 1.857261733340226e-06,
+      "loss": 0.1486,
+      "step": 13031
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.7382927712047969,
+      "learning_rate": 1.85656113633292e-06,
+      "loss": 0.1732,
+      "step": 13032
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.6363440974756613,
+      "learning_rate": 1.8558606413634495e-06,
+      "loss": 0.1937,
+      "step": 13033
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.0200870285235775,
+      "learning_rate": 1.8551602484545484e-06,
+      "loss": 0.1803,
+      "step": 13034
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.7905057224937506,
+      "learning_rate": 1.8544599576289547e-06,
+      "loss": 0.2001,
+      "step": 13035
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.9903859524708049,
+      "learning_rate": 1.8537597689093988e-06,
+      "loss": 0.194,
+      "step": 13036
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.9844795807279989,
+      "learning_rate": 1.8530596823186103e-06,
+      "loss": 0.1719,
+      "step": 13037
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.8982966570943667,
+      "learning_rate": 1.8523596978793162e-06,
+      "loss": 0.225,
+      "step": 13038
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.7104718824535459,
+      "learning_rate": 1.851659815614234e-06,
+      "loss": 0.2293,
+      "step": 13039
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.3694691882784662,
+      "learning_rate": 1.850960035546085e-06,
+      "loss": 0.1958,
+      "step": 13040
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.5561638057539176,
+      "learning_rate": 1.8502603576975858e-06,
+      "loss": 0.1938,
+      "step": 13041
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.2131340286335555,
+      "learning_rate": 1.8495607820914451e-06,
+      "loss": 0.1796,
+      "step": 13042
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.0895470382791623,
+      "learning_rate": 1.8488613087503742e-06,
+      "loss": 0.1896,
+      "step": 13043
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.136941606102441,
+      "learning_rate": 1.8481619376970762e-06,
+      "loss": 0.2077,
+      "step": 13044
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.1902904037844297,
+      "learning_rate": 1.8474626689542535e-06,
+      "loss": 0.1688,
+      "step": 13045
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.1786739782877318,
+      "learning_rate": 1.8467635025446058e-06,
+      "loss": 0.2242,
+      "step": 13046
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.0194525231552483,
+      "learning_rate": 1.8460644384908293e-06,
+      "loss": 0.1787,
+      "step": 13047
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.3731069321066582,
+      "learning_rate": 1.8453654768156138e-06,
+      "loss": 0.2056,
+      "step": 13048
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.3634408267170979,
+      "learning_rate": 1.8446666175416484e-06,
+      "loss": 0.2442,
+      "step": 13049
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.1388131899684255,
+      "learning_rate": 1.8439678606916195e-06,
+      "loss": 0.1439,
+      "step": 13050
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.3293085434609173,
+      "learning_rate": 1.8432692062882103e-06,
+      "loss": 0.1421,
+      "step": 13051
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.1092278960576731,
+      "learning_rate": 1.8425706543540956e-06,
+      "loss": 0.1935,
+      "step": 13052
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.412236262064391,
+      "learning_rate": 1.8418722049119536e-06,
+      "loss": 0.1777,
+      "step": 13053
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.5637783447886302,
+      "learning_rate": 1.8411738579844558e-06,
+      "loss": 0.2227,
+      "step": 13054
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.6483688385390431,
+      "learning_rate": 1.8404756135942708e-06,
+      "loss": 0.2124,
+      "step": 13055
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.4329561353924156,
+      "learning_rate": 1.8397774717640666e-06,
+      "loss": 0.1766,
+      "step": 13056
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.4401822462753806,
+      "learning_rate": 1.8390794325165006e-06,
+      "loss": 0.1802,
+      "step": 13057
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.3447334161942521,
+      "learning_rate": 1.838381495874234e-06,
+      "loss": 0.2598,
+      "step": 13058
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.1270228055306817,
+      "learning_rate": 1.8376836618599236e-06,
+      "loss": 0.1854,
+      "step": 13059
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.634792806201141,
+      "learning_rate": 1.836985930496219e-06,
+      "loss": 0.2283,
+      "step": 13060
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.9205477239874692,
+      "learning_rate": 1.836288301805771e-06,
+      "loss": 0.1891,
+      "step": 13061
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.4059227827632628,
+      "learning_rate": 1.8355907758112229e-06,
+      "loss": 0.2029,
+      "step": 13062
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.3181918708919078,
+      "learning_rate": 1.8348933525352175e-06,
+      "loss": 0.2084,
+      "step": 13063
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.399700553409036,
+      "learning_rate": 1.8341960320003938e-06,
+      "loss": 0.1438,
+      "step": 13064
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.6105476037985342,
+      "learning_rate": 1.833498814229389e-06,
+      "loss": 0.1978,
+      "step": 13065
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.790897407082967,
+      "learning_rate": 1.8328016992448323e-06,
+      "loss": 0.199,
+      "step": 13066
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.525552827097044,
+      "learning_rate": 1.8321046870693531e-06,
+      "loss": 0.1915,
+      "step": 13067
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.3477765458085553,
+      "learning_rate": 1.8314077777255778e-06,
+      "loss": 0.1851,
+      "step": 13068
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.851054600777963,
+      "learning_rate": 1.83071097123613e-06,
+      "loss": 0.216,
+      "step": 13069
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.4267179103652912,
+      "learning_rate": 1.8300142676236249e-06,
+      "loss": 0.1623,
+      "step": 13070
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.4642752599675282,
+      "learning_rate": 1.8293176669106793e-06,
+      "loss": 0.1762,
+      "step": 13071
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.9729727613886097,
+      "learning_rate": 1.8286211691199051e-06,
+      "loss": 0.2182,
+      "step": 13072
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.4830549494118572,
+      "learning_rate": 1.827924774273912e-06,
+      "loss": 0.222,
+      "step": 13073
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.3727872774033116,
+      "learning_rate": 1.8272284823953063e-06,
+      "loss": 0.1671,
+      "step": 13074
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.5300757494068769,
+      "learning_rate": 1.826532293506687e-06,
+      "loss": 0.2123,
+      "step": 13075
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.2027510458177115,
+      "learning_rate": 1.825836207630654e-06,
+      "loss": 0.1664,
+      "step": 13076
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.1850456925528052,
+      "learning_rate": 1.8251402247898054e-06,
+      "loss": 0.1702,
+      "step": 13077
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.2947810889073121,
+      "learning_rate": 1.8244443450067285e-06,
+      "loss": 0.163,
+      "step": 13078
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.0421249891347246,
+      "learning_rate": 1.8237485683040158e-06,
+      "loss": 0.1568,
+      "step": 13079
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.324538251429246,
+      "learning_rate": 1.8230528947042496e-06,
+      "loss": 0.201,
+      "step": 13080
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.254135295115422,
+      "learning_rate": 1.822357324230013e-06,
+      "loss": 0.1566,
+      "step": 13081
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.082613886264349,
+      "learning_rate": 1.8216618569038851e-06,
+      "loss": 0.1981,
+      "step": 13082
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.4743078264536502,
+      "learning_rate": 1.8209664927484432e-06,
+      "loss": 0.1922,
+      "step": 13083
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.258477805093477,
+      "learning_rate": 1.8202712317862548e-06,
+      "loss": 0.1978,
+      "step": 13084
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.5660988846205348,
+      "learning_rate": 1.819576074039891e-06,
+      "loss": 0.1825,
+      "step": 13085
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.3002222079864294,
+      "learning_rate": 1.8188810195319163e-06,
+      "loss": 0.1761,
+      "step": 13086
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.1725380166342643,
+      "learning_rate": 1.8181860682848952e-06,
+      "loss": 0.146,
+      "step": 13087
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.3999697346694178,
+      "learning_rate": 1.8174912203213818e-06,
+      "loss": 0.1938,
+      "step": 13088
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.307473879219537,
+      "learning_rate": 1.8167964756639334e-06,
+      "loss": 0.1562,
+      "step": 13089
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.5623198071528017,
+      "learning_rate": 1.8161018343351023e-06,
+      "loss": 0.1935,
+      "step": 13090
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.29748096298436,
+      "learning_rate": 1.8154072963574365e-06,
+      "loss": 0.2256,
+      "step": 13091
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.5672235711036058,
+      "learning_rate": 1.8147128617534825e-06,
+      "loss": 0.243,
+      "step": 13092
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.4841215117998368,
+      "learning_rate": 1.8140185305457791e-06,
+      "loss": 0.1889,
+      "step": 13093
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.0781159607881512,
+      "learning_rate": 1.8133243027568664e-06,
+      "loss": 0.1644,
+      "step": 13094
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.0986641383473408,
+      "learning_rate": 1.8126301784092808e-06,
+      "loss": 0.1484,
+      "step": 13095
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.1561595874955455,
+      "learning_rate": 1.8119361575255513e-06,
+      "loss": 0.1895,
+      "step": 13096
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.0411837998436713,
+      "learning_rate": 1.8112422401282082e-06,
+      "loss": 0.1846,
+      "step": 13097
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.9683350897387061,
+      "learning_rate": 1.8105484262397743e-06,
+      "loss": 0.1626,
+      "step": 13098
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.2355145011469217,
+      "learning_rate": 1.8098547158827728e-06,
+      "loss": 0.1908,
+      "step": 13099
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.0607369061224239,
+      "learning_rate": 1.8091611090797217e-06,
+      "loss": 0.1306,
+      "step": 13100
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.65947785822012,
+      "learning_rate": 1.8084676058531376e-06,
+      "loss": 0.1881,
+      "step": 13101
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.2848320062085083,
+      "learning_rate": 1.8077742062255281e-06,
+      "loss": 0.207,
+      "step": 13102
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.2389267283034253,
+      "learning_rate": 1.8070809102194042e-06,
+      "loss": 0.1792,
+      "step": 13103
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.3432645475301628,
+      "learning_rate": 1.8063877178572698e-06,
+      "loss": 0.2287,
+      "step": 13104
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.3634230695700083,
+      "learning_rate": 1.8056946291616279e-06,
+      "loss": 0.2071,
+      "step": 13105
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.409450806293956,
+      "learning_rate": 1.805001644154974e-06,
+      "loss": 0.2164,
+      "step": 13106
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.4443756907926497,
+      "learning_rate": 1.8043087628598033e-06,
+      "loss": 0.1705,
+      "step": 13107
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.2906161193680672,
+      "learning_rate": 1.8036159852986084e-06,
+      "loss": 0.1645,
+      "step": 13108
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.8697193648900017,
+      "learning_rate": 1.8029233114938766e-06,
+      "loss": 0.1552,
+      "step": 13109
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.6618639700940678,
+      "learning_rate": 1.802230741468094e-06,
+      "loss": 0.2375,
+      "step": 13110
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.328322098636422,
+      "learning_rate": 1.8015382752437389e-06,
+      "loss": 0.2229,
+      "step": 13111
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.2257925368568003,
+      "learning_rate": 1.8008459128432904e-06,
+      "loss": 0.188,
+      "step": 13112
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.1760019351900124,
+      "learning_rate": 1.800153654289225e-06,
+      "loss": 0.2108,
+      "step": 13113
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.0433661100397993,
+      "learning_rate": 1.7994614996040106e-06,
+      "loss": 0.1742,
+      "step": 13114
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.2141152433036548,
+      "learning_rate": 1.7987694488101175e-06,
+      "loss": 0.1931,
+      "step": 13115
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.9318998164180035,
+      "learning_rate": 1.7980775019300079e-06,
+      "loss": 0.1503,
+      "step": 13116
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.3516305233810688,
+      "learning_rate": 1.7973856589861433e-06,
+      "loss": 0.1919,
+      "step": 13117
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.5946488650585158,
+      "learning_rate": 1.796693920000982e-06,
+      "loss": 0.22,
+      "step": 13118
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.2999479929901585,
+      "learning_rate": 1.7960022849969799e-06,
+      "loss": 0.1589,
+      "step": 13119
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.1711292055347446,
+      "learning_rate": 1.7953107539965842e-06,
+      "loss": 0.1493,
+      "step": 13120
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.3498930212081315,
+      "learning_rate": 1.7946193270222444e-06,
+      "loss": 0.2099,
+      "step": 13121
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.9967460590788866,
+      "learning_rate": 1.7939280040964047e-06,
+      "loss": 0.1435,
+      "step": 13122
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.0708236587522784,
+      "learning_rate": 1.7932367852415066e-06,
+      "loss": 0.1484,
+      "step": 13123
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.0212269051904204,
+      "learning_rate": 1.7925456704799855e-06,
+      "loss": 0.2252,
+      "step": 13124
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.642506197557703,
+      "learning_rate": 1.7918546598342757e-06,
+      "loss": 0.218,
+      "step": 13125
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.6964383671771728,
+      "learning_rate": 1.7911637533268088e-06,
+      "loss": 0.1125,
+      "step": 13126
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.260052297931651,
+      "learning_rate": 1.7904729509800118e-06,
+      "loss": 0.172,
+      "step": 13127
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.334443199148046,
+      "learning_rate": 1.78978225281631e-06,
+      "loss": 0.2012,
+      "step": 13128
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.2545888920202333,
+      "learning_rate": 1.7890916588581208e-06,
+      "loss": 0.1979,
+      "step": 13129
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.9633264696532282,
+      "learning_rate": 1.7884011691278624e-06,
+      "loss": 0.1564,
+      "step": 13130
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.1490521440513948,
+      "learning_rate": 1.7877107836479507e-06,
+      "loss": 0.1732,
+      "step": 13131
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.3571607818889742,
+      "learning_rate": 1.7870205024407928e-06,
+      "loss": 0.2399,
+      "step": 13132
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.7988885751315002,
+      "learning_rate": 1.7863303255287977e-06,
+      "loss": 0.2054,
+      "step": 13133
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.0069776115308136,
+      "learning_rate": 1.7856402529343676e-06,
+      "loss": 0.2003,
+      "step": 13134
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.5913713647074113,
+      "learning_rate": 1.7849502846799027e-06,
+      "loss": 0.2031,
+      "step": 13135
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.6236488795784692,
+      "learning_rate": 1.7842604207878006e-06,
+      "loss": 0.148,
+      "step": 13136
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.0941674559795662,
+      "learning_rate": 1.783570661280456e-06,
+      "loss": 0.1203,
+      "step": 13137
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.3292732352128531,
+      "learning_rate": 1.7828810061802554e-06,
+      "loss": 0.2171,
+      "step": 13138
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.1734865234232617,
+      "learning_rate": 1.7821914555095876e-06,
+      "loss": 0.2301,
+      "step": 13139
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.3177528625121386,
+      "learning_rate": 1.7815020092908357e-06,
+      "loss": 0.1933,
+      "step": 13140
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.4324228145782494,
+      "learning_rate": 1.7808126675463806e-06,
+      "loss": 0.1665,
+      "step": 13141
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 7.875025737811217,
+      "learning_rate": 1.7801234302985958e-06,
+      "loss": 0.1735,
+      "step": 13142
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.237775353300713,
+      "learning_rate": 1.7794342975698565e-06,
+      "loss": 0.1731,
+      "step": 13143
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.6645359476177704,
+      "learning_rate": 1.778745269382532e-06,
+      "loss": 0.1961,
+      "step": 13144
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.1938205864144495,
+      "learning_rate": 1.7780563457589878e-06,
+      "loss": 0.157,
+      "step": 13145
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.286199702650201,
+      "learning_rate": 1.777367526721589e-06,
+      "loss": 0.1264,
+      "step": 13146
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.4388100521735778,
+      "learning_rate": 1.7766788122926921e-06,
+      "loss": 0.1564,
+      "step": 13147
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.0986905841472705,
+      "learning_rate": 1.775990202494654e-06,
+      "loss": 0.1983,
+      "step": 13148
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.3414300975334934,
+      "learning_rate": 1.7753016973498304e-06,
+      "loss": 0.254,
+      "step": 13149
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.4162230083270497,
+      "learning_rate": 1.7746132968805658e-06,
+      "loss": 0.2024,
+      "step": 13150
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.2476403943023529,
+      "learning_rate": 1.77392500110921e-06,
+      "loss": 0.1665,
+      "step": 13151
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.451833607473199,
+      "learning_rate": 1.7732368100581022e-06,
+      "loss": 0.2089,
+      "step": 13152
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.324874141851558,
+      "learning_rate": 1.772548723749583e-06,
+      "loss": 0.1864,
+      "step": 13153
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.0444993350189853,
+      "learning_rate": 1.771860742205988e-06,
+      "loss": 0.1394,
+      "step": 13154
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.2879389982016258,
+      "learning_rate": 1.771172865449652e-06,
+      "loss": 0.2076,
+      "step": 13155
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.4354164559718012,
+      "learning_rate": 1.770485093502899e-06,
+      "loss": 0.1676,
+      "step": 13156
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.4424039313565917,
+      "learning_rate": 1.7697974263880573e-06,
+      "loss": 0.207,
+      "step": 13157
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.2944579465712864,
+      "learning_rate": 1.7691098641274485e-06,
+      "loss": 0.1315,
+      "step": 13158
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.4331086504769304,
+      "learning_rate": 1.768422406743393e-06,
+      "loss": 0.1743,
+      "step": 13159
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.5385909495749992,
+      "learning_rate": 1.7677350542582033e-06,
+      "loss": 0.2024,
+      "step": 13160
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.0354761960757861,
+      "learning_rate": 1.7670478066941922e-06,
+      "loss": 0.1569,
+      "step": 13161
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.6962351169527866,
+      "learning_rate": 1.7663606640736681e-06,
+      "loss": 0.1218,
+      "step": 13162
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.1057071428745473,
+      "learning_rate": 1.7656736264189367e-06,
+      "loss": 0.1301,
+      "step": 13163
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.8658673315953116,
+      "learning_rate": 1.7649866937523007e-06,
+      "loss": 0.1586,
+      "step": 13164
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.4502321951443442,
+      "learning_rate": 1.7642998660960554e-06,
+      "loss": 0.2239,
+      "step": 13165
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.3720740701752925,
+      "learning_rate": 1.7636131434724969e-06,
+      "loss": 0.1476,
+      "step": 13166
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.2715828472158563,
+      "learning_rate": 1.7629265259039185e-06,
+      "loss": 0.2496,
+      "step": 13167
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.4142697719810593,
+      "learning_rate": 1.762240013412605e-06,
+      "loss": 0.183,
+      "step": 13168
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.2929223158874095,
+      "learning_rate": 1.7615536060208438e-06,
+      "loss": 0.2202,
+      "step": 13169
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.0067615722069785,
+      "learning_rate": 1.7608673037509134e-06,
+      "loss": 0.1496,
+      "step": 13170
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.4701086263760836,
+      "learning_rate": 1.760181106625093e-06,
+      "loss": 0.1846,
+      "step": 13171
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.4353257482545572,
+      "learning_rate": 1.759495014665657e-06,
+      "loss": 0.1464,
+      "step": 13172
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.5175924435483783,
+      "learning_rate": 1.7588090278948776e-06,
+      "loss": 0.1907,
+      "step": 13173
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.650944137299955,
+      "learning_rate": 1.7581231463350196e-06,
+      "loss": 0.1647,
+      "step": 13174
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.3013097130617497,
+      "learning_rate": 1.7574373700083485e-06,
+      "loss": 0.1519,
+      "step": 13175
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.0652852087198263,
+      "learning_rate": 1.7567516989371246e-06,
+      "loss": 0.1674,
+      "step": 13176
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.272149534612411,
+      "learning_rate": 1.7560661331436073e-06,
+      "loss": 0.1885,
+      "step": 13177
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.2777450701021922,
+      "learning_rate": 1.7553806726500471e-06,
+      "loss": 0.2066,
+      "step": 13178
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.3967354395400888,
+      "learning_rate": 1.7546953174786962e-06,
+      "loss": 0.1629,
+      "step": 13179
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.2363815567258116,
+      "learning_rate": 1.7540100676518019e-06,
+      "loss": 0.1742,
+      "step": 13180
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.1718932595419893,
+      "learning_rate": 1.7533249231916066e-06,
+      "loss": 0.1418,
+      "step": 13181
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.6628698095444099,
+      "learning_rate": 1.7526398841203534e-06,
+      "loss": 0.16,
+      "step": 13182
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.9991118500524567,
+      "learning_rate": 1.7519549504602751e-06,
+      "loss": 0.2431,
+      "step": 13183
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.2308392254612617,
+      "learning_rate": 1.7512701222336071e-06,
+      "loss": 0.1817,
+      "step": 13184
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.9148939174315125,
+      "learning_rate": 1.7505853994625805e-06,
+      "loss": 0.1432,
+      "step": 13185
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.0713166343736487,
+      "learning_rate": 1.7499007821694187e-06,
+      "loss": 0.122,
+      "step": 13186
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.2683196463088675,
+      "learning_rate": 1.7492162703763482e-06,
+      "loss": 0.1407,
+      "step": 13187
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.104953996655129,
+      "learning_rate": 1.7485318641055848e-06,
+      "loss": 0.1921,
+      "step": 13188
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.29143882544222,
+      "learning_rate": 1.7478475633793474e-06,
+      "loss": 0.1764,
+      "step": 13189
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.419590605721443,
+      "learning_rate": 1.747163368219848e-06,
+      "loss": 0.1464,
+      "step": 13190
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.2634720917069937,
+      "learning_rate": 1.746479278649298e-06,
+      "loss": 0.179,
+      "step": 13191
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.3263026918771506,
+      "learning_rate": 1.7457952946898998e-06,
+      "loss": 0.2385,
+      "step": 13192
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.7217995206486374,
+      "learning_rate": 1.7451114163638578e-06,
+      "loss": 0.2231,
+      "step": 13193
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.2595729473534234,
+      "learning_rate": 1.7444276436933706e-06,
+      "loss": 0.1627,
+      "step": 13194
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.4853546624085474,
+      "learning_rate": 1.7437439767006343e-06,
+      "loss": 0.1952,
+      "step": 13195
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.2457756784530518,
+      "learning_rate": 1.7430604154078423e-06,
+      "loss": 0.185,
+      "step": 13196
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.4977748330400302,
+      "learning_rate": 1.7423769598371809e-06,
+      "loss": 0.1492,
+      "step": 13197
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.1767580467635432,
+      "learning_rate": 1.7416936100108362e-06,
+      "loss": 0.2095,
+      "step": 13198
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.614850043500015,
+      "learning_rate": 1.7410103659509907e-06,
+      "loss": 0.1967,
+      "step": 13199
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.509348121940811,
+      "learning_rate": 1.740327227679825e-06,
+      "loss": 0.1903,
+      "step": 13200
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.8986558648817077,
+      "learning_rate": 1.7396441952195098e-06,
+      "loss": 0.1722,
+      "step": 13201
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.2617816289344743,
+      "learning_rate": 1.738961268592219e-06,
+      "loss": 0.1708,
+      "step": 13202
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.4605995083038392,
+      "learning_rate": 1.7382784478201225e-06,
+      "loss": 0.2021,
+      "step": 13203
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.671187544692119,
+      "learning_rate": 1.7375957329253812e-06,
+      "loss": 0.2032,
+      "step": 13204
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.758435736479552,
+      "learning_rate": 1.73691312393016e-06,
+      "loss": 0.2359,
+      "step": 13205
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.9851507657145738,
+      "learning_rate": 1.7362306208566138e-06,
+      "loss": 0.139,
+      "step": 13206
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.3005702274288546,
+      "learning_rate": 1.7355482237268983e-06,
+      "loss": 0.1685,
+      "step": 13207
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.308420615343526,
+      "learning_rate": 1.734865932563165e-06,
+      "loss": 0.1938,
+      "step": 13208
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.7522571049093245,
+      "learning_rate": 1.7341837473875623e-06,
+      "loss": 0.2005,
+      "step": 13209
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.3657420555594535,
+      "learning_rate": 1.7335016682222317e-06,
+      "loss": 0.2002,
+      "step": 13210
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.4649326551796609,
+      "learning_rate": 1.7328196950893155e-06,
+      "loss": 0.212,
+      "step": 13211
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.3844115984365113,
+      "learning_rate": 1.7321378280109508e-06,
+      "loss": 0.2133,
+      "step": 13212
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.4429674919961466,
+      "learning_rate": 1.7314560670092706e-06,
+      "loss": 0.2066,
+      "step": 13213
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.9675621856867344,
+      "learning_rate": 1.7307744121064085e-06,
+      "loss": 0.1445,
+      "step": 13214
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.9830460925789952,
+      "learning_rate": 1.7300928633244867e-06,
+      "loss": 0.1397,
+      "step": 13215
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.5379010140802305,
+      "learning_rate": 1.7294114206856317e-06,
+      "loss": 0.1786,
+      "step": 13216
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.2481966003488112,
+      "learning_rate": 1.7287300842119619e-06,
+      "loss": 0.1946,
+      "step": 13217
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.1931708603723217,
+      "learning_rate": 1.7280488539255962e-06,
+      "loss": 0.201,
+      "step": 13218
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.9226848942724393,
+      "learning_rate": 1.7273677298486452e-06,
+      "loss": 0.2302,
+      "step": 13219
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.1408820646449083,
+      "learning_rate": 1.7266867120032194e-06,
+      "loss": 0.2117,
+      "step": 13220
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.480121330220599,
+      "learning_rate": 1.7260058004114267e-06,
+      "loss": 0.1809,
+      "step": 13221
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.4684383792229176,
+      "learning_rate": 1.7253249950953672e-06,
+      "loss": 0.2193,
+      "step": 13222
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.4732049784841115,
+      "learning_rate": 1.7246442960771426e-06,
+      "loss": 0.1967,
+      "step": 13223
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.0642511088607671,
+      "learning_rate": 1.723963703378846e-06,
+      "loss": 0.1832,
+      "step": 13224
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.9654003460666245,
+      "learning_rate": 1.7232832170225716e-06,
+      "loss": 0.1271,
+      "step": 13225
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.4150813631483916,
+      "learning_rate": 1.7226028370304081e-06,
+      "loss": 0.2257,
+      "step": 13226
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.606071113206458,
+      "learning_rate": 1.721922563424443e-06,
+      "loss": 0.1737,
+      "step": 13227
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.8854875990308828,
+      "learning_rate": 1.721242396226755e-06,
+      "loss": 0.2162,
+      "step": 13228
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.1645615487230696,
+      "learning_rate": 1.7205623354594236e-06,
+      "loss": 0.1618,
+      "step": 13229
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.6782164852416228,
+      "learning_rate": 1.7198823811445254e-06,
+      "loss": 0.2121,
+      "step": 13230
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.2372434762026177,
+      "learning_rate": 1.7192025333041307e-06,
+      "loss": 0.1457,
+      "step": 13231
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.536661327551726,
+      "learning_rate": 1.7185227919603104e-06,
+      "loss": 0.1892,
+      "step": 13232
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.5113759946008227,
+      "learning_rate": 1.717843157135125e-06,
+      "loss": 0.1884,
+      "step": 13233
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.0968577749035537,
+      "learning_rate": 1.7171636288506388e-06,
+      "loss": 0.23,
+      "step": 13234
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.2769852238620683,
+      "learning_rate": 1.7164842071289085e-06,
+      "loss": 0.1691,
+      "step": 13235
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.315368095830793,
+      "learning_rate": 1.7158048919919906e-06,
+      "loss": 0.1614,
+      "step": 13236
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.7016820525808882,
+      "learning_rate": 1.7151256834619334e-06,
+      "loss": 0.165,
+      "step": 13237
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.1240468374910455,
+      "learning_rate": 1.7144465815607847e-06,
+      "loss": 0.2122,
+      "step": 13238
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.645070336480985,
+      "learning_rate": 1.713767586310592e-06,
+      "loss": 0.1092,
+      "step": 13239
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.4129143685224976,
+      "learning_rate": 1.7130886977333905e-06,
+      "loss": 0.1733,
+      "step": 13240
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.297726202450988,
+      "learning_rate": 1.712409915851222e-06,
+      "loss": 0.1566,
+      "step": 13241
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.6380330364781495,
+      "learning_rate": 1.7117312406861165e-06,
+      "loss": 0.1978,
+      "step": 13242
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.4699191958899493,
+      "learning_rate": 1.7110526722601057e-06,
+      "loss": 0.2427,
+      "step": 13243
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.2723956468274664,
+      "learning_rate": 1.7103742105952164e-06,
+      "loss": 0.1485,
+      "step": 13244
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.0258469937502428,
+      "learning_rate": 1.7096958557134734e-06,
+      "loss": 0.1705,
+      "step": 13245
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.1140465164108018,
+      "learning_rate": 1.7090176076368937e-06,
+      "loss": 0.2007,
+      "step": 13246
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.374399080963252,
+      "learning_rate": 1.7083394663874947e-06,
+      "loss": 0.1428,
+      "step": 13247
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.2973734862788,
+      "learning_rate": 1.7076614319872897e-06,
+      "loss": 0.1823,
+      "step": 13248
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.3349371784498612,
+      "learning_rate": 1.7069835044582878e-06,
+      "loss": 0.2219,
+      "step": 13249
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.0217801608542914,
+      "learning_rate": 1.7063056838224968e-06,
+      "loss": 0.2089,
+      "step": 13250
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.3093058555914618,
+      "learning_rate": 1.705627970101915e-06,
+      "loss": 0.1701,
+      "step": 13251
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.3078186928295963,
+      "learning_rate": 1.7049503633185443e-06,
+      "loss": 0.2371,
+      "step": 13252
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.566112122068757,
+      "learning_rate": 1.7042728634943789e-06,
+      "loss": 0.2017,
+      "step": 13253
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.5285847127437304,
+      "learning_rate": 1.703595470651414e-06,
+      "loss": 0.2237,
+      "step": 13254
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.4272675957954035,
+      "learning_rate": 1.7029181848116339e-06,
+      "loss": 0.1654,
+      "step": 13255
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.6084938437926863,
+      "learning_rate": 1.702241005997025e-06,
+      "loss": 0.1878,
+      "step": 13256
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.42246614471199,
+      "learning_rate": 1.7015639342295715e-06,
+      "loss": 0.1836,
+      "step": 13257
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.4160586330854168,
+      "learning_rate": 1.7008869695312474e-06,
+      "loss": 0.1944,
+      "step": 13258
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.4974194074240124,
+      "learning_rate": 1.700210111924031e-06,
+      "loss": 0.2388,
+      "step": 13259
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.3633718706414912,
+      "learning_rate": 1.6995333614298908e-06,
+      "loss": 0.1523,
+      "step": 13260
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.380506969628776,
+      "learning_rate": 1.6988567180707949e-06,
+      "loss": 0.2085,
+      "step": 13261
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.412928120962182,
+      "learning_rate": 1.6981801818687088e-06,
+      "loss": 0.201,
+      "step": 13262
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.6369848330945131,
+      "learning_rate": 1.697503752845594e-06,
+      "loss": 0.1711,
+      "step": 13263
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.2949418660457634,
+      "learning_rate": 1.696827431023405e-06,
+      "loss": 0.1373,
+      "step": 13264
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.110887742739038,
+      "learning_rate": 1.6961512164240967e-06,
+      "loss": 0.2272,
+      "step": 13265
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.170442180590381,
+      "learning_rate": 1.6954751090696203e-06,
+      "loss": 0.2079,
+      "step": 13266
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.3449738080592522,
+      "learning_rate": 1.6947991089819217e-06,
+      "loss": 0.1824,
+      "step": 13267
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.4391376455031908,
+      "learning_rate": 1.6941232161829464e-06,
+      "loss": 0.1728,
+      "step": 13268
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.5853134910163813,
+      "learning_rate": 1.6934474306946308e-06,
+      "loss": 0.1469,
+      "step": 13269
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.372233631554859,
+      "learning_rate": 1.6927717525389132e-06,
+      "loss": 0.2019,
+      "step": 13270
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.2171578583639997,
+      "learning_rate": 1.6920961817377262e-06,
+      "loss": 0.1613,
+      "step": 13271
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.1959495093682384,
+      "learning_rate": 1.6914207183130011e-06,
+      "loss": 0.1905,
+      "step": 13272
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.6013420976780237,
+      "learning_rate": 1.6907453622866605e-06,
+      "loss": 0.2193,
+      "step": 13273
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.7750066650460254,
+      "learning_rate": 1.690070113680628e-06,
+      "loss": 0.1912,
+      "step": 13274
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.407436363151996,
+      "learning_rate": 1.6893949725168246e-06,
+      "loss": 0.1776,
+      "step": 13275
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.1697510290461346,
+      "learning_rate": 1.6887199388171627e-06,
+      "loss": 0.2342,
+      "step": 13276
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.7383215910270224,
+      "learning_rate": 1.6880450126035576e-06,
+      "loss": 0.1832,
+      "step": 13277
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.3205071762843998,
+      "learning_rate": 1.6873701938979137e-06,
+      "loss": 0.1762,
+      "step": 13278
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.374242088744265,
+      "learning_rate": 1.6866954827221382e-06,
+      "loss": 0.2068,
+      "step": 13279
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.4529868183269892,
+      "learning_rate": 1.686020879098133e-06,
+      "loss": 0.1902,
+      "step": 13280
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.1032082053513876,
+      "learning_rate": 1.6853463830477972e-06,
+      "loss": 0.194,
+      "step": 13281
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.3773376360847889,
+      "learning_rate": 1.6846719945930224e-06,
+      "loss": 0.1705,
+      "step": 13282
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.9759971508078339,
+      "learning_rate": 1.6839977137557012e-06,
+      "loss": 0.1999,
+      "step": 13283
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.6099228343182348,
+      "learning_rate": 1.683323540557721e-06,
+      "loss": 0.196,
+      "step": 13284
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.9396624857253401,
+      "learning_rate": 1.6826494750209653e-06,
+      "loss": 0.1326,
+      "step": 13285
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.3144077502917078,
+      "learning_rate": 1.6819755171673174e-06,
+      "loss": 0.1886,
+      "step": 13286
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.1262788457806985,
+      "learning_rate": 1.6813016670186505e-06,
+      "loss": 0.1726,
+      "step": 13287
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.8124360615863917,
+      "learning_rate": 1.6806279245968399e-06,
+      "loss": 0.1628,
+      "step": 13288
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.170030171726686,
+      "learning_rate": 1.6799542899237558e-06,
+      "loss": 0.2191,
+      "step": 13289
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.1955224769015724,
+      "learning_rate": 1.679280763021266e-06,
+      "loss": 0.1749,
+      "step": 13290
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.2776775798810898,
+      "learning_rate": 1.6786073439112304e-06,
+      "loss": 0.2009,
+      "step": 13291
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.570466693500709,
+      "learning_rate": 1.6779340326155107e-06,
+      "loss": 0.2037,
+      "step": 13292
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.3138594228924216,
+      "learning_rate": 1.6772608291559639e-06,
+      "loss": 0.2033,
+      "step": 13293
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.3542403280568767,
+      "learning_rate": 1.67658773355444e-06,
+      "loss": 0.1424,
+      "step": 13294
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.5191014915285215,
+      "learning_rate": 1.6759147458327913e-06,
+      "loss": 0.1658,
+      "step": 13295
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.0784618777402408,
+      "learning_rate": 1.6752418660128593e-06,
+      "loss": 0.1856,
+      "step": 13296
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.2197233616929704,
+      "learning_rate": 1.6745690941164884e-06,
+      "loss": 0.1987,
+      "step": 13297
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.5148346083720816,
+      "learning_rate": 1.6738964301655163e-06,
+      "loss": 0.1862,
+      "step": 13298
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.1443817086546602,
+      "learning_rate": 1.6732238741817812e-06,
+      "loss": 0.1727,
+      "step": 13299
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.1842306535396625,
+      "learning_rate": 1.67255142618711e-06,
+      "loss": 0.2074,
+      "step": 13300
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.5106509525735454,
+      "learning_rate": 1.6718790862033334e-06,
+      "loss": 0.2188,
+      "step": 13301
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.5973229940919158,
+      "learning_rate": 1.6712068542522753e-06,
+      "loss": 0.1827,
+      "step": 13302
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.388976974343143,
+      "learning_rate": 1.6705347303557567e-06,
+      "loss": 0.1344,
+      "step": 13303
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.3716727227130998,
+      "learning_rate": 1.6698627145355973e-06,
+      "loss": 0.1788,
+      "step": 13304
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.438783010871333,
+      "learning_rate": 1.6691908068136075e-06,
+      "loss": 0.2159,
+      "step": 13305
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.5295053683402162,
+      "learning_rate": 1.6685190072115992e-06,
+      "loss": 0.1388,
+      "step": 13306
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.6066088167135435,
+      "learning_rate": 1.6678473157513798e-06,
+      "loss": 0.2055,
+      "step": 13307
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.3546583132916112,
+      "learning_rate": 1.6671757324547544e-06,
+      "loss": 0.1509,
+      "step": 13308
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.270321805270039,
+      "learning_rate": 1.6665042573435192e-06,
+      "loss": 0.2162,
+      "step": 13309
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.0294550280160042,
+      "learning_rate": 1.6658328904394749e-06,
+      "loss": 0.1652,
+      "step": 13310
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.142168564020628,
+      "learning_rate": 1.6651616317644104e-06,
+      "loss": 0.1517,
+      "step": 13311
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.5090409991339606,
+      "learning_rate": 1.664490481340117e-06,
+      "loss": 0.1503,
+      "step": 13312
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.2218478732744282,
+      "learning_rate": 1.6638194391883822e-06,
+      "loss": 0.1978,
+      "step": 13313
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.7242488333009676,
+      "learning_rate": 1.6631485053309848e-06,
+      "loss": 0.1739,
+      "step": 13314
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.6215815396237798,
+      "learning_rate": 1.6624776797897064e-06,
+      "loss": 0.2265,
+      "step": 13315
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.3808595152858896,
+      "learning_rate": 1.6618069625863215e-06,
+      "loss": 0.1966,
+      "step": 13316
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.2830984996139243,
+      "learning_rate": 1.6611363537426039e-06,
+      "loss": 0.1798,
+      "step": 13317
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.9604298716207591,
+      "learning_rate": 1.660465853280318e-06,
+      "loss": 0.1254,
+      "step": 13318
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.4058417787375463,
+      "learning_rate": 1.659795461221232e-06,
+      "loss": 0.1508,
+      "step": 13319
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.0317876815704048,
+      "learning_rate": 1.6591251775871053e-06,
+      "loss": 0.1655,
+      "step": 13320
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.8747514083100079,
+      "learning_rate": 1.6584550023996965e-06,
+      "loss": 0.2096,
+      "step": 13321
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.334175420125202,
+      "learning_rate": 1.6577849356807618e-06,
+      "loss": 0.2026,
+      "step": 13322
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.3241646960993838,
+      "learning_rate": 1.657114977452048e-06,
+      "loss": 0.1777,
+      "step": 13323
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.8648709466064316,
+      "learning_rate": 1.6564451277353045e-06,
+      "loss": 0.1408,
+      "step": 13324
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.7011943469660076,
+      "learning_rate": 1.6557753865522748e-06,
+      "loss": 0.235,
+      "step": 13325
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.331473533014573,
+      "learning_rate": 1.6551057539247006e-06,
+      "loss": 0.2069,
+      "step": 13326
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.616322438061659,
+      "learning_rate": 1.6544362298743161e-06,
+      "loss": 0.2329,
+      "step": 13327
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.7143768639507446,
+      "learning_rate": 1.6537668144228563e-06,
+      "loss": 0.2495,
+      "step": 13328
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.14646991545155,
+      "learning_rate": 1.6530975075920486e-06,
+      "loss": 0.1783,
+      "step": 13329
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.8444327609674873,
+      "learning_rate": 1.6524283094036202e-06,
+      "loss": 0.1626,
+      "step": 13330
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.2473126880142706,
+      "learning_rate": 1.6517592198792958e-06,
+      "loss": 0.1866,
+      "step": 13331
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.7369337624063579,
+      "learning_rate": 1.6510902390407912e-06,
+      "loss": 0.2176,
+      "step": 13332
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.1320257365116833,
+      "learning_rate": 1.6504213669098223e-06,
+      "loss": 0.1425,
+      "step": 13333
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.6835480570954118,
+      "learning_rate": 1.6497526035081029e-06,
+      "loss": 0.2288,
+      "step": 13334
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.034254066018763,
+      "learning_rate": 1.6490839488573418e-06,
+      "loss": 0.1535,
+      "step": 13335
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.1816914129770966,
+      "learning_rate": 1.6484154029792403e-06,
+      "loss": 0.1629,
+      "step": 13336
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.2564332519977097,
+      "learning_rate": 1.647746965895503e-06,
+      "loss": 0.2214,
+      "step": 13337
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.4378886345370525,
+      "learning_rate": 1.6470786376278263e-06,
+      "loss": 0.1533,
+      "step": 13338
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.5491197799718408,
+      "learning_rate": 1.6464104181979058e-06,
+      "loss": 0.1594,
+      "step": 13339
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.1626819129862518,
+      "learning_rate": 1.6457423076274327e-06,
+      "loss": 0.2028,
+      "step": 13340
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.0313302066707704,
+      "learning_rate": 1.6450743059380913e-06,
+      "loss": 0.161,
+      "step": 13341
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.3184497717489458,
+      "learning_rate": 1.6444064131515674e-06,
+      "loss": 0.233,
+      "step": 13342
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.9582135784449244,
+      "learning_rate": 1.6437386292895414e-06,
+      "loss": 0.2192,
+      "step": 13343
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.1040229475158418,
+      "learning_rate": 1.6430709543736905e-06,
+      "loss": 0.1751,
+      "step": 13344
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.156067898303116,
+      "learning_rate": 1.6424033884256857e-06,
+      "loss": 0.1311,
+      "step": 13345
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.4210996767143824,
+      "learning_rate": 1.6417359314671987e-06,
+      "loss": 0.2198,
+      "step": 13346
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.9017765836797136,
+      "learning_rate": 1.641068583519893e-06,
+      "loss": 0.1626,
+      "step": 13347
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.4651707551264002,
+      "learning_rate": 1.640401344605433e-06,
+      "loss": 0.1614,
+      "step": 13348
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.4568108706472762,
+      "learning_rate": 1.6397342147454787e-06,
+      "loss": 0.2175,
+      "step": 13349
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.5082052490171993,
+      "learning_rate": 1.639067193961682e-06,
+      "loss": 0.186,
+      "step": 13350
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.127544069634214,
+      "learning_rate": 1.6384002822756972e-06,
+      "loss": 0.2226,
+      "step": 13351
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.247853895499556,
+      "learning_rate": 1.6377334797091726e-06,
+      "loss": 0.1454,
+      "step": 13352
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.5329573659072695,
+      "learning_rate": 1.637066786283754e-06,
+      "loss": 0.1826,
+      "step": 13353
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.0807627117358891,
+      "learning_rate": 1.6364002020210795e-06,
+      "loss": 0.1967,
+      "step": 13354
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.3900218362773975,
+      "learning_rate": 1.6357337269427892e-06,
+      "loss": 0.1719,
+      "step": 13355
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.0938789019239892,
+      "learning_rate": 1.6350673610705164e-06,
+      "loss": 0.1254,
+      "step": 13356
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.5866349852261348,
+      "learning_rate": 1.6344011044258917e-06,
+      "loss": 0.1738,
+      "step": 13357
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.430398683378908,
+      "learning_rate": 1.6337349570305444e-06,
+      "loss": 0.1639,
+      "step": 13358
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.4078385877256137,
+      "learning_rate": 1.633068918906095e-06,
+      "loss": 0.2123,
+      "step": 13359
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.2117042528954018,
+      "learning_rate": 1.632402990074164e-06,
+      "loss": 0.1435,
+      "step": 13360
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.3065165302553543,
+      "learning_rate": 1.6317371705563689e-06,
+      "loss": 0.1961,
+      "step": 13361
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.1214352429880727,
+      "learning_rate": 1.6310714603743233e-06,
+      "loss": 0.1449,
+      "step": 13362
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.9943368339646929,
+      "learning_rate": 1.6304058595496342e-06,
+      "loss": 0.2047,
+      "step": 13363
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.0329022429138686,
+      "learning_rate": 1.6297403681039097e-06,
+      "loss": 0.1764,
+      "step": 13364
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.0424126864336156,
+      "learning_rate": 1.6290749860587497e-06,
+      "loss": 0.1995,
+      "step": 13365
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.9056227699584525,
+      "learning_rate": 1.6284097134357535e-06,
+      "loss": 0.1696,
+      "step": 13366
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.9856974105385112,
+      "learning_rate": 1.6277445502565192e-06,
+      "loss": 0.1389,
+      "step": 13367
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.3721194655550129,
+      "learning_rate": 1.6270794965426339e-06,
+      "loss": 0.2083,
+      "step": 13368
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.2749828896122404,
+      "learning_rate": 1.6264145523156878e-06,
+      "loss": 0.167,
+      "step": 13369
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.9506438393453456,
+      "learning_rate": 1.6257497175972654e-06,
+      "loss": 0.1442,
+      "step": 13370
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.424597250046304,
+      "learning_rate": 1.6250849924089485e-06,
+      "loss": 0.2178,
+      "step": 13371
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.2453139863908769,
+      "learning_rate": 1.6244203767723126e-06,
+      "loss": 0.2073,
+      "step": 13372
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.5223297982690966,
+      "learning_rate": 1.6237558707089319e-06,
+      "loss": 0.2533,
+      "step": 13373
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.7458065701400598,
+      "learning_rate": 1.6230914742403775e-06,
+      "loss": 0.1985,
+      "step": 13374
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.1999611808537936,
+      "learning_rate": 1.6224271873882153e-06,
+      "loss": 0.1791,
+      "step": 13375
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.304004956069806,
+      "learning_rate": 1.6217630101740106e-06,
+      "loss": 0.166,
+      "step": 13376
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.4499889804980417,
+      "learning_rate": 1.6210989426193196e-06,
+      "loss": 0.199,
+      "step": 13377
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.4216925640233185,
+      "learning_rate": 1.6204349847457e-06,
+      "loss": 0.1655,
+      "step": 13378
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.2019561719189427,
+      "learning_rate": 1.6197711365747043e-06,
+      "loss": 0.1602,
+      "step": 13379
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.2196557823883398,
+      "learning_rate": 1.6191073981278827e-06,
+      "loss": 0.1536,
+      "step": 13380
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.03174439774599,
+      "learning_rate": 1.6184437694267779e-06,
+      "loss": 0.1801,
+      "step": 13381
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.9474873391465543,
+      "learning_rate": 1.6177802504929341e-06,
+      "loss": 0.132,
+      "step": 13382
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.5035085672483952,
+      "learning_rate": 1.6171168413478865e-06,
+      "loss": 0.2053,
+      "step": 13383
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.217355208173453,
+      "learning_rate": 1.6164535420131717e-06,
+      "loss": 0.165,
+      "step": 13384
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.3009290067132955,
+      "learning_rate": 1.6157903525103225e-06,
+      "loss": 0.1474,
+      "step": 13385
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.351051986616196,
+      "learning_rate": 1.6151272728608625e-06,
+      "loss": 0.1913,
+      "step": 13386
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.4665209463683768,
+      "learning_rate": 1.6144643030863177e-06,
+      "loss": 0.2099,
+      "step": 13387
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.4916676856671036,
+      "learning_rate": 1.613801443208209e-06,
+      "loss": 0.214,
+      "step": 13388
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.392992733713075,
+      "learning_rate": 1.6131386932480535e-06,
+      "loss": 0.198,
+      "step": 13389
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.6782051109926246,
+      "learning_rate": 1.6124760532273626e-06,
+      "loss": 0.1974,
+      "step": 13390
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.5652509076792593,
+      "learning_rate": 1.6118135231676469e-06,
+      "loss": 0.1677,
+      "step": 13391
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.776756306916306,
+      "learning_rate": 1.611151103090412e-06,
+      "loss": 0.1786,
+      "step": 13392
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.1367544647059487,
+      "learning_rate": 1.6104887930171615e-06,
+      "loss": 0.1739,
+      "step": 13393
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.1856761780161522,
+      "learning_rate": 1.6098265929693957e-06,
+      "loss": 0.1458,
+      "step": 13394
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.7057341311045224,
+      "learning_rate": 1.6091645029686059e-06,
+      "loss": 0.2284,
+      "step": 13395
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.1941044568295642,
+      "learning_rate": 1.6085025230362866e-06,
+      "loss": 0.1165,
+      "step": 13396
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.489806047431603,
+      "learning_rate": 1.6078406531939261e-06,
+      "loss": 0.1936,
+      "step": 13397
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.274032401159071,
+      "learning_rate": 1.6071788934630094e-06,
+      "loss": 0.1957,
+      "step": 13398
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.5727804821839988,
+      "learning_rate": 1.6065172438650156e-06,
+      "loss": 0.1639,
+      "step": 13399
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.310095207507452,
+      "learning_rate": 1.6058557044214251e-06,
+      "loss": 0.1453,
+      "step": 13400
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.3114906642111064,
+      "learning_rate": 1.6051942751537087e-06,
+      "loss": 0.1492,
+      "step": 13401
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.0777941693776492,
+      "learning_rate": 1.6045329560833385e-06,
+      "loss": 0.198,
+      "step": 13402
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.5590637282418103,
+      "learning_rate": 1.6038717472317822e-06,
+      "loss": 0.2091,
+      "step": 13403
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.4445353686912004,
+      "learning_rate": 1.6032106486205002e-06,
+      "loss": 0.17,
+      "step": 13404
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.4470086558136805,
+      "learning_rate": 1.602549660270954e-06,
+      "loss": 0.1414,
+      "step": 13405
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.2873199938890145,
+      "learning_rate": 1.6018887822045998e-06,
+      "loss": 0.1904,
+      "step": 13406
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.4076609738482686,
+      "learning_rate": 1.601228014442891e-06,
+      "loss": 0.1436,
+      "step": 13407
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.2797444965492366,
+      "learning_rate": 1.600567357007274e-06,
+      "loss": 0.2109,
+      "step": 13408
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.4507455315042033,
+      "learning_rate": 1.599906809919195e-06,
+      "loss": 0.1561,
+      "step": 13409
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.5087714530633232,
+      "learning_rate": 1.5992463732000968e-06,
+      "loss": 0.1659,
+      "step": 13410
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.3914165387209636,
+      "learning_rate": 1.5985860468714159e-06,
+      "loss": 0.198,
+      "step": 13411
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.6269757290995213,
+      "learning_rate": 1.5979258309545903e-06,
+      "loss": 0.2237,
+      "step": 13412
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.444134448623554,
+      "learning_rate": 1.5972657254710467e-06,
+      "loss": 0.2165,
+      "step": 13413
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.2748928617594524,
+      "learning_rate": 1.5966057304422144e-06,
+      "loss": 0.1644,
+      "step": 13414
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.3237336878745234,
+      "learning_rate": 1.5959458458895172e-06,
+      "loss": 0.2237,
+      "step": 13415
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.3638668961123817,
+      "learning_rate": 1.5952860718343772e-06,
+      "loss": 0.1818,
+      "step": 13416
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.317308580754616,
+      "learning_rate": 1.594626408298207e-06,
+      "loss": 0.2131,
+      "step": 13417
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.42627472633322,
+      "learning_rate": 1.5939668553024234e-06,
+      "loss": 0.188,
+      "step": 13418
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.2449762242126665,
+      "learning_rate": 1.5933074128684333e-06,
+      "loss": 0.2176,
+      "step": 13419
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.1428128368214463,
+      "learning_rate": 1.592648081017643e-06,
+      "loss": 0.1668,
+      "step": 13420
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.3521604097446007,
+      "learning_rate": 1.5919888597714572e-06,
+      "loss": 0.186,
+      "step": 13421
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.389302047696695,
+      "learning_rate": 1.5913297491512713e-06,
+      "loss": 0.1787,
+      "step": 13422
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.0676142170371243,
+      "learning_rate": 1.5906707491784817e-06,
+      "loss": 0.1653,
+      "step": 13423
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.2052038674311651,
+      "learning_rate": 1.5900118598744802e-06,
+      "loss": 0.1114,
+      "step": 13424
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.168029872864184,
+      "learning_rate": 1.5893530812606555e-06,
+      "loss": 0.1706,
+      "step": 13425
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.3913335388136527,
+      "learning_rate": 1.5886944133583903e-06,
+      "loss": 0.1815,
+      "step": 13426
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.3062518900077325,
+      "learning_rate": 1.5880358561890657e-06,
+      "loss": 0.1917,
+      "step": 13427
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.557272224134792,
+      "learning_rate": 1.5873774097740597e-06,
+      "loss": 0.2046,
+      "step": 13428
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.6807028606930412,
+      "learning_rate": 1.5867190741347449e-06,
+      "loss": 0.1578,
+      "step": 13429
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.4707755771334254,
+      "learning_rate": 1.5860608492924933e-06,
+      "loss": 0.1983,
+      "step": 13430
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.2462860555838464,
+      "learning_rate": 1.5854027352686684e-06,
+      "loss": 0.2647,
+      "step": 13431
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.388715864040101,
+      "learning_rate": 1.5847447320846337e-06,
+      "loss": 0.2348,
+      "step": 13432
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.662990018505551,
+      "learning_rate": 1.5840868397617499e-06,
+      "loss": 0.2062,
+      "step": 13433
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.5167534746576978,
+      "learning_rate": 1.5834290583213725e-06,
+      "loss": 0.2245,
+      "step": 13434
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.417517913523517,
+      "learning_rate": 1.5827713877848505e-06,
+      "loss": 0.1981,
+      "step": 13435
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.1205491062193425,
+      "learning_rate": 1.5821138281735365e-06,
+      "loss": 0.1308,
+      "step": 13436
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.5218271229358777,
+      "learning_rate": 1.5814563795087717e-06,
+      "loss": 0.1819,
+      "step": 13437
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.1446788744714729,
+      "learning_rate": 1.580799041811898e-06,
+      "loss": 0.1763,
+      "step": 13438
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.0888725314843115,
+      "learning_rate": 1.5801418151042559e-06,
+      "loss": 0.1938,
+      "step": 13439
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.4004949320701796,
+      "learning_rate": 1.5794846994071744e-06,
+      "loss": 0.255,
+      "step": 13440
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.5341345660824364,
+      "learning_rate": 1.5788276947419872e-06,
+      "loss": 0.2022,
+      "step": 13441
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.3648097533526662,
+      "learning_rate": 1.5781708011300207e-06,
+      "loss": 0.1739,
+      "step": 13442
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.2862580802467198,
+      "learning_rate": 1.577514018592597e-06,
+      "loss": 0.1332,
+      "step": 13443
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.3922065758174247,
+      "learning_rate": 1.576857347151038e-06,
+      "loss": 0.2156,
+      "step": 13444
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.243816662923606,
+      "learning_rate": 1.5762007868266567e-06,
+      "loss": 0.2089,
+      "step": 13445
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.4622375955905011,
+      "learning_rate": 1.5755443376407664e-06,
+      "loss": 0.1799,
+      "step": 13446
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.5496238261636341,
+      "learning_rate": 1.5748879996146765e-06,
+      "loss": 0.1748,
+      "step": 13447
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.24625169715651,
+      "learning_rate": 1.5742317727696926e-06,
+      "loss": 0.2026,
+      "step": 13448
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.2344580875874218,
+      "learning_rate": 1.5735756571271144e-06,
+      "loss": 0.2048,
+      "step": 13449
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.296658994886218,
+      "learning_rate": 1.5729196527082402e-06,
+      "loss": 0.2373,
+      "step": 13450
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.1512031244153458,
+      "learning_rate": 1.5722637595343655e-06,
+      "loss": 0.1673,
+      "step": 13451
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.6850035665895993,
+      "learning_rate": 1.5716079776267812e-06,
+      "loss": 0.1956,
+      "step": 13452
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.2434910376442065,
+      "learning_rate": 1.5709523070067718e-06,
+      "loss": 0.1527,
+      "step": 13453
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.5217956841497102,
+      "learning_rate": 1.5702967476956243e-06,
+      "loss": 0.2037,
+      "step": 13454
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.6323365562764638,
+      "learning_rate": 1.5696412997146148e-06,
+      "loss": 0.1803,
+      "step": 13455
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.1292384064883445,
+      "learning_rate": 1.5689859630850213e-06,
+      "loss": 0.1854,
+      "step": 13456
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.1492050972703918,
+      "learning_rate": 1.5683307378281182e-06,
+      "loss": 0.1995,
+      "step": 13457
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.4557980798049632,
+      "learning_rate": 1.5676756239651708e-06,
+      "loss": 0.1756,
+      "step": 13458
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.0443846489136983,
+      "learning_rate": 1.5670206215174466e-06,
+      "loss": 0.1591,
+      "step": 13459
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.1981396930512014,
+      "learning_rate": 1.566365730506207e-06,
+      "loss": 0.1288,
+      "step": 13460
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.2417394651790157,
+      "learning_rate": 1.5657109509527103e-06,
+      "loss": 0.214,
+      "step": 13461
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.4773733647453076,
+      "learning_rate": 1.565056282878213e-06,
+      "loss": 0.1599,
+      "step": 13462
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.2832056623227655,
+      "learning_rate": 1.5644017263039617e-06,
+      "loss": 0.208,
+      "step": 13463
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.484580648130561,
+      "learning_rate": 1.563747281251206e-06,
+      "loss": 0.1844,
+      "step": 13464
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.3962682516745855,
+      "learning_rate": 1.5630929477411899e-06,
+      "loss": 0.136,
+      "step": 13465
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.146053714221947,
+      "learning_rate": 1.5624387257951552e-06,
+      "loss": 0.1904,
+      "step": 13466
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.6156345180843523,
+      "learning_rate": 1.5617846154343335e-06,
+      "loss": 0.1977,
+      "step": 13467
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.3274562442253064,
+      "learning_rate": 1.561130616679961e-06,
+      "loss": 0.1659,
+      "step": 13468
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.2818826125489449,
+      "learning_rate": 1.5604767295532657e-06,
+      "loss": 0.1818,
+      "step": 13469
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.384550618642257,
+      "learning_rate": 1.5598229540754756e-06,
+      "loss": 0.2314,
+      "step": 13470
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.0969192178757943,
+      "learning_rate": 1.559169290267809e-06,
+      "loss": 0.1322,
+      "step": 13471
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.656288846028219,
+      "learning_rate": 1.5585157381514875e-06,
+      "loss": 0.2148,
+      "step": 13472
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.3688017987766186,
+      "learning_rate": 1.5578622977477226e-06,
+      "loss": 0.1717,
+      "step": 13473
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.5793335767166328,
+      "learning_rate": 1.5572089690777264e-06,
+      "loss": 0.1906,
+      "step": 13474
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.192713161386549,
+      "learning_rate": 1.556555752162709e-06,
+      "loss": 0.2201,
+      "step": 13475
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.8978377992327824,
+      "learning_rate": 1.5559026470238702e-06,
+      "loss": 0.2165,
+      "step": 13476
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.9707406495361376,
+      "learning_rate": 1.5552496536824118e-06,
+      "loss": 0.1381,
+      "step": 13477
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.0360897767156596,
+      "learning_rate": 1.5545967721595307e-06,
+      "loss": 0.1498,
+      "step": 13478
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.319391211333761,
+      "learning_rate": 1.5539440024764196e-06,
+      "loss": 0.1947,
+      "step": 13479
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.3536294996570957,
+      "learning_rate": 1.5532913446542692e-06,
+      "loss": 0.2158,
+      "step": 13480
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.1590816967367734,
+      "learning_rate": 1.5526387987142621e-06,
+      "loss": 0.1455,
+      "step": 13481
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.6203245728097226,
+      "learning_rate": 1.5519863646775823e-06,
+      "loss": 0.2183,
+      "step": 13482
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.4163314677557834,
+      "learning_rate": 1.5513340425654077e-06,
+      "loss": 0.1616,
+      "step": 13483
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.43119487874945,
+      "learning_rate": 1.5506818323989148e-06,
+      "loss": 0.1892,
+      "step": 13484
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.3994757724527251,
+      "learning_rate": 1.550029734199272e-06,
+      "loss": 0.199,
+      "step": 13485
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.690107816291556,
+      "learning_rate": 1.549377747987647e-06,
+      "loss": 0.1834,
+      "step": 13486
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.4013705675569503,
+      "learning_rate": 1.548725873785205e-06,
+      "loss": 0.1862,
+      "step": 13487
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.3779269190705712,
+      "learning_rate": 1.5480741116131077e-06,
+      "loss": 0.1977,
+      "step": 13488
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.2659982790136768,
+      "learning_rate": 1.5474224614925076e-06,
+      "loss": 0.2293,
+      "step": 13489
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.1544250959256919,
+      "learning_rate": 1.5467709234445617e-06,
+      "loss": 0.1724,
+      "step": 13490
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.4158788391343116,
+      "learning_rate": 1.546119497490416e-06,
+      "loss": 0.1927,
+      "step": 13491
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.9765664653697569,
+      "learning_rate": 1.5454681836512169e-06,
+      "loss": 0.1471,
+      "step": 13492
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.261984611113241,
+      "learning_rate": 1.544816981948109e-06,
+      "loss": 0.1478,
+      "step": 13493
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.622211245776554,
+      "learning_rate": 1.5441658924022273e-06,
+      "loss": 0.1733,
+      "step": 13494
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.6663131940051923,
+      "learning_rate": 1.5435149150347073e-06,
+      "loss": 0.1646,
+      "step": 13495
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.8094027956868357,
+      "learning_rate": 1.5428640498666814e-06,
+      "loss": 0.1753,
+      "step": 13496
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.1485089746927113,
+      "learning_rate": 1.5422132969192766e-06,
+      "loss": 0.1456,
+      "step": 13497
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.5557097663559476,
+      "learning_rate": 1.5415626562136177e-06,
+      "loss": 0.1819,
+      "step": 13498
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.3344924037573513,
+      "learning_rate": 1.5409121277708222e-06,
+      "loss": 0.1878,
+      "step": 13499
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.5207730541519877,
+      "learning_rate": 1.5402617116120084e-06,
+      "loss": 0.2248,
+      "step": 13500
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.902936678691635,
+      "learning_rate": 1.5396114077582885e-06,
+      "loss": 0.1398,
+      "step": 13501
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.1827893403531058,
+      "learning_rate": 1.5389612162307738e-06,
+      "loss": 0.1675,
+      "step": 13502
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.2019331899731003,
+      "learning_rate": 1.538311137050567e-06,
+      "loss": 0.1782,
+      "step": 13503
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.5177563693950165,
+      "learning_rate": 1.537661170238771e-06,
+      "loss": 0.1606,
+      "step": 13504
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.9308007230036772,
+      "learning_rate": 1.537011315816485e-06,
+      "loss": 0.2139,
+      "step": 13505
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.4398636148524064,
+      "learning_rate": 1.5363615738048044e-06,
+      "loss": 0.1565,
+      "step": 13506
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.6266739366452296,
+      "learning_rate": 1.535711944224817e-06,
+      "loss": 0.2324,
+      "step": 13507
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.1985517939709895,
+      "learning_rate": 1.5350624270976132e-06,
+      "loss": 0.184,
+      "step": 13508
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.1141597861818404,
+      "learning_rate": 1.5344130224442744e-06,
+      "loss": 0.1251,
+      "step": 13509
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.202341806337051,
+      "learning_rate": 1.533763730285882e-06,
+      "loss": 0.156,
+      "step": 13510
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.4444246916102983,
+      "learning_rate": 1.5331145506435136e-06,
+      "loss": 0.2331,
+      "step": 13511
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.2236612777869704,
+      "learning_rate": 1.5324654835382386e-06,
+      "loss": 0.1379,
+      "step": 13512
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.3531744418240739,
+      "learning_rate": 1.531816528991128e-06,
+      "loss": 0.191,
+      "step": 13513
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.1614740638637011,
+      "learning_rate": 1.5311676870232473e-06,
+      "loss": 0.1842,
+      "step": 13514
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.8007980643788615,
+      "learning_rate": 1.5305189576556584e-06,
+      "loss": 0.141,
+      "step": 13515
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.1376207875740787,
+      "learning_rate": 1.5298703409094207e-06,
+      "loss": 0.1387,
+      "step": 13516
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.1861619375099413,
+      "learning_rate": 1.5292218368055856e-06,
+      "loss": 0.1685,
+      "step": 13517
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.0942628918135207,
+      "learning_rate": 1.5285734453652057e-06,
+      "loss": 0.16,
+      "step": 13518
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.1502540002076358,
+      "learning_rate": 1.5279251666093276e-06,
+      "loss": 0.1683,
+      "step": 13519
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.404612058835028,
+      "learning_rate": 1.527277000558997e-06,
+      "loss": 0.2079,
+      "step": 13520
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.1015148964193664,
+      "learning_rate": 1.5266289472352503e-06,
+      "loss": 0.1233,
+      "step": 13521
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.2427965026854142,
+      "learning_rate": 1.5259810066591253e-06,
+      "loss": 0.1438,
+      "step": 13522
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.399292780218607,
+      "learning_rate": 1.5253331788516545e-06,
+      "loss": 0.2091,
+      "step": 13523
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.0948603375630401,
+      "learning_rate": 1.5246854638338682e-06,
+      "loss": 0.1293,
+      "step": 13524
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.6252922703093338,
+      "learning_rate": 1.5240378616267887e-06,
+      "loss": 0.2102,
+      "step": 13525
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.3435053879967342,
+      "learning_rate": 1.523390372251441e-06,
+      "loss": 0.1872,
+      "step": 13526
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.5757744074073847,
+      "learning_rate": 1.5227429957288387e-06,
+      "loss": 0.1702,
+      "step": 13527
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.0738855642843241,
+      "learning_rate": 1.5220957320799989e-06,
+      "loss": 0.1744,
+      "step": 13528
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.1835067830380903,
+      "learning_rate": 1.5214485813259333e-06,
+      "loss": 0.1629,
+      "step": 13529
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.9889632574494165,
+      "learning_rate": 1.520801543487645e-06,
+      "loss": 0.1252,
+      "step": 13530
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.1522524102428502,
+      "learning_rate": 1.5201546185861394e-06,
+      "loss": 0.1622,
+      "step": 13531
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.521552327980549,
+      "learning_rate": 1.5195078066424162e-06,
+      "loss": 0.2398,
+      "step": 13532
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.340692068080723,
+      "learning_rate": 1.5188611076774705e-06,
+      "loss": 0.161,
+      "step": 13533
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.7675696104065899,
+      "learning_rate": 1.5182145217122973e-06,
+      "loss": 0.168,
+      "step": 13534
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.181003108537719,
+      "learning_rate": 1.5175680487678807e-06,
+      "loss": 0.1441,
+      "step": 13535
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.1552598231747635,
+      "learning_rate": 1.5169216888652083e-06,
+      "loss": 0.2057,
+      "step": 13536
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.4056799872470136,
+      "learning_rate": 1.51627544202526e-06,
+      "loss": 0.156,
+      "step": 13537
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.1619979708831538,
+      "learning_rate": 1.5156293082690166e-06,
+      "loss": 0.1456,
+      "step": 13538
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.5707577885372872,
+      "learning_rate": 1.5149832876174476e-06,
+      "loss": 0.2307,
+      "step": 13539
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.2434755760588967,
+      "learning_rate": 1.514337380091525e-06,
+      "loss": 0.1401,
+      "step": 13540
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.1500788850783734,
+      "learning_rate": 1.5136915857122159e-06,
+      "loss": 0.1415,
+      "step": 13541
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.615730599193353,
+      "learning_rate": 1.5130459045004836e-06,
+      "loss": 0.2209,
+      "step": 13542
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.300462941582214,
+      "learning_rate": 1.512400336477285e-06,
+      "loss": 0.1469,
+      "step": 13543
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.486142651969913,
+      "learning_rate": 1.5117548816635792e-06,
+      "loss": 0.2017,
+      "step": 13544
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.3050640156818725,
+      "learning_rate": 1.5111095400803132e-06,
+      "loss": 0.1684,
+      "step": 13545
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.4422947644302329,
+      "learning_rate": 1.5104643117484385e-06,
+      "loss": 0.1786,
+      "step": 13546
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.6732519152837084,
+      "learning_rate": 1.5098191966889003e-06,
+      "loss": 0.2512,
+      "step": 13547
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.2469881698332883,
+      "learning_rate": 1.5091741949226362e-06,
+      "loss": 0.1752,
+      "step": 13548
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.5112963290800165,
+      "learning_rate": 1.5085293064705853e-06,
+      "loss": 0.1811,
+      "step": 13549
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.658497944785723,
+      "learning_rate": 1.507884531353681e-06,
+      "loss": 0.1795,
+      "step": 13550
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.1282426903483869,
+      "learning_rate": 1.5072398695928526e-06,
+      "loss": 0.1896,
+      "step": 13551
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.0294678236644517,
+      "learning_rate": 1.5065953212090284e-06,
+      "loss": 0.1536,
+      "step": 13552
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.1561338520995923,
+      "learning_rate": 1.5059508862231271e-06,
+      "loss": 0.1508,
+      "step": 13553
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.3378504559132158,
+      "learning_rate": 1.5053065646560695e-06,
+      "loss": 0.1834,
+      "step": 13554
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.4165921057218864,
+      "learning_rate": 1.5046623565287698e-06,
+      "loss": 0.1967,
+      "step": 13555
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.5918174625305745,
+      "learning_rate": 1.5040182618621423e-06,
+      "loss": 0.217,
+      "step": 13556
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.503732069258201,
+      "learning_rate": 1.5033742806770901e-06,
+      "loss": 0.1734,
+      "step": 13557
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.12727510853968,
+      "learning_rate": 1.5027304129945203e-06,
+      "loss": 0.187,
+      "step": 13558
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.0215682110199413,
+      "learning_rate": 1.5020866588353334e-06,
+      "loss": 0.2223,
+      "step": 13559
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.3039176177025864,
+      "learning_rate": 1.501443018220423e-06,
+      "loss": 0.1575,
+      "step": 13560
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.5827207640398704,
+      "learning_rate": 1.5007994911706846e-06,
+      "loss": 0.196,
+      "step": 13561
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.4028322649383616,
+      "learning_rate": 1.5001560777070085e-06,
+      "loss": 0.1655,
+      "step": 13562
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.443495340098955,
+      "learning_rate": 1.4995127778502772e-06,
+      "loss": 0.2034,
+      "step": 13563
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.213299809723511,
+      "learning_rate": 1.4988695916213735e-06,
+      "loss": 0.1711,
+      "step": 13564
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.0792335118882652,
+      "learning_rate": 1.4982265190411782e-06,
+      "loss": 0.1511,
+      "step": 13565
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.5775872390718806,
+      "learning_rate": 1.4975835601305622e-06,
+      "loss": 0.1261,
+      "step": 13566
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.2751663735241086,
+      "learning_rate": 1.4969407149103977e-06,
+      "loss": 0.1749,
+      "step": 13567
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.3974255917902192,
+      "learning_rate": 1.4962979834015518e-06,
+      "loss": 0.1573,
+      "step": 13568
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.6530795792395376,
+      "learning_rate": 1.4956553656248885e-06,
+      "loss": 0.2303,
+      "step": 13569
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.7509352364798654,
+      "learning_rate": 1.4950128616012687e-06,
+      "loss": 0.2237,
+      "step": 13570
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.3747664545863885,
+      "learning_rate": 1.494370471351545e-06,
+      "loss": 0.1832,
+      "step": 13571
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.121431934375619,
+      "learning_rate": 1.493728194896572e-06,
+      "loss": 0.244,
+      "step": 13572
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.2798292377307203,
+      "learning_rate": 1.4930860322571976e-06,
+      "loss": 0.1705,
+      "step": 13573
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.3000113611687711,
+      "learning_rate": 1.4924439834542698e-06,
+      "loss": 0.1788,
+      "step": 13574
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.7338102039673846,
+      "learning_rate": 1.491802048508625e-06,
+      "loss": 0.2011,
+      "step": 13575
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.33217934375318,
+      "learning_rate": 1.491160227441103e-06,
+      "loss": 0.1748,
+      "step": 13576
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.5248357811434317,
+      "learning_rate": 1.4905185202725392e-06,
+      "loss": 0.198,
+      "step": 13577
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.2611437393714715,
+      "learning_rate": 1.4898769270237611e-06,
+      "loss": 0.1651,
+      "step": 13578
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.37143172475679,
+      "learning_rate": 1.489235447715598e-06,
+      "loss": 0.2008,
+      "step": 13579
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.1854106001483617,
+      "learning_rate": 1.4885940823688695e-06,
+      "loss": 0.1956,
+      "step": 13580
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.5596837508306152,
+      "learning_rate": 1.4879528310043962e-06,
+      "loss": 0.1831,
+      "step": 13581
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.200995629258981,
+      "learning_rate": 1.4873116936429938e-06,
+      "loss": 0.2239,
+      "step": 13582
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.602379599755599,
+      "learning_rate": 1.486670670305475e-06,
+      "loss": 0.2297,
+      "step": 13583
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.1336080980082999,
+      "learning_rate": 1.486029761012645e-06,
+      "loss": 0.1846,
+      "step": 13584
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.8814421373272506,
+      "learning_rate": 1.48538896578531e-06,
+      "loss": 0.1429,
+      "step": 13585
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.0510383986178737,
+      "learning_rate": 1.4847482846442695e-06,
+      "loss": 0.171,
+      "step": 13586
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.4397663153588194,
+      "learning_rate": 1.4841077176103218e-06,
+      "loss": 0.204,
+      "step": 13587
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.7827525577615082,
+      "learning_rate": 1.4834672647042608e-06,
+      "loss": 0.2056,
+      "step": 13588
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.198402440703446,
+      "learning_rate": 1.4828269259468725e-06,
+      "loss": 0.1773,
+      "step": 13589
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.0049943479496353,
+      "learning_rate": 1.4821867013589454e-06,
+      "loss": 0.2002,
+      "step": 13590
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.1115072203267373,
+      "learning_rate": 1.4815465909612604e-06,
+      "loss": 0.141,
+      "step": 13591
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.3149805470394818,
+      "learning_rate": 1.4809065947745977e-06,
+      "loss": 0.186,
+      "step": 13592
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.1507868999075763,
+      "learning_rate": 1.4802667128197295e-06,
+      "loss": 0.1933,
+      "step": 13593
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.5019572819725422,
+      "learning_rate": 1.4796269451174278e-06,
+      "loss": 0.2143,
+      "step": 13594
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.8523196310132672,
+      "learning_rate": 1.4789872916884613e-06,
+      "loss": 0.2243,
+      "step": 13595
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.1796229973268015,
+      "learning_rate": 1.47834775255359e-06,
+      "loss": 0.19,
+      "step": 13596
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.711967898282331,
+      "learning_rate": 1.4777083277335774e-06,
+      "loss": 0.1705,
+      "step": 13597
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.2753195107511992,
+      "learning_rate": 1.477069017249177e-06,
+      "loss": 0.1687,
+      "step": 13598
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.2251253516425655,
+      "learning_rate": 1.4764298211211414e-06,
+      "loss": 0.1422,
+      "step": 13599
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.7279846022279153,
+      "learning_rate": 1.4757907393702208e-06,
+      "loss": 0.2029,
+      "step": 13600
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.3407427215908532,
+      "learning_rate": 1.4751517720171605e-06,
+      "loss": 0.1654,
+      "step": 13601
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.7200880737543691,
+      "learning_rate": 1.474512919082699e-06,
+      "loss": 0.1715,
+      "step": 13602
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.1302221344253143,
+      "learning_rate": 1.4738741805875757e-06,
+      "loss": 0.1862,
+      "step": 13603
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.289035409584909,
+      "learning_rate": 1.4732355565525236e-06,
+      "loss": 0.1872,
+      "step": 13604
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.4787908673302506,
+      "learning_rate": 1.4725970469982736e-06,
+      "loss": 0.1981,
+      "step": 13605
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.1932502484412246,
+      "learning_rate": 1.4719586519455536e-06,
+      "loss": 0.1717,
+      "step": 13606
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.3089296507659014,
+      "learning_rate": 1.471320371415083e-06,
+      "loss": 0.2116,
+      "step": 13607
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.8932556068725949,
+      "learning_rate": 1.4706822054275821e-06,
+      "loss": 0.1818,
+      "step": 13608
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.5173530706308513,
+      "learning_rate": 1.4700441540037663e-06,
+      "loss": 0.1685,
+      "step": 13609
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.3793185079774666,
+      "learning_rate": 1.4694062171643492e-06,
+      "loss": 0.2263,
+      "step": 13610
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.3374680680179838,
+      "learning_rate": 1.468768394930034e-06,
+      "loss": 0.2272,
+      "step": 13611
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.4589828787391272,
+      "learning_rate": 1.4681306873215284e-06,
+      "loss": 0.1609,
+      "step": 13612
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.1370060056342843,
+      "learning_rate": 1.4674930943595323e-06,
+      "loss": 0.1629,
+      "step": 13613
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.1791318222593843,
+      "learning_rate": 1.4668556160647407e-06,
+      "loss": 0.1434,
+      "step": 13614
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.985432302713986,
+      "learning_rate": 1.4662182524578488e-06,
+      "loss": 0.1527,
+      "step": 13615
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.599156405226482,
+      "learning_rate": 1.465581003559543e-06,
+      "loss": 0.2071,
+      "step": 13616
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.3368542966447805,
+      "learning_rate": 1.4649438693905105e-06,
+      "loss": 0.1403,
+      "step": 13617
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.1401627204756375,
+      "learning_rate": 1.4643068499714325e-06,
+      "loss": 0.1777,
+      "step": 13618
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.0744409816790492,
+      "learning_rate": 1.4636699453229891e-06,
+      "loss": 0.1758,
+      "step": 13619
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.1415233079441378,
+      "learning_rate": 1.463033155465851e-06,
+      "loss": 0.2044,
+      "step": 13620
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.392861049817684,
+      "learning_rate": 1.4623964804206909e-06,
+      "loss": 0.2231,
+      "step": 13621
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.3473717624203558,
+      "learning_rate": 1.461759920208175e-06,
+      "loss": 0.2263,
+      "step": 13622
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.505843324489131,
+      "learning_rate": 1.4611234748489666e-06,
+      "loss": 0.1949,
+      "step": 13623
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.6196810162234339,
+      "learning_rate": 1.460487144363727e-06,
+      "loss": 0.1719,
+      "step": 13624
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.8131523027307224,
+      "learning_rate": 1.4598509287731082e-06,
+      "loss": 0.1705,
+      "step": 13625
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.244664601367481,
+      "learning_rate": 1.4592148280977647e-06,
+      "loss": 0.2025,
+      "step": 13626
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.2936428508293618,
+      "learning_rate": 1.4585788423583437e-06,
+      "loss": 0.1628,
+      "step": 13627
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.2134444343787265,
+      "learning_rate": 1.457942971575491e-06,
+      "loss": 0.1919,
+      "step": 13628
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.9983476130930065,
+      "learning_rate": 1.4573072157698454e-06,
+      "loss": 0.1933,
+      "step": 13629
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.297984378353697,
+      "learning_rate": 1.4566715749620441e-06,
+      "loss": 0.1727,
+      "step": 13630
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.5897207517166436,
+      "learning_rate": 1.4560360491727233e-06,
+      "loss": 0.223,
+      "step": 13631
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.581488292440304,
+      "learning_rate": 1.4554006384225084e-06,
+      "loss": 0.2282,
+      "step": 13632
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.5383179407892624,
+      "learning_rate": 1.4547653427320284e-06,
+      "loss": 0.1902,
+      "step": 13633
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.2267251848346743,
+      "learning_rate": 1.4541301621219024e-06,
+      "loss": 0.1756,
+      "step": 13634
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.9718306151064591,
+      "learning_rate": 1.453495096612751e-06,
+      "loss": 0.1121,
+      "step": 13635
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.9232762666355712,
+      "learning_rate": 1.4528601462251873e-06,
+      "loss": 0.1325,
+      "step": 13636
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.25523471870496,
+      "learning_rate": 1.452225310979825e-06,
+      "loss": 0.1974,
+      "step": 13637
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.407671305512088,
+      "learning_rate": 1.4515905908972677e-06,
+      "loss": 0.1822,
+      "step": 13638
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.5824830764013231,
+      "learning_rate": 1.4509559859981204e-06,
+      "loss": 0.1668,
+      "step": 13639
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.1685626137292655,
+      "learning_rate": 1.4503214963029828e-06,
+      "loss": 0.1338,
+      "step": 13640
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.2426933991807736,
+      "learning_rate": 1.4496871218324498e-06,
+      "loss": 0.1274,
+      "step": 13641
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.4668815121061205,
+      "learning_rate": 1.4490528626071165e-06,
+      "loss": 0.1933,
+      "step": 13642
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.5162360444177503,
+      "learning_rate": 1.4484187186475678e-06,
+      "loss": 0.1858,
+      "step": 13643
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.3315620434985647,
+      "learning_rate": 1.4477846899743896e-06,
+      "loss": 0.1966,
+      "step": 13644
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.6720903739175765,
+      "learning_rate": 1.4471507766081634e-06,
+      "loss": 0.1821,
+      "step": 13645
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.2120373622439695,
+      "learning_rate": 1.4465169785694672e-06,
+      "loss": 0.1869,
+      "step": 13646
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.7212691366621609,
+      "learning_rate": 1.4458832958788715e-06,
+      "loss": 0.1804,
+      "step": 13647
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.1587320442001885,
+      "learning_rate": 1.4452497285569484e-06,
+      "loss": 0.1671,
+      "step": 13648
