diff --git "a/trainer_state.json" "b/trainer_state.json"
new file mode 100644--- /dev/null
+++ "b/trainer_state.json"
@@ -0,0 +1,52500 @@
+{
+  "best_metric": null,
+  "best_model_checkpoint": null,
+  "epoch": 1.0,
+  "eval_steps": 15.0,
+  "global_step": 7494,
+  "is_hyper_param_search": false,
+  "is_local_process_zero": true,
+  "is_world_process_zero": true,
+  "log_history": [
+    {
+      "epoch": 0.00013344008540165466,
+      "grad_norm": 4.080643084961509,
+      "learning_rate": 8.88888888888889e-08,
+      "loss": 1.8736,
+      "step": 1
+    },
+    {
+      "epoch": 0.0002668801708033093,
+      "grad_norm": 5.08292352366484,
+      "learning_rate": 1.777777777777778e-07,
+      "loss": 1.9167,
+      "step": 2
+    },
+    {
+      "epoch": 0.00040032025620496394,
+      "grad_norm": 4.314269479207965,
+      "learning_rate": 2.666666666666667e-07,
+      "loss": 1.868,
+      "step": 3
+    },
+    {
+      "epoch": 0.0005337603416066186,
+      "grad_norm": 4.5997627346122325,
+      "learning_rate": 3.555555555555556e-07,
+      "loss": 1.8812,
+      "step": 4
+    },
+    {
+      "epoch": 0.0006672004270082733,
+      "grad_norm": 4.366746923746356,
+      "learning_rate": 4.444444444444445e-07,
+      "loss": 1.8816,
+      "step": 5
+    },
+    {
+      "epoch": 0.0008006405124099279,
+      "grad_norm": 6.417272891063272,
+      "learning_rate": 5.333333333333335e-07,
+      "loss": 1.9164,
+      "step": 6
+    },
+    {
+      "epoch": 0.0009340805978115826,
+      "grad_norm": 4.276397654879367,
+      "learning_rate": 6.222222222222223e-07,
+      "loss": 1.8699,
+      "step": 7
+    },
+    {
+      "epoch": 0.0010675206832132373,
+      "grad_norm": 40.21908327926398,
+      "learning_rate": 7.111111111111112e-07,
+      "loss": 1.9019,
+      "step": 8
+    },
+    {
+      "epoch": 0.0012009607686148918,
+      "grad_norm": 4.191772462937108,
+      "learning_rate": 8.000000000000001e-07,
+      "loss": 1.8959,
+      "step": 9
+    },
+    {
+      "epoch": 0.0013344008540165466,
+      "grad_norm": 4.39886202719947,
+      "learning_rate": 8.88888888888889e-07,
+      "loss": 1.8733,
+      "step": 10
+    },
+    {
+      "epoch": 0.0014678409394182012,
+      "grad_norm": 3.872194087105484,
+      "learning_rate": 9.77777777777778e-07,
+      "loss": 1.8535,
+      "step": 11
+    },
+    {
+      "epoch": 0.0016012810248198558,
+      "grad_norm": 3.738315413208318,
+      "learning_rate": 1.066666666666667e-06,
+      "loss": 1.8182,
+      "step": 12
+    },
+    {
+      "epoch": 0.0017347211102215106,
+      "grad_norm": 5.806590893863183,
+      "learning_rate": 1.1555555555555556e-06,
+      "loss": 1.841,
+      "step": 13
+    },
+    {
+      "epoch": 0.0018681611956231651,
+      "grad_norm": 3.9977833987634748,
+      "learning_rate": 1.2444444444444445e-06,
+      "loss": 1.8551,
+      "step": 14
+    },
+    {
+      "epoch": 0.0020016012810248197,
+      "grad_norm": 3.6521269631944073,
+      "learning_rate": 1.3333333333333334e-06,
+      "loss": 1.8873,
+      "step": 15
+    },
+    {
+      "epoch": 0.0021350413664264745,
+      "grad_norm": 2.9986950145773097,
+      "learning_rate": 1.4222222222222223e-06,
+      "loss": 1.8,
+      "step": 16
+    },
+    {
+      "epoch": 0.0022684814518281293,
+      "grad_norm": 3.1157009896387544,
+      "learning_rate": 1.5111111111111112e-06,
+      "loss": 1.8309,
+      "step": 17
+    },
+    {
+      "epoch": 0.0024019215372297837,
+      "grad_norm": 3.7767942592540855,
+      "learning_rate": 1.6000000000000001e-06,
+      "loss": 1.8749,
+      "step": 18
+    },
+    {
+      "epoch": 0.0025353616226314385,
+      "grad_norm": 3.480294979458449,
+      "learning_rate": 1.688888888888889e-06,
+      "loss": 1.8126,
+      "step": 19
+    },
+    {
+      "epoch": 0.0026688017080330933,
+      "grad_norm": 2.9012896594189366,
+      "learning_rate": 1.777777777777778e-06,
+      "loss": 1.7983,
+      "step": 20
+    },
+    {
+      "epoch": 0.0028022417934347476,
+      "grad_norm": 4.745562235186556,
+      "learning_rate": 1.8666666666666669e-06,
+      "loss": 1.7862,
+      "step": 21
+    },
+    {
+      "epoch": 0.0029356818788364024,
+      "grad_norm": 4.88191672119221,
+      "learning_rate": 1.955555555555556e-06,
+      "loss": 1.7786,
+      "step": 22
+    },
+    {
+      "epoch": 0.003069121964238057,
+      "grad_norm": 41.74795779070356,
+      "learning_rate": 2.0444444444444447e-06,
+      "loss": 1.7671,
+      "step": 23
+    },
+    {
+      "epoch": 0.0032025620496397116,
+      "grad_norm": 3.02827351605321,
+      "learning_rate": 2.133333333333334e-06,
+      "loss": 1.821,
+      "step": 24
+    },
+    {
+      "epoch": 0.0033360021350413663,
+      "grad_norm": 2.7722420080290804,
+      "learning_rate": 2.222222222222222e-06,
+      "loss": 1.7749,
+      "step": 25
+    },
+    {
+      "epoch": 0.003469442220443021,
+      "grad_norm": 3.669445898983361,
+      "learning_rate": 2.311111111111111e-06,
+      "loss": 1.8066,
+      "step": 26
+    },
+    {
+      "epoch": 0.003602882305844676,
+      "grad_norm": 2.928887643981388,
+      "learning_rate": 2.4000000000000003e-06,
+      "loss": 1.7647,
+      "step": 27
+    },
+    {
+      "epoch": 0.0037363223912463303,
+      "grad_norm": 3.7543465480647957,
+      "learning_rate": 2.488888888888889e-06,
+      "loss": 1.8005,
+      "step": 28
+    },
+    {
+      "epoch": 0.003869762476647985,
+      "grad_norm": 3.251679830837901,
+      "learning_rate": 2.577777777777778e-06,
+      "loss": 1.7779,
+      "step": 29
+    },
+    {
+      "epoch": 0.0040032025620496394,
+      "grad_norm": 21.468244588308337,
+      "learning_rate": 2.666666666666667e-06,
+      "loss": 1.7694,
+      "step": 30
+    },
+    {
+      "epoch": 0.004136642647451294,
+      "grad_norm": 4.000265636508565,
+      "learning_rate": 2.755555555555556e-06,
+      "loss": 1.7384,
+      "step": 31
+    },
+    {
+      "epoch": 0.004270082732852949,
+      "grad_norm": 3.0634285680156577,
+      "learning_rate": 2.8444444444444446e-06,
+      "loss": 1.7271,
+      "step": 32
+    },
+    {
+      "epoch": 0.004403522818254604,
+      "grad_norm": 2.406246700648424,
+      "learning_rate": 2.9333333333333338e-06,
+      "loss": 1.7338,
+      "step": 33
+    },
+    {
+      "epoch": 0.004536962903656259,
+      "grad_norm": 2.542980268594405,
+      "learning_rate": 3.0222222222222225e-06,
+      "loss": 1.7457,
+      "step": 34
+    },
+    {
+      "epoch": 0.004670402989057913,
+      "grad_norm": 2.7984146843006226,
+      "learning_rate": 3.1111111111111116e-06,
+      "loss": 1.7539,
+      "step": 35
+    },
+    {
+      "epoch": 0.004803843074459567,
+      "grad_norm": 3.013085830269592,
+      "learning_rate": 3.2000000000000003e-06,
+      "loss": 1.6781,
+      "step": 36
+    },
+    {
+      "epoch": 0.004937283159861222,
+      "grad_norm": 2.153019671959159,
+      "learning_rate": 3.2888888888888894e-06,
+      "loss": 1.744,
+      "step": 37
+    },
+    {
+      "epoch": 0.005070723245262877,
+      "grad_norm": 2.419384656819244,
+      "learning_rate": 3.377777777777778e-06,
+      "loss": 1.7563,
+      "step": 38
+    },
+    {
+      "epoch": 0.005204163330664532,
+      "grad_norm": 2.431242306767295,
+      "learning_rate": 3.4666666666666672e-06,
+      "loss": 1.7474,
+      "step": 39
+    },
+    {
+      "epoch": 0.0053376034160661865,
+      "grad_norm": 2.58042762585277,
+      "learning_rate": 3.555555555555556e-06,
+      "loss": 1.7052,
+      "step": 40
+    },
+    {
+      "epoch": 0.005471043501467841,
+      "grad_norm": 2.903485309507233,
+      "learning_rate": 3.644444444444445e-06,
+      "loss": 1.7613,
+      "step": 41
+    },
+    {
+      "epoch": 0.005604483586869495,
+      "grad_norm": 3.1612531609911123,
+      "learning_rate": 3.7333333333333337e-06,
+      "loss": 1.7141,
+      "step": 42
+    },
+    {
+      "epoch": 0.00573792367227115,
+      "grad_norm": 2.060069555889511,
+      "learning_rate": 3.8222222222222224e-06,
+      "loss": 1.6964,
+      "step": 43
+    },
+    {
+      "epoch": 0.005871363757672805,
+      "grad_norm": 2.0784087729038885,
+      "learning_rate": 3.911111111111112e-06,
+      "loss": 1.7272,
+      "step": 44
+    },
+    {
+      "epoch": 0.00600480384307446,
+      "grad_norm": 2.046333038993166,
+      "learning_rate": 4.000000000000001e-06,
+      "loss": 1.7155,
+      "step": 45
+    },
+    {
+      "epoch": 0.006138243928476114,
+      "grad_norm": 4.060174091987868,
+      "learning_rate": 4.088888888888889e-06,
+      "loss": 1.7394,
+      "step": 46
+    },
+    {
+      "epoch": 0.006271684013877769,
+      "grad_norm": 2.194133473755285,
+      "learning_rate": 4.177777777777778e-06,
+      "loss": 1.7473,
+      "step": 47
+    },
+    {
+      "epoch": 0.006405124099279423,
+      "grad_norm": 2.086936433450861,
+      "learning_rate": 4.266666666666668e-06,
+      "loss": 1.7482,
+      "step": 48
+    },
+    {
+      "epoch": 0.006538564184681078,
+      "grad_norm": 2.055176025616284,
+      "learning_rate": 4.3555555555555555e-06,
+      "loss": 1.6989,
+      "step": 49
+    },
+    {
+      "epoch": 0.006672004270082733,
+      "grad_norm": 4.2705916173620935,
+      "learning_rate": 4.444444444444444e-06,
+      "loss": 1.7113,
+      "step": 50
+    },
+    {
+      "epoch": 0.0068054443554843875,
+      "grad_norm": 3.062440642397753,
+      "learning_rate": 4.533333333333334e-06,
+      "loss": 1.6994,
+      "step": 51
+    },
+    {
+      "epoch": 0.006938884440886042,
+      "grad_norm": 3.6513027631899724,
+      "learning_rate": 4.622222222222222e-06,
+      "loss": 1.7435,
+      "step": 52
+    },
+    {
+      "epoch": 0.007072324526287697,
+      "grad_norm": 4.339915391562868,
+      "learning_rate": 4.711111111111111e-06,
+      "loss": 1.7093,
+      "step": 53
+    },
+    {
+      "epoch": 0.007205764611689352,
+      "grad_norm": 2.300952067433821,
+      "learning_rate": 4.800000000000001e-06,
+      "loss": 1.6687,
+      "step": 54
+    },
+    {
+      "epoch": 0.007339204697091006,
+      "grad_norm": 2.3827551772304814,
+      "learning_rate": 4.888888888888889e-06,
+      "loss": 1.748,
+      "step": 55
+    },
+    {
+      "epoch": 0.007472644782492661,
+      "grad_norm": 1.9554846454820922,
+      "learning_rate": 4.977777777777778e-06,
+      "loss": 1.6707,
+      "step": 56
+    },
+    {
+      "epoch": 0.007606084867894315,
+      "grad_norm": 2.208610815386933,
+      "learning_rate": 5.0666666666666676e-06,
+      "loss": 1.6908,
+      "step": 57
+    },
+    {
+      "epoch": 0.00773952495329597,
+      "grad_norm": 2.431678252426931,
+      "learning_rate": 5.155555555555556e-06,
+      "loss": 1.7041,
+      "step": 58
+    },
+    {
+      "epoch": 0.007872965038697624,
+      "grad_norm": 1.926978176003599,
+      "learning_rate": 5.244444444444445e-06,
+      "loss": 1.6928,
+      "step": 59
+    },
+    {
+      "epoch": 0.008006405124099279,
+      "grad_norm": 2.75648074491875,
+      "learning_rate": 5.333333333333334e-06,
+      "loss": 1.7009,
+      "step": 60
+    },
+    {
+      "epoch": 0.008139845209500934,
+      "grad_norm": 1.9054991774370111,
+      "learning_rate": 5.422222222222223e-06,
+      "loss": 1.6937,
+      "step": 61
+    },
+    {
+      "epoch": 0.008273285294902588,
+      "grad_norm": 2.7419834376889174,
+      "learning_rate": 5.511111111111112e-06,
+      "loss": 1.7218,
+      "step": 62
+    },
+    {
+      "epoch": 0.008406725380304243,
+      "grad_norm": 1.8897989454015407,
+      "learning_rate": 5.600000000000001e-06,
+      "loss": 1.7069,
+      "step": 63
+    },
+    {
+      "epoch": 0.008540165465705898,
+      "grad_norm": 2.159740439666748,
+      "learning_rate": 5.688888888888889e-06,
+      "loss": 1.7143,
+      "step": 64
+    },
+    {
+      "epoch": 0.008673605551107553,
+      "grad_norm": 2.1199099802628396,
+      "learning_rate": 5.777777777777778e-06,
+      "loss": 1.6872,
+      "step": 65
+    },
+    {
+      "epoch": 0.008807045636509208,
+      "grad_norm": 2.275963855081617,
+      "learning_rate": 5.8666666666666675e-06,
+      "loss": 1.8008,
+      "step": 66
+    },
+    {
+      "epoch": 0.008940485721910862,
+      "grad_norm": 2.208827668777784,
+      "learning_rate": 5.955555555555555e-06,
+      "loss": 1.686,
+      "step": 67
+    },
+    {
+      "epoch": 0.009073925807312517,
+      "grad_norm": 3.117561539877217,
+      "learning_rate": 6.044444444444445e-06,
+      "loss": 1.6589,
+      "step": 68
+    },
+    {
+      "epoch": 0.009207365892714172,
+      "grad_norm": 2.3085392935053766,
+      "learning_rate": 6.133333333333334e-06,
+      "loss": 1.6787,
+      "step": 69
+    },
+    {
+      "epoch": 0.009340805978115827,
+      "grad_norm": 2.1582405446873554,
+      "learning_rate": 6.222222222222223e-06,
+      "loss": 1.7265,
+      "step": 70
+    },
+    {
+      "epoch": 0.00947424606351748,
+      "grad_norm": 2.0969226907995053,
+      "learning_rate": 6.311111111111111e-06,
+      "loss": 1.6737,
+      "step": 71
+    },
+    {
+      "epoch": 0.009607686148919135,
+      "grad_norm": 4.200722941416422,
+      "learning_rate": 6.4000000000000006e-06,
+      "loss": 1.7188,
+      "step": 72
+    },
+    {
+      "epoch": 0.00974112623432079,
+      "grad_norm": 2.2485101560203105,
+      "learning_rate": 6.488888888888889e-06,
+      "loss": 1.6912,
+      "step": 73
+    },
+    {
+      "epoch": 0.009874566319722444,
+      "grad_norm": 2.763908223216233,
+      "learning_rate": 6.577777777777779e-06,
+      "loss": 1.6571,
+      "step": 74
+    },
+    {
+      "epoch": 0.010008006405124099,
+      "grad_norm": 1.9040632627929543,
+      "learning_rate": 6.666666666666667e-06,
+      "loss": 1.6861,
+      "step": 75
+    },
+    {
+      "epoch": 0.010141446490525754,
+      "grad_norm": 28.864482563980474,
+      "learning_rate": 6.755555555555556e-06,
+      "loss": 1.7206,
+      "step": 76
+    },
+    {
+      "epoch": 0.010274886575927409,
+      "grad_norm": 2.8893679461430732,
+      "learning_rate": 6.844444444444445e-06,
+      "loss": 1.651,
+      "step": 77
+    },
+    {
+      "epoch": 0.010408326661329063,
+      "grad_norm": 2.6807260850078705,
+      "learning_rate": 6.9333333333333344e-06,
+      "loss": 1.7404,
+      "step": 78
+    },
+    {
+      "epoch": 0.010541766746730718,
+      "grad_norm": 2.616916412887755,
+      "learning_rate": 7.022222222222222e-06,
+      "loss": 1.7453,
+      "step": 79
+    },
+    {
+      "epoch": 0.010675206832132373,
+      "grad_norm": 2.030396506610222,
+      "learning_rate": 7.111111111111112e-06,
+      "loss": 1.6951,
+      "step": 80
+    },
+    {
+      "epoch": 0.010808646917534028,
+      "grad_norm": 2.276338612685183,
+      "learning_rate": 7.2000000000000005e-06,
+      "loss": 1.6665,
+      "step": 81
+    },
+    {
+      "epoch": 0.010942087002935683,
+      "grad_norm": 4.873556416423791,
+      "learning_rate": 7.28888888888889e-06,
+      "loss": 1.6902,
+      "step": 82
+    },
+    {
+      "epoch": 0.011075527088337337,
+      "grad_norm": 2.480438858137466,
+      "learning_rate": 7.377777777777778e-06,
+      "loss": 1.6992,
+      "step": 83
+    },
+    {
+      "epoch": 0.01120896717373899,
+      "grad_norm": 2.0278767294473234,
+      "learning_rate": 7.4666666666666675e-06,
+      "loss": 1.6769,
+      "step": 84
+    },
+    {
+      "epoch": 0.011342407259140645,
+      "grad_norm": 2.7060597428983937,
+      "learning_rate": 7.555555555555556e-06,
+      "loss": 1.6538,
+      "step": 85
+    },
+    {
+      "epoch": 0.0114758473445423,
+      "grad_norm": 2.7423909405968963,
+      "learning_rate": 7.644444444444445e-06,
+      "loss": 1.7279,
+      "step": 86
+    },
+    {
+      "epoch": 0.011609287429943955,
+      "grad_norm": 2.2930045021509446,
+      "learning_rate": 7.733333333333334e-06,
+      "loss": 1.6586,
+      "step": 87
+    },
+    {
+      "epoch": 0.01174272751534561,
+      "grad_norm": 1.9855479277235186,
+      "learning_rate": 7.822222222222224e-06,
+      "loss": 1.7009,
+      "step": 88
+    },
+    {
+      "epoch": 0.011876167600747264,
+      "grad_norm": 2.7213236928751554,
+      "learning_rate": 7.911111111111112e-06,
+      "loss": 1.6854,
+      "step": 89
+    },
+    {
+      "epoch": 0.01200960768614892,
+      "grad_norm": 2.7306900665495757,
+      "learning_rate": 8.000000000000001e-06,
+      "loss": 1.7225,
+      "step": 90
+    },
+    {
+      "epoch": 0.012143047771550574,
+      "grad_norm": 2.107447235822766,
+      "learning_rate": 8.08888888888889e-06,
+      "loss": 1.7019,
+      "step": 91
+    },
+    {
+      "epoch": 0.012276487856952229,
+      "grad_norm": 2.570127439213928,
+      "learning_rate": 8.177777777777779e-06,
+      "loss": 1.6494,
+      "step": 92
+    },
+    {
+      "epoch": 0.012409927942353884,
+      "grad_norm": 1.8760712243264193,
+      "learning_rate": 8.266666666666667e-06,
+      "loss": 1.6933,
+      "step": 93
+    },
+    {
+      "epoch": 0.012543368027755538,
+      "grad_norm": 2.1290912423121253,
+      "learning_rate": 8.355555555555556e-06,
+      "loss": 1.6705,
+      "step": 94
+    },
+    {
+      "epoch": 0.012676808113157193,
+      "grad_norm": 2.488673234475764,
+      "learning_rate": 8.444444444444446e-06,
+      "loss": 1.7204,
+      "step": 95
+    },
+    {
+      "epoch": 0.012810248198558846,
+      "grad_norm": 2.1122071965215015,
+      "learning_rate": 8.533333333333335e-06,
+      "loss": 1.7021,
+      "step": 96
+    },
+    {
+      "epoch": 0.012943688283960501,
+      "grad_norm": 2.200704929120579,
+      "learning_rate": 8.622222222222223e-06,
+      "loss": 1.6436,
+      "step": 97
+    },
+    {
+      "epoch": 0.013077128369362156,
+      "grad_norm": 2.2007715935735606,
+      "learning_rate": 8.711111111111111e-06,
+      "loss": 1.6646,
+      "step": 98
+    },
+    {
+      "epoch": 0.01321056845476381,
+      "grad_norm": 2.2699598373044383,
+      "learning_rate": 8.8e-06,
+      "loss": 1.6495,
+      "step": 99
+    },
+    {
+      "epoch": 0.013344008540165465,
+      "grad_norm": 2.328713098955483,
+      "learning_rate": 8.888888888888888e-06,
+      "loss": 1.7009,
+      "step": 100
+    },
+    {
+      "epoch": 0.01347744862556712,
+      "grad_norm": 2.218972713645272,
+      "learning_rate": 8.977777777777778e-06,
+      "loss": 1.7082,
+      "step": 101
+    },
+    {
+      "epoch": 0.013610888710968775,
+      "grad_norm": 2.618934756244986,
+      "learning_rate": 9.066666666666667e-06,
+      "loss": 1.6985,
+      "step": 102
+    },
+    {
+      "epoch": 0.01374432879637043,
+      "grad_norm": 3.2314521781408074,
+      "learning_rate": 9.155555555555557e-06,
+      "loss": 1.7344,
+      "step": 103
+    },
+    {
+      "epoch": 0.013877768881772085,
+      "grad_norm": 3.697972156481625,
+      "learning_rate": 9.244444444444445e-06,
+      "loss": 1.683,
+      "step": 104
+    },
+    {
+      "epoch": 0.01401120896717374,
+      "grad_norm": 2.8531262360061698,
+      "learning_rate": 9.333333333333334e-06,
+      "loss": 1.6776,
+      "step": 105
+    },
+    {
+      "epoch": 0.014144649052575394,
+      "grad_norm": 39.794103633027675,
+      "learning_rate": 9.422222222222222e-06,
+      "loss": 1.7134,
+      "step": 106
+    },
+    {
+      "epoch": 0.014278089137977049,
+      "grad_norm": 2.9264673716827954,
+      "learning_rate": 9.511111111111112e-06,
+      "loss": 1.6754,
+      "step": 107
+    },
+    {
+      "epoch": 0.014411529223378704,
+      "grad_norm": 2.414594446024647,
+      "learning_rate": 9.600000000000001e-06,
+      "loss": 1.6917,
+      "step": 108
+    },
+    {
+      "epoch": 0.014544969308780357,
+      "grad_norm": 2.792320442984017,
+      "learning_rate": 9.688888888888889e-06,
+      "loss": 1.6453,
+      "step": 109
+    },
+    {
+      "epoch": 0.014678409394182012,
+      "grad_norm": 3.358788179969882,
+      "learning_rate": 9.777777777777779e-06,
+      "loss": 1.713,
+      "step": 110
+    },
+    {
+      "epoch": 0.014811849479583666,
+      "grad_norm": 2.044760856305306,
+      "learning_rate": 9.866666666666668e-06,
+      "loss": 1.6851,
+      "step": 111
+    },
+    {
+      "epoch": 0.014945289564985321,
+      "grad_norm": 2.4828242965545386,
+      "learning_rate": 9.955555555555556e-06,
+      "loss": 1.7159,
+      "step": 112
+    },
+    {
+      "epoch": 0.015078729650386976,
+      "grad_norm": 46.56955438722204,
+      "learning_rate": 1.0044444444444446e-05,
+      "loss": 1.7581,
+      "step": 113
+    },
+    {
+      "epoch": 0.01521216973578863,
+      "grad_norm": 4.705272454001238,
+      "learning_rate": 1.0133333333333335e-05,
+      "loss": 1.6542,
+      "step": 114
+    },
+    {
+      "epoch": 0.015345609821190286,
+      "grad_norm": 7.73608658858232,
+      "learning_rate": 1.0222222222222223e-05,
+      "loss": 1.6793,
+      "step": 115
+    },
+    {
+      "epoch": 0.01547904990659194,
+      "grad_norm": 5.229243538124936,
+      "learning_rate": 1.0311111111111113e-05,
+      "loss": 1.6866,
+      "step": 116
+    },
+    {
+      "epoch": 0.015612489991993595,
+      "grad_norm": 4.340274190286637,
+      "learning_rate": 1.04e-05,
+      "loss": 1.7025,
+      "step": 117
+    },
+    {
+      "epoch": 0.015745930077395248,
+      "grad_norm": 2.211270582628492,
+      "learning_rate": 1.048888888888889e-05,
+      "loss": 1.6903,
+      "step": 118
+    },
+    {
+      "epoch": 0.015879370162796905,
+      "grad_norm": 2.2975450110438334,
+      "learning_rate": 1.0577777777777778e-05,
+      "loss": 1.7088,
+      "step": 119
+    },
+    {
+      "epoch": 0.016012810248198558,
+      "grad_norm": 17.328579385026632,
+      "learning_rate": 1.0666666666666667e-05,
+      "loss": 1.7141,
+      "step": 120
+    },
+    {
+      "epoch": 0.016146250333600214,
+      "grad_norm": 3.877950326292574,
+      "learning_rate": 1.0755555555555557e-05,
+      "loss": 1.6759,
