diff --git "a/trainer_state.json" "b/trainer_state.json"
new file mode 100644--- /dev/null
+++ "b/trainer_state.json"
@@ -0,0 +1,44072 @@
+{
+  "best_metric": null,
+  "best_model_checkpoint": null,
+  "epoch": 0.350058407965734,
+  "eval_steps": 500,
+  "global_step": 6293,
+  "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
+    }
+  ],
+  "logging_steps": 1.0,
+  "max_steps": 17977,
+  "num_input_tokens_seen": 0,
+  "num_train_epochs": 1,
+  "save_steps": 899,
+  "total_flos": 365718765027328.0,
+  "train_batch_size": 2,
+  "trial_name": null,
+  "trial_params": null
+}