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.3514935690881709,
+      "learning_rate": 1.4446162766242644e-06,
+      "loss": 0.2092,
+      "step": 13649
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.811728831366372,
+      "learning_rate": 1.4439829401013788e-06,
+      "loss": 0.1969,
+      "step": 13650
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.0631028789520887,
+      "learning_rate": 1.443349719008854e-06,
+      "loss": 0.1115,
+      "step": 13651
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.0485493024745964,
+      "learning_rate": 1.4427166133672404e-06,
+      "loss": 0.1621,
+      "step": 13652
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.7576349804568216,
+      "learning_rate": 1.4420836231970914e-06,
+      "loss": 0.2014,
+      "step": 13653
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.610019568648684,
+      "learning_rate": 1.4414507485189537e-06,
+      "loss": 0.2119,
+      "step": 13654
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.154865434691017,
+      "learning_rate": 1.440817989353373e-06,
+      "loss": 0.1997,
+      "step": 13655
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.4271631102527476,
+      "learning_rate": 1.4401853457208853e-06,
+      "loss": 0.1608,
+      "step": 13656
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.3780216535107834,
+      "learning_rate": 1.439552817642028e-06,
+      "loss": 0.2082,
+      "step": 13657
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.6372485179747338,
+      "learning_rate": 1.438920405137334e-06,
+      "loss": 0.2038,
+      "step": 13658
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.5652657635413967,
+      "learning_rate": 1.4382881082273315e-06,
+      "loss": 0.2256,
+      "step": 13659
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.4643103683868424,
+      "learning_rate": 1.4376559269325462e-06,
+      "loss": 0.1921,
+      "step": 13660
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.3602080204383387,
+      "learning_rate": 1.4370238612734972e-06,
+      "loss": 0.2406,
+      "step": 13661
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.5053826998876636,
+      "learning_rate": 1.4363919112707019e-06,
+      "loss": 0.1625,
+      "step": 13662
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.5742577815364065,
+      "learning_rate": 1.4357600769446744e-06,
+      "loss": 0.202,
+      "step": 13663
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.7424959157789137,
+      "learning_rate": 1.4351283583159265e-06,
+      "loss": 0.2147,
+      "step": 13664
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.4151056878179191,
+      "learning_rate": 1.4344967554049599e-06,
+      "loss": 0.1998,
+      "step": 13665
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.3073412851274135,
+      "learning_rate": 1.433865268232279e-06,
+      "loss": 0.1592,
+      "step": 13666
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.2541058224063726,
+      "learning_rate": 1.433233896818384e-06,
+      "loss": 0.151,
+      "step": 13667
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.5030511381912781,
+      "learning_rate": 1.4326026411837663e-06,
+      "loss": 0.2553,
+      "step": 13668
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.3821965047088893,
+      "learning_rate": 1.4319715013489194e-06,
+      "loss": 0.1275,
+      "step": 13669
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.3280248772523704,
+      "learning_rate": 1.4313404773343277e-06,
+      "loss": 0.1557,
+      "step": 13670
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.2459096055009438,
+      "learning_rate": 1.4307095691604767e-06,
+      "loss": 0.1925,
+      "step": 13671
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.2380392578202648,
+      "learning_rate": 1.4300787768478452e-06,
+      "loss": 0.2008,
+      "step": 13672
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.6460208806084315,
+      "learning_rate": 1.4294481004169113e-06,
+      "loss": 0.1851,
+      "step": 13673
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.591585653134023,
+      "learning_rate": 1.4288175398881432e-06,
+      "loss": 0.22,
+      "step": 13674
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.6262672599767916,
+      "learning_rate": 1.428187095282012e-06,
+      "loss": 0.1596,
+      "step": 13675
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.1398243984830894,
+      "learning_rate": 1.427556766618981e-06,
+      "loss": 0.1445,
+      "step": 13676
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.2714349294678777,
+      "learning_rate": 1.426926553919512e-06,
+      "loss": 0.1607,
+      "step": 13677
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.13013425780577,
+      "learning_rate": 1.4262964572040633e-06,
+      "loss": 0.1664,
+      "step": 13678
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.1357914462112706,
+      "learning_rate": 1.4256664764930849e-06,
+      "loss": 0.1715,
+      "step": 13679
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.7754814200271094,
+      "learning_rate": 1.4250366118070285e-06,
+      "loss": 0.1891,
+      "step": 13680
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.1255448862930557,
+      "learning_rate": 1.4244068631663394e-06,
+      "loss": 0.1805,
+      "step": 13681
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.2100401907050877,
+      "learning_rate": 1.4237772305914594e-06,
+      "loss": 0.134,
+      "step": 13682
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.3431806578005496,
+      "learning_rate": 1.4231477141028282e-06,
+      "loss": 0.214,
+      "step": 13683
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.5316476062400546,
+      "learning_rate": 1.4225183137208775e-06,
+      "loss": 0.2353,
+      "step": 13684
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.2289720606173304,
+      "learning_rate": 1.4218890294660416e-06,
+      "loss": 0.1686,
+      "step": 13685
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.3040764656982806,
+      "learning_rate": 1.4212598613587437e-06,
+      "loss": 0.2022,
+      "step": 13686
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.1119216145430346,
+      "learning_rate": 1.42063080941941e-06,
+      "loss": 0.1566,
+      "step": 13687
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.4021008501640195,
+      "learning_rate": 1.4200018736684567e-06,
+      "loss": 0.1275,
+      "step": 13688
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.6769804676171787,
+      "learning_rate": 1.4193730541263017e-06,
+      "loss": 0.1992,
+      "step": 13689
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.5485060983623358,
+      "learning_rate": 1.4187443508133564e-06,
+      "loss": 0.1994,
+      "step": 13690
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.0903199994268558,
+      "learning_rate": 1.4181157637500288e-06,
+      "loss": 0.157,
+      "step": 13691
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.2681638484116278,
+      "learning_rate": 1.4174872929567246e-06,
+      "loss": 0.1162,
+      "step": 13692
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.13152041246091,
+      "learning_rate": 1.4168589384538412e-06,
+      "loss": 0.1953,
+      "step": 13693
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.2868274240976711,
+      "learning_rate": 1.4162307002617775e-06,
+      "loss": 0.1783,
+      "step": 13694
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.1239524076474696,
+      "learning_rate": 1.4156025784009258e-06,
+      "loss": 0.1741,
+      "step": 13695
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.4107154044057193,
+      "learning_rate": 1.4149745728916776e-06,
+      "loss": 0.1667,
+      "step": 13696
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.1404390575444083,
+      "learning_rate": 1.414346683754414e-06,
+      "loss": 0.1659,
+      "step": 13697
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.5203871633185349,
+      "learning_rate": 1.4137189110095196e-06,
+      "loss": 0.2255,
+      "step": 13698
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.1959050243295348,
+      "learning_rate": 1.413091254677371e-06,
+      "loss": 0.1662,
+      "step": 13699
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.4923539842732552,
+      "learning_rate": 1.4124637147783431e-06,
+      "loss": 0.2086,
+      "step": 13700
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.5122634734280185,
+      "learning_rate": 1.4118362913328082e-06,
+      "loss": 0.1935,
+      "step": 13701
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.2075605990918743,
+      "learning_rate": 1.4112089843611282e-06,
+      "loss": 0.1897,
+      "step": 13702
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.2902019608687911,
+      "learning_rate": 1.4105817938836703e-06,
+      "loss": 0.1587,
+      "step": 13703
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.9607676456720942,
+      "learning_rate": 1.4099547199207896e-06,
+      "loss": 0.13,
+      "step": 13704
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.7245963453619313,
+      "learning_rate": 1.4093277624928454e-06,
+      "loss": 0.1666,
+      "step": 13705
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.4793855962345608,
+      "learning_rate": 1.4087009216201846e-06,
+      "loss": 0.1691,
+      "step": 13706
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.552411670859569,
+      "learning_rate": 1.408074197323157e-06,
+      "loss": 0.2092,
+      "step": 13707
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.2807827917613037,
+      "learning_rate": 1.4074475896221067e-06,
+      "loss": 0.1852,
+      "step": 13708
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.2507100055807223,
+      "learning_rate": 1.406821098537373e-06,
+      "loss": 0.184,
+      "step": 13709
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.4605714778047179,
+      "learning_rate": 1.4061947240892947e-06,
+      "loss": 0.1998,
+      "step": 13710
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.2512486926413993,
+      "learning_rate": 1.405568466298201e-06,
+      "loss": 0.1671,
+      "step": 13711
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.3435576169211614,
+      "learning_rate": 1.4049423251844213e-06,
+      "loss": 0.2138,
+      "step": 13712
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.4146608488261825,
+      "learning_rate": 1.4043163007682814e-06,
+      "loss": 0.1607,
+      "step": 13713
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.7343862722219558,
+      "learning_rate": 1.4036903930701034e-06,
+      "loss": 0.2291,
+      "step": 13714
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.09032910138936,
+      "learning_rate": 1.4030646021102018e-06,
+      "loss": 0.1952,
+      "step": 13715
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.3082235296554043,
+      "learning_rate": 1.4024389279088913e-06,
+      "loss": 0.236,
+      "step": 13716
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.4867958079948564,
+      "learning_rate": 1.4018133704864822e-06,
+      "loss": 0.227,
+      "step": 13717
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.5559783493551556,
+      "learning_rate": 1.4011879298632803e-06,
+      "loss": 0.1953,
+      "step": 13718
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.4167122094346374,
+      "learning_rate": 1.4005626060595888e-06,
+      "loss": 0.1677,
+      "step": 13719
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.3422387018973396,
+      "learning_rate": 1.3999373990957039e-06,
+      "loss": 0.1954,
+      "step": 13720
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.3082536169942858,
+      "learning_rate": 1.399312308991922e-06,
+      "loss": 0.1378,
+      "step": 13721
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.4124558631786623,
+      "learning_rate": 1.398687335768532e-06,
+      "loss": 0.1932,
+      "step": 13722
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.5509148018932537,
+      "learning_rate": 1.3980624794458237e-06,
+      "loss": 0.2226,
+      "step": 13723
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.5980663426121895,
+      "learning_rate": 1.397437740044077e-06,
+      "loss": 0.2042,
+      "step": 13724
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.5270534346786095,
+      "learning_rate": 1.3968131175835726e-06,
+      "loss": 0.1554,
+      "step": 13725
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.6181698823776978,
+      "learning_rate": 1.3961886120845869e-06,
+      "loss": 0.1925,
+      "step": 13726
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.271917808360604,
+      "learning_rate": 1.3955642235673905e-06,
+      "loss": 0.1212,
+      "step": 13727
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.4538001983681015,
+      "learning_rate": 1.3949399520522544e-06,
+      "loss": 0.1555,
+      "step": 13728
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.5729711833141935,
+      "learning_rate": 1.3943157975594385e-06,
+      "loss": 0.2158,
+      "step": 13729
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.0993339854768434,
+      "learning_rate": 1.3936917601092053e-06,
+      "loss": 0.1766,
+      "step": 13730
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.4436826537472167,
+      "learning_rate": 1.3930678397218112e-06,
+      "loss": 0.2111,
+      "step": 13731
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.2777165647529032,
+      "learning_rate": 1.3924440364175107e-06,
+      "loss": 0.1691,
+      "step": 13732
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.138961127104309,
+      "learning_rate": 1.39182035021655e-06,
+      "loss": 0.2231,
+      "step": 13733
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 3.0253393133157536,
+      "learning_rate": 1.3911967811391751e-06,
+      "loss": 0.2112,
+      "step": 13734
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.191792153568224,
+      "learning_rate": 1.3905733292056285e-06,
+      "loss": 0.1833,
+      "step": 13735
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.34707832660368,
+      "learning_rate": 1.3899499944361471e-06,
+      "loss": 0.1739,
+      "step": 13736
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.7045505861183432,
+      "learning_rate": 1.389326776850966e-06,
+      "loss": 0.2327,
+      "step": 13737
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.2445607456656016,
+      "learning_rate": 1.3887036764703126e-06,
+      "loss": 0.1968,
+      "step": 13738
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.9883732960171169,
+      "learning_rate": 1.3880806933144164e-06,
+      "loss": 0.1755,
+      "step": 13739
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.016833697519491,
+      "learning_rate": 1.387457827403496e-06,
+      "loss": 0.1653,
+      "step": 13740
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.157599673099264,
+      "learning_rate": 1.3868350787577734e-06,
+      "loss": 0.2066,
+      "step": 13741
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.2839720728615325,
+      "learning_rate": 1.3862124473974604e-06,
+      "loss": 0.1694,
+      "step": 13742
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.487820711744178,
+      "learning_rate": 1.3855899333427698e-06,
+      "loss": 0.2087,
+      "step": 13743
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.2717101311265628,
+      "learning_rate": 1.3849675366139075e-06,
+      "loss": 0.1872,
+      "step": 13744
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.6389293718411764,
+      "learning_rate": 1.3843452572310784e-06,
+      "loss": 0.2252,
+      "step": 13745
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.0586284934391743,
+      "learning_rate": 1.3837230952144832e-06,
+      "loss": 0.1749,
+      "step": 13746
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.6524393897187715,
+      "learning_rate": 1.383101050584314e-06,
+      "loss": 0.162,
+      "step": 13747
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.2230862723541314,
+      "learning_rate": 1.3824791233607643e-06,
+      "loss": 0.2281,
+      "step": 13748
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.647428538258818,
+      "learning_rate": 1.3818573135640228e-06,
+      "loss": 0.1967,
+      "step": 13749
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.252668429317033,
+      "learning_rate": 1.3812356212142747e-06,
+      "loss": 0.1948,
+      "step": 13750
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.3269237358964499,
+      "learning_rate": 1.380614046331698e-06,
+      "loss": 0.1999,
+      "step": 13751
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.2629772971865854,
+      "learning_rate": 1.3799925889364707e-06,
+      "loss": 0.154,
+      "step": 13752
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.1387113757340543,
+      "learning_rate": 1.3793712490487655e-06,
+      "loss": 0.1706,
+      "step": 13753
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.792476112653468,
+      "learning_rate": 1.3787500266887515e-06,
+      "loss": 0.2567,
+      "step": 13754
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.2807321548685606,
+      "learning_rate": 1.3781289218765952e-06,
+      "loss": 0.1552,
+      "step": 13755
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.2352744700384655,
+      "learning_rate": 1.377507934632456e-06,
+      "loss": 0.1948,
+      "step": 13756
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.4527428032101553,
+      "learning_rate": 1.3768870649764937e-06,
+      "loss": 0.1493,
+      "step": 13757
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.4381712921729248,
+      "learning_rate": 1.376266312928859e-06,
+      "loss": 0.1677,
+      "step": 13758
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.5717909576215134,
+      "learning_rate": 1.375645678509705e-06,
+      "loss": 0.1833,
+      "step": 13759
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.2683139569674515,
+      "learning_rate": 1.3750251617391752e-06,
+      "loss": 0.1991,
+      "step": 13760
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.3088218148611703,
+      "learning_rate": 1.3744047626374129e-06,
+      "loss": 0.2013,
+      "step": 13761
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.2906049054869853,
+      "learning_rate": 1.3737844812245571e-06,
+      "loss": 0.1851,
+      "step": 13762
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.9489202088783926,
+      "learning_rate": 1.3731643175207422e-06,
+      "loss": 0.1668,
+      "step": 13763
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.5581328396498582,
+      "learning_rate": 1.3725442715461e-06,
+      "loss": 0.1878,
+      "step": 13764
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.4471178069412465,
+      "learning_rate": 1.3719243433207564e-06,
+      "loss": 0.1669,
+      "step": 13765
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.5241237781877672,
+      "learning_rate": 1.3713045328648338e-06,
+      "loss": 0.2011,
+      "step": 13766
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.1334728552450077,
+      "learning_rate": 1.3706848401984536e-06,
+      "loss": 0.1792,
+      "step": 13767
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 2.1049884963230934,
+      "learning_rate": 1.3700652653417313e-06,
+      "loss": 0.2366,
+      "step": 13768
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.7011375204793442,
+      "learning_rate": 1.3694458083147766e-06,
+      "loss": 0.1929,
+      "step": 13769
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.6364231662099902,
+      "learning_rate": 1.3688264691376985e-06,
+      "loss": 0.1799,
+      "step": 13770
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.2326917549297378,
+      "learning_rate": 1.3682072478306018e-06,
+      "loss": 0.1509,
+      "step": 13771
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.9922703648303747,
+      "learning_rate": 1.3675881444135859e-06,
+      "loss": 0.1317,
+      "step": 13772
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.7567316971155613,
+      "learning_rate": 1.3669691589067491e-06,
+      "loss": 0.1578,
+      "step": 13773
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.4112624138580911,
+      "learning_rate": 1.3663502913301812e-06,
+      "loss": 0.1934,
+      "step": 13774
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 2.957807028433388,
+      "learning_rate": 1.3657315417039734e-06,
+      "loss": 0.223,
+      "step": 13775
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.220276295616969,
+      "learning_rate": 1.3651129100482086e-06,
+      "loss": 0.1786,
+      "step": 13776
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.1827845435322046,
+      "learning_rate": 1.3644943963829705e-06,
+      "loss": 0.1422,
+      "step": 13777
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.6032429322706028,
+      "learning_rate": 1.363876000728333e-06,
+      "loss": 0.1807,
+      "step": 13778
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.3456358982784125,
+      "learning_rate": 1.3632577231043714e-06,
+      "loss": 0.1835,
+      "step": 13779
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.2747781313898547,
+      "learning_rate": 1.3626395635311557e-06,
+      "loss": 0.1746,
+      "step": 13780
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.3154731409135754,
+      "learning_rate": 1.3620215220287515e-06,
+      "loss": 0.161,
+      "step": 13781
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.2396633545774547,
+      "learning_rate": 1.361403598617222e-06,
+      "loss": 0.1769,
+      "step": 13782
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.4925165562534093,
+      "learning_rate": 1.3607857933166224e-06,
+      "loss": 0.1821,
+      "step": 13783
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.4209682329477293,
+      "learning_rate": 1.3601681061470085e-06,
+      "loss": 0.1984,
+      "step": 13784
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.3814370917570893,
+      "learning_rate": 1.3595505371284313e-06,
+      "loss": 0.1718,
+      "step": 13785
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.243950345686081,
+      "learning_rate": 1.3589330862809386e-06,
+      "loss": 0.1791,
+      "step": 13786
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.7453726574563015,
+      "learning_rate": 1.35831575362457e-06,
+      "loss": 0.221,
+      "step": 13787
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.2907237590107692,
+      "learning_rate": 1.3576985391793662e-06,
+      "loss": 0.1653,
+      "step": 13788
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.6745766131304263,
+      "learning_rate": 1.3570814429653623e-06,
+      "loss": 0.2107,
+      "step": 13789
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.2366955619416595,
+      "learning_rate": 1.3564644650025894e-06,
+      "loss": 0.1586,
+      "step": 13790
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.2242563764076189,
+      "learning_rate": 1.3558476053110769e-06,
+      "loss": 0.1816,
+      "step": 13791
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.5903622018577528,
+      "learning_rate": 1.3552308639108452e-06,
+      "loss": 0.1759,
+      "step": 13792
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.4210672494557528,
+      "learning_rate": 1.354614240821917e-06,
+      "loss": 0.2066,
+      "step": 13793
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.6672924600397763,
+      "learning_rate": 1.3539977360643054e-06,
+      "loss": 0.2033,
+      "step": 13794
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.6588580272715472,
+      "learning_rate": 1.353381349658025e-06,
+      "loss": 0.18,
+      "step": 13795
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.5402445538131506,
+      "learning_rate": 1.3527650816230813e-06,
+      "loss": 0.1668,
+      "step": 13796
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.5821984285161403,
+      "learning_rate": 1.3521489319794812e-06,
+      "loss": 0.1999,
+      "step": 13797
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.4044755893575802,
+      "learning_rate": 1.351532900747224e-06,
+      "loss": 0.1786,
+      "step": 13798
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.9836134689693552,
+      "learning_rate": 1.3509169879463062e-06,
+      "loss": 0.1276,
+      "step": 13799
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.5031622867162835,
+      "learning_rate": 1.3503011935967237e-06,
+      "loss": 0.1744,
+      "step": 13800
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.3874361627086753,
+      "learning_rate": 1.3496855177184614e-06,
+      "loss": 0.2313,
+      "step": 13801
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.1903353302434443,
+      "learning_rate": 1.349069960331506e-06,
+      "loss": 0.1836,
+      "step": 13802
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.4301279124450437,
+      "learning_rate": 1.3484545214558399e-06,
+      "loss": 0.2182,
+      "step": 13803
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.4110577837259521,
+      "learning_rate": 1.3478392011114406e-06,
+      "loss": 0.176,
+      "step": 13804
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.0287633667769265,
+      "learning_rate": 1.3472239993182802e-06,
+      "loss": 0.1519,
+      "step": 13805
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.7822152291551574,
+      "learning_rate": 1.3466089160963285e-06,
+      "loss": 0.2315,
+      "step": 13806
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.3795111925872925,
+      "learning_rate": 1.3459939514655523e-06,
+      "loss": 0.1897,
+      "step": 13807
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.523483759837952,
+      "learning_rate": 1.3453791054459154e-06,
+      "loss": 0.17,
+      "step": 13808
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.2078863898569985,
+      "learning_rate": 1.3447643780573727e-06,
+      "loss": 0.2128,
+      "step": 13809
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.5148292940186787,
+      "learning_rate": 1.3441497693198807e-06,
+      "loss": 0.1737,
+      "step": 13810
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.9956922731822095,
+      "learning_rate": 1.34353527925339e-06,
+      "loss": 0.1369,
+      "step": 13811
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.0283915177170175,
+      "learning_rate": 1.3429209078778461e-06,
+      "loss": 0.1646,
+      "step": 13812
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.95138764332986,
+      "learning_rate": 1.3423066552131937e-06,
+      "loss": 0.1883,
+      "step": 13813
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.4173710748323936,
+      "learning_rate": 1.3416925212793691e-06,
+      "loss": 0.2045,
+      "step": 13814
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.2763296791642076,
+      "learning_rate": 1.3410785060963088e-06,
+      "loss": 0.1812,
+      "step": 13815
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.5010419048315315,
+      "learning_rate": 1.3404646096839446e-06,
+      "loss": 0.1963,
+      "step": 13816
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.705354368853827,
+      "learning_rate": 1.3398508320622034e-06,
+      "loss": 0.1757,
+      "step": 13817
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.258995335194185,
+      "learning_rate": 1.33923717325101e-06,
+      "loss": 0.1968,
+      "step": 13818
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.1424865131656845,
+      "learning_rate": 1.3386236332702817e-06,
+      "loss": 0.1811,
+      "step": 13819
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.6153892433620713,
+      "learning_rate": 1.3380102121399357e-06,
+      "loss": 0.1871,
+      "step": 13820
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.2942734097225508,
+      "learning_rate": 1.3373969098798845e-06,
+      "loss": 0.1576,
+      "step": 13821
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.3087813171786014,
+      "learning_rate": 1.336783726510037e-06,
+      "loss": 0.2109,
+      "step": 13822
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.0698384606559483,
+      "learning_rate": 1.3361706620502946e-06,
+      "loss": 0.164,
+      "step": 13823
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.0338992854473679,
+      "learning_rate": 1.3355577165205597e-06,
+      "loss": 0.1831,
+      "step": 13824
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.023895117683033,
+      "learning_rate": 1.334944889940728e-06,
+      "loss": 0.1355,
+      "step": 13825
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 2.200577067422228,
+      "learning_rate": 1.3343321823306943e-06,
+      "loss": 0.2408,
+      "step": 13826
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.3391127190804364,
+      "learning_rate": 1.3337195937103448e-06,
+      "loss": 0.1555,
+      "step": 13827
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.3450219211600067,
+      "learning_rate": 1.3331071240995668e-06,
+      "loss": 0.2018,
+      "step": 13828
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.4105361469022462,
+      "learning_rate": 1.3324947735182386e-06,
+      "loss": 0.1635,
+      "step": 13829
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.8948093013631366,
+      "learning_rate": 1.3318825419862392e-06,
+      "loss": 0.2136,
+      "step": 13830
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.2080817450532253,
+      "learning_rate": 1.331270429523443e-06,
+      "loss": 0.1794,
+      "step": 13831
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.4071071568802473,
+      "learning_rate": 1.3306584361497171e-06,
+      "loss": 0.2135,
+      "step": 13832
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.368692790103038,
+      "learning_rate": 1.3300465618849283e-06,
+      "loss": 0.219,
+      "step": 13833
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.4416907011722477,
+      "learning_rate": 1.3294348067489383e-06,
+      "loss": 0.1935,
+      "step": 13834
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.580978721598516,
+      "learning_rate": 1.3288231707616055e-06,
+      "loss": 0.2003,
+      "step": 13835
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.403280290790838,
+      "learning_rate": 1.3282116539427852e-06,
+      "loss": 0.2093,
+      "step": 13836
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.382650441301844,
+      "learning_rate": 1.3276002563123242e-06,
+      "loss": 0.2006,
+      "step": 13837
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.18611152311981,
+      "learning_rate": 1.3269889778900708e-06,
+      "loss": 0.202,
+      "step": 13838
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.297335419447197,
+      "learning_rate": 1.3263778186958676e-06,
+      "loss": 0.1759,
+      "step": 13839
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.4539878498446015,
+      "learning_rate": 1.3257667787495537e-06,
+      "loss": 0.2058,
+      "step": 13840
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.1640134583453847,
+      "learning_rate": 1.3251558580709623e-06,
+      "loss": 0.1557,
+      "step": 13841
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.9663390915709685,
+      "learning_rate": 1.3245450566799245e-06,
+      "loss": 0.1483,
+      "step": 13842
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.2887175213978983,
+      "learning_rate": 1.323934374596268e-06,
+      "loss": 0.214,
+      "step": 13843
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.4037868637393407,
+      "learning_rate": 1.3233238118398166e-06,
+      "loss": 0.1535,
+      "step": 13844
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.670547225488989,
+      "learning_rate": 1.3227133684303872e-06,
+      "loss": 0.2206,
+      "step": 13845
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.0413856922132365,
+      "learning_rate": 1.322103044387798e-06,
+      "loss": 0.1272,
+      "step": 13846
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.42571244531031,
+      "learning_rate": 1.3214928397318578e-06,
+      "loss": 0.159,
+      "step": 13847
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.1992175711088566,
+      "learning_rate": 1.3208827544823743e-06,
+      "loss": 0.2022,
+      "step": 13848
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.9783697076605364,
+      "learning_rate": 1.320272788659155e-06,
+      "loss": 0.1737,
+      "step": 13849
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.3100602970688047,
+      "learning_rate": 1.3196629422819945e-06,
+      "loss": 0.1398,
+      "step": 13850
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.8552841094654127,
+      "learning_rate": 1.3190532153706915e-06,
+      "loss": 0.1843,
+      "step": 13851
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.3518980662573314,
+      "learning_rate": 1.318443607945038e-06,
+      "loss": 0.176,
+      "step": 13852
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.2118691767296408,
+      "learning_rate": 1.3178341200248219e-06,
+      "loss": 0.1673,
+      "step": 13853
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.44733798660816,
+      "learning_rate": 1.3172247516298293e-06,
+      "loss": 0.2386,
+      "step": 13854
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.4003617617560165,
+      "learning_rate": 1.3166155027798371e-06,
+      "loss": 0.1467,
+      "step": 13855
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.5357526198344151,
+      "learning_rate": 1.3160063734946244e-06,
+      "loss": 0.1896,
+      "step": 13856
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.5664163241930515,
+      "learning_rate": 1.3153973637939632e-06,
+      "loss": 0.1715,
+      "step": 13857
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.5780407297655061,
+      "learning_rate": 1.3147884736976234e-06,
+      "loss": 0.1626,
+      "step": 13858
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.3823349338171453,
+      "learning_rate": 1.3141797032253677e-06,
+      "loss": 0.1861,
+      "step": 13859
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.3715847907576704,
+      "learning_rate": 1.313571052396958e-06,
+      "loss": 0.1794,
+      "step": 13860
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.295454166152205,
+      "learning_rate": 1.3129625212321523e-06,
+      "loss": 0.2426,
+      "step": 13861
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.303071274438891,
+      "learning_rate": 1.3123541097507048e-06,
+      "loss": 0.1523,
+      "step": 13862
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.557711640845476,
+      "learning_rate": 1.3117458179723614e-06,
+      "loss": 0.1591,
+      "step": 13863
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.2339004527913817,
+      "learning_rate": 1.3111376459168718e-06,
+      "loss": 0.1869,
+      "step": 13864
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.2772787293383676,
+      "learning_rate": 1.3105295936039735e-06,
+      "loss": 0.1751,
+      "step": 13865
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.669293558697038,
+      "learning_rate": 1.3099216610534065e-06,
+      "loss": 0.2121,
+      "step": 13866
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.4177200267383316,
+      "learning_rate": 1.3093138482849054e-06,
+      "loss": 0.2389,
+      "step": 13867
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.2919304703945715,
+      "learning_rate": 1.308706155318198e-06,
+      "loss": 0.2051,
+      "step": 13868
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.8404078175725889,
+      "learning_rate": 1.3080985821730113e-06,
+      "loss": 0.1845,
+      "step": 13869
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.9675864492280845,
+      "learning_rate": 1.3074911288690679e-06,
+      "loss": 0.1773,
+      "step": 13870
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.4257492375372638,
+      "learning_rate": 1.3068837954260848e-06,
+      "loss": 0.2037,
+      "step": 13871
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.507083321201253,
+      "learning_rate": 1.3062765818637802e-06,
+      "loss": 0.1848,
+      "step": 13872
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.2427483677977098,
+      "learning_rate": 1.3056694882018595e-06,
+      "loss": 0.1909,
+      "step": 13873
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.014984750680107,
+      "learning_rate": 1.3050625144600316e-06,
+      "loss": 0.1314,
+      "step": 13874
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 2.1889759670147693,
+      "learning_rate": 1.3044556606579994e-06,
+      "loss": 0.2283,
+      "step": 13875
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.6712603999108884,
+      "learning_rate": 1.3038489268154631e-06,
+      "loss": 0.215,
+      "step": 13876
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.5976510689052148,
+      "learning_rate": 1.3032423129521148e-06,
+      "loss": 0.2024,
+      "step": 13877
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.4038227803715908,
+      "learning_rate": 1.3026358190876466e-06,
+      "loss": 0.2322,
+      "step": 13878
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.0989177758143835,
+      "learning_rate": 1.302029445241746e-06,
+      "loss": 0.1363,
+      "step": 13879
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.0892638232479293,
+      "learning_rate": 1.3014231914340969e-06,
+      "loss": 0.1397,
+      "step": 13880
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.3531732883162537,
+      "learning_rate": 1.300817057684377e-06,
+      "loss": 0.2218,
+      "step": 13881
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.2129098949038035,
+      "learning_rate": 1.3002110440122639e-06,
+      "loss": 0.1655,
+      "step": 13882
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.1040617781279618,
+      "learning_rate": 1.2996051504374263e-06,
+      "loss": 0.1288,
+      "step": 13883
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 2.0427738702140603,
+      "learning_rate": 1.2989993769795333e-06,
+      "loss": 0.1828,
+      "step": 13884
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.3674352857299017,
+      "learning_rate": 1.2983937236582506e-06,
+      "loss": 0.1602,
+      "step": 13885
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.3134897564647214,
+      "learning_rate": 1.297788190493235e-06,
+      "loss": 0.166,
+      "step": 13886
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.4283661077464127,
+      "learning_rate": 1.297182777504143e-06,
+      "loss": 0.1966,
+      "step": 13887
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.5952233255030672,
+      "learning_rate": 1.2965774847106273e-06,
+      "loss": 0.2106,
+      "step": 13888
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.5199654272678385,
+      "learning_rate": 1.2959723121323365e-06,
+      "loss": 0.1501,
+      "step": 13889
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.973541550103971,
+      "learning_rate": 1.2953672597889155e-06,
+      "loss": 0.168,
+      "step": 13890
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 2.0111488572075524,
+      "learning_rate": 1.2947623277000021e-06,
+      "loss": 0.2127,
+      "step": 13891
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.0804417741266958,
+      "learning_rate": 1.2941575158852343e-06,
+      "loss": 0.1306,
+      "step": 13892
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.2671486489940709,
+      "learning_rate": 1.2935528243642449e-06,
+      "loss": 0.1659,
+      "step": 13893
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.3087367819434645,
+      "learning_rate": 1.2929482531566634e-06,
+      "loss": 0.2056,
+      "step": 13894
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.2472771075084854,
+      "learning_rate": 1.2923438022821117e-06,
+      "loss": 0.224,
+      "step": 13895
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.1930017762774459,
+      "learning_rate": 1.2917394717602123e-06,
+      "loss": 0.1472,
+      "step": 13896
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.9487054245294203,
+      "learning_rate": 1.2911352616105821e-06,
+      "loss": 0.1363,
+      "step": 13897
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.5131353104093161,
+      "learning_rate": 1.2905311718528356e-06,
+      "loss": 0.2055,
+      "step": 13898
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.4302349553706373,
+      "learning_rate": 1.2899272025065783e-06,
+      "loss": 0.2066,
+      "step": 13899
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.1176575425485982,
+      "learning_rate": 1.2893233535914196e-06,
+      "loss": 0.1431,
+      "step": 13900
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.737108430718486,
+      "learning_rate": 1.2887196251269573e-06,
+      "loss": 0.2238,
+      "step": 13901
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.5389733990665586,
+      "learning_rate": 1.2881160171327895e-06,
+      "loss": 0.1537,
+      "step": 13902
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.5046344207995772,
+      "learning_rate": 1.2875125296285124e-06,
+      "loss": 0.2216,
+      "step": 13903
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.1037205699904675,
+      "learning_rate": 1.2869091626337116e-06,
+      "loss": 0.1678,
+      "step": 13904
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.5994301999529108,
+      "learning_rate": 1.2863059161679742e-06,
+      "loss": 0.2132,
+      "step": 13905
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.4100932374015784,
+      "learning_rate": 1.2857027902508829e-06,
+      "loss": 0.2158,
+      "step": 13906
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.2208882153378628,
+      "learning_rate": 1.2850997849020148e-06,
+      "loss": 0.2001,
+      "step": 13907
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.372584703347054,
+      "learning_rate": 1.2844969001409447e-06,
+      "loss": 0.2073,
+      "step": 13908
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.3691880494092823,
+      "learning_rate": 1.283894135987241e-06,
+      "loss": 0.1819,
+      "step": 13909
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.1862238124975648,
+      "learning_rate": 1.2832914924604705e-06,
+      "loss": 0.2027,
+      "step": 13910
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.095068161113461,
+      "learning_rate": 1.2826889695801953e-06,
+      "loss": 0.1559,
+      "step": 13911
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.5003361970920701,
+      "learning_rate": 1.2820865673659748e-06,
+      "loss": 0.2053,
+      "step": 13912
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.3782842978709848,
+      "learning_rate": 1.2814842858373615e-06,
+      "loss": 0.2158,
+      "step": 13913
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.994298948560207,
+      "learning_rate": 1.2808821250139063e-06,
+      "loss": 0.1423,
+      "step": 13914
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.5166102603778384,
+      "learning_rate": 1.2802800849151558e-06,
+      "loss": 0.1731,
+      "step": 13915
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.65344201781817,
+      "learning_rate": 1.2796781655606544e-06,
+      "loss": 0.1849,
+      "step": 13916
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.13130043665536,
+      "learning_rate": 1.279076366969938e-06,
+      "loss": 0.2027,
+      "step": 13917
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.509800475040325,
+      "learning_rate": 1.2784746891625432e-06,
+      "loss": 0.206,
+      "step": 13918
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.354256546876845,
+      "learning_rate": 1.277873132157999e-06,
+      "loss": 0.1368,
+      "step": 13919
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.3815858486578003,
+      "learning_rate": 1.277271695975833e-06,
+      "loss": 0.2087,
+      "step": 13920
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.6616807647223077,
+      "learning_rate": 1.2766703806355702e-06,
+      "loss": 0.2013,
+      "step": 13921
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.1288236370019895,
+      "learning_rate": 1.2760691861567258e-06,
+      "loss": 0.0977,
+      "step": 13922
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.5103295742335034,
+      "learning_rate": 1.2754681125588182e-06,
+      "loss": 0.1482,
+      "step": 13923
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.3567742126798614,
+      "learning_rate": 1.2748671598613565e-06,
+      "loss": 0.1516,
+      "step": 13924
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.3431323761202714,
+      "learning_rate": 1.2742663280838491e-06,
+      "loss": 0.1698,
+      "step": 13925
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.8851770913683025,
+      "learning_rate": 1.2736656172458006e-06,
+      "loss": 0.1334,
+      "step": 13926
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.2379302691692744,
+      "learning_rate": 1.2730650273667077e-06,
+      "loss": 0.1636,
+      "step": 13927
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.2089738429385726,
+      "learning_rate": 1.2724645584660672e-06,
+      "loss": 0.1685,
+      "step": 13928
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.9705729673142038,
+      "learning_rate": 1.27186421056337e-06,
+      "loss": 0.2288,
+      "step": 13929
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.365120311332554,
+      "learning_rate": 1.2712639836781048e-06,
+      "loss": 0.2267,
+      "step": 13930
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.4014974500345092,
+      "learning_rate": 1.2706638778297563e-06,
+      "loss": 0.2037,
+      "step": 13931
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.0156089231245675,
+      "learning_rate": 1.2700638930378017e-06,
+      "loss": 0.1322,
+      "step": 13932
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.1203751670197168,
+      "learning_rate": 1.2694640293217175e-06,
+      "loss": 0.1646,
+      "step": 13933
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.277393958361953,
+      "learning_rate": 1.268864286700978e-06,
+      "loss": 0.1466,
+      "step": 13934
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.0113017279523513,
+      "learning_rate": 1.2682646651950476e-06,
+      "loss": 0.1231,
+      "step": 13935
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.2027071220925543,
+      "learning_rate": 1.2676651648233933e-06,
+      "loss": 0.1455,
+      "step": 13936
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.3965753485651313,
+      "learning_rate": 1.267065785605473e-06,
+      "loss": 0.1997,
+      "step": 13937
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.2731179204958385,
+      "learning_rate": 1.2664665275607435e-06,
+      "loss": 0.2161,
+      "step": 13938
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.9528996873540222,
+      "learning_rate": 1.2658673907086577e-06,
+      "loss": 0.1355,
+      "step": 13939
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.2803453706018482,
+      "learning_rate": 1.265268375068665e-06,
+      "loss": 0.1802,
+      "step": 13940
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.8057061000157935,
+      "learning_rate": 1.2646694806602073e-06,
+      "loss": 0.21,
+      "step": 13941
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.5013915501515183,
+      "learning_rate": 1.264070707502726e-06,
+      "loss": 0.1676,
+      "step": 13942
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.0520345240623736,
+      "learning_rate": 1.2634720556156577e-06,
+      "loss": 0.1348,
+      "step": 13943
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.1738715774694217,
+      "learning_rate": 1.2628735250184372e-06,
+      "loss": 0.1798,
+      "step": 13944
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.3562215863604736,
+      "learning_rate": 1.262275115730489e-06,
+      "loss": 0.1757,
+      "step": 13945
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 3.051724921697757,
+      "learning_rate": 1.2616768277712398e-06,
+      "loss": 0.2369,
+      "step": 13946
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.3309286705811572,
+      "learning_rate": 1.2610786611601105e-06,
+      "loss": 0.157,
+      "step": 13947
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.1775293946798617,
+      "learning_rate": 1.2604806159165184e-06,
+      "loss": 0.1551,
+      "step": 13948
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.8879451122665287,
+      "learning_rate": 1.2598826920598773e-06,
+      "loss": 0.2517,
+      "step": 13949
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.6101983218964153,
+      "learning_rate": 1.2592848896095928e-06,
+      "loss": 0.2172,
+      "step": 13950
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.1281185867201244,
+      "learning_rate": 1.258687208585072e-06,
+      "loss": 0.1548,
+      "step": 13951
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.4847275566232834,
+      "learning_rate": 1.2580896490057176e-06,
+      "loss": 0.2047,
+      "step": 13952
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.2218942674839697,
+      "learning_rate": 1.2574922108909232e-06,
+      "loss": 0.1642,
+      "step": 13953
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.2180227408392414,
+      "learning_rate": 1.256894894260085e-06,
+      "loss": 0.1425,
+      "step": 13954
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.2466390247816408,
+      "learning_rate": 1.2562976991325898e-06,
+      "loss": 0.1884,
+      "step": 13955
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.3650297467253232,
+      "learning_rate": 1.2557006255278243e-06,
+      "loss": 0.1518,
+      "step": 13956
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.2181157882835354,
+      "learning_rate": 1.2551036734651688e-06,
+      "loss": 0.2031,
+      "step": 13957
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.5003192457577021,
+      "learning_rate": 1.254506842964004e-06,
+      "loss": 0.184,
+      "step": 13958
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.4352540943763112,
+      "learning_rate": 1.2539101340436987e-06,
+      "loss": 0.1443,
+      "step": 13959
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.2968378636202147,
+      "learning_rate": 1.253313546723625e-06,
+      "loss": 0.1943,
+      "step": 13960
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.449937780870481,
+      "learning_rate": 1.2527170810231483e-06,
+      "loss": 0.1909,
+      "step": 13961
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.319992734909836,
+      "learning_rate": 1.2521207369616307e-06,
+      "loss": 0.143,
+      "step": 13962
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.3950286266763428,
+      "learning_rate": 1.251524514558428e-06,
+      "loss": 0.2118,
+      "step": 13963
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.715711934331204,
+      "learning_rate": 1.2509284138328953e-06,
+      "loss": 0.2162,
+      "step": 13964
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.6917199269043395,
+      "learning_rate": 1.2503324348043821e-06,
+      "loss": 0.2252,
+      "step": 13965
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.437397015033856,
+      "learning_rate": 1.2497365774922342e-06,
+      "loss": 0.1848,
+      "step": 13966
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.5222142810693675,
+      "learning_rate": 1.2491408419157951e-06,
+      "loss": 0.1848,
+      "step": 13967
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.49818343692703,
+      "learning_rate": 1.2485452280943994e-06,
+      "loss": 0.2016,
+      "step": 13968
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.074177578223785,
+      "learning_rate": 1.247949736047383e-06,
+      "loss": 0.1816,
+      "step": 13969
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.5396577826330897,
+      "learning_rate": 1.2473543657940774e-06,
+      "loss": 0.1549,
+      "step": 13970
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 2.7725776790214485,
+      "learning_rate": 1.2467591173538051e-06,
+      "loss": 0.1969,
+      "step": 13971
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.533266515696974,
+      "learning_rate": 1.2461639907458922e-06,
+      "loss": 0.162,
+      "step": 13972
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.6783489393991124,
+      "learning_rate": 1.245568985989653e-06,
+      "loss": 0.1878,
+      "step": 13973
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.042719750559389,
+      "learning_rate": 1.2449741031044037e-06,
+      "loss": 0.124,
+      "step": 13974
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.2978917123570368,
+      "learning_rate": 1.2443793421094542e-06,
+      "loss": 0.1885,
+      "step": 13975
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.4197215761482063,
+      "learning_rate": 1.2437847030241118e-06,
+      "loss": 0.2058,
+      "step": 13976
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.3799881810523553,
+      "learning_rate": 1.2431901858676771e-06,
+      "loss": 0.186,
+      "step": 13977
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.8028851546262277,
+      "learning_rate": 1.2425957906594488e-06,
+      "loss": 0.2056,
+      "step": 13978
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.6769254997176353,
+      "learning_rate": 1.2420015174187223e-06,
+      "loss": 0.1458,
+      "step": 13979
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.8592712252988988,
+      "learning_rate": 1.2414073661647891e-06,
+      "loss": 0.17,
+      "step": 13980
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.6068357833726836,
+      "learning_rate": 1.2408133369169324e-06,
+      "loss": 0.249,
+      "step": 13981
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.4231214300118402,
+      "learning_rate": 1.2402194296944364e-06,
+      "loss": 0.1573,
+      "step": 13982
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.8589156368762825,
+      "learning_rate": 1.2396256445165794e-06,
+      "loss": 0.2239,
+      "step": 13983
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.453712298551366,
+      "learning_rate": 1.2390319814026369e-06,
+      "loss": 0.1685,
+      "step": 13984
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.953448116277914,
+      "learning_rate": 1.2384384403718803e-06,
+      "loss": 0.1965,
+      "step": 13985
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.0016474936213862,
+      "learning_rate": 1.2378450214435733e-06,
+      "loss": 0.1522,
+      "step": 13986
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.5024275507890894,
+      "learning_rate": 1.2372517246369808e-06,
+      "loss": 0.1844,
+      "step": 13987
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.513999883736766,
+      "learning_rate": 1.2366585499713618e-06,
+      "loss": 0.1982,
+      "step": 13988
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.0221885165212692,
+      "learning_rate": 1.2360654974659698e-06,
+      "loss": 0.1517,
+      "step": 13989
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.692911509033347,
+      "learning_rate": 1.2354725671400569e-06,
+      "loss": 0.1878,
+      "step": 13990
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.051490057154644,
+      "learning_rate": 1.234879759012868e-06,
+      "loss": 0.1997,
+      "step": 13991
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.0788812472839124,
+      "learning_rate": 1.2342870731036472e-06,
+      "loss": 0.1933,
+      "step": 13992
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.9936193699719819,
+      "learning_rate": 1.2336945094316332e-06,
+      "loss": 0.1748,
+      "step": 13993
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.5326725229944695,
+      "learning_rate": 1.2331020680160633e-06,
+      "loss": 0.2088,
+      "step": 13994
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.074364530901806,
+      "learning_rate": 1.2325097488761646e-06,
+      "loss": 0.1664,
+      "step": 13995
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.501201277759317,
+      "learning_rate": 1.2319175520311655e-06,
+      "loss": 0.1725,
+      "step": 13996
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.3129096073230013,
+      "learning_rate": 1.2313254775002902e-06,
+      "loss": 0.2113,
+      "step": 13997
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.154082165353493,
+      "learning_rate": 1.230733525302758e-06,
+      "loss": 0.1463,
+      "step": 13998
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.3041162568001918,
+      "learning_rate": 1.2301416954577815e-06,
+      "loss": 0.1703,
+      "step": 13999
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.3557627136538961,
+      "learning_rate": 1.229549987984574e-06,
+      "loss": 0.1525,
+      "step": 14000
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.6527115297553074,
+      "learning_rate": 1.2289584029023415e-06,
+      "loss": 0.2047,
+      "step": 14001
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.4517425914034725,
+      "learning_rate": 1.2283669402302878e-06,
+      "loss": 0.2268,
+      "step": 14002
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.721011498971113,
+      "learning_rate": 1.227775599987614e-06,
+      "loss": 0.1807,
+      "step": 14003
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.027653221469852,
+      "learning_rate": 1.2271843821935114e-06,
+      "loss": 0.0928,
+      "step": 14004
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.1630711654634642,
+      "learning_rate": 1.2265932868671737e-06,
+      "loss": 0.1699,
+      "step": 14005
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.179904879821617,
+      "learning_rate": 1.2260023140277887e-06,
+      "loss": 0.16,
+      "step": 14006
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.3830605139008953,
+      "learning_rate": 1.2254114636945375e-06,
+      "loss": 0.1819,
+      "step": 14007
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.4906476099070933,
+      "learning_rate": 1.2248207358866026e-06,
+      "loss": 0.1965,
+      "step": 14008
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.4894645072207973,
+      "learning_rate": 1.2242301306231557e-06,
+      "loss": 0.2009,
+      "step": 14009
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.0771871717168642,
+      "learning_rate": 1.2236396479233703e-06,
+      "loss": 0.125,
+      "step": 14010
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.4693744985481423,
+      "learning_rate": 1.2230492878064132e-06,
+      "loss": 0.1997,
+      "step": 14011
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.0832317660782584,
+      "learning_rate": 1.2224590502914497e-06,
+      "loss": 0.1567,
+      "step": 14012
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.3371150978137534,
+      "learning_rate": 1.221868935397636e-06,
+      "loss": 0.1689,
+      "step": 14013
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.4146080650710746,
+      "learning_rate": 1.221278943144129e-06,
+      "loss": 0.159,
+      "step": 14014
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.4523420737073724,
+      "learning_rate": 1.2206890735500809e-06,
+      "loss": 0.1907,
+      "step": 14015
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.9685684272359271,
+      "learning_rate": 1.2200993266346401e-06,
+      "loss": 0.1582,
+      "step": 14016
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.4966850755572887,
+      "learning_rate": 1.2195097024169473e-06,
+      "loss": 0.1907,
+      "step": 14017
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.5553258036020925,
+      "learning_rate": 1.2189202009161433e-06,
+      "loss": 0.2343,
+      "step": 14018
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.4744252070941737,
+      "learning_rate": 1.218330822151364e-06,
+      "loss": 0.1759,
+      "step": 14019
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.4016216246468085,
+      "learning_rate": 1.2177415661417412e-06,
+      "loss": 0.1935,
+      "step": 14020
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.4490051112061804,
+      "learning_rate": 1.2171524329064032e-06,
+      "loss": 0.2082,
+      "step": 14021
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.1933261785556883,
+      "learning_rate": 1.2165634224644712e-06,
+      "loss": 0.177,
+      "step": 14022
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.0798060469406132,
+      "learning_rate": 1.215974534835066e-06,
+      "loss": 0.1551,
+      "step": 14023
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.3689058839962724,
+      "learning_rate": 1.2153857700373045e-06,
+      "loss": 0.1771,
+      "step": 14024
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.528076922746685,
+      "learning_rate": 1.2147971280902964e-06,
+      "loss": 0.1923,
+      "step": 14025
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.3045285921918428,
+      "learning_rate": 1.214208609013151e-06,
+      "loss": 0.1745,
+      "step": 14026
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.5326237355230057,
+      "learning_rate": 1.2136202128249702e-06,
+      "loss": 0.1915,
+      "step": 14027
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.3329932299364766,
+      "learning_rate": 1.2130319395448548e-06,
+      "loss": 0.2033,
+      "step": 14028
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.6155595233052613,
+      "learning_rate": 1.2124437891918995e-06,
+      "loss": 0.1992,
+      "step": 14029
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.9271984961950176,
+      "learning_rate": 1.2118557617851988e-06,
+      "loss": 0.1559,
+      "step": 14030
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.7017673803450377,
+      "learning_rate": 1.2112678573438369e-06,
+      "loss": 0.206,
+      "step": 14031
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.4289292529190631,
+      "learning_rate": 1.2106800758868991e-06,
+      "loss": 0.1902,
+      "step": 14032
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.2367363100739066,
+      "learning_rate": 1.2100924174334649e-06,
+      "loss": 0.1853,
+      "step": 14033
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.3046821320017867,
+      "learning_rate": 1.2095048820026118e-06,
+      "loss": 0.1617,
+      "step": 14034
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.030400242165864,
+      "learning_rate": 1.2089174696134083e-06,
+      "loss": 0.1572,
+      "step": 14035
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.1194682681888193,
+      "learning_rate": 1.2083301802849234e-06,
+      "loss": 0.1747,
+      "step": 14036
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.0715207568315164,
+      "learning_rate": 1.2077430140362217e-06,
+      "loss": 0.1536,
+      "step": 14037
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.0620643058999009,
+      "learning_rate": 1.2071559708863628e-06,
+      "loss": 0.1723,
+      "step": 14038
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.6299377258448653,
+      "learning_rate": 1.2065690508544032e-06,
+      "loss": 0.2387,
+      "step": 14039
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.5077524222625063,
+      "learning_rate": 1.2059822539593917e-06,
+      "loss": 0.1925,
+      "step": 14040
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 2.0075182791931137,
+      "learning_rate": 1.2053955802203787e-06,
+      "loss": 0.16,
+      "step": 14041
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.106371814007105,
+      "learning_rate": 1.204809029656408e-06,
+      "loss": 0.1909,
+      "step": 14042
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.1045281179235023,
+      "learning_rate": 1.2042226022865179e-06,
+      "loss": 0.1888,
+      "step": 14043
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.389120799279776,
+      "learning_rate": 1.2036362981297455e-06,
+      "loss": 0.2163,
+      "step": 14044
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.2375091167075603,
+      "learning_rate": 1.2030501172051206e-06,
+      "loss": 0.2179,
+      "step": 14045
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.0528815171946857,
+      "learning_rate": 1.2024640595316727e-06,
+      "loss": 0.1414,
+      "step": 14046
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.2370579389955336,
+      "learning_rate": 1.201878125128425e-06,
+      "loss": 0.1837,
+      "step": 14047
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.0485260526665692,
+      "learning_rate": 1.201292314014399e-06,
+      "loss": 0.1898,
+      "step": 14048
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.0988444247428548,
+      "learning_rate": 1.2007066262086076e-06,
+      "loss": 0.1548,
+      "step": 14049
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.4351541721617598,
+      "learning_rate": 1.2001210617300634e-06,
+      "loss": 0.2146,
+      "step": 14050
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.5955005643252453,
+      "learning_rate": 1.1995356205977748e-06,
+      "loss": 0.189,
+      "step": 14051
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.1134289827537625,
+      "learning_rate": 1.198950302830747e-06,
+      "loss": 0.1816,
+      "step": 14052
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.2868490751840838,
+      "learning_rate": 1.1983651084479763e-06,
+      "loss": 0.1761,
+      "step": 14053
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.9682587985688698,
+      "learning_rate": 1.1977800374684606e-06,
+      "loss": 0.1528,
+      "step": 14054
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.545700494331626,
+      "learning_rate": 1.197195089911191e-06,
+      "loss": 0.2298,
+      "step": 14055
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.0596287636521866,
+      "learning_rate": 1.1966102657951562e-06,
+      "loss": 0.1458,
+      "step": 14056
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.1035128888796755,
+      "learning_rate": 1.1960255651393398e-06,
+      "loss": 0.151,
+      "step": 14057
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.305569393517778,
+      "learning_rate": 1.1954409879627204e-06,
+      "loss": 0.1711,
+      "step": 14058
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.3137254216864946,
+      "learning_rate": 1.1948565342842739e-06,
+      "loss": 0.1482,
+      "step": 14059
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.0931291452965022,
+      "learning_rate": 1.1942722041229736e-06,
+      "loss": 0.1804,
+      "step": 14060
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.6139301999165618,
+      "learning_rate": 1.1936879974977855e-06,
+      "loss": 0.1885,
+      "step": 14061
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.2902720065610023,
+      "learning_rate": 1.1931039144276746e-06,
+      "loss": 0.1953,
+      "step": 14062
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.8926033593100717,
+      "learning_rate": 1.192519954931598e-06,
+      "loss": 0.1891,
+      "step": 14063
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.2750535230294877,
+      "learning_rate": 1.1919361190285133e-06,
+      "loss": 0.1959,
+      "step": 14064
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.7078948618825527,
+      "learning_rate": 1.1913524067373721e-06,
+      "loss": 0.196,
+      "step": 14065
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.1830081781616957,
+      "learning_rate": 1.1907688180771237e-06,
+      "loss": 0.1366,
+      "step": 14066
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.988260223707801,
+      "learning_rate": 1.1901853530667078e-06,
+      "loss": 0.1556,
+      "step": 14067
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.2264260623151466,
+      "learning_rate": 1.1896020117250668e-06,
+      "loss": 0.176,
+      "step": 14068
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.5937246900306516,
+      "learning_rate": 1.189018794071135e-06,
+      "loss": 0.1612,
+      "step": 14069
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 2.3764276605459616,
+      "learning_rate": 1.1884357001238466e-06,
+      "loss": 0.2011,
+      "step": 14070
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.4555564810930508,
+      "learning_rate": 1.187852729902126e-06,
+      "loss": 0.1303,
+      "step": 14071
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.5180283201488456,
+      "learning_rate": 1.1872698834248969e-06,
+      "loss": 0.1463,
+      "step": 14072
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.291831547906522,
+      "learning_rate": 1.1866871607110808e-06,
+      "loss": 0.1448,
+      "step": 14073
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.2863291342204537,
+      "learning_rate": 1.1861045617795919e-06,
+      "loss": 0.1694,
+      "step": 14074
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.2507114025153965,
+      "learning_rate": 1.1855220866493433e-06,
+      "loss": 0.1869,
+      "step": 14075
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.325543798448818,
+      "learning_rate": 1.18493973533924e-06,
+      "loss": 0.1232,
+      "step": 14076
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.127233103545345,
+      "learning_rate": 1.184357507868188e-06,
+      "loss": 0.1823,
+      "step": 14077
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.509886767621264,
+      "learning_rate": 1.1837754042550841e-06,
+      "loss": 0.1622,
+      "step": 14078
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.409545792721455,
+      "learning_rate": 1.1831934245188253e-06,
+      "loss": 0.1436,
+      "step": 14079
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.4663767003296748,
+      "learning_rate": 1.1826115686783041e-06,
+      "loss": 0.2076,
+      "step": 14080
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.5421093732529578,
+      "learning_rate": 1.1820298367524053e-06,
+      "loss": 0.1504,
+      "step": 14081
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.1647666107363928,
+      "learning_rate": 1.1814482287600133e-06,
+      "loss": 0.1647,
+      "step": 14082
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.2847694464186252,
+      "learning_rate": 1.1808667447200073e-06,
+      "loss": 0.1691,
+      "step": 14083
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.8328758480835186,
+      "learning_rate": 1.1802853846512647e-06,
+      "loss": 0.2076,
+      "step": 14084
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.7212825323561223,
+      "learning_rate": 1.1797041485726535e-06,
+      "loss": 0.1874,
+      "step": 14085
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.5723647234470404,
+      "learning_rate": 1.179123036503042e-06,
+      "loss": 0.1191,
+      "step": 14086
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.6784744761898194,
+      "learning_rate": 1.1785420484612947e-06,
+      "loss": 0.2185,
+      "step": 14087
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.3543927871986627,
+      "learning_rate": 1.1779611844662709e-06,
+      "loss": 0.175,
+      "step": 14088
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.2449085315076622,
+      "learning_rate": 1.1773804445368236e-06,
+      "loss": 0.1801,
+      "step": 14089
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.6407639353824566,
+      "learning_rate": 1.1767998286918054e-06,
+      "loss": 0.1906,
+      "step": 14090
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.3200270877550446,
+      "learning_rate": 1.176219336950063e-06,
+      "loss": 0.1637,
+      "step": 14091
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.4182087351008714,
+      "learning_rate": 1.1756389693304399e-06,
+      "loss": 0.1759,
+      "step": 14092
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.4342080289192918,
+      "learning_rate": 1.1750587258517764e-06,
+      "loss": 0.1785,
+      "step": 14093
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.157728300495044,
+      "learning_rate": 1.1744786065329045e-06,
+      "loss": 0.1615,
+      "step": 14094
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.789079312074786,
+      "learning_rate": 1.1738986113926587e-06,
+      "loss": 0.206,
+      "step": 14095
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.631827472548385,
+      "learning_rate": 1.1733187404498624e-06,
+      "loss": 0.1311,
+      "step": 14096
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.3552752749744916,
+      "learning_rate": 1.172738993723342e-06,
+      "loss": 0.1461,
+      "step": 14097
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.4516353919428302,
+      "learning_rate": 1.1721593712319128e-06,
+      "loss": 0.1557,
+      "step": 14098
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.8826514662661772,
+      "learning_rate": 1.1715798729943923e-06,
+      "loss": 0.1212,
+      "step": 14099
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.438619623366897,
+      "learning_rate": 1.1710004990295897e-06,
+      "loss": 0.1669,
+      "step": 14100
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.1734667648648505,
+      "learning_rate": 1.1704212493563132e-06,
+      "loss": 0.1693,
+      "step": 14101
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 2.171782440161728,
+      "learning_rate": 1.1698421239933672e-06,
+      "loss": 0.1763,
+      "step": 14102
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.4048532755183762,
+      "learning_rate": 1.1692631229595458e-06,
+      "loss": 0.1907,
+      "step": 14103
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.2046245048864728,
+      "learning_rate": 1.1686842462736469e-06,
+      "loss": 0.1666,
+      "step": 14104
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.8926093509330195,
+      "learning_rate": 1.1681054939544607e-06,
+      "loss": 0.197,
+      "step": 14105
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.6889341053256823,
+      "learning_rate": 1.1675268660207744e-06,
+      "loss": 0.1405,
+      "step": 14106
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.482180393538951,
+      "learning_rate": 1.1669483624913692e-06,
+      "loss": 0.2064,
+      "step": 14107
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.5729693952358077,
+      "learning_rate": 1.166369983385024e-06,
+      "loss": 0.1492,
+      "step": 14108
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.1741652583757263,
+      "learning_rate": 1.1657917287205133e-06,
+      "loss": 0.179,
+      "step": 14109
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.3470661474658259,
+      "learning_rate": 1.1652135985166084e-06,
+      "loss": 0.1576,
+      "step": 14110
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 2.023653070462382,
+      "learning_rate": 1.1646355927920762e-06,
+      "loss": 0.1559,
+      "step": 14111
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.805306910446254,
+      "learning_rate": 1.164057711565677e-06,
+      "loss": 0.1592,
+      "step": 14112
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.3865349607503328,
+      "learning_rate": 1.163479954856172e-06,
+      "loss": 0.1883,
+      "step": 14113
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.2352919432234593,
+      "learning_rate": 1.1629023226823122e-06,
+      "loss": 0.1427,
+      "step": 14114
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.0742840764382795,
+      "learning_rate": 1.1623248150628507e-06,
+      "loss": 0.125,
+      "step": 14115
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 2.5973455831285674,
+      "learning_rate": 1.1617474320165313e-06,
+      "loss": 0.1798,
+      "step": 14116
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 2.181296660407877,
+      "learning_rate": 1.1611701735620973e-06,
+      "loss": 0.196,
+      "step": 14117
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.9099831609957314,
+      "learning_rate": 1.160593039718287e-06,
+      "loss": 0.1956,
+      "step": 14118
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.7304536633662684,
+      "learning_rate": 1.1600160305038343e-06,
+      "loss": 0.1883,
+      "step": 14119
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.276163877421839,
+      "learning_rate": 1.159439145937471e-06,
+      "loss": 0.1957,
+      "step": 14120
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.8429107048851878,
+      "learning_rate": 1.15886238603792e-06,
+      "loss": 0.1645,
+      "step": 14121
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.143715636200938,
+      "learning_rate": 1.1582857508239049e-06,
+      "loss": 0.1678,
+      "step": 14122
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.9656996635484465,
+      "learning_rate": 1.1577092403141431e-06,
+      "loss": 0.2293,
+      "step": 14123
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.4930409592154033,
+      "learning_rate": 1.1571328545273507e-06,
+      "loss": 0.1347,
+      "step": 14124
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.5879289967310042,
+      "learning_rate": 1.1565565934822343e-06,
+      "loss": 0.186,
+      "step": 14125
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.374338947502431,
+      "learning_rate": 1.155980457197501e-06,
+      "loss": 0.1735,
+      "step": 14126
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.5965276009924254,
+      "learning_rate": 1.1554044456918518e-06,
+      "loss": 0.1771,
+      "step": 14127
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.5243224972406424,
+      "learning_rate": 1.1548285589839859e-06,
+      "loss": 0.2055,
+      "step": 14128
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.790861690982206,
+      "learning_rate": 1.154252797092597e-06,
+      "loss": 0.1983,
+      "step": 14129
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.3326069099737765,
+      "learning_rate": 1.1536771600363727e-06,
+      "loss": 0.185,
+      "step": 14130
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.582859302969802,
+      "learning_rate": 1.1531016478340007e-06,
+      "loss": 0.2295,
+      "step": 14131
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.6185810746603346,
+      "learning_rate": 1.1525262605041603e-06,
+      "loss": 0.2261,
+      "step": 14132
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.6082293497411575,
+      "learning_rate": 1.1519509980655308e-06,
+      "loss": 0.1923,
+      "step": 14133
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.0186784948810739,
+      "learning_rate": 1.1513758605367836e-06,
+      "loss": 0.153,
+      "step": 14134
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.2410975928969272,
+      "learning_rate": 1.1508008479365896e-06,
+      "loss": 0.1387,
+      "step": 14135
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.4989796635203836,
+      "learning_rate": 1.150225960283613e-06,
+      "loss": 0.1667,
+      "step": 14136
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.9016477459157948,
+      "learning_rate": 1.1496511975965152e-06,
+      "loss": 0.2292,
+      "step": 14137
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.3235935603323918,
+      "learning_rate": 1.1490765598939557e-06,
+      "loss": 0.1616,
+      "step": 14138
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.479613109720674,
+      "learning_rate": 1.1485020471945834e-06,
+      "loss": 0.1778,
+      "step": 14139
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 2.163480473826758,
+      "learning_rate": 1.1479276595170502e-06,
+      "loss": 0.1837,
+      "step": 14140
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 2.337866621284623,
+      "learning_rate": 1.1473533968799994e-06,
+      "loss": 0.1778,
+      "step": 14141
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.6246895952132345,
+      "learning_rate": 1.1467792593020748e-06,
+      "loss": 0.1831,
+      "step": 14142
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.4192472216773981,
+      "learning_rate": 1.1462052468019097e-06,
+      "loss": 0.1833,
+      "step": 14143
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 2.188355700837831,
+      "learning_rate": 1.1456313593981383e-06,
+      "loss": 0.213,
+      "step": 14144
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.44721073556203,
+      "learning_rate": 1.1450575971093897e-06,
+      "loss": 0.1673,
+      "step": 14145
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.2382635061153995,
+      "learning_rate": 1.1444839599542878e-06,
+      "loss": 0.1635,
+      "step": 14146
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.413509782498543,
+      "learning_rate": 1.143910447951455e-06,
+      "loss": 0.1887,
+      "step": 14147
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.257261345041289,
+      "learning_rate": 1.1433370611195055e-06,
+      "loss": 0.1615,
+      "step": 14148
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.348451982068989,
+      "learning_rate": 1.142763799477054e-06,
+      "loss": 0.1749,
+      "step": 14149
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.5602062770328051,
+      "learning_rate": 1.142190663042706e-06,
+      "loss": 0.1706,
+      "step": 14150
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.532342336254726,
+      "learning_rate": 1.1416176518350692e-06,
+      "loss": 0.165,
+      "step": 14151
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.1563512461384036,
+      "learning_rate": 1.1410447658727402e-06,
+      "loss": 0.186,
+      "step": 14152
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.1991673157189613,
+      "learning_rate": 1.1404720051743178e-06,
+      "loss": 0.1412,
+      "step": 14153
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.497027631077813,
+      "learning_rate": 1.1398993697583933e-06,
+      "loss": 0.2331,
+      "step": 14154
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.469455488686661,
+      "learning_rate": 1.1393268596435546e-06,
+      "loss": 0.1704,
+      "step": 14155
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.3312392691058859,
+      "learning_rate": 1.138754474848388e-06,
+      "loss": 0.192,
+      "step": 14156
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.5668535664054482,
+      "learning_rate": 1.1381822153914695e-06,
+      "loss": 0.1732,
+      "step": 14157
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.378028636275123,
+      "learning_rate": 1.1376100812913776e-06,
+      "loss": 0.1478,
+      "step": 14158
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 2.000172250005885,
+      "learning_rate": 1.1370380725666836e-06,
+      "loss": 0.2294,
+      "step": 14159
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.2883031342395959,
+      "learning_rate": 1.1364661892359557e-06,
+      "loss": 0.2194,
+      "step": 14160
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 2.0994993385046805,
+      "learning_rate": 1.1358944313177566e-06,
+      "loss": 0.1913,
+      "step": 14161
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.109858273294733,
+      "learning_rate": 1.1353227988306453e-06,
+      "loss": 0.153,
+      "step": 14162
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.3619146520191285,
+      "learning_rate": 1.1347512917931792e-06,
+      "loss": 0.2075,
+      "step": 14163
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.4400357706706737,
+      "learning_rate": 1.1341799102239081e-06,
+      "loss": 0.191,
+      "step": 14164
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.2472848610569935,
+      "learning_rate": 1.1336086541413822e-06,
+      "loss": 0.1821,
+      "step": 14165
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.140745525655012,
+      "learning_rate": 1.1330375235641405e-06,
+      "loss": 0.2034,
+      "step": 14166
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.984912487156855,
+      "learning_rate": 1.1324665185107264e-06,
+      "loss": 0.1442,
+      "step": 14167
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.2737611932608985,
+      "learning_rate": 1.1318956389996716e-06,
+      "loss": 0.1826,
+      "step": 14168
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.5857616853272922,
+      "learning_rate": 1.1313248850495102e-06,
+      "loss": 0.1958,
+      "step": 14169
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.13723850493325,
+      "learning_rate": 1.130754256678766e-06,
+      "loss": 0.1886,
+      "step": 14170
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.6165735396947163,
+      "learning_rate": 1.130183753905964e-06,
+      "loss": 0.162,
+      "step": 14171
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.111897069847086,
+      "learning_rate": 1.1296133767496227e-06,
+      "loss": 0.1957,
+      "step": 14172
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.0926250769544557,
+      "learning_rate": 1.1290431252282563e-06,
+      "loss": 0.1573,
+      "step": 14173
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.6847277095074074,
+      "learning_rate": 1.1284729993603777e-06,
+      "loss": 0.1546,
+      "step": 14174
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.4830116460824223,
+      "learning_rate": 1.1279029991644902e-06,
+      "loss": 0.1562,
+      "step": 14175
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 2.106977145384722,
+      "learning_rate": 1.1273331246590984e-06,
+      "loss": 0.2261,
+      "step": 14176
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.0865140249011584,
+      "learning_rate": 1.1267633758627e-06,
+      "loss": 0.1831,
+      "step": 14177
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.521076064357207,
+      "learning_rate": 1.12619375279379e-06,
+      "loss": 0.2237,
+      "step": 14178
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.554191845808359,
+      "learning_rate": 1.1256242554708591e-06,
+      "loss": 0.1739,
+      "step": 14179
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.3620457830481918,
+      "learning_rate": 1.125054883912392e-06,
+      "loss": 0.2215,
+      "step": 14180
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.6574512390081684,
+      "learning_rate": 1.1244856381368712e-06,
+      "loss": 0.2034,
+      "step": 14181
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.426283865350236,
+      "learning_rate": 1.123916518162776e-06,
+      "loss": 0.204,
+      "step": 14182
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.6229022130294004,
+      "learning_rate": 1.12334752400858e-06,
+      "loss": 0.1885,
+      "step": 14183
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.3303487225171875,
+      "learning_rate": 1.1227786556927517e-06,
+      "loss": 0.1453,
+      "step": 14184
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.3774659358466879,
+      "learning_rate": 1.1222099132337587e-06,
+      "loss": 0.1656,
+      "step": 14185
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.1172379368954308,
+      "learning_rate": 1.1216412966500612e-06,
+      "loss": 0.1529,
+      "step": 14186
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.3731593741158736,
+      "learning_rate": 1.1210728059601168e-06,
+      "loss": 0.1826,
+      "step": 14187
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.7431937591844113,
+      "learning_rate": 1.1205044411823813e-06,
+      "loss": 0.166,
+      "step": 14188
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.7675475544683252,
+      "learning_rate": 1.119936202335301e-06,
+      "loss": 0.1652,
+      "step": 14189
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.4340926869054644,
+      "learning_rate": 1.1193680894373227e-06,
+      "loss": 0.1815,
+      "step": 14190
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.850870204219349,
+      "learning_rate": 1.118800102506888e-06,
+      "loss": 0.2427,
+      "step": 14191
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.5647183311355028,
+      "learning_rate": 1.1182322415624353e-06,
+      "loss": 0.163,
+      "step": 14192
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.150845902252526,
+      "learning_rate": 1.1176645066223946e-06,
+      "loss": 0.1326,
+      "step": 14193
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.4114044403213222,
+      "learning_rate": 1.1170968977051966e-06,
+      "loss": 0.1734,
+      "step": 14194
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.6583712012992948,
+      "learning_rate": 1.116529414829266e-06,
+      "loss": 0.1666,
+      "step": 14195
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.5117365370302827,
+      "learning_rate": 1.1159620580130238e-06,
+      "loss": 0.1939,
+      "step": 14196
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.4160956471950394,
+      "learning_rate": 1.1153948272748877e-06,
+      "loss": 0.1584,
+      "step": 14197
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.2340102954688892,
+      "learning_rate": 1.1148277226332682e-06,
+      "loss": 0.1914,
+      "step": 14198
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.6369549073345369,
+      "learning_rate": 1.114260744106575e-06,
+      "loss": 0.1718,
+      "step": 14199
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.7958812204586323,
+      "learning_rate": 1.1136938917132127e-06,
+      "loss": 0.1969,
+      "step": 14200
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.2260038476435537,
+      "learning_rate": 1.113127165471583e-06,
+      "loss": 0.1798,
+      "step": 14201
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.4334831221093853,
+      "learning_rate": 1.1125605654000788e-06,
+      "loss": 0.2258,
+      "step": 14202
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.9212890289948819,
+      "learning_rate": 1.1119940915170958e-06,
+      "loss": 0.1975,
+      "step": 14203
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.5705414790129262,
+      "learning_rate": 1.1114277438410186e-06,
+      "loss": 0.2208,
+      "step": 14204
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.2509176819618284,
+      "learning_rate": 1.110861522390233e-06,
+      "loss": 0.1889,
+      "step": 14205
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.682253496624752,
+      "learning_rate": 1.11029542718312e-06,
+      "loss": 0.1906,
+      "step": 14206
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.1650923902996166,
+      "learning_rate": 1.1097294582380536e-06,
+      "loss": 0.185,
+      "step": 14207
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.097871073814303,
+      "learning_rate": 1.1091636155734053e-06,
+      "loss": 0.1831,
+      "step": 14208
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.407443618182957,
+      "learning_rate": 1.1085978992075436e-06,
+      "loss": 0.1784,
+      "step": 14209
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.342867125970736,
+      "learning_rate": 1.1080323091588335e-06,
+      "loss": 0.2065,
+      "step": 14210
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.5424471685830765,
+      "learning_rate": 1.1074668454456305e-06,
+      "loss": 0.1911,
+      "step": 14211
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.5099236182368907,
+      "learning_rate": 1.1069015080862927e-06,
+      "loss": 0.2115,
+      "step": 14212
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.6307195363653337,
+      "learning_rate": 1.106336297099171e-06,
+      "loss": 0.166,
+      "step": 14213
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.9429315143939303,
+      "learning_rate": 1.1057712125026116e-06,
+      "loss": 0.2312,
+      "step": 14214
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.6237180403191718,
+      "learning_rate": 1.1052062543149595e-06,
+      "loss": 0.1854,
+      "step": 14215
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.466683191600097,
+      "learning_rate": 1.1046414225545505e-06,
+      "loss": 0.1653,
+      "step": 14216
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.9500352542384125,
+      "learning_rate": 1.104076717239721e-06,
+      "loss": 0.152,
+      "step": 14217
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.2397086944287656,
+      "learning_rate": 1.1035121383888013e-06,
+      "loss": 0.189,
+      "step": 14218
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.265979099188158,
+      "learning_rate": 1.1029476860201198e-06,
+      "loss": 0.2032,
+      "step": 14219
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.228372068159425,
+      "learning_rate": 1.1023833601519963e-06,
+      "loss": 0.1584,
+      "step": 14220
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.484248346647473,
+      "learning_rate": 1.1018191608027511e-06,
+      "loss": 0.2019,
+      "step": 14221
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.5613205511326032,
+      "learning_rate": 1.101255087990697e-06,
+      "loss": 0.2187,
+      "step": 14222
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.5035915937331934,
+      "learning_rate": 1.1006911417341436e-06,
+      "loss": 0.2013,
+      "step": 14223
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.3043640786606188,
+      "learning_rate": 1.1001273220514004e-06,
+      "loss": 0.1727,
+      "step": 14224
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.209461161282016,
+      "learning_rate": 1.0995636289607647e-06,
+      "loss": 0.201,
+      "step": 14225
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.629504521595422,
+      "learning_rate": 1.0990000624805368e-06,
+      "loss": 0.2009,
+      "step": 14226
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.4024923771805755,
+      "learning_rate": 1.0984366226290105e-06,
+      "loss": 0.1708,
+      "step": 14227
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.3591922823561804,
+      "learning_rate": 1.097873309424476e-06,
+      "loss": 0.1918,
+      "step": 14228
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.0848487088643508,
+      "learning_rate": 1.0973101228852163e-06,
+      "loss": 0.1754,
+      "step": 14229
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.018189160747273,
+      "learning_rate": 1.0967470630295146e-06,
+      "loss": 0.1914,
+      "step": 14230
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.0972284793507683,
+      "learning_rate": 1.0961841298756475e-06,
+      "loss": 0.1786,
+      "step": 14231
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.5074894267434191,
+      "learning_rate": 1.0956213234418883e-06,
+      "loss": 0.2115,
+      "step": 14232
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.383026214291865,
+      "learning_rate": 1.0950586437465076e-06,
+      "loss": 0.242,
+      "step": 14233
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.530453063413215,
+      "learning_rate": 1.0944960908077673e-06,
+      "loss": 0.1878,
+      "step": 14234
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.3971197510190918,
+      "learning_rate": 1.0939336646439303e-06,
+      "loss": 0.1963,
+      "step": 14235
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.2328312914713107,
+      "learning_rate": 1.093371365273252e-06,
+      "loss": 0.169,
+      "step": 14236
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.288566204825254,
+      "learning_rate": 1.0928091927139872e-06,
+      "loss": 0.1907,
+      "step": 14237
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.7522044517196458,
+      "learning_rate": 1.0922471469843815e-06,
+      "loss": 0.2453,
+      "step": 14238
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.875619563419302,
+      "learning_rate": 1.0916852281026818e-06,
+      "loss": 0.1798,
+      "step": 14239
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.6438249353363636,
+      "learning_rate": 1.091123436087126e-06,
+      "loss": 0.1792,
+      "step": 14240
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.2723514629824964,
+      "learning_rate": 1.0905617709559512e-06,
+      "loss": 0.2102,
+      "step": 14241
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.4532478675375478,
+      "learning_rate": 1.0900002327273905e-06,
+      "loss": 0.1524,
+      "step": 14242
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.5296888275452445,
+      "learning_rate": 1.0894388214196695e-06,
+      "loss": 0.2048,
+      "step": 14243
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.2754377032203303,
+      "learning_rate": 1.088877537051013e-06,
+      "loss": 0.1693,
+      "step": 14244
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.208786314977968,
+      "learning_rate": 1.0883163796396406e-06,
+      "loss": 0.172,
+      "step": 14245
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.3022009030351054,
+      "learning_rate": 1.087755349203769e-06,
+      "loss": 0.1617,
+      "step": 14246
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.0496835186250526,
+      "learning_rate": 1.087194445761608e-06,
+      "loss": 0.138,
+      "step": 14247
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.1848298533028516,
+      "learning_rate": 1.0866336693313645e-06,
+      "loss": 0.1799,
+      "step": 14248
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.0463739875876965,
+      "learning_rate": 1.0860730199312426e-06,
+      "loss": 0.1422,
+      "step": 14249
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.4174495723877525,
+      "learning_rate": 1.0855124975794412e-06,
+      "loss": 0.225,
+      "step": 14250
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.6328409365865402,
+      "learning_rate": 1.0849521022941566e-06,
+      "loss": 0.2055,
+      "step": 14251
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.2182108040912973,
+      "learning_rate": 1.0843918340935766e-06,
+      "loss": 0.1817,
+      "step": 14252
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.5243361365802033,
+      "learning_rate": 1.0838316929958891e-06,
+      "loss": 0.2144,
+      "step": 14253
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.1247082305422198,
+      "learning_rate": 1.0832716790192771e-06,
+      "loss": 0.1665,
+      "step": 14254
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.0612277274439945,
+      "learning_rate": 1.0827117921819203e-06,
+      "loss": 0.1636,
+      "step": 14255
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 2.3990047231427347,
+      "learning_rate": 1.0821520325019897e-06,
+      "loss": 0.2019,
+      "step": 14256
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.477007839114604,
+      "learning_rate": 1.081592399997658e-06,
+      "loss": 0.1494,
+      "step": 14257
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.466567020210146,
+      "learning_rate": 1.081032894687089e-06,
+      "loss": 0.195,
+      "step": 14258
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.320971538907418,
+      "learning_rate": 1.0804735165884462e-06,
+      "loss": 0.1637,
+      "step": 14259
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.8295940198563052,
+      "learning_rate": 1.0799142657198886e-06,
+      "loss": 0.1367,
+      "step": 14260
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.382568876713328,
+      "learning_rate": 1.0793551420995663e-06,
+      "loss": 0.218,
+      "step": 14261
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.9535504548471632,
+      "learning_rate": 1.0787961457456302e-06,
+      "loss": 0.1275,
+      "step": 14262
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.1282447374947229,
+      "learning_rate": 1.0782372766762267e-06,
+      "loss": 0.1722,
+      "step": 14263
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.104045868809608,
+      "learning_rate": 1.0776785349094975e-06,
+      "loss": 0.1524,
+      "step": 14264
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.9842147696742052,
+      "learning_rate": 1.0771199204635774e-06,
+      "loss": 0.1416,
+      "step": 14265
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.5163707864614921,