+      "step": 121
+    },
+    {
+      "epoch": 0.016279690419001867,
+      "grad_norm": 4.011366710334245,
+      "learning_rate": 1.0844444444444446e-05,
+      "loss": 1.699,
+      "step": 122
+    },
+    {
+      "epoch": 0.016413130504403524,
+      "grad_norm": 3.2602156984936417,
+      "learning_rate": 1.0933333333333334e-05,
+      "loss": 1.7003,
+      "step": 123
+    },
+    {
+      "epoch": 0.016546570589805177,
+      "grad_norm": 2.630659747532619,
+      "learning_rate": 1.1022222222222224e-05,
+      "loss": 1.6344,
+      "step": 124
+    },
+    {
+      "epoch": 0.016680010675206833,
+      "grad_norm": 2.458919675792434,
+      "learning_rate": 1.1111111111111113e-05,
+      "loss": 1.6945,
+      "step": 125
+    },
+    {
+      "epoch": 0.016813450760608487,
+      "grad_norm": 3.984975390451759,
+      "learning_rate": 1.1200000000000001e-05,
+      "loss": 1.7746,
+      "step": 126
+    },
+    {
+      "epoch": 0.016946890846010143,
+      "grad_norm": 3.8315471470735787,
+      "learning_rate": 1.1288888888888889e-05,
+      "loss": 1.6974,
+      "step": 127
+    },
+    {
+      "epoch": 0.017080330931411796,
+      "grad_norm": 2.499188895634502,
+      "learning_rate": 1.1377777777777779e-05,
+      "loss": 1.6892,
+      "step": 128
+    },
+    {
+      "epoch": 0.01721377101681345,
+      "grad_norm": 2.0452097919611005,
+      "learning_rate": 1.1466666666666668e-05,
+      "loss": 1.7397,
+      "step": 129
+    },
+    {
+      "epoch": 0.017347211102215106,
+      "grad_norm": 2.298966367562689,
+      "learning_rate": 1.1555555555555556e-05,
+      "loss": 1.6836,
+      "step": 130
+    },
+    {
+      "epoch": 0.01748065118761676,
+      "grad_norm": 2.5420025536951942,
+      "learning_rate": 1.1644444444444446e-05,
+      "loss": 1.7454,
+      "step": 131
+    },
+    {
+      "epoch": 0.017614091273018415,
+      "grad_norm": 13.604138953612503,
+      "learning_rate": 1.1733333333333335e-05,
+      "loss": 1.7288,
+      "step": 132
+    },
+    {
+      "epoch": 0.01774753135842007,
+      "grad_norm": 2.3801527003336487,
+      "learning_rate": 1.1822222222222225e-05,
+      "loss": 1.6876,
+      "step": 133
+    },
+    {
+      "epoch": 0.017880971443821725,
+      "grad_norm": 2.460244538839727,
+      "learning_rate": 1.191111111111111e-05,
+      "loss": 1.7061,
+      "step": 134
+    },
+    {
+      "epoch": 0.018014411529223378,
+      "grad_norm": 2.4874810432246095,
+      "learning_rate": 1.2e-05,
+      "loss": 1.6963,
+      "step": 135
+    },
+    {
+      "epoch": 0.018147851614625034,
+      "grad_norm": 2.409965122688259,
+      "learning_rate": 1.208888888888889e-05,
+      "loss": 1.7093,
+      "step": 136
+    },
+    {
+      "epoch": 0.018281291700026688,
+      "grad_norm": 2.5548220927651393,
+      "learning_rate": 1.217777777777778e-05,
+      "loss": 1.6531,
+      "step": 137
+    },
+    {
+      "epoch": 0.018414731785428344,
+      "grad_norm": 2.7637181173708676,
+      "learning_rate": 1.2266666666666667e-05,
+      "loss": 1.6484,
+      "step": 138
+    },
+    {
+      "epoch": 0.018548171870829997,
+      "grad_norm": 2.108441938061504,
+      "learning_rate": 1.2355555555555557e-05,
+      "loss": 1.6831,
+      "step": 139
+    },
+    {
+      "epoch": 0.018681611956231654,
+      "grad_norm": 1.843873869065762,
+      "learning_rate": 1.2444444444444446e-05,
+      "loss": 1.7209,
+      "step": 140
+    },
+    {
+      "epoch": 0.018815052041633307,
+      "grad_norm": 1.852059082824261,
+      "learning_rate": 1.2533333333333336e-05,
+      "loss": 1.6356,
+      "step": 141
+    },
+    {
+      "epoch": 0.01894849212703496,
+      "grad_norm": 1.8903070568688063,
+      "learning_rate": 1.2622222222222222e-05,
+      "loss": 1.666,
+      "step": 142
+    },
+    {
+      "epoch": 0.019081932212436616,
+      "grad_norm": 2.7546690000389145,
+      "learning_rate": 1.2711111111111112e-05,
+      "loss": 1.6859,
+      "step": 143
+    },
+    {
+      "epoch": 0.01921537229783827,
+      "grad_norm": 1.9120514050425437,
+      "learning_rate": 1.2800000000000001e-05,
+      "loss": 1.616,
+      "step": 144
+    },
+    {
+      "epoch": 0.019348812383239926,
+      "grad_norm": 1.9059883225343819,
+      "learning_rate": 1.288888888888889e-05,
+      "loss": 1.7452,
+      "step": 145
+    },
+    {
+      "epoch": 0.01948225246864158,
+      "grad_norm": 1.8505183157858194,
+      "learning_rate": 1.2977777777777779e-05,
+      "loss": 1.6603,
+      "step": 146
+    },
+    {
+      "epoch": 0.019615692554043235,
+      "grad_norm": 3.2853784344984027,
+      "learning_rate": 1.3066666666666668e-05,
+      "loss": 1.6544,
+      "step": 147
+    },
+    {
+      "epoch": 0.01974913263944489,
+      "grad_norm": 2.09698441082926,
+      "learning_rate": 1.3155555555555558e-05,
+      "loss": 1.725,
+      "step": 148
+    },
+    {
+      "epoch": 0.019882572724846545,
+      "grad_norm": 2.244317832313048,
+      "learning_rate": 1.3244444444444447e-05,
+      "loss": 1.6843,
+      "step": 149
+    },
+    {
+      "epoch": 0.020016012810248198,
+      "grad_norm": 1.906767252092642,
+      "learning_rate": 1.3333333333333333e-05,
+      "loss": 1.6561,
+      "step": 150
+    },
+    {
+      "epoch": 0.020149452895649855,
+      "grad_norm": 13.009656729569588,
+      "learning_rate": 1.3422222222222223e-05,
+      "loss": 1.717,
+      "step": 151
+    },
+    {
+      "epoch": 0.020282892981051508,
+      "grad_norm": 1.939458338988826,
+      "learning_rate": 1.3511111111111112e-05,
+      "loss": 1.6941,
+      "step": 152
+    },
+    {
+      "epoch": 0.020416333066453164,
+      "grad_norm": 1.9081248684144936,
+      "learning_rate": 1.3600000000000002e-05,
+      "loss": 1.7335,
+      "step": 153
+    },
+    {
+      "epoch": 0.020549773151854817,
+      "grad_norm": 3.46048722855863,
+      "learning_rate": 1.368888888888889e-05,
+      "loss": 1.6904,
+      "step": 154
+    },
+    {
+      "epoch": 0.02068321323725647,
+      "grad_norm": 3.2150826776743093,
+      "learning_rate": 1.377777777777778e-05,
+      "loss": 1.7326,
+      "step": 155
+    },
+    {
+      "epoch": 0.020816653322658127,
+      "grad_norm": 2.682239347552553,
+      "learning_rate": 1.3866666666666669e-05,
+      "loss": 1.6991,
+      "step": 156
+    },
+    {
+      "epoch": 0.02095009340805978,
+      "grad_norm": 2.9892152568563337,
+      "learning_rate": 1.3955555555555558e-05,
+      "loss": 1.6674,
+      "step": 157
+    },
+    {
+      "epoch": 0.021083533493461436,
+      "grad_norm": 1.9370449545076165,
+      "learning_rate": 1.4044444444444445e-05,
+      "loss": 1.7093,
+      "step": 158
+    },
+    {
+      "epoch": 0.02121697357886309,
+      "grad_norm": 2.8348396699033085,
+      "learning_rate": 1.4133333333333334e-05,
+      "loss": 1.6561,
+      "step": 159
+    },
+    {
+      "epoch": 0.021350413664264746,
+      "grad_norm": 5.362105984011789,
+      "learning_rate": 1.4222222222222224e-05,
+      "loss": 1.7225,
+      "step": 160
+    },
+    {
+      "epoch": 0.0214838537496664,
+      "grad_norm": 2.2489346309856972,
+      "learning_rate": 1.4311111111111111e-05,
+      "loss": 1.7205,
+      "step": 161
+    },
+    {
+      "epoch": 0.021617293835068056,
+      "grad_norm": 2.5792958960212813,
+      "learning_rate": 1.4400000000000001e-05,
+      "loss": 1.7106,
+      "step": 162
+    },
+    {
+      "epoch": 0.02175073392046971,
+      "grad_norm": 2.799373167012452,
+      "learning_rate": 1.448888888888889e-05,
+      "loss": 1.7011,
+      "step": 163
+    },
+    {
+      "epoch": 0.021884174005871365,
+      "grad_norm": 2.7270269575589587,
+      "learning_rate": 1.457777777777778e-05,
+      "loss": 1.7108,
+      "step": 164
+    },
+    {
+      "epoch": 0.022017614091273018,
+      "grad_norm": 2.14261161418117,
+      "learning_rate": 1.4666666666666666e-05,
+      "loss": 1.6779,
+      "step": 165
+    },
+    {
+      "epoch": 0.022151054176674675,
+      "grad_norm": 1.9467025797614437,
+      "learning_rate": 1.4755555555555556e-05,
+      "loss": 1.7194,
+      "step": 166
+    },
+    {
+      "epoch": 0.022284494262076328,
+      "grad_norm": 2.0951052041172034,
+      "learning_rate": 1.4844444444444445e-05,
+      "loss": 1.7088,
+      "step": 167
+    },
+    {
+      "epoch": 0.02241793434747798,
+      "grad_norm": 1.792897219038739,
+      "learning_rate": 1.4933333333333335e-05,
+      "loss": 1.6647,
+      "step": 168
+    },
+    {
+      "epoch": 0.022551374432879637,
+      "grad_norm": 2.057396723030578,
+      "learning_rate": 1.5022222222222223e-05,
+      "loss": 1.6376,
+      "step": 169
+    },
+    {
+      "epoch": 0.02268481451828129,
+      "grad_norm": 1.7777783661568858,
+      "learning_rate": 1.5111111111111112e-05,
+      "loss": 1.7434,
+      "step": 170
+    },
+    {
+      "epoch": 0.022818254603682947,
+      "grad_norm": 2.583723235252364,
+      "learning_rate": 1.5200000000000002e-05,
+      "loss": 1.6701,
+      "step": 171
+    },
+    {
+      "epoch": 0.0229516946890846,
+      "grad_norm": 2.708960473066722,
+      "learning_rate": 1.528888888888889e-05,
+      "loss": 1.7109,
+      "step": 172
+    },
+    {
+      "epoch": 0.023085134774486257,
+      "grad_norm": 2.179043118704419,
+      "learning_rate": 1.537777777777778e-05,
+      "loss": 1.6601,
+      "step": 173
+    },
+    {
+      "epoch": 0.02321857485988791,
+      "grad_norm": 30.22918083463127,
+      "learning_rate": 1.546666666666667e-05,
+      "loss": 1.8007,
+      "step": 174
+    },
+    {
+      "epoch": 0.023352014945289566,
+      "grad_norm": 2.230452217009043,
+      "learning_rate": 1.555555555555556e-05,
+      "loss": 1.7362,
+      "step": 175
+    },
+    {
+      "epoch": 0.02348545503069122,
+      "grad_norm": 2.9922017986999276,
+      "learning_rate": 1.5644444444444448e-05,
+      "loss": 1.7398,
+      "step": 176
+    },
+    {
+      "epoch": 0.023618895116092876,
+      "grad_norm": 3.914186394323879,
+      "learning_rate": 1.5733333333333334e-05,
+      "loss": 1.7256,
+      "step": 177
+    },
+    {
+      "epoch": 0.02375233520149453,
+      "grad_norm": 1.9015888556101328,
+      "learning_rate": 1.5822222222222224e-05,
+      "loss": 1.6762,
+      "step": 178
+    },
+    {
+      "epoch": 0.023885775286896185,
+      "grad_norm": 1.7266594765283543,
+      "learning_rate": 1.5911111111111113e-05,
+      "loss": 1.672,
+      "step": 179
+    },
+    {
+      "epoch": 0.02401921537229784,
+      "grad_norm": 2.070151343179939,
+      "learning_rate": 1.6000000000000003e-05,
+      "loss": 1.7933,
+      "step": 180
+    },
+    {
+      "epoch": 0.02415265545769949,
+      "grad_norm": 2.51353756811448,
+      "learning_rate": 1.608888888888889e-05,
+      "loss": 1.6875,
+      "step": 181
+    },
+    {
+      "epoch": 0.024286095543101148,
+      "grad_norm": 11.935615730367939,
+      "learning_rate": 1.617777777777778e-05,
+      "loss": 1.7512,
+      "step": 182
+    },
+    {
+      "epoch": 0.0244195356285028,
+      "grad_norm": 3.9274893534399635,
+      "learning_rate": 1.6266666666666668e-05,
+      "loss": 1.6877,
+      "step": 183
+    },
+    {
+      "epoch": 0.024552975713904458,
+      "grad_norm": 2.8888619941000377,
+      "learning_rate": 1.6355555555555557e-05,
+      "loss": 1.6988,
+      "step": 184
+    },
+    {
+      "epoch": 0.02468641579930611,
+      "grad_norm": 2.335255800684397,
+      "learning_rate": 1.6444444444444444e-05,
+      "loss": 1.7197,
+      "step": 185
+    },
+    {
+      "epoch": 0.024819855884707767,
+      "grad_norm": 2.132564305791294,
+      "learning_rate": 1.6533333333333333e-05,
+      "loss": 1.7189,
+      "step": 186
+    },
+    {
+      "epoch": 0.02495329597010942,
+      "grad_norm": 2.344308980206617,
+      "learning_rate": 1.6622222222222223e-05,
+      "loss": 1.6521,
+      "step": 187
+    },
+    {
+      "epoch": 0.025086736055511077,
+      "grad_norm": 1.7158444729386582,
+      "learning_rate": 1.6711111111111112e-05,
+      "loss": 1.7004,
+      "step": 188
+    },
+    {
+      "epoch": 0.02522017614091273,
+      "grad_norm": 3.0317644581837344,
+      "learning_rate": 1.6800000000000002e-05,
+      "loss": 1.7271,
+      "step": 189
+    },
+    {
+      "epoch": 0.025353616226314386,
+      "grad_norm": 1.9353267422280926,
+      "learning_rate": 1.688888888888889e-05,
+      "loss": 1.6795,
+      "step": 190
+    },
+    {
+      "epoch": 0.02548705631171604,
+      "grad_norm": 1.7304826902324242,
+      "learning_rate": 1.697777777777778e-05,
+      "loss": 1.6588,
+      "step": 191
+    },
+    {
+      "epoch": 0.025620496397117692,
+      "grad_norm": 1.9159549442018844,
+      "learning_rate": 1.706666666666667e-05,
+      "loss": 1.7432,
+      "step": 192
+    },
+    {
+      "epoch": 0.02575393648251935,
+      "grad_norm": 1.8910667470096478,
+      "learning_rate": 1.7155555555555557e-05,
+      "loss": 1.7513,
+      "step": 193
+    },
+    {
+      "epoch": 0.025887376567921002,
+      "grad_norm": 2.569829238717817,
+      "learning_rate": 1.7244444444444446e-05,
+      "loss": 1.6466,
+      "step": 194
+    },
+    {
+      "epoch": 0.02602081665332266,
+      "grad_norm": 2.078018415198727,
+      "learning_rate": 1.7333333333333336e-05,
+      "loss": 1.7358,
+      "step": 195
+    },
+    {
+      "epoch": 0.02615425673872431,
+      "grad_norm": 1.8817118359807499,
+      "learning_rate": 1.7422222222222222e-05,
+      "loss": 1.7516,
+      "step": 196
+    },
+    {
+      "epoch": 0.026287696824125968,
+      "grad_norm": 1.77364702683218,
+      "learning_rate": 1.751111111111111e-05,
+      "loss": 1.689,
+      "step": 197
+    },
+    {
+      "epoch": 0.02642113690952762,
+      "grad_norm": 2.2161414310718466,
+      "learning_rate": 1.76e-05,
+      "loss": 1.7242,
+      "step": 198
+    },
+    {
+      "epoch": 0.026554576994929278,
+      "grad_norm": 2.0643613848156392,
+      "learning_rate": 1.768888888888889e-05,
+      "loss": 1.6685,
+      "step": 199
+    },
+    {
+      "epoch": 0.02668801708033093,
+      "grad_norm": 14.87718153231451,
+      "learning_rate": 1.7777777777777777e-05,
+      "loss": 1.6533,
+      "step": 200
+    },
+    {
+      "epoch": 0.026821457165732587,
+      "grad_norm": 2.272950745857619,
+      "learning_rate": 1.7866666666666666e-05,
+      "loss": 1.7366,
+      "step": 201
+    },
+    {
+      "epoch": 0.02695489725113424,
+      "grad_norm": 2.8727770433260242,
+      "learning_rate": 1.7955555555555556e-05,
+      "loss": 1.7315,
+      "step": 202
+    },
+    {
+      "epoch": 0.027088337336535897,
+      "grad_norm": 2.3209623388005958,
+      "learning_rate": 1.8044444444444445e-05,
+      "loss": 1.7321,
+      "step": 203
+    },
+    {
+      "epoch": 0.02722177742193755,
+      "grad_norm": 1.897431656822105,
+      "learning_rate": 1.8133333333333335e-05,
+      "loss": 1.7697,
+      "step": 204
+    },
+    {
+      "epoch": 0.027355217507339203,
+      "grad_norm": 2.083186577685706,
+      "learning_rate": 1.8222222222222224e-05,
+      "loss": 1.7103,
+      "step": 205
+    },
+    {
+      "epoch": 0.02748865759274086,
+      "grad_norm": 2.4252062792709395,
+      "learning_rate": 1.8311111111111114e-05,
+      "loss": 1.7401,
+      "step": 206
+    },
+    {
+      "epoch": 0.027622097678142513,
+      "grad_norm": 5.5398864573778726,
+      "learning_rate": 1.8400000000000003e-05,
+      "loss": 1.7618,
+      "step": 207
+    },
+    {
+      "epoch": 0.02775553776354417,
+      "grad_norm": 2.12561692151565,
+      "learning_rate": 1.848888888888889e-05,
+      "loss": 1.698,
+      "step": 208
+    },
+    {
+      "epoch": 0.027888977848945822,
+      "grad_norm": 2.122353896017312,
+      "learning_rate": 1.857777777777778e-05,
+      "loss": 1.7519,
+      "step": 209
+    },
+    {
+      "epoch": 0.02802241793434748,
+      "grad_norm": 1.946448922291358,
+      "learning_rate": 1.866666666666667e-05,
+      "loss": 1.6719,
+      "step": 210
+    },
+    {
+      "epoch": 0.028155858019749132,
+      "grad_norm": 2.4301306367972813,
+      "learning_rate": 1.8755555555555558e-05,
+      "loss": 1.7039,
+      "step": 211
+    },
+    {
+      "epoch": 0.02828929810515079,
+      "grad_norm": 1.841018211290893,
+      "learning_rate": 1.8844444444444444e-05,
+      "loss": 1.7406,
+      "step": 212
+    },
+    {
+      "epoch": 0.02842273819055244,
+      "grad_norm": 1.9703359299596948,
+      "learning_rate": 1.8933333333333334e-05,
+      "loss": 1.6828,
+      "step": 213
+    },
+    {
+      "epoch": 0.028556178275954098,
+      "grad_norm": 2.0318091145704646,
+      "learning_rate": 1.9022222222222223e-05,
+      "loss": 1.6784,
+      "step": 214
+    },
+    {
+      "epoch": 0.02868961836135575,
+      "grad_norm": 1.8820707489427722,
+      "learning_rate": 1.9111111111111113e-05,
+      "loss": 1.7024,
+      "step": 215
+    },
+    {
+      "epoch": 0.028823058446757407,
+      "grad_norm": 2.4241888613922145,
+      "learning_rate": 1.9200000000000003e-05,
+      "loss": 1.685,
+      "step": 216
+    },
+    {
+      "epoch": 0.02895649853215906,
+      "grad_norm": 2.2092388050397456,
+      "learning_rate": 1.928888888888889e-05,
+      "loss": 1.7553,
+      "step": 217
+    },
+    {
+      "epoch": 0.029089938617560714,
+      "grad_norm": 2.165125467884484,
+      "learning_rate": 1.9377777777777778e-05,
+      "loss": 1.6919,
+      "step": 218
+    },
+    {
+      "epoch": 0.02922337870296237,
+      "grad_norm": 2.4113038080825624,
+      "learning_rate": 1.9466666666666668e-05,
+      "loss": 1.6618,
+      "step": 219
+    },
+    {
+      "epoch": 0.029356818788364023,
+      "grad_norm": 2.3316936837660576,
+      "learning_rate": 1.9555555555555557e-05,
+      "loss": 1.7472,
+      "step": 220
+    },
+    {
+      "epoch": 0.02949025887376568,
+      "grad_norm": 3.2727415100596855,
+      "learning_rate": 1.9644444444444447e-05,
+      "loss": 1.6679,
+      "step": 221
+    },
+    {
+      "epoch": 0.029623698959167333,
+      "grad_norm": 2.0492799095337784,
+      "learning_rate": 1.9733333333333336e-05,
+      "loss": 1.7506,
+      "step": 222
+    },
+    {
+      "epoch": 0.02975713904456899,
+      "grad_norm": 2.576771291880995,
+      "learning_rate": 1.9822222222222226e-05,
+      "loss": 1.737,
+      "step": 223
+    },
+    {
+      "epoch": 0.029890579129970642,
+      "grad_norm": 3.2741913701257257,
+      "learning_rate": 1.9911111111111112e-05,
+      "loss": 1.7066,
+      "step": 224
+    },
+    {
+      "epoch": 0.0300240192153723,
+      "grad_norm": 2.001886242005137,
+      "learning_rate": 2e-05,
+      "loss": 1.7444,
+      "step": 225
+    },
+    {
+      "epoch": 0.030157459300773952,
+      "grad_norm": 2.6698368689644334,
+      "learning_rate": 1.9999999066056927e-05,
+      "loss": 1.6765,
+      "step": 226
+    },
+    {
+      "epoch": 0.03029089938617561,
+      "grad_norm": 3.3466160854176628,
+      "learning_rate": 1.9999996264227883e-05,
+      "loss": 1.7286,
+      "step": 227
+    },
+    {
+      "epoch": 0.03042433947157726,
+      "grad_norm": 2.393225998646352,
+      "learning_rate": 1.999999159451339e-05,
+      "loss": 1.7539,
+      "step": 228
+    },
+    {
+      "epoch": 0.030557779556978918,
+      "grad_norm": 2.5849019495304777,
+      "learning_rate": 1.9999985056914326e-05,
+      "loss": 1.7202,
+      "step": 229
+    },
+    {
+      "epoch": 0.03069121964238057,
+      "grad_norm": 2.761192723900784,
+      "learning_rate": 1.9999976651431904e-05,
+      "loss": 1.7445,
+      "step": 230
+    },
+    {
+      "epoch": 0.030824659727782224,
+      "grad_norm": 2.440029715237141,
+      "learning_rate": 1.9999966378067696e-05,
+      "loss": 1.7564,
+      "step": 231
+    },
+    {
+      "epoch": 0.03095809981318388,
+      "grad_norm": 2.0847710036036307,
+      "learning_rate": 1.9999954236823625e-05,
+      "loss": 1.7053,
+      "step": 232
+    },
+    {
+      "epoch": 0.031091539898585534,
+      "grad_norm": 2.097727759259449,
+      "learning_rate": 1.9999940227701952e-05,
+      "loss": 1.7045,
+      "step": 233
+    },
+    {
+      "epoch": 0.03122497998398719,
+      "grad_norm": 15.81909104706428,
+      "learning_rate": 1.99999243507053e-05,
+      "loss": 1.8102,
+      "step": 234
+    },
+    {
+      "epoch": 0.03135842006938885,
+      "grad_norm": 2.223134069922936,
+      "learning_rate": 1.9999906605836637e-05,
+      "loss": 1.738,
+      "step": 235
+    },
+    {
+      "epoch": 0.031491860154790496,
+      "grad_norm": 2.3651382675222528,
+      "learning_rate": 1.9999886993099273e-05,
+      "loss": 1.7467,
+      "step": 236
+    },
+    {
+      "epoch": 0.03162530024019215,
+      "grad_norm": 1.8263105543794043,
+      "learning_rate": 1.9999865512496867e-05,
+      "loss": 1.7548,
+      "step": 237
+    },
+    {
+      "epoch": 0.03175874032559381,
+      "grad_norm": 3.9716970412217805,
+      "learning_rate": 1.9999842164033435e-05,
+      "loss": 1.7206,
+      "step": 238
+    },
+    {
+      "epoch": 0.031892180410995466,
+      "grad_norm": 2.0120218395868634,
+      "learning_rate": 1.9999816947713345e-05,
+      "loss": 1.7401,
+      "step": 239
+    },
+    {
+      "epoch": 0.032025620496397116,
+      "grad_norm": 2.3384151798445036,
+      "learning_rate": 1.9999789863541295e-05,
+      "loss": 1.7301,
+      "step": 240
+    },
+    {
+      "epoch": 0.03215906058179877,
+      "grad_norm": 3.940580388558925,
+      "learning_rate": 1.9999760911522355e-05,
+      "loss": 1.6965,
+      "step": 241
+    },
+    {
+      "epoch": 0.03229250066720043,
+      "grad_norm": 1.8224155291794617,
+      "learning_rate": 1.9999730091661928e-05,
+      "loss": 1.6705,
+      "step": 242
+    },
+    {
+      "epoch": 0.03242594075260208,
+      "grad_norm": 1.9045524865773107,
+      "learning_rate": 1.999969740396577e-05,
+      "loss": 1.7543,
+      "step": 243
+    },
+    {
+      "epoch": 0.032559380838003735,