+      "learning_rate": 1.0765614333566e-06,
+      "loss": 0.1703,
+      "step": 14266
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.4223173679493626,
+      "learning_rate": 1.0760030736066952e-06,
+      "loss": 0.1736,
+      "step": 14267
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.7533364421743796,
+      "learning_rate": 1.0754448412319867e-06,
+      "loss": 0.2257,
+      "step": 14268
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.591107393498661,
+      "learning_rate": 1.074886736250597e-06,
+      "loss": 0.1961,
+      "step": 14269
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.2538785366905576,
+      "learning_rate": 1.0743287586806394e-06,
+      "loss": 0.2041,
+      "step": 14270
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 2.1130453461079353,
+      "learning_rate": 1.0737709085402276e-06,
+      "loss": 0.2442,
+      "step": 14271
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.1374969014712462,
+      "learning_rate": 1.0732131858474705e-06,
+      "loss": 0.2023,
+      "step": 14272
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.082336100521952,
+      "learning_rate": 1.0726555906204721e-06,
+      "loss": 0.1913,
+      "step": 14273
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.7943126321103133,
+      "learning_rate": 1.072098122877331e-06,
+      "loss": 0.2078,
+      "step": 14274
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.0205376981164793,
+      "learning_rate": 1.071540782636144e-06,
+      "loss": 0.1062,
+      "step": 14275
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.282543878157909,
+      "learning_rate": 1.0709835699150012e-06,
+      "loss": 0.1952,
+      "step": 14276
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.9792598703679211,
+      "learning_rate": 1.0704264847319911e-06,
+      "loss": 0.1646,
+      "step": 14277
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.180893023103314,
+      "learning_rate": 1.0698695271051983e-06,
+      "loss": 0.1987,
+      "step": 14278
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.1040780516219384,
+      "learning_rate": 1.0693126970526995e-06,
+      "loss": 0.1673,
+      "step": 14279
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.3334390634349225,
+      "learning_rate": 1.0687559945925708e-06,
+      "loss": 0.1597,
+      "step": 14280
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.908404440768739,
+      "learning_rate": 1.068199419742883e-06,
+      "loss": 0.1621,
+      "step": 14281
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.1907941590437707,
+      "learning_rate": 1.0676429725217047e-06,
+      "loss": 0.1848,
+      "step": 14282
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.374854863157986,
+      "learning_rate": 1.0670866529470953e-06,
+      "loss": 0.1648,
+      "step": 14283
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.3667571507666854,
+      "learning_rate": 1.0665304610371142e-06,
+      "loss": 0.13,
+      "step": 14284
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.1609419849560696,
+      "learning_rate": 1.0659743968098168e-06,
+      "loss": 0.1397,
+      "step": 14285
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.2140293887617346,
+      "learning_rate": 1.0654184602832523e-06,
+      "loss": 0.1651,
+      "step": 14286
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.1171652634948597,
+      "learning_rate": 1.064862651475469e-06,
+      "loss": 0.1669,
+      "step": 14287
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.250769330741708,
+      "learning_rate": 1.064306970404505e-06,
+      "loss": 0.1349,
+      "step": 14288
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.2830830231873664,
+      "learning_rate": 1.0637514170884e-06,
+      "loss": 0.1616,
+      "step": 14289
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.2392655080040813,
+      "learning_rate": 1.0631959915451879e-06,
+      "loss": 0.1678,
+      "step": 14290
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.2840295333975236,
+      "learning_rate": 1.062640693792899e-06,
+      "loss": 0.1779,
+      "step": 14291
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.710829783667598,
+      "learning_rate": 1.062085523849556e-06,
+      "loss": 0.2409,
+      "step": 14292
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.6139416809137195,
+      "learning_rate": 1.0615304817331822e-06,
+      "loss": 0.1929,
+      "step": 14293
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.11904610572573,
+      "learning_rate": 1.0609755674617927e-06,
+      "loss": 0.1232,
+      "step": 14294
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.3953235256276335,
+      "learning_rate": 1.0604207810534011e-06,
+      "loss": 0.1892,
+      "step": 14295
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.326041494710611,
+      "learning_rate": 1.0598661225260182e-06,
+      "loss": 0.1631,
+      "step": 14296
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.137889495126181,
+      "learning_rate": 1.0593115918976455e-06,
+      "loss": 0.1817,
+      "step": 14297
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 2.0839498029020267,
+      "learning_rate": 1.058757189186284e-06,
+      "loss": 0.1877,
+      "step": 14298
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.3397852298201525,
+      "learning_rate": 1.058202914409931e-06,
+      "loss": 0.1892,
+      "step": 14299
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.4783168543209313,
+      "learning_rate": 1.0576487675865792e-06,
+      "loss": 0.1861,
+      "step": 14300
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.6412604100593766,
+      "learning_rate": 1.0570947487342143e-06,
+      "loss": 0.1546,
+      "step": 14301
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.222620006791915,
+      "learning_rate": 1.0565408578708209e-06,
+      "loss": 0.178,
+      "step": 14302
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.385927925807908,
+      "learning_rate": 1.0559870950143786e-06,
+      "loss": 0.129,
+      "step": 14303
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.1158592891713113,
+      "learning_rate": 1.0554334601828635e-06,
+      "loss": 0.2115,
+      "step": 14304
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.692824859031784,
+      "learning_rate": 1.0548799533942484e-06,
+      "loss": 0.1902,
+      "step": 14305
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.5243957904549037,
+      "learning_rate": 1.0543265746664966e-06,
+      "loss": 0.1696,
+      "step": 14306
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.6343197140359935,
+      "learning_rate": 1.0537733240175736e-06,
+      "loss": 0.1971,
+      "step": 14307
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.92939792421721,
+      "learning_rate": 1.0532202014654374e-06,
+      "loss": 0.1851,
+      "step": 14308
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.2001633604931905,
+      "learning_rate": 1.0526672070280443e-06,
+      "loss": 0.1248,
+      "step": 14309
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.9306631208850684,
+      "learning_rate": 1.0521143407233424e-06,
+      "loss": 0.2273,
+      "step": 14310
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.2336510092339419,
+      "learning_rate": 1.0515616025692803e-06,
+      "loss": 0.1988,
+      "step": 14311
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 2.0029286757029077,
+      "learning_rate": 1.051008992583798e-06,
+      "loss": 0.2438,
+      "step": 14312
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.4922864267001708,
+      "learning_rate": 1.0504565107848346e-06,
+      "loss": 0.112,
+      "step": 14313
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.1229994689223832,
+      "learning_rate": 1.0499041571903256e-06,
+      "loss": 0.2035,
+      "step": 14314
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.0135326767679902,
+      "learning_rate": 1.0493519318181977e-06,
+      "loss": 0.1487,
+      "step": 14315
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 2.062480290636622,
+      "learning_rate": 1.0487998346863775e-06,
+      "loss": 0.2127,
+      "step": 14316
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.5299772305757764,
+      "learning_rate": 1.048247865812787e-06,
+      "loss": 0.1963,
+      "step": 14317
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.194719145488182,
+      "learning_rate": 1.0476960252153446e-06,
+      "loss": 0.1592,
+      "step": 14318
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.5318473118023301,
+      "learning_rate": 1.0471443129119603e-06,
+      "loss": 0.1655,
+      "step": 14319
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.1498318401517644,
+      "learning_rate": 1.0465927289205452e-06,
+      "loss": 0.182,
+      "step": 14320
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.2026310253434465,
+      "learning_rate": 1.0460412732590031e-06,
+      "loss": 0.1405,
+      "step": 14321
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.154692873987435,
+      "learning_rate": 1.045489945945235e-06,
+      "loss": 0.1577,
+      "step": 14322
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.4534652680734266,
+      "learning_rate": 1.0449387469971383e-06,
+      "loss": 0.2055,
+      "step": 14323
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.2479531994738315,
+      "learning_rate": 1.0443876764326033e-06,
+      "loss": 0.1697,
+      "step": 14324
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.7125150645073763,
+      "learning_rate": 1.0438367342695188e-06,
+      "loss": 0.1761,
+      "step": 14325
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.3737181735842743,
+      "learning_rate": 1.0432859205257707e-06,
+      "loss": 0.1915,
+      "step": 14326
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.0496163043952949,
+      "learning_rate": 1.042735235219235e-06,
+      "loss": 0.1647,
+      "step": 14327
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.6890478629325842,
+      "learning_rate": 1.0421846783677892e-06,
+      "loss": 0.1817,
+      "step": 14328
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.318698997550853,
+      "learning_rate": 1.0416342499893056e-06,
+      "loss": 0.2229,
+      "step": 14329
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.2690373120066263,
+      "learning_rate": 1.0410839501016502e-06,
+      "loss": 0.1462,
+      "step": 14330
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.5420578501118756,
+      "learning_rate": 1.0405337787226855e-06,
+      "loss": 0.2027,
+      "step": 14331
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.4798087004960667,
+      "learning_rate": 1.0399837358702725e-06,
+      "loss": 0.2176,
+      "step": 14332
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.8115395023425351,
+      "learning_rate": 1.0394338215622636e-06,
+      "loss": 0.2073,
+      "step": 14333
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.2596201376480904,
+      "learning_rate": 1.0388840358165102e-06,
+      "loss": 0.1872,
+      "step": 14334
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.2513101804420355,
+      "learning_rate": 1.0383343786508587e-06,
+      "loss": 0.2003,
+      "step": 14335
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.2460148705655938,
+      "learning_rate": 1.0377848500831527e-06,
+      "loss": 0.1567,
+      "step": 14336
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.2536011678263916,
+      "learning_rate": 1.0372354501312281e-06,
+      "loss": 0.1734,
+      "step": 14337
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.5716572976701164,
+      "learning_rate": 1.036686178812919e-06,
+      "loss": 0.161,
+      "step": 14338
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.1658067628006286,
+      "learning_rate": 1.036137036146056e-06,
+      "loss": 0.1465,
+      "step": 14339
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.3876579864872516,
+      "learning_rate": 1.0355880221484637e-06,
+      "loss": 0.1375,
+      "step": 14340
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 2.516577522539003,
+      "learning_rate": 1.0350391368379658e-06,
+      "loss": 0.211,
+      "step": 14341
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.183041459452619,
+      "learning_rate": 1.0344903802323757e-06,
+      "loss": 0.1343,
+      "step": 14342
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.5282340857218382,
+      "learning_rate": 1.0339417523495082e-06,
+      "loss": 0.2298,
+      "step": 14343
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.0181093165733481,
+      "learning_rate": 1.0333932532071738e-06,
+      "loss": 0.1653,
+      "step": 14344
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.0843345145327465,
+      "learning_rate": 1.032844882823174e-06,
+      "loss": 0.1756,
+      "step": 14345
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.0144949384789148,
+      "learning_rate": 1.0322966412153118e-06,
+      "loss": 0.1839,
+      "step": 14346
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.877888552591604,
+      "learning_rate": 1.0317485284013806e-06,
+      "loss": 0.1889,
+      "step": 14347
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.3260119907817283,
+      "learning_rate": 1.031200544399174e-06,
+      "loss": 0.2082,
+      "step": 14348
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.5016985008978971,
+      "learning_rate": 1.0306526892264807e-06,
+      "loss": 0.1724,
+      "step": 14349
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.6030231792989618,
+      "learning_rate": 1.0301049629010845e-06,
+      "loss": 0.189,
+      "step": 14350
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.5018442955323283,
+      "learning_rate": 1.0295573654407626e-06,
+      "loss": 0.2273,
+      "step": 14351
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.446663925766912,
+      "learning_rate": 1.0290098968632917e-06,
+      "loss": 0.1562,
+      "step": 14352
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.951675609673464,
+      "learning_rate": 1.0284625571864438e-06,
+      "loss": 0.23,
+      "step": 14353
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.2019273785550828,
+      "learning_rate": 1.0279153464279855e-06,
+      "loss": 0.1789,
+      "step": 14354
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.4296477994982804,
+      "learning_rate": 1.027368264605678e-06,
+      "loss": 0.1851,
+      "step": 14355
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.7047843461538172,
+      "learning_rate": 1.0268213117372816e-06,
+      "loss": 0.1777,
+      "step": 14356
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.4646400081088267,
+      "learning_rate": 1.02627448784055e-06,
+      "loss": 0.1772,
+      "step": 14357
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.6336655190974336,
+      "learning_rate": 1.0257277929332332e-06,
+      "loss": 0.256,
+      "step": 14358
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.179972033260179,
+      "learning_rate": 1.025181227033079e-06,
+      "loss": 0.1641,
+      "step": 14359
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.5600313258884873,
+      "learning_rate": 1.024634790157827e-06,
+      "loss": 0.1508,
+      "step": 14360
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.5165387869552518,
+      "learning_rate": 1.0240884823252151e-06,
+      "loss": 0.1883,
+      "step": 14361
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.534209252940076,
+      "learning_rate": 1.023542303552979e-06,
+      "loss": 0.147,
+      "step": 14362
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.3198102300412773,
+      "learning_rate": 1.022996253858845e-06,
+      "loss": 0.1889,
+      "step": 14363
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.0963804171251899,
+      "learning_rate": 1.0224503332605413e-06,
+      "loss": 0.1797,
+      "step": 14364
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.1423361888912327,
+      "learning_rate": 1.021904541775785e-06,
+      "loss": 0.1478,
+      "step": 14365
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.2351334631867048,
+      "learning_rate": 1.0213588794222951e-06,
+      "loss": 0.1739,
+      "step": 14366
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.6674701919350685,
+      "learning_rate": 1.0208133462177838e-06,
+      "loss": 0.2267,
+      "step": 14367
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.252899692857248,
+      "learning_rate": 1.020267942179961e-06,
+      "loss": 0.1653,
+      "step": 14368
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.4933771663706574,
+      "learning_rate": 1.0197226673265281e-06,
+      "loss": 0.1769,
+      "step": 14369
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.955013514792596,
+      "learning_rate": 1.0191775216751858e-06,
+      "loss": 0.1244,
+      "step": 14370
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.5315753970549828,
+      "learning_rate": 1.0186325052436308e-06,
+      "loss": 0.1464,
+      "step": 14371
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.2736851913522573,
+      "learning_rate": 1.0180876180495553e-06,
+      "loss": 0.1334,
+      "step": 14372
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.4117089913500123,
+      "learning_rate": 1.0175428601106441e-06,
+      "loss": 0.1995,
+      "step": 14373
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.6274160319632656,
+      "learning_rate": 1.0169982314445813e-06,
+      "loss": 0.1879,
+      "step": 14374
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.4924297211065418,
+      "learning_rate": 1.0164537320690465e-06,
+      "loss": 0.2751,
+      "step": 14375
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.5377613124688683,
+      "learning_rate": 1.0159093620017146e-06,
+      "loss": 0.1869,
+      "step": 14376
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.5499223257029842,
+      "learning_rate": 1.0153651212602573e-06,
+      "loss": 0.2125,
+      "step": 14377
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.5668860388002905,
+      "learning_rate": 1.0148210098623374e-06,
+      "loss": 0.1403,
+      "step": 14378
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.7847294411985337,
+      "learning_rate": 1.0142770278256197e-06,
+      "loss": 0.2281,
+      "step": 14379
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.4726022158962386,
+      "learning_rate": 1.0137331751677632e-06,
+      "loss": 0.1944,
+      "step": 14380
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.4048953737386523,
+      "learning_rate": 1.0131894519064183e-06,
+      "loss": 0.1771,
+      "step": 14381
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.9943238072731606,
+      "learning_rate": 1.0126458580592374e-06,
+      "loss": 0.1587,
+      "step": 14382
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.6695518395988882,
+      "learning_rate": 1.012102393643864e-06,
+      "loss": 0.2032,
+      "step": 14383
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.5182132349328887,
+      "learning_rate": 1.01155905867794e-06,
+      "loss": 0.204,
+      "step": 14384
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.3606162871387144,
+      "learning_rate": 1.0110158531791025e-06,
+      "loss": 0.1928,
+      "step": 14385
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.6007412037417843,
+      "learning_rate": 1.0104727771649859e-06,
+      "loss": 0.202,
+      "step": 14386
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.5340995840508092,
+      "learning_rate": 1.0099298306532151e-06,
+      "loss": 0.161,
+      "step": 14387
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 2.085869327245526,
+      "learning_rate": 1.009387013661417e-06,
+      "loss": 0.2513,
+      "step": 14388
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.4211575799440848,
+      "learning_rate": 1.0088443262072112e-06,
+      "loss": 0.1839,
+      "step": 14389
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.7847835059803336,
+      "learning_rate": 1.0083017683082148e-06,
+      "loss": 0.1824,
+      "step": 14390
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.9903708068641411,
+      "learning_rate": 1.0077593399820368e-06,
+      "loss": 0.1617,
+      "step": 14391
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.6307534554629257,
+      "learning_rate": 1.0072170412462866e-06,
+      "loss": 0.1519,
+      "step": 14392
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.5598942916386989,
+      "learning_rate": 1.006674872118567e-06,
+      "loss": 0.1631,
+      "step": 14393
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.581507494847487,
+      "learning_rate": 1.0061328326164783e-06,
+      "loss": 0.233,
+      "step": 14394
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.607820227431105,
+      "learning_rate": 1.005590922757615e-06,
+      "loss": 0.2012,
+      "step": 14395
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.145539246735172,
+      "learning_rate": 1.005049142559566e-06,
+      "loss": 0.1737,
+      "step": 14396
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.279761228661554,
+      "learning_rate": 1.004507492039919e-06,
+      "loss": 0.1813,
+      "step": 14397
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.7617556616728842,
+      "learning_rate": 1.003965971216258e-06,
+      "loss": 0.1878,
+      "step": 14398
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.7470732401550906,
+      "learning_rate": 1.0034245801061582e-06,
+      "loss": 0.209,
+      "step": 14399
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.7833767333238522,
+      "learning_rate": 1.0028833187271958e-06,
+      "loss": 0.1733,
+      "step": 14400
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.0573319942632047,
+      "learning_rate": 1.002342187096938e-06,
+      "loss": 0.1245,
+      "step": 14401
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.9499417208689421,
+      "learning_rate": 1.0018011852329517e-06,
+      "loss": 0.1553,
+      "step": 14402
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.2910889332051398,
+      "learning_rate": 1.0012603131527982e-06,
+      "loss": 0.154,
+      "step": 14403
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.1685721423990805,
+      "learning_rate": 1.0007195708740358e-06,
+      "loss": 0.144,
+      "step": 14404
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.654046193915805,
+      "learning_rate": 1.000178958414214e-06,
+      "loss": 0.1809,
+      "step": 14405
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.5134837445895903,
+      "learning_rate": 9.996384757908838e-07,
+      "loss": 0.1353,
+      "step": 14406
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.4611119507859522,
+      "learning_rate": 9.990981230215891e-07,
+      "loss": 0.1797,
+      "step": 14407
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.335462862325151,
+      "learning_rate": 9.98557900123871e-07,
+      "loss": 0.1946,
+      "step": 14408
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.4347682164576698,
+      "learning_rate": 9.980178071152629e-07,
+      "loss": 0.1702,
+      "step": 14409
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.2568430037103469,
+      "learning_rate": 9.974778440132987e-07,
+      "loss": 0.1922,
+      "step": 14410
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.3238296162092333,
+      "learning_rate": 9.969380108355048e-07,
+      "loss": 0.197,
+      "step": 14411
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.792702977214779,
+      "learning_rate": 9.963983075994054e-07,
+      "loss": 0.2033,
+      "step": 14412
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.51619690005669,
+      "learning_rate": 9.958587343225202e-07,
+      "loss": 0.2077,
+      "step": 14413
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.4942532390964969,
+      "learning_rate": 9.95319291022362e-07,
+      "loss": 0.1558,
+      "step": 14414
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.522579509040203,
+      "learning_rate": 9.947799777164424e-07,
+      "loss": 0.1426,
+      "step": 14415
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.2564118758451492,
+      "learning_rate": 9.942407944222687e-07,
+      "loss": 0.1553,
+      "step": 14416
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.3235345496361557,
+      "learning_rate": 9.937017411573418e-07,
+      "loss": 0.1861,
+      "step": 14417
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.1155078505868603,
+      "learning_rate": 9.93162817939161e-07,
+      "loss": 0.1513,
+      "step": 14418
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 7.075643883511378,
+      "learning_rate": 9.926240247852181e-07,
+      "loss": 0.2172,
+      "step": 14419
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.3301751133627433,
+      "learning_rate": 9.920853617130033e-07,
+      "loss": 0.1717,
+      "step": 14420
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.4094986108591732,
+      "learning_rate": 9.915468287400027e-07,
+      "loss": 0.2062,
+      "step": 14421
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.2532440171836134,
+      "learning_rate": 9.910084258836983e-07,
+      "loss": 0.2209,
+      "step": 14422
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.1168972351890014,
+      "learning_rate": 9.904701531615647e-07,
+      "loss": 0.1645,
+      "step": 14423
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.9768218573443571,
+      "learning_rate": 9.899320105910754e-07,
+      "loss": 0.1604,
+      "step": 14424
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.2560369624087062,
+      "learning_rate": 9.893939981896988e-07,
+      "loss": 0.1683,
+      "step": 14425
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.392455709568883,
+      "learning_rate": 9.888561159748995e-07,
+      "loss": 0.1652,
+      "step": 14426
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.6079130546295797,
+      "learning_rate": 9.883183639641386e-07,
+      "loss": 0.1971,
+      "step": 14427
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.5824555687482418,
+      "learning_rate": 9.877807421748692e-07,
+      "loss": 0.1559,
+      "step": 14428
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.6335995616107095,
+      "learning_rate": 9.872432506245439e-07,
+      "loss": 0.2176,
+      "step": 14429
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.1001423413123868,
+      "learning_rate": 9.867058893306104e-07,
+      "loss": 0.1304,
+      "step": 14430
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.7050590667009071,
+      "learning_rate": 9.86168658310513e-07,
+      "loss": 0.192,
+      "step": 14431
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.4459372639290027,
+      "learning_rate": 9.856315575816871e-07,
+      "loss": 0.2449,
+      "step": 14432
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.3364728557643037,
+      "learning_rate": 9.850945871615697e-07,
+      "loss": 0.2162,
+      "step": 14433
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.1391324539838588,
+      "learning_rate": 9.845577470675922e-07,
+      "loss": 0.1775,
+      "step": 14434
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.409456660702522,
+      "learning_rate": 9.840210373171776e-07,
+      "loss": 0.1621,
+      "step": 14435
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.3069019968545537,
+      "learning_rate": 9.834844579277508e-07,
+      "loss": 0.2045,
+      "step": 14436
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.4030544955715054,
+      "learning_rate": 9.82948008916727e-07,
+      "loss": 0.1767,
+      "step": 14437
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.4306327894522883,
+      "learning_rate": 9.824116903015208e-07,
+      "loss": 0.1742,
+      "step": 14438
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.2347967055215623,
+      "learning_rate": 9.818755020995413e-07,
+      "loss": 0.1983,
+      "step": 14439
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.5895513973386053,
+      "learning_rate": 9.813394443281954e-07,
+      "loss": 0.1959,
+      "step": 14440
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.5153786567191379,
+      "learning_rate": 9.808035170048801e-07,
+      "loss": 0.161,
+      "step": 14441
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.4641619405288424,
+      "learning_rate": 9.80267720146994e-07,
+      "loss": 0.1565,
+      "step": 14442
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.5280915792980234,
+      "learning_rate": 9.797320537719295e-07,
+      "loss": 0.1989,
+      "step": 14443
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.3731927048144086,
+      "learning_rate": 9.791965178970742e-07,
+      "loss": 0.1895,
+      "step": 14444
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.3795401167209265,
+      "learning_rate": 9.786611125398133e-07,
+      "loss": 0.1373,
+      "step": 14445
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.3100070324738122,
+      "learning_rate": 9.781258377175235e-07,
+      "loss": 0.2032,
+      "step": 14446
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.2372003464635253,
+      "learning_rate": 9.775906934475821e-07,
+      "loss": 0.2,
+      "step": 14447
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.5224532053165685,
+      "learning_rate": 9.770556797473602e-07,
+      "loss": 0.1412,
+      "step": 14448
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.894508165041881,
+      "learning_rate": 9.76520796634225e-07,
+      "loss": 0.1739,
+      "step": 14449
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.1853445436625305,
+      "learning_rate": 9.759860441255369e-07,
+      "loss": 0.1884,
+      "step": 14450
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.3062183170948056,
+      "learning_rate": 9.754514222386564e-07,
+      "loss": 0.2032,
+      "step": 14451
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.5348748271886743,
+      "learning_rate": 9.749169309909385e-07,
+      "loss": 0.1901,
+      "step": 14452
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.9869464446945175,
+      "learning_rate": 9.743825703997295e-07,
+      "loss": 0.115,
+      "step": 14453
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.2206921752438589,
+      "learning_rate": 9.73848340482379e-07,
+      "loss": 0.2153,
+      "step": 14454
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.088583679240242,
+      "learning_rate": 9.733142412562253e-07,
+      "loss": 0.1621,
+      "step": 14455
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.3965755512910654,
+      "learning_rate": 9.727802727386066e-07,
+      "loss": 0.1978,
+      "step": 14456
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.5995448359137636,
+      "learning_rate": 9.722464349468563e-07,
+      "loss": 0.2002,
+      "step": 14457
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.5398523269484126,
+      "learning_rate": 9.717127278983041e-07,
+      "loss": 0.1862,
+      "step": 14458
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 2.154626310999755,
+      "learning_rate": 9.711791516102713e-07,
+      "loss": 0.2583,
+      "step": 14459
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.6788678764406286,
+      "learning_rate": 9.706457061000806e-07,
+      "loss": 0.2147,
+      "step": 14460
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.738534051750089,
+      "learning_rate": 9.701123913850474e-07,
+      "loss": 0.1726,
+      "step": 14461
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.5575861769312767,
+      "learning_rate": 9.695792074824827e-07,
+      "loss": 0.231,
+      "step": 14462
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.3284464783635443,
+      "learning_rate": 9.690461544096968e-07,
+      "loss": 0.1372,
+      "step": 14463
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.6557379911854953,
+      "learning_rate": 9.685132321839885e-07,
+      "loss": 0.1847,
+      "step": 14464
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.9022992655665991,
+      "learning_rate": 9.679804408226594e-07,
+      "loss": 0.1666,
+      "step": 14465
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.0007010811195107,
+      "learning_rate": 9.674477803430038e-07,
+      "loss": 0.1493,
+      "step": 14466
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.2270179012460947,
+      "learning_rate": 9.66915250762313e-07,
+      "loss": 0.1676,
+      "step": 14467
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.3300693192379558,
+      "learning_rate": 9.66382852097872e-07,
+      "loss": 0.2186,
+      "step": 14468
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.6216772660865977,
+      "learning_rate": 9.658505843669625e-07,
+      "loss": 0.1448,
+      "step": 14469
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.2741580691410324,
+      "learning_rate": 9.653184475868638e-07,
+      "loss": 0.1645,
+      "step": 14470
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.2947091723356194,
+      "learning_rate": 9.647864417748476e-07,
+      "loss": 0.1771,
+      "step": 14471
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 3.3433364897688027,
+      "learning_rate": 9.642545669481852e-07,
+      "loss": 0.2021,
+      "step": 14472
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.440253559271373,
+      "learning_rate": 9.637228231241385e-07,
+      "loss": 0.2078,
+      "step": 14473
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.3438710280185746,
+      "learning_rate": 9.631912103199703e-07,
+      "loss": 0.2423,
+      "step": 14474
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.0215213251086581,
+      "learning_rate": 9.626597285529371e-07,
+      "loss": 0.2179,
+      "step": 14475
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.1291105158183126,
+      "learning_rate": 9.621283778402912e-07,
+      "loss": 0.148,
+      "step": 14476
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.193262383473067,
+      "learning_rate": 9.615971581992794e-07,
+      "loss": 0.1809,
+      "step": 14477
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.1792057203030686,
+      "learning_rate": 9.610660696471464e-07,
+      "loss": 0.1783,
+      "step": 14478
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.9865293127657969,
+      "learning_rate": 9.605351122011308e-07,
+      "loss": 0.1679,
+      "step": 14479
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.1352114991501936,
+      "learning_rate": 9.600042858784681e-07,
+      "loss": 0.2107,
+      "step": 14480
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.4870821359883788,
+      "learning_rate": 9.594735906963915e-07,
+      "loss": 0.1691,
+      "step": 14481
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.4109768962851792,
+      "learning_rate": 9.589430266721239e-07,
+      "loss": 0.1567,
+      "step": 14482
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.4752925124104246,
+      "learning_rate": 9.584125938228894e-07,
+      "loss": 0.222,
+      "step": 14483
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.3843616601202227,
+      "learning_rate": 9.57882292165906e-07,
+      "loss": 0.2083,
+      "step": 14484
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.9693633241491372,
+      "learning_rate": 9.573521217183896e-07,
+      "loss": 0.159,
+      "step": 14485
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.3073318774155078,
+      "learning_rate": 9.56822082497546e-07,
+      "loss": 0.1617,
+      "step": 14486
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.0110948141001344,
+      "learning_rate": 9.562921745205833e-07,
+      "loss": 0.1217,
+      "step": 14487
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.2515959207784988,
+      "learning_rate": 9.557623978047026e-07,
+      "loss": 0.1769,
+      "step": 14488
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.2323981402207926,
+      "learning_rate": 9.55232752367099e-07,
+      "loss": 0.181,
+      "step": 14489
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.516486935480971,
+      "learning_rate": 9.547032382249678e-07,
+      "loss": 0.1683,
+      "step": 14490
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 2.03456152428568,
+      "learning_rate": 9.541738553954943e-07,
+      "loss": 0.173,
+      "step": 14491
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.8162290014928946,
+      "learning_rate": 9.536446038958635e-07,
+      "loss": 0.187,
+      "step": 14492
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.2045204690007136,
+      "learning_rate": 9.53115483743256e-07,
+      "loss": 0.1298,
+      "step": 14493
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.6112389279733115,
+      "learning_rate": 9.525864949548485e-07,
+      "loss": 0.2336,
+      "step": 14494
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.1573301309201676,
+      "learning_rate": 9.520576375478092e-07,
+      "loss": 0.1372,
+      "step": 14495
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.3403906220517638,
+      "learning_rate": 9.515289115393072e-07,
+      "loss": 0.2021,
+      "step": 14496
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.7489072610938576,
+      "learning_rate": 9.510003169465043e-07,
+      "loss": 0.2266,
+      "step": 14497
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.4134396345673592,
+      "learning_rate": 9.504718537865593e-07,
+      "loss": 0.1913,
+      "step": 14498
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.9614723624380002,
+      "learning_rate": 9.499435220766284e-07,
+      "loss": 0.1327,
+      "step": 14499
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.8310747557494018,
+      "learning_rate": 9.49415321833858e-07,
+      "loss": 0.1932,
+      "step": 14500
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.571204297592973,
+      "learning_rate": 9.488872530753956e-07,
+      "loss": 0.2246,
+      "step": 14501
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.7192273474109705,
+      "learning_rate": 9.483593158183829e-07,
+      "loss": 0.1815,
+      "step": 14502
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.5608886039032914,
+      "learning_rate": 9.478315100799574e-07,
+      "loss": 0.2276,
+      "step": 14503
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.1392029559241745,
+      "learning_rate": 9.473038358772502e-07,
+      "loss": 0.1571,
+      "step": 14504
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.3864727930432228,
+      "learning_rate": 9.46776293227391e-07,
+      "loss": 0.2119,
+      "step": 14505
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.3218489346741287,
+      "learning_rate": 9.462488821475052e-07,
+      "loss": 0.1556,
+      "step": 14506
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.2754372534180347,
+      "learning_rate": 9.457216026547105e-07,
+      "loss": 0.1316,
+      "step": 14507
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.0986449719921723,
+      "learning_rate": 9.451944547661257e-07,
+      "loss": 0.1974,
+      "step": 14508
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.502439402819746,
+      "learning_rate": 9.446674384988591e-07,
+      "loss": 0.1612,
+      "step": 14509
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.4213559597176937,
+      "learning_rate": 9.441405538700193e-07,
+      "loss": 0.1843,
+      "step": 14510
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 3.3078474443884636,
+      "learning_rate": 9.436138008967093e-07,
+      "loss": 0.1971,
+      "step": 14511
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.2963257108941868,
+      "learning_rate": 9.430871795960289e-07,
+      "loss": 0.1237,
+      "step": 14512
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.252002054528906,
+      "learning_rate": 9.425606899850708e-07,
+      "loss": 0.1711,
+      "step": 14513
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.1526113813576393,
+      "learning_rate": 9.420343320809261e-07,
+      "loss": 0.1867,
+      "step": 14514
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.3212719371331487,
+      "learning_rate": 9.415081059006798e-07,
+      "loss": 0.1473,
+      "step": 14515
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.8083841725606786,
+      "learning_rate": 9.409820114614144e-07,
+      "loss": 0.1903,
+      "step": 14516
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.21679060491964,
+      "learning_rate": 9.404560487802084e-07,
+      "loss": 0.1637,
+      "step": 14517
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.6908685653303512,
+      "learning_rate": 9.399302178741321e-07,
+      "loss": 0.1855,
+      "step": 14518
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.5262773796112605,
+      "learning_rate": 9.394045187602557e-07,
+      "loss": 0.2106,
+      "step": 14519
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 2.8263930027942465,
+      "learning_rate": 9.38878951455644e-07,
+      "loss": 0.1456,
+      "step": 14520
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.4433194156220546,
+      "learning_rate": 9.383535159773577e-07,
+      "loss": 0.1694,
+      "step": 14521
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.478682257913667,
+      "learning_rate": 9.378282123424503e-07,
+      "loss": 0.1777,
+      "step": 14522
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.9317439119996278,
+      "learning_rate": 9.373030405679756e-07,
+      "loss": 0.1664,
+      "step": 14523
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.078445578680009,
+      "learning_rate": 9.367780006709815e-07,
+      "loss": 0.1362,
+      "step": 14524
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.3282452248405436,
+      "learning_rate": 9.362530926685087e-07,
+      "loss": 0.1967,
+      "step": 14525
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.1441382556915496,
+      "learning_rate": 9.357283165775988e-07,
+      "loss": 0.1784,
+      "step": 14526
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.1835420238409502,
+      "learning_rate": 9.352036724152836e-07,
+      "loss": 0.1687,
+      "step": 14527
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.6078942718323634,
+      "learning_rate": 9.346791601985944e-07,
+      "loss": 0.2142,
+      "step": 14528
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.6399955313973489,
+      "learning_rate": 9.341547799445572e-07,
+      "loss": 0.1761,
+      "step": 14529
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.286258329321878,
+      "learning_rate": 9.336305316701955e-07,
+      "loss": 0.1585,
+      "step": 14530
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.3309480508250426,
+      "learning_rate": 9.33106415392524e-07,
+      "loss": 0.1369,
+      "step": 14531
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.6408881021887418,
+      "learning_rate": 9.325824311285564e-07,
+      "loss": 0.1792,
+      "step": 14532
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.5532150196414163,
+      "learning_rate": 9.320585788953018e-07,
+      "loss": 0.2057,
+      "step": 14533
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.4926012898464285,
+      "learning_rate": 9.315348587097656e-07,
+      "loss": 0.243,
+      "step": 14534
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.6025565063345437,
+      "learning_rate": 9.310112705889484e-07,
+      "loss": 0.2078,
+      "step": 14535
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.680073063977372,
+      "learning_rate": 9.304878145498442e-07,
+      "loss": 0.1997,
+      "step": 14536
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.4798849409152781,
+      "learning_rate": 9.299644906094451e-07,
+      "loss": 0.1601,
+      "step": 14537
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.4803634646377855,
+      "learning_rate": 9.294412987847395e-07,
+      "loss": 0.2017,
+      "step": 14538
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.6558678024088336,
+      "learning_rate": 9.289182390927115e-07,
+      "loss": 0.1643,
+      "step": 14539
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.345556420023989,
+      "learning_rate": 9.283953115503369e-07,
+      "loss": 0.1765,
+      "step": 14540
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.640404513847533,
+      "learning_rate": 9.278725161745922e-07,
+      "loss": 0.1701,
+      "step": 14541
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.8093977556131864,
+      "learning_rate": 9.273498529824482e-07,
+      "loss": 0.2264,
+      "step": 14542
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.8723927729072725,
+      "learning_rate": 9.268273219908685e-07,
+      "loss": 0.1925,
+      "step": 14543
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.5985406908490265,
+      "learning_rate": 9.263049232168176e-07,
+      "loss": 0.1833,
+      "step": 14544
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.2086387041506266,
+      "learning_rate": 9.257826566772505e-07,
+      "loss": 0.1431,
+      "step": 14545
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.1235000956343886,
+      "learning_rate": 9.252605223891209e-07,
+      "loss": 0.144,
+      "step": 14546
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.4688202886807094,
+      "learning_rate": 9.247385203693776e-07,
+      "loss": 0.1993,
+      "step": 14547
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.9435328650550412,
+      "learning_rate": 9.24216650634967e-07,
+      "loss": 0.2166,
+      "step": 14548
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.3375145543723705,
+      "learning_rate": 9.236949132028267e-07,
+      "loss": 0.1796,
+      "step": 14549
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.3258714349629905,
+      "learning_rate": 9.23173308089893e-07,
+      "loss": 0.1728,
+      "step": 14550
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.9702641187114607,
+      "learning_rate": 9.226518353130981e-07,
+      "loss": 0.174,
+      "step": 14551
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.441441563946433,
+      "learning_rate": 9.221304948893695e-07,
+      "loss": 0.1633,
+      "step": 14552
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.4562106902829555,
+      "learning_rate": 9.216092868356314e-07,
+      "loss": 0.1775,
+      "step": 14553
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.3285415164585914,
+      "learning_rate": 9.210882111687997e-07,
+      "loss": 0.1279,
+      "step": 14554
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.5764853060663975,
+      "learning_rate": 9.205672679057898e-07,
+      "loss": 0.1845,
+      "step": 14555
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.5245072252640952,
+      "learning_rate": 9.200464570635126e-07,
+      "loss": 0.1494,
+      "step": 14556
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.6818746920032142,
+      "learning_rate": 9.195257786588746e-07,
+      "loss": 0.1998,
+      "step": 14557
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.6600441928519174,
+      "learning_rate": 9.190052327087751e-07,
+      "loss": 0.214,
+      "step": 14558
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.4394974100037279,
+      "learning_rate": 9.184848192301127e-07,
+      "loss": 0.1553,
+      "step": 14559
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.3586016296632033,
+      "learning_rate": 9.179645382397812e-07,
+      "loss": 0.134,
+      "step": 14560
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.52481538376221,
+      "learning_rate": 9.174443897546669e-07,
+      "loss": 0.217,
+      "step": 14561
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.5083972212220855,
+      "learning_rate": 9.169243737916566e-07,
+      "loss": 0.2119,
+      "step": 14562
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.241401225584604,
+      "learning_rate": 9.164044903676278e-07,
+      "loss": 0.1712,
+      "step": 14563
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.1548088313250544,
+      "learning_rate": 9.158847394994574e-07,
+      "loss": 0.182,
+      "step": 14564
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.5722970829061618,
+      "learning_rate": 9.153651212040172e-07,
+      "loss": 0.2003,
+      "step": 14565
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.099909113114884,
+      "learning_rate": 9.14845635498175e-07,
+      "loss": 0.2037,
+      "step": 14566
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.2500860750841842,
+      "learning_rate": 9.143262823987909e-07,
+      "loss": 0.1695,
+      "step": 14567
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.684894099423805,
+      "learning_rate": 9.138070619227258e-07,
+      "loss": 0.1893,
+      "step": 14568
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.244584302376105,
+      "learning_rate": 9.132879740868327e-07,
+      "loss": 0.2022,
+      "step": 14569
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.31976213213906,
+      "learning_rate": 9.127690189079625e-07,
+      "loss": 0.1643,
+      "step": 14570
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.2846312682572445,
+      "learning_rate": 9.122501964029612e-07,
+      "loss": 0.1317,
+      "step": 14571
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.6438736223649142,
+      "learning_rate": 9.117315065886678e-07,
+      "loss": 0.1685,
+      "step": 14572
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.397815198266751,
+      "learning_rate": 9.112129494819211e-07,
+      "loss": 0.122,
+      "step": 14573
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.0743047194325923,
+      "learning_rate": 9.10694525099553e-07,
+      "loss": 0.1489,
+      "step": 14574
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.3113119220386007,
+      "learning_rate": 9.10176233458393e-07,
+      "loss": 0.2004,
+      "step": 14575
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.54025799653111,
+      "learning_rate": 9.096580745752637e-07,
+      "loss": 0.1641,
+      "step": 14576
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.2864585701431284,
+      "learning_rate": 9.091400484669848e-07,
+      "loss": 0.15,
+      "step": 14577
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.4647524283871278,
+      "learning_rate": 9.086221551503737e-07,
+      "loss": 0.1791,
+      "step": 14578
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.2891660301942165,
+      "learning_rate": 9.081043946422391e-07,
+      "loss": 0.1755,
+      "step": 14579
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.3780553964524085,
+      "learning_rate": 9.075867669593902e-07,
+      "loss": 0.1861,
+      "step": 14580
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.9834541562612065,
+      "learning_rate": 9.07069272118627e-07,
+      "loss": 0.1022,
+      "step": 14581
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.242730621746619,
+      "learning_rate": 9.065519101367487e-07,
+      "loss": 0.2367,
+      "step": 14582
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.2432371299349048,
+      "learning_rate": 9.060346810305493e-07,
+      "loss": 0.1726,
+      "step": 14583
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.3761758436654152,
+      "learning_rate": 9.055175848168191e-07,
+      "loss": 0.1857,
+      "step": 14584
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.8092922934321373,
+      "learning_rate": 9.050006215123419e-07,
+      "loss": 0.113,
+      "step": 14585
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.1810752304155328,
+      "learning_rate": 9.044837911338994e-07,
+      "loss": 0.159,
+      "step": 14586
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.198693875066969,
+      "learning_rate": 9.039670936982675e-07,
+      "loss": 0.2136,
+      "step": 14587
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.682372435726176,
+      "learning_rate": 9.034505292222195e-07,
+      "loss": 0.2103,
+      "step": 14588
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.7402784468253838,
+      "learning_rate": 9.029340977225248e-07,
+      "loss": 0.2042,
+      "step": 14589
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.5999937422451194,
+      "learning_rate": 9.024177992159433e-07,
+      "loss": 0.2097,
+      "step": 14590
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.1757591473602569,
+      "learning_rate": 9.019016337192365e-07,
+      "loss": 0.1632,
+      "step": 14591
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.9014345224739178,
+      "learning_rate": 9.01385601249159e-07,
+      "loss": 0.1732,
+      "step": 14592
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.7860743709657878,
+      "learning_rate": 9.008697018224638e-07,
+      "loss": 0.2303,
+      "step": 14593
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.3524236416371334,
+      "learning_rate": 9.003539354558932e-07,
+      "loss": 0.1405,
+      "step": 14594
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.0616337053007612,
+      "learning_rate": 8.998383021661927e-07,
+      "loss": 0.1857,
+      "step": 14595
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.44030255290044,
+      "learning_rate": 8.993228019700978e-07,
+      "loss": 0.1769,
+      "step": 14596
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.1503245427009847,
+      "learning_rate": 8.988074348843423e-07,
+      "loss": 0.2387,
+      "step": 14597
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.519573490565773,
+      "learning_rate": 8.982922009256573e-07,
+      "loss": 0.1625,
+      "step": 14598
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.7675577742105835,
+      "learning_rate": 8.97777100110765e-07,
+      "loss": 0.2268,
+      "step": 14599
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.4250587095748344,
+      "learning_rate": 8.972621324563868e-07,
+      "loss": 0.204,
+      "step": 14600
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.3484201947336139,
+      "learning_rate": 8.967472979792391e-07,
+      "loss": 0.2028,
+      "step": 14601
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.3167612095305534,
+      "learning_rate": 8.96232596696035e-07,
+      "loss": 0.1691,
+      "step": 14602
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.2548103044394443,
+      "learning_rate": 8.957180286234796e-07,
+      "loss": 0.1779,
+      "step": 14603
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.076434530893632,
+      "learning_rate": 8.952035937782766e-07,
+      "loss": 0.193,
+      "step": 14604
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.3696568804109337,
+      "learning_rate": 8.946892921771255e-07,
+      "loss": 0.1856,
+      "step": 14605
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.6478815135469715,
+      "learning_rate": 8.941751238367208e-07,
+      "loss": 0.1707,
+      "step": 14606
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.723922428514002,
+      "learning_rate": 8.93661088773754e-07,
+      "loss": 0.1678,
+      "step": 14607
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.540720591084625,
+      "learning_rate": 8.93147187004908e-07,
+      "loss": 0.1826,
+      "step": 14608
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.5431167531555983,
+      "learning_rate": 8.92633418546866e-07,
+      "loss": 0.1765,
+      "step": 14609
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.9821499700412553,
+      "learning_rate": 8.921197834163053e-07,
+      "loss": 0.1584,
+      "step": 14610
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.2202248269505047,
+      "learning_rate": 8.916062816298998e-07,
+      "loss": 0.181,
+      "step": 14611
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.139649901924402,
+      "learning_rate": 8.910929132043155e-07,
+      "loss": 0.1738,
+      "step": 14612
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.3041397532478043,
+      "learning_rate": 8.905796781562198e-07,
+      "loss": 0.1385,
+      "step": 14613
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.629960256557152,
+      "learning_rate": 8.900665765022692e-07,
+      "loss": 0.201,
+      "step": 14614
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.492931250515966,
+      "learning_rate": 8.895536082591205e-07,
+      "loss": 0.1973,
+      "step": 14615
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.310389263380027,
+      "learning_rate": 8.89040773443427e-07,
+      "loss": 0.1688,
+      "step": 14616
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.692583977948653,
+      "learning_rate": 8.885280720718331e-07,
+      "loss": 0.2236,
+      "step": 14617
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.2803354662920305,
+      "learning_rate": 8.880155041609817e-07,
+      "loss": 0.1285,
+      "step": 14618
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.235186958655098,
+      "learning_rate": 8.875030697275116e-07,
+      "loss": 0.173,
+      "step": 14619
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.830656540372016,
+      "learning_rate": 8.86990768788058e-07,
+      "loss": 0.1753,
+      "step": 14620
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.9361241285317334,
+      "learning_rate": 8.864786013592486e-07,
+      "loss": 0.2168,
+      "step": 14621
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.3252463137680264,
+      "learning_rate": 8.859665674577084e-07,
+      "loss": 0.1583,
+      "step": 14622
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.0624317154259255,
+      "learning_rate": 8.854546671000597e-07,
+      "loss": 0.1704,
+      "step": 14623
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.3464203903714733,
+      "learning_rate": 8.849429003029181e-07,
+      "loss": 0.2256,
+      "step": 14624
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.9593686345911248,
+      "learning_rate": 8.844312670828975e-07,
+      "loss": 0.1278,
+      "step": 14625
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.5678443490607168,
+      "learning_rate": 8.839197674566035e-07,
+      "loss": 0.1375,
+      "step": 14626
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.40581759056581,
+      "learning_rate": 8.834084014406413e-07,
+      "loss": 0.1933,
+      "step": 14627
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 2.282885898974094,
+      "learning_rate": 8.828971690516091e-07,
+      "loss": 0.2274,
+      "step": 14628
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.3823347020534933,
+      "learning_rate": 8.823860703061032e-07,
+      "loss": 0.1698,
+      "step": 14629
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 4.249289113083044,
+      "learning_rate": 8.818751052207125e-07,
+      "loss": 0.2004,
+      "step": 14630
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.2218094824642904,
+      "learning_rate": 8.813642738120254e-07,
+      "loss": 0.159,
+      "step": 14631
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.9161979674076386,
+      "learning_rate": 8.808535760966208e-07,
+      "loss": 0.1952,
+      "step": 14632
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.445382771201333,
+      "learning_rate": 8.803430120910778e-07,
+      "loss": 0.1546,
+      "step": 14633
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.2052502130580636,
+      "learning_rate": 8.798325818119707e-07,
+      "loss": 0.155,
+      "step": 14634
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.2767790718637044,
+      "learning_rate": 8.793222852758665e-07,
+      "loss": 0.2036,
+      "step": 14635
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 2.1195674976099523,
+      "learning_rate": 8.788121224993301e-07,
+      "loss": 0.1698,
+      "step": 14636
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.3867000283065485,
+      "learning_rate": 8.783020934989223e-07,
+      "loss": 0.1687,
+      "step": 14637
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.0935483405993085,
+      "learning_rate": 8.777921982911996e-07,
+      "loss": 0.1851,
+      "step": 14638
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.72696667251502,
+      "learning_rate": 8.772824368927113e-07,
+      "loss": 0.2024,
+      "step": 14639
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.326319628702514,
+      "learning_rate": 8.767728093200061e-07,
+      "loss": 0.1803,
+      "step": 14640
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.587624948758232,
+      "learning_rate": 8.762633155896266e-07,
+      "loss": 0.217,
+      "step": 14641
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.3328769214351186,
+      "learning_rate": 8.75753955718111e-07,
+      "loss": 0.1866,
+      "step": 14642
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.2185799162121402,
+      "learning_rate": 8.752447297219946e-07,
+      "loss": 0.1788,
+      "step": 14643
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.3715619757636721,
+      "learning_rate": 8.747356376178057e-07,
+      "loss": 0.2113,
+      "step": 14644
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.374677566090059,
+      "learning_rate": 8.742266794220694e-07,
+      "loss": 0.1724,
+      "step": 14645
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.2134732614592347,
+      "learning_rate": 8.737178551513081e-07,
+      "loss": 0.1489,
+      "step": 14646
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.2769198808050573,
+      "learning_rate": 8.732091648220392e-07,
+      "loss": 0.1598,
+      "step": 14647
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.5393371821012305,
+      "learning_rate": 8.727006084507722e-07,
+      "loss": 0.202,
+      "step": 14648
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.452319644976233,
+      "learning_rate": 8.721921860540184e-07,
+      "loss": 0.1813,
+      "step": 14649
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.060096427280997,
+      "learning_rate": 8.716838976482789e-07,
+      "loss": 0.1607,
+      "step": 14650
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.8699434045269812,
+      "learning_rate": 8.711757432500533e-07,
+      "loss": 0.182,
+      "step": 14651
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 2.7270371425632534,
+      "learning_rate": 8.706677228758393e-07,
+      "loss": 0.2053,
+      "step": 14652
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.639105747332378,
+      "learning_rate": 8.701598365421238e-07,
+      "loss": 0.2269,
+      "step": 14653
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.7103389755644907,
+      "learning_rate": 8.69652084265395e-07,
+      "loss": 0.2255,
+      "step": 14654
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.8083341259742207,
+      "learning_rate": 8.691444660621351e-07,
+      "loss": 0.235,
+      "step": 14655
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.8090096442480572,
+      "learning_rate": 8.686369819488222e-07,
+      "loss": 0.1919,
+      "step": 14656
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.4707788394724641,
+      "learning_rate": 8.681296319419269e-07,
+      "loss": 0.1951,
+      "step": 14657
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.2609129538149988,
+      "learning_rate": 8.6762241605792e-07,
+      "loss": 0.1926,
+      "step": 14658
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.1121150785406433,
+      "learning_rate": 8.67115334313266e-07,
+      "loss": 0.183,
+      "step": 14659
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.5142045591179674,
+      "learning_rate": 8.666083867244246e-07,
+      "loss": 0.2043,
+      "step": 14660
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.4251087666921927,
+      "learning_rate": 8.661015733078531e-07,
+      "loss": 0.1588,
+      "step": 14661
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.6721406351279449,
+      "learning_rate": 8.655948940800007e-07,
+      "loss": 0.1906,
+      "step": 14662
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.8351122848031782,
+      "learning_rate": 8.650883490573153e-07,
+      "loss": 0.1629,
+      "step": 14663
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.1406230403935569,
+      "learning_rate": 8.645819382562398e-07,
+      "loss": 0.1871,
+      "step": 14664
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.3457978773047674,
+      "learning_rate": 8.640756616932139e-07,
+      "loss": 0.1583,
+      "step": 14665
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.3648690647688284,
+      "learning_rate": 8.635695193846688e-07,
+      "loss": 0.1781,
+      "step": 14666
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.6747197076797202,
+      "learning_rate": 8.630635113470376e-07,
+      "loss": 0.1347,
+      "step": 14667
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.3397986790942087,
+      "learning_rate": 8.625576375967421e-07,
+      "loss": 0.1846,
+      "step": 14668
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.3242186205928592,
+      "learning_rate": 8.620518981502046e-07,
+      "loss": 0.1791,
+      "step": 14669
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.1681097895967312,
+      "learning_rate": 8.615462930238438e-07,
+      "loss": 0.1952,
+      "step": 14670
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.7662866753552078,
+      "learning_rate": 8.610408222340683e-07,
+      "loss": 0.1754,
+      "step": 14671
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 2.056121601281658,
+      "learning_rate": 8.605354857972881e-07,
+      "loss": 0.2046,
+      "step": 14672
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.2084842596534293,
+      "learning_rate": 8.600302837299063e-07,
+      "loss": 0.1991,
+      "step": 14673
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.3621805113936725,
+      "learning_rate": 8.595252160483225e-07,
+      "loss": 0.1521,
+      "step": 14674
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.2060905049677344,
+      "learning_rate": 8.59020282768932e-07,
+      "loss": 0.1769,
+      "step": 14675
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.150533039105407,
+      "learning_rate": 8.585154839081234e-07,
+      "loss": 0.1384,
+      "step": 14676
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.6315074318333078,
+      "learning_rate": 8.580108194822834e-07,
+      "loss": 0.2149,
+      "step": 14677
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.5204630252854634,
+      "learning_rate": 8.57506289507794e-07,
+      "loss": 0.2338,
+      "step": 14678
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.3020390166052502,
+      "learning_rate": 8.570018940010338e-07,
+      "loss": 0.1722,
+      "step": 14679
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.4344360725017102,
+      "learning_rate": 8.564976329783736e-07,
+      "loss": 0.1924,
+      "step": 14680
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.4771432390516224,
+      "learning_rate": 8.559935064561825e-07,
+      "loss": 0.2021,
+      "step": 14681
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.3013907714245694,
+      "learning_rate": 8.554895144508257e-07,
+      "loss": 0.1692,
+      "step": 14682
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.31333846036258,
+      "learning_rate": 8.549856569786636e-07,
+      "loss": 0.1953,
+      "step": 14683
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.78448416497897,
+      "learning_rate": 8.54481934056049e-07,
+      "loss": 0.2071,
+      "step": 14684
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.2758188186085015,
+      "learning_rate": 8.539783456993361e-07,
+      "loss": 0.1893,
+      "step": 14685
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.3652911740604443,
+      "learning_rate": 8.534748919248692e-07,
+      "loss": 0.1463,
+      "step": 14686
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.6051008772937534,
+      "learning_rate": 8.529715727489912e-07,
+      "loss": 0.1895,
+      "step": 14687
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.4103423759225706,
+      "learning_rate": 8.524683881880425e-07,
+      "loss": 0.2274,
+      "step": 14688
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.2610971259954715,
+      "learning_rate": 8.519653382583531e-07,
+      "loss": 0.1865,
+      "step": 14689
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.63472816123557,
+      "learning_rate": 8.514624229762547e-07,
+      "loss": 0.185,
+      "step": 14690
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.3692627986973556,
+      "learning_rate": 8.509596423580712e-07,
+      "loss": 0.1698,
+      "step": 14691
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.606124873798207,
+      "learning_rate": 8.504569964201237e-07,
+      "loss": 0.2025,
+      "step": 14692
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.3911086002376873,
+      "learning_rate": 8.499544851787294e-07,
+      "loss": 0.1841,
+      "step": 14693
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.513723478360553,
+      "learning_rate": 8.494521086501984e-07,
+      "loss": 0.1592,
+      "step": 14694
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.2641528834033924,
+      "learning_rate": 8.489498668508378e-07,
+      "loss": 0.1789,
+      "step": 14695
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.5214928417435438,
+      "learning_rate": 8.48447759796952e-07,
+      "loss": 0.2066,
+      "step": 14696
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.3184960584075276,
+      "learning_rate": 8.479457875048402e-07,
+      "loss": 0.1907,
+      "step": 14697
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.4449358217299018,
+      "learning_rate": 8.47443949990795e-07,
+      "loss": 0.1873,
+      "step": 14698
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.1339686117676497,
+      "learning_rate": 8.469422472711064e-07,
+      "loss": 0.1853,
+      "step": 14699
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.5305059922084507,
+      "learning_rate": 8.464406793620611e-07,
+      "loss": 0.2074,
+      "step": 14700
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.3928179969020091,
+      "learning_rate": 8.45939246279941e-07,
+      "loss": 0.217,
+      "step": 14701
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.6670672729959317,
+      "learning_rate": 8.454379480410202e-07,
+      "loss": 0.169,
+      "step": 14702
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.4418420520485495,
+      "learning_rate": 8.449367846615741e-07,
+      "loss": 0.1717,
+      "step": 14703
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.3364787093006933,
+      "learning_rate": 8.444357561578681e-07,
+      "loss": 0.193,
+      "step": 14704
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.407546007800216,
+      "learning_rate": 8.439348625461669e-07,
+      "loss": 0.1656,
+      "step": 14705
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.69164913098716,
+      "learning_rate": 8.434341038427318e-07,
+      "loss": 0.169,
+      "step": 14706
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.127892775158657,
+      "learning_rate": 8.429334800638139e-07,
+      "loss": 0.1756,
+      "step": 14707
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.2874047514950753,
+      "learning_rate": 8.424329912256662e-07,
+      "loss": 0.1871,
+      "step": 14708
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.440829086535549,
+      "learning_rate": 8.419326373445341e-07,
+      "loss": 0.1713,
+      "step": 14709
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.5033942329438537,
+      "learning_rate": 8.414324184366595e-07,
+      "loss": 0.214,
+      "step": 14710
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.590873380165777,
+      "learning_rate": 8.409323345182818e-07,
+      "loss": 0.2028,
+      "step": 14711
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.9548767127101463,
+      "learning_rate": 8.404323856056301e-07,
+      "loss": 0.1711,
+      "step": 14712
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.2112262381593795,
+      "learning_rate": 8.399325717149359e-07,
+      "loss": 0.1643,
+      "step": 14713
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.289336033901223,
+      "learning_rate": 8.394328928624223e-07,
+      "loss": 0.1317,
+      "step": 14714
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.3112411251986014,
+      "learning_rate": 8.389333490643103e-07,
+      "loss": 0.1922,
+      "step": 14715
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.3312718304558988,
+      "learning_rate": 8.384339403368142e-07,
+      "loss": 0.1975,
+      "step": 14716
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.2211247074744993,
+      "learning_rate": 8.379346666961446e-07,
+      "loss": 0.1605,
+      "step": 14717
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.0672689064723238,
+      "learning_rate": 8.374355281585094e-07,
+      "loss": 0.1577,
+      "step": 14718
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.191135400721682,
+      "learning_rate": 8.369365247401123e-07,
+      "loss": 0.1843,
+      "step": 14719
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.9186724752737963,
+      "learning_rate": 8.364376564571475e-07,
+      "loss": 0.196,
+      "step": 14720
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.4873005096116376,
+      "learning_rate": 8.359389233258125e-07,
+      "loss": 0.2404,
+      "step": 14721
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.6086769719802891,
+      "learning_rate": 8.354403253622928e-07,
+      "loss": 0.197,
+      "step": 14722
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.534812791759303,
+      "learning_rate": 8.349418625827755e-07,
+      "loss": 0.2041,
+      "step": 14723
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.850850366708398,
+      "learning_rate": 8.344435350034413e-07,
+      "loss": 0.232,
+      "step": 14724
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.491678875224094,
+      "learning_rate": 8.339453426404642e-07,
+      "loss": 0.2071,
+      "step": 14725
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.101504304038499,
+      "learning_rate": 8.334472855100168e-07,
+      "loss": 0.1298,
+      "step": 14726
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.3797868936924509,
+      "learning_rate": 8.329493636282665e-07,
+      "loss": 0.1872,
+      "step": 14727
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.3337666569843194,
+      "learning_rate": 8.324515770113762e-07,
+      "loss": 0.1844,
+      "step": 14728
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.2374699037437116,
+      "learning_rate": 8.319539256755049e-07,
+      "loss": 0.121,
+      "step": 14729
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 2.832531325814335,
+      "learning_rate": 8.314564096368056e-07,
+      "loss": 0.2046,
+      "step": 14730
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.6238515582858895,
+      "learning_rate": 8.309590289114278e-07,
+      "loss": 0.2606,
+      "step": 14731
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.0704187980827136,
+      "learning_rate": 8.304617835155177e-07,
+      "loss": 0.1738,
+      "step": 14732
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.7248253390095634,
+      "learning_rate": 8.299646734652162e-07,
+      "loss": 0.216,
+      "step": 14733
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.9294132060437137,
+      "learning_rate": 8.294676987766587e-07,
+      "loss": 0.1649,
+      "step": 14734
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.5767356733971971,
+      "learning_rate": 8.289708594659779e-07,
+      "loss": 0.2123,
+      "step": 14735
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.358673808243592,
+      "learning_rate": 8.284741555493014e-07,
+      "loss": 0.2115,
+      "step": 14736
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.081844713127602,
+      "learning_rate": 8.279775870427543e-07,
+      "loss": 0.1025,
+      "step": 14737
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.3062927227699617,
+      "learning_rate": 8.274811539624516e-07,
+      "loss": 0.1897,
+      "step": 14738
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.3479971882880302,
+      "learning_rate": 8.26984856324512e-07,
+      "loss": 0.2095,
+      "step": 14739
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.4196686656658977,
+      "learning_rate": 8.264886941450423e-07,
+      "loss": 0.1773,
+      "step": 14740
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.8249243815315148,
+      "learning_rate": 8.259926674401491e-07,
+      "loss": 0.2553,
+      "step": 14741
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.2779173423500791,
+      "learning_rate": 8.254967762259358e-07,
+      "loss": 0.1894,
+      "step": 14742
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.3279381816966578,
+      "learning_rate": 8.250010205184966e-07,
+      "loss": 0.1672,
+      "step": 14743
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.573950358908561,
+      "learning_rate": 8.245054003339247e-07,
+      "loss": 0.1401,
+      "step": 14744
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.725394324138287,
+      "learning_rate": 8.240099156883085e-07,
+      "loss": 0.1304,
+      "step": 14745
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.8497728076177815,
+      "learning_rate": 8.235145665977323e-07,
+      "loss": 0.1845,
+      "step": 14746
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.6221586934814791,
+      "learning_rate": 8.23019353078276e-07,
+      "loss": 0.1867,
+      "step": 14747
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.2710747212707176,
+      "learning_rate": 8.225242751460117e-07,
+      "loss": 0.1609,
+      "step": 14748
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.4542044045361795,
+      "learning_rate": 8.22029332817012e-07,
+      "loss": 0.2014,
+      "step": 14749
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.1811281243530334,
+      "learning_rate": 8.215345261073427e-07,
+      "loss": 0.1954,
+      "step": 14750
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.0389410966206112,
+      "learning_rate": 8.210398550330667e-07,
+      "loss": 0.1646,
+      "step": 14751
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.5182782525644123,
+      "learning_rate": 8.205453196102386e-07,
+      "loss": 0.2069,
+      "step": 14752
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.5347946702126802,
+      "learning_rate": 8.200509198549134e-07,
+      "loss": 0.212,
+      "step": 14753
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.1831539079587492,
+      "learning_rate": 8.19556655783138e-07,
+      "loss": 0.1817,
+      "step": 14754
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.4033791059204772,
+      "learning_rate": 8.190625274109598e-07,
+      "loss": 0.146,
+      "step": 14755
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.3829667980510887,
+      "learning_rate": 8.185685347544142e-07,
+      "loss": 0.1817,
+      "step": 14756
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.5526659584567244,
+      "learning_rate": 8.180746778295401e-07,
+      "loss": 0.1739,
+      "step": 14757
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.2867066617304068,
+      "learning_rate": 8.175809566523652e-07,
+      "loss": 0.1727,
+      "step": 14758
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.4995441538843906,
+      "learning_rate": 8.170873712389177e-07,
+      "loss": 0.1768,
+      "step": 14759
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.091694361287486,
+      "learning_rate": 8.165939216052204e-07,
+      "loss": 0.207,
+      "step": 14760
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.4754472486857684,
+      "learning_rate": 8.161006077672889e-07,
+      "loss": 0.1701,
+      "step": 14761
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.8101140794052513,
+      "learning_rate": 8.156074297411375e-07,
+      "loss": 0.1515,
+      "step": 14762
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.1439505607741203,
+      "learning_rate": 8.151143875427759e-07,
+      "loss": 0.1713,
+      "step": 14763
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.4509308051292331,
+      "learning_rate": 8.146214811882069e-07,
+      "loss": 0.1775,
+      "step": 14764
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.2643252191676486,
+      "learning_rate": 8.141287106934331e-07,
+      "loss": 0.2293,
+      "step": 14765
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.1896599995275616,
+      "learning_rate": 8.136360760744471e-07,
+      "loss": 0.1801,
+      "step": 14766
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.2372147693794018,
+      "learning_rate": 8.131435773472413e-07,
+      "loss": 0.1411,
+      "step": 14767
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.4217676578078868,
+      "learning_rate": 8.126512145278026e-07,
+      "loss": 0.2228,
+      "step": 14768
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.6865676140756274,
+      "learning_rate": 8.121589876321146e-07,
+      "loss": 0.2023,
+      "step": 14769
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.5810671564921719,
+      "learning_rate": 8.116668966761526e-07,
+      "loss": 0.2264,
+      "step": 14770
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.2697909456344363,
+      "learning_rate": 8.111749416758919e-07,
+      "loss": 0.2121,
+      "step": 14771
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.1050438418148987,
+      "learning_rate": 8.106831226473011e-07,
+      "loss": 0.1177,
+      "step": 14772
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.6169549341722018,
+      "learning_rate": 8.101914396063465e-07,
+      "loss": 0.2386,
+      "step": 14773
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.3141478514196245,
+      "learning_rate": 8.096998925689859e-07,
+      "loss": 0.1624,
+      "step": 14774
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.4189177140914437,
+      "learning_rate": 8.092084815511774e-07,
+      "loss": 0.222,
+      "step": 14775
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.3578223810835057,
+      "learning_rate": 8.087172065688709e-07,
+      "loss": 0.1602,
+      "step": 14776
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.7545282624532048,
+      "learning_rate": 8.082260676380132e-07,
+      "loss": 0.2253,
+      "step": 14777
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.4650803794471239,
+      "learning_rate": 8.077350647745496e-07,
+      "loss": 0.1148,
+      "step": 14778
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.0913955278354612,
+      "learning_rate": 8.072441979944151e-07,
+      "loss": 0.1555,
+      "step": 14779
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.4493230283498633,
+      "learning_rate": 8.067534673135446e-07,
+      "loss": 0.1732,
+      "step": 14780
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.4438490598308615,
+      "learning_rate": 8.062628727478683e-07,
+      "loss": 0.1582,
+      "step": 14781
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.5559585876782671,
+      "learning_rate": 8.057724143133106e-07,
+      "loss": 0.1875,
+      "step": 14782
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.252905508686682,
+      "learning_rate": 8.052820920257936e-07,
+      "loss": 0.1644,
+      "step": 14783
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.2912720124112362,
+      "learning_rate": 8.047919059012304e-07,
+      "loss": 0.1866,
+      "step": 14784
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.2506857659839392,
+      "learning_rate": 8.043018559555349e-07,
+      "loss": 0.1505,
+      "step": 14785
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.9778484413303614,
+      "learning_rate": 8.038119422046131e-07,
+      "loss": 0.1828,
+      "step": 14786
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.4904704525641217,
+      "learning_rate": 8.033221646643702e-07,
+      "loss": 0.2521,
+      "step": 14787
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.5106311897873868,
+      "learning_rate": 8.028325233507022e-07,
+      "loss": 0.1862,
+      "step": 14788
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.6504278957514948,
+      "learning_rate": 8.023430182795039e-07,
+      "loss": 0.1875,
+      "step": 14789
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.54135071624932,
+      "learning_rate": 8.018536494666645e-07,
+      "loss": 0.1858,
+      "step": 14790
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.4840625885919032,
+      "learning_rate": 8.013644169280715e-07,
+      "loss": 0.1583,
+      "step": 14791
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.2221230755073196,
+      "learning_rate": 8.008753206796027e-07,
+      "loss": 0.1931,
+      "step": 14792
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 2.4777937454839805,
+      "learning_rate": 8.003863607371371e-07,
+      "loss": 0.192,
+      "step": 14793
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.5600595250373175,
+      "learning_rate": 7.998975371165435e-07,
+      "loss": 0.2164,
+      "step": 14794
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.46481780474256,
+      "learning_rate": 7.994088498336911e-07,
+      "loss": 0.1732,
+      "step": 14795
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.1134335882127273,
+      "learning_rate": 7.989202989044442e-07,
+      "loss": 0.1635,
+      "step": 14796
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.4959129293944877,
+      "learning_rate": 7.984318843446593e-07,
+      "loss": 0.2035,
+      "step": 14797
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.3533046194634677,
+      "learning_rate": 7.979436061701912e-07,
+      "loss": 0.1516,
+      "step": 14798
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.3392261967036376,
+      "learning_rate": 7.974554643968901e-07,
+      "loss": 0.1753,
+      "step": 14799
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.455316977130539,
+      "learning_rate": 7.969674590406012e-07,
+      "loss": 0.2113,
+      "step": 14800
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 2.0536725471017467,
+      "learning_rate": 7.964795901171662e-07,
+      "loss": 0.1914,
+      "step": 14801
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.3965685145499769,
+      "learning_rate": 7.959918576424203e-07,
+      "loss": 0.1889,
+      "step": 14802
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 2.186176381545627,
+      "learning_rate": 7.955042616321962e-07,
+      "loss": 0.1864,
+      "step": 14803
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.5150700246212467,
+      "learning_rate": 7.950168021023213e-07,
+      "loss": 0.2201,
+      "step": 14804
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.5659373811209045,
+      "learning_rate": 7.945294790686203e-07,
+      "loss": 0.2323,
+      "step": 14805
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.1759823661981572,
+      "learning_rate": 7.940422925469099e-07,
+      "loss": 0.1563,
+      "step": 14806
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.508182732188083,
+      "learning_rate": 7.935552425530052e-07,
+      "loss": 0.1416,
+      "step": 14807
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.5000405157101118,
+      "learning_rate": 7.930683291027158e-07,
+      "loss": 0.1916,
+      "step": 14808
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.5232875945847058,
+      "learning_rate": 7.925815522118491e-07,
+      "loss": 0.1778,
+      "step": 14809
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.5440925899477402,
+      "learning_rate": 7.920949118962029e-07,
+      "loss": 0.1953,
+      "step": 14810
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.3405231554965402,
+      "learning_rate": 7.916084081715775e-07,
+      "loss": 0.1599,
+      "step": 14811
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 2.4551035167318758,
+      "learning_rate": 7.911220410537618e-07,
+      "loss": 0.1811,
+      "step": 14812
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.3787861169241662,
+      "learning_rate": 7.906358105585443e-07,
+      "loss": 0.2155,
+      "step": 14813
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.5193970432196737,
+      "learning_rate": 7.901497167017108e-07,
+      "loss": 0.199,
+      "step": 14814
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.9245886182691294,
+      "learning_rate": 7.89663759499037e-07,
+      "loss": 0.139,
+      "step": 14815
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.2485101583198956,
+      "learning_rate": 7.89177938966299e-07,
+      "loss": 0.152,
+      "step": 14816
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.759610320593469,
+      "learning_rate": 7.886922551192661e-07,
+      "loss": 0.2155,
+      "step": 14817
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.94443495233429,
+      "learning_rate": 7.882067079737044e-07,
+      "loss": 0.2246,
+      "step": 14818
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.1814222654244206,
+      "learning_rate": 7.877212975453758e-07,
+      "loss": 0.1904,
+      "step": 14819
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.690132710039976,
+      "learning_rate": 7.872360238500354e-07,
+      "loss": 0.1899,
+      "step": 14820
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.4767822324148683,
+      "learning_rate": 7.867508869034357e-07,
+      "loss": 0.1628,
+      "step": 14821
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.067255286497631,
+      "learning_rate": 7.862658867213252e-07,
+      "loss": 0.1555,
+      "step": 14822
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.1734548029023908,
+      "learning_rate": 7.857810233194485e-07,
+      "loss": 0.1771,
+      "step": 14823
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.0070433194936872,
+      "learning_rate": 7.852962967135419e-07,
+      "loss": 0.1288,
+      "step": 14824
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.5588486255105853,
+      "learning_rate": 7.848117069193406e-07,
+      "loss": 0.1669,
+      "step": 14825
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 2.5192020171580474,
+      "learning_rate": 7.843272539525759e-07,
+      "loss": 0.1758,
+      "step": 14826
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.5479269592983487,
+      "learning_rate": 7.838429378289736e-07,
+      "loss": 0.1794,
+      "step": 14827
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.320381670612768,
+      "learning_rate": 7.833587585642527e-07,
+      "loss": 0.2135,
+      "step": 14828
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.1238552401828026,
+      "learning_rate": 7.828747161741323e-07,
+      "loss": 0.206,
+      "step": 14829
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.26800971185375,
+      "learning_rate": 7.823908106743228e-07,
+      "loss": 0.1706,
+      "step": 14830
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.6323163497055035,
+      "learning_rate": 7.81907042080533e-07,
+      "loss": 0.145,
+      "step": 14831
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.2889588083119696,
+      "learning_rate": 7.814234104084673e-07,
+      "loss": 0.1897,
+      "step": 14832
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.5309794634619587,
+      "learning_rate": 7.809399156738218e-07,
+      "loss": 0.2159,
+      "step": 14833
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.1916757122554569,
+      "learning_rate": 7.804565578922935e-07,
+      "loss": 0.1871,
+      "step": 14834
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.1936374721117187,
+      "learning_rate": 7.799733370795714e-07,
+      "loss": 0.1845,
+      "step": 14835
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.1433252386191441,
+      "learning_rate": 7.794902532513415e-07,
+      "loss": 0.203,
+      "step": 14836
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.164235646216516,
+      "learning_rate": 7.790073064232862e-07,
+      "loss": 0.1796,
+      "step": 14837
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.8720204444800875,
+      "learning_rate": 7.785244966110794e-07,
+      "loss": 0.1276,
+      "step": 14838
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.9345929871522005,
+      "learning_rate": 7.780418238303955e-07,
+      "loss": 0.1135,
+      "step": 14839
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.60922077051042,
+      "learning_rate": 7.775592880969013e-07,
+      "loss": 0.1847,
+      "step": 14840
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.1744122519606823,
+      "learning_rate": 7.770768894262626e-07,
+      "loss": 0.1629,
+      "step": 14841
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.1880826461824163,
+      "learning_rate": 7.765946278341346e-07,
+      "loss": 0.211,
+      "step": 14842
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.2384617249773584,
+      "learning_rate": 7.761125033361738e-07,
+      "loss": 0.144,
+      "step": 14843
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.0486893340047623,
+      "learning_rate": 7.756305159480299e-07,
+      "loss": 0.1937,
+      "step": 14844
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.2496551425395395,
+      "learning_rate": 7.751486656853502e-07,
+      "loss": 0.1552,
+      "step": 14845
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.3571560249438224,
+      "learning_rate": 7.746669525637729e-07,
+      "loss": 0.2009,
+      "step": 14846
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.4055786967676092,
+      "learning_rate": 7.741853765989371e-07,
+      "loss": 0.1749,
+      "step": 14847
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.3145607426702857,
+      "learning_rate": 7.737039378064731e-07,
+      "loss": 0.199,
+      "step": 14848
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 2.8014128011524444,
+      "learning_rate": 7.73222636202009e-07,
+      "loss": 0.1553,
+      "step": 14849
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.432828407864885,
+      "learning_rate": 7.727414718011706e-07,
+      "loss": 0.2087,
+      "step": 14850
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.424966432987522,
+      "learning_rate": 7.72260444619573e-07,
+      "loss": 0.1634,
+      "step": 14851
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.6629138169789526,
+      "learning_rate": 7.717795546728329e-07,
+      "loss": 0.1914,
+      "step": 14852
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.4936051001158912,
+      "learning_rate": 7.7129880197656e-07,
+      "loss": 0.1583,
+      "step": 14853
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.5899776899074103,
+      "learning_rate": 7.708181865463593e-07,
+      "loss": 0.2237,
+      "step": 14854
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.3341780950542066,
+      "learning_rate": 7.703377083978336e-07,
+      "loss": 0.1716,
+      "step": 14855
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.7017306648434822,
+      "learning_rate": 7.69857367546577e-07,
+      "loss": 0.2012,
+      "step": 14856
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.4550020133984456,
+      "learning_rate": 7.693771640081826e-07,
+      "loss": 0.1766,
+      "step": 14857
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.7773321801112947,
+      "learning_rate": 7.688970977982385e-07,
+      "loss": 0.2203,
+      "step": 14858
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.0609467632411789,
+      "learning_rate": 7.684171689323289e-07,
+      "loss": 0.1347,
+      "step": 14859
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.766001644302782,
+      "learning_rate": 7.679373774260296e-07,
+      "loss": 0.1981,
+      "step": 14860
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.2453595811853682,
+      "learning_rate": 7.674577232949171e-07,
+      "loss": 0.1589,
+      "step": 14861
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.278142189572496,
+      "learning_rate": 7.669782065545616e-07,
+      "loss": 0.1715,
+      "step": 14862
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.5983024959660594,
+      "learning_rate": 7.66498827220527e-07,
+      "loss": 0.2233,
+      "step": 14863
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.1399546606528081,
+      "learning_rate": 7.660195853083746e-07,
+      "loss": 0.1619,
+      "step": 14864
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.329611459725154,
+      "learning_rate": 7.655404808336619e-07,
+      "loss": 0.1619,
+      "step": 14865
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.5499681219545263,
+      "learning_rate": 7.6506151381194e-07,
+      "loss": 0.1535,
+      "step": 14866
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.587882209652306,
+      "learning_rate": 7.645826842587556e-07,
+      "loss": 0.1821,
+      "step": 14867
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.1430857517018194,
+      "learning_rate": 7.641039921896548e-07,
+      "loss": 0.1561,
+      "step": 14868
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.551822111263238,
+      "learning_rate": 7.636254376201724e-07,
+      "loss": 0.1891,
+      "step": 14869
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.3838718526837326,
+      "learning_rate": 7.631470205658447e-07,
+      "loss": 0.2015,
+      "step": 14870
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.2869262478208552,
+      "learning_rate": 7.626687410422012e-07,
+      "loss": 0.2082,
+      "step": 14871
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.3754681473011003,
+      "learning_rate": 7.621905990647671e-07,
+      "loss": 0.142,
+      "step": 14872
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.4921535518040001,
+      "learning_rate": 7.617125946490639e-07,
+      "loss": 0.2203,
+      "step": 14873
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.443428017270673,
+      "learning_rate": 7.612347278106064e-07,
+      "loss": 0.2073,
+      "step": 14874
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 2.848759153727161,
+      "learning_rate": 7.607569985649066e-07,
+      "loss": 0.1823,
+      "step": 14875
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.038422129066992,
+      "learning_rate": 7.602794069274722e-07,
+      "loss": 0.1178,
+      "step": 14876
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.7660256834193018,
+      "learning_rate": 7.598019529138078e-07,
+      "loss": 0.1744,
+      "step": 14877
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.460238161845943,
+      "learning_rate": 7.593246365394091e-07,
+      "loss": 0.1583,
+      "step": 14878
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.5837993960667347,
+      "learning_rate": 7.588474578197713e-07,
+      "loss": 0.1566,
+      "step": 14879
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.4035368707812834,
+      "learning_rate": 7.583704167703848e-07,
+      "loss": 0.1838,
+      "step": 14880
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.13650058507684,
+      "learning_rate": 7.578935134067328e-07,
+      "loss": 0.1478,
+      "step": 14881
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.6326425819872628,
+      "learning_rate": 7.574167477442978e-07,
+      "loss": 0.2013,
+      "step": 14882
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.266573532841403,
+      "learning_rate": 7.569401197985537e-07,
+      "loss": 0.1446,
+      "step": 14883
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.541108509353899,
+      "learning_rate": 7.564636295849731e-07,
+      "loss": 0.189,
+      "step": 14884
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.2363030429603927,
+      "learning_rate": 7.559872771190235e-07,
+      "loss": 0.1985,
+      "step": 14885
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.3953690670119328,
+      "learning_rate": 7.555110624161683e-07,