+      "grad_norm": 1.6984600198344175,
+      "learning_rate": 1.9999662848439988e-05,
+      "loss": 1.7126,
+      "step": 244
+    },
+    {
+      "epoch": 0.03269282092340539,
+      "grad_norm": 1.9269958388703856,
+      "learning_rate": 1.9999626425091035e-05,
+      "loss": 1.7016,
+      "step": 245
+    },
+    {
+      "epoch": 0.03282626100880705,
+      "grad_norm": 2.4719605600430095,
+      "learning_rate": 1.9999588133925715e-05,
+      "loss": 1.733,
+      "step": 246
+    },
+    {
+      "epoch": 0.0329597010942087,
+      "grad_norm": 2.129823068185752,
+      "learning_rate": 1.9999547974951184e-05,
+      "loss": 1.6882,
+      "step": 247
+    },
+    {
+      "epoch": 0.033093141179610354,
+      "grad_norm": 2.0555832511145145,
+      "learning_rate": 1.9999505948174943e-05,
+      "loss": 1.7154,
+      "step": 248
+    },
+    {
+      "epoch": 0.03322658126501201,
+      "grad_norm": 1.8053396394120758,
+      "learning_rate": 1.9999462053604836e-05,
+      "loss": 1.6796,
+      "step": 249
+    },
+    {
+      "epoch": 0.03336002135041367,
+      "grad_norm": 5.575233121908451,
+      "learning_rate": 1.9999416291249064e-05,
+      "loss": 1.7884,
+      "step": 250
+    },
+    {
+      "epoch": 0.03349346143581532,
+      "grad_norm": 2.2350909040289673,
+      "learning_rate": 1.9999368661116177e-05,
+      "loss": 1.7342,
+      "step": 251
+    },
+    {
+      "epoch": 0.03362690152121697,
+      "grad_norm": 1.6497870642338468,
+      "learning_rate": 1.9999319163215075e-05,
+      "loss": 1.71,
+      "step": 252
+    },
+    {
+      "epoch": 0.03376034160661863,
+      "grad_norm": 1.993739573845127,
+      "learning_rate": 1.9999267797554995e-05,
+      "loss": 1.7162,
+      "step": 253
+    },
+    {
+      "epoch": 0.033893781692020286,
+      "grad_norm": 1.8123408280559632,
+      "learning_rate": 1.999921456414554e-05,
+      "loss": 1.7696,
+      "step": 254
+    },
+    {
+      "epoch": 0.034027221777421936,
+      "grad_norm": 1.6594530678698969,
+      "learning_rate": 1.9999159462996646e-05,
+      "loss": 1.6971,
+      "step": 255
+    },
+    {
+      "epoch": 0.03416066186282359,
+      "grad_norm": 1.7999533008081343,
+      "learning_rate": 1.9999102494118613e-05,
+      "loss": 1.7056,
+      "step": 256
+    },
+    {
+      "epoch": 0.03429410194822525,
+      "grad_norm": 1.5795266580620457,
+      "learning_rate": 1.999904365752208e-05,
+      "loss": 1.7091,
+      "step": 257
+    },
+    {
+      "epoch": 0.0344275420336269,
+      "grad_norm": 2.9298656896602355,
+      "learning_rate": 1.9998982953218028e-05,
+      "loss": 1.7193,
+      "step": 258
+    },
+    {
+      "epoch": 0.034560982119028555,
+      "grad_norm": 1.764460603536837,
+      "learning_rate": 1.9998920381217807e-05,
+      "loss": 1.7333,
+      "step": 259
+    },
+    {
+      "epoch": 0.03469442220443021,
+      "grad_norm": 2.3109901205414314,
+      "learning_rate": 1.99988559415331e-05,
+      "loss": 1.7506,
+      "step": 260
+    },
+    {
+      "epoch": 0.03482786228983187,
+      "grad_norm": 10.94371168313979,
+      "learning_rate": 1.9998789634175945e-05,
+      "loss": 1.7554,
+      "step": 261
+    },
+    {
+      "epoch": 0.03496130237523352,
+      "grad_norm": 1.8407021375815145,
+      "learning_rate": 1.999872145915873e-05,
+      "loss": 1.7242,
+      "step": 262
+    },
+    {
+      "epoch": 0.035094742460635174,
+      "grad_norm": 2.974143978244417,
+      "learning_rate": 1.999865141649418e-05,
+      "loss": 1.7209,
+      "step": 263
+    },
+    {
+      "epoch": 0.03522818254603683,
+      "grad_norm": 2.2722279313309475,
+      "learning_rate": 1.9998579506195384e-05,
+      "loss": 1.7143,
+      "step": 264
+    },
+    {
+      "epoch": 0.03536162263143849,
+      "grad_norm": 1.6990878341793165,
+      "learning_rate": 1.999850572827578e-05,
+      "loss": 1.7197,
+      "step": 265
+    },
+    {
+      "epoch": 0.03549506271684014,
+      "grad_norm": 1.6642794979330895,
+      "learning_rate": 1.999843008274914e-05,
+      "loss": 1.6542,
+      "step": 266
+    },
+    {
+      "epoch": 0.03562850280224179,
+      "grad_norm": 2.007102879639598,
+      "learning_rate": 1.9998352569629596e-05,
+      "loss": 1.7283,
+      "step": 267
+    },
+    {
+      "epoch": 0.03576194288764345,
+      "grad_norm": 1.603072629822543,
+      "learning_rate": 1.9998273188931628e-05,
+      "loss": 1.6958,
+      "step": 268
+    },
+    {
+      "epoch": 0.0358953829730451,
+      "grad_norm": 1.83795235651287,
+      "learning_rate": 1.9998191940670068e-05,
+      "loss": 1.7116,
+      "step": 269
+    },
+    {
+      "epoch": 0.036028823058446756,
+      "grad_norm": 1.9425709100774724,
+      "learning_rate": 1.999810882486008e-05,
+      "loss": 1.7268,
+      "step": 270
+    },
+    {
+      "epoch": 0.03616226314384841,
+      "grad_norm": 1.6308935939887939,
+      "learning_rate": 1.99980238415172e-05,
+      "loss": 1.7157,
+      "step": 271
+    },
+    {
+      "epoch": 0.03629570322925007,
+      "grad_norm": 1.668857008376373,
+      "learning_rate": 1.99979369906573e-05,
+      "loss": 1.7032,
+      "step": 272
+    },
+    {
+      "epoch": 0.03642914331465172,
+      "grad_norm": 2.1917958281972925,
+      "learning_rate": 1.9997848272296594e-05,
+      "loss": 1.666,
+      "step": 273
+    },
+    {
+      "epoch": 0.036562583400053375,
+      "grad_norm": 2.7045086446698567,
+      "learning_rate": 1.9997757686451668e-05,
+      "loss": 1.7166,
+      "step": 274
+    },
+    {
+      "epoch": 0.03669602348545503,
+      "grad_norm": 1.776827654276239,
+      "learning_rate": 1.9997665233139433e-05,
+      "loss": 1.7162,
+      "step": 275
+    },
+    {
+      "epoch": 0.03682946357085669,
+      "grad_norm": 1.8454111862717217,
+      "learning_rate": 1.999757091237716e-05,
+      "loss": 1.7304,
+      "step": 276
+    },
+    {
+      "epoch": 0.03696290365625834,
+      "grad_norm": 1.7745973768422731,
+      "learning_rate": 1.9997474724182465e-05,
+      "loss": 1.7309,
+      "step": 277
+    },
+    {
+      "epoch": 0.037096343741659994,
+      "grad_norm": 1.768404847121436,
+      "learning_rate": 1.9997376668573317e-05,
+      "loss": 1.7116,
+      "step": 278
+    },
+    {
+      "epoch": 0.03722978382706165,
+      "grad_norm": 1.6641524050216268,
+      "learning_rate": 1.9997276745568036e-05,
+      "loss": 1.718,
+      "step": 279
+    },
+    {
+      "epoch": 0.03736322391246331,
+      "grad_norm": 2.2160286554301094,
+      "learning_rate": 1.9997174955185276e-05,
+      "loss": 1.6983,
+      "step": 280
+    },
+    {
+      "epoch": 0.03749666399786496,
+      "grad_norm": 1.830594152900142,
+      "learning_rate": 1.9997071297444062e-05,
+      "loss": 1.7126,
+      "step": 281
+    },
+    {
+      "epoch": 0.03763010408326661,
+      "grad_norm": 1.9590317837706277,
+      "learning_rate": 1.9996965772363747e-05,
+      "loss": 1.7766,
+      "step": 282
+    },
+    {
+      "epoch": 0.03776354416866827,
+      "grad_norm": 17.291654078400484,
+      "learning_rate": 1.999685837996405e-05,
+      "loss": 1.7718,
+      "step": 283
+    },
+    {
+      "epoch": 0.03789698425406992,
+      "grad_norm": 1.9837536918624412,
+      "learning_rate": 1.9996749120265023e-05,
+      "loss": 1.7488,
+      "step": 284
+    },
+    {
+      "epoch": 0.038030424339471576,
+      "grad_norm": 2.143999288457606,
+      "learning_rate": 1.9996637993287072e-05,
+      "loss": 1.756,
+      "step": 285
+    },
+    {
+      "epoch": 0.03816386442487323,
+      "grad_norm": 2.0362899269691144,
+      "learning_rate": 1.9996524999050966e-05,
+      "loss": 1.7582,
+      "step": 286
+    },
+    {
+      "epoch": 0.03829730451027489,
+      "grad_norm": 1.90769408022856,
+      "learning_rate": 1.9996410137577806e-05,
+      "loss": 1.7833,
+      "step": 287
+    },
+    {
+      "epoch": 0.03843074459567654,
+      "grad_norm": 1.8427410428015334,
+      "learning_rate": 1.9996293408889046e-05,
+      "loss": 1.746,
+      "step": 288
+    },
+    {
+      "epoch": 0.038564184681078195,
+      "grad_norm": 1.6678979095618587,
+      "learning_rate": 1.9996174813006488e-05,
+      "loss": 1.7362,
+      "step": 289
+    },
+    {
+      "epoch": 0.03869762476647985,
+      "grad_norm": 1.8342284170270304,
+      "learning_rate": 1.9996054349952283e-05,
+      "loss": 1.7476,
+      "step": 290
+    },
+    {
+      "epoch": 0.03883106485188151,
+      "grad_norm": 1.71509522487389,
+      "learning_rate": 1.999593201974894e-05,
+      "loss": 1.7033,
+      "step": 291
+    },
+    {
+      "epoch": 0.03896450493728316,
+      "grad_norm": 1.6721230982503732,
+      "learning_rate": 1.9995807822419296e-05,
+      "loss": 1.7231,
+      "step": 292
+    },
+    {
+      "epoch": 0.039097945022684814,
+      "grad_norm": 1.6080547725886836,
+      "learning_rate": 1.9995681757986563e-05,
+      "loss": 1.7517,
+      "step": 293
+    },
+    {
+      "epoch": 0.03923138510808647,
+      "grad_norm": 3.219421965283889,
+      "learning_rate": 1.9995553826474282e-05,
+      "loss": 1.771,
+      "step": 294
+    },
+    {
+      "epoch": 0.03936482519348812,
+      "grad_norm": 1.6715306645770436,
+      "learning_rate": 1.9995424027906348e-05,
+      "loss": 1.7727,
+      "step": 295
+    },
+    {
+      "epoch": 0.03949826527888978,
+      "grad_norm": 2.0944536026816047,
+      "learning_rate": 1.999529236230701e-05,
+      "loss": 1.6944,
+      "step": 296
+    },
+    {
+      "epoch": 0.039631705364291434,
+      "grad_norm": 1.8183245740053557,
+      "learning_rate": 1.9995158829700857e-05,
+      "loss": 1.7002,
+      "step": 297
+    },
+    {
+      "epoch": 0.03976514544969309,
+      "grad_norm": 1.5947407157115723,
+      "learning_rate": 1.9995023430112838e-05,
+      "loss": 1.694,
+      "step": 298
+    },
+    {
+      "epoch": 0.03989858553509474,
+      "grad_norm": 1.732145875042339,
+      "learning_rate": 1.9994886163568234e-05,
+      "loss": 1.7074,
+      "step": 299
+    },
+    {
+      "epoch": 0.040032025620496396,
+      "grad_norm": 1.9283377201797984,
+      "learning_rate": 1.9994747030092694e-05,
+      "loss": 1.6803,
+      "step": 300
+    },
+    {
+      "epoch": 0.04016546570589805,
+      "grad_norm": 2.0214640496563603,
+      "learning_rate": 1.9994606029712204e-05,
+      "loss": 1.6983,
+      "step": 301
+    },
+    {
+      "epoch": 0.04029890579129971,
+      "grad_norm": 1.6234729327561015,
+      "learning_rate": 1.9994463162453098e-05,
+      "loss": 1.7016,
+      "step": 302
+    },
+    {
+      "epoch": 0.04043234587670136,
+      "grad_norm": 1.9270631511486271,
+      "learning_rate": 1.9994318428342066e-05,
+      "loss": 1.7469,
+      "step": 303
+    },
+    {
+      "epoch": 0.040565785962103015,
+      "grad_norm": 1.6310686042908182,
+      "learning_rate": 1.9994171827406143e-05,
+      "loss": 1.7147,
+      "step": 304
+    },
+    {
+      "epoch": 0.04069922604750467,
+      "grad_norm": 1.6867380171369333,
+      "learning_rate": 1.999402335967271e-05,
+      "loss": 1.7216,
+      "step": 305
+    },
+    {
+      "epoch": 0.04083266613290633,
+      "grad_norm": 2.1857500165121295,
+      "learning_rate": 1.99938730251695e-05,
+      "loss": 1.7171,
+      "step": 306
+    },
+    {
+      "epoch": 0.04096610621830798,
+      "grad_norm": 2.3135110454799332,
+      "learning_rate": 1.999372082392459e-05,
+      "loss": 1.7334,
+      "step": 307
+    },
+    {
+      "epoch": 0.041099546303709635,
+      "grad_norm": 1.589398670927812,
+      "learning_rate": 1.9993566755966414e-05,
+      "loss": 1.6869,
+      "step": 308
+    },
+    {
+      "epoch": 0.04123298638911129,
+      "grad_norm": 2.4365266348869428,
+      "learning_rate": 1.999341082132375e-05,
+      "loss": 1.7227,
+      "step": 309
+    },
+    {
+      "epoch": 0.04136642647451294,
+      "grad_norm": 1.6296732392377482,
+      "learning_rate": 1.999325302002572e-05,
+      "loss": 1.7589,
+      "step": 310
+    },
+    {
+      "epoch": 0.0414998665599146,
+      "grad_norm": 2.305895115990403,
+      "learning_rate": 1.999309335210181e-05,
+      "loss": 1.6795,
+      "step": 311
+    },
+    {
+      "epoch": 0.041633306645316254,
+      "grad_norm": 2.2267067402414136,
+      "learning_rate": 1.9992931817581836e-05,
+      "loss": 1.7257,
+      "step": 312
+    },
+    {
+      "epoch": 0.04176674673071791,
+      "grad_norm": 2.64734565806814,
+      "learning_rate": 1.999276841649597e-05,
+      "loss": 1.6765,
+      "step": 313
+    },
+    {
+      "epoch": 0.04190018681611956,
+      "grad_norm": 2.871860278561938,
+      "learning_rate": 1.9992603148874735e-05,
+      "loss": 1.7917,
+      "step": 314
+    },
+    {
+      "epoch": 0.042033626901521216,
+      "grad_norm": 1.680479043931703,
+      "learning_rate": 1.9992436014749002e-05,
+      "loss": 1.7058,
+      "step": 315
+    },
+    {
+      "epoch": 0.04216706698692287,
+      "grad_norm": 2.3517294932157986,
+      "learning_rate": 1.9992267014149992e-05,
+      "loss": 1.7037,
+      "step": 316
+    },
+    {
+      "epoch": 0.04230050707232453,
+      "grad_norm": 2.3384834771184315,
+      "learning_rate": 1.999209614710927e-05,
+      "loss": 1.7229,
+      "step": 317
+    },
+    {
+      "epoch": 0.04243394715772618,
+      "grad_norm": 3.0716993591543162,
+      "learning_rate": 1.9991923413658752e-05,
+      "loss": 1.7066,
+      "step": 318
+    },
+    {
+      "epoch": 0.042567387243127836,
+      "grad_norm": 22.39020312724416,
+      "learning_rate": 1.99917488138307e-05,
+      "loss": 1.7471,
+      "step": 319
+    },
+    {
+      "epoch": 0.04270082732852949,
+      "grad_norm": 4.513453647912413,
+      "learning_rate": 1.999157234765773e-05,
+      "loss": 1.7847,
+      "step": 320
+    },
+    {
+      "epoch": 0.04283426741393114,
+      "grad_norm": 2.8952574314494406,
+      "learning_rate": 1.9991394015172806e-05,
+      "loss": 1.7865,
+      "step": 321
+    },
+    {
+      "epoch": 0.0429677074993328,
+      "grad_norm": 2.3072174733434263,
+      "learning_rate": 1.9991213816409235e-05,
+      "loss": 1.6652,
+      "step": 322
+    },
+    {
+      "epoch": 0.043101147584734455,
+      "grad_norm": 3.501051033962392,
+      "learning_rate": 1.9991031751400678e-05,
+      "loss": 1.7621,
+      "step": 323
+    },
+    {
+      "epoch": 0.04323458767013611,
+      "grad_norm": 2.757581328302007,
+      "learning_rate": 1.9990847820181143e-05,
+      "loss": 1.7351,
+      "step": 324
+    },
+    {
+      "epoch": 0.04336802775553776,
+      "grad_norm": 2.208520128826526,
+      "learning_rate": 1.9990662022784984e-05,
+      "loss": 1.7927,
+      "step": 325
+    },
+    {
+      "epoch": 0.04350146784093942,
+      "grad_norm": 2.9115362978660366,
+      "learning_rate": 1.9990474359246907e-05,
+      "loss": 1.7272,
+      "step": 326
+    },
+    {
+      "epoch": 0.043634907926341074,
+      "grad_norm": 2.2116484952074638,
+      "learning_rate": 1.9990284829601965e-05,
+      "loss": 1.7165,
+      "step": 327
+    },
+    {
+      "epoch": 0.04376834801174273,
+      "grad_norm": 2.127585292605601,
+      "learning_rate": 1.999009343388556e-05,
+      "loss": 1.7696,
+      "step": 328
+    },
+    {
+      "epoch": 0.04390178809714438,
+      "grad_norm": 2.1126461502594407,
+      "learning_rate": 1.9989900172133446e-05,
+      "loss": 1.7299,
+      "step": 329
+    },
+    {
+      "epoch": 0.044035228182546036,
+      "grad_norm": 3.589659926709359,
+      "learning_rate": 1.9989705044381717e-05,
+      "loss": 1.7291,
+      "step": 330
+    },
+    {
+      "epoch": 0.04416866826794769,
+      "grad_norm": 2.5882477625445977,
+      "learning_rate": 1.998950805066682e-05,
+      "loss": 1.74,
+      "step": 331
+    },
+    {
+      "epoch": 0.04430210835334935,
+      "grad_norm": 2.102193237422039,
+      "learning_rate": 1.998930919102556e-05,
+      "loss": 1.7445,
+      "step": 332
+    },
+    {
+      "epoch": 0.044435548438751,
+      "grad_norm": 1.5603391863630198,
+      "learning_rate": 1.9989108465495074e-05,
+      "loss": 1.7013,
+      "step": 333
+    },
+    {
+      "epoch": 0.044568988524152656,
+      "grad_norm": 2.0093620838637096,
+      "learning_rate": 1.9988905874112853e-05,
+      "loss": 1.6947,
+      "step": 334
+    },
+    {
+      "epoch": 0.04470242860955431,
+      "grad_norm": 2.0641135398132873,
+      "learning_rate": 1.998870141691675e-05,
+      "loss": 1.6937,
+      "step": 335
+    },
+    {
+      "epoch": 0.04483586869495596,
+      "grad_norm": 1.9939169534113899,
+      "learning_rate": 1.9988495093944942e-05,
+      "loss": 1.7692,
+      "step": 336
+    },
+    {
+      "epoch": 0.04496930878035762,
+      "grad_norm": 1.5839856511709431,
+      "learning_rate": 1.998828690523597e-05,
+      "loss": 1.6663,
+      "step": 337
+    },
+    {
+      "epoch": 0.045102748865759275,
+      "grad_norm": 1.7578902574401223,
+      "learning_rate": 1.9988076850828736e-05,
+      "loss": 1.7076,
+      "step": 338
+    },
+    {
+      "epoch": 0.04523618895116093,
+      "grad_norm": 1.8132930759338493,
+      "learning_rate": 1.9987864930762456e-05,
+      "loss": 1.673,
+      "step": 339
+    },
+    {
+      "epoch": 0.04536962903656258,
+      "grad_norm": 1.4712573812370557,
+      "learning_rate": 1.9987651145076724e-05,
+      "loss": 1.7153,
+      "step": 340
+    },
+    {
+      "epoch": 0.04550306912196424,
+      "grad_norm": 1.7206193008312878,
+      "learning_rate": 1.9987435493811477e-05,
+      "loss": 1.7383,
+      "step": 341
+    },
+    {
+      "epoch": 0.045636509207365894,
+      "grad_norm": 1.7893043936529118,
+      "learning_rate": 1.998721797700699e-05,
+      "loss": 1.7391,
+      "step": 342
+    },
+    {
+      "epoch": 0.04576994929276755,
+      "grad_norm": 1.6899544517809348,
+      "learning_rate": 1.9986998594703887e-05,
+      "loss": 1.7279,
+      "step": 343
+    },
+    {
+      "epoch": 0.0459033893781692,
+      "grad_norm": 1.6139144085071184,
+      "learning_rate": 1.9986777346943157e-05,
+      "loss": 1.6811,
+      "step": 344
+    },
+    {
+      "epoch": 0.04603682946357086,
+      "grad_norm": 1.7986249828416747,
+      "learning_rate": 1.9986554233766125e-05,
+      "loss": 1.7152,
+      "step": 345
+    },
+    {
+      "epoch": 0.04617026954897251,
+      "grad_norm": 1.7071224522178132,
+      "learning_rate": 1.9986329255214457e-05,
+      "loss": 1.6777,
+      "step": 346
+    },
+    {
+      "epoch": 0.04630370963437416,
+      "grad_norm": 1.726135209124258,
+      "learning_rate": 1.9986102411330187e-05,
+      "loss": 1.6931,
+      "step": 347
+    },
+    {
+      "epoch": 0.04643714971977582,
+      "grad_norm": 1.640238498392757,
+      "learning_rate": 1.9985873702155684e-05,
+      "loss": 1.7056,
+      "step": 348
+    },
+    {
+      "epoch": 0.046570589805177476,
+      "grad_norm": 1.4745370494240224,
+      "learning_rate": 1.9985643127733664e-05,
+      "loss": 1.7318,
+      "step": 349
+    },
+    {
+      "epoch": 0.04670402989057913,
+      "grad_norm": 1.665538578112985,
+      "learning_rate": 1.99854106881072e-05,
+      "loss": 1.6866,
+      "step": 350
+    },
+    {
+      "epoch": 0.04683746997598078,
+      "grad_norm": 1.7265148941792416,
+      "learning_rate": 1.9985176383319706e-05,
+      "loss": 1.7242,
+      "step": 351
+    },
+    {
+      "epoch": 0.04697091006138244,
+      "grad_norm": 1.7372398018580704,
+      "learning_rate": 1.998494021341495e-05,
+      "loss": 1.6928,
+      "step": 352
+    },
+    {
+      "epoch": 0.047104350146784095,
+      "grad_norm": 1.9848537709359042,
+      "learning_rate": 1.998470217843705e-05,
+      "loss": 1.7095,
+      "step": 353
+    },
+    {
+      "epoch": 0.04723779023218575,
+      "grad_norm": 1.6273684847266405,
+      "learning_rate": 1.9984462278430454e-05,
+      "loss": 1.7052,
+      "step": 354
+    },
+    {
+      "epoch": 0.0473712303175874,
+      "grad_norm": 2.691629432371725,
+      "learning_rate": 1.9984220513439987e-05,
+      "loss": 1.7216,
+      "step": 355
+    },
+    {
+      "epoch": 0.04750467040298906,
+      "grad_norm": 3.0017537794303872,
+      "learning_rate": 1.9983976883510802e-05,
+      "loss": 1.6843,
+      "step": 356
+    },
+    {
+      "epoch": 0.047638110488390714,
+      "grad_norm": 1.6968617974676807,
+      "learning_rate": 1.9983731388688405e-05,
+      "loss": 1.7386,
+      "step": 357
+    },
+    {
+      "epoch": 0.04777155057379237,
+      "grad_norm": 1.7186160692409527,
+      "learning_rate": 1.998348402901866e-05,
+      "loss": 1.7086,
+      "step": 358
+    },
+    {
+      "epoch": 0.04790499065919402,
+      "grad_norm": 1.6381474039682424,
+      "learning_rate": 1.998323480454776e-05,
+      "loss": 1.7258,
+      "step": 359
+    },
+    {
+      "epoch": 0.04803843074459568,
+      "grad_norm": 1.5835971815447754,
+      "learning_rate": 1.998298371532226e-05,
+      "loss": 1.7738,
+      "step": 360
+    },
+    {
+      "epoch": 0.04817187082999733,
+      "grad_norm": 1.5878715756664898,
+      "learning_rate": 1.9982730761389063e-05,
+      "loss": 1.7094,
+      "step": 361
+    },
+    {
+      "epoch": 0.04830531091539898,
+      "grad_norm": 1.519758329640418,
+      "learning_rate": 1.9982475942795418e-05,
+      "loss": 1.7268,
+      "step": 362
+    },
+    {
+      "epoch": 0.04843875100080064,
+      "grad_norm": 1.514307211931559,
+      "learning_rate": 1.9982219259588925e-05,
+      "loss": 1.6696,
+      "step": 363
+    },
+    {
+      "epoch": 0.048572191086202296,
+      "grad_norm": 2.1724863692721352,
+      "learning_rate": 1.9981960711817524e-05,
+      "loss": 1.6908,
+      "step": 364
+    },
+    {
+      "epoch": 0.04870563117160395,
+      "grad_norm": 1.56228704416919,
+      "learning_rate": 1.998170029952951e-05,
+      "loss": 1.7315,
+      "step": 365
+    },
+    {
+      "epoch": 0.0488390712570056,
+      "grad_norm": 1.85473949983945,
+      "learning_rate": 1.9981438022773526e-05,
+      "loss": 1.7355,
+      "step": 366
+    },
+    {
+      "epoch": 0.04897251134240726,
+      "grad_norm": 11.783330160673703,
+      "learning_rate": 1.9981173881598563e-05,
+      "loss": 1.6998,
+      "step": 367
+    },
+    {
+      "epoch": 0.049105951427808915,
+      "grad_norm": 2.0242537102814415,
+      "learning_rate": 1.998090787605396e-05,
+      "loss": 1.6909,
+      "step": 368
+    },
+    {
+      "epoch": 0.04923939151321057,
+      "grad_norm": 1.6717583123155606,
+      "learning_rate": 1.99806400061894e-05,
+      "loss": 1.7593,
+      "step": 369
+    },
+    {
+      "epoch": 0.04937283159861222,
+      "grad_norm": 1.5943321940814934,
+      "learning_rate": 1.998037027205492e-05,
+      "loss": 1.7278,
+      "step": 370
+    },
+    {
+      "epoch": 0.04950627168401388,
+      "grad_norm": 1.8339002788527259,
+      "learning_rate": 1.998009867370091e-05,
+      "loss": 1.7573,
+      "step": 371
+    },
+    {
+      "epoch": 0.049639711769415534,
+      "grad_norm": 1.6495978655725259,
+      "learning_rate": 1.997982521117809e-05,
+      "loss": 1.7594,
+      "step": 372
+    },
+    {
+      "epoch": 0.049773151854817184,
+      "grad_norm": 1.9817095312631359,
+      "learning_rate": 1.9979549884537545e-05,
+      "loss": 1.699,
+      "step": 373
+    },
+    {
+      "epoch": 0.04990659194021884,
+      "grad_norm": 1.803977303801047,
+      "learning_rate": 1.997927269383071e-05,
+      "loss": 1.733,
+      "step": 374
+    },
+    {
+      "epoch": 0.0500400320256205,
+      "grad_norm": 1.6078905730093185,
+      "learning_rate": 1.9978993639109344e-05,
+      "loss": 1.741,
+      "step": 375
+    },
+    {
+      "epoch": 0.050173472111022153,
+      "grad_norm": 1.5995695562728156,
+      "learning_rate": 1.997871272042559e-05,
+      "loss": 1.7242,
+      "step": 376
+    },
+    {
+      "epoch": 0.0503069121964238,
+      "grad_norm": 1.6340423898197851,
+      "learning_rate": 1.9978429937831905e-05,
+      "loss": 1.7679,
+      "step": 377
+    },
+    {
+      "epoch": 0.05044035228182546,
+      "grad_norm": 1.6634270572428624,
+      "learning_rate": 1.997814529138112e-05,
+      "loss": 1.7056,
+      "step": 378
+    },
+    {
+      "epoch": 0.050573792367227116,
+      "grad_norm": 1.7876625580664276,
+      "learning_rate": 1.99778587811264e-05,
+      "loss": 1.7285,
+      "step": 379
+    },
+    {
+      "epoch": 0.05070723245262877,
+      "grad_norm": 1.5691636161666582,
+      "learning_rate": 1.9977570407121258e-05,
+      "loss": 1.7767,
+      "step": 380
+    },
+    {
+      "epoch": 0.05084067253803042,
+      "grad_norm": 2.0869743704839894,
+      "learning_rate": 1.9977280169419567e-05,
+      "loss": 1.7575,
+      "step": 381
+    },
+    {
+      "epoch": 0.05097411262343208,
+      "grad_norm": 1.5252100639836486,
+      "learning_rate": 1.9976988068075536e-05,
+      "loss": 1.8007,
+      "step": 382
+    },
+    {
+      "epoch": 0.051107552708833735,
+      "grad_norm": 1.4730478361850414,
+      "learning_rate": 1.9976694103143725e-05,
+      "loss": 1.7411,
+      "step": 383
+    },
+    {
+      "epoch": 0.051240992794235385,
+      "grad_norm": 2.0844587485090043,
+      "learning_rate": 1.9976398274679044e-05,
+      "loss": 1.7389,
+      "step": 384
+    },
+    {
+      "epoch": 0.05137443287963704,
+      "grad_norm": 1.6654951502904705,
+      "learning_rate": 1.9976100582736753e-05,
+      "loss": 1.7117,
+      "step": 385
+    },
+    {
+      "epoch": 0.0515078729650387,
+      "grad_norm": 1.4065668432108118,
+      "learning_rate": 1.997580102737245e-05,
+      "loss": 1.7272,
+      "step": 386
+    },
+    {
+      "epoch": 0.051641313050440354,
+      "grad_norm": 1.6267242440943979,
+      "learning_rate": 1.99754996086421e-05,
+      "loss": 1.7481,
+      "step": 387
+    },
+    {
+      "epoch": 0.051774753135842004,
+      "grad_norm": 2.2148633010781773,
+      "learning_rate": 1.9975196326601997e-05,
+      "loss": 1.7006,
+      "step": 388
+    },
+    {
+      "epoch": 0.05190819322124366,
+      "grad_norm": 6.191778325835291,
+      "learning_rate": 1.9974891181308788e-05,
+      "loss": 1.7091,
+      "step": 389
+    },
+    {
+      "epoch": 0.05204163330664532,
+      "grad_norm": 13.595285838337412,
+      "learning_rate": 1.9974584172819478e-05,
+      "loss": 1.7493,
+      "step": 390
+    },
+    {
+      "epoch": 0.052175073392046974,
+      "grad_norm": 2.062696345964478,
+      "learning_rate": 1.997427530119141e-05,
+      "loss": 1.7039,
+      "step": 391
+    },
+    {
+      "epoch": 0.05230851347744862,
+      "grad_norm": 2.0114191550940075,
+      "learning_rate": 1.9973964566482276e-05,
+      "loss": 1.7469,
+      "step": 392
+    },
+    {
+      "epoch": 0.05244195356285028,
+      "grad_norm": 1.7826740746863319,
+      "learning_rate": 1.997365196875012e-05,
+      "loss": 1.6988,
+      "step": 393
+    },
+    {
+      "epoch": 0.052575393648251936,
+      "grad_norm": 1.8065043559103795,
+      "learning_rate": 1.9973337508053328e-05,
+      "loss": 1.6763,
+      "step": 394
+    },
+    {
+      "epoch": 0.05270883373365359,
+      "grad_norm": 1.618657676406062,
+      "learning_rate": 1.9973021184450644e-05,
+      "loss": 1.664,
+      "step": 395
+    },
+    {
+      "epoch": 0.05284227381905524,
+      "grad_norm": 1.669337246084699,
+      "learning_rate": 1.9972702998001145e-05,
+      "loss": 1.6838,
+      "step": 396
+    },
+    {
+      "epoch": 0.0529757139044569,
+      "grad_norm": 1.6318231642923446,
+      "learning_rate": 1.9972382948764274e-05,
+      "loss": 1.7319,
+      "step": 397
+    },
+    {
+      "epoch": 0.053109153989858555,
+      "grad_norm": 1.68424778338556,
+      "learning_rate": 1.9972061036799805e-05,
+      "loss": 1.7526,
+      "step": 398
+    },
+    {
+      "epoch": 0.053242594075260205,
+      "grad_norm": 10.099486327210181,
+      "learning_rate": 1.9971737262167873e-05,
+      "loss": 1.7284,
+      "step": 399
+    },
+    {
+      "epoch": 0.05337603416066186,
+      "grad_norm": 2.4830108610902792,
+      "learning_rate": 1.9971411624928952e-05,
+      "loss": 1.6935,
+      "step": 400
+    },
+    {
+      "epoch": 0.05350947424606352,
+      "grad_norm": 2.157249041077982,
+      "learning_rate": 1.997108412514387e-05,
+      "loss": 1.7532,
+      "step": 401
+    },
+    {
+      "epoch": 0.053642914331465175,
+      "grad_norm": 17.029332948986212,
+      "learning_rate": 1.9970754762873793e-05,
+      "loss": 1.7565,
+      "step": 402
+    },
+    {
+      "epoch": 0.053776354416866824,
+      "grad_norm": 2.2002999089958517,
+      "learning_rate": 1.9970423538180253e-05,
+      "loss": 1.7299,
+      "step": 403
+    },
+    {
+      "epoch": 0.05390979450226848,
+      "grad_norm": 1.9526294546905574,
+      "learning_rate": 1.997009045112511e-05,
+      "loss": 1.7624,
+      "step": 404
+    },
+    {
+      "epoch": 0.05404323458767014,
+      "grad_norm": 3.451322381710603,
+      "learning_rate": 1.9969755501770588e-05,
+      "loss": 1.7655,
+      "step": 405
+    },
+    {
+      "epoch": 0.054176674673071794,
+      "grad_norm": 3.4396767676243116,
+      "learning_rate": 1.9969418690179245e-05,
+      "loss": 1.7311,
+      "step": 406
+    },
+    {
+      "epoch": 0.05431011475847344,
+      "grad_norm": 2.047158976072317,
+      "learning_rate": 1.9969080016413998e-05,
+      "loss": 1.823,
+      "step": 407
+    },
+    {
+      "epoch": 0.0544435548438751,
+      "grad_norm": 1.7003670198468013,
+      "learning_rate": 1.9968739480538106e-05,
+      "loss": 1.6803,
+      "step": 408
+    },
+    {
+      "epoch": 0.054576994929276756,
+      "grad_norm": 1.5275767558995044,
+      "learning_rate": 1.9968397082615177e-05,
+      "loss": 1.6911,
+      "step": 409
+    },
+    {
+      "epoch": 0.054710435014678406,
+      "grad_norm": 1.5021865116802673,
+      "learning_rate": 1.9968052822709168e-05,
+      "loss": 1.6727,
+      "step": 410
+    },
+    {
+      "epoch": 0.05484387510008006,
+      "grad_norm": 1.6346663207881,
+      "learning_rate": 1.9967706700884383e-05,
+      "loss": 1.7605,
+      "step": 411
+    },
+    {
+      "epoch": 0.05497731518548172,
+      "grad_norm": 1.4986173436934926,
+      "learning_rate": 1.9967358717205473e-05,
+      "loss": 1.7654,
+      "step": 412
+    },
+    {
+      "epoch": 0.055110755270883376,
+      "grad_norm": 9.343403487653784,
+      "learning_rate": 1.9967008871737435e-05,
+      "loss": 1.7153,
+      "step": 413
+    },
+    {
+      "epoch": 0.055244195356285025,
+      "grad_norm": 2.0824830761991167,
+      "learning_rate": 1.996665716454562e-05,
+      "loss": 1.738,
+      "step": 414
+    },
+    {
+      "epoch": 0.05537763544168668,
+      "grad_norm": 2.05226788462729,
+      "learning_rate": 1.996630359569572e-05,
+      "loss": 1.6989,
+      "step": 415
+    },
+    {
+      "epoch": 0.05551107552708834,
+      "grad_norm": 2.801402642121381,
+      "learning_rate": 1.9965948165253783e-05,
+      "loss": 1.7352,
+      "step": 416
+    },
+    {
+      "epoch": 0.055644515612489995,
+      "grad_norm": 1.5863635183628055,
+      "learning_rate": 1.996559087328619e-05,
+      "loss": 1.7482,
+      "step": 417
+    },
+    {
+      "epoch": 0.055777955697891644,
+      "grad_norm": 1.5347479681973688,
+      "learning_rate": 1.9965231719859686e-05,
+      "loss": 1.7804,
+      "step": 418
+    },
+    {
+      "epoch": 0.0559113957832933,
+      "grad_norm": 5.941069172684277,
+      "learning_rate": 1.9964870705041356e-05,
+      "loss": 1.6768,
+      "step": 419
+    },
+    {
+      "epoch": 0.05604483586869496,
+      "grad_norm": 1.9058886708572287,
+      "learning_rate": 1.996450782889863e-05,
+      "loss": 1.8138,
+      "step": 420
+    },
+    {
+      "epoch": 0.056178275954096614,
+      "grad_norm": 1.83905231786784,
+      "learning_rate": 1.9964143091499296e-05,
+      "loss": 1.696,
+      "step": 421
+    },
+    {
+      "epoch": 0.056311716039498264,
+      "grad_norm": 21.111853586038833,
+      "learning_rate": 1.996377649291148e-05,
+      "loss": 1.8305,
+      "step": 422
+    },
+    {
+      "epoch": 0.05644515612489992,
+      "grad_norm": 2.0473455232249957,
+      "learning_rate": 1.9963408033203652e-05,
+      "loss": 1.6751,
+      "step": 423
+    },
+    {
+      "epoch": 0.05657859621030158,
+      "grad_norm": 2.3640946700664305,
+      "learning_rate": 1.9963037712444643e-05,
+      "loss": 1.6767,
+      "step": 424
+    },
+    {
+      "epoch": 0.056712036295703226,
+      "grad_norm": 2.639814143002119,
+      "learning_rate": 1.9962665530703623e-05,
+      "loss": 1.7202,
+      "step": 425
+    },
+    {
+      "epoch": 0.05684547638110488,
+      "grad_norm": 2.035709803937704,
+      "learning_rate": 1.996229148805011e-05,
+      "loss": 1.7528,
+      "step": 426
+    },
+    {
+      "epoch": 0.05697891646650654,
+      "grad_norm": 4.8723906405798605,
+      "learning_rate": 1.996191558455397e-05,
+      "loss": 1.7948,
+      "step": 427
+    },
+    {
+      "epoch": 0.057112356551908196,
+      "grad_norm": 1.9755270680940058,
+      "learning_rate": 1.9961537820285425e-05,
+      "loss": 1.739,
+      "step": 428
+    },
+    {
+      "epoch": 0.057245796637309845,
+      "grad_norm": 1.36715317359475,
+      "learning_rate": 1.996115819531503e-05,
+      "loss": 1.7144,
+      "step": 429
+    },
+    {
+      "epoch": 0.0573792367227115,
+      "grad_norm": 2.099518986464087,
+      "learning_rate": 1.9960776709713695e-05,
+      "loss": 1.7191,
+      "step": 430
+    },
+    {
+      "epoch": 0.05751267680811316,
+      "grad_norm": 1.5524531892573008,
+      "learning_rate": 1.9960393363552677e-05,
+      "loss": 1.6912,
+      "step": 431
+    },
+    {
+      "epoch": 0.057646116893514815,
+      "grad_norm": 1.979055446110802,
+      "learning_rate": 1.9960008156903584e-05,
+      "loss": 1.6863,
+      "step": 432
+    },
+    {
+      "epoch": 0.057779556978916465,
+      "grad_norm": 1.700547443693333,
+      "learning_rate": 1.9959621089838363e-05,
+      "loss": 1.6982,
+      "step": 433
+    },
+    {
+      "epoch": 0.05791299706431812,
+      "grad_norm": 1.5225244263250537,
+      "learning_rate": 1.995923216242932e-05,
+      "loss": 1.7289,
+      "step": 434
+    },
+    {
+      "epoch": 0.05804643714971978,
+      "grad_norm": 1.8075354741575524,
+      "learning_rate": 1.9958841374749102e-05,
+      "loss": 1.7704,
+      "step": 435
+    },
+    {
+      "epoch": 0.05817987723512143,
+      "grad_norm": 1.5109107308638536,
+      "learning_rate": 1.9958448726870695e-05,
+      "loss": 1.7511,
+      "step": 436
+    },
+    {
+      "epoch": 0.058313317320523084,
+      "grad_norm": 1.722880833771426,
+      "learning_rate": 1.995805421886745e-05,
+      "loss": 1.7224,
+      "step": 437
+    },
+    {
+      "epoch": 0.05844675740592474,
+      "grad_norm": 1.507934531634378,
+      "learning_rate": 1.995765785081305e-05,
+      "loss": 1.7202,
+      "step": 438
+    },
+    {
+      "epoch": 0.0585801974913264,
+      "grad_norm": 1.4464233282667935,
+      "learning_rate": 1.995725962278154e-05,
+      "loss": 1.6983,
+      "step": 439
+    },
+    {
+      "epoch": 0.058713637576728046,
+      "grad_norm": 3.165760428882322,
+      "learning_rate": 1.99568595348473e-05,
+      "loss": 1.6845,
+      "step": 440
+    },
+    {
+      "epoch": 0.0588470776621297,
+      "grad_norm": 10.453161040701552,
+      "learning_rate": 1.995645758708506e-05,
+      "loss": 1.7234,
+      "step": 441
+    },
+    {
+      "epoch": 0.05898051774753136,
+      "grad_norm": 1.8410106428998834,
+      "learning_rate": 1.9956053779569905e-05,
+      "loss": 1.7281,
+      "step": 442
+    },
+    {
+      "epoch": 0.059113957832933016,
+      "grad_norm": 1.442847521907269,
+      "learning_rate": 1.9955648112377254e-05,
+      "loss": 1.7504,
+      "step": 443
+    },
+    {
+      "epoch": 0.059247397918334666,
+      "grad_norm": 1.4761935818962095,
+      "learning_rate": 1.9955240585582887e-05,
+      "loss": 1.6651,
+      "step": 444
+    },
+    {
+      "epoch": 0.05938083800373632,
+      "grad_norm": 1.4515253586484216,
+      "learning_rate": 1.995483119926292e-05,
+      "loss": 1.7229,
+      "step": 445
+    },
+    {
+      "epoch": 0.05951427808913798,
+      "grad_norm": 1.5540039321519692,
+      "learning_rate": 1.9954419953493827e-05,
+      "loss": 1.7037,
+      "step": 446
+    },
+    {
+      "epoch": 0.059647718174539635,
+      "grad_norm": 1.5484860948366919,
+      "learning_rate": 1.9954006848352423e-05,
+      "loss": 1.7149,
+      "step": 447
+    },
+    {
+      "epoch": 0.059781158259941285,
+      "grad_norm": 1.9290588477973933,
+      "learning_rate": 1.995359188391587e-05,
+      "loss": 1.7327,
+      "step": 448
+    },
+    {
+      "epoch": 0.05991459834534294,
+      "grad_norm": 1.5388782815766935,
+      "learning_rate": 1.9953175060261677e-05,
+      "loss": 1.7035,
+      "step": 449
+    },
+    {
+      "epoch": 0.0600480384307446,
+      "grad_norm": 1.4965138673427807,
+      "learning_rate": 1.9952756377467706e-05,
+      "loss": 1.7023,
+      "step": 450
+    },
+    {
+      "epoch": 0.06018147851614625,
+      "grad_norm": 1.7236071176748005,
+      "learning_rate": 1.995233583561216e-05,
+      "loss": 1.7184,
+      "step": 451
+    },
+    {
+      "epoch": 0.060314918601547904,
+      "grad_norm": 1.431141636104607,
+      "learning_rate": 1.9951913434773592e-05,
+      "loss": 1.7301,
+      "step": 452
+    },
+    {
+      "epoch": 0.06044835868694956,
+      "grad_norm": 1.6645336690164663,
+      "learning_rate": 1.9951489175030902e-05,
+      "loss": 1.7039,
+      "step": 453
+    },
+    {
+      "epoch": 0.06058179877235122,
+      "grad_norm": 2.517722128154126,
+      "learning_rate": 1.995106305646333e-05,
+      "loss": 1.7634,
+      "step": 454
+    },
+    {
+      "epoch": 0.060715238857752866,
+      "grad_norm": 1.6794801287390608,
+      "learning_rate": 1.9950635079150483e-05,
+      "loss": 1.7285,
+      "step": 455
+    },
+    {
+      "epoch": 0.06084867894315452,
+      "grad_norm": 1.7417705396361953,
+      "learning_rate": 1.995020524317229e-05,
+      "loss": 1.7344,
+      "step": 456
+    },
+    {
+      "epoch": 0.06098211902855618,
+      "grad_norm": 1.6141113551164898,
+      "learning_rate": 1.994977354860905e-05,
+      "loss": 1.7246,
+      "step": 457
+    },
+    {
+      "epoch": 0.061115559113957836,
+      "grad_norm": 1.468263394815004,
+      "learning_rate": 1.9949339995541393e-05,
+      "loss": 1.7539,
+      "step": 458
+    },
+    {
+      "epoch": 0.061248999199359486,
+      "grad_norm": 1.665408795940273,
+      "learning_rate": 1.99489045840503e-05,
+      "loss": 1.6866,
+      "step": 459
+    },
+    {
+      "epoch": 0.06138243928476114,
+      "grad_norm": 1.9816359801419392,
+      "learning_rate": 1.9948467314217104e-05,
+      "loss": 1.6644,
+      "step": 460
+    },
+    {
+      "epoch": 0.0615158793701628,
+      "grad_norm": 1.5840211934856465,
+      "learning_rate": 1.9948028186123482e-05,
+      "loss": 1.6853,
+      "step": 461
+    },
+    {
+      "epoch": 0.06164931945556445,
+      "grad_norm": 1.5472589180288636,
+      "learning_rate": 1.9947587199851454e-05,
+      "loss": 1.7171,
+      "step": 462
+    },
+    {
+      "epoch": 0.061782759540966105,
+      "grad_norm": 1.531878299898292,
+      "learning_rate": 1.99471443554834e-05,
+      "loss": 1.7135,
+      "step": 463
+    },
+    {
+      "epoch": 0.06191619962636776,
+      "grad_norm": 1.5698651966216552,
+      "learning_rate": 1.9946699653102027e-05,
+      "loss": 1.7386,
+      "step": 464
+    },
+    {
+      "epoch": 0.06204963971176942,
+      "grad_norm": 1.3931585707012093,
+      "learning_rate": 1.994625309279041e-05,
+      "loss": 1.7277,
+      "step": 465
+    },
+    {
+      "epoch": 0.06218307979717107,
+      "grad_norm": 1.360573940164653,
+      "learning_rate": 1.994580467463196e-05,
+      "loss": 1.7149,
+      "step": 466
+    },
+    {
+      "epoch": 0.062316519882572724,
+      "grad_norm": 2.827001353046576,
+      "learning_rate": 1.9945354398710427e-05,
+      "loss": 1.7028,
+      "step": 467
+    },
+    {
+      "epoch": 0.06244995996797438,
+      "grad_norm": 1.5514492612805253,
+      "learning_rate": 1.994490226510993e-05,
+      "loss": 1.75,
+      "step": 468
+    },
+    {
+      "epoch": 0.06258340005337604,
+      "grad_norm": 1.3811202391730824,
+      "learning_rate": 1.9944448273914917e-05,
+      "loss": 1.6524,
+      "step": 469
+    },
+    {
+      "epoch": 0.0627168401387777,
+      "grad_norm": 1.460693904274723,
+      "learning_rate": 1.9943992425210186e-05,
+      "loss": 1.7228,
+      "step": 470
+    },
+    {
+      "epoch": 0.06285028022417935,
+      "grad_norm": 1.5719369846193096,
+      "learning_rate": 1.9943534719080885e-05,
+      "loss": 1.6843,
+      "step": 471
+    },
+    {
+      "epoch": 0.06298372030958099,
+      "grad_norm": 1.4562562982659377,
+      "learning_rate": 1.9943075155612513e-05,
+      "loss": 1.7289,
+      "step": 472
+    },
+    {
+      "epoch": 0.06311716039498265,
+      "grad_norm": 1.426871294088278,
+      "learning_rate": 1.9942613734890908e-05,
+      "loss": 1.7302,
+      "step": 473
+    },
+    {
+      "epoch": 0.0632506004803843,
+      "grad_norm": 1.4065664674136416,
+      "learning_rate": 1.9942150457002258e-05,
+      "loss": 1.6727,
+      "step": 474
+    },
+    {
+      "epoch": 0.06338404056578596,
+      "grad_norm": 1.4239048483373207,
+      "learning_rate": 1.9941685322033095e-05,
+      "loss": 1.7427,
+      "step": 475
+    },
+    {
+      "epoch": 0.06351748065118762,
+      "grad_norm": 1.4154094770828864,
+      "learning_rate": 1.9941218330070305e-05,
+      "loss": 1.7506,
+      "step": 476
+    },
+    {
+      "epoch": 0.06365092073658928,
+      "grad_norm": 1.3948618761716254,
+      "learning_rate": 1.994074948120112e-05,
+      "loss": 1.6968,
+      "step": 477
+    },
+    {
+      "epoch": 0.06378436082199093,
+      "grad_norm": 1.4269960440031193,
+      "learning_rate": 1.994027877551311e-05,
+      "loss": 1.7036,
+      "step": 478
+    },
+    {
+      "epoch": 0.06391780090739257,
+      "grad_norm": 1.561320058873614,
+      "learning_rate": 1.9939806213094196e-05,
+      "loss": 1.7169,
+      "step": 479
+    },
+    {
+      "epoch": 0.06405124099279423,
+      "grad_norm": 1.3666224092243797,
+      "learning_rate": 1.9939331794032655e-05,
+      "loss": 1.6997,
+      "step": 480
+    },
+    {
+      "epoch": 0.06418468107819589,
+      "grad_norm": 1.5591971808566425,
+      "learning_rate": 1.9938855518417096e-05,
+      "loss": 1.7198,
+      "step": 481
+    },
+    {
+      "epoch": 0.06431812116359754,
+      "grad_norm": 1.436666270843483,
+      "learning_rate": 1.9938377386336483e-05,
+      "loss": 1.7764,
+      "step": 482
+    },
+    {
+      "epoch": 0.0644515612489992,
+      "grad_norm": 2.5574706627508386,
+      "learning_rate": 1.993789739788013e-05,
+      "loss": 1.7151,
+      "step": 483
+    },
+    {
+      "epoch": 0.06458500133440086,
+      "grad_norm": 1.682482407939873,
+      "learning_rate": 1.9937415553137686e-05,
+      "loss": 1.7123,
+      "step": 484
+    },
+    {
+      "epoch": 0.06471844141980251,
+      "grad_norm": 1.4561970319846727,
+      "learning_rate": 1.993693185219916e-05,
+      "loss": 1.7404,
+      "step": 485
+    },
+    {
+      "epoch": 0.06485188150520416,
+      "grad_norm": 1.503219984664679,
+      "learning_rate": 1.9936446295154902e-05,
+      "loss": 1.7044,
+      "step": 486
+    },
+    {
+      "epoch": 0.06498532159060581,
+      "grad_norm": 1.4575867517583754,
+      "learning_rate": 1.9935958882095607e-05,
+      "loss": 1.6829,
+      "step": 487
+    },
+    {
+      "epoch": 0.06511876167600747,
+      "grad_norm": 1.5458704457306156,
+      "learning_rate": 1.9935469613112318e-05,
+      "loss": 1.7147,
+      "step": 488
+    },