+      "loss": 0.1791,
+      "step": 14886
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.535186019880596,
+      "learning_rate": 7.550349854918637e-07,
+      "loss": 0.1622,
+      "step": 14887
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.239109234812429,
+      "learning_rate": 7.545590463615643e-07,
+      "loss": 0.1481,
+      "step": 14888
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.9572571779720708,
+      "learning_rate": 7.540832450407198e-07,
+      "loss": 0.1205,
+      "step": 14889
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.442238926409739,
+      "learning_rate": 7.536075815447747e-07,
+      "loss": 0.2088,
+      "step": 14890
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.0714109728707546,
+      "learning_rate": 7.531320558891708e-07,
+      "loss": 0.1519,
+      "step": 14891
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.5276003229392237,
+      "learning_rate": 7.526566680893411e-07,
+      "loss": 0.209,
+      "step": 14892
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.0710538851152276,
+      "learning_rate": 7.521814181607178e-07,
+      "loss": 0.1648,
+      "step": 14893
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.4207087806419174,
+      "learning_rate": 7.517063061187291e-07,
+      "loss": 0.2178,
+      "step": 14894
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.1259397978789645,
+      "learning_rate": 7.512313319787972e-07,
+      "loss": 0.1356,
+      "step": 14895
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.35377211019049,
+      "learning_rate": 7.50756495756339e-07,
+      "loss": 0.2323,
+      "step": 14896
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.4742166362478428,
+      "learning_rate": 7.502817974667681e-07,
+      "loss": 0.1922,
+      "step": 14897
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.1759550593471975,
+      "learning_rate": 7.498072371254949e-07,
+      "loss": 0.1666,
+      "step": 14898
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.2616334183598306,
+      "learning_rate": 7.493328147479217e-07,
+      "loss": 0.1513,
+      "step": 14899
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.345132377298354,
+      "learning_rate": 7.488585303494511e-07,
+      "loss": 0.157,
+      "step": 14900
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.4452667383733468,
+      "learning_rate": 7.483843839454758e-07,
+      "loss": 0.1901,
+      "step": 14901
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.2815512965823506,
+      "learning_rate": 7.479103755513878e-07,
+      "loss": 0.205,
+      "step": 14902
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.3063367709858205,
+      "learning_rate": 7.474365051825749e-07,
+      "loss": 0.199,
+      "step": 14903
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.5182563464140288,
+      "learning_rate": 7.469627728544187e-07,
+      "loss": 0.1851,
+      "step": 14904
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.6371296023277713,
+      "learning_rate": 7.464891785822964e-07,
+      "loss": 0.1709,
+      "step": 14905
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.4832016925538616,
+      "learning_rate": 7.460157223815806e-07,
+      "loss": 0.124,
+      "step": 14906
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.0838287551109227,
+      "learning_rate": 7.455424042676412e-07,
+      "loss": 0.1296,
+      "step": 14907
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.917055522476814,
+      "learning_rate": 7.450692242558416e-07,
+      "loss": 0.2141,
+      "step": 14908
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.99243370891135,
+      "learning_rate": 7.445961823615427e-07,
+      "loss": 0.171,
+      "step": 14909
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.3077843226652595,
+      "learning_rate": 7.441232786000979e-07,
+      "loss": 0.1391,
+      "step": 14910
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.2418334984336181,
+      "learning_rate": 7.436505129868587e-07,
+      "loss": 0.153,
+      "step": 14911
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.4088569636384718,
+      "learning_rate": 7.431778855371713e-07,
+      "loss": 0.163,
+      "step": 14912
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.1079974486863031,
+      "learning_rate": 7.427053962663778e-07,
+      "loss": 0.1505,
+      "step": 14913
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.599232043428616,
+      "learning_rate": 7.422330451898157e-07,
+      "loss": 0.2138,
+      "step": 14914
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.2244013802849294,
+      "learning_rate": 7.41760832322817e-07,
+      "loss": 0.1599,
+      "step": 14915
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.4178266637984633,
+      "learning_rate": 7.412887576807115e-07,
+      "loss": 0.1753,
+      "step": 14916
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.5930434091161843,
+      "learning_rate": 7.408168212788203e-07,
+      "loss": 0.2247,
+      "step": 14917
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.426297567236312,
+      "learning_rate": 7.403450231324655e-07,
+      "loss": 0.1943,
+      "step": 14918
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.2755647514148067,
+      "learning_rate": 7.398733632569599e-07,
+      "loss": 0.1397,
+      "step": 14919
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.9823906957119355,
+      "learning_rate": 7.394018416676141e-07,
+      "loss": 0.2197,
+      "step": 14920
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.405585778513103,
+      "learning_rate": 7.389304583797352e-07,
+      "loss": 0.1698,
+      "step": 14921
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.1967171537530845,
+      "learning_rate": 7.384592134086233e-07,
+      "loss": 0.1354,
+      "step": 14922
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.4267032829585131,
+      "learning_rate": 7.379881067695776e-07,
+      "loss": 0.1638,
+      "step": 14923
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.4869052453270082,
+      "learning_rate": 7.375171384778874e-07,
+      "loss": 0.222,
+      "step": 14924
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 2.1454028111263113,
+      "learning_rate": 7.370463085488416e-07,
+      "loss": 0.1571,
+      "step": 14925
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.3301489610896866,
+      "learning_rate": 7.365756169977245e-07,
+      "loss": 0.2208,
+      "step": 14926
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.8104612212922917,
+      "learning_rate": 7.361050638398154e-07,
+      "loss": 0.1726,
+      "step": 14927
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.3904740018517159,
+      "learning_rate": 7.356346490903871e-07,
+      "loss": 0.1514,
+      "step": 14928
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.3227128550206402,
+      "learning_rate": 7.351643727647101e-07,
+      "loss": 0.2121,
+      "step": 14929
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.3017048578158643,
+      "learning_rate": 7.3469423487805e-07,
+      "loss": 0.1193,
+      "step": 14930
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 2.0237890954332314,
+      "learning_rate": 7.342242354456675e-07,
+      "loss": 0.2047,
+      "step": 14931
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.634908626595866,
+      "learning_rate": 7.337543744828207e-07,
+      "loss": 0.2076,
+      "step": 14932
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.4551005214529262,
+      "learning_rate": 7.332846520047587e-07,
+      "loss": 0.169,
+      "step": 14933
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.1442754283819132,
+      "learning_rate": 7.328150680267321e-07,
+      "loss": 0.1631,
+      "step": 14934
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.2506576774880502,
+      "learning_rate": 7.323456225639808e-07,
+      "loss": 0.193,
+      "step": 14935
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.7397110619927143,
+      "learning_rate": 7.31876315631746e-07,
+      "loss": 0.159,
+      "step": 14936
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.111378190935836,
+      "learning_rate": 7.314071472452589e-07,
+      "loss": 0.1479,
+      "step": 14937
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.2549352553278337,
+      "learning_rate": 7.309381174197505e-07,
+      "loss": 0.1629,
+      "step": 14938
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.0628778455471397,
+      "learning_rate": 7.304692261704455e-07,
+      "loss": 0.2256,
+      "step": 14939
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.2811708425865105,
+      "learning_rate": 7.300004735125649e-07,
+      "loss": 0.1767,
+      "step": 14940
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.5355230189342957,
+      "learning_rate": 7.295318594613254e-07,
+      "loss": 0.1574,
+      "step": 14941
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.0599734334945394,
+      "learning_rate": 7.290633840319361e-07,
+      "loss": 0.1398,
+      "step": 14942
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.3117715539806327,
+      "learning_rate": 7.285950472396059e-07,
+      "loss": 0.1988,
+      "step": 14943
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.1910609886310777,
+      "learning_rate": 7.281268490995357e-07,
+      "loss": 0.1529,
+      "step": 14944
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.0945017904609622,
+      "learning_rate": 7.276587896269266e-07,
+      "loss": 0.1431,
+      "step": 14945
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.3072623468685203,
+      "learning_rate": 7.271908688369683e-07,
+      "loss": 0.1694,
+      "step": 14946
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.2846081733199204,
+      "learning_rate": 7.267230867448516e-07,
+      "loss": 0.2031,
+      "step": 14947
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.586057639621783,
+      "learning_rate": 7.262554433657609e-07,
+      "loss": 0.1649,
+      "step": 14948
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.192901891997061,
+      "learning_rate": 7.257879387148759e-07,
+      "loss": 0.1486,
+      "step": 14949
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.9894351480193524,
+      "learning_rate": 7.253205728073736e-07,
+      "loss": 0.1529,
+      "step": 14950
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.5893102402118768,
+      "learning_rate": 7.248533456584222e-07,
+      "loss": 0.1987,
+      "step": 14951
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.118419401507886,
+      "learning_rate": 7.243862572831911e-07,
+      "loss": 0.158,
+      "step": 14952
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.164556475291537,
+      "learning_rate": 7.239193076968393e-07,
+      "loss": 0.1761,
+      "step": 14953
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.4759123899919362,
+      "learning_rate": 7.234524969145267e-07,
+      "loss": 0.1684,
+      "step": 14954
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.5567900754629154,
+      "learning_rate": 7.229858249514044e-07,
+      "loss": 0.21,
+      "step": 14955
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.7630946776863416,
+      "learning_rate": 7.225192918226215e-07,
+      "loss": 0.1442,
+      "step": 14956
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.051374980281811,
+      "learning_rate": 7.220528975433221e-07,
+      "loss": 0.1521,
+      "step": 14957
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.3399853423014383,
+      "learning_rate": 7.21586642128646e-07,
+      "loss": 0.173,
+      "step": 14958
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.2665235293901629,
+      "learning_rate": 7.211205255937287e-07,
+      "loss": 0.1589,
+      "step": 14959
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 2.7357036959070737,
+      "learning_rate": 7.206545479536986e-07,
+      "loss": 0.1603,
+      "step": 14960
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.6215943999652962,
+      "learning_rate": 7.201887092236826e-07,
+      "loss": 0.1758,
+      "step": 14961
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.3858419198297585,
+      "learning_rate": 7.197230094188018e-07,
+      "loss": 0.1545,
+      "step": 14962
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.6407365897239619,
+      "learning_rate": 7.192574485541754e-07,
+      "loss": 0.204,
+      "step": 14963
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.9771323105683897,
+      "learning_rate": 7.187920266449122e-07,
+      "loss": 0.1191,
+      "step": 14964
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.258321718242335,
+      "learning_rate": 7.183267437061214e-07,
+      "loss": 0.1573,
+      "step": 14965
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.779248418053935,
+      "learning_rate": 7.17861599752907e-07,
+      "loss": 0.1397,
+      "step": 14966
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.6224336257459804,
+      "learning_rate": 7.173965948003675e-07,
+      "loss": 0.2445,
+      "step": 14967
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.6882592630079982,
+      "learning_rate": 7.169317288635979e-07,
+      "loss": 0.2007,
+      "step": 14968
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.5397450293990131,
+      "learning_rate": 7.164670019576869e-07,
+      "loss": 0.184,
+      "step": 14969
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.3505500959714443,
+      "learning_rate": 7.160024140977207e-07,
+      "loss": 0.1658,
+      "step": 14970
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.1419341465185322,
+      "learning_rate": 7.155379652987787e-07,
+      "loss": 0.2,
+      "step": 14971
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.7296775404067035,
+      "learning_rate": 7.150736555759396e-07,
+      "loss": 0.206,
+      "step": 14972
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.5965604498434354,
+      "learning_rate": 7.14609484944272e-07,
+      "loss": 0.1698,
+      "step": 14973
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.7013721851796157,
+      "learning_rate": 7.141454534188452e-07,
+      "loss": 0.1827,
+      "step": 14974
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.7151426432819514,
+      "learning_rate": 7.136815610147213e-07,
+      "loss": 0.1736,
+      "step": 14975
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 2.1917542505955154,
+      "learning_rate": 7.132178077469587e-07,
+      "loss": 0.1851,
+      "step": 14976
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.9915189706016222,
+      "learning_rate": 7.127541936306126e-07,
+      "loss": 0.1647,
+      "step": 14977
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.3362158546688487,
+      "learning_rate": 7.122907186807293e-07,
+      "loss": 0.181,
+      "step": 14978
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.9755375767082394,
+      "learning_rate": 7.118273829123546e-07,
+      "loss": 0.1634,
+      "step": 14979
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.5791625776045035,
+      "learning_rate": 7.11364186340529e-07,
+      "loss": 0.1585,
+      "step": 14980
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.0690243933441976,
+      "learning_rate": 7.109011289802898e-07,
+      "loss": 0.1531,
+      "step": 14981
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 2.362126016029874,
+      "learning_rate": 7.104382108466645e-07,
+      "loss": 0.1784,
+      "step": 14982
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.279822345005565,
+      "learning_rate": 7.099754319546819e-07,
+      "loss": 0.1706,
+      "step": 14983
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.3796092475611526,
+      "learning_rate": 7.09512792319364e-07,
+      "loss": 0.1968,
+      "step": 14984
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.206215280196226,
+      "learning_rate": 7.090502919557279e-07,
+      "loss": 0.1779,
+      "step": 14985
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.257516254463142,
+      "learning_rate": 7.085879308787874e-07,
+      "loss": 0.197,
+      "step": 14986
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.1107609613937213,
+      "learning_rate": 7.081257091035498e-07,
+      "loss": 0.1088,
+      "step": 14987
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.2241524204282572,
+      "learning_rate": 7.07663626645021e-07,
+      "loss": 0.1877,
+      "step": 14988
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.7742431004011,
+      "learning_rate": 7.072016835181978e-07,
+      "loss": 0.2059,
+      "step": 14989
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.3698715647783082,
+      "learning_rate": 7.06739879738077e-07,
+      "loss": 0.205,
+      "step": 14990
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.3950428331543294,
+      "learning_rate": 7.062782153196479e-07,
+      "loss": 0.148,
+      "step": 14991
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.7177672264069241,
+      "learning_rate": 7.058166902778974e-07,
+      "loss": 0.1974,
+      "step": 14992
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.3474734914946565,
+      "learning_rate": 7.053553046278061e-07,
+      "loss": 0.1544,
+      "step": 14993
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.3582598507883965,
+      "learning_rate": 7.048940583843511e-07,
+      "loss": 0.1774,
+      "step": 14994
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.6503020494155811,
+      "learning_rate": 7.044329515625059e-07,
+      "loss": 0.1905,
+      "step": 14995
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.8036765780478268,
+      "learning_rate": 7.039719841772363e-07,
+      "loss": 0.1999,
+      "step": 14996
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.103310017293774,
+      "learning_rate": 7.035111562435066e-07,
+      "loss": 0.1602,
+      "step": 14997
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.3407596593864863,
+      "learning_rate": 7.030504677762751e-07,
+      "loss": 0.1637,
+      "step": 14998
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.3642284523952382,
+      "learning_rate": 7.025899187904972e-07,
+      "loss": 0.1924,
+      "step": 14999
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.4958829706380832,
+      "learning_rate": 7.02129509301121e-07,
+      "loss": 0.1711,
+      "step": 15000
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.5394165580644295,
+      "learning_rate": 7.016692393230929e-07,
+      "loss": 0.1961,
+      "step": 15001
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.4705486697771852,
+      "learning_rate": 7.012091088713524e-07,
+      "loss": 0.1352,
+      "step": 15002
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.6097249881953577,
+      "learning_rate": 7.007491179608362e-07,
+      "loss": 0.1923,
+      "step": 15003
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.1931942546132372,
+      "learning_rate": 7.00289266606477e-07,
+      "loss": 0.1399,
+      "step": 15004
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.1647395302901113,
+      "learning_rate": 6.998295548231998e-07,
+      "loss": 0.164,
+      "step": 15005
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.5186326476075749,
+      "learning_rate": 6.993699826259287e-07,
+      "loss": 0.1742,
+      "step": 15006
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.3277369549404396,
+      "learning_rate": 6.989105500295807e-07,
+      "loss": 0.1673,
+      "step": 15007
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.2113022285751442,
+      "learning_rate": 6.984512570490697e-07,
+      "loss": 0.1442,
+      "step": 15008
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.3788467723997548,
+      "learning_rate": 6.979921036993042e-07,
+      "loss": 0.1535,
+      "step": 15009
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.2404207164720504,
+      "learning_rate": 6.975330899951882e-07,
+      "loss": 0.1846,
+      "step": 15010
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.318493447727259,
+      "learning_rate": 6.970742159516225e-07,
+      "loss": 0.206,
+      "step": 15011
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.0076535750176285,
+      "learning_rate": 6.966154815835019e-07,
+      "loss": 0.1463,
+      "step": 15012
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.2240745678624039,
+      "learning_rate": 6.961568869057189e-07,
+      "loss": 0.1687,
+      "step": 15013
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.221827835748376,
+      "learning_rate": 6.956984319331572e-07,
+      "loss": 0.1368,
+      "step": 15014
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.4184562223501083,
+      "learning_rate": 6.952401166806988e-07,
+      "loss": 0.1615,
+      "step": 15015
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.5384336193639399,
+      "learning_rate": 6.947819411632223e-07,
+      "loss": 0.1769,
+      "step": 15016
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.3624436502440782,
+      "learning_rate": 6.943239053956008e-07,
+      "loss": 0.1626,
+      "step": 15017
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.3967848207819977,
+      "learning_rate": 6.938660093926997e-07,
+      "loss": 0.2239,
+      "step": 15018
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.1968913963629024,
+      "learning_rate": 6.934082531693842e-07,
+      "loss": 0.1409,
+      "step": 15019
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.2786361502328014,
+      "learning_rate": 6.929506367405136e-07,
+      "loss": 0.1963,
+      "step": 15020
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.4204195454360258,
+      "learning_rate": 6.924931601209417e-07,
+      "loss": 0.2095,
+      "step": 15021
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.2508292487404142,
+      "learning_rate": 6.920358233255203e-07,
+      "loss": 0.1763,
+      "step": 15022
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.2931596148353492,
+      "learning_rate": 6.915786263690921e-07,
+      "loss": 0.2413,
+      "step": 15023
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.3073090068330768,
+      "learning_rate": 6.911215692665002e-07,
+      "loss": 0.1807,
+      "step": 15024
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.6664799337286325,
+      "learning_rate": 6.906646520325788e-07,
+      "loss": 0.1836,
+      "step": 15025
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.496514517675464,
+      "learning_rate": 6.902078746821622e-07,
+      "loss": 0.1818,
+      "step": 15026
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.239668757703501,
+      "learning_rate": 6.897512372300752e-07,
+      "loss": 0.1774,
+      "step": 15027
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.3740034935648617,
+      "learning_rate": 6.89294739691141e-07,
+      "loss": 0.1772,
+      "step": 15028
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.7144564667028375,
+      "learning_rate": 6.888383820801792e-07,
+      "loss": 0.172,
+      "step": 15029
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 3.5296267432885253,
+      "learning_rate": 6.883821644120021e-07,
+      "loss": 0.2065,
+      "step": 15030
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.1155860500054466,
+      "learning_rate": 6.879260867014203e-07,
+      "loss": 0.159,
+      "step": 15031
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.3096619075458247,
+      "learning_rate": 6.874701489632369e-07,
+      "loss": 0.1687,
+      "step": 15032
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.4753717535911266,
+      "learning_rate": 6.870143512122518e-07,
+      "loss": 0.1922,
+      "step": 15033
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.4292181891513958,
+      "learning_rate": 6.865586934632612e-07,
+      "loss": 0.1534,
+      "step": 15034
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.1774752760476246,
+      "learning_rate": 6.861031757310565e-07,
+      "loss": 0.1491,
+      "step": 15035
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.2999329249335763,
+      "learning_rate": 6.856477980304227e-07,
+      "loss": 0.1979,
+      "step": 15036
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.6761915864943362,
+      "learning_rate": 6.851925603761423e-07,
+      "loss": 0.2594,
+      "step": 15037
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.390141928274123,
+      "learning_rate": 6.847374627829928e-07,
+      "loss": 0.1622,
+      "step": 15038
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.0720206538461914,
+      "learning_rate": 6.842825052657459e-07,
+      "loss": 0.201,
+      "step": 15039
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.6425321846977443,
+      "learning_rate": 6.838276878391726e-07,
+      "loss": 0.1294,
+      "step": 15040
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.7580068438437981,
+      "learning_rate": 6.833730105180331e-07,
+      "loss": 0.2187,
+      "step": 15041
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.671001165856172,
+      "learning_rate": 6.829184733170891e-07,
+      "loss": 0.1982,
+      "step": 15042
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.3867003856033941,
+      "learning_rate": 6.824640762510926e-07,
+      "loss": 0.2153,
+      "step": 15043
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.729021229047162,
+      "learning_rate": 6.820098193347963e-07,
+      "loss": 0.2382,
+      "step": 15044
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.3518453864259183,
+      "learning_rate": 6.815557025829428e-07,
+      "loss": 0.2192,
+      "step": 15045
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.052700268647656,
+      "learning_rate": 6.811017260102748e-07,
+      "loss": 0.1402,
+      "step": 15046
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.9943183498412136,
+      "learning_rate": 6.806478896315282e-07,
+      "loss": 0.1484,
+      "step": 15047
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.489142931900411,
+      "learning_rate": 6.801941934614348e-07,
+      "loss": 0.184,
+      "step": 15048
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.6103427116377518,
+      "learning_rate": 6.79740637514723e-07,
+      "loss": 0.1853,
+      "step": 15049
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.3385211238571124,
+      "learning_rate": 6.792872218061137e-07,
+      "loss": 0.1532,
+      "step": 15050
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.6896729078274786,
+      "learning_rate": 6.788339463503257e-07,
+      "loss": 0.1569,
+      "step": 15051
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.955417112064155,
+      "learning_rate": 6.783808111620727e-07,
+      "loss": 0.1701,
+      "step": 15052
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.9366575686142036,
+      "learning_rate": 6.779278162560643e-07,
+      "loss": 0.1871,
+      "step": 15053
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.4030169701353703,
+      "learning_rate": 6.774749616470039e-07,
+      "loss": 0.2007,
+      "step": 15054
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.5498588182386213,
+      "learning_rate": 6.770222473495913e-07,
+      "loss": 0.1838,
+      "step": 15055
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.3905042317273582,
+      "learning_rate": 6.765696733785226e-07,
+      "loss": 0.2008,
+      "step": 15056
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.0982244453170693,
+      "learning_rate": 6.761172397484889e-07,
+      "loss": 0.1556,
+      "step": 15057
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.1369214450719034,
+      "learning_rate": 6.756649464741766e-07,
+      "loss": 0.1226,
+      "step": 15058
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.5127784307213599,
+      "learning_rate": 6.752127935702657e-07,
+      "loss": 0.2075,
+      "step": 15059
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.4918459955855115,
+      "learning_rate": 6.747607810514362e-07,
+      "loss": 0.1815,
+      "step": 15060
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.4646570213830392,
+      "learning_rate": 6.743089089323573e-07,
+      "loss": 0.1635,
+      "step": 15061
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.9830823427620425,
+      "learning_rate": 6.738571772276997e-07,
+      "loss": 0.2629,
+      "step": 15062
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.305364692850944,
+      "learning_rate": 6.734055859521249e-07,
+      "loss": 0.1678,
+      "step": 15063
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.3050353163383637,
+      "learning_rate": 6.729541351202929e-07,
+      "loss": 0.1942,
+      "step": 15064
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.6112039229179989,
+      "learning_rate": 6.725028247468573e-07,
+      "loss": 0.1419,
+      "step": 15065
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.7479227416221284,
+      "learning_rate": 6.72051654846469e-07,
+      "loss": 0.1755,
+      "step": 15066
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.6092056166759927,
+      "learning_rate": 6.716006254337737e-07,
+      "loss": 0.2004,
+      "step": 15067
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.8932684238846569,
+      "learning_rate": 6.711497365234104e-07,
+      "loss": 0.15,
+      "step": 15068
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.2364379960545868,
+      "learning_rate": 6.706989881300152e-07,
+      "loss": 0.1583,
+      "step": 15069
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.1372011609579058,
+      "learning_rate": 6.702483802682208e-07,
+      "loss": 0.1762,
+      "step": 15070
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.18982529729209,
+      "learning_rate": 6.69797912952655e-07,
+      "loss": 0.1637,
+      "step": 15071
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.0816688412684625,
+      "learning_rate": 6.693475861979376e-07,
+      "loss": 0.1419,
+      "step": 15072
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.3780593756969413,
+      "learning_rate": 6.688974000186876e-07,
+      "loss": 0.1891,
+      "step": 15073
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.2429316289167187,
+      "learning_rate": 6.684473544295195e-07,
+      "loss": 0.1441,
+      "step": 15074
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.5536955251834421,
+      "learning_rate": 6.679974494450403e-07,
+      "loss": 0.1405,
+      "step": 15075
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.9166691683324437,
+      "learning_rate": 6.675476850798562e-07,
+      "loss": 0.2033,
+      "step": 15076
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.5088198760506433,
+      "learning_rate": 6.670980613485644e-07,
+      "loss": 0.1499,
+      "step": 15077
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.6683229476821082,
+      "learning_rate": 6.666485782657628e-07,
+      "loss": 0.2351,
+      "step": 15078
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.4110217674412724,
+      "learning_rate": 6.661992358460384e-07,
+      "loss": 0.2192,
+      "step": 15079
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.3403622371973474,
+      "learning_rate": 6.657500341039802e-07,
+      "loss": 0.1945,
+      "step": 15080
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.6679487249525782,
+      "learning_rate": 6.653009730541676e-07,
+      "loss": 0.2076,
+      "step": 15081
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.2674564015115128,
+      "learning_rate": 6.648520527111773e-07,
+      "loss": 0.1898,
+      "step": 15082
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.882909305347365,
+      "learning_rate": 6.644032730895828e-07,
+      "loss": 0.1734,
+      "step": 15083
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.6932244604587618,
+      "learning_rate": 6.63954634203951e-07,
+      "loss": 0.1792,
+      "step": 15084
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.9658221165686192,
+      "learning_rate": 6.635061360688466e-07,
+      "loss": 0.1951,
+      "step": 15085
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.3801032428199467,
+      "learning_rate": 6.630577786988252e-07,
+      "loss": 0.1807,
+      "step": 15086
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.114211761031153,
+      "learning_rate": 6.626095621084422e-07,
+      "loss": 0.1797,
+      "step": 15087
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.4132403096289072,
+      "learning_rate": 6.621614863122472e-07,
+      "loss": 0.1805,
+      "step": 15088
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.394261881780929,
+      "learning_rate": 6.617135513247858e-07,
+      "loss": 0.1369,
+      "step": 15089
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.8698978863211966,
+      "learning_rate": 6.612657571605963e-07,
+      "loss": 0.2131,
+      "step": 15090
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.1619889269045136,
+      "learning_rate": 6.608181038342148e-07,
+      "loss": 0.1643,
+      "step": 15091
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.22364135520005,
+      "learning_rate": 6.603705913601732e-07,
+      "loss": 0.1962,
+      "step": 15092
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.1349593898255197,
+      "learning_rate": 6.599232197529981e-07,
+      "loss": 0.1389,
+      "step": 15093
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.7447436957146347,
+      "learning_rate": 6.594759890272112e-07,
+      "loss": 0.2008,
+      "step": 15094
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.2033272765158858,
+      "learning_rate": 6.590288991973292e-07,
+      "loss": 0.1931,
+      "step": 15095
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.6986729633261317,
+      "learning_rate": 6.585819502778662e-07,
+      "loss": 0.2003,
+      "step": 15096
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.8302107918610087,
+      "learning_rate": 6.581351422833287e-07,
+      "loss": 0.172,
+      "step": 15097
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.2930684468996752,
+      "learning_rate": 6.576884752282225e-07,
+      "loss": 0.1474,
+      "step": 15098
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.6533541284427533,
+      "learning_rate": 6.572419491270438e-07,
+      "loss": 0.1598,
+      "step": 15099
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.2673201805484395,
+      "learning_rate": 6.567955639942891e-07,
+      "loss": 0.1535,
+      "step": 15100
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.5719396976026203,
+      "learning_rate": 6.563493198444476e-07,
+      "loss": 0.191,
+      "step": 15101
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.0982570632939288,
+      "learning_rate": 6.559032166920054e-07,
+      "loss": 0.1154,
+      "step": 15102
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.6901709911221647,
+      "learning_rate": 6.554572545514432e-07,
+      "loss": 0.2449,
+      "step": 15103
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.8283244153806916,
+      "learning_rate": 6.550114334372366e-07,
+      "loss": 0.1248,
+      "step": 15104
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.8674652619239693,
+      "learning_rate": 6.545657533638567e-07,
+      "loss": 0.2352,
+      "step": 15105
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.6750159616919698,
+      "learning_rate": 6.541202143457714e-07,
+      "loss": 0.2112,
+      "step": 15106
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 2.590458631522777,
+      "learning_rate": 6.536748163974444e-07,
+      "loss": 0.2101,
+      "step": 15107
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.2378281988965214,
+      "learning_rate": 6.532295595333304e-07,
+      "loss": 0.147,
+      "step": 15108
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.6527085093288922,
+      "learning_rate": 6.527844437678854e-07,
+      "loss": 0.2134,
+      "step": 15109
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.3210547352900264,
+      "learning_rate": 6.523394691155566e-07,
+      "loss": 0.2059,
+      "step": 15110
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.8278701295553882,
+      "learning_rate": 6.518946355907896e-07,
+      "loss": 0.1351,
+      "step": 15111
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.5796779036112516,
+      "learning_rate": 6.514499432080223e-07,
+      "loss": 0.1589,
+      "step": 15112
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.4819531761522158,
+      "learning_rate": 6.510053919816905e-07,
+      "loss": 0.116,
+      "step": 15113
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.5859303309837631,
+      "learning_rate": 6.50560981926226e-07,
+      "loss": 0.157,
+      "step": 15114
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.7245207404130685,
+      "learning_rate": 6.501167130560515e-07,
+      "loss": 0.186,
+      "step": 15115
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.5826284152775127,
+      "learning_rate": 6.496725853855907e-07,
+      "loss": 0.2269,
+      "step": 15116
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.4623648764470767,
+      "learning_rate": 6.49228598929259e-07,
+      "loss": 0.1932,
+      "step": 15117
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.9849208612588849,
+      "learning_rate": 6.487847537014685e-07,
+      "loss": 0.1314,
+      "step": 15118
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.091239295483115,
+      "learning_rate": 6.483410497166271e-07,
+      "loss": 0.1482,
+      "step": 15119
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.6921942199122744,
+      "learning_rate": 6.478974869891375e-07,
+      "loss": 0.2075,
+      "step": 15120
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.3591387093921854,
+      "learning_rate": 6.474540655333988e-07,
+      "loss": 0.1988,
+      "step": 15121
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.9986197139687468,
+      "learning_rate": 6.470107853638036e-07,
+      "loss": 0.1448,
+      "step": 15122
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 2.205001045819335,
+      "learning_rate": 6.465676464947413e-07,
+      "loss": 0.2171,
+      "step": 15123
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.3083215800545058,
+      "learning_rate": 6.461246489405964e-07,
+      "loss": 0.1803,
+      "step": 15124
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.604194887540547,
+      "learning_rate": 6.456817927157505e-07,
+      "loss": 0.2147,
+      "step": 15125
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.3517686212617648,
+      "learning_rate": 6.452390778345763e-07,
+      "loss": 0.1908,
+      "step": 15126
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.1580556547255054,
+      "learning_rate": 6.447965043114457e-07,
+      "loss": 0.1906,
+      "step": 15127
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.1704470072050681,
+      "learning_rate": 6.443540721607245e-07,
+      "loss": 0.1864,
+      "step": 15128
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 2.5002730220485794,
+      "learning_rate": 6.439117813967765e-07,
+      "loss": 0.1958,
+      "step": 15129
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.8793049823441008,
+      "learning_rate": 6.434696320339557e-07,
+      "loss": 0.1809,
+      "step": 15130
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.2630617829858328,
+      "learning_rate": 6.430276240866173e-07,
+      "loss": 0.145,
+      "step": 15131
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.2084245448654847,
+      "learning_rate": 6.425857575691064e-07,
+      "loss": 0.181,
+      "step": 15132
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.5085479889403424,
+      "learning_rate": 6.421440324957672e-07,
+      "loss": 0.2018,
+      "step": 15133
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 4.7780283782800055,
+      "learning_rate": 6.417024488809398e-07,
+      "loss": 0.1591,
+      "step": 15134
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.6536843923643836,
+      "learning_rate": 6.412610067389563e-07,
+      "loss": 0.1615,
+      "step": 15135
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.536828612844928,
+      "learning_rate": 6.408197060841465e-07,
+      "loss": 0.165,
+      "step": 15136
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.769066991808751,
+      "learning_rate": 6.40378546930836e-07,
+      "loss": 0.1866,
+      "step": 15137
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.8159469351903303,
+      "learning_rate": 6.399375292933452e-07,
+      "loss": 0.1751,
+      "step": 15138
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.5827366421927742,
+      "learning_rate": 6.394966531859903e-07,
+      "loss": 0.1891,
+      "step": 15139
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.5255067807956157,
+      "learning_rate": 6.390559186230799e-07,
+      "loss": 0.1363,
+      "step": 15140
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.639880729065704,
+      "learning_rate": 6.386153256189225e-07,
+      "loss": 0.172,
+      "step": 15141
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.3683029147910875,
+      "learning_rate": 6.381748741878196e-07,
+      "loss": 0.1656,
+      "step": 15142
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.1989695168870373,
+      "learning_rate": 6.377345643440696e-07,
+      "loss": 0.1568,
+      "step": 15143
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.201915355804292,
+      "learning_rate": 6.372943961019634e-07,
+      "loss": 0.2085,
+      "step": 15144
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.7645549992204752,
+      "learning_rate": 6.368543694757895e-07,
+      "loss": 0.2152,
+      "step": 15145
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.514706131545269,
+      "learning_rate": 6.364144844798314e-07,
+      "loss": 0.1733,
+      "step": 15146
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.1990291526348777,
+      "learning_rate": 6.359747411283701e-07,
+      "loss": 0.1559,
+      "step": 15147
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.6608124995452165,
+      "learning_rate": 6.355351394356768e-07,
+      "loss": 0.1793,
+      "step": 15148
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.8362574197140151,
+      "learning_rate": 6.35095679416024e-07,
+      "loss": 0.1585,
+      "step": 15149
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 2.459827215021244,
+      "learning_rate": 6.346563610836736e-07,
+      "loss": 0.1746,
+      "step": 15150
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.3470049569701423,
+      "learning_rate": 6.342171844528888e-07,
+      "loss": 0.177,
+      "step": 15151
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 2.1479024463581204,
+      "learning_rate": 6.337781495379252e-07,
+      "loss": 0.1786,
+      "step": 15152
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.4150290215703996,
+      "learning_rate": 6.333392563530328e-07,
+      "loss": 0.1748,
+      "step": 15153
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.5173323197789166,
+      "learning_rate": 6.329005049124593e-07,
+      "loss": 0.1989,
+      "step": 15154
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.4809404693593067,
+      "learning_rate": 6.324618952304463e-07,
+      "loss": 0.1708,
+      "step": 15155
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.2140378456116252,
+      "learning_rate": 6.320234273212323e-07,
+      "loss": 0.1833,
+      "step": 15156
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.2071605909361767,
+      "learning_rate": 6.315851011990499e-07,
+      "loss": 0.1865,
+      "step": 15157
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.4629444932378117,
+      "learning_rate": 6.311469168781265e-07,
+      "loss": 0.1559,
+      "step": 15158
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.3048884414074107,
+      "learning_rate": 6.307088743726864e-07,
+      "loss": 0.1934,
+      "step": 15159
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.6730434488051678,
+      "learning_rate": 6.302709736969486e-07,
+      "loss": 0.2063,
+      "step": 15160
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.4014760417064231,
+      "learning_rate": 6.298332148651282e-07,
+      "loss": 0.1606,
+      "step": 15161
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.9857089897885913,
+      "learning_rate": 6.293955978914351e-07,
+      "loss": 0.1704,
+      "step": 15162
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.903897772806523,
+      "learning_rate": 6.289581227900737e-07,
+      "loss": 0.2083,
+      "step": 15163
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.6987620693687495,
+      "learning_rate": 6.285207895752449e-07,
+      "loss": 0.2033,
+      "step": 15164
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.1409751211027526,
+      "learning_rate": 6.280835982611461e-07,
+      "loss": 0.1143,
+      "step": 15165
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.5500376862128349,
+      "learning_rate": 6.276465488619671e-07,
+      "loss": 0.1773,
+      "step": 15166
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.7796099290026755,
+      "learning_rate": 6.27209641391896e-07,
+      "loss": 0.2039,
+      "step": 15167
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.2424165526523563,
+      "learning_rate": 6.267728758651131e-07,
+      "loss": 0.1995,
+      "step": 15168
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.7785608672535917,
+      "learning_rate": 6.263362522957977e-07,
+      "loss": 0.2041,
+      "step": 15169
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.3309785816939845,
+      "learning_rate": 6.258997706981229e-07,
+      "loss": 0.1439,
+      "step": 15170
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 3.7671415195641473,
+      "learning_rate": 6.254634310862578e-07,
+      "loss": 0.2668,
+      "step": 15171
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.429950898301154,
+      "learning_rate": 6.25027233474364e-07,
+      "loss": 0.163,
+      "step": 15172
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.4215596132213493,
+      "learning_rate": 6.245911778766023e-07,
+      "loss": 0.1724,
+      "step": 15173
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.2238912895454175,
+      "learning_rate": 6.241552643071269e-07,
+      "loss": 0.1584,
+      "step": 15174
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.5707942071978098,
+      "learning_rate": 6.237194927800888e-07,
+      "loss": 0.2065,
+      "step": 15175
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.3485480854398895,
+      "learning_rate": 6.232838633096317e-07,
+      "loss": 0.1413,
+      "step": 15176
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.7613081887158133,
+      "learning_rate": 6.228483759098969e-07,
+      "loss": 0.2003,
+      "step": 15177
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.4203122322577464,
+      "learning_rate": 6.22413030595021e-07,
+      "loss": 0.163,
+      "step": 15178
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.3928227845039707,
+      "learning_rate": 6.219778273791349e-07,
+      "loss": 0.1615,
+      "step": 15179
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.712191827395743,
+      "learning_rate": 6.215427662763679e-07,
+      "loss": 0.1876,
+      "step": 15180
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.4698462554757812,
+      "learning_rate": 6.211078473008392e-07,
+      "loss": 0.234,
+      "step": 15181
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.096676662573375,
+      "learning_rate": 6.206730704666675e-07,
+      "loss": 0.1596,
+      "step": 15182
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 7.088252588376229,
+      "learning_rate": 6.202384357879676e-07,
+      "loss": 0.2037,
+      "step": 15183
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.493524371947308,
+      "learning_rate": 6.198039432788456e-07,
+      "loss": 0.1704,
+      "step": 15184
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.4179227259289082,
+      "learning_rate": 6.193695929534077e-07,
+      "loss": 0.1406,
+      "step": 15185
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.9605199361661643,
+      "learning_rate": 6.189353848257501e-07,
+      "loss": 0.1988,
+      "step": 15186
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.0363520930423453,
+      "learning_rate": 6.185013189099698e-07,
+      "loss": 0.1646,
+      "step": 15187
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.479240340650185,
+      "learning_rate": 6.18067395220156e-07,
+      "loss": 0.1944,
+      "step": 15188
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.6420531369455864,
+      "learning_rate": 6.176336137703953e-07,
+      "loss": 0.1769,
+      "step": 15189
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.4102477522140493,
+      "learning_rate": 6.171999745747664e-07,
+      "loss": 0.2328,
+      "step": 15190
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.4719812379145378,
+      "learning_rate": 6.16766477647347e-07,
+      "loss": 0.1687,
+      "step": 15191
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.27725016983167,
+      "learning_rate": 6.163331230022079e-07,
+      "loss": 0.1303,
+      "step": 15192
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.518277560641109,
+      "learning_rate": 6.15899910653418e-07,
+      "loss": 0.201,
+      "step": 15193
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.925715411939553,
+      "learning_rate": 6.154668406150366e-07,
+      "loss": 0.1092,
+      "step": 15194
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.5791629692032672,
+      "learning_rate": 6.150339129011229e-07,
+      "loss": 0.2115,
+      "step": 15195
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.391766567151833,
+      "learning_rate": 6.146011275257296e-07,
+      "loss": 0.2099,
+      "step": 15196
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.572647654874218,
+      "learning_rate": 6.141684845029056e-07,
+      "loss": 0.1319,
+      "step": 15197
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.415594061877636,
+      "learning_rate": 6.137359838466961e-07,
+      "loss": 0.1761,
+      "step": 15198
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.174612532174701,
+      "learning_rate": 6.133036255711377e-07,
+      "loss": 0.1472,
+      "step": 15199
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.3206290255014475,
+      "learning_rate": 6.12871409690266e-07,
+      "loss": 0.2332,
+      "step": 15200
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.3981395649140669,
+      "learning_rate": 6.12439336218112e-07,
+      "loss": 0.1793,
+      "step": 15201
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.3239879744648435,
+      "learning_rate": 6.120074051686992e-07,
+      "loss": 0.1758,
+      "step": 15202
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.696539250416723,
+      "learning_rate": 6.115756165560505e-07,
+      "loss": 0.2209,
+      "step": 15203
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.3400981539213186,
+      "learning_rate": 6.111439703941796e-07,
+      "loss": 0.195,
+      "step": 15204
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.9075600234420838,
+      "learning_rate": 6.107124666970993e-07,
+      "loss": 0.159,
+      "step": 15205
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.2343551265951664,
+      "learning_rate": 6.102811054788166e-07,
+      "loss": 0.1871,
+      "step": 15206
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.7587718888651105,
+      "learning_rate": 6.098498867533343e-07,
+      "loss": 0.1954,
+      "step": 15207
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.34541582381993,
+      "learning_rate": 6.094188105346477e-07,
+      "loss": 0.1921,
+      "step": 15208
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.3734045200489335,
+      "learning_rate": 6.089878768367518e-07,
+      "loss": 0.1122,
+      "step": 15209
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.2928705206952675,
+      "learning_rate": 6.085570856736345e-07,
+      "loss": 0.1367,
+      "step": 15210
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.5988488304797543,
+      "learning_rate": 6.081264370592805e-07,
+      "loss": 0.2573,
+      "step": 15211
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.5262949676903206,
+      "learning_rate": 6.076959310076664e-07,
+      "loss": 0.1573,
+      "step": 15212
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.083686278998738,
+      "learning_rate": 6.072655675327682e-07,
+      "loss": 0.1598,
+      "step": 15213
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.5220692040688908,
+      "learning_rate": 6.06835346648556e-07,
+      "loss": 0.1673,
+      "step": 15214
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.368276609138685,
+      "learning_rate": 6.064052683689947e-07,
+      "loss": 0.1825,
+      "step": 15215
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.3918442899264885,
+      "learning_rate": 6.059753327080458e-07,
+      "loss": 0.156,
+      "step": 15216
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.5117318993127147,
+      "learning_rate": 6.055455396796633e-07,
+      "loss": 0.2161,
+      "step": 15217
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.8143009659908658,
+      "learning_rate": 6.051158892977993e-07,
+      "loss": 0.2059,
+      "step": 15218
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.4070596006884857,
+      "learning_rate": 6.046863815764026e-07,
+      "loss": 0.2191,
+      "step": 15219
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.4715500023295562,
+      "learning_rate": 6.042570165294115e-07,
+      "loss": 0.2184,
+      "step": 15220
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.2716725996778149,
+      "learning_rate": 6.038277941707671e-07,
+      "loss": 0.1825,
+      "step": 15221
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.542363630013449,
+      "learning_rate": 6.033987145143988e-07,
+      "loss": 0.1654,
+      "step": 15222
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.4803612954390264,
+      "learning_rate": 6.029697775742371e-07,
+      "loss": 0.1543,
+      "step": 15223
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.6293921992188372,
+      "learning_rate": 6.025409833642043e-07,
+      "loss": 0.1752,
+      "step": 15224
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.7071211330746827,
+      "learning_rate": 6.021123318982213e-07,
+      "loss": 0.1807,
+      "step": 15225
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.1349504257973144,
+      "learning_rate": 6.016838231902001e-07,
+      "loss": 0.156,
+      "step": 15226
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.3775189123352554,
+      "learning_rate": 6.012554572540508e-07,
+      "loss": 0.1773,
+      "step": 15227
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.1600556763490997,
+      "learning_rate": 6.008272341036792e-07,
+      "loss": 0.1865,
+      "step": 15228
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.6708132411655623,
+      "learning_rate": 6.003991537529857e-07,
+      "loss": 0.2041,
+      "step": 15229
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.582562585778786,
+      "learning_rate": 5.999712162158649e-07,
+      "loss": 0.1228,
+      "step": 15230
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.459171270428669,
+      "learning_rate": 5.99543421506209e-07,
+      "loss": 0.1931,
+      "step": 15231
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.5478563374827767,
+      "learning_rate": 5.991157696379035e-07,
+      "loss": 0.1964,
+      "step": 15232
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.4245536588291459,
+      "learning_rate": 5.986882606248312e-07,
+      "loss": 0.1939,
+      "step": 15233
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.1933544083911791,
+      "learning_rate": 5.982608944808699e-07,
+      "loss": 0.099,
+      "step": 15234
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.4289218436631779,
+      "learning_rate": 5.978336712198902e-07,
+      "loss": 0.152,
+      "step": 15235
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.321334804242185,
+      "learning_rate": 5.974065908557608e-07,
+      "loss": 0.1746,
+      "step": 15236
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 2.1672528399386044,
+      "learning_rate": 5.969796534023459e-07,
+      "loss": 0.2526,
+      "step": 15237
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.3213970199430398,
+      "learning_rate": 5.965528588735025e-07,
+      "loss": 0.1697,
+      "step": 15238
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.5119027511238599,
+      "learning_rate": 5.96126207283087e-07,
+      "loss": 0.1755,
+      "step": 15239
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 2.2622964150950113,
+      "learning_rate": 5.95699698644946e-07,
+      "loss": 0.1823,
+      "step": 15240
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.305563541214894,
+      "learning_rate": 5.952733329729249e-07,
+      "loss": 0.2065,
+      "step": 15241
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.274546105570779,
+      "learning_rate": 5.948471102808645e-07,
+      "loss": 0.1586,
+      "step": 15242
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 2.027629130275361,
+      "learning_rate": 5.944210305826015e-07,
+      "loss": 0.1661,
+      "step": 15243
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.412574955124582,
+      "learning_rate": 5.939950938919636e-07,
+      "loss": 0.1807,
+      "step": 15244
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.2392529426641352,
+      "learning_rate": 5.935693002227788e-07,
+      "loss": 0.1585,
+      "step": 15245
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.8576510274991591,
+      "learning_rate": 5.931436495888688e-07,
+      "loss": 0.2187,
+      "step": 15246
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.6954904981418344,
+      "learning_rate": 5.927181420040506e-07,
+      "loss": 0.1978,
+      "step": 15247
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.917462245154183,
+      "learning_rate": 5.922927774821352e-07,
+      "loss": 0.2122,
+      "step": 15248
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.2673323559075425,
+      "learning_rate": 5.918675560369308e-07,
+      "loss": 0.162,
+      "step": 15249
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.4341107405534912,
+      "learning_rate": 5.914424776822402e-07,
+      "loss": 0.1682,
+      "step": 15250
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.1695202693793223,
+      "learning_rate": 5.91017542431862e-07,
+      "loss": 0.1897,
+      "step": 15251
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.33636152498354,
+      "learning_rate": 5.905927502995906e-07,
+      "loss": 0.1825,
+      "step": 15252
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.3949840889642031,
+      "learning_rate": 5.901681012992133e-07,
+      "loss": 0.2155,
+      "step": 15253
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.5865161995615231,
+      "learning_rate": 5.897435954445152e-07,
+      "loss": 0.1568,
+      "step": 15254
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.2160861754913663,
+      "learning_rate": 5.89319232749277e-07,
+      "loss": 0.1571,
+      "step": 15255
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.6609034393923543,
+      "learning_rate": 5.88895013227272e-07,
+      "loss": 0.1899,
+      "step": 15256
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.494058753868438,
+      "learning_rate": 5.88470936892272e-07,
+      "loss": 0.1522,
+      "step": 15257
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.680469339850343,
+      "learning_rate": 5.880470037580416e-07,
+      "loss": 0.1713,
+      "step": 15258
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.4625003533485181,
+      "learning_rate": 5.876232138383425e-07,
+      "loss": 0.1827,
+      "step": 15259
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.6026349315060362,
+      "learning_rate": 5.871995671469311e-07,
+      "loss": 0.1953,
+      "step": 15260
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.2338703427664042,
+      "learning_rate": 5.8677606369756e-07,
+      "loss": 0.1965,
+      "step": 15261
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.3624788672006978,
+      "learning_rate": 5.863527035039751e-07,
+      "loss": 0.2183,
+      "step": 15262
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.1944370710421957,
+      "learning_rate": 5.859294865799192e-07,
+      "loss": 0.1557,
+      "step": 15263
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.4486951781919075,
+      "learning_rate": 5.855064129391302e-07,
+      "loss": 0.1757,
+      "step": 15264
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.7106865041210804,
+      "learning_rate": 5.850834825953433e-07,
+      "loss": 0.2091,
+      "step": 15265
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.4438778122973686,
+      "learning_rate": 5.846606955622841e-07,
+      "loss": 0.1972,
+      "step": 15266
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.644498090794813,
+      "learning_rate": 5.842380518536772e-07,
+      "loss": 0.1905,
+      "step": 15267
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.2272449769328944,
+      "learning_rate": 5.838155514832428e-07,
+      "loss": 0.1864,
+      "step": 15268
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.4266572273198739,
+      "learning_rate": 5.833931944646947e-07,
+      "loss": 0.1626,
+      "step": 15269
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.3300857319663335,
+      "learning_rate": 5.829709808117451e-07,
+      "loss": 0.1682,
+      "step": 15270
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.6720254063911872,
+      "learning_rate": 5.825489105380955e-07,
+      "loss": 0.1861,
+      "step": 15271
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.9944048663382933,
+      "learning_rate": 5.821269836574489e-07,
+      "loss": 0.2018,
+      "step": 15272
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.403441666125695,
+      "learning_rate": 5.817052001835016e-07,
+      "loss": 0.1909,
+      "step": 15273
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.742184455082033,
+      "learning_rate": 5.812835601299438e-07,
+      "loss": 0.2405,
+      "step": 15274
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.2698507053752246,
+      "learning_rate": 5.808620635104634e-07,
+      "loss": 0.1232,
+      "step": 15275
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.6970477606361212,
+      "learning_rate": 5.8044071033874e-07,
+      "loss": 0.1677,
+      "step": 15276
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.6413608127135113,
+      "learning_rate": 5.800195006284531e-07,
+      "loss": 0.1585,
+      "step": 15277
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.6399630346435412,
+      "learning_rate": 5.795984343932748e-07,
+      "loss": 0.2297,
+      "step": 15278
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.5935062193954153,
+      "learning_rate": 5.791775116468739e-07,
+      "loss": 0.2396,
+      "step": 15279
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.689677908171543,
+      "learning_rate": 5.787567324029125e-07,
+      "loss": 0.1981,
+      "step": 15280
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.4521346973418183,
+      "learning_rate": 5.783360966750501e-07,
+      "loss": 0.1673,
+      "step": 15281
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.9100564255793047,
+      "learning_rate": 5.779156044769402e-07,
+      "loss": 0.178,
+      "step": 15282
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.9570681331300341,
+      "learning_rate": 5.774952558222341e-07,
+      "loss": 0.1731,
+      "step": 15283
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.4829403519657067,
+      "learning_rate": 5.770750507245737e-07,
+      "loss": 0.1887,
+      "step": 15284
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.517516980602526,
+      "learning_rate": 5.766549891976008e-07,
+      "loss": 0.188,
+      "step": 15285
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.3199234902901769,
+      "learning_rate": 5.762350712549503e-07,
+      "loss": 0.2028,
+      "step": 15286
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.3547057442058068,
+      "learning_rate": 5.758152969102538e-07,
+      "loss": 0.1732,
+      "step": 15287
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.2967841277400016,
+      "learning_rate": 5.75395666177137e-07,
+      "loss": 0.2033,
+      "step": 15288
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.3334635476017163,
+      "learning_rate": 5.74976179069221e-07,
+      "loss": 0.1416,
+      "step": 15289
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.415012893803355,
+      "learning_rate": 5.745568356001224e-07,
+      "loss": 0.2082,
+      "step": 15290
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.2460602304380834,
+      "learning_rate": 5.74137635783455e-07,
+      "loss": 0.17,
+      "step": 15291
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.001584555967764,
+      "learning_rate": 5.737185796328238e-07,
+      "loss": 0.1414,
+      "step": 15292
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.0987523128368004,
+      "learning_rate": 5.732996671618335e-07,
+      "loss": 0.1554,
+      "step": 15293
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.2094944200032214,
+      "learning_rate": 5.728808983840812e-07,
+      "loss": 0.2317,
+      "step": 15294
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.7224540418769525,
+      "learning_rate": 5.724622733131602e-07,
+      "loss": 0.2106,
+      "step": 15295
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.381924424859774,
+      "learning_rate": 5.720437919626598e-07,
+      "loss": 0.1245,
+      "step": 15296
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.3715011230593488,
+      "learning_rate": 5.716254543461653e-07,
+      "loss": 0.1958,
+      "step": 15297
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.1986220114564572,
+      "learning_rate": 5.712072604772539e-07,
+      "loss": 0.1493,
+      "step": 15298
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.719565505848209,
+      "learning_rate": 5.707892103695017e-07,
+      "loss": 0.1599,
+      "step": 15299
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.2970733490896733,
+      "learning_rate": 5.703713040364789e-07,
+      "loss": 0.1725,
+      "step": 15300
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.93192052404351,
+      "learning_rate": 5.699535414917517e-07,
+      "loss": 0.185,
+      "step": 15301
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.2748281690335255,
+      "learning_rate": 5.69535922748879e-07,
+      "loss": 0.1638,
+      "step": 15302
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.5843334542419631,
+      "learning_rate": 5.691184478214179e-07,
+      "loss": 0.1745,
+      "step": 15303
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.8854277998093424,
+      "learning_rate": 5.6870111672292e-07,
+      "loss": 0.1472,
+      "step": 15304
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.1713679392852898,
+      "learning_rate": 5.682839294669317e-07,
+      "loss": 0.1664,
+      "step": 15305
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.3127964684030335,
+      "learning_rate": 5.678668860669966e-07,
+      "loss": 0.1659,
+      "step": 15306
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.4620520221489839,
+      "learning_rate": 5.674499865366507e-07,
+      "loss": 0.213,
+      "step": 15307
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.209813534653569,
+      "learning_rate": 5.670332308894261e-07,
+      "loss": 0.1365,
+      "step": 15308
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.2412802904203637,
+      "learning_rate": 5.666166191388533e-07,
+      "loss": 0.1883,
+      "step": 15309
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.5805524578545527,
+      "learning_rate": 5.662001512984539e-07,
+      "loss": 0.1864,
+      "step": 15310
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.367133534592553,
+      "learning_rate": 5.657838273817478e-07,
+      "loss": 0.1804,
+      "step": 15311
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.248722329429782,
+      "learning_rate": 5.653676474022473e-07,
+      "loss": 0.1795,
+      "step": 15312
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.2913456493543083,
+      "learning_rate": 5.649516113734638e-07,
+      "loss": 0.1674,
+      "step": 15313
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.3565678963359435,
+      "learning_rate": 5.645357193089008e-07,
+      "loss": 0.1438,
+      "step": 15314
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.3343282650082315,
+      "learning_rate": 5.641199712220602e-07,
+      "loss": 0.1918,
+      "step": 15315
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.5467973940099085,
+      "learning_rate": 5.637043671264348e-07,
+      "loss": 0.2053,
+      "step": 15316
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.137785826917568,
+      "learning_rate": 5.632889070355169e-07,
+      "loss": 0.1689,
+      "step": 15317
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.2760225281610118,
+      "learning_rate": 5.628735909627919e-07,
+      "loss": 0.2003,
+      "step": 15318
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.3101273162139984,
+      "learning_rate": 5.624584189217435e-07,
+      "loss": 0.1482,
+      "step": 15319
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.4018343316334547,
+      "learning_rate": 5.62043390925845e-07,
+      "loss": 0.2101,
+      "step": 15320
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.3850683196204483,
+      "learning_rate": 5.616285069885702e-07,
+      "loss": 0.1961,
+      "step": 15321
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.1307181630887,
+      "learning_rate": 5.612137671233864e-07,
+      "loss": 0.2112,
+      "step": 15322
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.5795257366679292,
+      "learning_rate": 5.607991713437561e-07,
+      "loss": 0.1848,
+      "step": 15323
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.409909303340349,
+      "learning_rate": 5.603847196631385e-07,
+      "loss": 0.1577,
+      "step": 15324
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.7207382666195967,
+      "learning_rate": 5.599704120949851e-07,
+      "loss": 0.1959,
+      "step": 15325
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.4354852883758646,
+      "learning_rate": 5.595562486527445e-07,
+      "loss": 0.1525,
+      "step": 15326
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.4597415618913927,
+      "learning_rate": 5.591422293498633e-07,
+      "loss": 0.1765,
+      "step": 15327
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.2609317557240693,
+      "learning_rate": 5.587283541997779e-07,
+      "loss": 0.1791,
+      "step": 15328
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.3722531017032817,
+      "learning_rate": 5.583146232159248e-07,
+      "loss": 0.1425,
+      "step": 15329
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.2929030110683721,
+      "learning_rate": 5.579010364117321e-07,
+      "loss": 0.1395,
+      "step": 15330
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.7483846071780753,
+      "learning_rate": 5.574875938006263e-07,
+      "loss": 0.1739,
+      "step": 15331
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.1547271713957752,
+      "learning_rate": 5.570742953960279e-07,
+      "loss": 0.1503,
+      "step": 15332
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.8756313214804095,
+      "learning_rate": 5.566611412113537e-07,
+      "loss": 0.213,
+      "step": 15333
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.68465804604108,
+      "learning_rate": 5.562481312600132e-07,
+      "loss": 0.193,
+      "step": 15334
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.4745055509484117,
+      "learning_rate": 5.558352655554133e-07,
+      "loss": 0.1894,
+      "step": 15335
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.9197080692988253,
+      "learning_rate": 5.554225441109562e-07,
+      "loss": 0.1877,
+      "step": 15336
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.1832647590288983,
+      "learning_rate": 5.550099669400405e-07,
+      "loss": 0.1656,
+      "step": 15337
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.5030509101706042,
+      "learning_rate": 5.545975340560561e-07,
+      "loss": 0.1751,
+      "step": 15338
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.2454633942032949,
+      "learning_rate": 5.541852454723923e-07,
+      "loss": 0.1957,
+      "step": 15339
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.7483062977862704,
+      "learning_rate": 5.537731012024317e-07,
+      "loss": 0.2212,
+      "step": 15340
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.8986171904243871,
+      "learning_rate": 5.53361101259553e-07,
+      "loss": 0.2032,
+      "step": 15341
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.231695822783409,
+      "learning_rate": 5.529492456571306e-07,
+      "loss": 0.1573,
+      "step": 15342
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.5615931072521962,
+      "learning_rate": 5.525375344085327e-07,
+      "loss": 0.2079,
+      "step": 15343
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.8107625180640863,
+      "learning_rate": 5.521259675271235e-07,
+      "loss": 0.195,
+      "step": 15344
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.3778808282290962,
+      "learning_rate": 5.517145450262639e-07,
+      "loss": 0.209,
+      "step": 15345
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.8707459510780817,
+      "learning_rate": 5.513032669193074e-07,
+      "loss": 0.1987,
+      "step": 15346
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.5670832552711624,
+      "learning_rate": 5.50892133219606e-07,
+      "loss": 0.1859,
+      "step": 15347
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 2.000188364158968,
+      "learning_rate": 5.504811439405033e-07,
+      "loss": 0.2059,
+      "step": 15348
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.9033554410555196,
+      "learning_rate": 5.500702990953416e-07,
+      "loss": 0.1957,
+      "step": 15349
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.2004138353402218,
+      "learning_rate": 5.496595986974568e-07,
+      "loss": 0.1586,
+      "step": 15350
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 2.3513942227400415,
+      "learning_rate": 5.492490427601816e-07,
+      "loss": 0.1798,
+      "step": 15351
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.2379141753855474,
+      "learning_rate": 5.488386312968403e-07,
+      "loss": 0.1605,
+      "step": 15352
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.9974788928677287,
+      "learning_rate": 5.484283643207572e-07,
+      "loss": 0.2121,
+      "step": 15353
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.2192407897507875,
+      "learning_rate": 5.48018241845249e-07,
+      "loss": 0.1549,
+      "step": 15354
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.3891459968797277,
+      "learning_rate": 5.476082638836294e-07,
+      "loss": 0.1357,
+      "step": 15355
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.131031517406737,
+      "learning_rate": 5.471984304492057e-07,
+      "loss": 0.187,
+      "step": 15356
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.649502950004427,
+      "learning_rate": 5.467887415552808e-07,
+      "loss": 0.2389,
+      "step": 15357
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.6488091470204147,
+      "learning_rate": 5.463791972151544e-07,
+      "loss": 0.2143,
+      "step": 15358
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.3308488066678859,
+      "learning_rate": 5.459697974421202e-07,
+      "loss": 0.2243,
+      "step": 15359
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.3218942568663683,
+      "learning_rate": 5.455605422494686e-07,
+      "loss": 0.1789,
+      "step": 15360
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.476510046980934,
+      "learning_rate": 5.451514316504824e-07,
+      "loss": 0.2268,
+      "step": 15361
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.838826758962836,
+      "learning_rate": 5.447424656584422e-07,
+      "loss": 0.188,
+      "step": 15362
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.5148860024456272,
+      "learning_rate": 5.443336442866248e-07,
+      "loss": 0.1909,
+      "step": 15363
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.5416835840903644,
+      "learning_rate": 5.439249675482982e-07,
+      "loss": 0.1835,
+      "step": 15364
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.9243118684875369,
+      "learning_rate": 5.435164354567307e-07,
+      "loss": 0.1882,
+      "step": 15365
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.5924745182135078,
+      "learning_rate": 5.431080480251816e-07,
+      "loss": 0.2344,
+      "step": 15366
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.158171970147829,
+      "learning_rate": 5.426998052669081e-07,
+      "loss": 0.1701,
+      "step": 15367
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.6390902743245503,
+      "learning_rate": 5.422917071951622e-07,
+      "loss": 0.1683,
+      "step": 15368
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.4792466970293705,
+      "learning_rate": 5.418837538231914e-07,
+      "loss": 0.1289,
+      "step": 15369
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.0781396298176074,
+      "learning_rate": 5.414759451642365e-07,
+      "loss": 0.1232,
+      "step": 15370
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.06204760962162,
+      "learning_rate": 5.410682812315366e-07,
+      "loss": 0.1176,
+      "step": 15371
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.7260195538660805,
+      "learning_rate": 5.406607620383241e-07,
+      "loss": 0.2612,
+      "step": 15372
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.2801125641404922,
+      "learning_rate": 5.402533875978289e-07,
+      "loss": 0.204,
+      "step": 15373
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.4959340570544484,
+      "learning_rate": 5.398461579232722e-07,
+      "loss": 0.2245,
+      "step": 15374
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.599782689804536,
+      "learning_rate": 5.394390730278742e-07,
+      "loss": 0.1917,
+      "step": 15375
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.3317032127633226,
+      "learning_rate": 5.390321329248482e-07,
+      "loss": 0.1861,
+      "step": 15376
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.3506095977939423,
+      "learning_rate": 5.386253376274047e-07,
+      "loss": 0.1202,
+      "step": 15377
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.8384642643255298,
+      "learning_rate": 5.382186871487494e-07,
+      "loss": 0.2249,
+      "step": 15378
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.3342640054490185,
+      "learning_rate": 5.378121815020804e-07,
+      "loss": 0.1668,
+      "step": 15379
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.4791600323470078,
+      "learning_rate": 5.374058207005945e-07,
+      "loss": 0.1922,
+      "step": 15380
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.2725885141377051,
+      "learning_rate": 5.369996047574811e-07,
+      "loss": 0.1531,
+      "step": 15381
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.584172803252186,
+      "learning_rate": 5.365935336859268e-07,
+      "loss": 0.1714,
+      "step": 15382
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.2910095509026973,
+      "learning_rate": 5.36187607499114e-07,
+      "loss": 0.1393,
+      "step": 15383
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.5447997926299817,
+      "learning_rate": 5.357818262102176e-07,
+      "loss": 0.1426,
+      "step": 15384
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.3823054887425028,
+      "learning_rate": 5.3537618983241e-07,
+      "loss": 0.1813,
+      "step": 15385
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.7558161186007437,
+      "learning_rate": 5.349706983788589e-07,
+      "loss": 0.208,
+      "step": 15386
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.7729658684438585,
+      "learning_rate": 5.345653518627269e-07,
+      "loss": 0.1644,
+      "step": 15387
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.2089340372036472,
+      "learning_rate": 5.34160150297171e-07,
+      "loss": 0.1667,
+      "step": 15388
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.7924337108211144,
+      "learning_rate": 5.337550936953445e-07,
+      "loss": 0.2333,
+      "step": 15389
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.5616858077665832,
+      "learning_rate": 5.333501820703957e-07,
+      "loss": 0.1757,
+      "step": 15390
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.4124934648894312,
+      "learning_rate": 5.329454154354696e-07,
+      "loss": 0.1633,
+      "step": 15391
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.9450666604486252,
+      "learning_rate": 5.32540793803703e-07,
+      "loss": 0.1582,
+      "step": 15392
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.4191128867452318,
+      "learning_rate": 5.321363171882316e-07,
+      "loss": 0.2089,
+      "step": 15393
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.7086467038754156,
+      "learning_rate": 5.317319856021841e-07,
+      "loss": 0.2261,
+      "step": 15394
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 3.049155658183556,
+      "learning_rate": 5.313277990586857e-07,
+      "loss": 0.2164,
+      "step": 15395
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.5674390170754224,
+      "learning_rate": 5.309237575708581e-07,
+      "loss": 0.2141,
+      "step": 15396
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.3532775295176915,
+      "learning_rate": 5.305198611518142e-07,
+      "loss": 0.1843,
+      "step": 15397
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.6771698913588076,
+      "learning_rate": 5.301161098146662e-07,
+      "loss": 0.1848,
+      "step": 15398
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.5335155095932085,
+      "learning_rate": 5.297125035725192e-07,
+      "loss": 0.1743,
+      "step": 15399
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.1711225382680974,
+      "learning_rate": 5.293090424384745e-07,
+      "loss": 0.1545,
+      "step": 15400
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.6416272099549116,
+      "learning_rate": 5.289057264256303e-07,
+      "loss": 0.1667,
+      "step": 15401
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.0759652591928004,
+      "learning_rate": 5.285025555470763e-07,
+      "loss": 0.1561,
+      "step": 15402
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.4507469592255655,
+      "learning_rate": 5.280995298159008e-07,
+      "loss": 0.1909,
+      "step": 15403
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.3994695754929873,
+      "learning_rate": 5.276966492451857e-07,
+      "loss": 0.1527,
+      "step": 15404
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.8000948543192554,
+      "learning_rate": 5.272939138480104e-07,
+      "loss": 0.168,
+      "step": 15405
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.510761899517451,
+      "learning_rate": 5.268913236374456e-07,
+      "loss": 0.1826,
+      "step": 15406
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.7320011492476495,
+      "learning_rate": 5.264888786265609e-07,
+      "loss": 0.2154,
+      "step": 15407
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.4670584352741123,
+      "learning_rate": 5.260865788284191e-07,
+      "loss": 0.1435,
+      "step": 15408
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.7074429432396696,
+      "learning_rate": 5.256844242560804e-07,
+      "loss": 0.1807,
+      "step": 15409
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.4241374690999038,
+      "learning_rate": 5.252824149225988e-07,
+      "loss": 0.1114,
+      "step": 15410
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.5037406433873486,
+      "learning_rate": 5.248805508410221e-07,
+      "loss": 0.1463,
+      "step": 15411
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.7059338662786125,
+      "learning_rate": 5.24478832024396e-07,
+      "loss": 0.2121,
+      "step": 15412
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.3241342725786527,
+      "learning_rate": 5.240772584857607e-07,
+      "loss": 0.1921,
+      "step": 15413
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.3764283878063075,
+      "learning_rate": 5.236758302381523e-07,
+      "loss": 0.2302,
+      "step": 15414
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.6039737506975569,
+      "learning_rate": 5.232745472945999e-07,
+      "loss": 0.1643,
+      "step": 15415
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.4884147133645755,
+      "learning_rate": 5.228734096681304e-07,
+      "loss": 0.1663,
+      "step": 15416
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.374263137915929,
+      "learning_rate": 5.224724173717643e-07,
+      "loss": 0.1335,
+      "step": 15417
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.7680375993094346,
+      "learning_rate": 5.220715704185175e-07,
+      "loss": 0.1792,
+      "step": 15418
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.1348713646635236,
+      "learning_rate": 5.21670868821404e-07,
+      "loss": 0.1449,
+      "step": 15419
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.6466027937871306,
+      "learning_rate": 5.212703125934287e-07,
+      "loss": 0.2299,
+      "step": 15420
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.581389586691899,
+      "learning_rate": 5.208699017475938e-07,
+      "loss": 0.1681,
+      "step": 15421
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.4001614956437705,
+      "learning_rate": 5.204696362968981e-07,
+      "loss": 0.1484,
+      "step": 15422
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.695907549870406,
+      "learning_rate": 5.200695162543345e-07,
+      "loss": 0.1868,
+      "step": 15423
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 2.148918749921928,
+      "learning_rate": 5.196695416328895e-07,
+      "loss": 0.1887,
+      "step": 15424
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.0725570762058787,
+      "learning_rate": 5.192697124455481e-07,
+      "loss": 0.1568,
+      "step": 15425
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 2.3997141332548027,
+      "learning_rate": 5.188700287052878e-07,
+      "loss": 0.2094,
+      "step": 15426
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.4253876129645453,
+      "learning_rate": 5.184704904250837e-07,
+      "loss": 0.1909,
+      "step": 15427
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.1651819530003695,
+      "learning_rate": 5.180710976179054e-07,
+      "loss": 0.1379,
+      "step": 15428
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.4823974235438038,
+      "learning_rate": 5.176718502967154e-07,
+      "loss": 0.1596,
+      "step": 15429
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.1539770028460752,
+      "learning_rate": 5.17272748474475e-07,
+      "loss": 0.1784,
+      "step": 15430
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.3979818464889178,
+      "learning_rate": 5.168737921641387e-07,
+      "loss": 0.1891,
+      "step": 15431
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.6941299401197656,
+      "learning_rate": 5.16474981378658e-07,
+      "loss": 0.1985,
+      "step": 15432
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.1644694493840915,
+      "learning_rate": 5.160763161309768e-07,
+      "loss": 0.2131,
+      "step": 15433
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.4967140674687067,
+      "learning_rate": 5.156777964340376e-07,
+      "loss": 0.1723,
+      "step": 15434
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.6999230251696689,
+      "learning_rate": 5.152794223007751e-07,
+      "loss": 0.1928,
+      "step": 15435
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.2949041765249207,
+      "learning_rate": 5.148811937441212e-07,
+      "loss": 0.1591,
+      "step": 15436
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.250024452566348,
+      "learning_rate": 5.144831107770037e-07,
+      "loss": 0.1453,
+      "step": 15437
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.6486571378057806,
+      "learning_rate": 5.140851734123431e-07,
+      "loss": 0.2111,
+      "step": 15438
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.5073170762160073,
+      "learning_rate": 5.136873816630572e-07,
+      "loss": 0.2275,
+      "step": 15439
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.3379759461978837,
+      "learning_rate": 5.132897355420591e-07,
+      "loss": 0.2066,
+      "step": 15440
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.3597093861474832,
+      "learning_rate": 5.128922350622567e-07,
+      "loss": 0.1509,
+      "step": 15441
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.306589607417413,
+      "learning_rate": 5.12494880236552e-07,
+      "loss": 0.1619,
+      "step": 15442
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.4577638839800746,
+      "learning_rate": 5.120976710778436e-07,
+      "loss": 0.1824,
+      "step": 15443
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.0992617964025941,
+      "learning_rate": 5.11700607599026e-07,
+      "loss": 0.1508,
+      "step": 15444
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 2.502509532466343,
+      "learning_rate": 5.113036898129875e-07,
+      "loss": 0.192,
+      "step": 15445
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.1545299546439323,
+      "learning_rate": 5.109069177326131e-07,
+      "loss": 0.1678,
+      "step": 15446
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.2372416455377397,
+      "learning_rate": 5.10510291370781e-07,
+      "loss": 0.1485,
+      "step": 15447
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.450159790426034,
+      "learning_rate": 5.101138107403658e-07,
+      "loss": 0.1522,
+      "step": 15448
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.3083171723009506,
+      "learning_rate": 5.097174758542389e-07,
+      "loss": 0.1774,
+      "step": 15449
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.30213201686129,
+      "learning_rate": 5.093212867252656e-07,
+      "loss": 0.1723,
+      "step": 15450
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.3436392084180133,
+      "learning_rate": 5.089252433663045e-07,
+      "loss": 0.137,
+      "step": 15451
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.5576922311911876,
+      "learning_rate": 5.085293457902135e-07,
+      "loss": 0.2313,
+      "step": 15452
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.9437828160100513,
+      "learning_rate": 5.081335940098419e-07,
+      "loss": 0.1794,
+      "step": 15453
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.8453950009165534,
+      "learning_rate": 5.077379880380367e-07,
+      "loss": 0.1682,
+      "step": 15454
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 2.170686300451742,
+      "learning_rate": 5.07342527887641e-07,
+      "loss": 0.2167,
+      "step": 15455
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.7202973206601384,
+      "learning_rate": 5.069472135714892e-07,
+      "loss": 0.1554,
+      "step": 15456
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.2080554104727765,
+      "learning_rate": 5.065520451024142e-07,
+      "loss": 0.1601,
+      "step": 15457
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.1958879008398253,
+      "learning_rate": 5.061570224932444e-07,
+      "loss": 0.1928,
+      "step": 15458
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.6434949386897082,
+      "learning_rate": 5.057621457568018e-07,
+      "loss": 0.1715,
+      "step": 15459
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.616231900911858,
+      "learning_rate": 5.05367414905904e-07,
+      "loss": 0.2013,
+      "step": 15460
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.2671020742425145,
+      "learning_rate": 5.049728299533646e-07,
+      "loss": 0.1545,
+      "step": 15461
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.5837901663108946,
+      "learning_rate": 5.045783909119917e-07,
+      "loss": 0.1821,
+      "step": 15462
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.3009557692843752,
+      "learning_rate": 5.041840977945894e-07,
+      "loss": 0.1541,
+      "step": 15463
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.5991125674420834,
+      "learning_rate": 5.037899506139577e-07,
+      "loss": 0.197,
+      "step": 15464
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.3095869480018036,
+      "learning_rate": 5.033959493828883e-07,
+      "loss": 0.1883,
+      "step": 15465
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.3339398154640023,
+      "learning_rate": 5.030020941141723e-07,
+      "loss": 0.13,
+      "step": 15466
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.6971111337039502,
+      "learning_rate": 5.026083848205943e-07,
+      "loss": 0.2141,
+      "step": 15467
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.303675316027502,
+      "learning_rate": 5.022148215149353e-07,
+      "loss": 0.1671,
+      "step": 15468
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.104522181878564,
+      "learning_rate": 5.018214042099684e-07,
+      "loss": 0.1149,
+      "step": 15469
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.400222119098567,
+      "learning_rate": 5.014281329184667e-07,
+      "loss": 0.19,
+      "step": 15470
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.40657406357879,
+      "learning_rate": 5.010350076531934e-07,
+      "loss": 0.2177,
+      "step": 15471
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.2621226712337899,
+      "learning_rate": 5.006420284269109e-07,
+      "loss": 0.1667,
+      "step": 15472
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.7095212561752202,
+      "learning_rate": 5.002491952523769e-07,
+      "loss": 0.1973,
+      "step": 15473
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.431195459201197,
+      "learning_rate": 4.998565081423401e-07,
+      "loss": 0.1799,
+      "step": 15474
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.5139144801485924,
+      "learning_rate": 4.994639671095491e-07,
+      "loss": 0.177,
+      "step": 15475
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.4171529425011038,
+      "learning_rate": 4.990715721667455e-07,
+      "loss": 0.1747,
+      "step": 15476
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.034739781219335,
+      "learning_rate": 4.986793233266679e-07,
+      "loss": 0.1359,
+      "step": 15477
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.3187568093752315,
+      "learning_rate": 4.982872206020467e-07,
+      "loss": 0.2025,
+      "step": 15478
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.5001345266178414,
+      "learning_rate": 4.978952640056117e-07,
+      "loss": 0.2022,
+      "step": 15479
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.0931636158047886,
+      "learning_rate": 4.975034535500844e-07,
+      "loss": 0.1513,
+      "step": 15480
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.9236420207365321,
+      "learning_rate": 4.971117892481847e-07,
+      "loss": 0.1825,
+      "step": 15481
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.4055097220928343,
+      "learning_rate": 4.967202711126268e-07,
+      "loss": 0.163,
+      "step": 15482
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.397271277395178,
+      "learning_rate": 4.963288991561172e-07,
+      "loss": 0.1602,
+      "step": 15483
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.1948411548257578,
+      "learning_rate": 4.959376733913613e-07,
+      "loss": 0.1387,
+      "step": 15484
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.6965166553058637,
+      "learning_rate": 4.955465938310583e-07,
+      "loss": 0.1693,
+      "step": 15485
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.3186147456617048,
+      "learning_rate": 4.951556604879049e-07,
+      "loss": 0.2097,
+      "step": 15486
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.0359843162959022,
+      "learning_rate": 4.947648733745875e-07,
+      "loss": 0.1483,
+      "step": 15487
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.2513026780978194,
+      "learning_rate": 4.943742325037942e-07,
+      "loss": 0.16,
+      "step": 15488
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 2.1250255176470634,
+      "learning_rate": 4.939837378882034e-07,
+      "loss": 0.2037,
+      "step": 15489
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.2167061363174818,
+      "learning_rate": 4.935933895404915e-07,
+      "loss": 0.1638,
+      "step": 15490
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.1069953675737434,
+      "learning_rate": 4.932031874733306e-07,
+      "loss": 0.2266,
+      "step": 15491
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.4791561538251063,
+      "learning_rate": 4.928131316993845e-07,
+      "loss": 0.2412,
+      "step": 15492
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.0820685803690886,
+      "learning_rate": 4.924232222313163e-07,
+      "loss": 0.1844,
+      "step": 15493
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.445535405471245,
+      "learning_rate": 4.920334590817821e-07,
+      "loss": 0.1635,
+      "step": 15494
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.1493202599013463,
+      "learning_rate": 4.916438422634356e-07,
+      "loss": 0.1479,
+      "step": 15495
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.5454093993367395,
+      "learning_rate": 4.912543717889207e-07,
+      "loss": 0.1532,
+      "step": 15496
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.9654169933838284,
+      "learning_rate": 4.90865047670882e-07,
+      "loss": 0.1894,
+      "step": 15497
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.2916404608662542,
+      "learning_rate": 4.90475869921957e-07,
+      "loss": 0.1635,
+      "step": 15498
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.5567605368498203,
+      "learning_rate": 4.900868385547786e-07,
+      "loss": 0.2138,
+      "step": 15499
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.3710022974624554,
+      "learning_rate": 4.896979535819751e-07,
+      "loss": 0.2114,
+      "step": 15500
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.4645220136173143,
+      "learning_rate": 4.893092150161688e-07,
+      "loss": 0.157,
+      "step": 15501
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.15342740006677,
+      "learning_rate": 4.889206228699794e-07,
+      "loss": 0.1317,
+      "step": 15502
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.8765814445950749,
+      "learning_rate": 4.885321771560208e-07,
+      "loss": 0.1833,
+      "step": 15503
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.2639423063305935,
+      "learning_rate": 4.88143877886903e-07,
+      "loss": 0.1695,
+      "step": 15504
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.7064147464459285,
+      "learning_rate": 4.87755725075228e-07,
+      "loss": 0.1975,
+      "step": 15505
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.8458725947329973,
+      "learning_rate": 4.873677187335985e-07,
+      "loss": 0.1655,
+      "step": 15506
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.3239415871766391,