+    {
+      "epoch": 0.06525220176140913,
+      "grad_norm": 1.4324636862239977,
+      "learning_rate": 1.9934978488296423e-05,
+      "loss": 1.699,
+      "step": 489
+    },
+    {
+      "epoch": 0.06538564184681078,
+      "grad_norm": 1.609515083336598,
+      "learning_rate": 1.993448550773966e-05,
+      "loss": 1.7071,
+      "step": 490
+    },
+    {
+      "epoch": 0.06551908193221244,
+      "grad_norm": 1.63784929721931,
+      "learning_rate": 1.9933990671534116e-05,
+      "loss": 1.7445,
+      "step": 491
+    },
+    {
+      "epoch": 0.0656525220176141,
+      "grad_norm": 1.7557428701106073,
+      "learning_rate": 1.9933493979772215e-05,
+      "loss": 1.7638,
+      "step": 492
+    },
+    {
+      "epoch": 0.06578596210301575,
+      "grad_norm": 1.6719589515914446,
+      "learning_rate": 1.9932995432546733e-05,
+      "loss": 1.6957,
+      "step": 493
+    },
+    {
+      "epoch": 0.0659194021884174,
+      "grad_norm": 1.3831309866724475,
+      "learning_rate": 1.99324950299508e-05,
+      "loss": 1.7108,
+      "step": 494
+    },
+    {
+      "epoch": 0.06605284227381905,
+      "grad_norm": 1.6438709703821648,
+      "learning_rate": 1.9931992772077877e-05,
+      "loss": 1.7459,
+      "step": 495
+    },
+    {
+      "epoch": 0.06618628235922071,
+      "grad_norm": 1.556799980817237,
+      "learning_rate": 1.993148865902179e-05,
+      "loss": 1.7165,
+      "step": 496
+    },
+    {
+      "epoch": 0.06631972244462236,
+      "grad_norm": 1.6005504681612104,
+      "learning_rate": 1.993098269087669e-05,
+      "loss": 1.6964,
+      "step": 497
+    },
+    {
+      "epoch": 0.06645316253002402,
+      "grad_norm": 1.6767116156256117,
+      "learning_rate": 1.9930474867737093e-05,
+      "loss": 1.7101,
+      "step": 498
+    },
+    {
+      "epoch": 0.06658660261542568,
+      "grad_norm": 1.437357072942104,
+      "learning_rate": 1.9929965189697855e-05,
+      "loss": 1.7503,
+      "step": 499
+    },
+    {
+      "epoch": 0.06672004270082733,
+      "grad_norm": 1.2785056943422897,
+      "learning_rate": 1.9929453656854177e-05,
+      "loss": 1.7404,
+      "step": 500
+    },
+    {
+      "epoch": 0.06685348278622898,
+      "grad_norm": 1.4081248663083363,
+      "learning_rate": 1.9928940269301607e-05,
+      "loss": 1.6389,
+      "step": 501
+    },
+    {
+      "epoch": 0.06698692287163063,
+      "grad_norm": 1.2600630608207999,
+      "learning_rate": 1.9928425027136038e-05,
+      "loss": 1.6954,
+      "step": 502
+    },
+    {
+      "epoch": 0.06712036295703229,
+      "grad_norm": 1.323168623645999,
+      "learning_rate": 1.992790793045371e-05,
+      "loss": 1.7456,
+      "step": 503
+    },
+    {
+      "epoch": 0.06725380304243395,
+      "grad_norm": 1.337670730789513,
+      "learning_rate": 1.9927388979351222e-05,
+      "loss": 1.7088,
+      "step": 504
+    },
+    {
+      "epoch": 0.0673872431278356,
+      "grad_norm": 1.3450109564421233,
+      "learning_rate": 1.9926868173925496e-05,
+      "loss": 1.7313,
+      "step": 505
+    },
+    {
+      "epoch": 0.06752068321323726,
+      "grad_norm": 6.8172313878010415,
+      "learning_rate": 1.9926345514273817e-05,
+      "loss": 1.7266,
+      "step": 506
+    },
+    {
+      "epoch": 0.06765412329863892,
+      "grad_norm": 1.4477937313346818,
+      "learning_rate": 1.992582100049381e-05,
+      "loss": 1.7154,
+      "step": 507
+    },
+    {
+      "epoch": 0.06778756338404057,
+      "grad_norm": 1.4944076158515036,
+      "learning_rate": 1.9925294632683454e-05,
+      "loss": 1.681,
+      "step": 508
+    },
+    {
+      "epoch": 0.06792100346944221,
+      "grad_norm": 1.4995222349055868,
+      "learning_rate": 1.992476641094106e-05,
+      "loss": 1.7149,
+      "step": 509
+    },
+    {
+      "epoch": 0.06805444355484387,
+      "grad_norm": 1.489364590881723,
+      "learning_rate": 1.99242363353653e-05,
+      "loss": 1.7353,
+      "step": 510
+    },
+    {
+      "epoch": 0.06818788364024553,
+      "grad_norm": 1.6809529621977222,
+      "learning_rate": 1.9923704406055185e-05,
+      "loss": 1.7274,
+      "step": 511
+    },
+    {
+      "epoch": 0.06832132372564718,
+      "grad_norm": 3.583760089378779,
+      "learning_rate": 1.992317062311007e-05,
+      "loss": 1.7217,
+      "step": 512
+    },
+    {
+      "epoch": 0.06845476381104884,
+      "grad_norm": 1.3874294319409661,
+      "learning_rate": 1.9922634986629667e-05,
+      "loss": 1.7285,
+      "step": 513
+    },
+    {
+      "epoch": 0.0685882038964505,
+      "grad_norm": 1.3167043130129317,
+      "learning_rate": 1.992209749671402e-05,
+      "loss": 1.7086,
+      "step": 514
+    },
+    {
+      "epoch": 0.06872164398185215,
+      "grad_norm": 1.6269952933086236,
+      "learning_rate": 1.9921558153463526e-05,
+      "loss": 1.7474,
+      "step": 515
+    },
+    {
+      "epoch": 0.0688550840672538,
+      "grad_norm": 2.1627161830488304,
+      "learning_rate": 1.992101695697893e-05,
+      "loss": 1.7087,
+      "step": 516
+    },
+    {
+      "epoch": 0.06898852415265545,
+      "grad_norm": 1.6239043871414123,
+      "learning_rate": 1.9920473907361324e-05,
+      "loss": 1.7406,
+      "step": 517
+    },
+    {
+      "epoch": 0.06912196423805711,
+      "grad_norm": 1.3153898428508883,
+      "learning_rate": 1.9919929004712137e-05,
+      "loss": 1.7223,
+      "step": 518
+    },
+    {
+      "epoch": 0.06925540432345877,
+      "grad_norm": 1.368634326229657,
+      "learning_rate": 1.9919382249133158e-05,
+      "loss": 1.6969,
+      "step": 519
+    },
+    {
+      "epoch": 0.06938884440886042,
+      "grad_norm": 1.6237538448200741,
+      "learning_rate": 1.991883364072651e-05,
+      "loss": 1.724,
+      "step": 520
+    },
+    {
+      "epoch": 0.06952228449426208,
+      "grad_norm": 1.3169440768502187,
+      "learning_rate": 1.991828317959467e-05,
+      "loss": 1.7785,
+      "step": 521
+    },
+    {
+      "epoch": 0.06965572457966374,
+      "grad_norm": 1.3815810002503197,
+      "learning_rate": 1.9917730865840453e-05,
+      "loss": 1.737,
+      "step": 522
+    },
+    {
+      "epoch": 0.06978916466506539,
+      "grad_norm": 19.153199421244327,
+      "learning_rate": 1.991717669956703e-05,
+      "loss": 1.6833,
+      "step": 523
+    },
+    {
+      "epoch": 0.06992260475046704,
+      "grad_norm": 1.6039439835379967,
+      "learning_rate": 1.991662068087791e-05,
+      "loss": 1.7322,
+      "step": 524
+    },
+    {
+      "epoch": 0.07005604483586869,
+      "grad_norm": 1.6305057820452133,
+      "learning_rate": 1.991606280987695e-05,
+      "loss": 1.7338,
+      "step": 525
+    },
+    {
+      "epoch": 0.07018948492127035,
+      "grad_norm": 1.6126054306261002,
+      "learning_rate": 1.9915503086668358e-05,
+      "loss": 1.7227,
+      "step": 526
+    },
+    {
+      "epoch": 0.070322925006672,
+      "grad_norm": 1.396267469940587,
+      "learning_rate": 1.991494151135668e-05,
+      "loss": 1.7336,
+      "step": 527
+    },
+    {
+      "epoch": 0.07045636509207366,
+      "grad_norm": 1.3519617857019466,
+      "learning_rate": 1.9914378084046814e-05,
+      "loss": 1.685,
+      "step": 528
+    },
+    {
+      "epoch": 0.07058980517747532,
+      "grad_norm": 1.369805017409349,
+      "learning_rate": 1.9913812804844003e-05,
+      "loss": 1.6849,
+      "step": 529
+    },
+    {
+      "epoch": 0.07072324526287697,
+      "grad_norm": 1.3456126243717488,
+      "learning_rate": 1.9913245673853833e-05,
+      "loss": 1.6779,
+      "step": 530
+    },
+    {
+      "epoch": 0.07085668534827862,
+      "grad_norm": 1.4115469932734985,
+      "learning_rate": 1.9912676691182237e-05,
+      "loss": 1.6661,
+      "step": 531
+    },
+    {
+      "epoch": 0.07099012543368027,
+      "grad_norm": 1.452789685185474,
+      "learning_rate": 1.9912105856935497e-05,
+      "loss": 1.722,
+      "step": 532
+    },
+    {
+      "epoch": 0.07112356551908193,
+      "grad_norm": 1.4137246122718905,
+      "learning_rate": 1.9911533171220234e-05,
+      "loss": 1.6956,
+      "step": 533
+    },
+    {
+      "epoch": 0.07125700560448359,
+      "grad_norm": 1.657715031131543,
+      "learning_rate": 1.991095863414342e-05,
+      "loss": 1.7233,
+      "step": 534
+    },
+    {
+      "epoch": 0.07139044568988524,
+      "grad_norm": 1.5491300325617336,
+      "learning_rate": 1.9910382245812376e-05,
+      "loss": 1.7357,
+      "step": 535
+    },
+    {
+      "epoch": 0.0715238857752869,
+      "grad_norm": 1.3876706174173905,
+      "learning_rate": 1.990980400633476e-05,
+      "loss": 1.6843,
+      "step": 536
+    },
+    {
+      "epoch": 0.07165732586068856,
+      "grad_norm": 1.5151426749768526,
+      "learning_rate": 1.9909223915818584e-05,
+      "loss": 1.7651,
+      "step": 537
+    },
+    {
+      "epoch": 0.0717907659460902,
+      "grad_norm": 1.3575479856663701,
+      "learning_rate": 1.99086419743722e-05,
+      "loss": 1.6813,
+      "step": 538
+    },
+    {
+      "epoch": 0.07192420603149186,
+      "grad_norm": 1.2693638194156773,
+      "learning_rate": 1.9908058182104317e-05,
+      "loss": 1.7248,
+      "step": 539
+    },
+    {
+      "epoch": 0.07205764611689351,
+      "grad_norm": 1.6457462336156898,
+      "learning_rate": 1.9907472539123968e-05,
+      "loss": 1.752,
+      "step": 540
+    },
+    {
+      "epoch": 0.07219108620229517,
+      "grad_norm": 1.678698655916284,
+      "learning_rate": 1.9906885045540547e-05,
+      "loss": 1.7071,
+      "step": 541
+    },
+    {
+      "epoch": 0.07232452628769682,
+      "grad_norm": 1.650324255621724,
+      "learning_rate": 1.99062957014638e-05,
+      "loss": 1.7447,
+      "step": 542
+    },
+    {
+      "epoch": 0.07245796637309848,
+      "grad_norm": 1.417893991120254,
+      "learning_rate": 1.9905704507003794e-05,
+      "loss": 1.6948,
+      "step": 543
+    },
+    {
+      "epoch": 0.07259140645850014,
+      "grad_norm": 1.4999406881312796,
+      "learning_rate": 1.9905111462270976e-05,
+      "loss": 1.648,
+      "step": 544
+    },
+    {
+      "epoch": 0.0727248465439018,
+      "grad_norm": 1.4453828984993826,
+      "learning_rate": 1.9904516567376105e-05,
+      "loss": 1.7441,
+      "step": 545
+    },
+    {
+      "epoch": 0.07285828662930344,
+      "grad_norm": 1.466543506749187,
+      "learning_rate": 1.990391982243031e-05,
+      "loss": 1.736,
+      "step": 546
+    },
+    {
+      "epoch": 0.0729917267147051,
+      "grad_norm": 1.2757799214271062,
+      "learning_rate": 1.990332122754505e-05,
+      "loss": 1.7273,
+      "step": 547
+    },
+    {
+      "epoch": 0.07312516680010675,
+      "grad_norm": 1.786649049168531,
+      "learning_rate": 1.9902720782832136e-05,
+      "loss": 1.7035,
+      "step": 548
+    },
+    {
+      "epoch": 0.0732586068855084,
+      "grad_norm": 1.6705192239448357,
+      "learning_rate": 1.990211848840373e-05,
+      "loss": 1.7395,
+      "step": 549
+    },
+    {
+      "epoch": 0.07339204697091006,
+      "grad_norm": 1.4329159010944417,
+      "learning_rate": 1.990151434437233e-05,
+      "loss": 1.7018,
+      "step": 550
+    },
+    {
+      "epoch": 0.07352548705631172,
+      "grad_norm": 1.3424218662116187,
+      "learning_rate": 1.9900908350850784e-05,
+      "loss": 1.6933,
+      "step": 551
+    },
+    {
+      "epoch": 0.07365892714171338,
+      "grad_norm": 1.4293148650020853,
+      "learning_rate": 1.9900300507952282e-05,
+      "loss": 1.7374,
+      "step": 552
+    },
+    {
+      "epoch": 0.07379236722711502,
+      "grad_norm": 1.3488966110143432,
+      "learning_rate": 1.9899690815790365e-05,
+      "loss": 1.72,
+      "step": 553
+    },
+    {
+      "epoch": 0.07392580731251668,
+      "grad_norm": 1.2512462922247582,
+      "learning_rate": 1.9899079274478916e-05,
+      "loss": 1.6908,
+      "step": 554
+    },
+    {
+      "epoch": 0.07405924739791833,
+      "grad_norm": 1.4908815449799042,
+      "learning_rate": 1.9898465884132164e-05,
+      "loss": 1.6776,
+      "step": 555
+    },
+    {
+      "epoch": 0.07419268748331999,
+      "grad_norm": 1.340382784697227,
+      "learning_rate": 1.9897850644864683e-05,
+      "loss": 1.7624,
+      "step": 556
+    },
+    {
+      "epoch": 0.07432612756872164,
+      "grad_norm": 1.5972162605012559,
+      "learning_rate": 1.9897233556791392e-05,
+      "loss": 1.7383,
+      "step": 557
+    },
+    {
+      "epoch": 0.0744595676541233,
+      "grad_norm": 1.5896167899937395,
+      "learning_rate": 1.989661462002756e-05,
+      "loss": 1.7443,
+      "step": 558
+    },
+    {
+      "epoch": 0.07459300773952496,
+      "grad_norm": 1.257607085158065,
+      "learning_rate": 1.9895993834688792e-05,
+      "loss": 1.6755,
+      "step": 559
+    },
+    {
+      "epoch": 0.07472644782492661,
+      "grad_norm": 1.3197820287481992,
+      "learning_rate": 1.9895371200891045e-05,
+      "loss": 1.7056,
+      "step": 560
+    },
+    {
+      "epoch": 0.07485988791032826,
+      "grad_norm": 1.2908766480278888,
+      "learning_rate": 1.989474671875062e-05,
+      "loss": 1.6912,
+      "step": 561
+    },
+    {
+      "epoch": 0.07499332799572991,
+      "grad_norm": 1.3304422687884063,
+      "learning_rate": 1.9894120388384167e-05,
+      "loss": 1.6982,
+      "step": 562
+    },
+    {
+      "epoch": 0.07512676808113157,
+      "grad_norm": 1.270386568576679,
+      "learning_rate": 1.9893492209908673e-05,
+      "loss": 1.6924,
+      "step": 563
+    },
+    {
+      "epoch": 0.07526020816653323,
+      "grad_norm": 1.3181304828307967,
+      "learning_rate": 1.9892862183441475e-05,
+      "loss": 1.7527,
+      "step": 564
+    },
+    {
+      "epoch": 0.07539364825193488,
+      "grad_norm": 1.2811143921204557,
+      "learning_rate": 1.9892230309100257e-05,
+      "loss": 1.6951,
+      "step": 565
+    },
+    {
+      "epoch": 0.07552708833733654,
+      "grad_norm": 1.5238424880045898,
+      "learning_rate": 1.989159658700304e-05,
+      "loss": 1.7398,
+      "step": 566
+    },
+    {
+      "epoch": 0.0756605284227382,
+      "grad_norm": 1.4366895400044768,
+      "learning_rate": 1.9890961017268206e-05,
+      "loss": 1.7202,
+      "step": 567
+    },
+    {
+      "epoch": 0.07579396850813984,
+      "grad_norm": 1.3526495950339494,
+      "learning_rate": 1.9890323600014465e-05,
+      "loss": 1.732,
+      "step": 568
+    },
+    {
+      "epoch": 0.0759274085935415,
+      "grad_norm": 1.8603030771660538,
+      "learning_rate": 1.9889684335360883e-05,
+      "loss": 1.7233,
+      "step": 569
+    },
+    {
+      "epoch": 0.07606084867894315,
+      "grad_norm": 1.457863107278941,
+      "learning_rate": 1.9889043223426864e-05,
+      "loss": 1.6965,
+      "step": 570
+    },
+    {
+      "epoch": 0.07619428876434481,
+      "grad_norm": 1.391003183808799,
+      "learning_rate": 1.988840026433216e-05,
+      "loss": 1.6628,
+      "step": 571
+    },
+    {
+      "epoch": 0.07632772884974647,
+      "grad_norm": 1.5192166538443774,
+      "learning_rate": 1.988775545819687e-05,
+      "loss": 1.7319,
+      "step": 572
+    },
+    {
+      "epoch": 0.07646116893514812,
+      "grad_norm": 1.349717213694597,
+      "learning_rate": 1.988710880514144e-05,
+      "loss": 1.7236,
+      "step": 573
+    },
+    {
+      "epoch": 0.07659460902054978,
+      "grad_norm": 1.4799233730457397,
+      "learning_rate": 1.9886460305286653e-05,
+      "loss": 1.6907,
+      "step": 574
+    },
+    {
+      "epoch": 0.07672804910595143,
+      "grad_norm": 1.3506939293859863,
+      "learning_rate": 1.988580995875364e-05,
+      "loss": 1.7699,
+      "step": 575
+    },
+    {
+      "epoch": 0.07686148919135308,
+      "grad_norm": 1.407704376614465,
+      "learning_rate": 1.9885157765663882e-05,
+      "loss": 1.6905,
+      "step": 576
+    },
+    {
+      "epoch": 0.07699492927675473,
+      "grad_norm": 1.3193501358588269,
+      "learning_rate": 1.98845037261392e-05,
+      "loss": 1.6843,
+      "step": 577
+    },
+    {
+      "epoch": 0.07712836936215639,
+      "grad_norm": 1.4235546422730225,
+      "learning_rate": 1.9883847840301766e-05,
+      "loss": 1.6729,
+      "step": 578
+    },
+    {
+      "epoch": 0.07726180944755805,
+      "grad_norm": 1.416179755704021,
+      "learning_rate": 1.9883190108274083e-05,
+      "loss": 1.6728,
+      "step": 579
+    },
+    {
+      "epoch": 0.0773952495329597,
+      "grad_norm": 1.4443285178109353,
+      "learning_rate": 1.9882530530179013e-05,
+      "loss": 1.7048,
+      "step": 580
+    },
+    {
+      "epoch": 0.07752868961836136,
+      "grad_norm": 1.5186108478360427,
+      "learning_rate": 1.9881869106139756e-05,
+      "loss": 1.7514,
+      "step": 581
+    },
+    {
+      "epoch": 0.07766212970376302,
+      "grad_norm": 1.5545405165342625,
+      "learning_rate": 1.988120583627986e-05,
+      "loss": 1.7108,
+      "step": 582
+    },
+    {
+      "epoch": 0.07779556978916466,
+      "grad_norm": 1.3900902877777435,
+      "learning_rate": 1.9880540720723214e-05,
+      "loss": 1.7202,
+      "step": 583
+    },
+    {
+      "epoch": 0.07792900987456632,
+      "grad_norm": 1.5816584989791276,
+      "learning_rate": 1.9879873759594057e-05,
+      "loss": 1.6767,
+      "step": 584
+    },
+    {
+      "epoch": 0.07806244995996797,
+      "grad_norm": 1.6976660554047953,
+      "learning_rate": 1.9879204953016968e-05,
+      "loss": 1.7432,
+      "step": 585
+    },
+    {
+      "epoch": 0.07819589004536963,
+      "grad_norm": 1.3644155406532983,
+      "learning_rate": 1.987853430111687e-05,
+      "loss": 1.7208,
+      "step": 586
+    },
+    {
+      "epoch": 0.07832933013077129,
+      "grad_norm": 1.7167546825301851,
+      "learning_rate": 1.987786180401904e-05,
+      "loss": 1.7147,
+      "step": 587
+    },
+    {
+      "epoch": 0.07846277021617294,
+      "grad_norm": 1.3847591281625475,
+      "learning_rate": 1.9877187461849086e-05,
+      "loss": 1.7112,
+      "step": 588
+    },
+    {
+      "epoch": 0.0785962103015746,
+      "grad_norm": 1.5571722725576893,
+      "learning_rate": 1.9876511274732973e-05,
+      "loss": 1.7638,
+      "step": 589
+    },
+    {
+      "epoch": 0.07872965038697624,
+      "grad_norm": 1.394231398970201,
+      "learning_rate": 1.9875833242797e-05,
+      "loss": 1.7043,
+      "step": 590
+    },
+    {
+      "epoch": 0.0788630904723779,
+      "grad_norm": 1.3846937701988455,
+      "learning_rate": 1.987515336616782e-05,
+      "loss": 1.7134,
+      "step": 591
+    },
+    {
+      "epoch": 0.07899653055777955,
+      "grad_norm": 1.3032703638813365,
+      "learning_rate": 1.9874471644972423e-05,
+      "loss": 1.7126,
+      "step": 592
+    },
+    {
+      "epoch": 0.07912997064318121,
+      "grad_norm": 1.4886448086335287,
+      "learning_rate": 1.9873788079338147e-05,
+      "loss": 1.7153,
+      "step": 593
+    },
+    {
+      "epoch": 0.07926341072858287,
+      "grad_norm": 1.611970344713762,
+      "learning_rate": 1.9873102669392676e-05,
+      "loss": 1.7026,
+      "step": 594
+    },
+    {
+      "epoch": 0.07939685081398452,
+      "grad_norm": 1.341383594822625,
+      "learning_rate": 1.9872415415264036e-05,
+      "loss": 1.6898,
+      "step": 595
+    },
+    {
+      "epoch": 0.07953029089938618,
+      "grad_norm": 1.3360746005534239,
+      "learning_rate": 1.9871726317080596e-05,
+      "loss": 1.6745,
+      "step": 596
+    },
+    {
+      "epoch": 0.07966373098478784,
+      "grad_norm": 8.410282418991848,
+      "learning_rate": 1.9871035374971076e-05,
+      "loss": 1.7583,
+      "step": 597
+    },
+    {
+      "epoch": 0.07979717107018948,
+      "grad_norm": 1.6794637542120652,
+      "learning_rate": 1.9870342589064533e-05,
+      "loss": 1.6641,
+      "step": 598
+    },
+    {
+      "epoch": 0.07993061115559114,
+      "grad_norm": 1.3453314375845957,
+      "learning_rate": 1.9869647959490373e-05,
+      "loss": 1.6974,
+      "step": 599
+    },
+    {
+      "epoch": 0.08006405124099279,
+      "grad_norm": 1.6191652794215983,
+      "learning_rate": 1.9868951486378344e-05,
+      "loss": 1.7029,
+      "step": 600
+    },
+    {
+      "epoch": 0.08019749132639445,
+      "grad_norm": 7.014227307878178,
+      "learning_rate": 1.986825316985854e-05,
+      "loss": 1.6979,
+      "step": 601
+    },
+    {
+      "epoch": 0.0803309314117961,
+      "grad_norm": 1.666603092412851,
+      "learning_rate": 1.9867553010061397e-05,
+      "loss": 1.7212,
+      "step": 602
+    },
+    {
+      "epoch": 0.08046437149719776,
+      "grad_norm": 6.777435475671393,
+      "learning_rate": 1.98668510071177e-05,
+      "loss": 1.6779,
+      "step": 603
+    },
+    {
+      "epoch": 0.08059781158259942,
+      "grad_norm": 9.166357022667544,
+      "learning_rate": 1.9866147161158574e-05,
+      "loss": 1.7199,
+      "step": 604
+    },
+    {
+      "epoch": 0.08073125166800106,
+      "grad_norm": 1.8917080711861438,
+      "learning_rate": 1.9865441472315482e-05,
+      "loss": 1.6838,
+      "step": 605
+    },
+    {
+      "epoch": 0.08086469175340272,
+      "grad_norm": 1.5977719491010742,
+      "learning_rate": 1.986473394072025e-05,
+      "loss": 1.7179,
+      "step": 606
+    },
+    {
+      "epoch": 0.08099813183880437,
+      "grad_norm": 1.658856522654762,
+      "learning_rate": 1.9864024566505034e-05,
+      "loss": 1.6835,
+      "step": 607
+    },
+    {
+      "epoch": 0.08113157192420603,
+      "grad_norm": 1.499498737885197,
+      "learning_rate": 1.9863313349802333e-05,
+      "loss": 1.7349,
+      "step": 608
+    },
+    {
+      "epoch": 0.08126501200960769,
+      "grad_norm": 13.15546017312421,
+      "learning_rate": 1.9862600290744996e-05,
+      "loss": 1.7886,
+      "step": 609
+    },
+    {
+      "epoch": 0.08139845209500934,
+      "grad_norm": 1.9828316081894657,
+      "learning_rate": 1.9861885389466216e-05,
+      "loss": 1.731,
+      "step": 610
+    },
+    {
+      "epoch": 0.081531892180411,
+      "grad_norm": 1.7448728734068772,
+      "learning_rate": 1.9861168646099525e-05,
+      "loss": 1.6969,
+      "step": 611
+    },
+    {
+      "epoch": 0.08166533226581266,
+      "grad_norm": 1.6273856601272179,
+      "learning_rate": 1.9860450060778806e-05,
+      "loss": 1.7009,
+      "step": 612
+    },
+    {
+      "epoch": 0.0817987723512143,
+      "grad_norm": 1.624049525128813,
+      "learning_rate": 1.9859729633638278e-05,
+      "loss": 1.762,
+      "step": 613
+    },
+    {
+      "epoch": 0.08193221243661596,
+      "grad_norm": 2.6801910069846797,
+      "learning_rate": 1.9859007364812516e-05,
+      "loss": 1.64,
+      "step": 614
+    },
+    {
+      "epoch": 0.08206565252201761,
+      "grad_norm": 1.905005831673537,
+      "learning_rate": 1.9858283254436425e-05,
+      "loss": 1.6968,
+      "step": 615
+    },
+    {
+      "epoch": 0.08219909260741927,
+      "grad_norm": 1.7290318164845841,
+      "learning_rate": 1.9857557302645265e-05,
+      "loss": 1.669,
+      "step": 616
+    },
+    {
+      "epoch": 0.08233253269282093,
+      "grad_norm": 1.7447056074174758,
+      "learning_rate": 1.985682950957463e-05,
+      "loss": 1.6803,
+      "step": 617
+    },
+    {
+      "epoch": 0.08246597277822258,
+      "grad_norm": 1.7209313384190483,
+      "learning_rate": 1.9856099875360472e-05,
+      "loss": 1.7356,
+      "step": 618
+    },
+    {
+      "epoch": 0.08259941286362424,
+      "grad_norm": 1.6574968534579264,
+      "learning_rate": 1.9855368400139068e-05,
+      "loss": 1.714,
+      "step": 619
+    },
+    {
+      "epoch": 0.08273285294902588,
+      "grad_norm": 2.0519868971563344,
+      "learning_rate": 1.9854635084047055e-05,
+      "loss": 1.6875,
+      "step": 620
+    },
+    {
+      "epoch": 0.08286629303442754,
+      "grad_norm": 1.5958780040536966,
+      "learning_rate": 1.9853899927221408e-05,
+      "loss": 1.7272,
+      "step": 621
+    },
+    {
+      "epoch": 0.0829997331198292,
+      "grad_norm": 1.596271866444752,
+      "learning_rate": 1.9853162929799445e-05,
+      "loss": 1.6827,
+      "step": 622
+    },
+    {
+      "epoch": 0.08313317320523085,
+      "grad_norm": 1.7231315613385811,
+      "learning_rate": 1.9852424091918832e-05,
+      "loss": 1.7274,
+      "step": 623
+    },
+    {
+      "epoch": 0.08326661329063251,
+      "grad_norm": 1.4156414113003977,
+      "learning_rate": 1.985168341371757e-05,
+      "loss": 1.7028,
+      "step": 624
+    },
+    {
+      "epoch": 0.08340005337603416,
+      "grad_norm": 1.6053274294337398,
+      "learning_rate": 1.9850940895334014e-05,
+      "loss": 1.7042,
+      "step": 625
+    },
+    {
+      "epoch": 0.08353349346143582,
+      "grad_norm": 1.4585415436887006,
+      "learning_rate": 1.9850196536906857e-05,
+      "loss": 1.7034,
+      "step": 626
+    },
+    {
+      "epoch": 0.08366693354683746,
+      "grad_norm": 1.7610381200893763,
+      "learning_rate": 1.9849450338575132e-05,
+      "loss": 1.6512,
+      "step": 627
+    },
+    {
+      "epoch": 0.08380037363223912,
+      "grad_norm": 1.3605669698961698,
+      "learning_rate": 1.9848702300478227e-05,
+      "loss": 1.7041,
+      "step": 628
+    },
+    {
+      "epoch": 0.08393381371764078,
+      "grad_norm": 1.3831476932394524,
+      "learning_rate": 1.984795242275586e-05,
+      "loss": 1.6914,
+      "step": 629
+    },
+    {
+      "epoch": 0.08406725380304243,
+      "grad_norm": 1.3145903110830626,
+      "learning_rate": 1.9847200705548106e-05,
+      "loss": 1.7033,
+      "step": 630
+    },
+    {
+      "epoch": 0.08420069388844409,
+      "grad_norm": 1.509116622712505,
+      "learning_rate": 1.9846447148995374e-05,
+      "loss": 1.7358,
+      "step": 631
+    },
+    {
+      "epoch": 0.08433413397384575,
+      "grad_norm": 1.3781652631062695,
+      "learning_rate": 1.984569175323842e-05,
+      "loss": 1.6917,
+      "step": 632
+    },
+    {
+      "epoch": 0.0844675740592474,
+      "grad_norm": 1.264221299881922,
+      "learning_rate": 1.984493451841835e-05,
+      "loss": 1.6757,
+      "step": 633
+    },
+    {
+      "epoch": 0.08460101414464906,
+      "grad_norm": 1.4764352016979319,
+      "learning_rate": 1.9844175444676594e-05,
+      "loss": 1.6922,
+      "step": 634
+    },
+    {
+      "epoch": 0.0847344542300507,
+      "grad_norm": 1.2723999017947936,
+      "learning_rate": 1.9843414532154946e-05,
+      "loss": 1.6704,
+      "step": 635
+    },
+    {
+      "epoch": 0.08486789431545236,
+      "grad_norm": 1.3038264727029132,
+      "learning_rate": 1.9842651780995536e-05,
+      "loss": 1.7411,
+      "step": 636
+    },
+    {
+      "epoch": 0.08500133440085401,
+      "grad_norm": 1.2103071746870917,
+      "learning_rate": 1.9841887191340835e-05,
+      "loss": 1.7046,
+      "step": 637
+    },
+    {
+      "epoch": 0.08513477448625567,
+      "grad_norm": 1.4950179854172028,
+      "learning_rate": 1.9841120763333665e-05,
+      "loss": 1.6348,
+      "step": 638
+    },
+    {
+      "epoch": 0.08526821457165733,
+      "grad_norm": 1.3121278424473823,
+      "learning_rate": 1.9840352497117178e-05,
+      "loss": 1.6354,
+      "step": 639
+    },
+    {
+      "epoch": 0.08540165465705898,
+      "grad_norm": 1.2723164770543802,
+      "learning_rate": 1.9839582392834883e-05,
+      "loss": 1.7162,
+      "step": 640
+    },
+    {
+      "epoch": 0.08553509474246064,
+      "grad_norm": 1.2860827518610605,
+      "learning_rate": 1.983881045063062e-05,
+      "loss": 1.7034,
+      "step": 641
+    },
+    {
+      "epoch": 0.08566853482786228,
+      "grad_norm": 1.2951379952738893,
+      "learning_rate": 1.983803667064859e-05,
+      "loss": 1.6595,
+      "step": 642
+    },
+    {
+      "epoch": 0.08580197491326394,
+      "grad_norm": 1.3206243247305416,
+      "learning_rate": 1.9837261053033316e-05,
+      "loss": 1.6181,
+      "step": 643
+    },
+    {
+      "epoch": 0.0859354149986656,
+      "grad_norm": 1.2669833972811808,
+      "learning_rate": 1.983648359792968e-05,
+      "loss": 1.7101,
+      "step": 644
+    },
+    {
+      "epoch": 0.08606885508406725,
+      "grad_norm": 1.29975526714487,
+      "learning_rate": 1.9835704305482905e-05,
+      "loss": 1.7718,
+      "step": 645
+    },
+    {
+      "epoch": 0.08620229516946891,
+      "grad_norm": 1.292410669128173,
+      "learning_rate": 1.9834923175838543e-05,
+      "loss": 1.6673,
+      "step": 646
+    },
+    {
+      "epoch": 0.08633573525487057,
+      "grad_norm": 1.4355885884004973,
+      "learning_rate": 1.9834140209142507e-05,
+      "loss": 1.7111,
+      "step": 647
+    },
+    {
+      "epoch": 0.08646917534027222,
+      "grad_norm": 1.3556914239382971,
+      "learning_rate": 1.983335540554105e-05,
+      "loss": 1.7227,
+      "step": 648
+    },
+    {
+      "epoch": 0.08660261542567388,
+      "grad_norm": 1.3219454699436908,
+      "learning_rate": 1.9832568765180758e-05,
+      "loss": 1.6788,
+      "step": 649
+    },
+    {
+      "epoch": 0.08673605551107552,
+      "grad_norm": 1.3046640649028514,
+      "learning_rate": 1.9831780288208565e-05,
+      "loss": 1.7202,
+      "step": 650
+    },
+    {
+      "epoch": 0.08686949559647718,
+      "grad_norm": 1.2956980030658956,
+      "learning_rate": 1.983098997477176e-05,
+      "loss": 1.6821,
+      "step": 651
+    },
+    {
+      "epoch": 0.08700293568187883,
+      "grad_norm": 1.4884991397141505,
+      "learning_rate": 1.9830197825017955e-05,
+      "loss": 1.6731,
+      "step": 652
+    },
+    {
+      "epoch": 0.08713637576728049,
+      "grad_norm": 1.2808830107084233,
+      "learning_rate": 1.9829403839095115e-05,
+      "loss": 1.6643,
+      "step": 653
+    },
+    {
+      "epoch": 0.08726981585268215,
+      "grad_norm": 1.3169596863329627,
+      "learning_rate": 1.9828608017151554e-05,
+      "loss": 1.6755,
+      "step": 654
+    },
+    {
+      "epoch": 0.0874032559380838,
+      "grad_norm": 1.4063600492133352,
+      "learning_rate": 1.9827810359335914e-05,
+      "loss": 1.6977,
+      "step": 655
+    },
+    {
+      "epoch": 0.08753669602348546,
+      "grad_norm": 2.4592773035398823,
+      "learning_rate": 1.9827010865797196e-05,
+      "loss": 1.6971,
+      "step": 656
+    },
+    {
+      "epoch": 0.0876701361088871,
+      "grad_norm": 1.5217703143835495,
+      "learning_rate": 1.9826209536684732e-05,
+      "loss": 1.747,
+      "step": 657
+    },
+    {
+      "epoch": 0.08780357619428876,
+      "grad_norm": 1.2927816468462434,
+      "learning_rate": 1.98254063721482e-05,
+      "loss": 1.671,
+      "step": 658
+    },
+    {
+      "epoch": 0.08793701627969042,
+      "grad_norm": 1.4773678075247743,
+      "learning_rate": 1.9824601372337628e-05,
+      "loss": 1.7076,
+      "step": 659
+    },
+    {
+      "epoch": 0.08807045636509207,
+      "grad_norm": 1.4581860403342026,
+      "learning_rate": 1.9823794537403372e-05,
+      "loss": 1.7272,
+      "step": 660
+    },
+    {
+      "epoch": 0.08820389645049373,
+      "grad_norm": 1.3752337806151618,
+      "learning_rate": 1.982298586749615e-05,
+      "loss": 1.6682,
+      "step": 661
+    },
+    {
+      "epoch": 0.08833733653589539,
+      "grad_norm": 1.353951137249054,
+      "learning_rate": 1.9822175362767006e-05,
+      "loss": 1.6528,
+      "step": 662
+    },
+    {
+      "epoch": 0.08847077662129704,
+      "grad_norm": 1.3149849437010714,
+      "learning_rate": 1.9821363023367327e-05,
+      "loss": 1.7095,
+      "step": 663
+    },
+    {
+      "epoch": 0.0886042167066987,
+      "grad_norm": 1.3521374496959613,
+      "learning_rate": 1.982054884944886e-05,
+      "loss": 1.7091,
+      "step": 664
+    },
+    {
+      "epoch": 0.08873765679210034,
+      "grad_norm": 1.2635125106435583,
+      "learning_rate": 1.9819732841163683e-05,
+      "loss": 1.6917,
+      "step": 665
+    },
+    {
+      "epoch": 0.088871096877502,
+      "grad_norm": 1.4576045027521718,
+      "learning_rate": 1.981891499866421e-05,
+      "loss": 1.7307,
+      "step": 666
+    },
+    {
+      "epoch": 0.08900453696290365,
+      "grad_norm": 2.0342924247731333,
+      "learning_rate": 1.9818095322103207e-05,
+      "loss": 1.7208,
+      "step": 667
+    },
+    {
+      "epoch": 0.08913797704830531,
+      "grad_norm": 1.3227265586786,
+      "learning_rate": 1.981727381163378e-05,
+      "loss": 1.7141,
+      "step": 668
+    },
+    {
+      "epoch": 0.08927141713370697,
+      "grad_norm": 1.3798866135669063,
+      "learning_rate": 1.981645046740938e-05,
+      "loss": 1.7157,
+      "step": 669
+    },
+    {
+      "epoch": 0.08940485721910862,
+      "grad_norm": 1.399678731841314,
+      "learning_rate": 1.9815625289583802e-05,
+      "loss": 1.7082,
+      "step": 670
+    },
+    {
+      "epoch": 0.08953829730451028,
+      "grad_norm": 1.4295883763194557,
+      "learning_rate": 1.981479827831117e-05,
+      "loss": 1.7054,
+      "step": 671
+    },
+    {
+      "epoch": 0.08967173738991192,
+      "grad_norm": 1.2635430082079968,
+      "learning_rate": 1.9813969433745964e-05,
+      "loss": 1.7389,
+      "step": 672
+    },
+    {
+      "epoch": 0.08980517747531358,
+      "grad_norm": 1.2907473612580835,
+      "learning_rate": 1.981313875604301e-05,
+      "loss": 1.7277,
+      "step": 673
+    },
+    {
+      "epoch": 0.08993861756071524,
+      "grad_norm": 1.237003971445107,
+      "learning_rate": 1.981230624535746e-05,
+      "loss": 1.7385,
+      "step": 674
+    },
+    {
+      "epoch": 0.0900720576461169,
+      "grad_norm": 1.3068726757473064,
+      "learning_rate": 1.9811471901844818e-05,
+      "loss": 1.7012,
+      "step": 675
+    },
+    {
+      "epoch": 0.09020549773151855,
+      "grad_norm": 1.3009066481680596,
+      "learning_rate": 1.9810635725660934e-05,
+      "loss": 1.7133,
+      "step": 676
+    },
+    {
+      "epoch": 0.0903389378169202,
+      "grad_norm": 1.4485758632883985,
+      "learning_rate": 1.9809797716961995e-05,
+      "loss": 1.704,
+      "step": 677
+    },
+    {
+      "epoch": 0.09047237790232186,
+      "grad_norm": 1.2649229686508479,
+      "learning_rate": 1.980895787590453e-05,
+      "loss": 1.7537,
+      "step": 678
+    },
+    {
+      "epoch": 0.0906058179877235,
+      "grad_norm": 1.2115856539500358,
+      "learning_rate": 1.9808116202645414e-05,
+      "loss": 1.7245,
+      "step": 679
+    },
+    {
+      "epoch": 0.09073925807312516,
+      "grad_norm": 1.8074140919241495,
+      "learning_rate": 1.9807272697341862e-05,
+      "loss": 1.7238,
+      "step": 680
+    },
+    {
+      "epoch": 0.09087269815852682,
+      "grad_norm": 1.3795215862127743,
+      "learning_rate": 1.980642736015143e-05,
+      "loss": 1.7037,
+      "step": 681
+    },
+    {
+      "epoch": 0.09100613824392847,
+      "grad_norm": 1.2851711329581033,
+      "learning_rate": 1.9805580191232017e-05,
+      "loss": 1.6901,
+      "step": 682
+    },
+    {
+      "epoch": 0.09113957832933013,
+      "grad_norm": 1.2298242189849842,
+      "learning_rate": 1.9804731190741867e-05,
+      "loss": 1.6914,
+      "step": 683
+    },
+    {
+      "epoch": 0.09127301841473179,
+      "grad_norm": 1.2359913396040505,
+      "learning_rate": 1.980388035883956e-05,
+      "loss": 1.6764,
+      "step": 684
+    },
+    {
+      "epoch": 0.09140645850013344,
+      "grad_norm": 1.4724399232032126,
+      "learning_rate": 1.980302769568402e-05,
+      "loss": 1.6784,
+      "step": 685
+    },
+    {
+      "epoch": 0.0915398985855351,
+      "grad_norm": 1.4540529296606115,
+      "learning_rate": 1.9802173201434522e-05,
+      "loss": 1.726,
+      "step": 686
+    },
+    {
+      "epoch": 0.09167333867093674,
+      "grad_norm": 1.2409303185307805,
+      "learning_rate": 1.980131687625067e-05,
+      "loss": 1.7201,
+      "step": 687
+    },
+    {
+      "epoch": 0.0918067787563384,
+      "grad_norm": 1.2194142955503393,
+      "learning_rate": 1.980045872029242e-05,
+      "loss": 1.7108,
+      "step": 688
+    },
+    {
+      "epoch": 0.09194021884174006,
+      "grad_norm": 1.39809489648676,
+      "learning_rate": 1.979959873372006e-05,
+      "loss": 1.6765,
+      "step": 689
+    },
+    {
+      "epoch": 0.09207365892714171,
+      "grad_norm": 1.253703018642756,
+      "learning_rate": 1.9798736916694234e-05,
+      "loss": 1.7405,
+      "step": 690
+    },
+    {
+      "epoch": 0.09220709901254337,
+      "grad_norm": 1.4561652380463412,
+      "learning_rate": 1.979787326937591e-05,
+      "loss": 1.7281,
+      "step": 691
+    },
+    {
+      "epoch": 0.09234053909794503,
+      "grad_norm": 1.2475596641603122,
+      "learning_rate": 1.9797007791926416e-05,
+      "loss": 1.7424,
+      "step": 692
+    },
+    {
+      "epoch": 0.09247397918334668,
+      "grad_norm": 1.2389551154520175,
+      "learning_rate": 1.9796140484507407e-05,
+      "loss": 1.64,
+      "step": 693
+    },
+    {
+      "epoch": 0.09260741926874833,
+      "grad_norm": 1.2409009102177742,
+      "learning_rate": 1.979527134728089e-05,
+      "loss": 1.7022,
+      "step": 694
+    },
+    {
+      "epoch": 0.09274085935414998,
+      "grad_norm": 3.070113056380779,
+      "learning_rate": 1.979440038040921e-05,
+      "loss": 1.7089,
+      "step": 695
+    },
+    {
+      "epoch": 0.09287429943955164,
+      "grad_norm": 1.2847037546220312,
+      "learning_rate": 1.9793527584055048e-05,
+      "loss": 1.6702,
+      "step": 696
+    },
+    {
+      "epoch": 0.0930077395249533,
+      "grad_norm": 1.342430254516785,
+      "learning_rate": 1.9792652958381442e-05,
+      "loss": 1.6981,
+      "step": 697
+    },
+    {
+      "epoch": 0.09314117961035495,
+      "grad_norm": 1.285114160550707,
+      "learning_rate": 1.9791776503551753e-05,
+      "loss": 1.7194,
+      "step": 698
+    },
+    {
+      "epoch": 0.09327461969575661,
+      "grad_norm": 1.384740422514496,
+      "learning_rate": 1.97908982197297e-05,
+      "loss": 1.6947,
+      "step": 699
+    },
+    {
+      "epoch": 0.09340805978115826,
+      "grad_norm": 1.317742692774045,
+      "learning_rate": 1.9790018107079328e-05,
+      "loss": 1.7059,
+      "step": 700
+    },
+    {
+      "epoch": 0.09354149986655992,
+      "grad_norm": 1.3948823280878986,
+      "learning_rate": 1.978913616576504e-05,
+      "loss": 1.6838,
+      "step": 701
+    },
+    {
+      "epoch": 0.09367493995196156,
+      "grad_norm": 1.5694970220793347,
+      "learning_rate": 1.9788252395951572e-05,
+      "loss": 1.7468,
+      "step": 702
+    },
+    {
+      "epoch": 0.09380838003736322,
+      "grad_norm": 1.3817257421047109,
+      "learning_rate": 1.9787366797804e-05,
+      "loss": 1.7176,
+      "step": 703
+    },
+    {
+      "epoch": 0.09394182012276488,
+      "grad_norm": 1.6513313977834363,
+      "learning_rate": 1.978647937148774e-05,
+      "loss": 1.6944,
+      "step": 704
+    },
+    {
+      "epoch": 0.09407526020816653,
+      "grad_norm": 1.405768726367186,
+      "learning_rate": 1.9785590117168558e-05,
+      "loss": 1.6983,
+      "step": 705
+    },
+    {
+      "epoch": 0.09420870029356819,
+      "grad_norm": 1.3377994406213816,
+      "learning_rate": 1.9784699035012552e-05,
+      "loss": 1.668,
+      "step": 706
+    },
+    {
+      "epoch": 0.09434214037896985,
+      "grad_norm": 1.4374634387262166,
+      "learning_rate": 1.9783806125186176e-05,
+      "loss": 1.725,
+      "step": 707
+    },
+    {
+      "epoch": 0.0944755804643715,
+      "grad_norm": 1.2800462634526275,
+      "learning_rate": 1.9782911387856204e-05,
+      "loss": 1.6953,
+      "step": 708
+    },
+    {
+      "epoch": 0.09460902054977315,
+      "grad_norm": 1.364118854409637,
+      "learning_rate": 1.978201482318977e-05,
+      "loss": 1.7155,
+      "step": 709
+    },
+    {
+      "epoch": 0.0947424606351748,
+      "grad_norm": 1.2401749198659822,
+      "learning_rate": 1.9781116431354337e-05,
+      "loss": 1.7057,
+      "step": 710
+    },
+    {
+      "epoch": 0.09487590072057646,
+      "grad_norm": 1.8122777264311412,
+      "learning_rate": 1.9780216212517718e-05,
+      "loss": 1.7083,
+      "step": 711
+    },
+    {
+      "epoch": 0.09500934080597812,
+      "grad_norm": 1.3875021914221024,
+      "learning_rate": 1.9779314166848063e-05,
+      "loss": 1.6533,
+      "step": 712
+    },
+    {
+      "epoch": 0.09514278089137977,
+      "grad_norm": 1.2772834218394613,
+      "learning_rate": 1.9778410294513865e-05,
+      "loss": 1.6915,
+      "step": 713
+    },
+    {
+      "epoch": 0.09527622097678143,
+      "grad_norm": 1.4645542214685092,
+      "learning_rate": 1.977750459568395e-05,
+      "loss": 1.7061,
+      "step": 714
+    },
+    {
+      "epoch": 0.09540966106218308,
+      "grad_norm": 1.38108304312693,
+      "learning_rate": 1.9776597070527502e-05,
+      "loss": 1.721,
+      "step": 715
+    },
+    {
+      "epoch": 0.09554310114758474,
+      "grad_norm": 1.382167340023687,
+      "learning_rate": 1.977568771921403e-05,
+      "loss": 1.7469,
+      "step": 716
+    },
+    {
+      "epoch": 0.09567654123298638,
+      "grad_norm": 1.4198404214818579,
+      "learning_rate": 1.9774776541913394e-05,
+      "loss": 1.6709,
+      "step": 717
+    },
+    {
+      "epoch": 0.09580998131838804,
+      "grad_norm": 1.3599396711566136,
+      "learning_rate": 1.9773863538795787e-05,
+      "loss": 1.7425,
+      "step": 718
+    },
+    {
+      "epoch": 0.0959434214037897,
+      "grad_norm": 1.2499141359082184,
+      "learning_rate": 1.9772948710031754e-05,
+      "loss": 1.6758,
+      "step": 719
+    },
+    {
+      "epoch": 0.09607686148919135,
+      "grad_norm": 1.3452952575533974,
+      "learning_rate": 1.977203205579217e-05,
+      "loss": 1.6976,
+      "step": 720
+    },
+    {
+      "epoch": 0.09621030157459301,
+      "grad_norm": 1.5618388812813502,
+      "learning_rate": 1.9771113576248257e-05,
+      "loss": 1.7143,
+      "step": 721
+    },
+    {
+      "epoch": 0.09634374165999467,
+      "grad_norm": 1.4099658140052458,
+      "learning_rate": 1.9770193271571573e-05,
+      "loss": 1.704,
+      "step": 722
+    },
+    {
+      "epoch": 0.09647718174539632,
+      "grad_norm": 1.4101573255249529,
+      "learning_rate": 1.976927114193403e-05,
+      "loss": 1.7099,
+      "step": 723
+    },
+    {
+      "epoch": 0.09661062183079797,
+      "grad_norm": 2.4533761018966898,
+      "learning_rate": 1.976834718750786e-05,
+      "loss": 1.7659,
+      "step": 724
+    },
+    {
+      "epoch": 0.09674406191619962,
+      "grad_norm": 1.450517529596779,
+      "learning_rate": 1.9767421408465654e-05,
+      "loss": 1.7009,
+      "step": 725
+    },
+    {
+      "epoch": 0.09687750200160128,
+      "grad_norm": 1.268810752392761,
+      "learning_rate": 1.9766493804980335e-05,
+      "loss": 1.7101,
+      "step": 726
+    },
+    {
+      "epoch": 0.09701094208700294,
+      "grad_norm": 1.3332261405385217,
+      "learning_rate": 1.976556437722517e-05,
+      "loss": 1.7127,
+      "step": 727
+    },
+    {
+      "epoch": 0.09714438217240459,
+      "grad_norm": 1.2579594665184823,
+      "learning_rate": 1.976463312537376e-05,
+      "loss": 1.7202,
+      "step": 728
+    },
+    {
+      "epoch": 0.09727782225780625,
+      "grad_norm": 1.5697939160123118,
+      "learning_rate": 1.976370004960006e-05,
+      "loss": 1.7365,
+      "step": 729
+    },
+    {
+      "epoch": 0.0974112623432079,
+      "grad_norm": 1.437009988573765,
+      "learning_rate": 1.9762765150078356e-05,
+      "loss": 1.7322,
+      "step": 730
+    },
+    {
+      "epoch": 0.09754470242860955,
+      "grad_norm": 1.3169792642285323,
+      "learning_rate": 1.9761828426983275e-05,
+      "loss": 1.7273,
+      "step": 731
+    },
+    {
+      "epoch": 0.0976781425140112,
+      "grad_norm": 1.5676941938167916,
+      "learning_rate": 1.9760889880489785e-05,
+      "loss": 1.6898,
+      "step": 732
+    },
+    {
+      "epoch": 0.09781158259941286,
+      "grad_norm": 1.4697881996182383,
+      "learning_rate": 1.9759949510773198e-05,
+      "loss": 1.7192,
+      "step": 733
+    },
+    {
+      "epoch": 0.09794502268481452,