+      "learning_rate": 4.86979858874606e-07,
+      "loss": 0.1792,
+      "step": 15507
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.3974905430758693,
+      "learning_rate": 4.86592145510843e-07,
+      "loss": 0.1322,
+      "step": 15508
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.1882982708014793,
+      "learning_rate": 4.862045786548953e-07,
+      "loss": 0.1323,
+      "step": 15509
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.5425784619375187,
+      "learning_rate": 4.858171583193416e-07,
+      "loss": 0.1888,
+      "step": 15510
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.124104153065789,
+      "learning_rate": 4.854298845167587e-07,
+      "loss": 0.1442,
+      "step": 15511
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.4245415196786564,
+      "learning_rate": 4.850427572597183e-07,
+      "loss": 0.2402,
+      "step": 15512
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.4549949263782462,
+      "learning_rate": 4.846557765607868e-07,
+      "loss": 0.1694,
+      "step": 15513
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.242204815965796,
+      "learning_rate": 4.842689424325242e-07,
+      "loss": 0.1884,
+      "step": 15514
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.3624682365879712,
+      "learning_rate": 4.838822548874883e-07,
+      "loss": 0.1384,
+      "step": 15515
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.4880578656896495,
+      "learning_rate": 4.834957139382318e-07,
+      "loss": 0.1553,
+      "step": 15516
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.3576350312315022,
+      "learning_rate": 4.831093195973014e-07,
+      "loss": 0.1672,
+      "step": 15517
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.3059238868637362,
+      "learning_rate": 4.827230718772413e-07,
+      "loss": 0.1771,
+      "step": 15518
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.7903801342837886,
+      "learning_rate": 4.823369707905861e-07,
+      "loss": 0.1342,
+      "step": 15519
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.1491861206494953,
+      "learning_rate": 4.819510163498714e-07,
+      "loss": 0.1529,
+      "step": 15520
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.2191902612557208,
+      "learning_rate": 4.815652085676243e-07,
+      "loss": 0.1645,
+      "step": 15521
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.5815643469274532,
+      "learning_rate": 4.811795474563702e-07,
+      "loss": 0.2249,
+      "step": 15522
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.0596863010751822,
+      "learning_rate": 4.807940330286253e-07,
+      "loss": 0.1867,
+      "step": 15523
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.5996269998289752,
+      "learning_rate": 4.804086652969059e-07,
+      "loss": 0.207,
+      "step": 15524
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.4520381072356676,
+      "learning_rate": 4.80023444273719e-07,
+      "loss": 0.1609,
+      "step": 15525
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.295175702978321,
+      "learning_rate": 4.796383699715706e-07,
+      "loss": 0.1989,
+      "step": 15526
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.4019437049300527,
+      "learning_rate": 4.792534424029605e-07,
+      "loss": 0.1917,
+      "step": 15527
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.6117484405711806,
+      "learning_rate": 4.788686615803829e-07,
+      "loss": 0.2158,
+      "step": 15528
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.4055499665658493,
+      "learning_rate": 4.784840275163282e-07,
+      "loss": 0.144,
+      "step": 15529
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.4898076677678909,
+      "learning_rate": 4.780995402232819e-07,
+      "loss": 0.1729,
+      "step": 15530
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.7150736331007697,
+      "learning_rate": 4.777151997137258e-07,
+      "loss": 0.2098,
+      "step": 15531
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.5055517928970696,
+      "learning_rate": 4.77331006000134e-07,
+      "loss": 0.172,
+      "step": 15532
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.1418496866505392,
+      "learning_rate": 4.769469590949783e-07,
+      "loss": 0.1721,
+      "step": 15533
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.3823978427372874,
+      "learning_rate": 4.765630590107251e-07,
+      "loss": 0.1861,
+      "step": 15534
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.058595104850099,
+      "learning_rate": 4.761793057598363e-07,
+      "loss": 0.1462,
+      "step": 15535
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.565857998229226,
+      "learning_rate": 4.7579569935476967e-07,
+      "loss": 0.2057,
+      "step": 15536
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.546669667034829,
+      "learning_rate": 4.754122398079752e-07,
+      "loss": 0.1601,
+      "step": 15537
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.4823257857430705,
+      "learning_rate": 4.750289271319014e-07,
+      "loss": 0.1203,
+      "step": 15538
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.6109888586594971,
+      "learning_rate": 4.7464576133899043e-07,
+      "loss": 0.1688,
+      "step": 15539
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.5680324503489973,
+      "learning_rate": 4.742627424416818e-07,
+      "loss": 0.2165,
+      "step": 15540
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.186179361571634,
+      "learning_rate": 4.738798704524056e-07,
+      "loss": 0.1878,
+      "step": 15541
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.613726492118846,
+      "learning_rate": 4.734971453835924e-07,
+      "loss": 0.185,
+      "step": 15542
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.4228323448836515,
+      "learning_rate": 4.7311456724766394e-07,
+      "loss": 0.1555,
+      "step": 15543
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.3751813216297524,
+      "learning_rate": 4.727321360570397e-07,
+      "loss": 0.2021,
+      "step": 15544
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.5285139357306226,
+      "learning_rate": 4.7234985182413485e-07,
+      "loss": 0.1923,
+      "step": 15545
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.0866695619169868,
+      "learning_rate": 4.719677145613566e-07,
+      "loss": 0.1544,
+      "step": 15546
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.5263218400092327,
+      "learning_rate": 4.7158572428111007e-07,
+      "loss": 0.1463,
+      "step": 15547
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.3671181470446865,
+      "learning_rate": 4.7120388099579473e-07,
+      "loss": 0.1923,
+      "step": 15548
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.4839270970034082,
+      "learning_rate": 4.708221847178063e-07,
+      "loss": 0.1901,
+      "step": 15549
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 2.001670780275593,
+      "learning_rate": 4.704406354595337e-07,
+      "loss": 0.2037,
+      "step": 15550
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.2394964290523602,
+      "learning_rate": 4.7005923323336254e-07,
+      "loss": 0.1451,
+      "step": 15551
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 2.0525976692651184,
+      "learning_rate": 4.696779780516741e-07,
+      "loss": 0.2456,
+      "step": 15552
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.7812601892280273,
+      "learning_rate": 4.6929686992684296e-07,
+      "loss": 0.1968,
+      "step": 15553
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.3673128288590535,
+      "learning_rate": 4.6891590887124197e-07,
+      "loss": 0.1609,
+      "step": 15554
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.2677949476026076,
+      "learning_rate": 4.6853509489723556e-07,
+      "loss": 0.2049,
+      "step": 15555
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.2227414790054318,
+      "learning_rate": 4.681544280171851e-07,
+      "loss": 0.1558,
+      "step": 15556
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.6708425961766449,
+      "learning_rate": 4.677739082434485e-07,
+      "loss": 0.2078,
+      "step": 15557
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.5519390588521877,
+      "learning_rate": 4.673935355883774e-07,
+      "loss": 0.2094,
+      "step": 15558
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.3248646772756125,
+      "learning_rate": 4.6701331006431816e-07,
+      "loss": 0.1558,
+      "step": 15559
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.3648459409926526,
+      "learning_rate": 4.666332316836142e-07,
+      "loss": 0.1624,
+      "step": 15560
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.6958885356891225,
+      "learning_rate": 4.662533004586012e-07,
+      "loss": 0.161,
+      "step": 15561
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.4207253576830312,
+      "learning_rate": 4.658735164016137e-07,
+      "loss": 0.1883,
+      "step": 15562
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.4013754920846322,
+      "learning_rate": 4.654938795249797e-07,
+      "loss": 0.1309,
+      "step": 15563
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.560248278337729,
+      "learning_rate": 4.6511438984102097e-07,
+      "loss": 0.1932,
+      "step": 15564
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.503139082194325,
+      "learning_rate": 4.647350473620571e-07,
+      "loss": 0.196,
+      "step": 15565
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.5458049926661885,
+      "learning_rate": 4.6435585210040157e-07,
+      "loss": 0.1955,
+      "step": 15566
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.015252931006225,
+      "learning_rate": 4.63976804068364e-07,
+      "loss": 0.1623,
+      "step": 15567
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.4825798171671414,
+      "learning_rate": 4.6359790327824726e-07,
+      "loss": 0.1627,
+      "step": 15568
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.3107234733517719,
+      "learning_rate": 4.6321914974235103e-07,
+      "loss": 0.1847,
+      "step": 15569
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.3848144727089327,
+      "learning_rate": 4.628405434729699e-07,
+      "loss": 0.185,
+      "step": 15570
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.7971681728859372,
+      "learning_rate": 4.624620844823935e-07,
+      "loss": 0.19,
+      "step": 15571
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.2951337661107771,
+      "learning_rate": 4.620837727829086e-07,
+      "loss": 0.1511,
+      "step": 15572
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.4995921544097068,
+      "learning_rate": 4.617056083867927e-07,
+      "loss": 0.1391,
+      "step": 15573
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.677239155095614,
+      "learning_rate": 4.61327591306322e-07,
+      "loss": 0.1845,
+      "step": 15574
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.936822662443969,
+      "learning_rate": 4.609497215537684e-07,
+      "loss": 0.1587,
+      "step": 15575
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.9678860539851173,
+      "learning_rate": 4.605719991413976e-07,
+      "loss": 0.1863,
+      "step": 15576
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.6384919969225322,
+      "learning_rate": 4.601944240814693e-07,
+      "loss": 0.2254,
+      "step": 15577
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.6558852693837258,
+      "learning_rate": 4.598169963862409e-07,
+      "loss": 0.1934,
+      "step": 15578
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.2222278502574337,
+      "learning_rate": 4.5943971606796314e-07,
+      "loss": 0.1584,
+      "step": 15579
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.0603421190972249,
+      "learning_rate": 4.590625831388834e-07,
+      "loss": 0.187,
+      "step": 15580
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.2060556944561047,
+      "learning_rate": 4.5868559761124423e-07,
+      "loss": 0.2109,
+      "step": 15581
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.7792740201686106,
+      "learning_rate": 4.583087594972807e-07,
+      "loss": 0.2043,
+      "step": 15582
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.1300061289654837,
+      "learning_rate": 4.57932068809227e-07,
+      "loss": 0.1675,
+      "step": 15583
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.8580779833191028,
+      "learning_rate": 4.5755552555931004e-07,
+      "loss": 0.1644,
+      "step": 15584
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.3630841066189938,
+      "learning_rate": 4.5717912975975333e-07,
+      "loss": 0.1553,
+      "step": 15585
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.4189058470307991,
+      "learning_rate": 4.568028814227737e-07,
+      "loss": 0.1681,
+      "step": 15586
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.3963367837660785,
+      "learning_rate": 4.5642678056058545e-07,
+      "loss": 0.173,
+      "step": 15587
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.6674908661413463,
+      "learning_rate": 4.560508271853964e-07,
+      "loss": 0.2112,
+      "step": 15588
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.2284222523391461,
+      "learning_rate": 4.5567502130941024e-07,
+      "loss": 0.1684,
+      "step": 15589
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.501458313470301,
+      "learning_rate": 4.552993629448271e-07,
+      "loss": 0.177,
+      "step": 15590
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.4736339905306264,
+      "learning_rate": 4.5492385210383904e-07,
+      "loss": 0.2133,
+      "step": 15591
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.9113892974140146,
+      "learning_rate": 4.545484887986368e-07,
+      "loss": 0.1505,
+      "step": 15592
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.1491888987633878,
+      "learning_rate": 4.541732730414039e-07,
+      "loss": 0.1289,
+      "step": 15593
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.4167759767836365,
+      "learning_rate": 4.537982048443218e-07,
+      "loss": 0.225,
+      "step": 15594
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 2.373631333955147,
+      "learning_rate": 4.534232842195635e-07,
+      "loss": 0.2073,
+      "step": 15595
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.2902896761844638,
+      "learning_rate": 4.5304851117930035e-07,
+      "loss": 0.1673,
+      "step": 15596
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.9649438979803212,
+      "learning_rate": 4.526738857356966e-07,
+      "loss": 0.1555,
+      "step": 15597
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.4152588393907861,
+      "learning_rate": 4.52299407900913e-07,
+      "loss": 0.1608,
+      "step": 15598
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.4963445030012776,
+      "learning_rate": 4.51925077687107e-07,
+      "loss": 0.1485,
+      "step": 15599
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.319724318731212,
+      "learning_rate": 4.5155089510642737e-07,
+      "loss": 0.1957,
+      "step": 15600
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.6082973575850925,
+      "learning_rate": 4.511768601710209e-07,
+      "loss": 0.1828,
+      "step": 15601
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.4089573442210366,
+      "learning_rate": 4.5080297289302966e-07,
+      "loss": 0.1841,
+      "step": 15602
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.1842057185117265,
+      "learning_rate": 4.504292332845911e-07,
+      "loss": 0.1344,
+      "step": 15603
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.2238724360524376,
+      "learning_rate": 4.5005564135783443e-07,
+      "loss": 0.1246,
+      "step": 15604
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.5183484148945507,
+      "learning_rate": 4.4968219712488825e-07,
+      "loss": 0.1562,
+      "step": 15605
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.4633790844473922,
+      "learning_rate": 4.49308900597874e-07,
+      "loss": 0.1583,
+      "step": 15606
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.590809189187931,
+      "learning_rate": 4.489357517889104e-07,
+      "loss": 0.2051,
+      "step": 15607
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.5601457025269798,
+      "learning_rate": 4.485627507101098e-07,
+      "loss": 0.1652,
+      "step": 15608
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.3712374837368477,
+      "learning_rate": 4.481898973735793e-07,
+      "loss": 0.2215,
+      "step": 15609
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.4349545690650378,
+      "learning_rate": 4.4781719179142145e-07,
+      "loss": 0.2062,
+      "step": 15610
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.786146772746412,
+      "learning_rate": 4.4744463397573543e-07,
+      "loss": 0.1195,
+      "step": 15611
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.4550018905024356,
+      "learning_rate": 4.4707222393861606e-07,
+      "loss": 0.1535,
+      "step": 15612
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.615275229106896,
+      "learning_rate": 4.466999616921486e-07,
+      "loss": 0.2078,
+      "step": 15613
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.5100787431509515,
+      "learning_rate": 4.4632784724842e-07,
+      "loss": 0.2226,
+      "step": 15614
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.6350992425202575,
+      "learning_rate": 4.4595588061950744e-07,
+      "loss": 0.1701,
+      "step": 15615
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.137830187616325,
+      "learning_rate": 4.455840618174856e-07,
+      "loss": 0.1911,
+      "step": 15616
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.5722579861094033,
+      "learning_rate": 4.452123908544248e-07,
+      "loss": 0.1916,
+      "step": 15617
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.2039031571454344,
+      "learning_rate": 4.448408677423882e-07,
+      "loss": 0.1893,
+      "step": 15618
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.5636273703911299,
+      "learning_rate": 4.444694924934362e-07,
+      "loss": 0.2044,
+      "step": 15619
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.3098026100896032,
+      "learning_rate": 4.4409826511962405e-07,
+      "loss": 0.123,
+      "step": 15620
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.4692254311534323,
+      "learning_rate": 4.437271856330033e-07,
+      "loss": 0.178,
+      "step": 15621
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.6722066541404197,
+      "learning_rate": 4.433562540456171e-07,
+      "loss": 0.1863,
+      "step": 15622
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.7413025947960803,
+      "learning_rate": 4.4298547036950743e-07,
+      "loss": 0.2078,
+      "step": 15623
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.3063878440678138,
+      "learning_rate": 4.4261483461670915e-07,
+      "loss": 0.1939,
+      "step": 15624
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.6122947937484498,
+      "learning_rate": 4.422443467992549e-07,
+      "loss": 0.2199,
+      "step": 15625
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.3736799687904988,
+      "learning_rate": 4.4187400692917047e-07,
+      "loss": 0.1505,
+      "step": 15626
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.2089627931113678,
+      "learning_rate": 4.4150381501847585e-07,
+      "loss": 0.2088,
+      "step": 15627
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.7519962832813436,
+      "learning_rate": 4.411337710791891e-07,
+      "loss": 0.1713,
+      "step": 15628
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 2.302119398625692,
+      "learning_rate": 4.407638751233223e-07,
+      "loss": 0.2669,
+      "step": 15629
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.149078216362258,
+      "learning_rate": 4.4039412716288087e-07,
+      "loss": 0.1379,
+      "step": 15630
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.5698715226564455,
+      "learning_rate": 4.4002452720986856e-07,
+      "loss": 0.1931,
+      "step": 15631
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.6793731129092313,
+      "learning_rate": 4.396550752762824e-07,
+      "loss": 0.2004,
+      "step": 15632
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.5020665223926462,
+      "learning_rate": 4.392857713741144e-07,
+      "loss": 0.1898,
+      "step": 15633
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.4319186493619864,
+      "learning_rate": 4.389166155153529e-07,
+      "loss": 0.2079,
+      "step": 15634
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.0626568468112108,
+      "learning_rate": 4.3854760771198156e-07,
+      "loss": 0.1345,
+      "step": 15635
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.2698171968477894,
+      "learning_rate": 4.381787479759775e-07,
+      "loss": 0.1211,
+      "step": 15636
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.8266112589467094,
+      "learning_rate": 4.3781003631931384e-07,
+      "loss": 0.1692,
+      "step": 15637
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.317368888357225,
+      "learning_rate": 4.374414727539605e-07,
+      "loss": 0.1311,
+      "step": 15638
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.6879967779509022,
+      "learning_rate": 4.370730572918813e-07,
+      "loss": 0.2345,
+      "step": 15639
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.291581273781236,
+      "learning_rate": 4.367047899450333e-07,
+      "loss": 0.1839,
+      "step": 15640
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.767373839317743,
+      "learning_rate": 4.363366707253719e-07,
+      "loss": 0.2207,
+      "step": 15641
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 2.444347215444324,
+      "learning_rate": 4.3596869964484703e-07,
+      "loss": 0.219,
+      "step": 15642
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.3763761393695904,
+      "learning_rate": 4.356008767154024e-07,
+      "loss": 0.1169,
+      "step": 15643
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.2004714939821406,
+      "learning_rate": 4.352332019489786e-07,
+      "loss": 0.1895,
+      "step": 15644
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.5754401530476032,
+      "learning_rate": 4.348656753575092e-07,
+      "loss": 0.1834,
+      "step": 15645
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.3805235652694587,
+      "learning_rate": 4.3449829695292545e-07,
+      "loss": 0.2027,
+      "step": 15646
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.7093422084571186,
+      "learning_rate": 4.3413106674715324e-07,
+      "loss": 0.1655,
+      "step": 15647
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.3326918762066124,
+      "learning_rate": 4.3376398475211077e-07,
+      "loss": 0.1842,
+      "step": 15648
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.5661270101989369,
+      "learning_rate": 4.333970509797164e-07,
+      "loss": 0.2113,
+      "step": 15649
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.4978278496089241,
+      "learning_rate": 4.330302654418789e-07,
+      "loss": 0.1632,
+      "step": 15650
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.526748453409009,
+      "learning_rate": 4.3266362815050536e-07,
+      "loss": 0.1882,
+      "step": 15651
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.382789309894487,
+      "learning_rate": 4.3229713911749625e-07,
+      "loss": 0.163,
+      "step": 15652
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.2602677047074602,
+      "learning_rate": 4.3193079835474996e-07,
+      "loss": 0.1669,
+      "step": 15653
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.2175095156693325,
+      "learning_rate": 4.315646058741557e-07,
+      "loss": 0.1404,
+      "step": 15654
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.1434096351057765,
+      "learning_rate": 4.311985616876013e-07,
+      "loss": 0.1753,
+      "step": 15655
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.4558031260128395,
+      "learning_rate": 4.308326658069689e-07,
+      "loss": 0.1636,
+      "step": 15656
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.3888381236123561,
+      "learning_rate": 4.3046691824413565e-07,
+      "loss": 0.2302,
+      "step": 15657
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.3906027277759514,
+      "learning_rate": 4.3010131901097483e-07,
+      "loss": 0.1473,
+      "step": 15658
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 2.0128305862625746,
+      "learning_rate": 4.2973586811935253e-07,
+      "loss": 0.2131,
+      "step": 15659
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.2266794197581006,
+      "learning_rate": 4.2937056558113144e-07,
+      "loss": 0.1748,
+      "step": 15660
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.208553995614156,
+      "learning_rate": 4.2900541140817044e-07,
+      "loss": 0.1592,
+      "step": 15661
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.2167088031276483,
+      "learning_rate": 4.2864040561232334e-07,
+      "loss": 0.1428,
+      "step": 15662
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.98032980587612,
+      "learning_rate": 4.282755482054357e-07,
+      "loss": 0.1685,
+      "step": 15663
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.6178209151619656,
+      "learning_rate": 4.279108391993536e-07,
+      "loss": 0.1493,
+      "step": 15664
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.2312955122368778,
+      "learning_rate": 4.2754627860591537e-07,
+      "loss": 0.1957,
+      "step": 15665
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.4131840037158854,
+      "learning_rate": 4.2718186643695315e-07,
+      "loss": 0.2127,
+      "step": 15666
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.5848619685178469,
+      "learning_rate": 4.2681760270429804e-07,
+      "loss": 0.1732,
+      "step": 15667
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.5428611476415108,
+      "learning_rate": 4.2645348741977286e-07,
+      "loss": 0.1546,
+      "step": 15668
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.6639789103615232,
+      "learning_rate": 4.2608952059519706e-07,
+      "loss": 0.1586,
+      "step": 15669
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.5822991980180274,
+      "learning_rate": 4.2572570224238606e-07,
+      "loss": 0.1898,
+      "step": 15670
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.2952139857232599,
+      "learning_rate": 4.253620323731494e-07,
+      "loss": 0.1925,
+      "step": 15671
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.5075608671056235,
+      "learning_rate": 4.2499851099929157e-07,
+      "loss": 0.1606,
+      "step": 15672
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.9865063312601492,
+      "learning_rate": 4.246351381326125e-07,
+      "loss": 0.2085,
+      "step": 15673
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.2263465800637017,
+      "learning_rate": 4.242719137849077e-07,
+      "loss": 0.1474,
+      "step": 15674
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.2011465116474733,
+      "learning_rate": 4.2390883796796787e-07,
+      "loss": 0.1794,
+      "step": 15675
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.8180022170621952,
+      "learning_rate": 4.2354591069357963e-07,
+      "loss": 0.2085,
+      "step": 15676
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.2264808852383433,
+      "learning_rate": 4.231831319735219e-07,
+      "loss": 0.2035,
+      "step": 15677
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.4643722436547644,
+      "learning_rate": 4.228205018195719e-07,
+      "loss": 0.1679,
+      "step": 15678
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.3595291138029266,
+      "learning_rate": 4.224580202434997e-07,
+      "loss": 0.1539,
+      "step": 15679
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.377441622609984,
+      "learning_rate": 4.2209568725707415e-07,
+      "loss": 0.1457,
+      "step": 15680
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.6281826735768332,
+      "learning_rate": 4.217335028720537e-07,
+      "loss": 0.1722,
+      "step": 15681
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.2476790218866842,
+      "learning_rate": 4.2137146710019616e-07,
+      "loss": 0.149,
+      "step": 15682
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.5154982346682082,
+      "learning_rate": 4.210095799532554e-07,
+      "loss": 0.1992,
+      "step": 15683
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.628101351267859,
+      "learning_rate": 4.206478414429754e-07,
+      "loss": 0.1664,
+      "step": 15684
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.2420197379590956,
+      "learning_rate": 4.202862515811007e-07,
+      "loss": 0.1477,
+      "step": 15685
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.3639904461998713,
+      "learning_rate": 4.1992481037936684e-07,
+      "loss": 0.1564,
+      "step": 15686
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.5051061426164225,
+      "learning_rate": 4.195635178495072e-07,
+      "loss": 0.172,
+      "step": 15687
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.4579307806638828,
+      "learning_rate": 4.1920237400325025e-07,
+      "loss": 0.167,
+      "step": 15688
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.5953982842944607,
+      "learning_rate": 4.188413788523188e-07,
+      "loss": 0.1644,
+      "step": 15689
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 2.1010608357954745,
+      "learning_rate": 4.184805324084301e-07,
+      "loss": 0.1904,
+      "step": 15690
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.4039858817228876,
+      "learning_rate": 4.1811983468329764e-07,
+      "loss": 0.1985,
+      "step": 15691
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.4013718953867662,
+      "learning_rate": 4.177592856886298e-07,
+      "loss": 0.182,
+      "step": 15692
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.3250009482758205,
+      "learning_rate": 4.1739888543613107e-07,
+      "loss": 0.0998,
+      "step": 15693
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.140919634246329,
+      "learning_rate": 4.1703863393750046e-07,
+      "loss": 0.1675,
+      "step": 15694
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.215844269788399,
+      "learning_rate": 4.1667853120442967e-07,
+      "loss": 0.1756,
+      "step": 15695
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.7075134399410135,
+      "learning_rate": 4.1631857724861003e-07,
+      "loss": 0.2068,
+      "step": 15696
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 2.4308914685111502,
+      "learning_rate": 4.159587720817249e-07,
+      "loss": 0.177,
+      "step": 15697
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.3364566608467054,
+      "learning_rate": 4.1559911571545544e-07,
+      "loss": 0.172,
+      "step": 15698
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.6891681982134459,
+      "learning_rate": 4.1523960816147345e-07,
+      "loss": 0.1789,
+      "step": 15699
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.169781235407481,
+      "learning_rate": 4.1488024943145023e-07,
+      "loss": 0.1561,
+      "step": 15700
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.733011306729938,
+      "learning_rate": 4.145210395370519e-07,
+      "loss": 0.2114,
+      "step": 15701
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.6905638114306123,
+      "learning_rate": 4.1416197848993643e-07,
+      "loss": 0.1902,
+      "step": 15702
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.3494448968718966,
+      "learning_rate": 4.1380306630176117e-07,
+      "loss": 0.1751,
+      "step": 15703
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.9990447415482477,
+      "learning_rate": 4.134443029841745e-07,
+      "loss": 0.2294,
+      "step": 15704
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.1356186869314062,
+      "learning_rate": 4.130856885488233e-07,
+      "loss": 0.2066,
+      "step": 15705
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.6056948314670278,
+      "learning_rate": 4.127272230073487e-07,
+      "loss": 0.1721,
+      "step": 15706
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 2.383184797875853,
+      "learning_rate": 4.123689063713865e-07,
+      "loss": 0.1834,
+      "step": 15707
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.204659202134478,
+      "learning_rate": 4.1201073865256734e-07,
+      "loss": 0.1291,
+      "step": 15708
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.4835849365968952,
+      "learning_rate": 4.116527198625181e-07,
+      "loss": 0.167,
+      "step": 15709
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.3297648628133618,
+      "learning_rate": 4.112948500128594e-07,
+      "loss": 0.224,
+      "step": 15710
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.706054198380189,
+      "learning_rate": 4.1093712911520867e-07,
+      "loss": 0.231,
+      "step": 15711
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.3529343092949455,
+      "learning_rate": 4.105795571811788e-07,
+      "loss": 0.2012,
+      "step": 15712
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.0585537723438907,
+      "learning_rate": 4.1022213422237446e-07,
+      "loss": 0.1273,
+      "step": 15713
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.3284874533695021,
+      "learning_rate": 4.098648602503996e-07,
+      "loss": 0.1433,
+      "step": 15714
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.1914778547187996,
+      "learning_rate": 4.095077352768501e-07,
+      "loss": 0.1769,
+      "step": 15715
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.1273136983502274,
+      "learning_rate": 4.091507593133204e-07,
+      "loss": 0.1989,
+      "step": 15716
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.3105394673031523,
+      "learning_rate": 4.0879393237139633e-07,
+      "loss": 0.1722,
+      "step": 15717
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.5426270891321792,
+      "learning_rate": 4.084372544626608e-07,
+      "loss": 0.1921,
+      "step": 15718
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.205168481094807,
+      "learning_rate": 4.080807255986935e-07,
+      "loss": 0.1805,
+      "step": 15719
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.4761789419141815,
+      "learning_rate": 4.0772434579106565e-07,
+      "loss": 0.1771,
+      "step": 15720
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.7096397058956576,
+      "learning_rate": 4.07368115051347e-07,
+      "loss": 0.2041,
+      "step": 15721
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.4620771196105944,
+      "learning_rate": 4.0701203339109984e-07,
+      "loss": 0.1967,
+      "step": 15722
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.4354822261032822,
+      "learning_rate": 4.066561008218828e-07,
+      "loss": 0.214,
+      "step": 15723
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.8495421029800736,
+      "learning_rate": 4.0630031735524987e-07,
+      "loss": 0.2284,
+      "step": 15724
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.357634004991706,
+      "learning_rate": 4.0594468300275136e-07,
+      "loss": 0.2182,
+      "step": 15725
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.7439462910172099,
+      "learning_rate": 4.0558919777592963e-07,
+      "loss": 0.2254,
+      "step": 15726
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.2247571719112054,
+      "learning_rate": 4.052338616863244e-07,
+      "loss": 0.1755,
+      "step": 15727
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.3780247569659694,
+      "learning_rate": 4.048786747454697e-07,
+      "loss": 0.1732,
+      "step": 15728
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.5278882180620335,
+      "learning_rate": 4.045236369648958e-07,
+      "loss": 0.1467,
+      "step": 15729
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.4252754045680354,
+      "learning_rate": 4.041687483561285e-07,
+      "loss": 0.1793,
+      "step": 15730
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.1497548164605627,
+      "learning_rate": 4.0381400893068523e-07,
+      "loss": 0.2004,
+      "step": 15731
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.3203642880141473,
+      "learning_rate": 4.034594187000823e-07,
+      "loss": 0.1766,
+      "step": 15732
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.3087980537919763,
+      "learning_rate": 4.0310497767583e-07,
+      "loss": 0.1486,
+      "step": 15733
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.2683510209382052,
+      "learning_rate": 4.0275068586943414e-07,
+      "loss": 0.2113,
+      "step": 15734
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.3991071919623705,
+      "learning_rate": 4.0239654329239377e-07,
+      "loss": 0.1603,
+      "step": 15735
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.6182973619163408,
+      "learning_rate": 4.020425499562053e-07,
+      "loss": 0.2097,
+      "step": 15736
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.394725091535719,
+      "learning_rate": 4.016887058723606e-07,
+      "loss": 0.1522,
+      "step": 15737
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.462501973370908,
+      "learning_rate": 4.0133501105234396e-07,
+      "loss": 0.1561,
+      "step": 15738
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.4145123888361824,
+      "learning_rate": 4.009814655076383e-07,
+      "loss": 0.1802,
+      "step": 15739
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.2642890036448176,
+      "learning_rate": 4.0062806924971775e-07,
+      "loss": 0.1531,
+      "step": 15740
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.1941449035178553,
+      "learning_rate": 4.002748222900554e-07,
+      "loss": 0.158,
+      "step": 15741
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.6748490041313533,
+      "learning_rate": 3.9992172464011657e-07,
+      "loss": 0.1905,
+      "step": 15742
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.3480291569357024,
+      "learning_rate": 3.9956877631136537e-07,
+      "loss": 0.195,
+      "step": 15743
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.7279769834071683,
+      "learning_rate": 3.99215977315256e-07,
+      "loss": 0.1755,
+      "step": 15744
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.2788075974712882,
+      "learning_rate": 3.988633276632414e-07,
+      "loss": 0.1738,
+      "step": 15745
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.9469113194979837,
+      "learning_rate": 3.9851082736676926e-07,
+      "loss": 0.1924,
+      "step": 15746
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 2.6086751718943835,
+      "learning_rate": 3.98158476437282e-07,
+      "loss": 0.2157,
+      "step": 15747
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.7367417494005575,
+      "learning_rate": 3.9780627488621714e-07,
+      "loss": 0.1994,
+      "step": 15748
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.5931455036114486,
+      "learning_rate": 3.9745422272500665e-07,
+      "loss": 0.1861,
+      "step": 15749
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 2.173379390770697,
+      "learning_rate": 3.9710231996507866e-07,
+      "loss": 0.2117,
+      "step": 15750
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.1600936269074411,
+      "learning_rate": 3.9675056661785563e-07,
+      "loss": 0.1517,
+      "step": 15751
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.728025403524249,
+      "learning_rate": 3.963989626947579e-07,
+      "loss": 0.2204,
+      "step": 15752
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 3.708481771153912,
+      "learning_rate": 3.960475082071957e-07,
+      "loss": 0.1592,
+      "step": 15753
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.4338274980866987,
+      "learning_rate": 3.9569620316657895e-07,
+      "loss": 0.1943,
+      "step": 15754
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.710502656346377,
+      "learning_rate": 3.9534504758431237e-07,
+      "loss": 0.1878,
+      "step": 15755
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.9951037211810931,
+      "learning_rate": 3.9499404147179234e-07,
+      "loss": 0.2488,
+      "step": 15756
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.3616473180842865,
+      "learning_rate": 3.9464318484041476e-07,
+      "loss": 0.1587,
+      "step": 15757
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.1533506035763519,
+      "learning_rate": 3.9429247770156664e-07,
+      "loss": 0.1367,
+      "step": 15758
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.4068026993000136,
+      "learning_rate": 3.9394192006663277e-07,
+      "loss": 0.1755,
+      "step": 15759
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.5219180232239045,
+      "learning_rate": 3.935915119469935e-07,
+      "loss": 0.2239,
+      "step": 15760
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.0960975521834668,
+      "learning_rate": 3.9324125335402307e-07,
+      "loss": 0.1898,
+      "step": 15761
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 2.774300261073232,
+      "learning_rate": 3.9289114429909013e-07,
+      "loss": 0.1453,
+      "step": 15762
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.5717802708371302,
+      "learning_rate": 3.925411847935595e-07,
+      "loss": 0.1625,
+      "step": 15763
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.804595292613318,
+      "learning_rate": 3.921913748487916e-07,
+      "loss": 0.2061,
+      "step": 15764
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.2169008072848215,
+      "learning_rate": 3.918417144761416e-07,
+      "loss": 0.1765,
+      "step": 15765
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.331104638666028,
+      "learning_rate": 3.9149220368696004e-07,
+      "loss": 0.1574,
+      "step": 15766
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.2646798737496068,
+      "learning_rate": 3.9114284249259114e-07,
+      "loss": 0.147,
+      "step": 15767
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.5471092441718475,
+      "learning_rate": 3.907936309043753e-07,
+      "loss": 0.1832,
+      "step": 15768
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 2.027281537874642,
+      "learning_rate": 3.9044456893364947e-07,
+      "loss": 0.2231,
+      "step": 15769
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 2.541756758630523,
+      "learning_rate": 3.900956565917435e-07,
+      "loss": 0.2097,
+      "step": 15770
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.9761708122574302,
+      "learning_rate": 3.8974689388998346e-07,
+      "loss": 0.178,
+      "step": 15771
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.6761874171300772,
+      "learning_rate": 3.893982808396896e-07,
+      "loss": 0.1757,
+      "step": 15772
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.4912139802613158,
+      "learning_rate": 3.8904981745218005e-07,
+      "loss": 0.1496,
+      "step": 15773
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.1374496784188117,
+      "learning_rate": 3.887015037387643e-07,
+      "loss": 0.1935,
+      "step": 15774
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.3666092548968924,
+      "learning_rate": 3.8835333971074974e-07,
+      "loss": 0.213,
+      "step": 15775
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.4476207704102653,
+      "learning_rate": 3.880053253794369e-07,
+      "loss": 0.2049,
+      "step": 15776
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.5171944367616517,
+      "learning_rate": 3.876574607561234e-07,
+      "loss": 0.1595,
+      "step": 15777
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.384529922290127,
+      "learning_rate": 3.8730974585210135e-07,
+      "loss": 0.1955,
+      "step": 15778
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 2.0352107862608158,
+      "learning_rate": 3.8696218067865776e-07,
+      "loss": 0.207,
+      "step": 15779
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.540714983991443,
+      "learning_rate": 3.8661476524707365e-07,
+      "loss": 0.1719,
+      "step": 15780
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.4157224364779504,
+      "learning_rate": 3.8626749956862773e-07,
+      "loss": 0.241,
+      "step": 15781
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.1597399366032215,
+      "learning_rate": 3.8592038365459105e-07,
+      "loss": 0.2142,
+      "step": 15782
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.6581487298497541,
+      "learning_rate": 3.855734175162323e-07,
+      "loss": 0.1616,
+      "step": 15783
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.4940874027920752,
+      "learning_rate": 3.852266011648154e-07,
+      "loss": 0.1832,
+      "step": 15784
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.880929579172285,
+      "learning_rate": 3.84879934611595e-07,
+      "loss": 0.1497,
+      "step": 15785
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.3302371004300395,
+      "learning_rate": 3.8453341786782615e-07,
+      "loss": 0.1969,
+      "step": 15786
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.5314331966894383,
+      "learning_rate": 3.841870509447565e-07,
+      "loss": 0.2104,
+      "step": 15787
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.9605923319606043,
+      "learning_rate": 3.8384083385363036e-07,
+      "loss": 0.1319,
+      "step": 15788
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 2.7281191625893135,
+      "learning_rate": 3.8349476660568486e-07,
+      "loss": 0.2046,
+      "step": 15789
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.5256770764956236,
+      "learning_rate": 3.831488492121532e-07,
+      "loss": 0.1256,
+      "step": 15790
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.3688026968945173,
+      "learning_rate": 3.8280308168426595e-07,
+      "loss": 0.1795,
+      "step": 15791
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.2372726942109609,
+      "learning_rate": 3.8245746403324515e-07,
+      "loss": 0.1698,
+      "step": 15792
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.144306886704314,
+      "learning_rate": 3.821119962703107e-07,
+      "loss": 0.1701,
+      "step": 15793
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.4294551076821604,
+      "learning_rate": 3.817666784066759e-07,
+      "loss": 0.1486,
+      "step": 15794
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.4940826155428029,
+      "learning_rate": 3.8142151045355015e-07,
+      "loss": 0.1738,
+      "step": 15795
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.9811884225974279,
+      "learning_rate": 3.810764924221377e-07,
+      "loss": 0.1507,
+      "step": 15796
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.2380436449700238,
+      "learning_rate": 3.8073162432363976e-07,
+      "loss": 0.222,
+      "step": 15797
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.7285721936640799,
+      "learning_rate": 3.803869061692489e-07,
+      "loss": 0.1457,
+      "step": 15798
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.1995241167781363,
+      "learning_rate": 3.8004233797015513e-07,
+      "loss": 0.1437,
+      "step": 15799
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.614999779451727,
+      "learning_rate": 3.796979197375433e-07,
+      "loss": 0.1903,
+      "step": 15800
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.5434396242573918,
+      "learning_rate": 3.793536514825946e-07,
+      "loss": 0.2099,
+      "step": 15801
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.227250638109361,
+      "learning_rate": 3.7900953321648394e-07,
+      "loss": 0.109,
+      "step": 15802
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.3308868302362014,
+      "learning_rate": 3.7866556495038065e-07,
+      "loss": 0.1597,
+      "step": 15803
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.2102260683973998,
+      "learning_rate": 3.783217466954503e-07,
+      "loss": 0.1175,
+      "step": 15804
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.572311054781784,
+      "learning_rate": 3.779780784628534e-07,
+      "loss": 0.1847,
+      "step": 15805
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.3684426785702317,
+      "learning_rate": 3.776345602637471e-07,
+      "loss": 0.1545,
+      "step": 15806
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.3898495054849525,
+      "learning_rate": 3.772911921092803e-07,
+      "loss": 0.1351,
+      "step": 15807
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.5282128366333658,
+      "learning_rate": 3.7694797401059914e-07,
+      "loss": 0.1914,
+      "step": 15808
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.035511826677905,
+      "learning_rate": 3.766049059788468e-07,
+      "loss": 0.1302,
+      "step": 15809
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.3761177965161548,
+      "learning_rate": 3.762619880251561e-07,
+      "loss": 0.1954,
+      "step": 15810
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.7968562747145493,
+      "learning_rate": 3.7591922016066153e-07,
+      "loss": 0.2292,
+      "step": 15811
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.2187866608901163,
+      "learning_rate": 3.7557660239648743e-07,
+      "loss": 0.1602,
+      "step": 15812
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.3305664958428869,
+      "learning_rate": 3.7523413474375546e-07,
+      "loss": 0.1972,
+      "step": 15813
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 2.0072596712899746,
+      "learning_rate": 3.7489181721358345e-07,
+      "loss": 0.1614,
+      "step": 15814
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.4375093138434496,
+      "learning_rate": 3.745496498170831e-07,
+      "loss": 0.1953,
+      "step": 15815
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 3.186241612084753,
+      "learning_rate": 3.742076325653604e-07,
+      "loss": 0.1626,
+      "step": 15816
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.3797242815292554,
+      "learning_rate": 3.7386576546951824e-07,
+      "loss": 0.1264,
+      "step": 15817
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 2.036352629654069,
+      "learning_rate": 3.735240485406527e-07,
+      "loss": 0.2081,
+      "step": 15818
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.491209663435536,
+      "learning_rate": 3.731824817898577e-07,
+      "loss": 0.2196,
+      "step": 15819
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.7542995445215865,
+      "learning_rate": 3.728410652282205e-07,
+      "loss": 0.1469,
+      "step": 15820
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.6521204804344727,
+      "learning_rate": 3.7249979886682217e-07,
+      "loss": 0.2112,
+      "step": 15821
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.9393821465142058,
+      "learning_rate": 3.721586827167417e-07,
+      "loss": 0.2033,
+      "step": 15822
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.9334416262192793,
+      "learning_rate": 3.718177167890513e-07,
+      "loss": 0.1837,
+      "step": 15823
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.3371984180799033,
+      "learning_rate": 3.714769010948205e-07,
+      "loss": 0.1741,
+      "step": 15824
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.4635799092990112,
+      "learning_rate": 3.711362356451098e-07,
+      "loss": 0.1976,
+      "step": 15825
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.3992974285282032,
+      "learning_rate": 3.7079572045097876e-07,
+      "loss": 0.1845,
+      "step": 15826
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 2.0327233839623062,
+      "learning_rate": 3.704553555234813e-07,
+      "loss": 0.1983,
+      "step": 15827
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.451510937845646,
+      "learning_rate": 3.7011514087366476e-07,
+      "loss": 0.1629,
+      "step": 15828
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.2280191591873597,
+      "learning_rate": 3.697750765125735e-07,
+      "loss": 0.1699,
+      "step": 15829
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.5482265319071182,
+      "learning_rate": 3.6943516245124545e-07,
+      "loss": 0.1951,
+      "step": 15830
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.4267143122773531,
+      "learning_rate": 3.6909539870071454e-07,
+      "loss": 0.2289,
+      "step": 15831
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.5733408313895159,
+      "learning_rate": 3.6875578527200974e-07,
+      "loss": 0.189,
+      "step": 15832
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.2834958478539464,
+      "learning_rate": 3.6841632217615666e-07,
+      "loss": 0.1664,
+      "step": 15833
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.241913777308962,
+      "learning_rate": 3.68077009424172e-07,
+      "loss": 0.1679,
+      "step": 15834
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.4015198627735623,
+      "learning_rate": 3.677378470270709e-07,
+      "loss": 0.1736,
+      "step": 15835
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.447805430121053,
+      "learning_rate": 3.673988349958635e-07,
+      "loss": 0.2109,
+      "step": 15836
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.3362126262298493,
+      "learning_rate": 3.670599733415536e-07,
+      "loss": 0.1455,
+      "step": 15837
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.5166918450752815,
+      "learning_rate": 3.66721262075142e-07,
+      "loss": 0.1759,
+      "step": 15838
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.29571688888825,
+      "learning_rate": 3.663827012076215e-07,
+      "loss": 0.2205,
+      "step": 15839
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.633399082299114,
+      "learning_rate": 3.6604429074998336e-07,
+      "loss": 0.1403,
+      "step": 15840
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.768920009718902,
+      "learning_rate": 3.657060307132121e-07,
+      "loss": 0.2095,
+      "step": 15841
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.4753951246593437,
+      "learning_rate": 3.653679211082889e-07,
+      "loss": 0.1673,
+      "step": 15842
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.5685551546115344,
+      "learning_rate": 3.650299619461867e-07,
+      "loss": 0.1786,
+      "step": 15843
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.7214151266810447,
+      "learning_rate": 3.6469215323787776e-07,
+      "loss": 0.1732,
+      "step": 15844
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.4227783770776974,
+      "learning_rate": 3.6435449499432785e-07,
+      "loss": 0.1815,
+      "step": 15845
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.1444652655527392,
+      "learning_rate": 3.640169872264959e-07,
+      "loss": 0.161,
+      "step": 15846
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.972352040317304,
+      "learning_rate": 3.6367962994533934e-07,
+      "loss": 0.1841,
+      "step": 15847
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.0047992007927293,
+      "learning_rate": 3.6334242316180714e-07,
+      "loss": 0.0985,
+      "step": 15848
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.3914665718860881,
+      "learning_rate": 3.630053668868461e-07,
+      "loss": 0.1259,
+      "step": 15849
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.7166703560551868,
+      "learning_rate": 3.626684611313974e-07,
+      "loss": 0.1824,
+      "step": 15850
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.7061269097416365,
+      "learning_rate": 3.6233170590639744e-07,
+      "loss": 0.2402,
+      "step": 15851
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.46529507868493,
+      "learning_rate": 3.619951012227768e-07,
+      "loss": 0.2078,
+      "step": 15852
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.4161543100778424,
+      "learning_rate": 3.616586470914624e-07,
+      "loss": 0.1449,
+      "step": 15853
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.5041997945083916,
+      "learning_rate": 3.6132234352337534e-07,
+      "loss": 0.1344,
+      "step": 15854
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.491155647097196,
+      "learning_rate": 3.609861905294326e-07,
+      "loss": 0.1597,
+      "step": 15855
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.0552374571534011,
+      "learning_rate": 3.6065018812054595e-07,
+      "loss": 0.1506,
+      "step": 15856
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.1403304301018145,
+      "learning_rate": 3.603143363076217e-07,
+      "loss": 0.1844,
+      "step": 15857
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.1805105369841649,
+      "learning_rate": 3.5997863510156216e-07,
+      "loss": 0.1593,
+      "step": 15858
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.741382506984373,
+      "learning_rate": 3.5964308451326425e-07,
+      "loss": 0.2113,
+      "step": 15859
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.0398780160785217,
+      "learning_rate": 3.5930768455362086e-07,
+      "loss": 0.1742,
+      "step": 15860
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.7293355602679232,
+      "learning_rate": 3.5897243523351776e-07,
+      "loss": 0.1785,
+      "step": 15861
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.640966043901465,
+      "learning_rate": 3.586373365638385e-07,
+      "loss": 0.1944,
+      "step": 15862
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.4284025734282835,
+      "learning_rate": 3.5830238855546097e-07,
+      "loss": 0.1897,
+      "step": 15863
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.7530733522501338,
+      "learning_rate": 3.579675912192565e-07,
+      "loss": 0.2006,
+      "step": 15864
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.4339438902133876,
+      "learning_rate": 3.5763294456609366e-07,
+      "loss": 0.1746,
+      "step": 15865
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.2356154631139071,
+      "learning_rate": 3.572984486068343e-07,
+      "loss": 0.1585,
+      "step": 15866
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.4438954236743704,
+      "learning_rate": 3.5696410335233743e-07,
+      "loss": 0.1755,
+      "step": 15867
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.4881387777543826,
+      "learning_rate": 3.5662990881345563e-07,
+      "loss": 0.1652,
+      "step": 15868
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.577972404731849,
+      "learning_rate": 3.562958650010373e-07,
+      "loss": 0.2111,
+      "step": 15869
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.4926170310106495,
+      "learning_rate": 3.559619719259255e-07,
+      "loss": 0.2001,
+      "step": 15870
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.6679123862255532,
+      "learning_rate": 3.556282295989583e-07,
+      "loss": 0.2182,
+      "step": 15871
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.4983852804987594,
+      "learning_rate": 3.552946380309691e-07,
+      "loss": 0.183,
+      "step": 15872
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.6268603560616264,
+      "learning_rate": 3.549611972327871e-07,
+      "loss": 0.1719,
+      "step": 15873
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.5621197523920405,
+      "learning_rate": 3.5462790721523697e-07,
+      "loss": 0.1476,
+      "step": 15874
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.9720424032621531,
+      "learning_rate": 3.5429476798913507e-07,
+      "loss": 0.1856,
+      "step": 15875
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.8815847643954313,
+      "learning_rate": 3.539617795652961e-07,
+      "loss": 0.1697,
+      "step": 15876
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.6138075646609045,
+      "learning_rate": 3.536289419545302e-07,
+      "loss": 0.2523,
+      "step": 15877
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.1677468321877111,
+      "learning_rate": 3.532962551676411e-07,
+      "loss": 0.155,
+      "step": 15878
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.5523182341307715,
+      "learning_rate": 3.5296371921542673e-07,
+      "loss": 0.2134,
+      "step": 15879
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 2.480602759398285,
+      "learning_rate": 3.5263133410868234e-07,
+      "loss": 0.2256,
+      "step": 15880
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.563646558661006,
+      "learning_rate": 3.5229909985819823e-07,
+      "loss": 0.2051,
+      "step": 15881
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.47235010137021,
+      "learning_rate": 3.5196701647475685e-07,
+      "loss": 0.1869,
+      "step": 15882
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.4354376147983492,
+      "learning_rate": 3.5163508396913957e-07,
+      "loss": 0.176,
+      "step": 15883
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.214630764508954,
+      "learning_rate": 3.5130330235212006e-07,
+      "loss": 0.1734,
+      "step": 15884
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.7712483106832586,
+      "learning_rate": 3.5097167163446855e-07,
+      "loss": 0.1841,
+      "step": 15885
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.9018005284807606,
+      "learning_rate": 3.5064019182694977e-07,
+      "loss": 0.1943,
+      "step": 15886
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.9692695855438818,
+      "learning_rate": 3.5030886294032507e-07,
+      "loss": 0.2376,
+      "step": 15887
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.347433961973434,
+      "learning_rate": 3.499776849853476e-07,
+      "loss": 0.1283,
+      "step": 15888
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.3263505525968415,
+      "learning_rate": 3.496466579727681e-07,
+      "loss": 0.1789,
+      "step": 15889
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.5009308102546208,
+      "learning_rate": 3.493157819133325e-07,
+      "loss": 0.1875,
+      "step": 15890
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.4928811942416558,
+      "learning_rate": 3.489850568177805e-07,
+      "loss": 0.1687,
+      "step": 15891
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.4165251558347745,
+      "learning_rate": 3.486544826968491e-07,
+      "loss": 0.2147,
+      "step": 15892
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.6039147943045218,
+      "learning_rate": 3.483240595612675e-07,
+      "loss": 0.1979,
+      "step": 15893
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.901461762993996,
+      "learning_rate": 3.479937874217615e-07,
+      "loss": 0.1716,
+      "step": 15894
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.3412283091199901,
+      "learning_rate": 3.476636662890526e-07,
+      "loss": 0.189,
+      "step": 15895
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.6806609681814149,
+      "learning_rate": 3.4733369617385657e-07,
+      "loss": 0.2086,
+      "step": 15896
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.3407847322659492,
+      "learning_rate": 3.470038770868839e-07,
+      "loss": 0.1803,
+      "step": 15897
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.147203768867057,
+      "learning_rate": 3.466742090388414e-07,
+      "loss": 0.1643,
+      "step": 15898
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.4750419436972886,
+      "learning_rate": 3.463446920404295e-07,
+      "loss": 0.1598,
+      "step": 15899
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.4099300340446088,
+      "learning_rate": 3.4601532610234525e-07,
+      "loss": 0.1974,
+      "step": 15900
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.8849926431345365,
+      "learning_rate": 3.4568611123527997e-07,
+      "loss": 0.1567,
+      "step": 15901
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.6480802854056917,
+      "learning_rate": 3.4535704744991907e-07,
+      "loss": 0.1715,
+      "step": 15902
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.4229050719829948,
+      "learning_rate": 3.450281347569456e-07,
+      "loss": 0.1491,
+      "step": 15903
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.7948170113466824,
+      "learning_rate": 3.4469937316703507e-07,
+      "loss": 0.1714,
+      "step": 15904
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.6869162326846043,
+      "learning_rate": 3.4437076269086047e-07,
+      "loss": 0.2044,
+      "step": 15905
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.6953327116222883,
+      "learning_rate": 3.440423033390883e-07,
+      "loss": 0.1518,
+      "step": 15906
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.4920026854460227,
+      "learning_rate": 3.4371399512238e-07,
+      "loss": 0.1563,
+      "step": 15907
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.8135740452183429,
+      "learning_rate": 3.433858380513927e-07,
+      "loss": 0.1793,
+      "step": 15908
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.41123227095639,
+      "learning_rate": 3.430578321367789e-07,
+      "loss": 0.1649,
+      "step": 15909
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.1337636703289518,
+      "learning_rate": 3.427299773891868e-07,
+      "loss": 0.1689,
+      "step": 15910
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.562770924921934,
+      "learning_rate": 3.424022738192567e-07,
+      "loss": 0.1804,
+      "step": 15911
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.7928063479074106,
+      "learning_rate": 3.420747214376269e-07,
+      "loss": 0.2241,
+      "step": 15912
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.5813778693625653,
+      "learning_rate": 3.417473202549304e-07,
+      "loss": 0.2077,
+      "step": 15913
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.42886884628905,
+      "learning_rate": 3.4142007028179546e-07,
+      "loss": 0.1643,
+      "step": 15914
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.7044855947532018,
+      "learning_rate": 3.4109297152884303e-07,
+      "loss": 0.2094,
+      "step": 15915
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.5595251135346597,
+      "learning_rate": 3.407660240066929e-07,
+      "loss": 0.2018,
+      "step": 15916
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.484429207866072,
+      "learning_rate": 3.404392277259555e-07,
+      "loss": 0.2431,
+      "step": 15917
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.534721759529485,
+      "learning_rate": 3.4011258269724067e-07,
+      "loss": 0.2051,
+      "step": 15918
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.170254015384322,
+      "learning_rate": 3.3978608893115163e-07,
+      "loss": 0.1849,
+      "step": 15919
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.6142466176055037,
+      "learning_rate": 3.3945974643828537e-07,
+      "loss": 0.1934,
+      "step": 15920
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.9794798645027403,
+      "learning_rate": 3.3913355522923573e-07,
+      "loss": 0.157,
+      "step": 15921
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.2020259983855954,
+      "learning_rate": 3.388075153145909e-07,
+      "loss": 0.1278,
+      "step": 15922
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.5145675863875503,
+      "learning_rate": 3.3848162670493456e-07,
+      "loss": 0.1867,
+      "step": 15923
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.0878978708460798,
+      "learning_rate": 3.381558894108455e-07,
+      "loss": 0.1582,
+      "step": 15924
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.2674674528032497,
+      "learning_rate": 3.37830303442897e-07,
+      "loss": 0.1521,
+      "step": 15925
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.3130201830238792,
+      "learning_rate": 3.375048688116572e-07,
+      "loss": 0.1496,
+      "step": 15926
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.641447705667254,
+      "learning_rate": 3.3717958552769103e-07,
+      "loss": 0.1862,
+      "step": 15927
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.4840905669902476,
+      "learning_rate": 3.368544536015572e-07,
+      "loss": 0.1649,
+      "step": 15928
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.8290836966762742,
+      "learning_rate": 3.3652947304380844e-07,
+      "loss": 0.2003,
+      "step": 15929
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.642235338002463,
+      "learning_rate": 3.362046438649952e-07,
+      "loss": 0.1855,
+      "step": 15930
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.672133840150144,
+      "learning_rate": 3.3587996607566063e-07,
+      "loss": 0.2029,
+      "step": 15931
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.3368633029349337,
+      "learning_rate": 3.355554396863453e-07,
+      "loss": 0.1544,
+      "step": 15932
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.862859458286675,
+      "learning_rate": 3.352310647075818e-07,
+      "loss": 0.1805,
+      "step": 15933
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.3942535294817622,
+      "learning_rate": 3.349068411499018e-07,
+      "loss": 0.1864,
+      "step": 15934
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.6527896486874718,
+      "learning_rate": 3.345827690238268e-07,
+      "loss": 0.1862,
+      "step": 15935
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.4908108577380985,
+      "learning_rate": 3.342588483398784e-07,
+      "loss": 0.1489,
+      "step": 15936
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.374881847680494,
+      "learning_rate": 3.33935079108571e-07,
+      "loss": 0.1484,
+      "step": 15937
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.6937927826619539,
+      "learning_rate": 3.33611461340414e-07,
+      "loss": 0.1892,
+      "step": 15938
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.2596640257563334,
+      "learning_rate": 3.332879950459117e-07,
+      "loss": 0.1598,
+      "step": 15939
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 2.0489555479268966,
+      "learning_rate": 3.329646802355652e-07,
+      "loss": 0.1516,
+      "step": 15940
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.2723751580813223,
+      "learning_rate": 3.3264151691986824e-07,
+      "loss": 0.165,
+      "step": 15941
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.1725901265470233,
+      "learning_rate": 3.3231850510931307e-07,
+      "loss": 0.1621,
+      "step": 15942
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.5758552375032973,
+      "learning_rate": 3.319956448143818e-07,
+      "loss": 0.1823,
+      "step": 15943
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.4002873357351855,
+      "learning_rate": 3.3167293604555673e-07,
+      "loss": 0.131,
+      "step": 15944
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.2391283855984598,
+      "learning_rate": 3.313503788133127e-07,
+      "loss": 0.2077,
+      "step": 15945
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.8177709723521767,
+      "learning_rate": 3.3102797312812075e-07,
+      "loss": 0.2479,
+      "step": 15946
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.6829224418634863,
+      "learning_rate": 3.307057190004448e-07,
+      "loss": 0.233,
+      "step": 15947
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.6691271982472056,
+      "learning_rate": 3.303836164407459e-07,
+      "loss": 0.1951,
+      "step": 15948
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.307010381638041,
+      "learning_rate": 3.300616654594807e-07,
+      "loss": 0.2063,
+      "step": 15949
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.660695035682979,
+      "learning_rate": 3.297398660670992e-07,
+      "loss": 0.1841,
+      "step": 15950
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.9971215873422168,
+      "learning_rate": 3.294182182740474e-07,
+      "loss": 0.1916,
+      "step": 15951
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.6496613013992245,
+      "learning_rate": 3.2909672209076596e-07,
+      "loss": 0.2169,
+      "step": 15952
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.6093770601203843,
+      "learning_rate": 3.2877537752769027e-07,
+      "loss": 0.2335,
+      "step": 15953
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.7463764926094993,
+      "learning_rate": 3.284541845952527e-07,
+      "loss": 0.1688,
+      "step": 15954
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.4350628095099591,
+      "learning_rate": 3.2813314330387866e-07,
+      "loss": 0.2161,
+      "step": 15955
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.3215464632365184,
+      "learning_rate": 3.2781225366398885e-07,
+      "loss": 0.1352,
+      "step": 15956
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.6411617944930534,
+      "learning_rate": 3.274915156859998e-07,
+      "loss": 0.153,
+      "step": 15957
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.455694285767038,
+      "learning_rate": 3.2717092938032327e-07,
+      "loss": 0.125,
+      "step": 15958
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.540324133333547,
+      "learning_rate": 3.2685049475736475e-07,
+      "loss": 0.1754,
+      "step": 15959
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.285278766670769,
+      "learning_rate": 3.265302118275282e-07,
+      "loss": 0.1658,
+      "step": 15960
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.1987837832029242,
+      "learning_rate": 3.2621008060120697e-07,
+      "loss": 0.1492,
+      "step": 15961
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.3944514756666067,
+      "learning_rate": 3.2589010108879437e-07,
+      "loss": 0.1891,
+      "step": 15962
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.7040869048179552,
+      "learning_rate": 3.255702733006766e-07,
+      "loss": 0.2396,
+      "step": 15963
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.5841604151427668,
+      "learning_rate": 3.25250597247237e-07,
+      "loss": 0.1661,
+      "step": 15964
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.7241087734546021,
+      "learning_rate": 3.2493107293884995e-07,
+      "loss": 0.179,
+      "step": 15965
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.478044059263187,
+      "learning_rate": 3.2461170038588896e-07,
+      "loss": 0.1856,
+      "step": 15966
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.4481853800646112,
+      "learning_rate": 3.242924795987207e-07,
+      "loss": 0.1654,
+      "step": 15967
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.1627580194850815,
+      "learning_rate": 3.239734105877079e-07,
+      "loss": 0.171,
+      "step": 15968
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.0152315789947817,
+      "learning_rate": 3.236544933632063e-07,
+      "loss": 0.1466,
+      "step": 15969
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.5178287549529776,
+      "learning_rate": 3.233357279355698e-07,
+      "loss": 0.1435,
+      "step": 15970
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.5590518096661303,
+      "learning_rate": 3.230171143151445e-07,
+      "loss": 0.1427,
+      "step": 15971
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.1907854870644614,
+      "learning_rate": 3.2269865251227275e-07,
+      "loss": 0.1526,
+      "step": 15972
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.0059447687436893,
+      "learning_rate": 3.223803425372929e-07,
+      "loss": 0.1567,
+      "step": 15973
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.7819358065949908,
+      "learning_rate": 3.2206218440053615e-07,
+      "loss": 0.1565,
+      "step": 15974
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.7414460122457947,
+      "learning_rate": 3.217441781123315e-07,
+      "loss": 0.2088,
+      "step": 15975
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.7466398703596582,
+      "learning_rate": 3.2142632368300067e-07,
+      "loss": 0.2207,
+      "step": 15976
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.4047106795048057,
+      "learning_rate": 3.2110862112286146e-07,
+      "loss": 0.1473,
+      "step": 15977
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 2.030272292540036,
+      "learning_rate": 3.2079107044222793e-07,
+      "loss": 0.1763,
+      "step": 15978
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.4815319893412355,
+      "learning_rate": 3.2047367165140676e-07,
+      "loss": 0.137,
+      "step": 15979
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.4615178230582573,
+      "learning_rate": 3.201564247607003e-07,
+      "loss": 0.2015,
+      "step": 15980
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.8755005962957676,
+      "learning_rate": 3.1983932978040813e-07,
+      "loss": 0.1758,
+      "step": 15981
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.285135895124368,
+      "learning_rate": 3.1952238672082246e-07,
+      "loss": 0.1642,
+      "step": 15982
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.4470116190335103,
+      "learning_rate": 3.1920559559223185e-07,
+      "loss": 0.2113,
+      "step": 15983
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.508120495464998,
+      "learning_rate": 3.188889564049186e-07,
+      "loss": 0.1727,
+      "step": 15984
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.3655972742020241,
+      "learning_rate": 3.185724691691616e-07,
+      "loss": 0.1917,
+      "step": 15985
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.8966466910833308,
+      "learning_rate": 3.182561338952356e-07,
+      "loss": 0.1753,
+      "step": 15986
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.2893056466796373,
+      "learning_rate": 3.1793995059340664e-07,
+      "loss": 0.164,
+      "step": 15987
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.2813949154102624,
+      "learning_rate": 3.1762391927393997e-07,
+      "loss": 0.1956,
+      "step": 15988
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.6183791906795044,
+      "learning_rate": 3.173080399470924e-07,
+      "loss": 0.1972,
+      "step": 15989
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.242811481133941,
+      "learning_rate": 3.169923126231189e-07,
+      "loss": 0.1536,
+      "step": 15990
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.6091172465606776,
+      "learning_rate": 3.1667673731226867e-07,
+      "loss": 0.1554,
+      "step": 15991
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.7491415336436484,
+      "learning_rate": 3.1636131402478345e-07,
+      "loss": 0.1849,
+      "step": 15992
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.2862062426020382,
+      "learning_rate": 3.160460427709039e-07,
+      "loss": 0.1792,
+      "step": 15993
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.5399059317936625,
+      "learning_rate": 3.1573092356086297e-07,
+      "loss": 0.1618,
+      "step": 15994
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.4371648428836163,
+      "learning_rate": 3.1541595640488974e-07,
+      "loss": 0.1478,
+      "step": 15995
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.2634844987436915,
+      "learning_rate": 3.1510114131320935e-07,
+      "loss": 0.1558,
+      "step": 15996
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.780089560851413,
+      "learning_rate": 3.1478647829603914e-07,
+      "loss": 0.1805,
+      "step": 15997
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.5213218866632823,
+      "learning_rate": 3.144719673635943e-07,
+      "loss": 0.1794,
+      "step": 15998
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.471367299744275,
+      "learning_rate": 3.1415760852608333e-07,
+      "loss": 0.1676,
+      "step": 15999
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.428965433103739,
+      "learning_rate": 3.1384340179371197e-07,
+      "loss": 0.1253,
+      "step": 16000
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.1594311924192646,
+      "learning_rate": 3.135293471766776e-07,
+      "loss": 0.1616,
+      "step": 16001
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.9074855732090996,
+      "learning_rate": 3.132154446851754e-07,
+      "loss": 0.178,
+      "step": 16002
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.9930999420082453,
+      "learning_rate": 3.1290169432939556e-07,
+      "loss": 0.106,
+      "step": 16003
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.3749358693252827,
+      "learning_rate": 3.125880961195221e-07,
+      "loss": 0.1381,
+      "step": 16004
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.2702646660567158,
+      "learning_rate": 3.122746500657342e-07,
+      "loss": 0.1845,
+      "step": 16005
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.555707141880805,
+      "learning_rate": 3.11961356178207e-07,
+      "loss": 0.2445,
+      "step": 16006
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.3024753530379747,
+      "learning_rate": 3.116482144671096e-07,
+      "loss": 0.1179,
+      "step": 16007
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.8172708566035303,
+      "learning_rate": 3.1133522494260714e-07,
+      "loss": 0.2136,
+      "step": 16008
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.2194121102806452,
+      "learning_rate": 3.110223876148605e-07,
+      "loss": 0.1639,
+      "step": 16009
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.687586662928181,
+      "learning_rate": 3.10709702494022e-07,
+      "loss": 0.1407,
+      "step": 16010
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.6865021263061657,
+      "learning_rate": 3.1039716959024355e-07,
+      "loss": 0.1577,
+      "step": 16011
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.3145188404451065,
+      "learning_rate": 3.1008478891366986e-07,
+      "loss": 0.2492,
+      "step": 16012
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.6808721822454182,
+      "learning_rate": 3.097725604744406e-07,
+      "loss": 0.1909,
+      "step": 16013
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.5651335718774606,
+      "learning_rate": 3.0946048428269203e-07,
+      "loss": 0.1979,
+      "step": 16014
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.4608898920064117,
+      "learning_rate": 3.091485603485522e-07,
+      "loss": 0.1508,
+      "step": 16015
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.4627462501738604,
+      "learning_rate": 3.088367886821481e-07,
+      "loss": 0.1576,
+      "step": 16016
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.4341022514707564,
+      "learning_rate": 3.085251692935992e-07,
+      "loss": 0.1739,
+      "step": 16017
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.7134073537191838,
+      "learning_rate": 3.0821370219302204e-07,
+      "loss": 0.1896,
+      "step": 16018
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.615822082024261,
+      "learning_rate": 3.0790238739052515e-07,
+      "loss": 0.185,
+      "step": 16019
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.5110770842526877,
+      "learning_rate": 3.0759122489621487e-07,
+      "loss": 0.1963,
+      "step": 16020
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.2440918773640766,
+      "learning_rate": 3.072802147201914e-07,
+      "loss": 0.1604,
+      "step": 16021
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.800788027074147,
+      "learning_rate": 3.069693568725518e-07,
+      "loss": 0.2156,
+      "step": 16022
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.228998050061938,
+      "learning_rate": 3.066586513633851e-07,
+      "loss": 0.147,
+      "step": 16023
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.7424579974876662,
+      "learning_rate": 3.0634809820277775e-07,
+      "loss": 0.1782,
+      "step": 16024
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.175231044454939,
+      "learning_rate": 3.0603769740080944e-07,
+      "loss": 0.1435,
+      "step": 16025
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.774734392576027,
+      "learning_rate": 3.0572744896755714e-07,
+      "loss": 0.1775,
+      "step": 16026
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.550226890777745,
+      "learning_rate": 3.054173529130916e-07,
+      "loss": 0.1936,
+      "step": 16027
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.3187824870126106,
+      "learning_rate": 3.051074092474776e-07,
+      "loss": 0.2023,
+      "step": 16028
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.204678158408434,
+      "learning_rate": 3.0479761798077713e-07,
+      "loss": 0.1786,
+      "step": 16029
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.1538703375660058,
+      "learning_rate": 3.044879791230454e-07,
+      "loss": 0.1926,
+      "step": 16030
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.2502675723751933,
+      "learning_rate": 3.0417849268433443e-07,
+      "loss": 0.1405,
+      "step": 16031
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.44970703125,
+      "learning_rate": 3.038691586746906e-07,
+      "loss": 0.1682,
+      "step": 16032
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.448931611603367,
+      "learning_rate": 3.0355997710415363e-07,
+      "loss": 0.148,
+      "step": 16033
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.5117744612962933,
+      "learning_rate": 3.0325094798276e-07,
+      "loss": 0.1789,
+      "step": 16034
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.0720111843822757,
+      "learning_rate": 3.0294207132054165e-07,
+      "loss": 0.1189,
+      "step": 16035
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.3631961954467462,
+      "learning_rate": 3.0263334712752554e-07,
+      "loss": 0.213,
+      "step": 16036
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.3212733976192559,
+      "learning_rate": 3.0232477541373153e-07,
+      "loss": 0.1779,
+      "step": 16037
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.2079434265888995,
+      "learning_rate": 3.0201635618917646e-07,
+      "loss": 0.1364,
+      "step": 16038
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.6369597136987954,
+      "learning_rate": 3.0170808946387185e-07,
+      "loss": 0.1472,
+      "step": 16039
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.2859561193423006,
+      "learning_rate": 3.013999752478258e-07,
+      "loss": 0.1365,
+      "step": 16040
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.6657496527356996,
+      "learning_rate": 3.01092013551037e-07,
+      "loss": 0.2395,
+      "step": 16041
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.1670060480164561,
+      "learning_rate": 3.007842043835046e-07,
+      "loss": 0.1225,
+      "step": 16042
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.1256629990280353,
+      "learning_rate": 3.0047654775521793e-07,
+      "loss": 0.155,
+      "step": 16043
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.1831111711025377,
+      "learning_rate": 3.0016904367616564e-07,
+      "loss": 0.1594,
+      "step": 16044
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.63357500588214,
+      "learning_rate": 2.9986169215632907e-07,
+      "loss": 0.1919,
+      "step": 16045
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.3462436957818267,
+      "learning_rate": 2.995544932056843e-07,
+      "loss": 0.1979,
+      "step": 16046
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.5120073777301364,
+      "learning_rate": 2.9924744683420324e-07,
+      "loss": 0.1992,
+      "step": 16047
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.098206701083581,
+      "learning_rate": 2.9894055305185355e-07,
+      "loss": 0.0934,
+      "step": 16048
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.9965869271919674,
+      "learning_rate": 2.9863381186859663e-07,
+      "loss": 0.1829,
+      "step": 16049
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.0922054386166529,
+      "learning_rate": 2.9832722329439015e-07,
+      "loss": 0.1319,
+      "step": 16050
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.287894922399134,
+      "learning_rate": 2.980207873391855e-07,
+      "loss": 0.167,
+      "step": 16051
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.682907300880377,
+      "learning_rate": 2.9771450401292976e-07,
+      "loss": 0.1866,
+      "step": 16052
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.4588538472770969,
+      "learning_rate": 2.9740837332556504e-07,
+      "loss": 0.1538,
+      "step": 16053
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.4297708059752425,
+      "learning_rate": 2.9710239528702944e-07,
+      "loss": 0.1598,
+      "step": 16054
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.6173713183400547,
+      "learning_rate": 2.9679656990725334e-07,
+      "loss": 0.2233,
+      "step": 16055
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.0645955121761472,
+      "learning_rate": 2.964908971961661e-07,
+      "loss": 0.181,
+      "step": 16056
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.37731202771142,
+      "learning_rate": 2.96185377163688e-07,
+      "loss": 0.1676,
+      "step": 16057
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.3734801519061737,
+      "learning_rate": 2.9588000981973896e-07,
+      "loss": 0.1923,
+      "step": 16058
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.1148166976672604,
+      "learning_rate": 2.955747951742288e-07,
+      "loss": 0.133,
+      "step": 16059
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 2.0715422450842182,
+      "learning_rate": 2.9526973323706685e-07,
+      "loss": 0.2093,
+      "step": 16060
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.589982422725519,
+      "learning_rate": 2.9496482401815405e-07,
+      "loss": 0.1505,
+      "step": 16061
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.466644487452604,
+      "learning_rate": 2.9466006752738865e-07,
+      "loss": 0.1961,
+      "step": 16062
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.464837605781385,
+      "learning_rate": 2.943554637746637e-07,
+      "loss": 0.1827,
+      "step": 16063
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.7392041788715142,
+      "learning_rate": 2.940510127698659e-07,
+      "loss": 0.1953,
+      "step": 16064
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.2277508737279013,
+      "learning_rate": 2.937467145228778e-07,
+      "loss": 0.1395,
+      "step": 16065
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.3482685688327374,
+      "learning_rate": 2.9344256904357827e-07,
+      "loss": 0.2058,
+      "step": 16066
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.0620815210313226,
+      "learning_rate": 2.9313857634183927e-07,
+      "loss": 0.1327,
+      "step": 16067
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.9989056424393687,
+      "learning_rate": 2.928347364275291e-07,
+      "loss": 0.2073,
+      "step": 16068
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.4541117896443463,
+      "learning_rate": 2.925310493105099e-07,
+      "loss": 0.1666,
+      "step": 16069
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 3.1904619517123853,
+      "learning_rate": 2.9222751500063983e-07,
+      "loss": 0.1908,
+      "step": 16070
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.2984010496199037,
+      "learning_rate": 2.919241335077716e-07,
+      "loss": 0.1374,
+      "step": 16071
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.03177510244124,
+      "learning_rate": 2.91620904841754e-07,
+      "loss": 0.1513,
+      "step": 16072
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.7160702839740243,
+      "learning_rate": 2.9131782901242857e-07,
+      "loss": 0.1747,
+      "step": 16073
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.4889477916778653,
+      "learning_rate": 2.910149060296347e-07,
+      "loss": 0.1638,
+      "step": 16074
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.7759445845358426,
+      "learning_rate": 2.907121359032039e-07,
+      "loss": 0.1796,
+      "step": 16075
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 2.0955705414114614,
+      "learning_rate": 2.9040951864296674e-07,
+      "loss": 0.1625,
+      "step": 16076
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.1879455928760638,
+      "learning_rate": 2.9010705425874363e-07,
+      "loss": 0.1402,
+      "step": 16077
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.3049176921045471,
+      "learning_rate": 2.8980474276035507e-07,
+      "loss": 0.1744,
+      "step": 16078
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.0033660652491847,
+      "learning_rate": 2.8950258415761154e-07,
+      "loss": 0.1337,
+      "step": 16079
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.5301961921909641,
+      "learning_rate": 2.8920057846032355e-07,
+      "loss": 0.2296,
+      "step": 16080
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.559690868516975,
+      "learning_rate": 2.888987256782944e-07,
+      "loss": 0.1687,
+      "step": 16081
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 2.3602411878013414,
+      "learning_rate": 2.885970258213211e-07,
+      "loss": 0.1776,
+      "step": 16082
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.3192904003538104,
+      "learning_rate": 2.8829547889919706e-07,
+      "loss": 0.1479,
+      "step": 16083
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.1191022643327067,
+      "learning_rate": 2.879940849217117e-07,
+      "loss": 0.1706,
+      "step": 16084
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.7917035361495937,
+      "learning_rate": 2.876928438986476e-07,
+      "loss": 0.1836,
+      "step": 16085
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.2515873516010936,
+      "learning_rate": 2.873917558397843e-07,
+      "loss": 0.1627,
+      "step": 16086
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.2135116105457455,
+      "learning_rate": 2.8709082075489394e-07,
+      "loss": 0.2066,
+      "step": 16087
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 2.0822901752638967,
+      "learning_rate": 2.8679003865374534e-07,
+      "loss": 0.1857,
+      "step": 16088
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.5507992568039608,
+      "learning_rate": 2.8648940954610294e-07,
+      "loss": 0.1747,
+      "step": 16089
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.5723153716563525,
+      "learning_rate": 2.86188933441725e-07,
+      "loss": 0.2092,
+      "step": 16090
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.2504691018596565,
+      "learning_rate": 2.858886103503644e-07,
+      "loss": 0.1903,
+      "step": 16091
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.51463297203329,
+      "learning_rate": 2.8558844028177044e-07,
+      "loss": 0.1531,
+      "step": 16092
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.1268853773780687,
+      "learning_rate": 2.852884232456865e-07,
+      "loss": 0.1288,
+      "step": 16093
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.6460572907221669,
+      "learning_rate": 2.849885592518525e-07,
+      "loss": 0.2208,
+      "step": 16094
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.517702350417026,
+      "learning_rate": 2.8468884831000024e-07,
+      "loss": 0.1487,
+      "step": 16095
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.7160862134225976,
+      "learning_rate": 2.843892904298601e-07,
+      "loss": 0.2001,
+      "step": 16096
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 2.1141777463910083,
+      "learning_rate": 2.8408988562115494e-07,
+      "loss": 0.1917,
+      "step": 16097
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 2.0046046045874637,
+      "learning_rate": 2.837906338936036e-07,
+      "loss": 0.1956,
+      "step": 16098
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.264393643078185,
+      "learning_rate": 2.834915352569206e-07,
+      "loss": 0.1632,
+      "step": 16099
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.3292981968597646,
+      "learning_rate": 2.831925897208143e-07,
+      "loss": 0.213,
+      "step": 16100
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 7.386970543970105,
+      "learning_rate": 2.828937972949891e-07,
+      "loss": 0.1676,
+      "step": 16101
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.3970167972400338,
+      "learning_rate": 2.825951579891434e-07,
+      "loss": 0.182,
+      "step": 16102
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.574172057129018,
+      "learning_rate": 2.8229667181297214e-07,
+      "loss": 0.1809,
+      "step": 16103
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.3298638993328913,
+      "learning_rate": 2.819983387761638e-07,
+      "loss": 0.1456,
+      "step": 16104
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.6357771684692428,
+      "learning_rate": 2.817001588884022e-07,
+      "loss": 0.1769,
+      "step": 16105
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.3411210591765605,
+      "learning_rate": 2.814021321593663e-07,
+      "loss": 0.1962,
+      "step": 16106
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.6749722766717035,
+      "learning_rate": 2.8110425859873057e-07,
+      "loss": 0.1615,
+      "step": 16107
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 2.137317399261943,
+      "learning_rate": 2.808065382161651e-07,
+      "loss": 0.1773,
+      "step": 16108
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.791268833111974,
+      "learning_rate": 2.8050897102133267e-07,
+      "loss": 0.2109,
+      "step": 16109
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.3607065869894173,
+      "learning_rate": 2.8021155702389223e-07,
+      "loss": 0.1295,
+      "step": 16110
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.5341529965279461,
+      "learning_rate": 2.799142962334994e-07,
+      "loss": 0.1997,
+      "step": 16111
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 2.284347247517428,
+      "learning_rate": 2.796171886598031e-07,
+      "loss": 0.2288,
+      "step": 16112
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.6624644311527457,
+      "learning_rate": 2.7932023431244683e-07,
+      "loss": 0.187,
+      "step": 16113
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.9611812531242812,
+      "learning_rate": 2.7902343320107104e-07,
+      "loss": 0.1907,
+      "step": 16114
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.4814482697046847,
+      "learning_rate": 2.7872678533530815e-07,
+      "loss": 0.2152,
+      "step": 16115
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.7214115083255253,
+      "learning_rate": 2.7843029072478934e-07,
+      "loss": 0.2199,
+      "step": 16116
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.4888705766367305,
+      "learning_rate": 2.781339493791385e-07,
+      "loss": 0.1643,
+      "step": 16117
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.8736018053915087,
+      "learning_rate": 2.778377613079747e-07,
+      "loss": 0.177,
+      "step": 16118
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 2.1624290967353694,
+      "learning_rate": 2.775417265209124e-07,
+      "loss": 0.2141,
+      "step": 16119
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 2.3738774545429986,
+      "learning_rate": 2.772458450275617e-07,
+      "loss": 0.1456,
+      "step": 16120
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 2.056493407918955,
+      "learning_rate": 2.7695011683752606e-07,
+      "loss": 0.193,
+      "step": 16121
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.1508904813784024,
+      "learning_rate": 2.7665454196040665e-07,
+      "loss": 0.2073,
+      "step": 16122
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.285515216216173,
+      "learning_rate": 2.763591204057958e-07,
+      "loss": 0.1582,
+      "step": 16123
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.6290545316026002,
+      "learning_rate": 2.7606385218328427e-07,
+      "loss": 0.1689,
+      "step": 16124
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.4072539401650648,
+      "learning_rate": 2.75768737302457e-07,
+      "loss": 0.1913,
+      "step": 16125
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.7582416222558641,
+      "learning_rate": 2.754737757728937e-07,
+      "loss": 0.2002,
+      "step": 16126
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.3828027369267701,
+      "learning_rate": 2.7517896760416776e-07,
+      "loss": 0.1458,
+      "step": 16127
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.2566547930164536,
+      "learning_rate": 2.7488431280584926e-07,
+      "loss": 0.1289,
+      "step": 16128
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.4691213331145125,
+      "learning_rate": 2.7458981138750283e-07,
+      "loss": 0.1803,
+      "step": 16129
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.402036415361816,
+      "learning_rate": 2.7429546335868973e-07,
+      "loss": 0.1703,
+      "step": 16130
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.7052847790540264,
+      "learning_rate": 2.7400126872896226e-07,
+      "loss": 0.2107,
+      "step": 16131
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.271406402775405,
+      "learning_rate": 2.7370722750787225e-07,
+      "loss": 0.1702,
+      "step": 16132
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.164463495786106,
+      "learning_rate": 2.7341333970496265e-07,
+      "loss": 0.1908,
+      "step": 16133
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.2135052252552057,
+      "learning_rate": 2.731196053297741e-07,
+      "loss": 0.1925,
+      "step": 16134
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.0981549524989067,
+      "learning_rate": 2.728260243918418e-07,
+      "loss": 0.1571,
+      "step": 16135
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.3545674232319869,
+      "learning_rate": 2.7253259690069424e-07,
+      "loss": 0.1674,
+      "step": 16136
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.4248472106737624,
+      "learning_rate": 2.7223932286585766e-07,
+      "loss": 0.1882,
+      "step": 16137
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.472284528125633,
+      "learning_rate": 2.719462022968505e-07,
+      "loss": 0.1523,
+      "step": 16138
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.2194695366072732,
+      "learning_rate": 2.716532352031892e-07,
+      "loss": 0.138,
+      "step": 16139
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.7365462656379504,
+      "learning_rate": 2.7136042159438325e-07,
+      "loss": 0.1876,
+      "step": 16140
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.5249235681234876,
+      "learning_rate": 2.7106776147993674e-07,
+      "loss": 0.1836,
+      "step": 16141
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.3895004472457286,
+      "learning_rate": 2.707752548693493e-07,
+      "loss": 0.162,
+      "step": 16142
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.3468689982552688,
+      "learning_rate": 2.7048290177211724e-07,
+      "loss": 0.1895,
+      "step": 16143
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.2877375874233092,
+      "learning_rate": 2.701907021977296e-07,
+      "loss": 0.1506,
+      "step": 16144
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.328065029361077,
+      "learning_rate": 2.698986561556727e-07,
+      "loss": 0.2141,
+      "step": 16145
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.398198573517573,
+      "learning_rate": 2.6960676365542446e-07,
+      "loss": 0.165,
+      "step": 16146
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.8001612259927071,
+      "learning_rate": 2.6931502470646077e-07,
+      "loss": 0.2044,
+      "step": 16147
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.9099232207516202,
+      "learning_rate": 2.6902343931825224e-07,
+      "loss": 0.1679,
+      "step": 16148
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.5711852348168973,
+      "learning_rate": 2.687320075002631e-07,
+      "loss": 0.1867,
+      "step": 16149
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.2268830354352596,
+      "learning_rate": 2.68440729261954e-07,
+      "loss": 0.178,
+      "step": 16150
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.4150305959002014,
+      "learning_rate": 2.6814960461277907e-07,
+      "loss": 0.1723,
+      "step": 16151
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.300509862684187,
+      "learning_rate": 2.678586335621891e-07,
+      "loss": 0.1652,
+      "step": 16152
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.5860378158487844,
+      "learning_rate": 2.6756781611962876e-07,
+      "loss": 0.1973,
+      "step": 16153
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.4271721574423477,
+      "learning_rate": 2.6727715229453944e-07,
+      "loss": 0.1748,
+      "step": 16154
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.60472134087379,
+      "learning_rate": 2.6698664209635464e-07,
+      "loss": 0.1968,
+      "step": 16155
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.4980856829494305,
+      "learning_rate": 2.666962855345051e-07,
+      "loss": 0.1844,
+      "step": 16156
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.5384728614110765,
+      "learning_rate": 2.664060826184156e-07,
+      "loss": 0.1851,
+      "step": 16157
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.963210995468715,
+      "learning_rate": 2.66116033357508e-07,
+      "loss": 0.2022,
+      "step": 16158
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.4074313049889848,
+      "learning_rate": 2.658261377611948e-07,
+      "loss": 0.1878,
+      "step": 16159
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.4071968334904195,
+      "learning_rate": 2.655363958388879e-07,
+      "loss": 0.2202,
+      "step": 16160
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.667972783568461,
+      "learning_rate": 2.65246807599992e-07,
+      "loss": 0.1397,
+      "step": 16161
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.240331925287851,
+      "learning_rate": 2.6495737305390736e-07,
+      "loss": 0.1434,
+      "step": 16162
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.4104372309937507,
+      "learning_rate": 2.646680922100298e-07,
+      "loss": 0.2104,
+      "step": 16163
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.1466434959713043,
+      "learning_rate": 2.643789650777484e-07,
+      "loss": 0.1565,
+      "step": 16164
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.6268329461488098,
+      "learning_rate": 2.6408999166644965e-07,
+      "loss": 0.2037,
+      "step": 16165
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.654017748022802,
+      "learning_rate": 2.638011719855127e-07,
+      "loss": 0.1801,
+      "step": 16166
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.590706101656395,
+      "learning_rate": 2.635125060443128e-07,
+      "loss": 0.175,
+      "step": 16167
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 2.084942183756407,
+      "learning_rate": 2.6322399385222185e-07,
+      "loss": 0.1938,
+      "step": 16168
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.1718905257150238,
+      "learning_rate": 2.6293563541860246e-07,
+      "loss": 0.136,
+      "step": 16169
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.1267521611265536,
+      "learning_rate": 2.626474307528171e-07,
+      "loss": 0.1257,
+      "step": 16170
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.4465218304537117,
+      "learning_rate": 2.623593798642199e-07,
+      "loss": 0.181,
+      "step": 16171
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.9214096979223683,
+      "learning_rate": 2.620714827621623e-07,
+      "loss": 0.1853,
+      "step": 16172
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.2807033669504033,
+      "learning_rate": 2.6178373945598803e-07,
+      "loss": 0.1669,
+      "step": 16173
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.5267203832095824,
+      "learning_rate": 2.6149614995503834e-07,
+      "loss": 0.1814,
+      "step": 16174
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.3229377199234875,
+      "learning_rate": 2.6120871426864866e-07,
+      "loss": 0.1244,
+      "step": 16175
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.7413749938459906,
+      "learning_rate": 2.6092143240614985e-07,
+      "loss": 0.1773,
+      "step": 16176
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.7259992742387742,
+      "learning_rate": 2.606343043768655e-07,
+      "loss": 0.1962,
+      "step": 16177
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.737713444587201,
+      "learning_rate": 2.6034733019011715e-07,
+      "loss": 0.2462,
+      "step": 16178
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.3949976336209589,
+      "learning_rate": 2.6006050985522e-07,
+      "loss": 0.197,
+      "step": 16179
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.4032542532369263,
+      "learning_rate": 2.597738433814845e-07,
+      "loss": 0.1582,
+      "step": 16180
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.4289911742338264,
+      "learning_rate": 2.5948733077821587e-07,
+      "loss": 0.1934,
+      "step": 16181
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.6979782948006727,
+      "learning_rate": 2.592009720547145e-07,
+      "loss": 0.1685,
+      "step": 16182
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.4667312465162596,
+      "learning_rate": 2.5891476722027633e-07,
+      "loss": 0.1801,
+      "step": 16183
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.4125679453399713,
+      "learning_rate": 2.5862871628418996e-07,
+      "loss": 0.1799,
+      "step": 16184
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.4270068087792978,
+      "learning_rate": 2.58342819255743e-07,
+      "loss": 0.1663,
+      "step": 16185
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.3155295964894926,
+      "learning_rate": 2.5805707614421414e-07,
+      "loss": 0.1862,
+      "step": 16186
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.1206244074399183,
+      "learning_rate": 2.577714869588793e-07,
+      "loss": 0.1257,
+      "step": 16187
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.6804752653874389,
+      "learning_rate": 2.5748605170900886e-07,
+      "loss": 0.1569,
+      "step": 16188
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.3341358093274882,
+      "learning_rate": 2.572007704038687e-07,
+      "loss": 0.1979,
+      "step": 16189
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.9553815775761074,
+      "learning_rate": 2.569156430527192e-07,
+      "loss": 0.2152,
+      "step": 16190
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.0830510391400752,
+      "learning_rate": 2.5663066966481465e-07,
+      "loss": 0.1393,
+      "step": 16191
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.4467170025553018,
+      "learning_rate": 2.563458502494065e-07,
+      "loss": 0.1876,
+      "step": 16192
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.3674314771671536,
+      "learning_rate": 2.560611848157396e-07,
+      "loss": 0.1906,
+      "step": 16193
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.5484681980004047,
+      "learning_rate": 2.5577667337305547e-07,
+      "loss": 0.2008,
+      "step": 16194
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.5534642261637932,
+      "learning_rate": 2.5549231593058777e-07,
+      "loss": 0.1724,
+      "step": 16195
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.7617890783765593,
+      "learning_rate": 2.5520811249756806e-07,
+      "loss": 0.1822,
+      "step": 16196
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.1467889644355946,
+      "learning_rate": 2.549240630832212e-07,
+      "loss": 0.1951,
+      "step": 16197
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.192998703616158,
+      "learning_rate": 2.546401676967675e-07,
+      "loss": 0.1495,
+      "step": 16198
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.4477009909932157,
+      "learning_rate": 2.543564263474241e-07,
+      "loss": 0.2145,
+      "step": 16199
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.7895037760682726,
+      "learning_rate": 2.5407283904439916e-07,
+      "loss": 0.1441,
+      "step": 16200
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.4142521841640239,
+      "learning_rate": 2.5378940579689925e-07,
+      "loss": 0.1604,
+      "step": 16201
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.2168534450943695,
+      "learning_rate": 2.535061266141242e-07,
+      "loss": 0.1468,
+      "step": 16202
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.196877399496955,
+      "learning_rate": 2.532230015052706e-07,
+      "loss": 0.162,
+      "step": 16203
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.0923753785552741,
+      "learning_rate": 2.529400304795271e-07,
+      "loss": 0.1527,
+      "step": 16204
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.2611571677214743,
+      "learning_rate": 2.5265721354607975e-07,
+      "loss": 0.1926,
+      "step": 16205
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.4035368123885257,
+      "learning_rate": 2.5237455071410957e-07,
+      "loss": 0.2247,
+      "step": 16206
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 2.228646651485751,
+      "learning_rate": 2.520920419927908e-07,
+      "loss": 0.1646,
+      "step": 16207
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.268590878153197,
+      "learning_rate": 2.518096873912962e-07,
+      "loss": 0.1654,
+      "step": 16208
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.1213311133500699,
+      "learning_rate": 2.5152748691878835e-07,
+      "loss": 0.153,
+      "step": 16209
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.7067115711318563,
+      "learning_rate": 2.512454405844289e-07,
+      "loss": 0.1674,
+      "step": 16210
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.0002757526716077,
+      "learning_rate": 2.5096354839737325e-07,
+      "loss": 0.1343,
+      "step": 16211
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.535615667851201,
+      "learning_rate": 2.5068181036677243e-07,
+      "loss": 0.16,
+      "step": 16212
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.6266495053843286,
+      "learning_rate": 2.504002265017702e-07,
+      "loss": 0.2078,
+      "step": 16213
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.3411544583484623,
+      "learning_rate": 2.5011879681150764e-07,
+      "loss": 0.1794,
+      "step": 16214
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.7802534954985982,
+      "learning_rate": 2.498375213051207e-07,
+      "loss": 0.1943,
+      "step": 16215
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.8535320413905598,
+      "learning_rate": 2.495563999917394e-07,
+      "loss": 0.1744,
+      "step": 16216
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.1119621157920918,
+      "learning_rate": 2.492754328804897e-07,
+      "loss": 0.1848,
+      "step": 16217
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.5002402619823108,
+      "learning_rate": 2.4899461998049043e-07,
+      "loss": 0.1472,
+      "step": 16218
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.4343091130966963,
+      "learning_rate": 2.487139613008588e-07,
+      "loss": 0.165,
+      "step": 16219
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.7818107977024744,
+      "learning_rate": 2.484334568507035e-07,
+      "loss": 0.1812,
+      "step": 16220
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.7227985254126081,
+      "learning_rate": 2.481531066391307e-07,
+      "loss": 0.2048,
+      "step": 16221
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.2714774428968247,
+      "learning_rate": 2.4787291067524035e-07,
+      "loss": 0.1525,
+      "step": 16222
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.3379499491883426,
+      "learning_rate": 2.4759286896812783e-07,
+      "loss": 0.1835,
+      "step": 16223
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.345116042924164,
+      "learning_rate": 2.473129815268838e-07,
+      "loss": 0.1474,
+      "step": 16224
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 2.570278086330457,
+      "learning_rate": 2.470332483605931e-07,
+      "loss": 0.171,
+      "step": 16225
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.3666380840991637,
+      "learning_rate": 2.467536694783373e-07,
+      "loss": 0.1569,
+      "step": 16226
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.500865388623583,
+      "learning_rate": 2.4647424488918935e-07,
+      "loss": 0.1795,
+      "step": 16227
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.3814085174669657,
+      "learning_rate": 2.4619497460222184e-07,
+      "loss": 0.1742,
+      "step": 16228
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.5408396118750325,
+      "learning_rate": 2.4591585862649816e-07,
+      "loss": 0.1763,
+      "step": 16229
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.7564923860733472,
+      "learning_rate": 2.4563689697108096e-07,
+      "loss": 0.1824,
+      "step": 16230
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.9953243019274751,
+      "learning_rate": 2.453580896450231e-07,
+      "loss": 0.1571,
+      "step": 16231
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.3172497409944104,
+      "learning_rate": 2.4507943665737553e-07,
+      "loss": 0.138,
+      "step": 16232
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.4151505662623474,
+      "learning_rate": 2.4480093801718395e-07,
+      "loss": 0.2056,
+      "step": 16233
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.2203928866022657,
+      "learning_rate": 2.445225937334883e-07,
+      "loss": 0.1813,
+      "step": 16234
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.4428612619959025,
+      "learning_rate": 2.4424440381532465e-07,
+      "loss": 0.1943,
+      "step": 16235
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.7795893096964894,
+      "learning_rate": 2.439663682717214e-07,
+      "loss": 0.2343,
+      "step": 16236
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.7233244449889233,
+      "learning_rate": 2.436884871117057e-07,
+      "loss": 0.2347,
+      "step": 16237
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.2370044702317688,
+      "learning_rate": 2.434107603442959e-07,
+      "loss": 0.1736,
+      "step": 16238
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.143498954424695,
+      "learning_rate": 2.431331879785087e-07,
+      "loss": 0.173,
+      "step": 16239
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.379969071096937,
+      "learning_rate": 2.428557700233536e-07,
+      "loss": 0.1318,
+      "step": 16240
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.8096102047767508,
+      "learning_rate": 2.42578506487835e-07,
+      "loss": 0.1912,
+      "step": 16241
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.738967674254353,
+      "learning_rate": 2.4230139738095405e-07,
+      "loss": 0.1825,
+      "step": 16242
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.1142106744155025,
+      "learning_rate": 2.420244427117058e-07,
+      "loss": 0.2223,
+      "step": 16243
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.528139004906445,
+      "learning_rate": 2.4174764248908077e-07,
+      "loss": 0.2082,
+      "step": 16244
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.2229880166184086,
+      "learning_rate": 2.414709967220624e-07,
+      "loss": 0.1875,
+      "step": 16245
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.4298203932903655,
+      "learning_rate": 2.411945054196324e-07,
+      "loss": 0.1609,
+      "step": 16246
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.4461273576900397,
+      "learning_rate": 2.4091816859076465e-07,
+      "loss": 0.1728,
+      "step": 16247
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.7420903123168212,
+      "learning_rate": 2.406419862444309e-07,
+      "loss": 0.2221,
+      "step": 16248
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.7248156047307504,
+      "learning_rate": 2.4036595838959453e-07,
+      "loss": 0.1666,
+      "step": 16249
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.558216049205766,
+      "learning_rate": 2.4009008503521557e-07,
+      "loss": 0.1443,
+      "step": 16250
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 2.145046960771236,
+      "learning_rate": 2.398143661902502e-07,
+      "loss": 0.2441,
+      "step": 16251
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.3149448625623266,
+      "learning_rate": 2.395388018636474e-07,
+      "loss": 0.1519,
+      "step": 16252
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.4021885554898639,
+      "learning_rate": 2.3926339206435334e-07,
+      "loss": 0.1808,
+      "step": 16253
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 3.6323412517769245,
+      "learning_rate": 2.3898813680130694e-07,
+      "loss": 0.2131,
+      "step": 16254
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.4849682877105428,
+      "learning_rate": 2.3871303608344333e-07,
+      "loss": 0.1659,
+      "step": 16255
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.6407264678095232,
+      "learning_rate": 2.384380899196925e-07,
+      "loss": 0.1954,
+      "step": 16256
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 2.129810288290729,
+      "learning_rate": 2.3816329831897967e-07,
+      "loss": 0.2379,
+      "step": 16257
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.8526849040778393,
+      "learning_rate": 2.378886612902237e-07,
+      "loss": 0.208,
+      "step": 16258
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.2269800168029152,
+      "learning_rate": 2.376141788423403e-07,
+      "loss": 0.1439,
+      "step": 16259
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.1989208311131148,
+      "learning_rate": 2.3733985098423952e-07,
+      "loss": 0.1398,
+      "step": 16260
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.0925396352950092,
+      "learning_rate": 2.3706567772482537e-07,
+      "loss": 0.1857,
+      "step": 16261
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.4262853802430833,
+      "learning_rate": 2.3679165907299905e-07,
+      "loss": 0.213,
+      "step": 16262
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.17361681872202,
+      "learning_rate": 2.3651779503765405e-07,
+      "loss": 0.16,
+      "step": 16263
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.9906932313085683,
+      "learning_rate": 2.3624408562768042e-07,
+      "loss": 0.1656,
+      "step": 16264
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.533332879960038,
+      "learning_rate": 2.3597053085196332e-07,
+      "loss": 0.1466,
+      "step": 16265
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.5964404192259622,
+      "learning_rate": 2.3569713071938282e-07,
+      "loss": 0.1944,
+      "step": 16266
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.416815044075174,
+      "learning_rate": 2.3542388523881243e-07,
+      "loss": 0.2032,
+      "step": 16267
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.6356377411704313,
+      "learning_rate": 2.3515079441912281e-07,
+      "loss": 0.218,
+      "step": 16268
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 2.1110615239956383,
+      "learning_rate": 2.3487785826917797e-07,
+      "loss": 0.2119,
+      "step": 16269
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.6202079306052566,
+      "learning_rate": 2.346050767978386e-07,
+      "loss": 0.1672,
+      "step": 16270
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.287787639328897,
+      "learning_rate": 2.3433245001395876e-07,
+      "loss": 0.2032,
+      "step": 16271
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.6788242295201496,
+      "learning_rate": 2.3405997792638746e-07,
+      "loss": 0.1817,
+      "step": 16272
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.719529789980763,
+      "learning_rate": 2.3378766054397095e-07,
+      "loss": 0.1696,
+      "step": 16273
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.5927214763124862,
+      "learning_rate": 2.335154978755466e-07,
+      "loss": 0.1676,
+      "step": 16274
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.6580183872079353,
+      "learning_rate": 2.3324348992995127e-07,
+      "loss": 0.205,
+      "step": 16275
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.205590836133362,
+      "learning_rate": 2.329716367160123e-07,
+      "loss": 0.1494,
+      "step": 16276
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.3691389979027673,
+      "learning_rate": 2.3269993824255543e-07,
+      "loss": 0.1483,
+      "step": 16277
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.8564426235478495,
+      "learning_rate": 2.3242839451840028e-07,
+      "loss": 0.1759,
+      "step": 16278
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.5154648922384308,
+      "learning_rate": 2.3215700555236032e-07,
+      "loss": 0.1576,
+      "step": 16279
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.180053243718874,
+      "learning_rate": 2.3188577135324686e-07,
+      "loss": 0.1808,
+      "step": 16280
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.6110261206388825,
+      "learning_rate": 2.316146919298623e-07,
+      "loss": 0.2066,
+      "step": 16281
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.2937058309948806,
+      "learning_rate": 2.3134376729100684e-07,
+      "loss": 0.2105,
+      "step": 16282
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.1098894148912064,
+      "learning_rate": 2.3107299744547452e-07,
+      "loss": 0.1314,
+      "step": 16283
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.3066912813730858,
+      "learning_rate": 2.3080238240205611e-07,
+      "loss": 0.1991,
+      "step": 16284
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.7204009103534696,
+      "learning_rate": 2.30531922169534e-07,
+      "loss": 0.1735,
+      "step": 16285
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.5914182715881862,
+      "learning_rate": 2.3026161675668845e-07,
+      "loss": 0.1794,
+      "step": 16286
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.8607145380869046,
+      "learning_rate": 2.299914661722935e-07,
+      "loss": 0.2503,
+      "step": 16287
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.6683567721436565,
+      "learning_rate": 2.2972147042511883e-07,
+      "loss": 0.1601,
+      "step": 16288
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.4076445817035061,
+      "learning_rate": 2.294516295239285e-07,
+      "loss": 0.1893,
+      "step": 16289
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.6726472443681624,
+      "learning_rate": 2.2918194347748113e-07,
+      "loss": 0.1913,
+      "step": 16290
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.4826721060627934,
+      "learning_rate": 2.2891241229453188e-07,
+      "loss": 0.2202,
+      "step": 16291
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.057250402221594,
+      "learning_rate": 2.2864303598382875e-07,
+      "loss": 0.161,
+      "step": 16292
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.719762815822268,
+      "learning_rate": 2.28373814554117e-07,
+      "loss": 0.1534,
+      "step": 16293
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.59525214724657,
+      "learning_rate": 2.2810474801413463e-07,
+      "loss": 0.1955,
+      "step": 16294
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.493631333461815,
+      "learning_rate": 2.2783583637261574e-07,
+      "loss": 0.2467,
+      "step": 16295
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.712488064063027,
+      "learning_rate": 2.2756707963829005e-07,
+      "loss": 0.183,
+      "step": 16296
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.5332077998193312,
+      "learning_rate": 2.272984778198817e-07,
+      "loss": 0.1998,
+      "step": 16297
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.229709075874171,
+      "learning_rate": 2.270300309261092e-07,
+      "loss": 0.1413,
+      "step": 16298
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.617485815904325,
+      "learning_rate": 2.267617389656862e-07,
+      "loss": 0.1447,
+      "step": 16299
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.5987987001135997,
+      "learning_rate": 2.264936019473224e-07,
+      "loss": 0.2168,
+      "step": 16300
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.1006720305299018,
+      "learning_rate": 2.262256198797208e-07,
+      "loss": 0.143,
+      "step": 16301
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.5798062112955238,
+      "learning_rate": 2.2595779277158114e-07,
+      "loss": 0.1531,
+      "step": 16302
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.7330705445404975,
+      "learning_rate": 2.2569012063159646e-07,
+      "loss": 0.1932,
+      "step": 16303
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.249218908289648,
+      "learning_rate": 2.254226034684559e-07,
+      "loss": 0.1442,
+      "step": 16304
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.9951959378544145,
+      "learning_rate": 2.251552412908431e-07,
+      "loss": 0.1271,
+      "step": 16305
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.3998653964491499,
+      "learning_rate": 2.2488803410743664e-07,
+      "loss": 0.1484,
+      "step": 16306
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.3275525149502425,
+      "learning_rate": 2.2462098192691185e-07,
+      "loss": 0.144,
+      "step": 16307
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.318497804508978,
+      "learning_rate": 2.2435408475793452e-07,
+      "loss": 0.1799,
+      "step": 16308
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.5911276551460567,
+      "learning_rate": 2.240873426091711e-07,
+      "loss": 0.2049,
+      "step": 16309
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.3789353786342051,
+      "learning_rate": 2.2382075548927796e-07,
+      "loss": 0.1685,
+      "step": 16310
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.6537550001580616,
+      "learning_rate": 2.2355432340691042e-07,
+      "loss": 0.1957,
+      "step": 16311
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.4789978410074442,
+      "learning_rate": 2.2328804637071545e-07,
+      "loss": 0.1605,
+      "step": 16312
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 2.1003623854086877,
+      "learning_rate": 2.2302192438933722e-07,
+      "loss": 0.1892,
+      "step": 16313
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.377168863047923,
+      "learning_rate": 2.2275595747141387e-07,
+      "loss": 0.1547,
+      "step": 16314
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.7166500527849518,
+      "learning_rate": 2.2249014562558013e-07,
+      "loss": 0.161,
+      "step": 16315
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.4028016939212262,
+      "learning_rate": 2.2222448886046354e-07,
+      "loss": 0.1733,
+      "step": 16316
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.5306841378880833,
+      "learning_rate": 2.2195898718468724e-07,
+      "loss": 0.1721,
+      "step": 16317
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.6939854503984086,
+      "learning_rate": 2.2169364060686927e-07,
+      "loss": 0.1833,
+      "step": 16318
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.7813930391300759,
+      "learning_rate": 2.214284491356239e-07,
+      "loss": 0.1825,
+      "step": 16319
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.342768560227293,
+      "learning_rate": 2.2116341277955922e-07,
+      "loss": 0.1887,
+      "step": 16320
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.3415048828476897,
+      "learning_rate": 2.208985315472778e-07,
+      "loss": 0.1685,
+      "step": 16321
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.721381673915303,
+      "learning_rate": 2.2063380544737834e-07,
+      "loss": 0.1798,
+      "step": 16322
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.5023368970186008,
+      "learning_rate": 2.2036923448845392e-07,
+      "loss": 0.1674,
+      "step": 16323
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.3522031621729,
+      "learning_rate": 2.2010481867909217e-07,
+      "loss": 0.167,
+      "step": 16324
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.5315863959925955,
+      "learning_rate": 2.1984055802787784e-07,
+      "loss": 0.1579,
+      "step": 16325
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.0161927616943465,
+      "learning_rate": 2.1957645254338633e-07,
+      "loss": 0.1513,
+      "step": 16326
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.6515711157076902,
+      "learning_rate": 2.1931250223419354e-07,
+      "loss": 0.1842,
+      "step": 16327
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.1692048870540817,
+      "learning_rate": 2.190487071088654e-07,
+      "loss": 0.161,
+      "step": 16328
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.4091825531463142,
+      "learning_rate": 2.187850671759656e-07,
+      "loss": 0.1958,
+      "step": 16329
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.4963641955602456,
+      "learning_rate": 2.185215824440512e-07,
+      "loss": 0.1973,
+      "step": 16330
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.4785071086764254,
+      "learning_rate": 2.1825825292167646e-07,
+      "loss": 0.191,
+      "step": 16331
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.5554799451669414,
+      "learning_rate": 2.179950786173879e-07,
+      "loss": 0.1507,
+      "step": 16332
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.1343030889695465,
+      "learning_rate": 2.177320595397292e-07,
+      "loss": 0.1431,
+      "step": 16333
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.414600917887582,
+      "learning_rate": 2.1746919569723858e-07,
+      "loss": 0.1519,
+      "step": 16334
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.4582275894405936,
+      "learning_rate": 2.172064870984475e-07,
+      "loss": 0.1721,
+      "step": 16335
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.6511903078175267,
+      "learning_rate": 2.169439337518836e-07,
+      "loss": 0.1771,
+      "step": 16336
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.3750349007858853,
+      "learning_rate": 2.1668153566607065e-07,
+      "loss": 0.196,
+      "step": 16337
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.137291806420097,
+      "learning_rate": 2.164192928495257e-07,
+      "loss": 0.159,
+      "step": 16338
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.7197896067443132,
+      "learning_rate": 2.1615720531076135e-07,
+      "loss": 0.2559,
+      "step": 16339
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.7450790785357246,
+      "learning_rate": 2.1589527305828473e-07,
+      "loss": 0.1782,
+      "step": 16340
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.7966960154414626,
+      "learning_rate": 2.1563349610059846e-07,
+      "loss": 0.1517,
+      "step": 16341
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.566405764838926,
+      "learning_rate": 2.1537187444619967e-07,
+      "loss": 0.1534,
+      "step": 16342
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.622680962511668,
+      "learning_rate": 2.1511040810358262e-07,
+      "loss": 0.2039,
+      "step": 16343
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 2.3994373759781684,
+      "learning_rate": 2.1484909708123226e-07,
+      "loss": 0.2149,
+      "step": 16344
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.2140638264289199,
+      "learning_rate": 2.1458794138763228e-07,
+      "loss": 0.139,
+      "step": 16345
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.1246305627340805,
+      "learning_rate": 2.1432694103125928e-07,
+      "loss": 0.1748,
+      "step": 16346
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.5364614863821318,
+      "learning_rate": 2.1406609602058648e-07,
+      "loss": 0.1922,
+      "step": 16347
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.0090030395222904,
+      "learning_rate": 2.1380540636407932e-07,
+      "loss": 0.1409,
+      "step": 16348
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.4088936645649666,
+      "learning_rate": 2.1354487207020101e-07,
+      "loss": 0.1832,
+      "step": 16349
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 2.2340814390939165,
+      "learning_rate": 2.1328449314740873e-07,
+      "loss": 0.1652,
+      "step": 16350
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.7578285301325283,
+      "learning_rate": 2.1302426960415402e-07,
+      "loss": 0.2076,
+      "step": 16351
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.4567231473822602,
+      "learning_rate": 2.127642014488851e-07,
+      "loss": 0.1758,
+      "step": 16352
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.2128533743084575,
+      "learning_rate": 2.1250428869004248e-07,
+      "loss": 0.1497,
+      "step": 16353
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 2.421538421869534,
+      "learning_rate": 2.1224453133606381e-07,
+      "loss": 0.1784,
+      "step": 16354
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.4263674066766192,
+      "learning_rate": 2.1198492939538017e-07,
+      "loss": 0.1515,
+      "step": 16355
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.4680407516094374,
+      "learning_rate": 2.1172548287642036e-07,
+      "loss": 0.1971,
+      "step": 16356
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.3618482857491658,
+      "learning_rate": 2.1146619178760374e-07,
+      "loss": 0.1897,
+      "step": 16357
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.2395571323906167,
+      "learning_rate": 2.1120705613734804e-07,
+      "loss": 0.1394,
+      "step": 16358
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.3365913486466723,
+      "learning_rate": 2.1094807593406542e-07,
+      "loss": 0.1335,
+      "step": 16359
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 2.0691137118008864,
+      "learning_rate": 2.106892511861619e-07,
+      "loss": 0.2222,
+      "step": 16360
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 2.192766743030762,
+      "learning_rate": 2.1043058190204025e-07,
+      "loss": 0.1989,
+      "step": 16361
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.5646725737619938,
+      "learning_rate": 2.101720680900954e-07,
+      "loss": 0.1665,
+      "step": 16362
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.809349248304255,
+      "learning_rate": 2.0991370975872006e-07,
+      "loss": 0.1712,
+      "step": 16363
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.6362992069897324,
+      "learning_rate": 2.0965550691629977e-07,
+      "loss": 0.0986,
+      "step": 16364
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.3306886333356653,
+      "learning_rate": 2.093974595712167e-07,
+      "loss": 0.1662,
+      "step": 16365
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.652599057974034,
+      "learning_rate": 2.0913956773184642e-07,
+      "loss": 0.1845,
+      "step": 16366
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.4561440525480207,
+      "learning_rate": 2.0888183140656104e-07,
+      "loss": 0.2113,
+      "step": 16367
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.4024881910143132,
+      "learning_rate": 2.0862425060372616e-07,
+      "loss": 0.1673,
+      "step": 16368
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.3491459587752066,
+      "learning_rate": 2.0836682533170337e-07,
+      "loss": 0.1701,
+      "step": 16369
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.4080141710552392,
+      "learning_rate": 2.0810955559884938e-07,
+      "loss": 0.2212,
+      "step": 16370
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.139629772461646,
+      "learning_rate": 2.0785244141351414e-07,
+      "loss": 0.1494,
+      "step": 16371
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.0211843483972716,
+      "learning_rate": 2.075954827840443e-07,
+      "loss": 0.1444,
+      "step": 16372
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.367669298564524,
+      "learning_rate": 2.0733867971878096e-07,
+      "loss": 0.1718,
+      "step": 16373
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.3853040382906459,
+      "learning_rate": 2.0708203222606082e-07,
+      "loss": 0.1741,
+      "step": 16374
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.7911785970939338,
+      "learning_rate": 2.068255403142133e-07,
+      "loss": 0.1987,
+      "step": 16375
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.277868693798423,
+      "learning_rate": 2.0656920399156511e-07,
+      "loss": 0.1882,
+      "step": 16376
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.4595985600777754,
+      "learning_rate": 2.0631302326643676e-07,
+      "loss": 0.2104,
+      "step": 16377
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.389268358035844,
+      "learning_rate": 2.0605699814714442e-07,
+      "loss": 0.1515,
+      "step": 16378
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.6333143795342584,
+      "learning_rate": 2.058011286419992e-07,
+      "loss": 0.1577,
+      "step": 16379
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.5718536500637406,
+      "learning_rate": 2.0554541475930556e-07,
+      "loss": 0.1959,
+      "step": 16380
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.1624894298052404,
+      "learning_rate": 2.0528985650736576e-07,
+      "loss": 0.1691,
+      "step": 16381
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.7966748665190204,
+      "learning_rate": 2.0503445389447318e-07,
+      "loss": 0.1891,
+      "step": 16382
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.372590674286997,
+      "learning_rate": 2.047792069289195e-07,
+      "loss": 0.1501,
+      "step": 16383
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.5578791000531413,
+      "learning_rate": 2.0452411561899143e-07,
+      "loss": 0.1829,
+      "step": 16384
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.3003279064955378,
+      "learning_rate": 2.0426917997296735e-07,
+      "loss": 0.1818,
+      "step": 16385
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.7144667964936897,
+      "learning_rate": 2.0401439999912288e-07,
+      "loss": 0.2266,
+      "step": 16386
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.317870534053654,
+      "learning_rate": 2.037597757057297e-07,
+      "loss": 0.2091,
+      "step": 16387
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.4626721523314112,
+      "learning_rate": 2.0350530710105232e-07,
+      "loss": 0.2033,
+      "step": 16388
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.5594826078267598,
+      "learning_rate": 2.0325099419335025e-07,
+      "loss": 0.1799,
+      "step": 16389
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 2.024090812406169,
+      "learning_rate": 2.0299683699087914e-07,
+      "loss": 0.1682,
+      "step": 16390
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.8270596962302001,
+      "learning_rate": 2.0274283550188956e-07,
+      "loss": 0.1781,
+      "step": 16391
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.3568278652756594,
+      "learning_rate": 2.0248898973462606e-07,
+      "loss": 0.1425,
+      "step": 16392
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.149428070164503,
+      "learning_rate": 2.0223529969732925e-07,
+      "loss": 0.1886,
+      "step": 16393
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.4629036020434198,
+      "learning_rate": 2.0198176539823255e-07,
+      "loss": 0.1518,
+      "step": 16394
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.4294568016404008,
+      "learning_rate": 2.0172838684556717e-07,
+      "loss": 0.1833,
+      "step": 16395
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.3101901209802793,
+      "learning_rate": 2.0147516404755763e-07,
+      "loss": 0.1583,
+      "step": 16396
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.2612243841233468,
+      "learning_rate": 2.0122209701242458e-07,
+      "loss": 0.1763,
+      "step": 16397
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.8072993464311322,
+      "learning_rate": 2.009691857483803e-07,
+      "loss": 0.156,
+      "step": 16398
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.354182165925263,
+      "learning_rate": 2.0071643026363719e-07,
+      "loss": 0.1885,
+      "step": 16399
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 2.588201957545083,
+      "learning_rate": 2.0046383056639752e-07,
+      "loss": 0.1953,
+      "step": 16400
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.38559380691925,
+      "learning_rate": 2.0021138666486196e-07,
+      "loss": 0.1384,
+      "step": 16401
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.6805466517255467,
+      "learning_rate": 1.9995909856722563e-07,
+      "loss": 0.1515,
+      "step": 16402
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.2010597393758289,
+      "learning_rate": 1.997069662816764e-07,
+      "loss": 0.1655,
+      "step": 16403
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.2828875578970311,
+      "learning_rate": 1.9945498981639944e-07,
+      "loss": 0.16,
+      "step": 16404
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.3832622809527653,
+      "learning_rate": 1.9920316917957373e-07,
+      "loss": 0.1212,
+      "step": 16405
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.4989048118747448,
+      "learning_rate": 1.9895150437937493e-07,
+      "loss": 0.1863,
+      "step": 16406
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.0625757057202228,
+      "learning_rate": 1.9869999542396988e-07,
+      "loss": 0.1704,
+      "step": 16407
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.671211708260729,
+      "learning_rate": 1.9844864232152484e-07,
+      "loss": 0.2025,
+      "step": 16408
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.2558149860896461,
+      "learning_rate": 1.9819744508019712e-07,
+      "loss": 0.1829,
+      "step": 16409
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.2238655690961737,
+      "learning_rate": 1.9794640370814188e-07,
+      "loss": 0.1675,
+      "step": 16410
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.7659080329155197,
+      "learning_rate": 1.9769551821350818e-07,
+      "loss": 0.2396,
+      "step": 16411
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 7.39408651809045,
+      "learning_rate": 1.9744478860443949e-07,
+      "loss": 0.209,
+      "step": 16412
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.7433446881399783,
+      "learning_rate": 1.9719421488907375e-07,
+      "loss": 0.1406,
+      "step": 16413
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.6432598028536236,
+      "learning_rate": 1.9694379707554668e-07,
+      "loss": 0.213,
+      "step": 16414
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.400262068589707,
+      "learning_rate": 1.9669353517198563e-07,
+      "loss": 0.1843,
+      "step": 16415
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.4718625234168223,
+      "learning_rate": 1.9644342918651417e-07,
+      "loss": 0.1688,
+      "step": 16416
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.7971703991435368,
+      "learning_rate": 1.9619347912725183e-07,
+      "loss": 0.1744,
+      "step": 16417
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.7295747604776437,
+      "learning_rate": 1.9594368500231054e-07,
+      "loss": 0.1879,
+      "step": 16418
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.7752018608906746,
+      "learning_rate": 1.9569404681980043e-07,
+      "loss": 0.223,
+      "step": 16419
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.4363042974168525,
+      "learning_rate": 1.9544456458782445e-07,
+      "loss": 0.1673,
+      "step": 16420
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.4740591601628503,
+      "learning_rate": 1.9519523831448005e-07,
+      "loss": 0.195,
+      "step": 16421
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.5045824289526062,
+      "learning_rate": 1.9494606800786185e-07,
+      "loss": 0.1558,
+      "step": 16422
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.291154250927572,
+      "learning_rate": 1.9469705367605675e-07,
+      "loss": 0.1928,
+      "step": 16423
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.2783082592820674,
+      "learning_rate": 1.944481953271493e-07,
+      "loss": 0.1911,
+      "step": 16424
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.184516128722291,
+      "learning_rate": 1.9419949296921593e-07,
+      "loss": 0.1716,
+      "step": 16425
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.6025846383020033,
+      "learning_rate": 1.939509466103312e-07,
+      "loss": 0.1873,
+      "step": 16426
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.3456534167494867,
+      "learning_rate": 1.9370255625856205e-07,
+      "loss": 0.1483,
+      "step": 16427
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.7973366849134156,
+      "learning_rate": 1.9345432192197144e-07,
+      "loss": 0.1255,
+      "step": 16428
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.1281491119840716,
+      "learning_rate": 1.932062436086185e-07,
+      "loss": 0.1636,
+      "step": 16429
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.2977016353766047,
+      "learning_rate": 1.9295832132655457e-07,
+      "loss": 0.1842,
+      "step": 16430
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.6740126446892412,
+      "learning_rate": 1.9271055508382708e-07,
+      "loss": 0.1748,
+      "step": 16431
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.6701110464997402,
+      "learning_rate": 1.924629448884807e-07,
+      "loss": 0.1694,
+      "step": 16432
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.4449303405299223,
+      "learning_rate": 1.9221549074855074e-07,
+      "loss": 0.1868,
+      "step": 16433
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.8595793155003089,
+      "learning_rate": 1.9196819267207122e-07,
+      "loss": 0.2152,
+      "step": 16434
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.3749881277872271,
+      "learning_rate": 1.917210506670686e-07,
+      "loss": 0.1777,
+      "step": 16435
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.951422315377612,
+      "learning_rate": 1.9147406474156528e-07,
+      "loss": 0.2237,
+      "step": 16436
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.6011354854084652,
+      "learning_rate": 1.9122723490357875e-07,
+      "loss": 0.1462,
+      "step": 16437
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.7373613559514578,
+      "learning_rate": 1.909805611611226e-07,
+      "loss": 0.1952,
+      "step": 16438
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.4119757751326552,
+      "learning_rate": 1.9073404352220215e-07,
+      "loss": 0.1718,
+      "step": 16439
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.421909101307833,
+      "learning_rate": 1.9048768199481983e-07,
+      "loss": 0.1536,
+      "step": 16440
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.9257436056113872,
+      "learning_rate": 1.9024147658697313e-07,
+      "loss": 0.2145,
+      "step": 16441
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.429122010423998,
+      "learning_rate": 1.8999542730665456e-07,
+      "loss": 0.1578,
+      "step": 16442
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 2.648376824238126,
+      "learning_rate": 1.8974953416184938e-07,
+      "loss": 0.1802,
+      "step": 16443
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.709699661973031,
+      "learning_rate": 1.8950379716054123e-07,
+      "loss": 0.1954,
+      "step": 16444
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.4621824376213686,
+      "learning_rate": 1.892582163107054e-07,
+      "loss": 0.1619,
+      "step": 16445
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.8366935796404367,
+      "learning_rate": 1.890127916203144e-07,
+      "loss": 0.1929,
+      "step": 16446
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.6610501384849992,
+      "learning_rate": 1.887675230973357e-07,
+      "loss": 0.2012,
+      "step": 16447
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.5185471611204413,
+      "learning_rate": 1.8852241074972856e-07,
+      "loss": 0.2104,
+      "step": 16448
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.2658564979856253,
+      "learning_rate": 1.8827745458545098e-07,
+      "loss": 0.202,
+      "step": 16449
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.3403011853502504,
+      "learning_rate": 1.8803265461245502e-07,
+      "loss": 0.1391,
+      "step": 16450
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.2142996075355301,
+      "learning_rate": 1.8778801083868536e-07,
+      "loss": 0.147,
+      "step": 16451
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.351686251492946,
+      "learning_rate": 1.8754352327208513e-07,
+      "loss": 0.1852,
+      "step": 16452
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.47146942621207,
+      "learning_rate": 1.8729919192058854e-07,
+      "loss": 0.1863,
+      "step": 16453
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.3393387802146814,
+      "learning_rate": 1.8705501679212757e-07,
+      "loss": 0.1496,
+      "step": 16454
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 2.973042567010289,
+      "learning_rate": 1.8681099789462864e-07,
+      "loss": 0.2078,
+      "step": 16455
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.296089802161718,
+      "learning_rate": 1.8656713523601267e-07,
+      "loss": 0.1966,
+      "step": 16456
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.514202886161175,
+      "learning_rate": 1.8632342882419552e-07,
+      "loss": 0.1589,
+      "step": 16457
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.4659038522437207,
+      "learning_rate": 1.8607987866708698e-07,
+      "loss": 0.1686,
+      "step": 16458
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.3170936103020805,
+      "learning_rate": 1.8583648477259464e-07,
+      "loss": 0.1688,
+      "step": 16459
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.8469511828063998,
+      "learning_rate": 1.8559324714861882e-07,
+      "loss": 0.1632,
+      "step": 16460
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.4341954312089638,
+      "learning_rate": 1.8535016580305377e-07,
+      "loss": 0.2045,
+      "step": 16461
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.8804542208127317,
+      "learning_rate": 1.8510724074379095e-07,
+      "loss": 0.1902,
+      "step": 16462
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.2527997134577398,
+      "learning_rate": 1.8486447197871626e-07,
+      "loss": 0.1527,
+      "step": 16463
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.693556830169797,
+      "learning_rate": 1.8462185951570898e-07,
+      "loss": 0.1846,
+      "step": 16464
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 2.0307390964327574,
+      "learning_rate": 1.843794033626467e-07,
+      "loss": 0.1596,
+      "step": 16465
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.5812126773455335,
+      "learning_rate": 1.84137103527397e-07,
+      "loss": 0.1578,
+      "step": 16466
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.6479041696678238,
+      "learning_rate": 1.8389496001782637e-07,
+      "loss": 0.2315,
+      "step": 16467
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.4251271197438191,
+      "learning_rate": 1.8365297284179517e-07,
+      "loss": 0.1599,
+      "step": 16468
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.1950116339869383,
+      "learning_rate": 1.834111420071577e-07,
+      "loss": 0.2171,
+      "step": 16469
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.3928328838829616,
+      "learning_rate": 1.831694675217649e-07,
+      "loss": 0.1253,
+      "step": 16470
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.4277094200865286,
+      "learning_rate": 1.8292794939346047e-07,
+      "loss": 0.1865,
+      "step": 16471
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.2059010272653012,
+      "learning_rate": 1.8268658763008484e-07,
+      "loss": 0.1812,
+      "step": 16472
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.8073891321188138,
+      "learning_rate": 1.8244538223947228e-07,
+      "loss": 0.2126,
+      "step": 16473
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.1428246843409802,
+      "learning_rate": 1.822043332294543e-07,
+      "loss": 0.1726,
+      "step": 16474
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.5836921309411665,
+      "learning_rate": 1.8196344060785298e-07,
+      "loss": 0.1754,
+      "step": 16475
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.5730294877657476,
+      "learning_rate": 1.8172270438248874e-07,
+      "loss": 0.1854,
+      "step": 16476
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.5866196579374023,
+      "learning_rate": 1.8148212456117697e-07,
+      "loss": 0.2516,
+      "step": 16477
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.3349838757166268,
+      "learning_rate": 1.8124170115172645e-07,
+      "loss": 0.1897,
+      "step": 16478
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.7888627586305756,
+      "learning_rate": 1.8100143416194037e-07,
+      "loss": 0.1867,
+      "step": 16479
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.2999677835195544,
+      "learning_rate": 1.8076132359961917e-07,
+      "loss": 0.143,
+      "step": 16480
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.8776860190621463,
+      "learning_rate": 1.805213694725566e-07,
+      "loss": 0.2159,
+      "step": 16481
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.7085878882287968,
+      "learning_rate": 1.80281571788542e-07,
+      "loss": 0.1956,
+      "step": 16482
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.4823428146730893,
+      "learning_rate": 1.8004193055535967e-07,
+      "loss": 0.1452,
+      "step": 16483
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.6241075467587252,
+      "learning_rate": 1.798024457807873e-07,
+      "loss": 0.2066,
+      "step": 16484
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.4275069988500049,
+      "learning_rate": 1.7956311747259926e-07,
+      "loss": 0.1893,
+      "step": 16485
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.2290800190008493,
+      "learning_rate": 1.793239456385648e-07,
+      "loss": 0.1357,
+      "step": 16486
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.6447713137577962,
+      "learning_rate": 1.790849302864467e-07,
+      "loss": 0.2184,
+      "step": 16487
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.6888266576088338,
+      "learning_rate": 1.788460714240048e-07,
+      "loss": 0.1769,
+      "step": 16488
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.6239173611367523,
+      "learning_rate": 1.7860736905899123e-07,
+      "loss": 0.1946,
+      "step": 16489
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.5021249650493478,
+      "learning_rate": 1.7836882319915427e-07,
+      "loss": 0.1588,
+      "step": 16490
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 2.0111411070877216,
+      "learning_rate": 1.7813043385223882e-07,
+      "loss": 0.2144,
+      "step": 16491
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.6193749013272796,
+      "learning_rate": 1.7789220102598258e-07,
+      "loss": 0.248,
+      "step": 16492
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.7083134407730591,
+      "learning_rate": 1.776541247281177e-07,
+      "loss": 0.1891,
+      "step": 16493
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.8004073625789274,
+      "learning_rate": 1.7741620496637246e-07,
+      "loss": 0.2008,
+      "step": 16494
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.5737536526592912,
+      "learning_rate": 1.771784417484712e-07,
+      "loss": 0.2184,
+      "step": 16495
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.5514523809948948,
+      "learning_rate": 1.769408350821311e-07,
+      "loss": 0.1765,
+      "step": 16496
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.0799008304985296,
+      "learning_rate": 1.767033849750649e-07,
+      "loss": 0.1543,
+      "step": 16497
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.3063257861625486,
+      "learning_rate": 1.7646609143498027e-07,
+      "loss": 0.177,
+      "step": 16498
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.707409229812104,
+      "learning_rate": 1.7622895446957945e-07,
+      "loss": 0.1832,
+      "step": 16499
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.525794480086245,
+      "learning_rate": 1.7599197408656123e-07,
+      "loss": 0.1625,
+      "step": 16500
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.4448141424479395,
+      "learning_rate": 1.757551502936178e-07,
+      "loss": 0.1603,
+      "step": 16501
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.125629781901376,
+      "learning_rate": 1.7551848309843632e-07,
+      "loss": 0.1785,
+      "step": 16502
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 2.0337336395018264,
+      "learning_rate": 1.752819725086985e-07,
+      "loss": 0.1955,
+      "step": 16503
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.1631936472641726,
+      "learning_rate": 1.7504561853208313e-07,
+      "loss": 0.1569,
+      "step": 16504
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.4844685123501955,
+      "learning_rate": 1.7480942117626076e-07,
+      "loss": 0.2019,
+      "step": 16505
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.6499845919467637,
+      "learning_rate": 1.7457338044890026e-07,
+      "loss": 0.2083,
+      "step": 16506
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.2541267581411424,
+      "learning_rate": 1.7433749635766163e-07,
+      "loss": 0.1989,
+      "step": 16507
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.4391953536432418,
+      "learning_rate": 1.7410176891020314e-07,
+      "loss": 0.1834,
+      "step": 16508
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.4305004455955048,
+      "learning_rate": 1.7386619811417595e-07,
+      "loss": 0.167,
+      "step": 16509
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.6454219807049328,
+      "learning_rate": 1.7363078397722833e-07,
+      "loss": 0.1566,
+      "step": 16510
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.9215179899713699,
+      "learning_rate": 1.7339552650699975e-07,
+      "loss": 0.1273,
+      "step": 16511
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.734928712864441,
+      "learning_rate": 1.7316042571112802e-07,
+      "loss": 0.1809,
+      "step": 16512
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 2.0244839808474935,
+      "learning_rate": 1.7292548159724475e-07,
+      "loss": 0.1861,
+      "step": 16513
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.6551854473121883,
+      "learning_rate": 1.7269069417297611e-07,
+      "loss": 0.202,
+      "step": 16514
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.290108655024167,
+      "learning_rate": 1.7245606344594324e-07,
+      "loss": 0.1699,
+      "step": 16515
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.4581419620516574,
+      "learning_rate": 1.7222158942376276e-07,
+      "loss": 0.2263,
+      "step": 16516
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.7200166759749909,
+      "learning_rate": 1.719872721140453e-07,
+      "loss": 0.2027,
+      "step": 16517
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.280377346684306,
+      "learning_rate": 1.7175311152439756e-07,
+      "loss": 0.1852,
+      "step": 16518
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.6928722556533544,
+      "learning_rate": 1.7151910766242064e-07,
+      "loss": 0.2098,
+      "step": 16519
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.3146044175028522,
+      "learning_rate": 1.712852605357096e-07,
+      "loss": 0.1839,
+      "step": 16520
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.1921859941185557,
+      "learning_rate": 1.7105157015185558e-07,
+      "loss": 0.1897,
+      "step": 16521
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.664063489492776,
+      "learning_rate": 1.708180365184453e-07,
+      "loss": 0.2049,
+      "step": 16522
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.7319660124582548,
+      "learning_rate": 1.705846596430577e-07,
+      "loss": 0.1881,
+      "step": 16523
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.5821712949956288,
+      "learning_rate": 1.7035143953327005e-07,
+      "loss": 0.1761,
+      "step": 16524
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.7311598902363703,
+      "learning_rate": 1.7011837619665072e-07,
+      "loss": 0.187,
+      "step": 16525
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.398637640073361,
+      "learning_rate": 1.6988546964076703e-07,
+      "loss": 0.1711,
+      "step": 16526
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.6375715542221896,
+      "learning_rate": 1.696527198731779e-07,
+      "loss": 0.1589,
+      "step": 16527
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.4705414296863935,
+      "learning_rate": 1.6942012690144004e-07,
+      "loss": 0.1854,
+      "step": 16528
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.5291013239413174,
+      "learning_rate": 1.6918769073310194e-07,
+      "loss": 0.138,
+      "step": 16529
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.4389968832259261,
+      "learning_rate": 1.6895541137570914e-07,
+      "loss": 0.1355,
+      "step": 16530
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.1657996944189857,
+      "learning_rate": 1.6872328883680233e-07,
+      "loss": 0.2082,
+      "step": 16531
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.794324358255734,
+      "learning_rate": 1.6849132312391547e-07,
+      "loss": 0.208,
+      "step": 16532
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.5311816599251156,
+      "learning_rate": 1.6825951424457866e-07,
+      "loss": 0.1878,
+      "step": 16533
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.4480351964429712,
+      "learning_rate": 1.680278622063164e-07,
+      "loss": 0.1866,
+      "step": 16534
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.354089662000809,
+      "learning_rate": 1.6779636701664826e-07,
+      "loss": 0.1816,
+      "step": 16535
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.6025653444095838,
+      "learning_rate": 1.6756502868308877e-07,
+      "loss": 0.1836,
+      "step": 16536
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.3819507160302713,
+      "learning_rate": 1.6733384721314749e-07,
+      "loss": 0.1771,
+      "step": 16537
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.5689033595767912,
+      "learning_rate": 1.671028226143284e-07,
+      "loss": 0.1664,
+      "step": 16538
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.6039459496744406,
+      "learning_rate": 1.6687195489413055e-07,
+      "loss": 0.1924,
+      "step": 16539
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.4840327771861819,
+      "learning_rate": 1.66641244060049e-07,
+      "loss": 0.1835,
+      "step": 16540
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.349372765741093,
+      "learning_rate": 1.6641069011957168e-07,
+      "loss": 0.1959,
+      "step": 16541
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 2.0774467766941216,
+      "learning_rate": 1.6618029308018313e-07,
+      "loss": 0.2017,
+      "step": 16542
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.6963498593834865,
+      "learning_rate": 1.6595005294936128e-07,
+      "loss": 0.2188,
+      "step": 16543
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.2983371868013107,
+      "learning_rate": 1.6571996973458072e-07,
+      "loss": 0.1663,
+      "step": 16544
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.6490852972380996,
+      "learning_rate": 1.654900434433093e-07,
+      "loss": 0.2115,
+      "step": 16545
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.3338644551877483,
+      "learning_rate": 1.6526027408301227e-07,
+      "loss": 0.1655,
+      "step": 16546
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.7570039373016098,
+      "learning_rate": 1.6503066166114633e-07,
+      "loss": 0.225,
+      "step": 16547
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.5673788765831131,
+      "learning_rate": 1.6480120618516559e-07,
+      "loss": 0.1754,
+      "step": 16548
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.7375322592779991,
+      "learning_rate": 1.6457190766251741e-07,
+      "loss": 0.1626,
+      "step": 16549
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.7357354713534072,
+      "learning_rate": 1.6434276610064692e-07,
+      "loss": 0.2303,
+      "step": 16550
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.2277226519411835,
+      "learning_rate": 1.6411378150698986e-07,
+      "loss": 0.1749,
+      "step": 16551
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.2340159739101104,
+      "learning_rate": 1.6388495388898085e-07,
+      "loss": 0.1708,
+      "step": 16552
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.2871223703750045,
+      "learning_rate": 1.636562832540467e-07,
+      "loss": 0.2004,
+      "step": 16553
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.3344827226579457,
+      "learning_rate": 1.6342776960961094e-07,
+      "loss": 0.1711,
+      "step": 16554
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.556869346442997,
+      "learning_rate": 1.6319941296309149e-07,
+      "loss": 0.1869,
+      "step": 16555
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.5313076426886478,
+      "learning_rate": 1.6297121332190025e-07,
+      "loss": 0.2031,
+      "step": 16556
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.646918813061121,
+      "learning_rate": 1.627431706934446e-07,
+      "loss": 0.1826,
+      "step": 16557
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.4408069698403179,
+      "learning_rate": 1.6251528508512747e-07,
+      "loss": 0.1812,
+      "step": 16558
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.0925172227586566,
+      "learning_rate": 1.622875565043458e-07,
+      "loss": 0.1561,
+      "step": 16559
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.4896854974706435,
+      "learning_rate": 1.620599849584925e-07,
+      "loss": 0.1572,
+      "step": 16560
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.8259266040939788,
+      "learning_rate": 1.6183257045495338e-07,
+      "loss": 0.1654,
+      "step": 16561
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 2.785102421051786,
+      "learning_rate": 1.6160531300111138e-07,
+      "loss": 0.2074,
+      "step": 16562
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.394129457479797,
+      "learning_rate": 1.6137821260434282e-07,
+      "loss": 0.1944,
+      "step": 16563
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.287601903475142,
+      "learning_rate": 1.6115126927202075e-07,
+      "loss": 0.166,
+      "step": 16564
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.061224206548082,
+      "learning_rate": 1.609244830115103e-07,
+      "loss": 0.1608,
+      "step": 16565
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.4194071281388227,
+      "learning_rate": 1.60697853830174e-07,
+      "loss": 0.1799,
+      "step": 16566
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.646611951996535,
+      "learning_rate": 1.6047138173536813e-07,
+      "loss": 0.2142,
+      "step": 16567
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 2.0442160524127284,
+      "learning_rate": 1.6024506673444463e-07,
+      "loss": 0.1723,
+      "step": 16568
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 2.167231707682427,
+      "learning_rate": 1.6001890883474868e-07,
+      "loss": 0.2078,
+      "step": 16569
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.6654229759098071,
+      "learning_rate": 1.5979290804362224e-07,
+      "loss": 0.1737,
+      "step": 16570
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.5194133036126825,
+      "learning_rate": 1.5956706436840108e-07,
+      "loss": 0.1671,
+      "step": 16571
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.6948989196685784,
+      "learning_rate": 1.5934137781641712e-07,
+      "loss": 0.2438,
+      "step": 16572
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.9095818644780234,
+      "learning_rate": 1.5911584839499562e-07,
+      "loss": 0.1887,
+      "step": 16573
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.3570901698716324,
+      "learning_rate": 1.588904761114568e-07,
+      "loss": 0.1751,
+      "step": 16574
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.6002598745181162,
+      "learning_rate": 1.5866526097311707e-07,
+      "loss": 0.1624,
+      "step": 16575
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.4613788009622186,
+      "learning_rate": 1.5844020298728779e-07,
+      "loss": 0.2031,
+      "step": 16576
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.461816311755392,
+      "learning_rate": 1.5821530216127256e-07,
+      "loss": 0.1967,
+      "step": 16577
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.2404972788985766,
+      "learning_rate": 1.5799055850237332e-07,
+      "loss": 0.224,
+      "step": 16578
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.3706542449089048,
+      "learning_rate": 1.577659720178848e-07,
+      "loss": 0.1601,
+      "step": 16579
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.7032966964688971,
+      "learning_rate": 1.5754154271509724e-07,
+      "loss": 0.1992,
+      "step": 16580
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.569841798295328,
+      "learning_rate": 1.57317270601296e-07,
+      "loss": 0.1913,
+      "step": 16581
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.6969252619315927,
+      "learning_rate": 1.5709315568376125e-07,
+      "loss": 0.1654,
+      "step": 16582
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.5795033450373497,
+      "learning_rate": 1.5686919796976675e-07,
+      "loss": 0.1459,
+      "step": 16583
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.5176775886216094,
+      "learning_rate": 1.5664539746658325e-07,
+      "loss": 0.2034,
+      "step": 16584
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.74010772757997,
+      "learning_rate": 1.5642175418147499e-07,
+      "loss": 0.167,
+      "step": 16585
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.490106134658931,
+      "learning_rate": 1.561982681217028e-07,
+      "loss": 0.1264,
+      "step": 16586
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.6592666396628226,
+      "learning_rate": 1.559749392945198e-07,
+      "loss": 0.1652,
+      "step": 16587
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.2656698336725312,
+      "learning_rate": 1.5575176770717572e-07,
+      "loss": 0.199,
+      "step": 16588
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.5924608495711647,
+      "learning_rate": 1.555287533669142e-07,
+      "loss": 0.1735,
+      "step": 16589
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 2.0252505836151644,
+      "learning_rate": 1.5530589628097615e-07,
+      "loss": 0.143,
+      "step": 16590
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.7573709823521229,
+      "learning_rate": 1.5508319645659465e-07,
+      "loss": 0.1746,
+      "step": 16591
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 2.754840173022843,
+      "learning_rate": 1.548606539009978e-07,
+      "loss": 0.1965,
+      "step": 16592
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.9493982006941561,
+      "learning_rate": 1.546382686214104e-07,
+      "loss": 0.1884,
+      "step": 16593
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.625026239586768,
+      "learning_rate": 1.5441604062505167e-07,
+      "loss": 0.1608,
+      "step": 16594
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.781344285778477,
+      "learning_rate": 1.541939699191347e-07,
+      "loss": 0.1827,
+      "step": 16595
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.3503399264236984,
+      "learning_rate": 1.5397205651086766e-07,
+      "loss": 0.1479,
+      "step": 16596
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.4143134574126621,
+      "learning_rate": 1.537503004074542e-07,
+      "loss": 0.1325,
+      "step": 16597
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.51989822075172,
+      "learning_rate": 1.53528701616093e-07,
+      "loss": 0.1668,
+      "step": 16598
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.4224418306776307,
+      "learning_rate": 1.5330726014397668e-07,
+      "loss": 0.2092,
+      "step": 16599
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.4745680645341435,
+      "learning_rate": 1.5308597599829444e-07,
+      "loss": 0.202,
+      "step": 16600
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.3556575368843153,
+      "learning_rate": 1.5286484918622834e-07,
+      "loss": 0.1224,
+      "step": 16601
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.2591264833310738,
+      "learning_rate": 1.5264387971495598e-07,
+      "loss": 0.1465,
+      "step": 16602
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.7002450145838492,
+      "learning_rate": 1.5242306759165105e-07,
+      "loss": 0.1394,
+      "step": 16603
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.648265160686305,
+      "learning_rate": 1.522024128234817e-07,
+      "loss": 0.1663,
+      "step": 16604
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.3687290601565647,
+      "learning_rate": 1.5198191541760887e-07,
+      "loss": 0.1622,
+      "step": 16605
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.69556095570399,
+      "learning_rate": 1.5176157538119074e-07,
+      "loss": 0.1608,
+      "step": 16606
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.6649389555697864,
+      "learning_rate": 1.515413927213799e-07,
+      "loss": 0.1994,
+      "step": 16607
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.1474297636951412,
+      "learning_rate": 1.5132136744532343e-07,
+      "loss": 0.1618,
+      "step": 16608
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.5528153164477803,
+      "learning_rate": 1.5110149956016452e-07,
+      "loss": 0.2103,
+      "step": 16609
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.6006410180809314,
+      "learning_rate": 1.5088178907303797e-07,
+      "loss": 0.1633,
+      "step": 16610
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.9140165284533337,
+      "learning_rate": 1.5066223599107755e-07,
+      "loss": 0.2003,
+      "step": 16611
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.249264703493462,
+      "learning_rate": 1.5044284032141033e-07,
+      "loss": 0.1601,
+      "step": 16612
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.2199459810101407,
+      "learning_rate": 1.5022360207115616e-07,
+      "loss": 0.1296,
+      "step": 16613
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.8608385945388255,
+      "learning_rate": 1.5000452124743326e-07,
+      "loss": 0.2085,
+      "step": 16614
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.2521309659337254,
+      "learning_rate": 1.4978559785735201e-07,
+      "loss": 0.186,
+      "step": 16615
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.2888254583488072,
+      "learning_rate": 1.4956683190801959e-07,
+      "loss": 0.1762,
+      "step": 16616
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.274130059528887,
+      "learning_rate": 1.4934822340653686e-07,
+      "loss": 0.2301,
+      "step": 16617
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.3486368233125947,
+      "learning_rate": 1.4912977236000047e-07,
+      "loss": 0.1877,
+      "step": 16618
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.0987531774071329,
+      "learning_rate": 1.4891147877550082e-07,
+      "loss": 0.1189,
+      "step": 16619
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.3922910140201208,
+      "learning_rate": 1.4869334266012392e-07,
+      "loss": 0.1974,
+      "step": 16620
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.370043186356322,
+      "learning_rate": 1.4847536402095076e-07,
+      "loss": 0.1603,
+      "step": 16621
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.857912061091285,
+      "learning_rate": 1.482575428650579e-07,
+      "loss": 0.2501,
+      "step": 16622
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.6679985168802633,
+      "learning_rate": 1.4803987919951467e-07,
+      "loss": 0.2009,
+      "step": 16623
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.6071250706399909,
+      "learning_rate": 1.4782237303138714e-07,
+      "loss": 0.1681,
+      "step": 16624
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.1071288883491535,
+      "learning_rate": 1.4760502436773516e-07,
+      "loss": 0.1286,
+      "step": 16625
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.1721621511236007,
+      "learning_rate": 1.4738783321561478e-07,
+      "loss": 0.1174,
+      "step": 16626
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 2.5291233017915324,
+      "learning_rate": 1.471707995820759e-07,
+      "loss": 0.191,
+      "step": 16627
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.4296937275318364,
+      "learning_rate": 1.4695392347416348e-07,
+      "loss": 0.1803,
+      "step": 16628
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.4236379347253876,
+      "learning_rate": 1.4673720489891686e-07,
+      "loss": 0.1739,
+      "step": 16629
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.2440689072231945,
+      "learning_rate": 1.4652064386337205e-07,
+      "loss": 0.1394,
+      "step": 16630
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.2922301531005782,
+      "learning_rate": 1.4630424037455792e-07,
+      "loss": 0.1801,
+      "step": 16631
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.0776719336898442,
+      "learning_rate": 1.4608799443949939e-07,
+      "loss": 0.1945,
+      "step": 16632
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.6463323864637067,
+      "learning_rate": 1.458719060652153e-07,
+      "loss": 0.1526,
+      "step": 16633
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.5826755247882864,
+      "learning_rate": 1.4565597525872056e-07,
+      "loss": 0.2122,
+      "step": 16634
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.6265313608155263,
+      "learning_rate": 1.454402020270246e-07,
+      "loss": 0.1711,
+      "step": 16635
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.362626309183456,
+      "learning_rate": 1.4522458637713121e-07,
+      "loss": 0.2033,
+      "step": 16636
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.3739692986402414,
+      "learning_rate": 1.450091283160393e-07,
+      "loss": 0.1716,
+      "step": 16637
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.7599130202537294,
+      "learning_rate": 1.447938278507427e-07,
+      "loss": 0.2332,
+      "step": 16638
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.1730087930834714,
+      "learning_rate": 1.4457868498823025e-07,
+      "loss": 0.1099,
+      "step": 16639
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.4757488471331546,
+      "learning_rate": 1.443636997354858e-07,
+      "loss": 0.1819,
+      "step": 16640
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.4621290585219489,
+      "learning_rate": 1.441488720994888e-07,
+      "loss": 0.1819,
+      "step": 16641
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.7271650997481023,
+      "learning_rate": 1.4393420208721087e-07,
+      "loss": 0.1608,
+      "step": 16642
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.1655274204594523,
+      "learning_rate": 1.4371968970562143e-07,
+      "loss": 0.1705,
+      "step": 16643
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.9400956701992975,
+      "learning_rate": 1.4350533496168273e-07,
+      "loss": 0.2241,
+      "step": 16644
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.269311865410892,
+      "learning_rate": 1.4329113786235472e-07,
+      "loss": 0.1561,
+      "step": 16645
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.9698750861409146,
+      "learning_rate": 1.430770984145885e-07,
+      "loss": 0.1487,
+      "step": 16646
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.1585091834609829,
+      "learning_rate": 1.428632166253324e-07,
+      "loss": 0.1662,
+      "step": 16647
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.9349620409009658,
+      "learning_rate": 1.4264949250153039e-07,
+      "loss": 0.2061,
+      "step": 16648
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.7375815665701209,
+      "learning_rate": 1.4243592605011792e-07,
+      "loss": 0.2105,
+      "step": 16649
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.8934014559870844,
+      "learning_rate": 1.422225172780295e-07,
+      "loss": 0.1853,
+      "step": 16650
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 2.004544387029177,
+      "learning_rate": 1.4200926619219124e-07,
+      "loss": 0.2222,
+      "step": 16651
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.0183888827995737,
+      "learning_rate": 1.417961727995254e-07,
+      "loss": 0.1531,
+      "step": 16652
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.5304998973969777,
+      "learning_rate": 1.415832371069492e-07,
+      "loss": 0.196,
+      "step": 16653
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.3885353654383175,
+      "learning_rate": 1.4137045912137603e-07,
+      "loss": 0.1762,
+      "step": 16654
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.3954877938716048,
+      "learning_rate": 1.4115783884971036e-07,
+      "loss": 0.1651,
+      "step": 16655
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.4468886210004361,
+      "learning_rate": 1.4094537629885608e-07,
+      "loss": 0.2182,
+      "step": 16656
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.5628984563126465,
+      "learning_rate": 1.407330714757088e-07,
+      "loss": 0.2345,
+      "step": 16657
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.7891035033612126,
+      "learning_rate": 1.4052092438716024e-07,
+      "loss": 0.1782,
+      "step": 16658
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.7504338084408162,
+      "learning_rate": 1.4030893504009713e-07,
+      "loss": 0.1821,
+      "step": 16659
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.6746978547747242,
+      "learning_rate": 1.4009710344140003e-07,
+      "loss": 0.2349,
+      "step": 16660
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.3038996613452034,
+      "learning_rate": 1.3988542959794627e-07,
+      "loss": 0.1735,
+      "step": 16661
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.134732533167717,
+      "learning_rate": 1.396739135166053e-07,
+      "loss": 0.1696,
+      "step": 16662
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.8182864107925933,
+      "learning_rate": 1.39462555204245e-07,
+      "loss": 0.2232,
+      "step": 16663
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.4970749707532103,
+      "learning_rate": 1.3925135466772488e-07,
+      "loss": 0.2036,
+      "step": 16664
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.812873744248991,
+      "learning_rate": 1.3904031191390056e-07,
+      "loss": 0.1696,
+      "step": 16665
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.909908481976541,
+      "learning_rate": 1.3882942694962377e-07,
+      "loss": 0.0802,
+      "step": 16666
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.4501629244594576,
+      "learning_rate": 1.3861869978173848e-07,
+      "loss": 0.1599,
+      "step": 16667
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.605760751979263,
+      "learning_rate": 1.384081304170859e-07,
+      "loss": 0.1947,
+      "step": 16668
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.9727775914103693,
+      "learning_rate": 1.3819771886250055e-07,
+      "loss": 0.25,
+      "step": 16669
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.850927606208735,
+      "learning_rate": 1.3798746512481364e-07,
+      "loss": 0.2028,
+      "step": 16670
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.6904227588848746,
+      "learning_rate": 1.3777736921084862e-07,
+      "loss": 0.1826,
+      "step": 16671
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.6276921108646192,
+      "learning_rate": 1.3756743112742722e-07,
+      "loss": 0.1804,
+      "step": 16672
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.3475092489813347,
+      "learning_rate": 1.3735765088136232e-07,
+      "loss": 0.1988,
+      "step": 16673
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.3857954201880829,
+      "learning_rate": 1.3714802847946462e-07,
+      "loss": 0.1804,
+      "step": 16674
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.451316590108012,
+      "learning_rate": 1.369385639285381e-07,
+      "loss": 0.1467,
+      "step": 16675
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.4453314805717634,
+      "learning_rate": 1.3672925723538234e-07,
+      "loss": 0.1828,
+      "step": 16676
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.366556828677022,
+      "learning_rate": 1.3652010840679186e-07,
+      "loss": 0.2222,
+      "step": 16677
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.7987050315665036,
+      "learning_rate": 1.3631111744955462e-07,
+      "loss": 0.2426,
+      "step": 16678
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.3063282386492738,
+      "learning_rate": 1.361022843704557e-07,
+      "loss": 0.1597,
+      "step": 16679
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.3951677704941143,
+      "learning_rate": 1.358936091762736e-07,
+      "loss": 0.1798,
+      "step": 16680
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.5349686728846956,
+      "learning_rate": 1.3568509187378233e-07,
+      "loss": 0.1881,
+      "step": 16681
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.3744078358071015,
+      "learning_rate": 1.354767324697498e-07,
+      "loss": 0.2001,
+      "step": 16682
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.887030613242046,
+      "learning_rate": 1.3526853097094006e-07,
+      "loss": 0.2145,
+      "step": 16683
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.5222254112752442,
+      "learning_rate": 1.3506048738411103e-07,
+      "loss": 0.1802,
+      "step": 16684
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.28679031330562,
+      "learning_rate": 1.3485260171601567e-07,
+      "loss": 0.1806,
+      "step": 16685
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.8379166750222173,
+      "learning_rate": 1.3464487397340353e-07,
+      "loss": 0.2065,
+      "step": 16686
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.9917306563207873,
+      "learning_rate": 1.3443730416301593e-07,
+      "loss": 0.218,
+      "step": 16687
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.4746187525571028,
+      "learning_rate": 1.3422989229159078e-07,
+      "loss": 0.2197,
+      "step": 16688
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.2753002578710488,
+      "learning_rate": 1.3402263836586216e-07,
+      "loss": 0.1869,
+      "step": 16689
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.2988240636317494,
+      "learning_rate": 1.338155423925569e-07,
+      "loss": 0.1819,
+      "step": 16690
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.401815078350133,
+      "learning_rate": 1.3360860437839684e-07,
+      "loss": 0.1267,
+      "step": 16691
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.315606131610894,
+      "learning_rate": 1.3340182433009995e-07,
+      "loss": 0.224,
+      "step": 16692
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.360074827768809,
+      "learning_rate": 1.3319520225437865e-07,
+      "loss": 0.1482,
+      "step": 16693
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.3380912463275376,
+      "learning_rate": 1.3298873815793922e-07,
+      "loss": 0.168,
+      "step": 16694
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.6699876337321637,
+      "learning_rate": 1.327824320474852e-07,
+      "loss": 0.2145,
+      "step": 16695
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.8013909604802407,
+      "learning_rate": 1.3257628392971123e-07,
+      "loss": 0.1752,
+      "step": 16696
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.2573125597959518,
+      "learning_rate": 1.3237029381131028e-07,
+      "loss": 0.2219,
+      "step": 16697
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.2398120686786607,
+      "learning_rate": 1.3216446169896867e-07,
+      "loss": 0.1668,
+      "step": 16698
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.616626577724158,
+      "learning_rate": 1.3195878759936831e-07,
+      "loss": 0.2184,
+      "step": 16699
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.3834610511906964,
+      "learning_rate": 1.3175327151918438e-07,
+      "loss": 0.1527,
+      "step": 16700
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.486506978927356,
+      "learning_rate": 1.3154791346508934e-07,
+      "loss": 0.1812,
+      "step": 16701
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.6315551278781248,
+      "learning_rate": 1.3134271344374783e-07,
+      "loss": 0.1737,
+      "step": 16702
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.6985459716817493,
+      "learning_rate": 1.3113767146182178e-07,
+      "loss": 0.2062,
+      "step": 16703
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.746355160937513,
+      "learning_rate": 1.3093278752596696e-07,
+      "loss": 0.1945,
+      "step": 16704
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.3124569675793452,
+      "learning_rate": 1.307280616428336e-07,
+      "loss": 0.1773,
+      "step": 16705
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.9700096354394148,
+      "learning_rate": 1.305234938190675e-07,
+      "loss": 0.1983,
+      "step": 16706
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.5924415546663409,
+      "learning_rate": 1.3031908406130833e-07,
+      "loss": 0.1912,
+      "step": 16707
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.7485921893582825,
+      "learning_rate": 1.3011483237619306e-07,
+      "loss": 0.1923,
+      "step": 16708
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.3178076939068892,
+      "learning_rate": 1.2991073877035021e-07,
+      "loss": 0.1814,
+      "step": 16709
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.5690891852860882,
+      "learning_rate": 1.2970680325040508e-07,
+      "loss": 0.1805,
+      "step": 16710
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.6539788328883982,
+      "learning_rate": 1.2950302582297735e-07,
+      "loss": 0.1883,
+      "step": 16711
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.6975100816418878,
+      "learning_rate": 1.2929940649468287e-07,
+      "loss": 0.1829,
+      "step": 16712
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.3009375343716771,
+      "learning_rate": 1.2909594527213077e-07,
+      "loss": 0.1189,
+      "step": 16713
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.1869109411859267,
+      "learning_rate": 1.2889264216192465e-07,
+      "loss": 0.1974,
+      "step": 16714
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.576476945625629,
+      "learning_rate": 1.2868949717066425e-07,
+      "loss": 0.1456,
+      "step": 16715
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.7477710143918408,
+      "learning_rate": 1.284865103049443e-07,
+      "loss": 0.2108,
+      "step": 16716
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.137120166178285,
+      "learning_rate": 1.2828368157135394e-07,
+      "loss": 0.1792,
+      "step": 16717
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.7105087718565863,
+      "learning_rate": 1.2808101097647573e-07,
+      "loss": 0.1988,
+      "step": 16718
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.4216107762244932,
+      "learning_rate": 1.278784985268905e-07,
+      "loss": 0.1462,
+      "step": 16719
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.2410374421869654,
+      "learning_rate": 1.276761442291702e-07,
+      "loss": 0.1377,
+      "step": 16720
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.4264671400355562,
+      "learning_rate": 1.2747394808988456e-07,
+      "loss": 0.1764,
+      "step": 16721
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.3153252960220696,
+      "learning_rate": 1.2727191011559614e-07,
+      "loss": 0.1427,
+      "step": 16722
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.5146549945347498,
+      "learning_rate": 1.2707003031286357e-07,
+      "loss": 0.1535,
+      "step": 16723
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.8188962625242935,
+      "learning_rate": 1.2686830868823997e-07,
+      "loss": 0.2042,
+      "step": 16724
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.3683247196209627,
+      "learning_rate": 1.2666674524827338e-07,
+      "loss": 0.1734,
+      "step": 16725
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.4645339536504556,
+      "learning_rate": 1.264653399995064e-07,
+      "loss": 0.1973,
+      "step": 16726
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.3774358241506663,
+      "learning_rate": 1.2626409294847707e-07,
+      "loss": 0.1904,
+      "step": 16727
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.4753869034182652,
+      "learning_rate": 1.26063004101718e-07,
+      "loss": 0.2232,
+      "step": 16728
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.6605097270966116,
+      "learning_rate": 1.2586207346575617e-07,
+      "loss": 0.2218,
+      "step": 16729
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.3812930816732591,
+      "learning_rate": 1.256613010471147e-07,
+      "loss": 0.1777,
+      "step": 16730
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.466448837824346,
+      "learning_rate": 1.2546068685231006e-07,
+      "loss": 0.168,
+      "step": 16731
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.4589570488470838,
+      "learning_rate": 1.2526023088785479e-07,
+      "loss": 0.1635,
+      "step": 16732
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.3132012174366545,
+      "learning_rate": 1.2505993316025533e-07,
+      "loss": 0.1751,
+      "step": 16733
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.2395867045808142,
+      "learning_rate": 1.2485979367601375e-07,
+      "loss": 0.132,
+      "step": 16734
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.2639878775563829,
+      "learning_rate": 1.2465981244162706e-07,
+      "loss": 0.1341,
+      "step": 16735
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.7859546703938312,
+      "learning_rate": 1.2445998946358617e-07,
+      "loss": 0.2325,
+      "step": 16736
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.4753326562184519,
+      "learning_rate": 1.2426032474837813e-07,
+      "loss": 0.2349,
+      "step": 16737
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.432588697902992,
+      "learning_rate": 1.240608183024833e-07,
+      "loss": 0.1595,
+      "step": 16738
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.6131173205276013,
+      "learning_rate": 1.238614701323776e-07,
+      "loss": 0.2125,
+      "step": 16739
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.4335643443072332,
+      "learning_rate": 1.2366228024453364e-07,
+      "loss": 0.1776,
+      "step": 16740
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.3950286854252338,
+      "learning_rate": 1.2346324864541515e-07,
+      "loss": 0.1939,
+      "step": 16741
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.4591397428706494,
+      "learning_rate": 1.2326437534148416e-07,
+      "loss": 0.1706,
+      "step": 16742
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.824565184737349,