+      "grad_norm": 1.2684891400607772,
+      "learning_rate": 1.9759007318009163e-05,
+      "loss": 1.7186,
+      "step": 734
+    },
+    {
+      "epoch": 0.09807846277021617,
+      "grad_norm": 1.2364993695259308,
+      "learning_rate": 1.9758063302373668e-05,
+      "loss": 1.7161,
+      "step": 735
+    },
+    {
+      "epoch": 0.09821190285561783,
+      "grad_norm": 1.2322024949845318,
+      "learning_rate": 1.9757117464043054e-05,
+      "loss": 1.6399,
+      "step": 736
+    },
+    {
+      "epoch": 0.09834534294101949,
+      "grad_norm": 1.2787037328868474,
+      "learning_rate": 1.9756169803193982e-05,
+      "loss": 1.6938,
+      "step": 737
+    },
+    {
+      "epoch": 0.09847878302642114,
+      "grad_norm": 1.4067488547142601,
+      "learning_rate": 1.9755220320003472e-05,
+      "loss": 1.6728,
+      "step": 738
+    },
+    {
+      "epoch": 0.09861222311182279,
+      "grad_norm": 1.2315172135495314,
+      "learning_rate": 1.9754269014648876e-05,
+      "loss": 1.6648,
+      "step": 739
+    },
+    {
+      "epoch": 0.09874566319722444,
+      "grad_norm": 1.267447733362674,
+      "learning_rate": 1.975331588730788e-05,
+      "loss": 1.706,
+      "step": 740
+    },
+    {
+      "epoch": 0.0988791032826261,
+      "grad_norm": 1.2625669177056562,
+      "learning_rate": 1.9752360938158522e-05,
+      "loss": 1.6826,
+      "step": 741
+    },
+    {
+      "epoch": 0.09901254336802776,
+      "grad_norm": 1.3204723066933017,
+      "learning_rate": 1.9751404167379175e-05,
+      "loss": 1.7471,
+      "step": 742
+    },
+    {
+      "epoch": 0.09914598345342941,
+      "grad_norm": 1.2156415623133245,
+      "learning_rate": 1.9750445575148557e-05,
+      "loss": 1.7028,
+      "step": 743
+    },
+    {
+      "epoch": 0.09927942353883107,
+      "grad_norm": 1.3803504882683986,
+      "learning_rate": 1.9749485161645715e-05,
+      "loss": 1.5917,
+      "step": 744
+    },
+    {
+      "epoch": 0.09941286362423273,
+      "grad_norm": 1.3249632567906648,
+      "learning_rate": 1.974852292705005e-05,
+      "loss": 1.7628,
+      "step": 745
+    },
+    {
+      "epoch": 0.09954630370963437,
+      "grad_norm": 1.2675581593990985,
+      "learning_rate": 1.974755887154129e-05,
+      "loss": 1.7161,
+      "step": 746
+    },
+    {
+      "epoch": 0.09967974379503602,
+      "grad_norm": 1.1665077868327602,
+      "learning_rate": 1.9746592995299515e-05,
+      "loss": 1.6748,
+      "step": 747
+    },
+    {
+      "epoch": 0.09981318388043768,
+      "grad_norm": 1.8639201171370587,
+      "learning_rate": 1.974562529850514e-05,
+      "loss": 1.697,
+      "step": 748
+    },
+    {
+      "epoch": 0.09994662396583934,
+      "grad_norm": 1.3196170477492044,
+      "learning_rate": 1.9744655781338913e-05,
+      "loss": 1.7133,
+      "step": 749
+    },
+    {
+      "epoch": 0.100080064051241,
+      "grad_norm": 1.2684243723887993,
+      "learning_rate": 1.9743684443981933e-05,
+      "loss": 1.7135,
+      "step": 750
+    },
+    {
+      "epoch": 0.10021350413664265,
+      "grad_norm": 1.1693884140497548,
+      "learning_rate": 1.9742711286615637e-05,
+      "loss": 1.7565,
+      "step": 751
+    },
+    {
+      "epoch": 0.10034694422204431,
+      "grad_norm": 1.2151625452030927,
+      "learning_rate": 1.97417363094218e-05,
+      "loss": 1.6717,
+      "step": 752
+    },
+    {
+      "epoch": 0.10048038430744596,
+      "grad_norm": 1.205035641499454,
+      "learning_rate": 1.974075951258253e-05,
+      "loss": 1.699,
+      "step": 753
+    },
+    {
+      "epoch": 0.1006138243928476,
+      "grad_norm": 1.547680875790279,
+      "learning_rate": 1.973978089628029e-05,
+      "loss": 1.7031,
+      "step": 754
+    },
+    {
+      "epoch": 0.10074726447824926,
+      "grad_norm": 1.5859192028100046,
+      "learning_rate": 1.9738800460697864e-05,
+      "loss": 1.7173,
+      "step": 755
+    },
+    {
+      "epoch": 0.10088070456365092,
+      "grad_norm": 1.5322336200687723,
+      "learning_rate": 1.9737818206018398e-05,
+      "loss": 1.6959,
+      "step": 756
+    },
+    {
+      "epoch": 0.10101414464905258,
+      "grad_norm": 1.2626433527759493,
+      "learning_rate": 1.9736834132425358e-05,
+      "loss": 1.7239,
+      "step": 757
+    },
+    {
+      "epoch": 0.10114758473445423,
+      "grad_norm": 1.505323027317607,
+      "learning_rate": 1.973584824010256e-05,
+      "loss": 1.6621,
+      "step": 758
+    },
+    {
+      "epoch": 0.10128102481985589,
+      "grad_norm": 1.286494859170549,
+      "learning_rate": 1.9734860529234158e-05,
+      "loss": 1.7242,
+      "step": 759
+    },
+    {
+      "epoch": 0.10141446490525755,
+      "grad_norm": 1.2306017189093845,
+      "learning_rate": 1.9733871000004643e-05,
+      "loss": 1.7357,
+      "step": 760
+    },
+    {
+      "epoch": 0.10154790499065919,
+      "grad_norm": 1.2061936583163555,
+      "learning_rate": 1.973287965259885e-05,
+      "loss": 1.7067,
+      "step": 761
+    },
+    {
+      "epoch": 0.10168134507606084,
+      "grad_norm": 1.2179611482678936,
+      "learning_rate": 1.9731886487201955e-05,
+      "loss": 1.685,
+      "step": 762
+    },
+    {
+      "epoch": 0.1018147851614625,
+      "grad_norm": 1.2683345257288643,
+      "learning_rate": 1.973089150399946e-05,
+      "loss": 1.6885,
+      "step": 763
+    },
+    {
+      "epoch": 0.10194822524686416,
+      "grad_norm": 1.2219892409057462,
+      "learning_rate": 1.9729894703177224e-05,
+      "loss": 1.6881,
+      "step": 764
+    },
+    {
+      "epoch": 0.10208166533226581,
+      "grad_norm": 1.458449697810168,
+      "learning_rate": 1.972889608492144e-05,
+      "loss": 1.7057,
+      "step": 765
+    },
+    {
+      "epoch": 0.10221510541766747,
+      "grad_norm": 1.3400740587143003,
+      "learning_rate": 1.972789564941863e-05,
+      "loss": 1.7171,
+      "step": 766
+    },
+    {
+      "epoch": 0.10234854550306913,
+      "grad_norm": 1.2400559343919726,
+      "learning_rate": 1.9726893396855675e-05,
+      "loss": 1.6838,
+      "step": 767
+    },
+    {
+      "epoch": 0.10248198558847077,
+      "grad_norm": 1.2051072425315403,
+      "learning_rate": 1.9725889327419773e-05,
+      "loss": 1.7223,
+      "step": 768
+    },
+    {
+      "epoch": 0.10261542567387243,
+      "grad_norm": 1.2802866462485152,
+      "learning_rate": 1.9724883441298482e-05,
+      "loss": 1.6311,
+      "step": 769
+    },
+    {
+      "epoch": 0.10274886575927408,
+      "grad_norm": 1.2328867984674214,
+      "learning_rate": 1.9723875738679684e-05,
+      "loss": 1.7446,
+      "step": 770
+    },
+    {
+      "epoch": 0.10288230584467574,
+      "grad_norm": 1.2341749669672337,
+      "learning_rate": 1.9722866219751606e-05,
+      "loss": 1.7008,
+      "step": 771
+    },
+    {
+      "epoch": 0.1030157459300774,
+      "grad_norm": 1.236991722122687,
+      "learning_rate": 1.972185488470282e-05,
+      "loss": 1.7141,
+      "step": 772
+    },
+    {
+      "epoch": 0.10314918601547905,
+      "grad_norm": 1.577844605519268,
+      "learning_rate": 1.9720841733722228e-05,
+      "loss": 1.7455,
+      "step": 773
+    },
+    {
+      "epoch": 0.10328262610088071,
+      "grad_norm": 1.3705708996585753,
+      "learning_rate": 1.9719826766999076e-05,
+      "loss": 1.7616,
+      "step": 774
+    },
+    {
+      "epoch": 0.10341606618628237,
+      "grad_norm": 1.221952982391811,
+      "learning_rate": 1.971880998472295e-05,
+      "loss": 1.7413,
+      "step": 775
+    },
+    {
+      "epoch": 0.10354950627168401,
+      "grad_norm": 1.537181900251082,
+      "learning_rate": 1.9717791387083772e-05,
+      "loss": 1.7011,
+      "step": 776
+    },
+    {
+      "epoch": 0.10368294635708566,
+      "grad_norm": 1.2027484662414312,
+      "learning_rate": 1.9716770974271803e-05,
+      "loss": 1.6332,
+      "step": 777
+    },
+    {
+      "epoch": 0.10381638644248732,
+      "grad_norm": 1.253964108899184,
+      "learning_rate": 1.9715748746477644e-05,
+      "loss": 1.7075,
+      "step": 778
+    },
+    {
+      "epoch": 0.10394982652788898,
+      "grad_norm": 1.309242214190543,
+      "learning_rate": 1.9714724703892238e-05,
+      "loss": 1.7086,
+      "step": 779
+    },
+    {
+      "epoch": 0.10408326661329063,
+      "grad_norm": 1.1977182401997006,
+      "learning_rate": 1.9713698846706865e-05,
+      "loss": 1.6517,
+      "step": 780
+    },
+    {
+      "epoch": 0.10421670669869229,
+      "grad_norm": 6.937492769576936,
+      "learning_rate": 1.971267117511314e-05,
+      "loss": 1.7718,
+      "step": 781
+    },
+    {
+      "epoch": 0.10435014678409395,
+      "grad_norm": 1.7070387484387968,
+      "learning_rate": 1.9711641689303024e-05,
+      "loss": 1.6522,
+      "step": 782
+    },
+    {
+      "epoch": 0.10448358686949559,
+      "grad_norm": 1.407052136214357,
+      "learning_rate": 1.971061038946881e-05,
+      "loss": 1.6851,
+      "step": 783
+    },
+    {
+      "epoch": 0.10461702695489725,
+      "grad_norm": 1.5040283593232824,
+      "learning_rate": 1.970957727580314e-05,
+      "loss": 1.64,
+      "step": 784
+    },
+    {
+      "epoch": 0.1047504670402989,
+      "grad_norm": 1.3258063153724626,
+      "learning_rate": 1.9708542348498975e-05,
+      "loss": 1.6948,
+      "step": 785
+    },
+    {
+      "epoch": 0.10488390712570056,
+      "grad_norm": 1.288600009166478,
+      "learning_rate": 1.970750560774964e-05,
+      "loss": 1.7193,
+      "step": 786
+    },
+    {
+      "epoch": 0.10501734721110222,
+      "grad_norm": 1.4250574606731519,
+      "learning_rate": 1.9706467053748782e-05,
+      "loss": 1.7096,
+      "step": 787
+    },
+    {
+      "epoch": 0.10515078729650387,
+      "grad_norm": 1.5215309159514565,
+      "learning_rate": 1.9705426686690387e-05,
+      "loss": 1.6971,
+      "step": 788
+    },
+    {
+      "epoch": 0.10528422738190553,
+      "grad_norm": 1.3671118437746754,
+      "learning_rate": 1.9704384506768788e-05,
+      "loss": 1.7509,
+      "step": 789
+    },
+    {
+      "epoch": 0.10541766746730719,
+      "grad_norm": 1.2902319309239332,
+      "learning_rate": 1.9703340514178656e-05,
+      "loss": 1.7539,
+      "step": 790
+    },
+    {
+      "epoch": 0.10555110755270883,
+      "grad_norm": 1.2681335491850911,
+      "learning_rate": 1.9702294709114987e-05,
+      "loss": 1.7117,
+      "step": 791
+    },
+    {
+      "epoch": 0.10568454763811048,
+      "grad_norm": 1.2418617832342838,
+      "learning_rate": 1.9701247091773135e-05,
+      "loss": 1.7345,
+      "step": 792
+    },
+    {
+      "epoch": 0.10581798772351214,
+      "grad_norm": 1.2895790697871072,
+      "learning_rate": 1.9700197662348777e-05,
+      "loss": 1.6623,
+      "step": 793
+    },
+    {
+      "epoch": 0.1059514278089138,
+      "grad_norm": 1.2684357393175132,
+      "learning_rate": 1.9699146421037935e-05,
+      "loss": 1.6487,
+      "step": 794
+    },
+    {
+      "epoch": 0.10608486789431545,
+      "grad_norm": 1.2423605419736183,
+      "learning_rate": 1.9698093368036976e-05,
+      "loss": 1.7129,
+      "step": 795
+    },
+    {
+      "epoch": 0.10621830797971711,
+      "grad_norm": 1.4752327188086216,
+      "learning_rate": 1.969703850354259e-05,
+      "loss": 1.6892,
+      "step": 796
+    },
+    {
+      "epoch": 0.10635174806511877,
+      "grad_norm": 1.4526839181300373,
+      "learning_rate": 1.9695981827751815e-05,
+      "loss": 1.7072,
+      "step": 797
+    },
+    {
+      "epoch": 0.10648518815052041,
+      "grad_norm": 1.285791119441606,
+      "learning_rate": 1.969492334086203e-05,
+      "loss": 1.7122,
+      "step": 798
+    },
+    {
+      "epoch": 0.10661862823592207,
+      "grad_norm": 1.2256787041306045,
+      "learning_rate": 1.9693863043070944e-05,
+      "loss": 1.6888,
+      "step": 799
+    },
+    {
+      "epoch": 0.10675206832132372,
+      "grad_norm": 1.228951068308044,
+      "learning_rate": 1.9692800934576607e-05,
+      "loss": 1.727,
+      "step": 800
+    },
+    {
+      "epoch": 0.10688550840672538,
+      "grad_norm": 1.352118087145811,
+      "learning_rate": 1.9691737015577418e-05,
+      "loss": 1.7125,
+      "step": 801
+    },
+    {
+      "epoch": 0.10701894849212704,
+      "grad_norm": 1.21157400011724,
+      "learning_rate": 1.96906712862721e-05,
+      "loss": 1.6923,
+      "step": 802
+    },
+    {
+      "epoch": 0.10715238857752869,
+      "grad_norm": 1.3003550717703736,
+      "learning_rate": 1.9689603746859714e-05,
+      "loss": 1.6511,
+      "step": 803
+    },
+    {
+      "epoch": 0.10728582866293035,
+      "grad_norm": 1.1564673283657656,
+      "learning_rate": 1.9688534397539666e-05,
+      "loss": 1.6726,
+      "step": 804
+    },
+    {
+      "epoch": 0.107419268748332,
+      "grad_norm": 1.2730332287589288,
+      "learning_rate": 1.9687463238511704e-05,
+      "loss": 1.6618,
+      "step": 805
+    },
+    {
+      "epoch": 0.10755270883373365,
+      "grad_norm": 1.3800666144026246,
+      "learning_rate": 1.9686390269975907e-05,
+      "loss": 1.7154,
+      "step": 806
+    },
+    {
+      "epoch": 0.1076861489191353,
+      "grad_norm": 1.2048204552645487,
+      "learning_rate": 1.968531549213269e-05,
+      "loss": 1.6448,
+      "step": 807
+    },
+    {
+      "epoch": 0.10781958900453696,
+      "grad_norm": 1.3937316868921632,
+      "learning_rate": 1.9684238905182807e-05,
+      "loss": 1.729,
+      "step": 808
+    },
+    {
+      "epoch": 0.10795302908993862,
+      "grad_norm": 1.3900561665275548,
+      "learning_rate": 1.968316050932736e-05,
+      "loss": 1.6828,
+      "step": 809
+    },
+    {
+      "epoch": 0.10808646917534027,
+      "grad_norm": 1.6114246988977743,
+      "learning_rate": 1.9682080304767775e-05,
+      "loss": 1.6792,
+      "step": 810
+    },
+    {
+      "epoch": 0.10821990926074193,
+      "grad_norm": 1.197150037360335,
+      "learning_rate": 1.9680998291705822e-05,
+      "loss": 1.6617,
+      "step": 811
+    },
+    {
+      "epoch": 0.10835334934614359,
+      "grad_norm": 1.2177958885989686,
+      "learning_rate": 1.9679914470343615e-05,
+      "loss": 1.7133,
+      "step": 812
+    },
+    {
+      "epoch": 0.10848678943154523,
+      "grad_norm": 1.494320786513006,
+      "learning_rate": 1.9678828840883592e-05,
+      "loss": 1.7004,
+      "step": 813
+    },
+    {
+      "epoch": 0.10862022951694689,
+      "grad_norm": 1.1694521285820114,
+      "learning_rate": 1.9677741403528538e-05,
+      "loss": 1.6699,
+      "step": 814
+    },
+    {
+      "epoch": 0.10875366960234854,
+      "grad_norm": 1.1962398552992284,
+      "learning_rate": 1.967665215848158e-05,
+      "loss": 1.6874,
+      "step": 815
+    },
+    {
+      "epoch": 0.1088871096877502,
+      "grad_norm": 1.5912303268847496,
+      "learning_rate": 1.9675561105946165e-05,
+      "loss": 1.6643,
+      "step": 816
+    },
+    {
+      "epoch": 0.10902054977315186,
+      "grad_norm": 1.2975078209899873,
+      "learning_rate": 1.96744682461261e-05,
+      "loss": 1.6789,
+      "step": 817
+    },
+    {
+      "epoch": 0.10915398985855351,
+      "grad_norm": 1.2145763409844237,
+      "learning_rate": 1.9673373579225514e-05,
+      "loss": 1.6771,
+      "step": 818
+    },
+    {
+      "epoch": 0.10928742994395517,
+      "grad_norm": 1.493883539659001,
+      "learning_rate": 1.9672277105448878e-05,
+      "loss": 1.6724,
+      "step": 819
+    },
+    {
+      "epoch": 0.10942087002935681,
+      "grad_norm": 1.644683081505485,
+      "learning_rate": 1.9671178825001002e-05,
+      "loss": 1.6167,
+      "step": 820
+    },
+    {
+      "epoch": 0.10955431011475847,
+      "grad_norm": 1.2719900635243897,
+      "learning_rate": 1.967007873808703e-05,
+      "loss": 1.6765,
+      "step": 821
+    },
+    {
+      "epoch": 0.10968775020016013,
+      "grad_norm": 1.2528047543957326,
+      "learning_rate": 1.966897684491245e-05,
+      "loss": 1.7102,
+      "step": 822
+    },
+    {
+      "epoch": 0.10982119028556178,
+      "grad_norm": 1.2419647024115608,
+      "learning_rate": 1.9667873145683082e-05,
+      "loss": 1.7077,
+      "step": 823
+    },
+    {
+      "epoch": 0.10995463037096344,
+      "grad_norm": 1.4501286698451352,
+      "learning_rate": 1.966676764060508e-05,
+      "loss": 1.6994,
+      "step": 824
+    },
+    {
+      "epoch": 0.1100880704563651,
+      "grad_norm": 1.3630120265958166,
+      "learning_rate": 1.9665660329884944e-05,
+      "loss": 1.6926,
+      "step": 825
+    },
+    {
+      "epoch": 0.11022151054176675,
+      "grad_norm": 1.460903039980999,
+      "learning_rate": 1.966455121372951e-05,
+      "loss": 1.7103,
+      "step": 826
+    },
+    {
+      "epoch": 0.11035495062716841,
+      "grad_norm": 1.5274078789068972,
+      "learning_rate": 1.9663440292345942e-05,
+      "loss": 1.6988,
+      "step": 827
+    },
+    {
+      "epoch": 0.11048839071257005,
+      "grad_norm": 1.2090669145048123,
+      "learning_rate": 1.9662327565941747e-05,
+      "loss": 1.6496,
+      "step": 828
+    },
+    {
+      "epoch": 0.1106218307979717,
+      "grad_norm": 1.2269483421344651,
+      "learning_rate": 1.9661213034724776e-05,
+      "loss": 1.7122,
+      "step": 829
+    },
+    {
+      "epoch": 0.11075527088337336,
+      "grad_norm": 1.4069578940447711,
+      "learning_rate": 1.9660096698903203e-05,
+      "loss": 1.7269,
+      "step": 830
+    },
+    {
+      "epoch": 0.11088871096877502,
+      "grad_norm": 1.2037399973430059,
+      "learning_rate": 1.9658978558685557e-05,
+      "loss": 1.7083,
+      "step": 831
+    },
+    {
+      "epoch": 0.11102215105417668,
+      "grad_norm": 1.2631008462452196,
+      "learning_rate": 1.9657858614280682e-05,
+      "loss": 1.6926,
+      "step": 832
+    },
+    {
+      "epoch": 0.11115559113957833,
+      "grad_norm": 1.3642794324735983,
+      "learning_rate": 1.965673686589778e-05,
+      "loss": 1.7298,
+      "step": 833
+    },
+    {
+      "epoch": 0.11128903122497999,
+      "grad_norm": 1.7518036471850427,
+      "learning_rate": 1.9655613313746378e-05,
+      "loss": 1.669,
+      "step": 834
+    },
+    {
+      "epoch": 0.11142247131038163,
+      "grad_norm": 1.3194015458910031,
+      "learning_rate": 1.965448795803634e-05,
+      "loss": 1.6799,
+      "step": 835
+    },
+    {
+      "epoch": 0.11155591139578329,
+      "grad_norm": 1.2408206570528904,
+      "learning_rate": 1.9653360798977872e-05,
+      "loss": 1.7045,
+      "step": 836
+    },
+    {
+      "epoch": 0.11168935148118495,
+      "grad_norm": 1.255528423044181,
+      "learning_rate": 1.9652231836781514e-05,
+      "loss": 1.6847,
+      "step": 837
+    },
+    {
+      "epoch": 0.1118227915665866,
+      "grad_norm": 1.5892490809881583,
+      "learning_rate": 1.965110107165815e-05,
+      "loss": 1.6491,
+      "step": 838
+    },
+    {
+      "epoch": 0.11195623165198826,
+      "grad_norm": 1.117194657692538,
+      "learning_rate": 1.964996850381898e-05,
+      "loss": 1.6832,
+      "step": 839
+    },
+    {
+      "epoch": 0.11208967173738991,
+      "grad_norm": 1.3292775964235752,
+      "learning_rate": 1.9648834133475564e-05,
+      "loss": 1.7325,
+      "step": 840
+    },
+    {
+      "epoch": 0.11222311182279157,
+      "grad_norm": 1.2706968194762298,
+      "learning_rate": 1.964769796083979e-05,
+      "loss": 1.7105,
+      "step": 841
+    },
+    {
+      "epoch": 0.11235655190819323,
+      "grad_norm": 1.1622186343437864,
+      "learning_rate": 1.964655998612388e-05,
+      "loss": 1.6583,
+      "step": 842
+    },
+    {
+      "epoch": 0.11248999199359487,
+      "grad_norm": 1.2930179210511994,
+      "learning_rate": 1.9645420209540394e-05,
+      "loss": 1.728,
+      "step": 843
+    },
+    {
+      "epoch": 0.11262343207899653,
+      "grad_norm": 1.3384278372678646,
+      "learning_rate": 1.964427863130223e-05,
+      "loss": 1.7889,
+      "step": 844
+    },
+    {
+      "epoch": 0.11275687216439818,
+      "grad_norm": 1.3828255437162351,
+      "learning_rate": 1.964313525162262e-05,
+      "loss": 1.6834,
+      "step": 845
+    },
+    {
+      "epoch": 0.11289031224979984,
+      "grad_norm": 1.151410588671105,
+      "learning_rate": 1.964199007071514e-05,
+      "loss": 1.7299,
+      "step": 846
+    },
+    {
+      "epoch": 0.1130237523352015,
+      "grad_norm": 1.1440279928005437,
+      "learning_rate": 1.9640843088793692e-05,
+      "loss": 1.7109,
+      "step": 847
+    },
+    {
+      "epoch": 0.11315719242060315,
+      "grad_norm": 1.4236141452119535,
+      "learning_rate": 1.9639694306072518e-05,
+      "loss": 1.7074,
+      "step": 848
+    },
+    {
+      "epoch": 0.11329063250600481,
+      "grad_norm": 1.2106461953286103,
+      "learning_rate": 1.96385437227662e-05,
+      "loss": 1.7274,
+      "step": 849
+    },
+    {
+      "epoch": 0.11342407259140645,
+      "grad_norm": 1.1556803406122926,
+      "learning_rate": 1.9637391339089655e-05,
+      "loss": 1.6884,
+      "step": 850
+    },
+    {
+      "epoch": 0.11355751267680811,
+      "grad_norm": 4.314358844478156,
+      "learning_rate": 1.9636237155258132e-05,
+      "loss": 1.7116,
+      "step": 851
+    },
+    {
+      "epoch": 0.11369095276220977,
+      "grad_norm": 1.3854953599475233,
+      "learning_rate": 1.9635081171487223e-05,
+      "loss": 1.7502,
+      "step": 852
+    },
+    {
+      "epoch": 0.11382439284761142,
+      "grad_norm": 1.3815136446017762,
+      "learning_rate": 1.9633923387992852e-05,
+      "loss": 1.6506,
+      "step": 853
+    },
+    {
+      "epoch": 0.11395783293301308,
+      "grad_norm": 1.2593646207844973,
+      "learning_rate": 1.9632763804991275e-05,
+      "loss": 1.6633,
+      "step": 854
+    },
+    {
+      "epoch": 0.11409127301841474,
+      "grad_norm": 1.1318090116685733,
+      "learning_rate": 1.9631602422699093e-05,
+      "loss": 1.6382,
+      "step": 855
+    },
+    {
+      "epoch": 0.11422471310381639,
+      "grad_norm": 1.523834563895417,
+      "learning_rate": 1.963043924133324e-05,