+      "learning_rate": 1.230656603391961e-07,
+      "loss": 0.174,
+      "step": 16743
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.6591127773936314,
+      "learning_rate": 1.228671036450013e-07,
+      "loss": 0.2308,
+      "step": 16744
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.1084105738845047,
+      "learning_rate": 1.2266870526534412e-07,
+      "loss": 0.1605,
+      "step": 16745
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.3025033794622822,
+      "learning_rate": 1.2247046520666606e-07,
+      "loss": 0.1348,
+      "step": 16746
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.4651014482798639,
+      "learning_rate": 1.2227238347540083e-07,
+      "loss": 0.1787,
+      "step": 16747
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.8531820079403927,
+      "learning_rate": 1.220744600779794e-07,
+      "loss": 0.1491,
+      "step": 16748
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 2.0943769469973565,
+      "learning_rate": 1.218766950208261e-07,
+      "loss": 0.2598,
+      "step": 16749
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.198205693939792,
+      "learning_rate": 1.2167908831036025e-07,
+      "loss": 0.1718,
+      "step": 16750
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.5969654725754783,
+      "learning_rate": 1.2148163995299668e-07,
+      "loss": 0.2016,
+      "step": 16751
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.649008624859619,
+      "learning_rate": 1.2128434995514414e-07,
+      "loss": 0.1808,
+      "step": 16752
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.3342889295687974,
+      "learning_rate": 1.2108721832320702e-07,
+      "loss": 0.1423,
+      "step": 16753
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.4982347482499816,
+      "learning_rate": 1.208902450635846e-07,
+      "loss": 0.1647,
+      "step": 16754
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.0560469907716843,
+      "learning_rate": 1.2069343018267065e-07,
+      "loss": 0.1844,
+      "step": 16755
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.1282579414993466,
+      "learning_rate": 1.2049677368685343e-07,
+      "loss": 0.1596,
+      "step": 16756
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.4060641271881307,
+      "learning_rate": 1.2030027558251666e-07,
+      "loss": 0.1667,
+      "step": 16757
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.0949928306971963,
+      "learning_rate": 1.2010393587603975e-07,
+      "loss": 0.1609,
+      "step": 16758
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.4730074005988272,
+      "learning_rate": 1.1990775457379478e-07,
+      "loss": 0.2141,
+      "step": 16759
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.300414775510686,
+      "learning_rate": 1.1971173168215e-07,
+      "loss": 0.1481,
+      "step": 16760
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.4587450201363048,
+      "learning_rate": 1.1951586720746867e-07,
+      "loss": 0.2209,
+      "step": 16761
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.4042020508395183,
+      "learning_rate": 1.1932016115610956e-07,
+      "loss": 0.1691,
+      "step": 16762
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.1807933417914256,
+      "learning_rate": 1.1912461353442428e-07,
+      "loss": 0.1401,
+      "step": 16763
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.296780605787496,
+      "learning_rate": 1.1892922434876053e-07,
+      "loss": 0.1515,
+      "step": 16764
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.2665909448324217,
+      "learning_rate": 1.18733993605461e-07,
+      "loss": 0.1605,
+      "step": 16765
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.6367066685633542,
+      "learning_rate": 1.1853892131086287e-07,
+      "loss": 0.1961,
+      "step": 16766
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 2.3604018269538436,
+      "learning_rate": 1.1834400747129882e-07,
+      "loss": 0.2082,
+      "step": 16767
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.3899349954639915,
+      "learning_rate": 1.1814925209309547e-07,
+      "loss": 0.1703,
+      "step": 16768
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.7268398113993828,
+      "learning_rate": 1.1795465518257442e-07,
+      "loss": 0.2168,
+      "step": 16769
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.4072038647358556,
+      "learning_rate": 1.1776021674605232e-07,
+      "loss": 0.1539,
+      "step": 16770
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.3183843370123085,
+      "learning_rate": 1.1756593678984185e-07,
+      "loss": 0.1255,
+      "step": 16771
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.3484295547534175,
+      "learning_rate": 1.17371815320248e-07,
+      "loss": 0.1655,
+      "step": 16772
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.3123129643279867,
+      "learning_rate": 1.1717785234357349e-07,
+      "loss": 0.1674,
+      "step": 16773
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.0171678964568733,
+      "learning_rate": 1.1698404786611273e-07,
+      "loss": 0.1428,
+      "step": 16774
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.2204935916359678,
+      "learning_rate": 1.1679040189415792e-07,
+      "loss": 0.1519,
+      "step": 16775
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.5041859057121563,
+      "learning_rate": 1.1659691443399567e-07,
+      "loss": 0.1889,
+      "step": 16776
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.1565762845598817,
+      "learning_rate": 1.1640358549190489e-07,
+      "loss": 0.1287,
+      "step": 16777
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.9320236690262989,
+      "learning_rate": 1.1621041507416164e-07,
+      "loss": 0.1638,
+      "step": 16778
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.7402066144002728,
+      "learning_rate": 1.1601740318703703e-07,
+      "loss": 0.2137,
+      "step": 16779
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.1882880695355498,
+      "learning_rate": 1.1582454983679658e-07,
+      "loss": 0.1463,
+      "step": 16780
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.1177989513545525,
+      "learning_rate": 1.156318550296992e-07,
+      "loss": 0.1691,
+      "step": 16781
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.2791182648119739,
+      "learning_rate": 1.1543931877200099e-07,
+      "loss": 0.1783,
+      "step": 16782
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.1162755518623375,
+      "learning_rate": 1.1524694106995082e-07,
+      "loss": 0.188,
+      "step": 16783
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.4051811076590934,
+      "learning_rate": 1.1505472192979372e-07,
+      "loss": 0.1682,
+      "step": 16784
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.4576782242407424,
+      "learning_rate": 1.1486266135777024e-07,
+      "loss": 0.1561,
+      "step": 16785
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.6777233438717587,
+      "learning_rate": 1.1467075936011318e-07,
+      "loss": 0.2015,
+      "step": 16786
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.535268386356147,
+      "learning_rate": 1.1447901594305311e-07,
+      "loss": 0.2293,
+      "step": 16787
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.090514868556196,
+      "learning_rate": 1.1428743111281282e-07,
+      "loss": 0.167,
+      "step": 16788
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.5832559060523128,
+      "learning_rate": 1.140960048756129e-07,
+      "loss": 0.1704,
+      "step": 16789
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.766671967460969,
+      "learning_rate": 1.139047372376656e-07,
+      "loss": 0.2003,
+      "step": 16790
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.1480806990161283,
+      "learning_rate": 1.1371362820518094e-07,
+      "loss": 0.1543,
+      "step": 16791
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.5912955214228908,
+      "learning_rate": 1.1352267778436121e-07,
+      "loss": 0.1725,
+      "step": 16792
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.4902803756393577,
+      "learning_rate": 1.1333188598140532e-07,
+      "loss": 0.1486,
+      "step": 16793
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.1119988802279863,
+      "learning_rate": 1.1314125280250666e-07,
+      "loss": 0.1303,
+      "step": 16794
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 2.142265317854316,
+      "learning_rate": 1.1295077825385248e-07,
+      "loss": 0.1931,
+      "step": 16795
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.8381489198250949,
+      "learning_rate": 1.1276046234162674e-07,
+      "loss": 0.1605,
+      "step": 16796
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.6370092603017243,
+      "learning_rate": 1.1257030507200672e-07,
+      "loss": 0.1984,
+      "step": 16797
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.7055602380710377,
+      "learning_rate": 1.1238030645116527e-07,
+      "loss": 0.1525,
+      "step": 16798
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.7826677502670694,
+      "learning_rate": 1.121904664852691e-07,
+      "loss": 0.2094,
+      "step": 16799
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.6102600256707402,
+      "learning_rate": 1.1200078518048108e-07,
+      "loss": 0.193,
+      "step": 16800
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.9168374252941125,
+      "learning_rate": 1.1181126254295904e-07,
+      "loss": 0.1591,
+      "step": 16801
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 2.196991071714162,
+      "learning_rate": 1.1162189857885364e-07,
+      "loss": 0.1871,
+      "step": 16802
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.130220674265742,
+      "learning_rate": 1.1143269329431327e-07,
+      "loss": 0.1617,
+      "step": 16803
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.8514664218617591,
+      "learning_rate": 1.1124364669547805e-07,
+      "loss": 0.2291,
+      "step": 16804
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.3565425247500325,
+      "learning_rate": 1.1105475878848526e-07,
+      "loss": 0.2158,
+      "step": 16805
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.685029826862421,
+      "learning_rate": 1.1086602957946668e-07,
+      "loss": 0.2258,
+      "step": 16806
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.6077350946298081,
+      "learning_rate": 1.1067745907454852e-07,
+      "loss": 0.1566,
+      "step": 16807
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.9588243696744863,
+      "learning_rate": 1.1048904727985089e-07,
+      "loss": 0.1532,
+      "step": 16808
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.9461269894965296,
+      "learning_rate": 1.103007942014911e-07,
+      "loss": 0.1894,
+      "step": 16809
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.0290445383996714,
+      "learning_rate": 1.1011269984557926e-07,
+      "loss": 0.1516,
+      "step": 16810
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.6864304066828244,
+      "learning_rate": 1.0992476421822052e-07,
+      "loss": 0.1779,
+      "step": 16811
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.2520163664414585,
+      "learning_rate": 1.0973698732551663e-07,
+      "loss": 0.1757,
+      "step": 16812
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.4343745784301802,
+      "learning_rate": 1.0954936917356163e-07,
+      "loss": 0.1611,
+      "step": 16813
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.648374655732431,
+      "learning_rate": 1.0936190976844674e-07,
+      "loss": 0.2012,
+      "step": 16814
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.8690161190200212,
+      "learning_rate": 1.09174609116256e-07,
+      "loss": 0.2203,
+      "step": 16815
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 2.163873974131562,
+      "learning_rate": 1.0898746722307063e-07,
+      "loss": 0.1931,
+      "step": 16816
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.7909094967281078,
+      "learning_rate": 1.0880048409496414e-07,
+      "loss": 0.2377,
+      "step": 16817
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 2.152289692776244,
+      "learning_rate": 1.0861365973800664e-07,
+      "loss": 0.2383,
+      "step": 16818
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.5211526707472467,
+      "learning_rate": 1.0842699415826219e-07,
+      "loss": 0.1797,
+      "step": 16819
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.6517219816527573,
+      "learning_rate": 1.0824048736179038e-07,
+      "loss": 0.2162,
+      "step": 16820
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.3929961461247002,
+      "learning_rate": 1.0805413935464637e-07,
+      "loss": 0.2117,
+      "step": 16821
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.442914525135194,
+      "learning_rate": 1.0786795014287699e-07,
+      "loss": 0.2034,
+      "step": 16822
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.5032172717999988,
+      "learning_rate": 1.076819197325274e-07,
+      "loss": 0.2257,
+      "step": 16823
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.3858466508886478,
+      "learning_rate": 1.074960481296361e-07,
+      "loss": 0.1566,
+      "step": 16824
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.483223111789312,
+      "learning_rate": 1.073103353402366e-07,
+      "loss": 0.1993,
+      "step": 16825
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.9804337635865189,
+      "learning_rate": 1.0712478137035687e-07,
+      "loss": 0.1643,
+      "step": 16826
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.7099477875278521,
+      "learning_rate": 1.069393862260204e-07,
+      "loss": 0.1941,
+      "step": 16827
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.5764290506462932,
+      "learning_rate": 1.0675414991324517e-07,
+      "loss": 0.2279,
+      "step": 16828
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.8050038114174627,
+      "learning_rate": 1.0656907243804415e-07,
+      "loss": 0.2078,
+      "step": 16829
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.5601831711770673,
+      "learning_rate": 1.0638415380642475e-07,
+      "loss": 0.1976,
+      "step": 16830
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.7620037973208216,
+      "learning_rate": 1.0619939402438994e-07,
+      "loss": 0.1668,
+      "step": 16831
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.5704620631717676,
+      "learning_rate": 1.0601479309793661e-07,
+      "loss": 0.1768,
+      "step": 16832
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.254034101201252,
+      "learning_rate": 1.0583035103305772e-07,
+      "loss": 0.1603,
+      "step": 16833
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.5763748397518529,
+      "learning_rate": 1.0564606783574017e-07,
+      "loss": 0.1897,
+      "step": 16834
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.323360254855049,
+      "learning_rate": 1.0546194351196526e-07,
+      "loss": 0.161,
+      "step": 16835
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.2216084350494105,
+      "learning_rate": 1.0527797806770989e-07,
+      "loss": 0.1961,
+      "step": 16836
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.3000108711540066,
+      "learning_rate": 1.0509417150894652e-07,
+      "loss": 0.172,
+      "step": 16837
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.3922033434264975,
+      "learning_rate": 1.0491052384164092e-07,
+      "loss": 0.1486,
+      "step": 16838
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.5288254948001334,
+      "learning_rate": 1.0472703507175497e-07,
+      "loss": 0.1754,
+      "step": 16839
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 2.159840800480573,
+      "learning_rate": 1.0454370520524449e-07,
+      "loss": 0.2013,
+      "step": 16840
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.7210951151866234,
+      "learning_rate": 1.043605342480597e-07,
+      "loss": 0.1792,
+      "step": 16841
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.4128382900122662,
+      "learning_rate": 1.0417752220614753e-07,
+      "loss": 0.1772,
+      "step": 16842
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.7397823581561498,
+      "learning_rate": 1.0399466908544875e-07,
+      "loss": 0.0786,
+      "step": 16843
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.5181500696175139,
+      "learning_rate": 1.0381197489189809e-07,
+      "loss": 0.2018,
+      "step": 16844
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.1060966420422114,
+      "learning_rate": 1.0362943963142636e-07,
+      "loss": 0.1755,
+      "step": 16845
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.1196199692087883,
+      "learning_rate": 1.0344706330995824e-07,
+      "loss": 0.1912,
+      "step": 16846
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.3049561515197587,
+      "learning_rate": 1.0326484593341401e-07,
+      "loss": 0.2022,
+      "step": 16847
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.4993600274849443,
+      "learning_rate": 1.030827875077095e-07,
+      "loss": 0.1526,
+      "step": 16848
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.8017214805020603,
+      "learning_rate": 1.0290088803875331e-07,
+      "loss": 0.2121,
+      "step": 16849
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.4788629786493142,
+      "learning_rate": 1.0271914753245016e-07,
+      "loss": 0.1715,
+      "step": 16850
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 2.049877323573845,
+      "learning_rate": 1.0253756599469922e-07,
+      "loss": 0.1798,
+      "step": 16851
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.8702229761227005,
+      "learning_rate": 1.0235614343139632e-07,
+      "loss": 0.2039,
+      "step": 16852
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.2909597018273546,
+      "learning_rate": 1.0217487984842845e-07,
+      "loss": 0.1651,
+      "step": 16853
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.8447648181250351,
+      "learning_rate": 1.019937752516803e-07,
+      "loss": 0.1528,
+      "step": 16854
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.7424634193236659,
+      "learning_rate": 1.0181282964703109e-07,
+      "loss": 0.1607,
+      "step": 16855
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.1886665298988595,
+      "learning_rate": 1.016320430403539e-07,
+      "loss": 0.125,
+      "step": 16856
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.7576257733668277,
+      "learning_rate": 1.0145141543751791e-07,
+      "loss": 0.1486,
+      "step": 16857
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.4222540926446097,
+      "learning_rate": 1.0127094684438509e-07,
+      "loss": 0.1788,
+      "step": 16858
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.7722327807996172,
+      "learning_rate": 1.0109063726681467e-07,
+      "loss": 0.2297,
+      "step": 16859
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.1487745063043988,
+      "learning_rate": 1.0091048671065917e-07,
+      "loss": 0.1399,
+      "step": 16860
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.326737632918093,
+      "learning_rate": 1.0073049518176669e-07,
+      "loss": 0.19,
+      "step": 16861
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.9463443241035547,
+      "learning_rate": 1.0055066268597924e-07,
+      "loss": 0.1791,
+      "step": 16862
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.4310767844592167,
+      "learning_rate": 1.0037098922913547e-07,
+      "loss": 0.1325,
+      "step": 16863
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.2125438126779162,
+      "learning_rate": 1.0019147481706626e-07,
+      "loss": 0.1057,
+      "step": 16864
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.8833292651624824,
+      "learning_rate": 1.0001211945559919e-07,
+      "loss": 0.1751,
+      "step": 16865
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.449727773598294,
+      "learning_rate": 9.983292315055681e-08,
+      "loss": 0.1431,
+      "step": 16866
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.059717430062178,
+      "learning_rate": 9.965388590775504e-08,
+      "loss": 0.1736,
+      "step": 16867
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.3711765677281746,
+      "learning_rate": 9.947500773300645e-08,
+      "loss": 0.1661,
+      "step": 16868
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.4964969380567574,
+      "learning_rate": 9.929628863211694e-08,
+      "loss": 0.1645,
+      "step": 16869
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.7894891913043076,
+      "learning_rate": 9.911772861088853e-08,
+      "loss": 0.1953,
+      "step": 16870
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.2855265875125166,
+      "learning_rate": 9.893932767511605e-08,
+      "loss": 0.2039,
+      "step": 16871
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.5513211614837943,
+      "learning_rate": 9.876108583059152e-08,
+      "loss": 0.1788,
+      "step": 16872
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.6732938708403697,
+      "learning_rate": 9.858300308310087e-08,
+      "loss": 0.1837,
+      "step": 16873
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.5686772880572895,
+      "learning_rate": 9.840507943842392e-08,
+      "loss": 0.1933,
+      "step": 16874
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.6143107881524759,
+      "learning_rate": 9.822731490233717e-08,
+      "loss": 0.1891,
+      "step": 16875
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.2475697794970804,
+      "learning_rate": 9.804970948061044e-08,
+      "loss": 0.1316,
+      "step": 16876
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.374672002577096,
+      "learning_rate": 9.78722631790091e-08,
+      "loss": 0.1836,
+      "step": 16877
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.5634854356369807,
+      "learning_rate": 9.769497600329248e-08,
+      "loss": 0.1927,
+      "step": 16878
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.1464770282724426,
+      "learning_rate": 9.751784795921648e-08,
+      "loss": 0.157,
+      "step": 16879
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.3357485308484682,
+      "learning_rate": 9.734087905252987e-08,
+      "loss": 0.1824,
+      "step": 16880
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.5199804549152705,
+      "learning_rate": 9.716406928897859e-08,
+      "loss": 0.1876,
+      "step": 16881
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.5654473068759251,
+      "learning_rate": 9.698741867429973e-08,
+      "loss": 0.2068,
+      "step": 16882
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.460362989506714,
+      "learning_rate": 9.681092721422925e-08,
+      "loss": 0.1317,
+      "step": 16883
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.4117918171891508,
+      "learning_rate": 9.663459491449589e-08,
+      "loss": 0.1655,
+      "step": 16884
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.0683704947497914,
+      "learning_rate": 9.645842178082232e-08,
+      "loss": 0.1405,
+      "step": 16885
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.5515884393404176,
+      "learning_rate": 9.628240781892839e-08,
+      "loss": 0.1878,
+      "step": 16886
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.4231584017808052,
+      "learning_rate": 9.610655303452788e-08,
+      "loss": 0.2066,
+      "step": 16887
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.3566245172820133,
+      "learning_rate": 9.593085743332787e-08,
+      "loss": 0.1919,
+      "step": 16888
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.2929310404185175,
+      "learning_rate": 9.575532102103324e-08,
+      "loss": 0.1129,
+      "step": 16889
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.2387600764183306,
+      "learning_rate": 9.557994380334001e-08,
+      "loss": 0.1765,
+      "step": 16890
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.745837728347648,
+      "learning_rate": 9.54047257859425e-08,
+      "loss": 0.2562,
+      "step": 16891
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.4361206620697293,
+      "learning_rate": 9.522966697452785e-08,
+      "loss": 0.1413,
+      "step": 16892
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.3315597158231456,
+      "learning_rate": 9.505476737477926e-08,
+      "loss": 0.1785,
+      "step": 16893
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.5484348630064713,
+      "learning_rate": 9.488002699237275e-08,
+      "loss": 0.1865,
+      "step": 16894
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.4031898581248956,
+      "learning_rate": 9.470544583298103e-08,
+      "loss": 0.1022,
+      "step": 16895
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.3580048112137952,
+      "learning_rate": 9.453102390227176e-08,
+      "loss": 0.1519,
+      "step": 16896
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.5642420451321832,
+      "learning_rate": 9.435676120590653e-08,
+      "loss": 0.1783,
+      "step": 16897
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.4499878500560919,
+      "learning_rate": 9.418265774954138e-08,
+      "loss": 0.1726,
+      "step": 16898
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.4410182720190003,
+      "learning_rate": 9.400871353882845e-08,
+      "loss": 0.183,
+      "step": 16899
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.245281247568871,
+      "learning_rate": 9.383492857941323e-08,
+      "loss": 0.1923,
+      "step": 16900
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.6116783732443074,
+      "learning_rate": 9.366130287693786e-08,
+      "loss": 0.167,
+      "step": 16901
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.903697693288407,
+      "learning_rate": 9.34878364370384e-08,
+      "loss": 0.2039,
+      "step": 16902
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.5040493438814255,
+      "learning_rate": 9.33145292653448e-08,
+      "loss": 0.2205,
+      "step": 16903
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.112984831984476,
+      "learning_rate": 9.314138136748308e-08,
+      "loss": 0.1607,
+      "step": 16904
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.6768125637315625,
+      "learning_rate": 9.296839274907377e-08,
+      "loss": 0.1897,
+      "step": 16905
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.3225180977933337,
+      "learning_rate": 9.279556341573181e-08,
+      "loss": 0.1966,
+      "step": 16906
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.275566018912348,
+      "learning_rate": 9.26228933730683e-08,
+      "loss": 0.2066,
+      "step": 16907
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.2258497250640075,
+      "learning_rate": 9.245038262668759e-08,
+      "loss": 0.1749,
+      "step": 16908
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.3538446893360543,
+      "learning_rate": 9.227803118218914e-08,
+      "loss": 0.1852,
+      "step": 16909
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.6005041393630497,
+      "learning_rate": 9.21058390451679e-08,
+      "loss": 0.1838,
+      "step": 16910
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.3605921848100357,
+      "learning_rate": 9.193380622121384e-08,
+      "loss": 0.1708,
+      "step": 16911
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.6765974675751083,
+      "learning_rate": 9.176193271591083e-08,
+      "loss": 0.1854,
+      "step": 16912
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.7436951456021625,
+      "learning_rate": 9.159021853483718e-08,
+      "loss": 0.2327,
+      "step": 16913
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.538177710323236,
+      "learning_rate": 9.141866368356788e-08,
+      "loss": 0.2194,
+      "step": 16914
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.5224011148245213,
+      "learning_rate": 9.124726816767182e-08,
+      "loss": 0.1727,
+      "step": 16915
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.3953025725674717,
+      "learning_rate": 9.107603199271176e-08,
+      "loss": 0.1865,
+      "step": 16916
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.1648480913527786,
+      "learning_rate": 9.090495516424713e-08,
+      "loss": 0.1778,
+      "step": 16917
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 2.524782229937665,
+      "learning_rate": 9.073403768783018e-08,
+      "loss": 0.1493,
+      "step": 16918
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.6487051985989627,
+      "learning_rate": 9.056327956900923e-08,
+      "loss": 0.1717,
+      "step": 16919
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.3104492640067102,
+      "learning_rate": 9.039268081332764e-08,
+      "loss": 0.1589,
+      "step": 16920
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.6353083385339495,
+      "learning_rate": 9.022224142632263e-08,
+      "loss": 0.1915,
+      "step": 16921
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.506795164435531,
+      "learning_rate": 9.005196141352646e-08,
+      "loss": 0.1528,
+      "step": 16922
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 2.226084888831805,
+      "learning_rate": 8.988184078046802e-08,
+      "loss": 0.2204,
+      "step": 16923
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.7547947686918357,
+      "learning_rate": 8.97118795326679e-08,
+      "loss": 0.1842,
+      "step": 16924
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.612445753500997,
+      "learning_rate": 8.954207767564394e-08,
+      "loss": 0.1925,
+      "step": 16925
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.6252552437227532,
+      "learning_rate": 8.937243521490779e-08,
+      "loss": 0.252,
+      "step": 16926
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.758563902579175,
+      "learning_rate": 8.920295215596675e-08,
+      "loss": 0.1858,
+      "step": 16927
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.615932378360597,
+      "learning_rate": 8.903362850432085e-08,
+      "loss": 0.1995,
+      "step": 16928
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.6952226412818114,
+      "learning_rate": 8.886446426546846e-08,
+      "loss": 0.2482,
+      "step": 16929
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.6093505538083728,
+      "learning_rate": 8.86954594448991e-08,
+      "loss": 0.1864,
+      "step": 16930
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.023473302200405,
+      "learning_rate": 8.852661404809892e-08,
+      "loss": 0.1543,
+      "step": 16931
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.2921282522878932,
+      "learning_rate": 8.835792808055022e-08,
+      "loss": 0.1824,
+      "step": 16932
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.4872802311570352,
+      "learning_rate": 8.818940154772693e-08,
+      "loss": 0.1896,
+      "step": 16933
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.612480592980517,
+      "learning_rate": 8.802103445509968e-08,
+      "loss": 0.1726,
+      "step": 16934
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.4785587904158317,
+      "learning_rate": 8.785282680813523e-08,
+      "loss": 0.1894,
+      "step": 16935
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.5154356936574078,
+      "learning_rate": 8.768477861229196e-08,
+      "loss": 0.1794,
+      "step": 16936
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.7604387628045135,
+      "learning_rate": 8.751688987302554e-08,
+      "loss": 0.1758,
+      "step": 16937
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.831081022389953,
+      "learning_rate": 8.734916059578657e-08,
+      "loss": 0.1027,
+      "step": 16938
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.3851616270344993,
+      "learning_rate": 8.718159078601795e-08,
+      "loss": 0.1887,
+      "step": 16939
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.4282223823742726,
+      "learning_rate": 8.701418044916032e-08,
+      "loss": 0.1593,
+      "step": 16940
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.6404650655629327,
+      "learning_rate": 8.684692959064767e-08,
+      "loss": 0.2079,
+      "step": 16941
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.2784781362210411,
+      "learning_rate": 8.667983821590842e-08,
+      "loss": 0.1872,
+      "step": 16942
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.5455075153459978,
+      "learning_rate": 8.651290633036768e-08,
+      "loss": 0.1187,
+      "step": 16943
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.3530332082664682,
+      "learning_rate": 8.63461339394428e-08,
+      "loss": 0.1456,
+      "step": 16944
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.3889338357593248,
+      "learning_rate": 8.617952104854832e-08,
+      "loss": 0.1904,
+      "step": 16945
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.4235413501932128,
+      "learning_rate": 8.601306766309214e-08,
+      "loss": 0.1755,
+      "step": 16946
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.2908872693497249,
+      "learning_rate": 8.584677378847827e-08,
+      "loss": 0.1784,
+      "step": 16947
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.3991853855911305,
+      "learning_rate": 8.568063943010296e-08,
+      "loss": 0.217,
+      "step": 16948
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.250029092688561,
+      "learning_rate": 8.551466459336022e-08,
+      "loss": 0.1659,
+      "step": 16949
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.2824357640016564,
+      "learning_rate": 8.534884928363796e-08,
+      "loss": 0.1726,
+      "step": 16950
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.3694653636101066,
+      "learning_rate": 8.518319350631798e-08,
+      "loss": 0.1226,
+      "step": 16951
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.520395356848112,
+      "learning_rate": 8.501769726677767e-08,
+      "loss": 0.1946,
+      "step": 16952
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.704834438882762,
+      "learning_rate": 8.485236057038937e-08,
+      "loss": 0.1726,
+      "step": 16953
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.4468800421064612,
+      "learning_rate": 8.468718342251936e-08,
+      "loss": 0.2184,
+      "step": 16954
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.1026008696755316,
+      "learning_rate": 8.452216582853001e-08,
+      "loss": 0.1451,
+      "step": 16955
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.5088155454021248,
+      "learning_rate": 8.435730779377815e-08,
+      "loss": 0.1768,
+      "step": 16956
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.7262340552583222,
+      "learning_rate": 8.41926093236145e-08,
+      "loss": 0.151,
+      "step": 16957
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.2851487481266428,
+      "learning_rate": 8.40280704233859e-08,
+      "loss": 0.1351,
+      "step": 16958
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.3550135125607286,
+      "learning_rate": 8.386369109843251e-08,
+      "loss": 0.153,
+      "step": 16959
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.2848621600062742,
+      "learning_rate": 8.369947135409062e-08,
+      "loss": 0.1559,
+      "step": 16960
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.6673528440905065,
+      "learning_rate": 8.353541119569153e-08,
+      "loss": 0.1582,
+      "step": 16961
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.2523972769356142,
+      "learning_rate": 8.337151062855985e-08,
+      "loss": 0.1601,
+      "step": 16962
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.2335600637580102,
+      "learning_rate": 8.320776965801636e-08,
+      "loss": 0.1765,
+      "step": 16963
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.3127195878488083,
+      "learning_rate": 8.304418828937566e-08,
+      "loss": 0.1504,
+      "step": 16964
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.7540733113406792,
+      "learning_rate": 8.288076652794851e-08,
+      "loss": 0.145,
+      "step": 16965
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.9567510286218568,
+      "learning_rate": 8.271750437903903e-08,
+      "loss": 0.2206,
+      "step": 16966
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.4774018883927906,
+      "learning_rate": 8.255440184794683e-08,
+      "loss": 0.1977,
+      "step": 16967
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.5783431874890381,
+      "learning_rate": 8.239145893996715e-08,
+      "loss": 0.1897,
+      "step": 16968
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.3614142129350622,
+      "learning_rate": 8.222867566038794e-08,
+      "loss": 0.1741,
+      "step": 16969
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.6639058975174932,
+      "learning_rate": 8.206605201449447e-08,
+      "loss": 0.1793,
+      "step": 16970
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.9744373200241274,
+      "learning_rate": 8.190358800756415e-08,
+      "loss": 0.1135,
+      "step": 16971
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.387849781692652,
+      "learning_rate": 8.174128364487166e-08,
+      "loss": 0.1634,
+      "step": 16972
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.3918348846255983,
+      "learning_rate": 8.157913893168501e-08,
+      "loss": 0.1609,
+      "step": 16973
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.7769151862608705,
+      "learning_rate": 8.141715387326887e-08,
+      "loss": 0.2024,
+      "step": 16974
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.5507714682013503,
+      "learning_rate": 8.125532847487905e-08,
+      "loss": 0.1792,
+      "step": 16975
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.576543860991649,
+      "learning_rate": 8.109366274177022e-08,
+      "loss": 0.19,
+      "step": 16976
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.312723657307402,
+      "learning_rate": 8.093215667918986e-08,
+      "loss": 0.146,
+      "step": 16977
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.7021483499567445,
+      "learning_rate": 8.077081029238044e-08,
+      "loss": 0.1932,
+      "step": 16978
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.7918442501360876,
+      "learning_rate": 8.060962358657942e-08,
+      "loss": 0.1468,
+      "step": 16979
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.5602625658299116,
+      "learning_rate": 8.04485965670182e-08,
+      "loss": 0.16,
+      "step": 16980
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 2.041305244011269,
+      "learning_rate": 8.02877292389248e-08,
+      "loss": 0.191,
+      "step": 16981
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.5821399133357168,
+      "learning_rate": 8.012702160752061e-08,
+      "loss": 0.1564,
+      "step": 16982
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.3762057068234248,
+      "learning_rate": 7.996647367802312e-08,
+      "loss": 0.1791,
+      "step": 16983
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.435724898737463,
+      "learning_rate": 7.980608545564206e-08,
+      "loss": 0.2021,
+      "step": 16984
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.138517536153088,
+      "learning_rate": 7.964585694558546e-08,
+      "loss": 0.1417,
+      "step": 16985
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.8980536210120218,
+      "learning_rate": 7.94857881530542e-08,
+      "loss": 0.1796,
+      "step": 16986
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.1998145831925051,
+      "learning_rate": 7.932587908324296e-08,
+      "loss": 0.1416,
+      "step": 16987
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 2.0496848090867044,
+      "learning_rate": 7.916612974134375e-08,
+      "loss": 0.2184,
+      "step": 16988
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.2156938847144738,
+      "learning_rate": 7.900654013254128e-08,
+      "loss": 0.1744,
+      "step": 16989
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.9502007260388312,
+      "learning_rate": 7.884711026201586e-08,
+      "loss": 0.2403,
+      "step": 16990
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.3451614950855224,
+      "learning_rate": 7.86878401349439e-08,
+      "loss": 0.1786,
+      "step": 16991
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.4281290868120546,
+      "learning_rate": 7.852872975649461e-08,
+      "loss": 0.1612,
+      "step": 16992
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.565657229672366,
+      "learning_rate": 7.836977913183219e-08,
+      "loss": 0.1792,
+      "step": 16993
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.6729909434600696,
+      "learning_rate": 7.821098826611751e-08,
+      "loss": 0.174,
+      "step": 16994
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.0989872475145401,
+      "learning_rate": 7.805235716450366e-08,
+      "loss": 0.126,
+      "step": 16995
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.767683348500675,
+      "learning_rate": 7.789388583214097e-08,
+      "loss": 0.1815,
+      "step": 16996
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 2.019548205778562,
+      "learning_rate": 7.773557427417366e-08,
+      "loss": 0.189,
+      "step": 16997
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.437610658242575,
+      "learning_rate": 7.75774224957393e-08,
+      "loss": 0.184,
+      "step": 16998
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.5618199776020862,
+      "learning_rate": 7.741943050197265e-08,
+      "loss": 0.2118,
+      "step": 16999
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.2645002536045633,
+      "learning_rate": 7.72615982980024e-08,
+      "loss": 0.1776,
+      "step": 17000
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.7096550938640127,
+      "learning_rate": 7.71039258889511e-08,
+      "loss": 0.1704,
+      "step": 17001
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.4000815046376172,
+      "learning_rate": 7.694641327993746e-08,
+      "loss": 0.174,
+      "step": 17002
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.1622853057788924,
+      "learning_rate": 7.678906047607404e-08,
+      "loss": 0.1691,
+      "step": 17003
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.4182924184961476,
+      "learning_rate": 7.663186748246898e-08,
+      "loss": 0.197,
+      "step": 17004
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.95449523925543,
+      "learning_rate": 7.64748343042243e-08,
+      "loss": 0.1808,
+      "step": 17005
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.5800750694669738,
+      "learning_rate": 7.631796094643818e-08,
+      "loss": 0.2307,
+      "step": 17006
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.4270864443733482,
+      "learning_rate": 7.616124741420206e-08,
+      "loss": 0.2016,
+      "step": 17007
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.855393141159696,
+      "learning_rate": 7.600469371260299e-08,
+      "loss": 0.195,
+      "step": 17008
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.60028199899603,
+      "learning_rate": 7.584829984672304e-08,
+      "loss": 0.1883,
+      "step": 17009
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.5300058351199761,
+      "learning_rate": 7.56920658216398e-08,
+      "loss": 0.1881,
+      "step": 17010
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.2067076980853289,
+      "learning_rate": 7.553599164242253e-08,
+      "loss": 0.1207,
+      "step": 17011
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.2555394276805227,
+      "learning_rate": 7.538007731413943e-08,
+      "loss": 0.1599,
+      "step": 17012
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.3933009643118217,
+      "learning_rate": 7.522432284185088e-08,
+      "loss": 0.2371,
+      "step": 17013
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.614365298921012,
+      "learning_rate": 7.506872823061229e-08,
+      "loss": 0.1427,
+      "step": 17014
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.572426672796584,
+      "learning_rate": 7.491329348547571e-08,
+      "loss": 0.1981,
+      "step": 17015
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.7255533730198058,
+      "learning_rate": 7.47580186114849e-08,
+      "loss": 0.181,
+      "step": 17016
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.1288166604577632,
+      "learning_rate": 7.46029036136814e-08,
+      "loss": 0.1387,
+      "step": 17017
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.7181137771420303,
+      "learning_rate": 7.444794849709946e-08,
+      "loss": 0.1974,
+      "step": 17018
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 2.304518800153913,
+      "learning_rate": 7.429315326677067e-08,
+      "loss": 0.2565,
+      "step": 17019
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.7884839717540297,
+      "learning_rate": 7.413851792771765e-08,
+      "loss": 0.1953,
+      "step": 17020
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.3544682484816761,
+      "learning_rate": 7.398404248496083e-08,
+      "loss": 0.1914,
+      "step": 17021
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.4613761855203742,
+      "learning_rate": 7.382972694351564e-08,
+      "loss": 0.1606,
+      "step": 17022
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.3951436949882852,
+      "learning_rate": 7.367557130838921e-08,
+      "loss": 0.2118,
+      "step": 17023
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.305170780604723,
+      "learning_rate": 7.35215755845875e-08,
+      "loss": 0.1543,
+      "step": 17024
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.5996174282295113,
+      "learning_rate": 7.336773977710764e-08,
+      "loss": 0.185,
+      "step": 17025
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.4821700081948268,
+      "learning_rate": 7.321406389094398e-08,
+      "loss": 0.1604,
+      "step": 17026
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.1143548406669888,
+      "learning_rate": 7.306054793108475e-08,
+      "loss": 0.1164,
+      "step": 17027
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.2987922921651553,
+      "learning_rate": 7.290719190251428e-08,
+      "loss": 0.1785,
+      "step": 17028
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.4397854880655432,
+      "learning_rate": 7.275399581020859e-08,
+      "loss": 0.1562,
+      "step": 17029
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.5067720727248446,
+      "learning_rate": 7.260095965914205e-08,
+      "loss": 0.2085,
+      "step": 17030
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.396227206045814,
+      "learning_rate": 7.244808345428178e-08,
+      "loss": 0.1475,
+      "step": 17031
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 2.028674840821755,
+      "learning_rate": 7.229536720058994e-08,
+      "loss": 0.1947,
+      "step": 17032
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 2.440421029236185,
+      "learning_rate": 7.214281090302477e-08,
+      "loss": 0.1824,
+      "step": 17033
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.1165320100686424,
+      "learning_rate": 7.199041456653733e-08,
+      "loss": 0.1631,
+      "step": 17034
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.3831756082731783,
+      "learning_rate": 7.183817819607474e-08,
+      "loss": 0.1716,
+      "step": 17035
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.3515671084992442,
+      "learning_rate": 7.168610179657864e-08,
+      "loss": 0.1745,
+      "step": 17036
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.859168049170463,
+      "learning_rate": 7.153418537298617e-08,
+      "loss": 0.2371,
+      "step": 17037
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.4696084922433805,
+      "learning_rate": 7.138242893022785e-08,
+      "loss": 0.2091,
+      "step": 17038
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.154295048333028,
+      "learning_rate": 7.123083247323026e-08,
+      "loss": 0.1462,
+      "step": 17039
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.3676469929888593,
+      "learning_rate": 7.107939600691394e-08,
+      "loss": 0.1906,
+      "step": 17040
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 3.2396509566205,
+      "learning_rate": 7.092811953619439e-08,
+      "loss": 0.1806,
+      "step": 17041
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.4948696277859037,
+      "learning_rate": 7.077700306598323e-08,
+      "loss": 0.1787,
+      "step": 17042
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.2639803340433946,
+      "learning_rate": 7.062604660118488e-08,
+      "loss": 0.1884,
+      "step": 17043
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.9395685382474044,
+      "learning_rate": 7.047525014669931e-08,
+      "loss": 0.1694,
+      "step": 17044
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.0853376378389257,
+      "learning_rate": 7.03246137074215e-08,
+      "loss": 0.1467,
+      "step": 17045
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.1109737810147213,
+      "learning_rate": 7.017413728824196e-08,
+      "loss": 0.1511,
+      "step": 17046
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.6550848299223349,
+      "learning_rate": 7.002382089404459e-08,
+      "loss": 0.1459,
+      "step": 17047
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.433642789329367,
+      "learning_rate": 6.987366452970934e-08,
+      "loss": 0.2056,
+      "step": 17048
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.532865358310362,
+      "learning_rate": 6.972366820010956e-08,
+      "loss": 0.1802,
+      "step": 17049
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.551478577463782,
+      "learning_rate": 6.957383191011469e-08,
+      "loss": 0.1631,
+      "step": 17050
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.2495173416518015,
+      "learning_rate": 6.94241556645886e-08,
+      "loss": 0.1477,
+      "step": 17051
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.372495673551498,
+      "learning_rate": 6.927463946838964e-08,
+      "loss": 0.2181,
+      "step": 17052
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.5247570485310054,
+      "learning_rate": 6.912528332637114e-08,
+      "loss": 0.1987,
+      "step": 17053
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.3607379832777045,
+      "learning_rate": 6.89760872433809e-08,
+      "loss": 0.164,
+      "step": 17054
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.6139576673853055,
+      "learning_rate": 6.882705122426337e-08,
+      "loss": 0.1673,
+      "step": 17055
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.4992285224711346,
+      "learning_rate": 6.867817527385467e-08,
+      "loss": 0.1934,
+      "step": 17056
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.5586435123616593,
+      "learning_rate": 6.852945939698874e-08,
+      "loss": 0.1454,
+      "step": 17057
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.5690062104692444,
+      "learning_rate": 6.838090359849225e-08,
+      "loss": 0.2537,
+      "step": 17058
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.3588436447482848,
+      "learning_rate": 6.823250788318692e-08,
+      "loss": 0.1534,
+      "step": 17059
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.0678807883047152,
+      "learning_rate": 6.80842722558911e-08,
+      "loss": 0.1526,
+      "step": 17060
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.2603327812570109,
+      "learning_rate": 6.793619672141538e-08,
+      "loss": 0.1694,
+      "step": 17061
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.2820652659086076,
+      "learning_rate": 6.778828128456704e-08,
+      "loss": 0.1554,
+      "step": 17062
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.9927178368854448,
+      "learning_rate": 6.764052595014725e-08,
+      "loss": 0.1887,
+      "step": 17063
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 2.037904322766274,
+      "learning_rate": 6.749293072295215e-08,
+      "loss": 0.2388,
+      "step": 17064
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.47239050756822,
+      "learning_rate": 6.734549560777292e-08,
+      "loss": 0.1847,
+      "step": 17065
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.6932419996930101,
+      "learning_rate": 6.719822060939573e-08,
+      "loss": 0.2066,
+      "step": 17066
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.2499830304424462,
+      "learning_rate": 6.705110573260066e-08,
+      "loss": 0.1751,
+      "step": 17067
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.5478932997795436,
+      "learning_rate": 6.690415098216274e-08,
+      "loss": 0.1753,
+      "step": 17068
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.5959874510319532,
+      "learning_rate": 6.675735636285375e-08,
+      "loss": 0.1852,
+      "step": 17069
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.299669690075466,
+      "learning_rate": 6.661072187943763e-08,
+      "loss": 0.1784,
+      "step": 17070
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.424246047737138,
+      "learning_rate": 6.646424753667391e-08,
+      "loss": 0.1895,
+      "step": 17071
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.2080631228022074,
+      "learning_rate": 6.631793333931769e-08,
+      "loss": 0.1705,
+      "step": 17072
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.6888964094075936,
+      "learning_rate": 6.617177929211904e-08,
+      "loss": 0.2131,
+      "step": 17073
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.5704278096996072,
+      "learning_rate": 6.60257853998214e-08,
+      "loss": 0.1607,
+      "step": 17074
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.6492615533837223,
+      "learning_rate": 6.587995166716377e-08,
+      "loss": 0.1888,
+      "step": 17075
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.4838356845285354,
+      "learning_rate": 6.573427809888067e-08,
+      "loss": 0.1724,
+      "step": 17076
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.4907545179403838,
+      "learning_rate": 6.558876469969944e-08,
+      "loss": 0.1464,
+      "step": 17077
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.2448634827884777,
+      "learning_rate": 6.544341147434518e-08,
+      "loss": 0.1452,
+      "step": 17078
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.125988605365387,
+      "learning_rate": 6.529821842753525e-08,
+      "loss": 0.185,
+      "step": 17079
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.5571919880311806,
+      "learning_rate": 6.515318556398254e-08,
+      "loss": 0.1653,
+      "step": 17080
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.35429772574214,
+      "learning_rate": 6.500831288839549e-08,
+      "loss": 0.1643,
+      "step": 17081
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 2.0149888358031682,
+      "learning_rate": 6.486360040547645e-08,
+      "loss": 0.1351,
+      "step": 17082
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.3618998592571026,
+      "learning_rate": 6.471904811992225e-08,
+      "loss": 0.1528,
+      "step": 17083
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.4726928332006832,
+      "learning_rate": 6.457465603642577e-08,
+      "loss": 0.1844,
+      "step": 17084
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.6966830220523992,
+      "learning_rate": 6.443042415967437e-08,
+      "loss": 0.1801,
+      "step": 17085
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.8893169181267553,
+      "learning_rate": 6.428635249434933e-08,
+      "loss": 0.1901,
+      "step": 17086
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.2953404459222215,
+      "learning_rate": 6.414244104512801e-08,
+      "loss": 0.1765,
+      "step": 17087
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.3782509173177941,
+      "learning_rate": 6.399868981668111e-08,
+      "loss": 0.1745,
+      "step": 17088
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.4697593000954463,
+      "learning_rate": 6.385509881367491e-08,
+      "loss": 0.2262,
+      "step": 17089
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.2663087410778537,
+      "learning_rate": 6.371166804077122e-08,
+      "loss": 0.1483,
+      "step": 17090
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.6146584882985187,
+      "learning_rate": 6.356839750262522e-08,
+      "loss": 0.1842,
+      "step": 17091
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.5264499712635549,
+      "learning_rate": 6.342528720388763e-08,
+      "loss": 0.1766,
+      "step": 17092
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.8385241406787811,
+      "learning_rate": 6.328233714920362e-08,
+      "loss": 0.2381,
+      "step": 17093
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.2051418046494828,
+      "learning_rate": 6.313954734321393e-08,
+      "loss": 0.1737,
+      "step": 17094
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.781557985081595,
+      "learning_rate": 6.299691779055372e-08,
+      "loss": 0.225,
+      "step": 17095
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.6553231831055168,
+      "learning_rate": 6.285444849585265e-08,
+      "loss": 0.1736,
+      "step": 17096
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.356659750744561,
+      "learning_rate": 6.271213946373478e-08,
+      "loss": 0.167,
+      "step": 17097
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.8372379810336177,
+      "learning_rate": 6.256999069881974e-08,
+      "loss": 0.2085,
+      "step": 17098
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.2122156012430452,
+      "learning_rate": 6.24280022057222e-08,
+      "loss": 0.1706,
+      "step": 17099
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.283538696097501,
+      "learning_rate": 6.22861739890518e-08,
+      "loss": 0.1456,
+      "step": 17100
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.5858675748696804,
+      "learning_rate": 6.214450605341093e-08,
+      "loss": 0.1865,
+      "step": 17101
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.2882660167989468,
+      "learning_rate": 6.200299840339874e-08,
+      "loss": 0.1718,
+      "step": 17102
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.101946212681967,
+      "learning_rate": 6.186165104360876e-08,
+      "loss": 0.1926,
+      "step": 17103
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.7244536263631016,
+      "learning_rate": 6.172046397862952e-08,
+      "loss": 0.1727,
+      "step": 17104
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.3826647857392196,
+      "learning_rate": 6.157943721304349e-08,
+      "loss": 0.2,
+      "step": 17105
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.437272012294837,
+      "learning_rate": 6.143857075142923e-08,
+      "loss": 0.1929,
+      "step": 17106
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.6227428367261532,
+      "learning_rate": 6.129786459835807e-08,
+      "loss": 0.2198,
+      "step": 17107
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.6369877505423784,
+      "learning_rate": 6.115731875839858e-08,
+      "loss": 0.1887,
+      "step": 17108
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.247446677008627,
+      "learning_rate": 6.101693323611269e-08,
+      "loss": 0.1955,
+      "step": 17109
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.46160615317107,
+      "learning_rate": 6.087670803605728e-08,
+      "loss": 0.1498,
+      "step": 17110
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.3398166464686143,
+      "learning_rate": 6.073664316278372e-08,
+      "loss": 0.1735,
+      "step": 17111
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.481925479169119,
+      "learning_rate": 6.059673862083948e-08,
+      "loss": 0.1529,
+      "step": 17112
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.6063336040766465,
+      "learning_rate": 6.045699441476538e-08,
+      "loss": 0.1775,
+      "step": 17113
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.4600015692016581,
+      "learning_rate": 6.031741054909779e-08,
+      "loss": 0.1746,
+      "step": 17114
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.4276804985599807,
+      "learning_rate": 6.017798702836697e-08,
+      "loss": 0.1735,
+      "step": 17115
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.8092715638989219,
+      "learning_rate": 6.003872385709985e-08,
+      "loss": 0.1794,
+      "step": 17116
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.3710420006444053,
+      "learning_rate": 5.989962103981618e-08,
+      "loss": 0.2353,
+      "step": 17117
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.4154007569866494,
+      "learning_rate": 5.976067858103174e-08,
+      "loss": 0.1196,
+      "step": 17118
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.1206010407005038,
+      "learning_rate": 5.962189648525629e-08,
+      "loss": 0.14,
+      "step": 17119
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.1718186396715138,
+      "learning_rate": 5.948327475699511e-08,
+      "loss": 0.1785,
+      "step": 17120
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.4096174663803454,
+      "learning_rate": 5.934481340074793e-08,
+      "loss": 0.1924,
+      "step": 17121
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.7576754792109657,
+      "learning_rate": 5.9206512421008944e-08,
+      "loss": 0.246,
+      "step": 17122
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.6406196866630618,
+      "learning_rate": 5.906837182226788e-08,
+      "loss": 0.1958,
+      "step": 17123
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.619365313025078,
+      "learning_rate": 5.893039160900893e-08,
+      "loss": 0.1543,
+      "step": 17124
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.380687416948697,
+      "learning_rate": 5.8792571785710185e-08,
+      "loss": 0.2055,
+      "step": 17125
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.6531588480608403,
+      "learning_rate": 5.8654912356846394e-08,
+      "loss": 0.219,
+      "step": 17126
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.3506279950547804,
+      "learning_rate": 5.8517413326885654e-08,
+      "loss": 0.2481,
+      "step": 17127
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.5404704039580557,
+      "learning_rate": 5.8380074700291055e-08,
+      "loss": 0.1565,
+      "step": 17128
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.997049920087568,
+      "learning_rate": 5.824289648152126e-08,
+      "loss": 0.1604,
+      "step": 17129
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.4368045254399375,
+      "learning_rate": 5.810587867502882e-08,
+      "loss": 0.1777,
+      "step": 17130
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.7311768643737562,
+      "learning_rate": 5.7969021285260716e-08,
+      "loss": 0.2083,
+      "step": 17131
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.3350836495827554,
+      "learning_rate": 5.7832324316660636e-08,
+      "loss": 0.144,
+      "step": 17132
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.5125680159416233,
+      "learning_rate": 5.7695787773664466e-08,
+      "loss": 0.2057,
+      "step": 17133
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.301725005117756,
+      "learning_rate": 5.755941166070589e-08,
+      "loss": 0.1722,
+      "step": 17134
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.6064386106774506,
+      "learning_rate": 5.7423195982210244e-08,
+      "loss": 0.1512,
+      "step": 17135
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.8167329884009948,
+      "learning_rate": 5.728714074260011e-08,
+      "loss": 0.2252,
+      "step": 17136
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.6795198135064677,
+      "learning_rate": 5.7151245946291956e-08,
+      "loss": 0.1806,
+      "step": 17137
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.5657429801056306,
+      "learning_rate": 5.701551159769614e-08,
+      "loss": 0.1864,
+      "step": 17138
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.5041569637036702,
+      "learning_rate": 5.6879937701219687e-08,
+      "loss": 0.1802,
+      "step": 17139
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.6732537654023298,
+      "learning_rate": 5.6744524261262426e-08,
+      "loss": 0.1529,
+      "step": 17140
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.6698980853968395,
+      "learning_rate": 5.660927128222138e-08,
+      "loss": 0.1777,
+      "step": 17141
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.3371338367592744,
+      "learning_rate": 5.647417876848526e-08,
+      "loss": 0.165,
+      "step": 17142
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.8354257885045568,
+      "learning_rate": 5.6339246724440577e-08,
+      "loss": 0.1979,
+      "step": 17143
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.518343679603614,
+      "learning_rate": 5.6204475154466034e-08,
+      "loss": 0.1486,
+      "step": 17144
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.4069026121465864,
+      "learning_rate": 5.6069864062937574e-08,
+      "loss": 0.229,
+      "step": 17145
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.8447075232456764,
+      "learning_rate": 5.593541345422448e-08,
+      "loss": 0.184,
+      "step": 17146
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.4426970814244313,
+      "learning_rate": 5.58011233326905e-08,
+      "loss": 0.1646,
+      "step": 17147
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.0489866964425751,
+      "learning_rate": 5.566699370269602e-08,
+      "loss": 0.1401,
+      "step": 17148
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.41166851342753,
+      "learning_rate": 5.5533024568593106e-08,
+      "loss": 0.1623,
+      "step": 17149
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.7253463051967683,
+      "learning_rate": 5.539921593473219e-08,
+      "loss": 0.1876,
+      "step": 17150
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.5451803364951238,
+      "learning_rate": 5.5265567805455333e-08,
+      "loss": 0.205,
+      "step": 17151
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.5579409080544917,
+      "learning_rate": 5.513208018510186e-08,
+      "loss": 0.175,
+      "step": 17152
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.9906430110386066,
+      "learning_rate": 5.499875307800495e-08,
+      "loss": 0.1886,
+      "step": 17153
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.2732120238240863,
+      "learning_rate": 5.486558648849172e-08,
+      "loss": 0.1351,
+      "step": 17154
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.362267234588213,
+      "learning_rate": 5.473258042088536e-08,
+      "loss": 0.1481,
+      "step": 17155
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.3818103370047408,
+      "learning_rate": 5.459973487950298e-08,
+      "loss": 0.1884,
+      "step": 17156
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.3235768363051197,
+      "learning_rate": 5.446704986865725e-08,
+      "loss": 0.1474,
+      "step": 17157
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.6262910876629038,
+      "learning_rate": 5.4334525392654716e-08,
+      "loss": 0.1673,
+      "step": 17158
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.7482015871599599,
+      "learning_rate": 5.4202161455798044e-08,
+      "loss": 0.138,
+      "step": 17159
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.6199106610206437,
+      "learning_rate": 5.4069958062382686e-08,
+      "loss": 0.1312,
+      "step": 17160
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.5978248479251491,
+      "learning_rate": 5.393791521670022e-08,
+      "loss": 0.1655,
+      "step": 17161
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.3378302790265266,
+      "learning_rate": 5.3806032923037765e-08,
+      "loss": 0.1825,
+      "step": 17162
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.7539016281780697,
+      "learning_rate": 5.367431118567579e-08,
+      "loss": 0.1652,
+      "step": 17163
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.9522381236556826,
+      "learning_rate": 5.3542750008889755e-08,
+      "loss": 0.1725,
+      "step": 17164
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.2382447225315538,
+      "learning_rate": 5.34113493969507e-08,
+      "loss": 0.1805,
+      "step": 17165
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.3392109725169083,
+      "learning_rate": 5.328010935412409e-08,
+      "loss": 0.1805,
+      "step": 17166
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.587412444275867,
+      "learning_rate": 5.31490298846693e-08,
+      "loss": 0.2237,
+      "step": 17167
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 2.088478084240085,
+      "learning_rate": 5.301811099284182e-08,
+      "loss": 0.1899,
+      "step": 17168
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.3086164273247136,
+      "learning_rate": 5.288735268289103e-08,
+      "loss": 0.157,
+      "step": 17169
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.3734885166140878,
+      "learning_rate": 5.2756754959061295e-08,
+      "loss": 0.1931,
+      "step": 17170
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.0943679256145196,
+      "learning_rate": 5.2626317825592025e-08,
+      "loss": 0.1634,
+      "step": 17171
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.3453932295977227,
+      "learning_rate": 5.2496041286717594e-08,
+      "loss": 0.1499,
+      "step": 17172
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.2829235881419654,
+      "learning_rate": 5.2365925346667404e-08,
+      "loss": 0.1718,
+      "step": 17173
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.3975002621836408,
+      "learning_rate": 5.223597000966363e-08,
+      "loss": 0.1983,
+      "step": 17174
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.466557484481432,
+      "learning_rate": 5.210617527992568e-08,
+      "loss": 0.1711,
+      "step": 17175
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.0922479090267867,
+      "learning_rate": 5.197654116166628e-08,
+      "loss": 0.1395,
+      "step": 17176
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.3792613925653474,
+      "learning_rate": 5.184706765909431e-08,
+      "loss": 0.1884,
+      "step": 17177
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.3287113101660954,
+      "learning_rate": 5.171775477641139e-08,
+      "loss": 0.1835,
+      "step": 17178
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.7433865103382329,
+      "learning_rate": 5.158860251781528e-08,
+      "loss": 0.2017,
+      "step": 17179
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.931687773928495,
+      "learning_rate": 5.145961088749929e-08,
+      "loss": 0.1902,
+      "step": 17180
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.1431946941151663,
+      "learning_rate": 5.133077988964952e-08,
+      "loss": 0.1551,
+      "step": 17181
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.9443680642279264,
+      "learning_rate": 5.120210952844873e-08,
+      "loss": 0.2124,
+      "step": 17182
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.2865831633736804,
+      "learning_rate": 5.1073599808073024e-08,
+      "loss": 0.1486,
+      "step": 17183
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.4437385959092275,
+      "learning_rate": 5.094525073269407e-08,
+      "loss": 0.1595,
+      "step": 17184
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.6287333170255245,
+      "learning_rate": 5.0817062306478534e-08,
+      "loss": 0.2065,
+      "step": 17185
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.353706891306041,
+      "learning_rate": 5.068903453358698e-08,
+      "loss": 0.1391,
+      "step": 17186
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.820941415287939,
+      "learning_rate": 5.056116741817496e-08,
+      "loss": 0.192,
+      "step": 17187
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.471264492666836,
+      "learning_rate": 5.0433460964394166e-08,
+      "loss": 0.1719,
+      "step": 17188
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 2.104051059593782,
+      "learning_rate": 5.0305915176389056e-08,
+      "loss": 0.211,
+      "step": 17189
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.3726340713836946,
+      "learning_rate": 5.0178530058300204e-08,
+      "loss": 0.1397,
+      "step": 17190
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.46342064457794,
+      "learning_rate": 5.005130561426263e-08,
+      "loss": 0.1571,
+      "step": 17191
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.466615221176118,
+      "learning_rate": 4.992424184840638e-08,
+      "loss": 0.1805,
+      "step": 17192
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.0559584077575455,
+      "learning_rate": 4.9797338764855373e-08,
+      "loss": 0.1428,
+      "step": 17193
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.261393927461164,
+      "learning_rate": 4.967059636772964e-08,
+      "loss": 0.1945,
+      "step": 17194
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.6677447711371978,
+      "learning_rate": 4.954401466114311e-08,
+      "loss": 0.1622,
+      "step": 17195
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.6498715007481004,
+      "learning_rate": 4.941759364920418e-08,
+      "loss": 0.2386,
+      "step": 17196
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.2153469374444872,
+      "learning_rate": 4.9291333336016764e-08,
+      "loss": 0.1859,
+      "step": 17197
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.3360654312360742,
+      "learning_rate": 4.916523372567983e-08,
+      "loss": 0.1929,
+      "step": 17198
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.3644696005048504,
+      "learning_rate": 4.9039294822286755e-08,
+      "loss": 0.1084,
+      "step": 17199
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.5871201225891423,
+      "learning_rate": 4.8913516629924826e-08,
+      "loss": 0.1455,
+      "step": 17200
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.4031423696152245,
+      "learning_rate": 4.8787899152677455e-08,
+      "loss": 0.1706,
+      "step": 17201
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.7449491790350034,
+      "learning_rate": 4.866244239462192e-08,
+      "loss": 0.172,
+      "step": 17202
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 2.055568092987414,
+      "learning_rate": 4.853714635982998e-08,
+      "loss": 0.2048,
+      "step": 17203
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.5548245834412837,
+      "learning_rate": 4.841201105237059e-08,
+      "loss": 0.1233,
+      "step": 17204
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 2.11130706466066,
+      "learning_rate": 4.828703647630439e-08,
+      "loss": 0.2322,
+      "step": 17205
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.774906062944758,
+      "learning_rate": 4.816222263568815e-08,
+      "loss": 0.205,
+      "step": 17206
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.4904346756419864,
+      "learning_rate": 4.803756953457361e-08,
+      "loss": 0.2238,
+      "step": 17207
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.5610037649254054,
+      "learning_rate": 4.791307717700755e-08,
+      "loss": 0.1666,
+      "step": 17208
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.1859242060322055,
+      "learning_rate": 4.7788745567030616e-08,
+      "loss": 0.1452,
+      "step": 17209
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.937753476205345,
+      "learning_rate": 4.766457470867847e-08,
+      "loss": 0.1712,
+      "step": 17210
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.7971251520927722,
+      "learning_rate": 4.754056460598178e-08,
+      "loss": 0.2035,
+      "step": 17211
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 3.362462518972366,
+      "learning_rate": 4.7416715262966764e-08,
+      "loss": 0.1535,
+      "step": 17212
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.4380286115550909,
+      "learning_rate": 4.7293026683652984e-08,
+      "loss": 0.1428,
+      "step": 17213
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 2.2475050166289425,
+      "learning_rate": 4.7169498872055575e-08,
+      "loss": 0.1667,
+      "step": 17214
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.3585754131156742,
+      "learning_rate": 4.704613183218409e-08,
+      "loss": 0.2114,
+      "step": 17215
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.2364173965404812,
+      "learning_rate": 4.6922925568043676e-08,
+      "loss": 0.1382,
+      "step": 17216
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.6691832219390146,
+      "learning_rate": 4.6799880083633344e-08,
+      "loss": 0.1683,
+      "step": 17217
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.3628306258933067,
+      "learning_rate": 4.6676995382946563e-08,
+      "loss": 0.21,
+      "step": 17218
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.8891737033558416,
+      "learning_rate": 4.655427146997349e-08,
+      "loss": 0.1727,
+      "step": 17219
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.2100180488513184,
+      "learning_rate": 4.6431708348697036e-08,
+      "loss": 0.1559,
+      "step": 17220
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.6728167198993058,
+      "learning_rate": 4.630930602309569e-08,
+      "loss": 0.2019,
+      "step": 17221
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.5570989625622933,
+      "learning_rate": 4.618706449714294e-08,
+      "loss": 0.1813,
+      "step": 17222
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.5652651447482873,
+      "learning_rate": 4.6064983774806726e-08,
+      "loss": 0.1915,
+      "step": 17223
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 4.380719642264466,
+      "learning_rate": 4.5943063860049984e-08,
+      "loss": 0.1755,
+      "step": 17224
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.391517674109408,
+      "learning_rate": 4.582130475683011e-08,
+      "loss": 0.186,
+      "step": 17225
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.491612798057319,
+      "learning_rate": 4.569970646909949e-08,
+      "loss": 0.1747,
+      "step": 17226
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.5757907657213042,
+      "learning_rate": 4.557826900080553e-08,
+      "loss": 0.2276,
+      "step": 17227
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.3236457406412199,
+      "learning_rate": 4.545699235588952e-08,
+      "loss": 0.1551,
+      "step": 17228
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.9211575634423184,
+      "learning_rate": 4.533587653828886e-08,
+      "loss": 0.2217,
+      "step": 17229
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.7057089845662148,
+      "learning_rate": 4.52149215519343e-08,
+      "loss": 0.1642,
+      "step": 17230
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 2.1526823694782506,
+      "learning_rate": 4.5094127400753255e-08,
+      "loss": 0.1786,
+      "step": 17231
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.5023862216796027,
+      "learning_rate": 4.4973494088665914e-08,
+      "loss": 0.1535,
+      "step": 17232
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.9698237229275763,
+      "learning_rate": 4.48530216195886e-08,
+      "loss": 0.1355,
+      "step": 17233
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.4780582340315542,
+      "learning_rate": 4.473270999743096e-08,
+      "loss": 0.1533,
+      "step": 17234
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.202628776471408,
+      "learning_rate": 4.461255922609986e-08,
+      "loss": 0.1786,
+      "step": 17235
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.4371463091729462,
+      "learning_rate": 4.449256930949386e-08,
+      "loss": 0.1878,
+      "step": 17236
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.0438928546340127,
+      "learning_rate": 4.437274025150928e-08,
+      "loss": 0.1628,
+      "step": 17237
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.556335172754529,
+      "learning_rate": 4.4253072056034686e-08,
+      "loss": 0.178,
+      "step": 17238
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.6728127247343265,
+      "learning_rate": 4.413356472695529e-08,
+      "loss": 0.1707,
+      "step": 17239
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.8136258698083474,
+      "learning_rate": 4.4014218268150775e-08,
+      "loss": 0.1885,
+      "step": 17240
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.5509826856030406,
+      "learning_rate": 4.389503268349415e-08,
+      "loss": 0.1944,
+      "step": 17241
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.5131015318350942,
+      "learning_rate": 4.377600797685455e-08,
+      "loss": 0.2008,
+      "step": 17242
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.1906765751418837,
+      "learning_rate": 4.3657144152095544e-08,
+      "loss": 0.1653,
+      "step": 17243
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.0040200981568732,
+      "learning_rate": 4.353844121307627e-08,
+      "loss": 0.1063,
+      "step": 17244
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.7513279006548776,
+      "learning_rate": 4.341989916364919e-08,
+      "loss": 0.1991,
+      "step": 17245
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.0500079432822815,
+      "learning_rate": 4.330151800766291e-08,
+      "loss": 0.1343,
+      "step": 17246
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.6638745842079765,
+      "learning_rate": 4.3183297748958796e-08,
+      "loss": 0.1699,
+      "step": 17247
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.3176721604349146,
+      "learning_rate": 4.306523839137544e-08,
+      "loss": 0.13,
+      "step": 17248
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.0178301566832135,
+      "learning_rate": 4.294733993874533e-08,
+      "loss": 0.1755,
+      "step": 17249
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 2.0256028832037685,
+      "learning_rate": 4.2829602394894866e-08,
+      "loss": 0.1796,
+      "step": 17250
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.3846766573757572,
+      "learning_rate": 4.271202576364597e-08,
+      "loss": 0.1596,
+      "step": 17251
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.4027050636690033,
+      "learning_rate": 4.25946100488156e-08,
+      "loss": 0.1848,
+      "step": 17252
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.5951277957515402,
+      "learning_rate": 4.247735525421459e-08,
+      "loss": 0.2306,
+      "step": 17253
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 2.1616280423990415,
+      "learning_rate": 4.2360261383649906e-08,
+      "loss": 0.1731,
+      "step": 17254
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.6080774706159555,
+      "learning_rate": 4.2243328440921825e-08,
+      "loss": 0.1829,
+      "step": 17255
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 2.1193284193233763,
+      "learning_rate": 4.212655642982566e-08,
+      "loss": 0.2041,
+      "step": 17256
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.664023475010279,
+      "learning_rate": 4.2009945354152816e-08,
+      "loss": 0.1936,
+      "step": 17257
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 4.387640536059977,
+      "learning_rate": 4.18934952176886e-08,
+      "loss": 0.2229,
+      "step": 17258
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.5728810782122224,
+      "learning_rate": 4.17772060242122e-08,
+      "loss": 0.1497,
+      "step": 17259
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.8729754564593715,
+      "learning_rate": 4.166107777749839e-08,
+      "loss": 0.1861,
+      "step": 17260
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.552491021429811,
+      "learning_rate": 4.154511048131804e-08,
+      "loss": 0.1977,
+      "step": 17261
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 2.1808790621107197,
+      "learning_rate": 4.142930413943425e-08,
+      "loss": 0.2007,
+      "step": 17262
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.9437742700932124,
+      "learning_rate": 4.131365875560678e-08,
+      "loss": 0.2083,
+      "step": 17263
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 2.3878716526560178,
+      "learning_rate": 4.1198174333589856e-08,
+      "loss": 0.1971,
+      "step": 17264
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.497071407389878,
+      "learning_rate": 4.108285087713104e-08,
+      "loss": 0.1513,
+      "step": 17265
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.1882070394061865,
+      "learning_rate": 4.096768838997456e-08,
+      "loss": 0.1464,
+      "step": 17266
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.69947790164727,
+      "learning_rate": 4.085268687585908e-08,
+      "loss": 0.1999,
+      "step": 17267
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.4661618723649825,
+      "learning_rate": 4.0737846338516626e-08,
+      "loss": 0.1694,
+      "step": 17268
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.5263777895392654,
+      "learning_rate": 4.062316678167533e-08,
+      "loss": 0.167,
+      "step": 17269
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.317788838451598,
+      "learning_rate": 4.050864820905831e-08,
+      "loss": 0.1508,
+      "step": 17270
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.4959949331293385,
+      "learning_rate": 4.0394290624382605e-08,
+      "loss": 0.1808,
+      "step": 17271
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.4198671091634372,
+      "learning_rate": 4.0280094031359685e-08,
+      "loss": 0.1772,
+      "step": 17272
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.4352590830368825,
+      "learning_rate": 4.0166058433697144e-08,
+      "loss": 0.2068,
+      "step": 17273
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.8573637069755735,
+      "learning_rate": 4.0052183835096455e-08,
+      "loss": 0.1936,
+      "step": 17274
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.5604447031770456,
+      "learning_rate": 3.9938470239254125e-08,
+      "loss": 0.1231,
+      "step": 17275
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.9661989501616666,
+      "learning_rate": 3.982491764986163e-08,
+      "loss": 0.1125,
+      "step": 17276
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.4397132152018686,
+      "learning_rate": 3.97115260706038e-08,
+      "loss": 0.1144,
+      "step": 17277
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.21669987495491,
+      "learning_rate": 3.95982955051627e-08,
+      "loss": 0.1572,
+      "step": 17278
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.1061865932536967,
+      "learning_rate": 3.9485225957213155e-08,
+      "loss": 0.1782,
+      "step": 17279
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.4715958275851704,
+      "learning_rate": 3.937231743042558e-08,
+      "loss": 0.1838,
+      "step": 17280
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.3526715650966044,
+      "learning_rate": 3.9259569928465914e-08,
+      "loss": 0.1823,
+      "step": 17281
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.70057344476964,
+      "learning_rate": 3.914698345499235e-08,
+      "loss": 0.1533,
+      "step": 17282
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.1776600524832859,
+      "learning_rate": 3.903455801366085e-08,
+      "loss": 0.1721,
+      "step": 17283
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.3453788643483493,
+      "learning_rate": 3.8922293608120164e-08,
+      "loss": 0.1957,
+      "step": 17284
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.2567402344541099,
+      "learning_rate": 3.8810190242015156e-08,
+      "loss": 0.1919,
+      "step": 17285
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.3055767352531424,
+      "learning_rate": 3.869824791898347e-08,
+      "loss": 0.1396,
+      "step": 17286
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.136666978397751,
+      "learning_rate": 3.858646664265997e-08,
+      "loss": 0.1708,
+      "step": 17287
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.2098923387393523,
+      "learning_rate": 3.8474846416672874e-08,
+      "loss": 0.1908,
+      "step": 17288
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.4536498388396728,
+      "learning_rate": 3.836338724464539e-08,
+      "loss": 0.1733,
+      "step": 17289
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.3427552877413298,
+      "learning_rate": 3.825208913019518e-08,
+      "loss": 0.1534,
+      "step": 17290
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.5427605065927765,
+      "learning_rate": 3.8140952076936024e-08,
+      "loss": 0.2214,
+      "step": 17291
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.2958934183030715,
+      "learning_rate": 3.802997608847392e-08,
+      "loss": 0.1843,
+      "step": 17292
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.3917599664848264,
+      "learning_rate": 3.7919161168412654e-08,
+      "loss": 0.1636,
+      "step": 17293
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.3954106931747439,
+      "learning_rate": 3.7808507320348797e-08,
+      "loss": 0.1914,
+      "step": 17294
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.3101795267256373,
+      "learning_rate": 3.769801454787447e-08,
+      "loss": 0.171,
+      "step": 17295
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.4262925576790468,
+      "learning_rate": 3.75876828545757e-08,
+      "loss": 0.1721,
+      "step": 17296
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.3252315080572672,
+      "learning_rate": 3.747751224403462e-08,
+      "loss": 0.2082,
+      "step": 17297
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.9057757772698778,
+      "learning_rate": 3.73675027198267e-08,
+      "loss": 0.1773,
+      "step": 17298
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.4965637753541539,
+      "learning_rate": 3.7257654285524084e-08,
+      "loss": 0.1703,
+      "step": 17299
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.279361595089173,
+      "learning_rate": 3.714796694469114e-08,
+      "loss": 0.1396,
+      "step": 17300
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.4353210427287277,
+      "learning_rate": 3.703844070088891e-08,
+      "loss": 0.1653,
+      "step": 17301
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.3641496641400184,
+      "learning_rate": 3.692907555767344e-08,
+      "loss": 0.1424,
+      "step": 17302
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.3287562300759406,
+      "learning_rate": 3.681987151859412e-08,
+      "loss": 0.1404,
+      "step": 17303
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.348937279918656,
+      "learning_rate": 3.671082858719532e-08,
+      "loss": 0.1416,
+      "step": 17304
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.9745916748762846,
+      "learning_rate": 3.660194676701756e-08,
+      "loss": 0.1894,
+      "step": 17305
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.830281705331859,
+      "learning_rate": 3.649322606159411e-08,
+      "loss": 0.1781,
+      "step": 17306
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.4741614412975625,
+      "learning_rate": 3.638466647445549e-08,
+      "loss": 0.1351,
+      "step": 17307
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.505157846596282,
+      "learning_rate": 3.6276268009124424e-08,
+      "loss": 0.2559,
+      "step": 17308
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.1277925944718699,
+      "learning_rate": 3.6168030669120333e-08,
+      "loss": 0.1547,
+      "step": 17309
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.1879991686879086,
+      "learning_rate": 3.6059954457956514e-08,
+      "loss": 0.1839,
+      "step": 17310
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.479896938240951,
+      "learning_rate": 3.5952039379140714e-08,
+      "loss": 0.2217,
+      "step": 17311
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 2.6658147603475406,
+      "learning_rate": 3.584428543617624e-08,
+      "loss": 0.1826,
+      "step": 17312
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.3396094537188687,
+      "learning_rate": 3.5736692632560855e-08,
+      "loss": 0.2018,
+      "step": 17313
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.33774466170834,
+      "learning_rate": 3.562926097178732e-08,
+      "loss": 0.1642,
+      "step": 17314
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.4875463513557654,
+      "learning_rate": 3.552199045734228e-08,
+      "loss": 0.1772,
+      "step": 17315
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.6617592914695145,
+      "learning_rate": 3.541488109270852e-08,
+      "loss": 0.1776,
+      "step": 17316
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.5566587841932347,
+      "learning_rate": 3.5307932881362695e-08,
+      "loss": 0.1725,
+      "step": 17317
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.8646005642128507,
+      "learning_rate": 3.520114582677592e-08,
+      "loss": 0.2113,
+      "step": 17318
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.1864126836631679,
+      "learning_rate": 3.5094519932415417e-08,
+      "loss": 0.2037,
+      "step": 17319
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.2244940647185452,
+      "learning_rate": 3.49880552017412e-08,
+      "loss": 0.1903,
+      "step": 17320
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.6463258786915447,
+      "learning_rate": 3.4881751638210504e-08,
+      "loss": 0.1986,
+      "step": 17321
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.4516746425322873,
+      "learning_rate": 3.477560924527279e-08,
+      "loss": 0.1909,
+      "step": 17322
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 2.4323444129294542,
+      "learning_rate": 3.4669628026373634e-08,
+      "loss": 0.2099,
+      "step": 17323
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.553373941622673,
+      "learning_rate": 3.4563807984954176e-08,
+      "loss": 0.1974,
+      "step": 17324
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.6320549973677072,
+      "learning_rate": 3.4458149124448894e-08,
+      "loss": 0.1912,
+      "step": 17325
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.4781764912659263,
+      "learning_rate": 3.435265144828726e-08,
+      "loss": 0.1731,
+      "step": 17326
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.5593963315800485,
+      "learning_rate": 3.424731495989431e-08,
+      "loss": 0.1963,
+      "step": 17327
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.3690509467474457,
+      "learning_rate": 3.4142139662688975e-08,
+      "loss": 0.1669,
+      "step": 17328
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.5962639783696788,
+      "learning_rate": 3.403712556008465e-08,
+      "loss": 0.1613,
+      "step": 17329
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.3474611583421459,
+      "learning_rate": 3.393227265549193e-08,
+      "loss": 0.2192,
+      "step": 17330
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.7818173876801093,
+      "learning_rate": 3.3827580952312534e-08,
+      "loss": 0.171,
+      "step": 17331
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.134395194202301,
+      "learning_rate": 3.372305045394597e-08,
+      "loss": 0.1359,
+      "step": 17332
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.8189956256662865,
+      "learning_rate": 3.3618681163784526e-08,
+      "loss": 0.2135,
+      "step": 17333
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.3027162250358255,
+      "learning_rate": 3.351447308521716e-08,
+      "loss": 0.1953,
+      "step": 17334
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.8936875814594407,
+      "learning_rate": 3.341042622162616e-08,
+      "loss": 0.174,
+      "step": 17335
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.9367882436430968,
+      "learning_rate": 3.3306540576387717e-08,
+      "loss": 0.1734,
+      "step": 17336
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.716696066820669,
+      "learning_rate": 3.320281615287579e-08,
+      "loss": 0.2334,
+      "step": 17337
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.4776318931286498,
+      "learning_rate": 3.309925295445604e-08,
+      "loss": 0.1853,
+      "step": 17338
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.5897225748507928,
+      "learning_rate": 3.299585098449132e-08,
+      "loss": 0.2129,
+      "step": 17339
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.221649791617027,
+      "learning_rate": 3.2892610246337275e-08,
+      "loss": 0.1668,
+      "step": 17340
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.3543373464957875,
+      "learning_rate": 3.278953074334512e-08,
+      "loss": 0.2089,
+      "step": 17341
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.2220817044729897,
+      "learning_rate": 3.268661247886162e-08,
+      "loss": 0.1748,
+      "step": 17342
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.743762074534679,
+      "learning_rate": 3.258385545622689e-08,
+      "loss": 0.208,
+      "step": 17343
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.5170875207977144,
+      "learning_rate": 3.248125967877658e-08,
+      "loss": 0.1194,
+      "step": 17344
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.544201500887105,
+      "learning_rate": 3.237882514984136e-08,
+      "loss": 0.1682,
+      "step": 17345
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.2650395381444237,
+      "learning_rate": 3.227655187274581e-08,
+      "loss": 0.1836,
+      "step": 17346
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.8085923823110464,
+      "learning_rate": 3.217443985081004e-08,
+      "loss": 0.2082,
+      "step": 17347
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.3669019673412224,
+      "learning_rate": 3.207248908734861e-08,
+      "loss": 0.2024,
+      "step": 17348
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.809033223058747,
+      "learning_rate": 3.197069958567112e-08,
+      "loss": 0.1766,
+      "step": 17349
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 2.5732684602789613,
+      "learning_rate": 3.1869071349081017e-08,
+      "loss": 0.1424,
+      "step": 17350
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.5655218657675625,
+      "learning_rate": 3.1767604380878446e-08,
+      "loss": 0.1708,
+      "step": 17351
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.6142904851783575,
+      "learning_rate": 3.166629868435578e-08,
+      "loss": 0.1823,
+      "step": 17352
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.3032357710817657,
+      "learning_rate": 3.15651542628026e-08,
+      "loss": 0.1961,
+      "step": 17353
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.6415192573787711,
+      "learning_rate": 3.146417111950129e-08,
+      "loss": 0.1877,
+      "step": 17354
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.3550273825282186,
+      "learning_rate": 3.1363349257730324e-08,
+      "loss": 0.147,
+      "step": 17355
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 2.0216862895578798,
+      "learning_rate": 3.1262688680761544e-08,
+      "loss": 0.1553,
+      "step": 17356
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.3638620205229182,
+      "learning_rate": 3.116218939186399e-08,
+      "loss": 0.1393,
+      "step": 17357
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.5572118201778617,
+      "learning_rate": 3.10618513942984e-08,
+      "loss": 0.1653,
+      "step": 17358
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.6058465740461043,
+      "learning_rate": 3.096167469132272e-08,
+      "loss": 0.1827,
+      "step": 17359
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.9489478112172385,
+      "learning_rate": 3.0861659286188785e-08,
+      "loss": 0.2199,
+      "step": 17360
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.1727083199699204,
+      "learning_rate": 3.0761805182142334e-08,
+      "loss": 0.1738,
+      "step": 17361
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.4736659436201829,
+      "learning_rate": 3.0662112382425777e-08,
+      "loss": 0.2112,
+      "step": 17362
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.5853003786534636,
+      "learning_rate": 3.056258089027431e-08,
+      "loss": 0.1997,
+      "step": 17363
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.503323484191681,
+      "learning_rate": 3.046321070891867e-08,
+      "loss": 0.169,
+      "step": 17364
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.079084342715584,
+      "learning_rate": 3.0364001841585166e-08,
+      "loss": 0.118,
+      "step": 17365
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 2.4484580043379656,
+      "learning_rate": 3.0264954291494006e-08,
+      "loss": 0.2394,
+      "step": 17366
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.7375991683213297,
+      "learning_rate": 3.01660680618604e-08,
+      "loss": 0.1621,
+      "step": 17367
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.2996331495077402,
+      "learning_rate": 3.006734315589399e-08,
+      "loss": 0.1787,
+      "step": 17368
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.5645692665527664,
+      "learning_rate": 2.996877957679945e-08,
+      "loss": 0.1866,
+      "step": 17369
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 2.0947584029758834,
+      "learning_rate": 2.9870377327776425e-08,
+      "loss": 0.1967,
+      "step": 17370
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.5846539550883794,
+      "learning_rate": 2.9772136412019038e-08,
+      "loss": 0.1099,
+      "step": 17371
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.513936271956435,
+      "learning_rate": 2.9674056832715847e-08,
+      "loss": 0.1478,
+      "step": 17372
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 2.2530460732718947,
+      "learning_rate": 2.9576138593050972e-08,
+      "loss": 0.2156,
+      "step": 17373
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.773295384228952,
+      "learning_rate": 2.9478381696202985e-08,
+      "loss": 0.2252,
+      "step": 17374
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.1092114563060147,
+      "learning_rate": 2.9380786145344896e-08,
+      "loss": 0.1267,
+      "step": 17375
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.2777009108851471,
+      "learning_rate": 2.9283351943645288e-08,
+      "loss": 0.1946,
+      "step": 17376
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.37260854895467,
+      "learning_rate": 2.918607909426552e-08,
+      "loss": 0.1894,
+      "step": 17377
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 2.0588580329241752,
+      "learning_rate": 2.9088967600364727e-08,
+      "loss": 0.1798,
+      "step": 17378
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.4278995559261976,
+      "learning_rate": 2.8992017465094278e-08,
+      "loss": 0.1618,
+      "step": 17379
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.3104212852825208,
+      "learning_rate": 2.8895228691601662e-08,
+      "loss": 0.1951,
+      "step": 17380
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.5906666680894508,
+      "learning_rate": 2.8798601283028804e-08,
+      "loss": 0.182,
+      "step": 17381
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.4157985547183511,
+      "learning_rate": 2.870213524251153e-08,
+      "loss": 0.176,
+      "step": 17382
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.8843562417812663,
+      "learning_rate": 2.8605830573181227e-08,
+      "loss": 0.2228,
+      "step": 17383
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.3447580964069341,
+      "learning_rate": 2.8509687278164834e-08,
+      "loss": 0.1499,
+      "step": 17384
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.1681968408835361,
+      "learning_rate": 2.8413705360583187e-08,
+      "loss": 0.1654,
+      "step": 17385
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.364648251105591,
+      "learning_rate": 2.8317884823551022e-08,
+      "loss": 0.156,
+      "step": 17386
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.6036907430317413,
+      "learning_rate": 2.8222225670179736e-08,
+      "loss": 0.2232,
+      "step": 17387
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 2.248229707894193,
+      "learning_rate": 2.8126727903573515e-08,
+      "loss": 0.1856,
+      "step": 17388
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.3481565348509776,
+      "learning_rate": 2.8031391526832653e-08,
+      "loss": 0.1606,
+      "step": 17389
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.666453743287798,
+      "learning_rate": 2.7936216543052452e-08,
+      "loss": 0.1885,
+      "step": 17390
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.3428920291203326,
+      "learning_rate": 2.7841202955321556e-08,
+      "loss": 0.1514,
+      "step": 17391
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.4663727879961581,
+      "learning_rate": 2.774635076672416e-08,
+      "loss": 0.1567,
+      "step": 17392
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.3746027589396477,
+      "learning_rate": 2.7651659980339474e-08,
+      "loss": 0.1547,
+      "step": 17393
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.7856112428294766,
+      "learning_rate": 2.75571305992417e-08,
+      "loss": 0.186,
+      "step": 17394
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.7283565297159607,
+      "learning_rate": 2.7462762626498386e-08,
+      "loss": 0.1421,
+      "step": 17395
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.4670857171777183,
+      "learning_rate": 2.7368556065173192e-08,
+      "loss": 0.2041,
+      "step": 17396
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.466864237654432,
+      "learning_rate": 2.7274510918324227e-08,
+      "loss": 0.1847,
+      "step": 17397
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.4082669945992834,
+      "learning_rate": 2.7180627189004606e-08,
+      "loss": 0.1571,
+      "step": 17398
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.093139089544804,
+      "learning_rate": 2.7086904880260778e-08,
+      "loss": 0.1316,
+      "step": 17399
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.45562015653028,
+      "learning_rate": 2.6993343995135866e-08,
+      "loss": 0.1888,
+      "step": 17400
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.2774249643061333,
+      "learning_rate": 2.6899944536666888e-08,
+      "loss": 0.14,
+      "step": 17401
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.5457496209968549,
+      "learning_rate": 2.6806706507885305e-08,
+      "loss": 0.2348,
+      "step": 17402
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.286204386046903,
+      "learning_rate": 2.6713629911817585e-08,
+      "loss": 0.1716,
+      "step": 17403
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.0644481658820026,
+      "learning_rate": 2.6620714751485754e-08,
+      "loss": 0.1396,
+      "step": 17404
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.118416640227578,
+      "learning_rate": 2.6527961029904626e-08,
+      "loss": 0.1349,
+      "step": 17405
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.795000309107337,
+      "learning_rate": 2.643536875008623e-08,
+      "loss": 0.2318,
+      "step": 17406
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 2.2470891720454085,
+      "learning_rate": 2.6342937915036504e-08,
+      "loss": 0.1624,
+      "step": 17407
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.593297520152562,
+      "learning_rate": 2.6250668527754153e-08,
+      "loss": 0.225,
+      "step": 17408
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.9097149266960973,
+      "learning_rate": 2.6158560591235672e-08,
+      "loss": 0.2406,
+      "step": 17409
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.912493790822141,
+      "learning_rate": 2.6066614108470335e-08,
+      "loss": 0.1971,
+      "step": 17410
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 2.1621989961294545,
+      "learning_rate": 2.5974829082442423e-08,
+      "loss": 0.2093,
+      "step": 17411
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.3462285537215757,
+      "learning_rate": 2.5883205516132883e-08,
+      "loss": 0.2084,
+      "step": 17412
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.6556815035442394,
+      "learning_rate": 2.579174341251378e-08,
+      "loss": 0.2153,
+      "step": 17413
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.265139999002601,
+      "learning_rate": 2.570044277455552e-08,
+      "loss": 0.1436,
+      "step": 17414
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.506286745988408,
+      "learning_rate": 2.5609303605221292e-08,
+      "loss": 0.1742,
+      "step": 17415
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.6292003164897628,
+      "learning_rate": 2.551832590746983e-08,
+      "loss": 0.1836,
+      "step": 17416
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.1456701483989684,
+      "learning_rate": 2.5427509684253782e-08,
+      "loss": 0.1569,
+      "step": 17417
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.509925818981673,
+      "learning_rate": 2.5336854938521337e-08,
+      "loss": 0.152,
+      "step": 17418
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.4701313107824343,
+      "learning_rate": 2.5246361673215148e-08,
+      "loss": 0.1792,
+      "step": 17419
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.9541873941630412,
+      "learning_rate": 2.515602989127286e-08,
+      "loss": 0.2061,
+      "step": 17420
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.4419532709658482,
+      "learning_rate": 2.5065859595626573e-08,
+      "loss": 0.143,
+      "step": 17421
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 2.041536678990003,
+      "learning_rate": 2.4975850789203394e-08,
+      "loss": 0.2514,
+      "step": 17422
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.0508004651209466,
+      "learning_rate": 2.4886003474924868e-08,
+      "loss": 0.1716,
+      "step": 17423
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.1798660793674465,
+      "learning_rate": 2.4796317655707002e-08,
+      "loss": 0.1771,
+      "step": 17424
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.3490204029703592,
+      "learning_rate": 2.4706793334462465e-08,
+      "loss": 0.1645,
+      "step": 17425
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.5322690511711963,
+      "learning_rate": 2.461743051409615e-08,
+      "loss": 0.1683,
+      "step": 17426
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.946286666110679,
+      "learning_rate": 2.4528229197508525e-08,
+      "loss": 0.2349,
+      "step": 17427
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.4450310020561234,
+      "learning_rate": 2.443918938759615e-08,
+      "loss": 0.1813,
+      "step": 17428
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.4357025295650112,
+      "learning_rate": 2.4350311087248946e-08,
+      "loss": 0.1907,
+      "step": 17429
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.6945378251957253,
+      "learning_rate": 2.4261594299351822e-08,
+      "loss": 0.1952,
+      "step": 17430
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.9487511144589607,
+      "learning_rate": 2.4173039026784694e-08,
+      "loss": 0.1784,
+      "step": 17431
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.4040526813101786,
+      "learning_rate": 2.4084645272421935e-08,
+      "loss": 0.1854,
+      "step": 17432
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.3177018002082996,
+      "learning_rate": 2.3996413039132917e-08,
+      "loss": 0.1454,
+      "step": 17433
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.494946246400786,
+      "learning_rate": 2.3908342329781453e-08,
+      "loss": 0.1901,
+      "step": 17434
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.4084097066557242,
+      "learning_rate": 2.382043314722693e-08,
+      "loss": 0.1541,
+      "step": 17435
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 2.2923951479438616,
+      "learning_rate": 2.3732685494322615e-08,
+      "loss": 0.2228,
+      "step": 17436
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 2.09299489539299,
+      "learning_rate": 2.364509937391679e-08,
+      "loss": 0.2048,
+      "step": 17437
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.5176094474162092,
+      "learning_rate": 2.3557674788852736e-08,
+      "loss": 0.1958,
+      "step": 17438
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.2673463536046352,
+      "learning_rate": 2.347041174196818e-08,
+      "loss": 0.1417,
+      "step": 17439
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 2.0012660526928,
+      "learning_rate": 2.3383310236095857e-08,
+      "loss": 0.1979,
+      "step": 17440
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.8213634322376828,
+      "learning_rate": 2.329637027406295e-08,
+      "loss": 0.3119,
+      "step": 17441
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.2019348760539221,
+      "learning_rate": 2.320959185869165e-08,
+      "loss": 0.1563,
+      "step": 17442
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.4012598933705225,
+      "learning_rate": 2.312297499279914e-08,
+      "loss": 0.2009,
+      "step": 17443
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.66268213984513,
+      "learning_rate": 2.303651967919651e-08,
+      "loss": 0.2171,
+      "step": 17444
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.3908811986854857,
+      "learning_rate": 2.2950225920690406e-08,
+      "loss": 0.1725,
+      "step": 17445
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.6069631095686205,
+      "learning_rate": 2.2864093720081914e-08,
+      "loss": 0.1781,
+      "step": 17446
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.3968423557947458,
+      "learning_rate": 2.2778123080167136e-08,
+      "loss": 0.1944,
+      "step": 17447
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.34304276973063,
+      "learning_rate": 2.2692314003736616e-08,
+      "loss": 0.1363,
+      "step": 17448
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 2.118857334028887,
+      "learning_rate": 2.26066664935759e-08,
+      "loss": 0.2075,
+      "step": 17449
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.1074895417600128,
+      "learning_rate": 2.252118055246444e-08,
+      "loss": 0.1249,
+      "step": 17450
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.2830550748517409,
+      "learning_rate": 2.243585618317834e-08,
+      "loss": 0.1197,
+      "step": 17451
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.9093039313667919,
+      "learning_rate": 2.23506933884865e-08,
+      "loss": 0.2167,
+      "step": 17452
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.1467847089539873,
+      "learning_rate": 2.2265692171153373e-08,
+      "loss": 0.1344,
+      "step": 17453
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.3887078968639412,
+      "learning_rate": 2.218085253393787e-08,
+      "loss": 0.1963,
+      "step": 17454
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.6385146237239654,
+      "learning_rate": 2.2096174479595002e-08,
+      "loss": 0.1862,
+      "step": 17455
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.423682944558238,
+      "learning_rate": 2.2011658010872573e-08,
+      "loss": 0.1569,
+      "step": 17456
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.9038145853297113,
+      "learning_rate": 2.1927303130514498e-08,
+      "loss": 0.2029,
+      "step": 17457
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.3870034046564483,
+      "learning_rate": 2.1843109841258036e-08,
+      "loss": 0.1736,
+      "step": 17458
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.5362048074641355,
+      "learning_rate": 2.1759078145837664e-08,
+      "loss": 0.1694,
+      "step": 17459
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.5242279154131928,
+      "learning_rate": 2.167520804698009e-08,
+      "loss": 0.2404,
+      "step": 17460
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.8986926123238432,
+      "learning_rate": 2.159149954740758e-08,
+      "loss": 0.18,
+      "step": 17461
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.3631679192685862,
+      "learning_rate": 2.1507952649838517e-08,
+      "loss": 0.1725,
+      "step": 17462
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.7413454586831418,
+      "learning_rate": 2.1424567356983504e-08,
+      "loss": 0.1945,
+      "step": 17463
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.7018409096647849,
+      "learning_rate": 2.134134367154983e-08,
+      "loss": 0.2187,
+      "step": 17464
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.8970562222999354,
+      "learning_rate": 2.125828159623977e-08,
+      "loss": 0.1618,
+      "step": 17465
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.615136509136535,
+      "learning_rate": 2.1175381133748397e-08,
+      "loss": 0.1938,
+      "step": 17466
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.4049051105623538,
+      "learning_rate": 2.109264228676744e-08,
+      "loss": 0.1616,
+      "step": 17467
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.7369379275076657,
+      "learning_rate": 2.101006505798253e-08,
+      "loss": 0.2081,
+      "step": 17468
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.8570195954775166,
+      "learning_rate": 2.092764945007375e-08,
+      "loss": 0.23,
+      "step": 17469
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.4402562104844299,
+      "learning_rate": 2.0845395465716177e-08,
+      "loss": 0.1393,
+      "step": 17470
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.441042412219099,
+      "learning_rate": 2.0763303107581012e-08,
+      "loss": 0.1767,
+      "step": 17471
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.4744987496823156,
+      "learning_rate": 2.068137237833223e-08,
+      "loss": 0.1539,
+      "step": 17472
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.8316502449432226,
+      "learning_rate": 2.0599603280628823e-08,
+      "loss": 0.1789,
+      "step": 17473
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 2.3752816434807533,
+      "learning_rate": 2.051799581712588e-08,
+      "loss": 0.2056,
+      "step": 17474
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.7225590546537772,
+      "learning_rate": 2.0436549990472952e-08,
+      "loss": 0.1828,
+      "step": 17475
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.799726752561715,
+      "learning_rate": 2.0355265803312373e-08,
+      "loss": 0.1567,
+      "step": 17476
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.5627392490324115,
+      "learning_rate": 2.0274143258283142e-08,
+      "loss": 0.1807,
+      "step": 17477
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.387138518078584,
+      "learning_rate": 2.019318235801926e-08,
+      "loss": 0.1536,
+      "step": 17478
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.6608473472495249,
+      "learning_rate": 2.011238310514807e-08,
+      "loss": 0.1984,
+      "step": 17479
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.5646913992298166,
+      "learning_rate": 2.003174550229248e-08,
+      "loss": 0.204,
+      "step": 17480
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.2720030797442756,
+      "learning_rate": 1.9951269552070384e-08,
+      "loss": 0.1662,
+      "step": 17481
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.3798326292854415,
+      "learning_rate": 1.9870955257093593e-08,
+      "loss": 0.1591,
+      "step": 17482
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.5073994975172573,
+      "learning_rate": 1.9790802619969462e-08,
+      "loss": 0.1704,
+      "step": 17483
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.4595663296504677,
+      "learning_rate": 1.9710811643299798e-08,
+      "loss": 0.2166,
+      "step": 17484
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.3146246788225455,
+      "learning_rate": 1.963098232968086e-08,
+      "loss": 0.1595,
+      "step": 17485
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.1568673586374065,
+      "learning_rate": 1.955131468170446e-08,
+      "loss": 0.1751,
+      "step": 17486
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.6511336510563246,
+      "learning_rate": 1.9471808701955754e-08,
+      "loss": 0.1478,
+      "step": 17487
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.8132351749060787,
+      "learning_rate": 1.9392464393017118e-08,
+      "loss": 0.1813,
+      "step": 17488
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.1009155700474467,
+      "learning_rate": 1.9313281757462054e-08,
+      "loss": 0.1572,
+      "step": 17489
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.607303628940861,
+      "learning_rate": 1.9234260797862392e-08,
+      "loss": 0.1807,
+      "step": 17490
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.8209627979119294,
+      "learning_rate": 1.9155401516782746e-08,
+      "loss": 0.1923,
+      "step": 17491
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.6765341501247675,
+      "learning_rate": 1.907670391678329e-08,
+      "loss": 0.1412,
+      "step": 17492
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.4501379881552487,
+      "learning_rate": 1.8998168000418093e-08,
+      "loss": 0.194,
+      "step": 17493
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.5220899539768362,
+      "learning_rate": 1.891979377023678e-08,
+      "loss": 0.1535,
+      "step": 17494
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.239592703081296,
+      "learning_rate": 1.884158122878288e-08,
+      "loss": 0.1688,
+      "step": 17495
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.7599474212869162,
+      "learning_rate": 1.8763530378595462e-08,
+      "loss": 0.2189,
+      "step": 17496
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.6051692265576725,
+      "learning_rate": 1.868564122220917e-08,
+      "loss": 0.1937,
+      "step": 17497
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.4315530872514335,
+      "learning_rate": 1.8607913762150875e-08,
+      "loss": 0.2023,
+      "step": 17498
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.1526775652473806,
+      "learning_rate": 1.853034800094411e-08,
+      "loss": 0.1799,
+      "step": 17499
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.4274304245145792,
+      "learning_rate": 1.845294394110686e-08,
+      "loss": 0.1225,
+      "step": 17500
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.6127409873205572,
+      "learning_rate": 1.837570158515156e-08,
+      "loss": 0.1639,
+      "step": 17501
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.14226088072883,
+      "learning_rate": 1.8298620935585655e-08,
+      "loss": 0.1124,
+      "step": 17502
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.5600090675671239,
+      "learning_rate": 1.8221701994911023e-08,
+      "loss": 0.1512,
+      "step": 17503
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.8710887407144219,
+      "learning_rate": 1.8144944765625117e-08,
+      "loss": 0.1689,
+      "step": 17504
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.5811744347711862,
+      "learning_rate": 1.8068349250219275e-08,
+      "loss": 0.1502,
+      "step": 17505
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.6015993532732298,
+      "learning_rate": 1.799191545117929e-08,
+      "loss": 0.1682,
+      "step": 17506
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.3182404471549902,
+      "learning_rate": 1.7915643370986503e-08,
+      "loss": 0.1381,
+      "step": 17507
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.5140574105790365,
+      "learning_rate": 1.7839533012116717e-08,
+      "loss": 0.1805,
+      "step": 17508
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.3081791918556687,
+      "learning_rate": 1.776358437704073e-08,
+      "loss": 0.1415,
+      "step": 17509
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.4334435865041393,
+      "learning_rate": 1.76877974682238e-08,
+      "loss": 0.1439,
+      "step": 17510
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.339841714751808,
+      "learning_rate": 1.7612172288126173e-08,
+      "loss": 0.1393,
+      "step": 17511
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.3470273167409266,
+      "learning_rate": 1.7536708839202554e-08,
+      "loss": 0.1349,
+      "step": 17512
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.2257281581839607,
+      "learning_rate": 1.7461407123902098e-08,
+      "loss": 0.1989,
+      "step": 17513
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.4521524445940872,
+      "learning_rate": 1.7386267144669512e-08,
+      "loss": 0.1783,
+      "step": 17514
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.5133138614699304,
+      "learning_rate": 1.7311288903944513e-08,
+      "loss": 0.2206,
+      "step": 17515
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.5209467259901799,
+      "learning_rate": 1.7236472404159598e-08,
+      "loss": 0.1727,
+      "step": 17516
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.6163329732408154,
+      "learning_rate": 1.7161817647744494e-08,
+      "loss": 0.158,
+      "step": 17517
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.140122874663967,
+      "learning_rate": 1.708732463712226e-08,
+      "loss": 0.1627,
+      "step": 17518
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.5389898544439384,
+      "learning_rate": 1.7012993374710408e-08,
+      "loss": 0.1456,
+      "step": 17519
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.770415398889119,
+      "learning_rate": 1.693882386292256e-08,
+      "loss": 0.1954,
+      "step": 17520
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.9204699489771948,
+      "learning_rate": 1.6864816104165683e-08,
+      "loss": 0.1804,
+      "step": 17521
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.5656967554000436,
+      "learning_rate": 1.679097010084285e-08,
+      "loss": 0.1783,
+      "step": 17522
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.236511507641056,
+      "learning_rate": 1.6717285855350485e-08,
+      "loss": 0.1566,
+      "step": 17523
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.8277607416320738,
+      "learning_rate": 1.6643763370080557e-08,
+      "loss": 0.1668,
+      "step": 17524
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.411210424275455,
+      "learning_rate": 1.6570402647419493e-08,
+      "loss": 0.1714,
+      "step": 17525
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.2129010370517102,
+      "learning_rate": 1.6497203689749276e-08,
+      "loss": 0.1869,
+      "step": 17526
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.2690947413112235,
+      "learning_rate": 1.6424166499445227e-08,
+      "loss": 0.112,
+      "step": 17527
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.522274331333948,
+      "learning_rate": 1.6351291078878228e-08,
+      "loss": 0.1728,
+      "step": 17528
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.3335588036077557,
+      "learning_rate": 1.627857743041472e-08,
+      "loss": 0.1677,
+      "step": 17529
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.6852165066376268,
+      "learning_rate": 1.620602555641393e-08,
+      "loss": 0.173,
+      "step": 17530
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.8267600270052624,
+      "learning_rate": 1.6133635459231744e-08,
+      "loss": 0.2312,
+      "step": 17531
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.4047550944697407,
+      "learning_rate": 1.60614071412174e-08,
+      "loss": 0.2293,
+      "step": 17532
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.5824266010365458,
+      "learning_rate": 1.598934060471624e-08,
+      "loss": 0.1854,
+      "step": 17533
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.5194081254123215,
+      "learning_rate": 1.5917435852066953e-08,
+      "loss": 0.207,
+      "step": 17534
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.2490720940747073,
+      "learning_rate": 1.5845692885603224e-08,
+      "loss": 0.1936,
+      "step": 17535
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.3899702609606368,
+      "learning_rate": 1.577411170765486e-08,
+      "loss": 0.2132,
+      "step": 17536
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.1862757797173369,
+      "learning_rate": 1.5702692320545e-08,
+      "loss": 0.1021,
+      "step": 17537
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.6999941229718665,
+      "learning_rate": 1.5631434726591786e-08,
+      "loss": 0.2131,
+      "step": 17538
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.5527705542230128,
+      "learning_rate": 1.5560338928108377e-08,
+      "loss": 0.177,
+      "step": 17539
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.708703803873028,
+      "learning_rate": 1.5489404927402918e-08,
+      "loss": 0.1875,
+      "step": 17540
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.5432443783032144,
+      "learning_rate": 1.5418632726777462e-08,
+      "loss": 0.1832,
+      "step": 17541
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.965161261869477,
+      "learning_rate": 1.534802232852961e-08,
+      "loss": 0.2402,
+      "step": 17542
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.6993474801947406,
+      "learning_rate": 1.5277573734950867e-08,
+      "loss": 0.1465,
+      "step": 17543
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.5365710909465382,
+      "learning_rate": 1.5207286948328847e-08,
+      "loss": 0.1666,
+      "step": 17544
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.6610642586656525,
+      "learning_rate": 1.5137161970945056e-08,
+      "loss": 0.1758,
+      "step": 17545
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.852583680045429,
+      "learning_rate": 1.50671988050749e-08,
+      "loss": 0.192,
+      "step": 17546
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.7806241960614078,
+      "learning_rate": 1.4997397452990447e-08,
+      "loss": 0.1803,
+      "step": 17547
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.6937754030891603,
+      "learning_rate": 1.4927757916957107e-08,
+      "loss": 0.1666,
+      "step": 17548
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.311301456165458,
+      "learning_rate": 1.4858280199234742e-08,
+      "loss": 0.186,
+      "step": 17549
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.432840138816705,
+      "learning_rate": 1.4788964302079877e-08,
+      "loss": 0.2076,
+      "step": 17550
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.490136009559635,
+      "learning_rate": 1.4719810227741826e-08,
+      "loss": 0.18,
+      "step": 17551
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.5825539655508407,
+      "learning_rate": 1.465081797846546e-08,
+      "loss": 0.1838,
+      "step": 17552
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.3192766121176644,
+      "learning_rate": 1.4581987556490095e-08,
+      "loss": 0.1537,
+      "step": 17553
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.7242153318066196,
+      "learning_rate": 1.451331896405006e-08,
+      "loss": 0.1873,
+      "step": 17554
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.4202192695578617,
+      "learning_rate": 1.4444812203375235e-08,
+      "loss": 0.1879,
+      "step": 17555
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.7400285532748854,
+      "learning_rate": 1.4376467276688288e-08,
+      "loss": 0.1616,
+      "step": 17556
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.8510615278698002,
+      "learning_rate": 1.4308284186207999e-08,
+      "loss": 0.2145,
+      "step": 17557
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.8933102792804857,
+      "learning_rate": 1.4240262934148153e-08,
+      "loss": 0.2115,
+      "step": 17558
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.5679193974565009,
+      "learning_rate": 1.4172403522716427e-08,
+      "loss": 0.1657,
+      "step": 17559
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.6297511829869904,
+      "learning_rate": 1.4104705954115506e-08,
+      "loss": 0.2254,
+      "step": 17560
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.2862712232397666,
+      "learning_rate": 1.4037170230542519e-08,
+      "loss": 0.1711,
+      "step": 17561
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 2.0632557134698253,
+      "learning_rate": 1.3969796354190713e-08,
+      "loss": 0.189,
+      "step": 17562
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.399526379930175,
+      "learning_rate": 1.3902584327246115e-08,
+      "loss": 0.1593,
+      "step": 17563
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.5827451719856045,
+      "learning_rate": 1.3835534151890872e-08,
+      "loss": 0.2206,
+      "step": 17564
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.1375517005489526,
+      "learning_rate": 1.3768645830301575e-08,
+      "loss": 0.1289,
+      "step": 17565
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.7910849621246112,
+      "learning_rate": 1.3701919364648708e-08,
+      "loss": 0.1686,
+      "step": 17566
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.5220692872844677,
+      "learning_rate": 1.3635354757099429e-08,
+      "loss": 0.1716,
+      "step": 17567
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.8482181751086562,
+      "learning_rate": 1.3568952009813674e-08,
+      "loss": 0.2043,
+      "step": 17568
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.287765937487343,
+      "learning_rate": 1.3502711124947498e-08,
+      "loss": 0.1267,
+      "step": 17569
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.168915023984947,
+      "learning_rate": 1.3436632104650848e-08,
+      "loss": 0.1654,
+      "step": 17570
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.227990418745462,
+      "learning_rate": 1.337071495106812e-08,
+      "loss": 0.1725,
+      "step": 17571
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.5938581168209707,
+      "learning_rate": 1.3304959666339267e-08,
+      "loss": 0.1812,
+      "step": 17572
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 2.1840599565522556,
+      "learning_rate": 1.3239366252599806e-08,
+      "loss": 0.1286,
+      "step": 17573
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.254842842148272,
+      "learning_rate": 1.3173934711977476e-08,
+      "loss": 0.1659,
+      "step": 17574
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.507093129486963,
+      "learning_rate": 1.3108665046597247e-08,
+      "loss": 0.1661,
+      "step": 17575
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.3834500486426489,
+      "learning_rate": 1.3043557258577422e-08,
+      "loss": 0.1822,
+      "step": 17576
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.4850365569932433,
+      "learning_rate": 1.2978611350031311e-08,
+      "loss": 0.1965,
+      "step": 17577
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.6346008144782234,
+      "learning_rate": 1.291382732306723e-08,
+      "loss": 0.1521,
+      "step": 17578
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.73794966546174,
+      "learning_rate": 1.2849205179787938e-08,
+      "loss": 0.1506,
+      "step": 17579
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.4439461461295313,
+      "learning_rate": 1.2784744922291204e-08,
+      "loss": 0.1634,
+      "step": 17580
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.4021577392941238,
+      "learning_rate": 1.2720446552669797e-08,
+      "loss": 0.1328,
+      "step": 17581
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.5461692452371019,
+      "learning_rate": 1.2656310073010381e-08,
+      "loss": 0.176,
+      "step": 17582
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.6679413232625517,
+      "learning_rate": 1.2592335485395179e-08,
+      "loss": 0.1928,
+      "step": 17583
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.302075323716006,
+      "learning_rate": 1.2528522791900311e-08,
+      "loss": 0.1093,
+      "step": 17584
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.7422760397817585,
+      "learning_rate": 1.2464871994598004e-08,
+      "loss": 0.1977,
+      "step": 17585
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.0338318089278495,
+      "learning_rate": 1.2401383095553832e-08,
+      "loss": 0.1602,
+      "step": 17586
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.4712642394635107,
+      "learning_rate": 1.2338056096829475e-08,
+      "loss": 0.2054,
+      "step": 17587
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.4604782157730771,
+      "learning_rate": 1.227489100047885e-08,
+      "loss": 0.2114,
+      "step": 17588
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.7325983527437705,
+      "learning_rate": 1.2211887808554202e-08,
+      "loss": 0.1588,
+      "step": 17589
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.0678452155907021,
+      "learning_rate": 1.214904652309945e-08,
+      "loss": 0.1234,
+      "step": 17590
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.9690195062517374,
+      "learning_rate": 1.2086367146155187e-08,
+      "loss": 0.1265,
+      "step": 17591
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.4421845275914404,
+      "learning_rate": 1.202384967975534e-08,
+      "loss": 0.1526,
+      "step": 17592
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.4426808291497795,
+      "learning_rate": 1.196149412592995e-08,
+      "loss": 0.1836,
+      "step": 17593
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.4689677310425324,
+      "learning_rate": 1.1899300486702959e-08,
+      "loss": 0.1527,
+      "step": 17594
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.3920990814148684,
+      "learning_rate": 1.1837268764092746e-08,
+      "loss": 0.18,
+      "step": 17595
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.3770978919147934,
+      "learning_rate": 1.1775398960113261e-08,
+      "loss": 0.1512,
+      "step": 17596
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.2558750964377605,
+      "learning_rate": 1.1713691076772337e-08,
+      "loss": 0.1537,
+      "step": 17597
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 2.2093243324696537,
+      "learning_rate": 1.1652145116073931e-08,
+      "loss": 0.1858,
+      "step": 17598
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.4539726512736153,
+      "learning_rate": 1.1590761080015334e-08,
+      "loss": 0.2149,
+      "step": 17599
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.581466385582127,
+      "learning_rate": 1.1529538970588839e-08,
+      "loss": 0.1744,
+      "step": 17600
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 2.445807239343453,
+      "learning_rate": 1.1468478789782855e-08,
+      "loss": 0.2263,
+      "step": 17601
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.8655398892552575,
+      "learning_rate": 1.1407580539578022e-08,
+      "loss": 0.2141,
+      "step": 17602
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.0651902262733317,
+      "learning_rate": 1.1346844221952203e-08,
+      "loss": 0.1567,
+      "step": 17603
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.88894224154827,
+      "learning_rate": 1.128626983887604e-08,
+      "loss": 0.1909,
+      "step": 17604
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.2253806683199984,
+      "learning_rate": 1.1225857392316298e-08,
+      "loss": 0.1587,
+      "step": 17605
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.5751820213102414,
+      "learning_rate": 1.1165606884234182e-08,
+      "loss": 0.1791,
+      "step": 17606
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.5527928851157964,
+      "learning_rate": 1.110551831658535e-08,
+      "loss": 0.1838,
+      "step": 17607
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.628501347776376,
+      "learning_rate": 1.1045591691319913e-08,
+      "loss": 0.2143,
+      "step": 17608
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.2884630312114418,
+      "learning_rate": 1.0985827010383531e-08,
+      "loss": 0.1412,
+      "step": 17609
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.320535409608244,
+      "learning_rate": 1.0926224275716323e-08,
+      "loss": 0.1406,
+      "step": 17610
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.4134506883159856,
+      "learning_rate": 1.0866783489252298e-08,
+      "loss": 0.1939,
+      "step": 17611
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.9945263615633104,
+      "learning_rate": 1.080750465292213e-08,
+      "loss": 0.2107,
+      "step": 17612
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.2357998603734293,
+      "learning_rate": 1.0748387768648727e-08,
+      "loss": 0.1545,
+      "step": 17613
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.7680051509061494,
+      "learning_rate": 1.068943283835222e-08,
+      "loss": 0.2009,
+      "step": 17614
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.667139137374589,
+      "learning_rate": 1.0630639863945525e-08,
+      "loss": 0.1888,
+      "step": 17615
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.6382961092426487,
+      "learning_rate": 1.0572008847337111e-08,
+      "loss": 0.1827,
+      "step": 17616
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.2742059442705793,
+      "learning_rate": 1.051353979043046e-08,
+      "loss": 0.1468,
+      "step": 17617
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.4192662201086075,
+      "learning_rate": 1.0455232695124051e-08,
+      "loss": 0.2002,
+      "step": 17618
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.577599636292609,
+      "learning_rate": 1.0397087563309704e-08,
+      "loss": 0.1782,
+      "step": 17619
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.624405896724841,
+      "learning_rate": 1.03391043968748e-08,
+      "loss": 0.1711,
+      "step": 17620
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.5209351749374962,
+      "learning_rate": 1.0281283197701719e-08,
+      "loss": 0.1602,
+      "step": 17621
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.1278451103438372,
+      "learning_rate": 1.0223623967667851e-08,
+      "loss": 0.1638,
+      "step": 17622
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.8615594381922296,
+      "learning_rate": 1.0166126708644475e-08,
+      "loss": 0.1934,
+      "step": 17623
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.5904981405186749,
+      "learning_rate": 1.0108791422497876e-08,
+      "loss": 0.1707,
+      "step": 17624
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.2180239703449756,
+      "learning_rate": 1.0051618111089346e-08,
+      "loss": 0.165,
+      "step": 17625
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.5086334938412644,
+      "learning_rate": 9.994606776274619e-09,
+      "loss": 0.1534,
+      "step": 17626
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.3154454559536521,
+      "learning_rate": 9.937757419904437e-09,
+      "loss": 0.1367,
+      "step": 17627
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.2276598644069867,
+      "learning_rate": 9.881070043824548e-09,
+      "loss": 0.1669,
+      "step": 17628
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.62374296536291,
+      "learning_rate": 9.824544649874035e-09,
+      "loss": 0.1633,
+      "step": 17629
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.6191414670861497,
+      "learning_rate": 9.768181239888653e-09,
+      "loss": 0.203,
+      "step": 17630
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.7472869947641716,
+      "learning_rate": 9.711979815698046e-09,
+      "loss": 0.2021,
+      "step": 17631
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.3611425768638437,
+      "learning_rate": 9.655940379125761e-09,
+      "loss": 0.1483,
+      "step": 17632
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.732068125702614,
+      "learning_rate": 9.60006293199145e-09,
+      "loss": 0.1871,
+      "step": 17633
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.5550951590966957,
+      "learning_rate": 9.544347476108662e-09,
+      "loss": 0.2175,
+      "step": 17634
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.83007478662795,
+      "learning_rate": 9.488794013285952e-09,
+      "loss": 0.1756,
+      "step": 17635
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.4866715985517136,
+      "learning_rate": 9.433402545326875e-09,
+      "loss": 0.1871,
+      "step": 17636
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.593259773405187,
+      "learning_rate": 9.378173074029439e-09,
+      "loss": 0.1716,
+      "step": 17637
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.5316605017689835,
+      "learning_rate": 9.3231056011861e-09,
+      "loss": 0.1701,
+      "step": 17638
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.1668861692146373,
+      "learning_rate": 9.26820012858487e-09,
+      "loss": 0.1551,
+      "step": 17639
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.4692438796783915,
+      "learning_rate": 9.213456658007102e-09,
+      "loss": 0.1955,
+      "step": 17640
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.5360641464145106,
+      "learning_rate": 9.158875191231375e-09,
+      "loss": 0.1802,
+      "step": 17641
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.067317133468182,
+      "learning_rate": 9.10445573002794e-09,
+      "loss": 0.1624,
+      "step": 17642
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 2.2464064272638833,
+      "learning_rate": 9.050198276163712e-09,
+      "loss": 0.1886,
+      "step": 17643
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.6529421432983202,
+      "learning_rate": 8.996102831400621e-09,
+      "loss": 0.1906,
+      "step": 17644
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.010988281071243,
+      "learning_rate": 8.942169397493372e-09,
+      "loss": 0.1103,
+      "step": 17645
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.4227942335404447,
+      "learning_rate": 8.888397976193897e-09,
+      "loss": 0.1843,
+      "step": 17646
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.441421591410884,
+      "learning_rate": 8.83478856924691e-09,
+      "loss": 0.1605,
+      "step": 17647
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.6238768391555098,
+      "learning_rate": 8.781341178393244e-09,
+      "loss": 0.1972,
+      "step": 17648
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.1855269215144273,
+      "learning_rate": 8.728055805366508e-09,
+      "loss": 0.1614,
+      "step": 17649
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.1346396936081093,
+      "learning_rate": 8.674932451898099e-09,
+      "loss": 0.1365,
+      "step": 17650
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.3149430550820955,
+      "learning_rate": 8.62197111971108e-09,
+      "loss": 0.1687,
+      "step": 17651
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.8125043244145862,
+      "learning_rate": 8.569171810525189e-09,
+      "loss": 0.161,
+      "step": 17652
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.3234262317326384,
+      "learning_rate": 8.51653452605461e-09,
+      "loss": 0.1667,
+      "step": 17653
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.6926330187744023,
+      "learning_rate": 8.464059268007418e-09,
+      "loss": 0.1495,
+      "step": 17654
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.6543277886214691,
+      "learning_rate": 8.411746038087255e-09,
+      "loss": 0.166,
+      "step": 17655
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.449800497956295,
+      "learning_rate": 8.359594837992202e-09,
+      "loss": 0.2153,
+      "step": 17656
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.1930184696831028,
+      "learning_rate": 8.307605669414797e-09,
+      "loss": 0.1594,
+      "step": 17657
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.5557805105761005,
+      "learning_rate": 8.255778534043134e-09,
+      "loss": 0.1875,
+      "step": 17658
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.6795544638179751,
+      "learning_rate": 8.204113433559202e-09,
+      "loss": 0.181,
+      "step": 17659
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.7225910615588798,
+      "learning_rate": 8.152610369640546e-09,
+      "loss": 0.1933,
+      "step": 17660
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.9704068568942654,
+      "learning_rate": 8.101269343958607e-09,
+      "loss": 0.1684,
+      "step": 17661
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.4261424613015403,
+      "learning_rate": 8.05009035817983e-09,
+      "loss": 0.2015,
+      "step": 17662
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.2976161550979781,
+      "learning_rate": 7.999073413965664e-09,
+      "loss": 0.142,
+      "step": 17663
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.2760982390598385,
+      "learning_rate": 7.94821851297256e-09,
+      "loss": 0.1691,
+      "step": 17664
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.3186246308400245,
+      "learning_rate": 7.897525656850868e-09,
+      "loss": 0.1634,
+      "step": 17665
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.0338839647561422,
+      "learning_rate": 7.846994847245937e-09,
+      "loss": 0.1599,
+      "step": 17666
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.5087885217758248,
+      "learning_rate": 7.79662608579812e-09,
+      "loss": 0.1847,
+      "step": 17667
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 2.0590032752871577,
+      "learning_rate": 7.746419374142777e-09,
+      "loss": 0.1954,
+      "step": 17668
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.9661441238766413,
+      "learning_rate": 7.69637471390916e-09,
+      "loss": 0.2019,
+      "step": 17669
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.5204868204949955,
+      "learning_rate": 7.646492106722637e-09,
+      "loss": 0.1486,
+      "step": 17670
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.418054972147326,
+      "learning_rate": 7.5967715542008e-09,
+      "loss": 0.1802,
+      "step": 17671
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.4021792833373585,
+      "learning_rate": 7.547213057959579e-09,
+      "loss": 0.1978,
+      "step": 17672
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.473580882154094,
+      "learning_rate": 7.497816619606025e-09,
+      "loss": 0.1198,
+      "step": 17673
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.3306288062346254,
+      "learning_rate": 7.448582240744407e-09,
+      "loss": 0.1821,
+      "step": 17674
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.3748187465845771,
+      "learning_rate": 7.399509922972892e-09,
+      "loss": 0.1367,
+      "step": 17675
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.187279398404566,
+      "learning_rate": 7.350599667884095e-09,
+      "loss": 0.1421,
+      "step": 17676
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.2062878093822296,
+      "learning_rate": 7.3018514770656354e-09,
+      "loss": 0.1571,
+      "step": 17677
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.7533880967920887,
+      "learning_rate": 7.253265352100691e-09,
+      "loss": 0.1597,
+      "step": 17678
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.4044726053569543,
+      "learning_rate": 7.204841294565224e-09,
+      "loss": 0.1625,
+      "step": 17679
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.1887454917393026,
+      "learning_rate": 7.156579306031863e-09,
+      "loss": 0.1664,
+      "step": 17680
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.3757614823177087,
+      "learning_rate": 7.10847938806658e-09,
+      "loss": 0.161,
+      "step": 17681
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.9840506528872281,
+      "learning_rate": 7.060541542232014e-09,
+      "loss": 0.2171,
+      "step": 17682
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.6554957417785023,
+      "learning_rate": 7.01276577008303e-09,
+      "loss": 0.18,
+      "step": 17683
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 2.3451250111905377,
+      "learning_rate": 6.965152073170611e-09,
+      "loss": 0.1777,
+      "step": 17684
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.2377332443485443,
+      "learning_rate": 6.917700453040743e-09,
+      "loss": 0.1772,
+      "step": 17685
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.7198124506145238,
+      "learning_rate": 6.870410911233305e-09,
+      "loss": 0.1537,
+      "step": 17686
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.4705300602906368,
+      "learning_rate": 6.8232834492842906e-09,
+      "loss": 0.19,
+      "step": 17687
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.6318472970855926,
+      "learning_rate": 6.776318068721921e-09,
+      "loss": 0.1879,
+      "step": 17688
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.306411737371864,
+      "learning_rate": 6.729514771072198e-09,
+      "loss": 0.122,
+      "step": 17689
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.991306169822796,
+      "learning_rate": 6.682873557853353e-09,
+      "loss": 0.2006,
+      "step": 17690
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.4034227613882622,
+      "learning_rate": 6.636394430580284e-09,
+      "loss": 0.163,
+      "step": 17691
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.051651781948377,
+      "learning_rate": 6.590077390760674e-09,
+      "loss": 0.1583,
+      "step": 17692
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.4475900747096362,
+      "learning_rate": 6.5439224398988755e-09,
+      "loss": 0.1993,
+      "step": 17693
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.7002635856649913,
+      "learning_rate": 6.497929579492579e-09,
+      "loss": 0.1816,
+      "step": 17694
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.142053225983412,
+      "learning_rate": 6.452098811035035e-09,
+      "loss": 0.1954,
+      "step": 17695
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.3262672839445186,
+      "learning_rate": 6.406430136013941e-09,
+      "loss": 0.1358,
+      "step": 17696
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 2.095289731737613,
+      "learning_rate": 6.3609235559120016e-09,
+      "loss": 0.2023,
+      "step": 17697
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.1205835876323187,
+      "learning_rate": 6.315579072205258e-09,
+      "loss": 0.1554,
+      "step": 17698
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.5967983750026877,
+      "learning_rate": 6.270396686366975e-09,
+      "loss": 0.1786,
+      "step": 17699
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.5069327480938488,
+      "learning_rate": 6.225376399863203e-09,
+      "loss": 0.1543,
+      "step": 17700
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.400529740152614,
+      "learning_rate": 6.180518214154995e-09,
+      "loss": 0.17,
+      "step": 17701
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.5801709858657913,
+      "learning_rate": 6.135822130698965e-09,
+      "loss": 0.2223,
+      "step": 17702
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.5412875323586859,
+      "learning_rate": 6.091288150946173e-09,
+      "loss": 0.1997,
+      "step": 17703
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.187503601369918,
+      "learning_rate": 6.046916276341574e-09,
+      "loss": 0.1801,
+      "step": 17704
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.4983828489900441,
+      "learning_rate": 6.0027065083256845e-09,
+      "loss": 0.1561,
+      "step": 17705
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.2987521844906098,
+      "learning_rate": 5.9586588483340204e-09,
+      "loss": 0.1688,
+      "step": 17706
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 2.2978783316943714,
+      "learning_rate": 5.91477329779544e-09,
+      "loss": 0.1991,
+      "step": 17707
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.6963317022416338,
+      "learning_rate": 5.871049858135469e-09,
+      "loss": 0.1774,
+      "step": 17708
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.618901878213332,
+      "learning_rate": 5.827488530773528e-09,
+      "loss": 0.1741,
+      "step": 17709
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.9966720482397342,
+      "learning_rate": 5.784089317122932e-09,
+      "loss": 0.2252,
+      "step": 17710
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.1311263327681884,
+      "learning_rate": 5.740852218591997e-09,
+      "loss": 0.1672,
+      "step": 17711
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.7406217099061072,
+      "learning_rate": 5.6977772365857105e-09,
+      "loss": 0.1681,
+      "step": 17712
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.4730047476365675,
+      "learning_rate": 5.654864372501289e-09,
+      "loss": 0.1376,
+      "step": 17713
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.518542971059855,
+      "learning_rate": 5.612113627731508e-09,
+      "loss": 0.2049,
+      "step": 17714
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.8865609744743332,
+      "learning_rate": 5.5695250036647e-09,
+      "loss": 0.1244,
+      "step": 17715
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.267728634966635,
+      "learning_rate": 5.527098501683648e-09,
+      "loss": 0.1334,
+      "step": 17716
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.5642980767806371,
+      "learning_rate": 5.4848341231644734e-09,
+      "loss": 0.1934,
+      "step": 17717
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.4528897412912745,
+      "learning_rate": 5.442731869479967e-09,
+      "loss": 0.1858,
+      "step": 17718
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.0417400079975305,
+      "learning_rate": 5.400791741995704e-09,
+      "loss": 0.1331,
+      "step": 17719
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.7392221883030923,
+      "learning_rate": 5.359013742074481e-09,
+      "loss": 0.2293,
+      "step": 17720
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.5103708512384117,
+      "learning_rate": 5.3173978710718835e-09,
+      "loss": 0.175,
+      "step": 17721
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.546096090283196,
+      "learning_rate": 5.275944130338495e-09,
+      "loss": 0.1833,
+      "step": 17722
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.552406093994639,
+      "learning_rate": 5.234652521220463e-09,
+      "loss": 0.1856,
+      "step": 17723
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.4925879545269995,
+      "learning_rate": 5.193523045057269e-09,
+      "loss": 0.1523,
+      "step": 17724
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.3500594011236497,
+      "learning_rate": 5.152555703184514e-09,
+      "loss": 0.1771,
+      "step": 17725
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.2897382467780292,
+      "learning_rate": 5.111750496932244e-09,
+      "loss": 0.1576,
+      "step": 17726
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.205977699200912,
+      "learning_rate": 5.071107427624955e-09,
+      "loss": 0.1645,
+      "step": 17727
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.9138107912809603,
+      "learning_rate": 5.030626496581592e-09,
+      "loss": 0.2198,
+      "step": 17728
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.4255496790704636,
+      "learning_rate": 4.99030770511666e-09,
+      "loss": 0.1679,
+      "step": 17729
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.1741824289771698,
+      "learning_rate": 4.9501510545385546e-09,
+      "loss": 0.1218,
+      "step": 17730
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.4751545485656512,
+      "learning_rate": 4.9101565461506795e-09,
+      "loss": 0.1398,
+      "step": 17731
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.646172517721418,
+      "learning_rate": 4.870324181251995e-09,
+      "loss": 0.1813,
+      "step": 17732
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.057540066514655,
+      "learning_rate": 4.8306539611348016e-09,
+      "loss": 0.1543,
+      "step": 17733
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.237227710716961,
+      "learning_rate": 4.791145887086956e-09,
+      "loss": 0.1573,
+      "step": 17734
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.1289499102798763,
+      "learning_rate": 4.751799960390768e-09,
+      "loss": 0.1436,
+      "step": 17735
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.6959224342055175,
+      "learning_rate": 4.712616182324103e-09,
+      "loss": 0.2278,
+      "step": 17736
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.6290537655309192,
+      "learning_rate": 4.6735945541581674e-09,
+      "loss": 0.2392,
+      "step": 17737
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.4750393474309156,
+      "learning_rate": 4.634735077159724e-09,
+      "loss": 0.1118,
+      "step": 17738
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.4304164320640622,
+      "learning_rate": 4.5960377525899876e-09,
+      "loss": 0.1793,
+      "step": 17739
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.3470743443363888,
+      "learning_rate": 4.557502581706286e-09,
+      "loss": 0.1856,
+      "step": 17740
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.3633878059607694,
+      "learning_rate": 4.519129565758173e-09,
+      "loss": 0.1825,
+      "step": 17741
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.916725633757655,
+      "learning_rate": 4.4809187059913216e-09,
+      "loss": 0.1595,
+      "step": 17742
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.8539774955373358,
+      "learning_rate": 4.4428700036469595e-09,
+      "loss": 0.1936,
+      "step": 17743
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.3661977564166994,
+      "learning_rate": 4.404983459959655e-09,
+      "loss": 0.1833,
+      "step": 17744
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.199337372520092,
+      "learning_rate": 4.367259076159536e-09,
+      "loss": 0.1724,
+      "step": 17745
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.4729029198436245,
+      "learning_rate": 4.329696853470622e-09,
+      "loss": 0.2258,
+      "step": 17746
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.2940634734705803,
+      "learning_rate": 4.292296793111939e-09,
+      "loss": 0.1756,
+      "step": 17747
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.6587084212622196,
+      "learning_rate": 4.255058896298625e-09,
+      "loss": 0.1662,
+      "step": 17748
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.892597729844158,
+      "learning_rate": 4.217983164238604e-09,
+      "loss": 0.1992,
+      "step": 17749
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.9330538448677852,
+      "learning_rate": 4.1810695981359115e-09,
+      "loss": 0.1975,
+      "step": 17750
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.7796412447711583,
+      "learning_rate": 4.14431819918848e-09,
+      "loss": 0.2114,
+      "step": 17751
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.4313920600220245,
+      "learning_rate": 4.107728968589242e-09,
+      "loss": 0.183,
+      "step": 17752
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.27327946694493,
+      "learning_rate": 4.0713019075261375e-09,
+      "loss": 0.172,
+      "step": 17753
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.287925585873027,
+      "learning_rate": 4.035037017180998e-09,
+      "loss": 0.2146,
+      "step": 17754
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.5280244681086932,
+      "learning_rate": 3.998934298731771e-09,
+      "loss": 0.1794,
+      "step": 17755
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.3382077363372549,
+      "learning_rate": 3.962993753349742e-09,
+      "loss": 0.1391,
+      "step": 17756
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.376977668705793,
+      "learning_rate": 3.927215382202309e-09,
+      "loss": 0.1613,
+      "step": 17757
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.1428410219488154,
+      "learning_rate": 3.891599186449657e-09,
+      "loss": 0.1589,
+      "step": 17758
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.489726253691237,
+      "learning_rate": 3.856145167249192e-09,
+      "loss": 0.1679,
+      "step": 17759
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.4111817453737503,
+      "learning_rate": 3.8208533257516614e-09,
+      "loss": 0.1711,
+      "step": 17760
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 2.1415217880928195,
+      "learning_rate": 3.785723663101703e-09,
+      "loss": 0.143,
+      "step": 17761
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.3785011842035597,
+      "learning_rate": 3.750756180440074e-09,
+      "loss": 0.1638,
+      "step": 17762
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.9756234543048203,
+      "learning_rate": 3.7159508789019748e-09,
+      "loss": 0.1837,
+      "step": 17763
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.4969399178310896,
+      "learning_rate": 3.6813077596176138e-09,
+      "loss": 0.1551,
+      "step": 17764
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.7036234624924167,
+      "learning_rate": 3.6468268237105364e-09,
+      "loss": 0.1676,
+      "step": 17765
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.9028718125372321,
+      "learning_rate": 3.6125080723009574e-09,
+      "loss": 0.2019,
+      "step": 17766
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.008470837751592,
+      "learning_rate": 3.578351506502431e-09,
+      "loss": 0.1238,
+      "step": 17767
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.7043218869199757,
+      "learning_rate": 3.5443571274240696e-09,
+      "loss": 0.139,
+      "step": 17768
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 2.3150602357269943,
+      "learning_rate": 3.5105249361683245e-09,
+      "loss": 0.2533,
+      "step": 17769
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.7176969683587147,
+      "learning_rate": 3.476854933834317e-09,
+      "loss": 0.2359,
+      "step": 17770
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.3593931361336906,
+      "learning_rate": 3.4433471215150616e-09,
+      "loss": 0.1884,
+      "step": 17771
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.6449381852637894,
+      "learning_rate": 3.4100015002980213e-09,
+      "loss": 0.1743,
+      "step": 17772
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.6663549375078202,
+      "learning_rate": 3.376818071265664e-09,
+      "loss": 0.1565,
+      "step": 17773
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.4126376856101048,
+      "learning_rate": 3.343796835494906e-09,
+      "loss": 0.187,
+      "step": 17774
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.1550112351783484,
+      "learning_rate": 3.310937794057667e-09,
+      "loss": 0.1997,
+      "step": 17775
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.5250738305839973,
+      "learning_rate": 3.2782409480208723e-09,
+      "loss": 0.1618,
+      "step": 17776
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.5646027816075057,
+      "learning_rate": 3.2457062984453393e-09,
+      "loss": 0.2134,
+      "step": 17777
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.539090074011398,
+      "learning_rate": 3.2133338463880006e-09,
+      "loss": 0.1672,
+      "step": 17778
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.5995164036078096,
+      "learning_rate": 3.181123592898572e-09,
+      "loss": 0.2232,
+      "step": 17779
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.3438394428027403,
+      "learning_rate": 3.1490755390239935e-09,
+      "loss": 0.2385,
+      "step": 17780
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.2761273614822526,
+      "learning_rate": 3.1171896858034345e-09,
+      "loss": 0.1721,
+      "step": 17781
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.608870644843655,
+      "learning_rate": 3.0854660342721778e-09,
+      "loss": 0.2274,
+      "step": 17782
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.2189433910303995,
+      "learning_rate": 3.0539045854605097e-09,
+      "loss": 0.1199,
+      "step": 17783
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.7159221447431454,
+      "learning_rate": 3.0225053403920568e-09,
+      "loss": 0.2039,
+      "step": 17784
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.9165824339505007,
+      "learning_rate": 2.9912683000865583e-09,
+      "loss": 0.1736,
+      "step": 17785
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 2.1703524982015567,
+      "learning_rate": 2.9601934655587584e-09,
+      "loss": 0.1953,
+      "step": 17786
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.5230998520933274,
+      "learning_rate": 2.929280837815629e-09,
+      "loss": 0.1991,
+      "step": 17787
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.9233497163193343,
+      "learning_rate": 2.8985304178619223e-09,
+      "loss": 0.1203,
+      "step": 17788
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.4668524588099205,
+      "learning_rate": 2.8679422066957287e-09,
+      "loss": 0.1695,
+      "step": 17789
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.289011671046741,
+      "learning_rate": 2.837516205309587e-09,
+      "loss": 0.1249,
+      "step": 17790
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 2.2885446906702835,
+      "learning_rate": 2.8072524146910417e-09,
+      "loss": 0.2053,
+      "step": 17791
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.3635875226706184,
+      "learning_rate": 2.7771508358226396e-09,
+      "loss": 0.1589,
+      "step": 17792
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.6766503932272478,
+      "learning_rate": 2.747211469681932e-09,
+      "loss": 0.2086,
+      "step": 17793
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.727498383417767,
+      "learning_rate": 2.7174343172398087e-09,
+      "loss": 0.214,
+      "step": 17794
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.5767380171541705,
+      "learning_rate": 2.6878193794638294e-09,
+      "loss": 0.1738,
+      "step": 17795
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.3607715333588843,
+      "learning_rate": 2.6583666573148925e-09,
+      "loss": 0.1419,
+      "step": 17796
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.7608043936828974,
+      "learning_rate": 2.6290761517488995e-09,
+      "loss": 0.2016,
+      "step": 17797
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.425624190785581,
+      "learning_rate": 2.5999478637167565e-09,
+      "loss": 0.1449,
+      "step": 17798
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.6347361918617482,
+      "learning_rate": 2.5709817941643734e-09,
+      "loss": 0.1504,
+      "step": 17799
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.6140771061992503,
+      "learning_rate": 2.542177944030999e-09,
+      "loss": 0.1765,
+      "step": 17800
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.1288889649331753,
+      "learning_rate": 2.5135363142525515e-09,
+      "loss": 0.1617,
+      "step": 17801
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.4514247011738122,
+      "learning_rate": 2.4850569057588425e-09,
+      "loss": 0.1687,
+      "step": 17802
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.5859395482844012,
+      "learning_rate": 2.4567397194735776e-09,
+      "loss": 0.175,
+      "step": 17803
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.097782344974049,
+      "learning_rate": 2.4285847563165763e-09,
+      "loss": 0.1505,
+      "step": 17804
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.127594390177761,
+      "learning_rate": 2.4005920172009976e-09,
+      "loss": 0.1727,
+      "step": 17805
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 2.612509940096549,
+      "learning_rate": 2.372761503036669e-09,
+      "loss": 0.1725,
+      "step": 17806
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.4022800409833218,
+      "learning_rate": 2.3450932147256467e-09,
+      "loss": 0.1223,
+      "step": 17807
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.3827269791704293,
+      "learning_rate": 2.3175871531672113e-09,
+      "loss": 0.1894,
+      "step": 17808
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.4406174463003725,
+      "learning_rate": 2.290243319253982e-09,
+      "loss": 0.1619,
+      "step": 17809
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.4579927631477871,
+      "learning_rate": 2.263061713873027e-09,
+      "loss": 0.1582,
+      "step": 17810
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.7967584986859246,
+      "learning_rate": 2.236042337906974e-09,
+      "loss": 0.2061,
+      "step": 17811
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.4217101515063342,
+      "learning_rate": 2.209185192232899e-09,
+      "loss": 0.1363,
+      "step": 17812
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.4180075847677778,
+      "learning_rate": 2.1824902777228818e-09,
+      "loss": 0.1774,
+      "step": 17813
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.1944544960247143,
+      "learning_rate": 2.155957595242897e-09,
+      "loss": 0.1498,
+      "step": 17814
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.5637565805218963,
+      "learning_rate": 2.129587145654477e-09,
+      "loss": 0.1742,
+      "step": 17815
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.5700730573002284,
+      "learning_rate": 2.1033789298141592e-09,
+      "loss": 0.1842,
+      "step": 17816
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.350001514177886,
+      "learning_rate": 2.0773329485718197e-09,
+      "loss": 0.1955,
+      "step": 17817
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.425072397099564,
+      "learning_rate": 2.0514492027728928e-09,
+      "loss": 0.171,
+      "step": 17818
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.201494415522038,
+      "learning_rate": 2.0257276932583724e-09,
+      "loss": 0.1524,
+      "step": 17819
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.3740981731681514,
+      "learning_rate": 2.0001684208625915e-09,
+      "loss": 0.1411,
+      "step": 17820
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 3.7346003316787297,
+      "learning_rate": 1.9747713864154418e-09,
+      "loss": 0.2084,
+      "step": 17821
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.6668925420369423,
+      "learning_rate": 1.949536590741263e-09,
+      "loss": 0.1846,
+      "step": 17822
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.550406441928631,
+      "learning_rate": 1.9244640346594014e-09,
+      "loss": 0.1999,
+      "step": 17823
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.560660729295261,
+      "learning_rate": 1.8995537189830937e-09,
+      "loss": 0.2325,
+      "step": 17824
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 2.072051458770098,
+      "learning_rate": 1.8748056445222486e-09,
+      "loss": 0.2491,
+      "step": 17825
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.428708590794355,
+      "learning_rate": 1.8502198120784464e-09,
+      "loss": 0.1911,
+      "step": 17826
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.186958029320143,
+      "learning_rate": 1.8257962224510484e-09,
+      "loss": 0.176,
+      "step": 17827
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.2799956124293856,
+      "learning_rate": 1.8015348764321983e-09,
+      "loss": 0.1482,
+      "step": 17828
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.52518589731545,
+      "learning_rate": 1.7774357748101545e-09,
+      "loss": 0.154,
+      "step": 17829
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.1369843909958626,
+      "learning_rate": 1.753498918366514e-09,
+      "loss": 0.1405,
+      "step": 17830
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.3852949220422877,
+      "learning_rate": 1.7297243078784332e-09,
+      "loss": 0.1763,
+      "step": 17831
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.370258582303878,
+      "learning_rate": 1.7061119441175167e-09,
+      "loss": 0.1816,
+      "step": 17832
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.4217116398292386,
+      "learning_rate": 1.6826618278509289e-09,
+      "loss": 0.1919,
+      "step": 17833
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.508931653837053,
+      "learning_rate": 1.6593739598386172e-09,
+      "loss": 0.163,
+      "step": 17834
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.4227120534645534,
+      "learning_rate": 1.636248340838309e-09,
+      "loss": 0.1732,
+      "step": 17835
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.2404121271547324,
+      "learning_rate": 1.6132849715988496e-09,
+      "loss": 0.1155,
+      "step": 17836
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.3960628166678262,
+      "learning_rate": 1.5904838528668643e-09,
+      "loss": 0.1599,
+      "step": 17837
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.10956166268556,
+      "learning_rate": 1.5678449853817612e-09,
+      "loss": 0.1716,
+      "step": 17838
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.276799071253517,
+      "learning_rate": 1.5453683698785082e-09,
+      "loss": 0.1583,
+      "step": 17839
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.3601565964121545,
+      "learning_rate": 1.523054007087632e-09,
+      "loss": 0.1422,
+      "step": 17840
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.2685783918484506,
+      "learning_rate": 1.500901897731888e-09,
+      "loss": 0.1454,
+      "step": 17841
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.1653420245439718,
+      "learning_rate": 1.4789120425318104e-09,
+      "loss": 0.1296,
+      "step": 17842
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.5598718092591002,
+      "learning_rate": 1.457084442200718e-09,
+      "loss": 0.2036,
+      "step": 17843
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.4695269908569526,
+      "learning_rate": 1.4354190974463778e-09,
+      "loss": 0.1967,
+      "step": 17844
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.6370628468649255,
+      "learning_rate": 1.4139160089732263e-09,
+      "loss": 0.129,
+      "step": 17845
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.3430372388240066,
+      "learning_rate": 1.3925751774784834e-09,
+      "loss": 0.1344,
+      "step": 17846
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.5468755683512558,
+      "learning_rate": 1.3713966036554838e-09,
+      "loss": 0.1741,
+      "step": 17847
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.7844458737970683,
+      "learning_rate": 1.350380288191455e-09,
+      "loss": 0.1783,
+      "step": 17848
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.5840546654699204,
+      "learning_rate": 1.3295262317680746e-09,
+      "loss": 0.2151,
+      "step": 17849
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.817593799542333,
+      "learning_rate": 1.3088344350631332e-09,
+      "loss": 0.2383,
+      "step": 17850
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.6333165371851912,
+      "learning_rate": 1.2883048987472057e-09,
+      "loss": 0.2251,
+      "step": 17851
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.4262035580545218,
+      "learning_rate": 1.2679376234880914e-09,
+      "loss": 0.1611,
+      "step": 17852
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.3683268459067932,
+      "learning_rate": 1.2477326099458176e-09,
+      "loss": 0.2256,
+      "step": 17853
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.6999097602924824,
+      "learning_rate": 1.2276898587765262e-09,
+      "loss": 0.1687,
+      "step": 17854
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.1788721298326021,
+      "learning_rate": 1.2078093706313633e-09,
+      "loss": 0.1291,
+      "step": 17855
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.4513324838636403,
+      "learning_rate": 1.1880911461548128e-09,
+      "loss": 0.1691,
+      "step": 17856
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.4682884099708013,
+      "learning_rate": 1.1685351859874739e-09,
+      "loss": 0.1776,
+      "step": 17857
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.4058983257098954,
+      "learning_rate": 1.149141490764394e-09,
+      "loss": 0.1799,
+      "step": 17858
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.698828096536692,
+      "learning_rate": 1.1299100611139591e-09,
+      "loss": 0.1551,
+      "step": 17859
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.492916408481659,
+      "learning_rate": 1.1108408976617802e-09,
+      "loss": 0.1972,
+      "step": 17860
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 2.0687230936761227,
+      "learning_rate": 1.0919340010262514e-09,
+      "loss": 0.1562,
+      "step": 17861
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.3315400870907148,
+      "learning_rate": 1.073189371820771e-09,
+      "loss": 0.18,
+      "step": 17862
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.384049772633413,
+      "learning_rate": 1.054607010654296e-09,
+      "loss": 0.1394,
+      "step": 17863
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.5467388738596122,
+      "learning_rate": 1.0361869181302331e-09,
+      "loss": 0.1969,
+      "step": 17864
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.3700342350357742,
+      "learning_rate": 1.017929094845882e-09,
+      "loss": 0.1607,
+      "step": 17865
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.5894727625539624,
+      "learning_rate": 9.998335413941018e-10,
+      "loss": 0.1717,
+      "step": 17866
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.6457496895439947,
+      "learning_rate": 9.819002583622006e-10,
+      "loss": 0.1771,
+      "step": 17867
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.4764710655361375,
+      "learning_rate": 9.641292463324903e-10,
+      "loss": 0.2009,
+      "step": 17868
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.6086351860827184,
+      "learning_rate": 9.465205058822868e-10,
+      "loss": 0.197,
+      "step": 17869
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.3224428922982352,
+      "learning_rate": 9.290740375822449e-10,
+      "loss": 0.1518,
+      "step": 17870
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.6066276840998381,
+      "learning_rate": 9.117898419991333e-10,
+      "loss": 0.1484,
+      "step": 17871
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.3628927185444868,
+      "learning_rate": 8.946679196941698e-10,
+      "loss": 0.2372,
+      "step": 17872
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.3742710587184173,
+      "learning_rate": 8.777082712230212e-10,
+      "loss": 0.1311,
+      "step": 17873
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.5262690126326781,
+      "learning_rate": 8.609108971363577e-10,
+      "loss": 0.1356,
+      "step": 17874
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.0530898108107436,
+      "learning_rate": 8.442757979787442e-10,
+      "loss": 0.1636,
+      "step": 17875
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.5717486864495254,
+      "learning_rate": 8.278029742908589e-10,
+      "loss": 0.1815,
+      "step": 17876
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.5302915785054412,
+      "learning_rate": 8.114924266072744e-10,
+      "loss": 0.1915,
+      "step": 17877
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.279028187265196,
+      "learning_rate": 7.953441554570118e-10,
+      "loss": 0.1741,
+      "step": 17878
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.8607004954789625,
+      "learning_rate": 7.793581613652069e-10,
+      "loss": 0.2132,
+      "step": 17879
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.341719579344338,
+      "learning_rate": 7.635344448497783e-10,
+      "loss": 0.1146,
+      "step": 17880
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.68495010296438,
+      "learning_rate": 7.478730064253148e-10,
+      "loss": 0.1516,
+      "step": 17881
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.5489528889572144,
+      "learning_rate": 7.323738465991881e-10,
+      "loss": 0.1804,
+      "step": 17882
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.5799218938707071,
+      "learning_rate": 7.170369658748844e-10,
+      "loss": 0.2208,
+      "step": 17883
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.5670237002056315,
+      "learning_rate": 7.018623647508937e-10,
+      "loss": 0.1658,
+      "step": 17884
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.2756618135495934,
+      "learning_rate": 6.8685004371849e-10,
+      "loss": 0.1747,
+      "step": 17885
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.174698365839384,
+      "learning_rate": 6.720000032661711e-10,
+      "loss": 0.1155,
+      "step": 17886
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.604135874250401,
+      "learning_rate": 6.57312243875774e-10,
+      "loss": 0.1522,
+      "step": 17887
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.8380476613913956,
+      "learning_rate": 6.427867660235843e-10,
+      "loss": 0.1859,
+      "step": 17888
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.502685229551652,
+      "learning_rate": 6.284235701814467e-10,
+      "loss": 0.1541,
+      "step": 17889
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.8011242091023407,
+      "learning_rate": 6.14222656815655e-10,
+      "loss": 0.1849,
+      "step": 17890
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.73051357857051,
+      "learning_rate": 6.001840263869519e-10,
+      "loss": 0.1681,
+      "step": 17891
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.0599978063612407,
+      "learning_rate": 5.863076793510836e-10,
+      "loss": 0.1617,
+      "step": 17892
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 2.041394270676722,
+      "learning_rate": 5.725936161582458e-10,
+      "loss": 0.177,
+      "step": 17893
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.2536292141440253,
+      "learning_rate": 5.590418372541928e-10,
+      "loss": 0.1607,
+      "step": 17894
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.7225634837528563,
+      "learning_rate": 5.456523430785731e-10,
+      "loss": 0.183,
+      "step": 17895
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.5225663163943064,
+      "learning_rate": 5.324251340660391e-10,
+      "loss": 0.1805,
+      "step": 17896
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.6500954900188805,
+      "learning_rate": 5.193602106456919e-10,
+      "loss": 0.1612,
+      "step": 17897
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.4400839283942701,
+      "learning_rate": 5.064575732421916e-10,
+      "loss": 0.1666,
+      "step": 17898
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.2767748046836682,
+      "learning_rate": 4.937172222735376e-10,
+      "loss": 0.1274,
+      "step": 17899
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.6317275243572738,
+      "learning_rate": 4.811391581538427e-10,
+      "loss": 0.1888,
+      "step": 17900
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.7508802073614982,
+      "learning_rate": 4.68723381291114e-10,
+      "loss": 0.1896,
+      "step": 17901
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.6791766343479573,
+      "learning_rate": 4.5646989208891767e-10,
+      "loss": 0.1643,
+      "step": 17902
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.6829606833072932,
+      "learning_rate": 4.4437869094471344e-10,
+      "loss": 0.1974,
+      "step": 17903
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.470449063468551,
+      "learning_rate": 4.3244977825041e-10,
+      "loss": 0.1623,
+      "step": 17904
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.5696869358530308,
+      "learning_rate": 4.206831543945855e-10,
+      "loss": 0.1665,
+      "step": 17905
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.3478594723491355,
+      "learning_rate": 4.090788197574913e-10,
+      "loss": 0.1477,
+      "step": 17906
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.4954683675019182,
+      "learning_rate": 3.9763677471715835e-10,
+      "loss": 0.1721,
+      "step": 17907
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.3480354770652756,
+      "learning_rate": 3.863570196444011e-10,
+      "loss": 0.2266,
+      "step": 17908
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 2.0312992676848864,
+      "learning_rate": 3.752395549055932e-10,
+      "loss": 0.2475,
+      "step": 17909
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.4243792604028678,
+      "learning_rate": 3.6428438086100194e-10,
+      "loss": 0.1689,
+      "step": 17910
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.3977233085942637,
+      "learning_rate": 3.5349149786700896e-10,
+      "loss": 0.1889,
+      "step": 17911
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.3126786825621954,
+      "learning_rate": 3.4286090627388967e-10,
+      "loss": 0.1582,
+      "step": 17912
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.5219491584146134,
+      "learning_rate": 3.323926064263683e-10,
+      "loss": 0.1636,
+      "step": 17913
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.5162311059744773,
+      "learning_rate": 3.220865986647281e-10,
+      "loss": 0.1761,
+      "step": 17914
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.6812752073022001,
+      "learning_rate": 3.1194288332259124e-10,
+      "loss": 0.1594,
+      "step": 17915
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.8367096271556056,
+      "learning_rate": 3.01961460730249e-10,
+      "loss": 0.1574,
+      "step": 17916
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.6672791388798607,
+      "learning_rate": 2.921423312113314e-10,
+      "loss": 0.1746,
+      "step": 17917
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.1968410292901555,
+      "learning_rate": 2.8248549508391733e-10,
+      "loss": 0.1265,
+      "step": 17918
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 2.223245304671502,
+      "learning_rate": 2.72990952662755e-10,
+      "loss": 0.189,
+      "step": 17919
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.5213109506434845,
+      "learning_rate": 2.636587042548211e-10,
+      "loss": 0.1539,
+      "step": 17920
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.6240946558546672,
+      "learning_rate": 2.5448875016376163e-10,
+      "loss": 0.196,
+      "step": 17921
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.6036764568375463,
+      "learning_rate": 2.4548109068711635e-10,
+      "loss": 0.1957,
+      "step": 17922
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.288467073195994,
+      "learning_rate": 2.3663572611742904e-10,
+      "loss": 0.1331,
+      "step": 17923
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.477260354046923,
+      "learning_rate": 2.2795265674113721e-10,
+      "loss": 0.1531,
+      "step": 17924
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.6205150812360052,
+      "learning_rate": 2.1943188284079264e-10,
+      "loss": 0.1642,
+      "step": 17925
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.5245820709840827,
+      "learning_rate": 2.1107340469284087e-10,
+      "loss": 0.1841,
+      "step": 17926
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.2620525125334956,
+      "learning_rate": 2.0287722256873143e-10,
+      "loss": 0.1315,
+      "step": 17927
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.3257201581376976,
+      "learning_rate": 1.9484333673380762e-10,
+      "loss": 0.1633,
+      "step": 17928
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 2.183979846483645,
+      "learning_rate": 1.8697174744952694e-10,
+      "loss": 0.204,
+      "step": 17929
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.1750105362784993,
+      "learning_rate": 1.7926245497179584e-10,
+      "loss": 0.111,
+      "step": 17930
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.547018015637016,
+      "learning_rate": 1.7171545954985934e-10,
+      "loss": 0.1859,
+      "step": 17931
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.6616012388215111,
+      "learning_rate": 1.643307614290768e-10,
+      "loss": 0.1256,
+      "step": 17932
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.2741324394927283,
+      "learning_rate": 1.571083608492563e-10,
+      "loss": 0.1467,
+      "step": 17933
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.48675770510596,
+      "learning_rate": 1.5004825804465494e-10,
+      "loss": 0.2075,
+      "step": 17934
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 2.203495846642513,
+      "learning_rate": 1.4315045324508892e-10,
+      "loss": 0.2594,
+      "step": 17935
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.4882506570149756,
+      "learning_rate": 1.3641494667371301e-10,
+      "loss": 0.1531,
+      "step": 17936
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.5543537260838378,
+      "learning_rate": 1.298417385492412e-10,
+      "loss": 0.1726,
+      "step": 17937
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.5789906705232821,
+      "learning_rate": 1.234308290853914e-10,
+      "loss": 0.1342,
+      "step": 17938
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.5075497867814722,
+      "learning_rate": 1.1718221849033041e-10,
+      "loss": 0.1766,
+      "step": 17939
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.624367361531456,
+      "learning_rate": 1.1109590696611883e-10,
+      "loss": 0.169,
+      "step": 17940
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.6779868276815622,
+      "learning_rate": 1.0517189471148659e-10,
+      "loss": 0.1865,
+      "step": 17941
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.7141078638700016,
+      "learning_rate": 9.941018191794716e-11,
+      "loss": 0.1862,
+      "step": 17942
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.297740808128472,
+      "learning_rate": 9.381076877257311e-11,
+      "loss": 0.1753,
+      "step": 17943
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.709355820752212,
+      "learning_rate": 8.837365545688592e-11,
+      "loss": 0.1618,
+      "step": 17944
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.6519344670650808,
+      "learning_rate": 8.309884214852127e-11,
+      "loss": 0.1959,
+      "step": 17945
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.5642229546663897,
+      "learning_rate": 7.798632901734326e-11,
+      "loss": 0.1555,
+      "step": 17946
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.3789909758217214,
+      "learning_rate": 7.303611623044049e-11,
+      "loss": 0.1797,
+      "step": 17947
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.4013415078392182,
+      "learning_rate": 6.824820394712994e-11,
+      "loss": 0.1542,
+      "step": 17948
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 4.093322076344518,
+      "learning_rate": 6.362259232450818e-11,
+      "loss": 0.1912,
+      "step": 17949
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.3413588462722879,
+      "learning_rate": 5.915928151134509e-11,
+      "loss": 0.1846,
+      "step": 17950
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.6693430516475338,
+      "learning_rate": 5.4858271653079885e-11,
+      "loss": 0.2136,
+      "step": 17951
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 2.4982049607160626,
+      "learning_rate": 5.071956288960067e-11,
+      "loss": 0.179,
+      "step": 17952
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.5107542356813306,
+      "learning_rate": 4.674315535468932e-11,
+      "loss": 0.1483,
+      "step": 17953
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.2696988039220045,
+      "learning_rate": 4.292904917768681e-11,
+      "loss": 0.2264,
+      "step": 17954
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.4153195584273415,
+      "learning_rate": 3.927724448238301e-11,
+      "loss": 0.2113,
+      "step": 17955
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.507554901927853,
+      "learning_rate": 3.5787741387571794e-11,
+      "loss": 0.219,
+      "step": 17956
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.959475888236047,
+      "learning_rate": 3.246054000594079e-11,
+      "loss": 0.239,
+      "step": 17957
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.3334572952978954,
+      "learning_rate": 2.929564044573674e-11,
+      "loss": 0.156,
+      "step": 17958
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.2065981486047839,
+      "learning_rate": 2.629304281021039e-11,
+      "loss": 0.186,
+      "step": 17959
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.9834891753776754,
+      "learning_rate": 2.3452747196506254e-11,
+      "loss": 0.1984,
+      "step": 17960
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.6849670960117122,
+      "learning_rate": 2.077475369621773e-11,
+      "loss": 0.1694,
+      "step": 17961
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.573860690948552,
+      "learning_rate": 1.825906239705244e-11,
+      "loss": 0.1868,
+      "step": 17962
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.8139799915774828,
+      "learning_rate": 1.5905673380056662e-11,
+      "loss": 0.2045,
+      "step": 17963
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.335187997323363,
+      "learning_rate": 1.3714586722390899e-11,
+      "loss": 0.1323,
+      "step": 17964
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.2112534695291186,
+      "learning_rate": 1.168580249455431e-11,
+      "loss": 0.1379,
+      "step": 17965
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.1080701804061914,
+      "learning_rate": 9.819320762605166e-12,
+      "loss": 0.1258,
+      "step": 17966
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.337852538743148,
+      "learning_rate": 8.115141587050623e-12,
+      "loss": 0.1409,
+      "step": 17967
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.3330543146650733,
+      "learning_rate": 6.573265023401831e-12,
+      "loss": 0.1636,
+      "step": 17968
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.4674282213687266,
+      "learning_rate": 5.193691121618827e-12,
+      "loss": 0.1932,
+      "step": 17969
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.130718538675683,
+      "learning_rate": 3.9764199261105305e-12,
+      "loss": 0.153,
+      "step": 17970
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.3316723935854942,
+      "learning_rate": 2.921451476844972e-12,
+      "loss": 0.1764,
+      "step": 17971
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.4938870183598352,
+      "learning_rate": 2.0287858082390645e-12,
+      "loss": 0.1937,
+      "step": 17972
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.343149353119868,
+      "learning_rate": 1.2984229491586064e-12,
+      "loss": 0.1615,
+      "step": 17973
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.0956259987625818,
+      "learning_rate": 7.303629223631703e-13,
+      "loss": 0.1381,
+      "step": 17974
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.983292952962602,
+      "learning_rate": 3.2460574783677037e-13,
+      "loss": 0.2322,
+      "step": 17975
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.3915595921662296,
+      "learning_rate": 8.115143779185985e-14,
+      "loss": 0.1667,
+      "step": 17976
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.2984514364236004,
+      "learning_rate": 0.0,
+      "loss": 0.1613,
+      "step": 17977
+    },
+    {
+      "epoch": 1.0,
+      "step": 17977,
+      "total_flos": 1044585525231616.0,
+      "train_loss": 0.026658210229302048,
+      "train_runtime": 56588.3556,
+      "train_samples_per_second": 5.083,
+      "train_steps_per_second": 0.318
+    }
+  ],
+  "logging_steps": 1.0,
+  "max_steps": 17977,
+  "num_input_tokens_seen": 0,
+  "num_train_epochs": 1,
+  "save_steps": 899,
+  "total_flos": 1044585525231616.0,
+  "train_batch_size": 2,
+  "trial_name": null,
+  "trial_params": null
+}