+      "loss": 1.7173,
+      "step": 856
+    },
+    {
+      "epoch": 0.11435815318921803,
+      "grad_norm": 1.369606548099637,
+      "learning_rate": 1.962927426111098e-05,
+      "loss": 1.7182,
+      "step": 857
+    },
+    {
+      "epoch": 0.11449159327461969,
+      "grad_norm": 1.2489405584762354,
+      "learning_rate": 1.9628107482249926e-05,
+      "loss": 1.6701,
+      "step": 858
+    },
+    {
+      "epoch": 0.11462503336002135,
+      "grad_norm": 1.3750560913985916,
+      "learning_rate": 1.9626938904968013e-05,
+      "loss": 1.725,
+      "step": 859
+    },
+    {
+      "epoch": 0.114758473445423,
+      "grad_norm": 1.3689472702896512,
+      "learning_rate": 1.962576852948352e-05,
+      "loss": 1.6338,
+      "step": 860
+    },
+    {
+      "epoch": 0.11489191353082466,
+      "grad_norm": 1.5090068885201111,
+      "learning_rate": 1.9624596356015057e-05,
+      "loss": 1.7569,
+      "step": 861
+    },
+    {
+      "epoch": 0.11502535361622632,
+      "grad_norm": 1.1857397963469596,
+      "learning_rate": 1.9623422384781575e-05,
+      "loss": 1.689,
+      "step": 862
+    },
+    {
+      "epoch": 0.11515879370162797,
+      "grad_norm": 4.383590101850042,
+      "learning_rate": 1.9622246616002362e-05,
+      "loss": 1.7869,
+      "step": 863
+    },
+    {
+      "epoch": 0.11529223378702963,
+      "grad_norm": 1.6301360882052789,
+      "learning_rate": 1.9621069049897026e-05,
+      "loss": 1.7169,
+      "step": 864
+    },
+    {
+      "epoch": 0.11542567387243127,
+      "grad_norm": 1.7846551885225532,
+      "learning_rate": 1.961988968668554e-05,
+      "loss": 1.731,
+      "step": 865
+    },
+    {
+      "epoch": 0.11555911395783293,
+      "grad_norm": 1.576969047359226,
+      "learning_rate": 1.9618708526588187e-05,
+      "loss": 1.7073,
+      "step": 866
+    },
+    {
+      "epoch": 0.11569255404323459,
+      "grad_norm": 1.4423014187494878,
+      "learning_rate": 1.961752556982559e-05,
+      "loss": 1.6984,
+      "step": 867
+    },
+    {
+      "epoch": 0.11582599412863624,
+      "grad_norm": 1.411158314477759,
+      "learning_rate": 1.9616340816618718e-05,
+      "loss": 1.6689,
+      "step": 868
+    },
+    {
+      "epoch": 0.1159594342140379,
+      "grad_norm": 1.2615885604648809,
+      "learning_rate": 1.9615154267188865e-05,
+      "loss": 1.7033,
+      "step": 869
+    },
+    {
+      "epoch": 0.11609287429943956,
+      "grad_norm": 1.3497153430677546,
+      "learning_rate": 1.9613965921757672e-05,
+      "loss": 1.7052,
+      "step": 870
+    },
+    {
+      "epoch": 0.11622631438484121,
+      "grad_norm": 1.3238657405928431,
+      "learning_rate": 1.96127757805471e-05,
+      "loss": 1.6803,
+      "step": 871
+    },
+    {
+      "epoch": 0.11635975447024285,
+      "grad_norm": 1.4113872862910883,
+      "learning_rate": 1.961158384377946e-05,
+      "loss": 1.6832,
+      "step": 872
+    },
+    {
+      "epoch": 0.11649319455564451,
+      "grad_norm": 1.1657735990335347,
+      "learning_rate": 1.9610390111677388e-05,
+      "loss": 1.6761,
+      "step": 873
+    },
+    {
+      "epoch": 0.11662663464104617,
+      "grad_norm": 1.4288925931679666,
+      "learning_rate": 1.9609194584463866e-05,
+      "loss": 1.7142,
+      "step": 874
+    },
+    {
+      "epoch": 0.11676007472644782,
+      "grad_norm": 1.3482735605608565,
+      "learning_rate": 1.9607997262362195e-05,
+      "loss": 1.7664,
+      "step": 875
+    },
+    {
+      "epoch": 0.11689351481184948,
+      "grad_norm": 1.2758965299420557,
+      "learning_rate": 1.960679814559603e-05,
+      "loss": 1.7643,
+      "step": 876
+    },
+    {
+      "epoch": 0.11702695489725114,
+      "grad_norm": 2.091988370893863,
+      "learning_rate": 1.9605597234389347e-05,
+      "loss": 1.6968,
+      "step": 877
+    },
+    {
+      "epoch": 0.1171603949826528,
+      "grad_norm": 1.7213048905302604,
+      "learning_rate": 1.9604394528966467e-05,
+      "loss": 1.6905,
+      "step": 878
+    },
+    {
+      "epoch": 0.11729383506805445,
+      "grad_norm": 1.3791112926314404,
+      "learning_rate": 1.9603190029552036e-05,
+      "loss": 1.6883,
+      "step": 879
+    },
+    {
+      "epoch": 0.11742727515345609,
+      "grad_norm": 1.3014168835710302,
+      "learning_rate": 1.9601983736371047e-05,
+      "loss": 1.747,
+      "step": 880
+    },
+    {
+      "epoch": 0.11756071523885775,
+      "grad_norm": 1.2352469046555101,
+      "learning_rate": 1.9600775649648818e-05,
+      "loss": 1.663,
+      "step": 881
+    },
+    {
+      "epoch": 0.1176941553242594,
+      "grad_norm": 1.3042902670559662,
+      "learning_rate": 1.9599565769611004e-05,
+      "loss": 1.6927,
+      "step": 882
+    },
+    {
+      "epoch": 0.11782759540966106,
+      "grad_norm": 1.3814487920433733,
+      "learning_rate": 1.95983540964836e-05,
+      "loss": 1.6648,
+      "step": 883
+    },
+    {
+      "epoch": 0.11796103549506272,
+      "grad_norm": 1.192223516720673,
+      "learning_rate": 1.9597140630492934e-05,
+      "loss": 1.6921,
+      "step": 884
+    },
+    {
+      "epoch": 0.11809447558046438,
+      "grad_norm": 1.2034614588136634,
+      "learning_rate": 1.9595925371865667e-05,
+      "loss": 1.6936,
+      "step": 885
+    },
+    {
+      "epoch": 0.11822791566586603,
+      "grad_norm": 1.1538622984795832,
+      "learning_rate": 1.959470832082879e-05,
+      "loss": 1.6821,
+      "step": 886
+    },
+    {
+      "epoch": 0.11836135575126767,
+      "grad_norm": 1.2580439984544938,
+      "learning_rate": 1.9593489477609646e-05,
+      "loss": 1.7528,
+      "step": 887
+    },
+    {
+      "epoch": 0.11849479583666933,
+      "grad_norm": 1.4294495361334751,
+      "learning_rate": 1.9592268842435884e-05,
+      "loss": 1.7465,
+      "step": 888
+    },
+    {
+      "epoch": 0.11862823592207099,
+      "grad_norm": 1.257681366416578,
+      "learning_rate": 1.959104641553552e-05,
+      "loss": 1.6859,
+      "step": 889
+    },
+    {
+      "epoch": 0.11876167600747264,
+      "grad_norm": 1.2143693832991094,
+      "learning_rate": 1.958982219713688e-05,
+      "loss": 1.7008,
+      "step": 890
+    },
+    {
+      "epoch": 0.1188951160928743,
+      "grad_norm": 1.2891059442807018,
+      "learning_rate": 1.9588596187468642e-05,
+      "loss": 1.7103,
+      "step": 891
+    },
+    {
+      "epoch": 0.11902855617827596,
+      "grad_norm": 1.2311129858964251,
+      "learning_rate": 1.9587368386759805e-05,
+      "loss": 1.714,
+      "step": 892
+    },
+    {
+      "epoch": 0.11916199626367761,
+      "grad_norm": 1.2118946144194225,
+      "learning_rate": 1.9586138795239708e-05,
+      "loss": 1.6704,
+      "step": 893
+    },
+    {
+      "epoch": 0.11929543634907927,
+      "grad_norm": 1.159008481780571,
+      "learning_rate": 1.9584907413138028e-05,
+      "loss": 1.6918,
+      "step": 894
+    },
+    {
+      "epoch": 0.11942887643448091,
+      "grad_norm": 1.2289280001259455,
+      "learning_rate": 1.9583674240684774e-05,
+      "loss": 1.7155,
+      "step": 895
+    },
+    {
+      "epoch": 0.11956231651988257,
+      "grad_norm": 1.2482207680648996,
+      "learning_rate": 1.9582439278110282e-05,
+      "loss": 1.6881,
+      "step": 896
+    },
+    {
+      "epoch": 0.11969575660528423,
+      "grad_norm": 1.435525566789562,
+      "learning_rate": 1.9581202525645232e-05,
+      "loss": 1.6873,
+      "step": 897
+    },
+    {
+      "epoch": 0.11982919669068588,
+      "grad_norm": 1.3060438169865851,
+      "learning_rate": 1.957996398352064e-05,
+      "loss": 1.6797,
+      "step": 898
+    },
+    {
+      "epoch": 0.11996263677608754,
+      "grad_norm": 1.2737667444400789,
+      "learning_rate": 1.9578723651967845e-05,
+      "loss": 1.6942,
+      "step": 899
+    },
+    {
+      "epoch": 0.1200960768614892,
+      "grad_norm": 1.6807278602860825,
+      "learning_rate": 1.957748153121853e-05,
+      "loss": 1.693,
+      "step": 900
+    },
+    {
+      "epoch": 0.12022951694689085,
+      "grad_norm": 1.2626260309636503,
+      "learning_rate": 1.957623762150471e-05,
+      "loss": 1.6766,
+      "step": 901
+    },
+    {
+      "epoch": 0.1203629570322925,
+      "grad_norm": 1.6130682344451919,
+      "learning_rate": 1.9574991923058735e-05,
+      "loss": 1.7133,
+      "step": 902
+    },
+    {
+      "epoch": 0.12049639711769415,
+      "grad_norm": 1.1868637434354896,
+      "learning_rate": 1.957374443611328e-05,
+      "loss": 1.719,
+      "step": 903
+    },
+    {
+      "epoch": 0.12062983720309581,
+      "grad_norm": 1.1629455637439938,
+      "learning_rate": 1.9572495160901365e-05,
+      "loss": 1.6718,
+      "step": 904
+    },
+    {
+      "epoch": 0.12076327728849746,
+      "grad_norm": 1.2126155193246908,
+      "learning_rate": 1.9571244097656343e-05,
+      "loss": 1.6731,
+      "step": 905
+    },
+    {
+      "epoch": 0.12089671737389912,
+      "grad_norm": 1.3426037774841502,
+      "learning_rate": 1.9569991246611897e-05,
+      "loss": 1.7082,
+      "step": 906
+    },
+    {
+      "epoch": 0.12103015745930078,
+      "grad_norm": 1.237669168766994,
+      "learning_rate": 1.9568736608002045e-05,
+      "loss": 1.698,
+      "step": 907
+    },
+    {
+      "epoch": 0.12116359754470243,
+      "grad_norm": 1.2019255673482718,
+      "learning_rate": 1.9567480182061134e-05,
+      "loss": 1.7086,
+      "step": 908
+    },
+    {
+      "epoch": 0.12129703763010408,
+      "grad_norm": 1.4253857823504195,
+      "learning_rate": 1.956622196902386e-05,
+      "loss": 1.7049,
+      "step": 909
+    },
+    {
+      "epoch": 0.12143047771550573,
+      "grad_norm": 1.681519114291726,
+      "learning_rate": 1.9564961969125235e-05,
+      "loss": 1.6735,
+      "step": 910
+    },
+    {
+      "epoch": 0.12156391780090739,
+      "grad_norm": 1.2989313363354473,
+      "learning_rate": 1.9563700182600612e-05,
+      "loss": 1.7477,
+      "step": 911
+    },
+    {
+      "epoch": 0.12169735788630905,
+      "grad_norm": 1.1509458671490507,
+      "learning_rate": 1.9562436609685685e-05,
+      "loss": 1.7121,
+      "step": 912
+    },
+    {
+      "epoch": 0.1218307979717107,
+      "grad_norm": 1.234465309616413,
+      "learning_rate": 1.9561171250616472e-05,
+      "loss": 1.7032,
+      "step": 913
+    },
+    {
+      "epoch": 0.12196423805711236,
+      "grad_norm": 1.286630945624901,
+      "learning_rate": 1.955990410562933e-05,
+      "loss": 1.6834,
+      "step": 914
+    },
+    {
+      "epoch": 0.12209767814251402,
+      "grad_norm": 1.2427074764646417,
+      "learning_rate": 1.9558635174960942e-05,
+      "loss": 1.6784,
+      "step": 915
+    },
+    {
+      "epoch": 0.12223111822791567,
+      "grad_norm": 1.2023142218836853,
+      "learning_rate": 1.9557364458848334e-05,
+      "loss": 1.6907,
+      "step": 916
+    },
+    {
+      "epoch": 0.12236455831331731,
+      "grad_norm": 1.1692669558611244,
+      "learning_rate": 1.955609195752886e-05,
+      "loss": 1.6604,
+      "step": 917
+    },
+    {
+      "epoch": 0.12249799839871897,
+      "grad_norm": 1.441083824408625,
+      "learning_rate": 1.9554817671240205e-05,
+      "loss": 1.7268,
+      "step": 918
+    },
+    {
+      "epoch": 0.12263143848412063,
+      "grad_norm": 1.1637128804063606,
+      "learning_rate": 1.9553541600220395e-05,
+      "loss": 1.7105,
+      "step": 919
+    },
+    {
+      "epoch": 0.12276487856952228,
+      "grad_norm": 1.1948097801622213,
+      "learning_rate": 1.955226374470779e-05,
+      "loss": 1.6994,
+      "step": 920
+    },
+    {
+      "epoch": 0.12289831865492394,
+      "grad_norm": 1.19037198249763,
+      "learning_rate": 1.955098410494107e-05,
+      "loss": 1.6805,
+      "step": 921
+    },
+    {
+      "epoch": 0.1230317587403256,
+      "grad_norm": 1.1965278157578167,
+      "learning_rate": 1.954970268115926e-05,
+      "loss": 1.7126,
+      "step": 922
+    },
+    {
+      "epoch": 0.12316519882572725,
+      "grad_norm": 12.043177426547418,
+      "learning_rate": 1.954841947360172e-05,
+      "loss": 1.6847,
+      "step": 923
+    },
+    {
+      "epoch": 0.1232986389111289,
+      "grad_norm": 1.4171849613590173,
+      "learning_rate": 1.9547134482508135e-05,
+      "loss": 1.6773,
+      "step": 924
+    },
+    {
+      "epoch": 0.12343207899653055,
+      "grad_norm": 1.277439153377438,
+      "learning_rate": 1.9545847708118524e-05,
+      "loss": 1.654,
+      "step": 925
+    },
+    {
+      "epoch": 0.12356551908193221,
+      "grad_norm": 1.5344343949821269,
+      "learning_rate": 1.954455915067325e-05,
+      "loss": 1.714,
+      "step": 926
+    },
+    {
+      "epoch": 0.12369895916733387,
+      "grad_norm": 1.3273336556361204,
+      "learning_rate": 1.954326881041299e-05,
+      "loss": 1.7272,
+      "step": 927
+    },
+    {
+      "epoch": 0.12383239925273552,
+      "grad_norm": 1.2408609710996044,
+      "learning_rate": 1.9541976687578773e-05,
+      "loss": 1.7073,
+      "step": 928
+    },
+    {
+      "epoch": 0.12396583933813718,
+      "grad_norm": 1.3534841024628121,
+      "learning_rate": 1.954068278241195e-05,
+      "loss": 1.7464,
+      "step": 929
+    },
+    {
+      "epoch": 0.12409927942353884,
+      "grad_norm": 1.8710842856544097,
+      "learning_rate": 1.9539387095154207e-05,
+      "loss": 1.6445,
+      "step": 930
+    },
+    {
+      "epoch": 0.12423271950894049,
+      "grad_norm": 1.3643772964473504,
+      "learning_rate": 1.9538089626047566e-05,
+      "loss": 1.7005,
+      "step": 931
+    },
+    {
+      "epoch": 0.12436615959434213,
+      "grad_norm": 1.2248623778543628,
+      "learning_rate": 1.953679037533438e-05,
+      "loss": 1.6574,
+      "step": 932
+    },
+    {
+      "epoch": 0.12449959967974379,
+      "grad_norm": 1.3022000288924167,
+      "learning_rate": 1.9535489343257326e-05,
+      "loss": 1.7176,
+      "step": 933
+    },
+    {
+      "epoch": 0.12463303976514545,
+      "grad_norm": 1.4725976531996385,
+      "learning_rate": 1.9534186530059434e-05,
+      "loss": 1.6851,
+      "step": 934
+    },
+    {
+      "epoch": 0.1247664798505471,
+      "grad_norm": 1.2400512724415127,
+      "learning_rate": 1.9532881935984046e-05,
+      "loss": 1.688,
+      "step": 935
+    },
+    {
+      "epoch": 0.12489991993594876,
+      "grad_norm": 1.4145930106236353,
+      "learning_rate": 1.9531575561274852e-05,
+      "loss": 1.6835,
+      "step": 936
+    },
+    {
+      "epoch": 0.1250333600213504,
+      "grad_norm": 1.3793980244786415,
+      "learning_rate": 1.953026740617586e-05,
+      "loss": 1.7196,
+      "step": 937
+    },
+    {
+      "epoch": 0.12516680010675207,
+      "grad_norm": 13.216726991500538,
+      "learning_rate": 1.9528957470931424e-05,
+      "loss": 1.7525,
+      "step": 938
+    },
+    {
+      "epoch": 0.12530024019215372,
+      "grad_norm": 1.3237742593449273,
+      "learning_rate": 1.9527645755786224e-05,
+      "loss": 1.6883,
+      "step": 939
+    },
+    {
+      "epoch": 0.1254336802775554,
+      "grad_norm": 1.445191972771836,
+      "learning_rate": 1.9526332260985275e-05,
+      "loss": 1.6686,
+      "step": 940
+    },
+    {
+      "epoch": 0.12556712036295703,
+      "grad_norm": 1.5185282218871532,
+      "learning_rate": 1.952501698677392e-05,
+      "loss": 1.6581,
+      "step": 941
+    },
+    {
+      "epoch": 0.1257005604483587,
+      "grad_norm": 1.3124983698296924,
+      "learning_rate": 1.9523699933397834e-05,
+      "loss": 1.6544,
+      "step": 942
+    },
+    {
+      "epoch": 0.12583400053376034,
+      "grad_norm": 1.183777382042402,
+      "learning_rate": 1.9522381101103038e-05,
+      "loss": 1.7612,
+      "step": 943
+    },
+    {
+      "epoch": 0.12596744061916199,
+      "grad_norm": 1.2415393099346195,
+      "learning_rate": 1.9521060490135865e-05,
+      "loss": 1.7291,
+      "step": 944
+    },
+    {
+      "epoch": 0.12610088070456366,
+      "grad_norm": 1.176856792014906,
+      "learning_rate": 1.9519738100742995e-05,
+      "loss": 1.6638,
+      "step": 945
+    },
+    {
+      "epoch": 0.1262343207899653,
+      "grad_norm": 1.1487711372941813,
+      "learning_rate": 1.9518413933171432e-05,
+      "loss": 1.6464,
+      "step": 946
+    },
+    {
+      "epoch": 0.12636776087536697,
+      "grad_norm": 1.4297554387238383,
+      "learning_rate": 1.9517087987668522e-05,
+      "loss": 1.6069,
+      "step": 947
+    },
+    {
+      "epoch": 0.1265012009607686,
+      "grad_norm": 1.31215843034171,
+      "learning_rate": 1.951576026448193e-05,
+      "loss": 1.689,
+      "step": 948
+    },
+    {
+      "epoch": 0.12663464104617028,
+      "grad_norm": 1.1689420027444986,
+      "learning_rate": 1.951443076385966e-05,
+      "loss": 1.6715,
+      "step": 949
+    },
+    {
+      "epoch": 0.12676808113157192,
+      "grad_norm": 1.680542791799617,
+      "learning_rate": 1.951309948605005e-05,
+      "loss": 1.6546,
+      "step": 950
+    },
+    {
+      "epoch": 0.12690152121697357,
+      "grad_norm": 1.1797862555326082,
+      "learning_rate": 1.9511766431301766e-05,
+      "loss": 1.7215,
+      "step": 951
+    },
+    {
+      "epoch": 0.12703496130237524,
+      "grad_norm": 1.5427071574060558,
+      "learning_rate": 1.951043159986381e-05,
+      "loss": 1.6322,
+      "step": 952
+    },
+    {
+      "epoch": 0.12716840138777688,
+      "grad_norm": 1.1779500188493681,
+      "learning_rate": 1.950909499198551e-05,
+      "loss": 1.6812,
+      "step": 953
+    },
+    {
+      "epoch": 0.12730184147317855,
+      "grad_norm": 1.1913258200020012,
+      "learning_rate": 1.950775660791653e-05,
+      "loss": 1.6755,
+      "step": 954
+    },
+    {
+      "epoch": 0.1274352815585802,
+      "grad_norm": 1.232490173323732,
+      "learning_rate": 1.950641644790687e-05,
+      "loss": 1.7031,
+      "step": 955
+    },
+    {
+      "epoch": 0.12756872164398186,
+      "grad_norm": 1.327617648590312,
+      "learning_rate": 1.9505074512206847e-05,
+      "loss": 1.6455,
+      "step": 956
+    },
+    {
+      "epoch": 0.1277021617293835,
+      "grad_norm": 1.1740821011806148,
+      "learning_rate": 1.9503730801067125e-05,
+      "loss": 1.6913,
+      "step": 957
+    },
+    {
+      "epoch": 0.12783560181478515,
+      "grad_norm": 1.3812646322593256,
+      "learning_rate": 1.9502385314738697e-05,
+      "loss": 1.6215,
+      "step": 958
+    },
+    {
+      "epoch": 0.12796904190018682,
+      "grad_norm": 1.3155996801138305,
+      "learning_rate": 1.9501038053472877e-05,
+      "loss": 1.6518,
+      "step": 959
+    },
+    {
+      "epoch": 0.12810248198558846,
+      "grad_norm": 1.3349048294702728,
+      "learning_rate": 1.9499689017521323e-05,
+      "loss": 1.6609,
+      "step": 960
+    },
+    {
+      "epoch": 0.12823592207099013,
+      "grad_norm": 1.2671901915804749,
+      "learning_rate": 1.9498338207136022e-05,
+      "loss": 1.725,
+      "step": 961
+    },
+    {
+      "epoch": 0.12836936215639178,
+      "grad_norm": 1.1359327570816697,
+      "learning_rate": 1.949698562256928e-05,
+      "loss": 1.6597,
+      "step": 962
+    },
+    {
+      "epoch": 0.12850280224179345,
+      "grad_norm": 1.2333203786654017,
+      "learning_rate": 1.9495631264073758e-05,
+      "loss": 1.6659,
+      "step": 963
+    },
+    {
+      "epoch": 0.1286362423271951,
+      "grad_norm": 1.2554849794053347,
+      "learning_rate": 1.9494275131902423e-05,
+      "loss": 1.6749,
+      "step": 964
+    },
+    {
+      "epoch": 0.12876968241259673,
+      "grad_norm": 1.3230800031319025,
+      "learning_rate": 1.949291722630859e-05,
+      "loss": 1.6923,
+      "step": 965
+    },
+    {
+      "epoch": 0.1289031224979984,
+      "grad_norm": 2.1623618851588327,
+      "learning_rate": 1.9491557547545903e-05,
+      "loss": 1.6905,
+      "step": 966
+    },
+    {
+      "epoch": 0.12903656258340004,
+      "grad_norm": 1.3456119660809565,
+      "learning_rate": 1.9490196095868328e-05,
+      "loss": 1.7063,
+      "step": 967
+    },
+    {
+      "epoch": 0.12917000266880171,
+      "grad_norm": 1.600288717290206,
+      "learning_rate": 1.9488832871530173e-05,
+      "loss": 1.7006,
+      "step": 968
+    },
+    {
+      "epoch": 0.12930344275420336,
+      "grad_norm": 1.3672986818133261,
+      "learning_rate": 1.9487467874786076e-05,
+      "loss": 1.7197,
+      "step": 969
+    },
+    {
+      "epoch": 0.12943688283960503,
+      "grad_norm": 1.2760213827762075,
+      "learning_rate": 1.9486101105890993e-05,
+      "loss": 1.6708,
+      "step": 970
+    },
+    {
+      "epoch": 0.12957032292500667,
+      "grad_norm": 1.393135659234334,
+      "learning_rate": 1.948473256510023e-05,
+      "loss": 1.6924,
+      "step": 971
+    },
+    {
+      "epoch": 0.1297037630104083,
+      "grad_norm": 1.367984927030445,
+      "learning_rate": 1.948336225266941e-05,
+      "loss": 1.7087,
+      "step": 972
+    },
+    {
+      "epoch": 0.12983720309580998,
+      "grad_norm": 1.2953603648125673,
+      "learning_rate": 1.9481990168854494e-05,
+      "loss": 1.7151,
+      "step": 973
+    },
+    {
+      "epoch": 0.12997064318121163,
+      "grad_norm": 1.167269554903983,
+      "learning_rate": 1.9480616313911774e-05,
+      "loss": 1.6787,
+      "step": 974
+    },
+    {
+      "epoch": 0.1301040832666133,
+      "grad_norm": 1.246260379852551,
+      "learning_rate": 1.9479240688097864e-05,
+      "loss": 1.7456,
+      "step": 975
+    },
+    {
+      "epoch": 0.13023752335201494,
+      "grad_norm": 1.2725941500644802,
+      "learning_rate": 1.9477863291669718e-05,
+      "loss": 1.6552,
+      "step": 976
+    },
+    {
+      "epoch": 0.1303709634374166,
+      "grad_norm": 1.2084106713506604,
+      "learning_rate": 1.9476484124884622e-05,
+      "loss": 1.6639,
+      "step": 977
+    },
+    {
+      "epoch": 0.13050440352281825,
+      "grad_norm": 1.194547259427426,
+      "learning_rate": 1.9475103188000183e-05,
+      "loss": 1.6877,
+      "step": 978
+    },
+    {
+      "epoch": 0.13063784360821992,
+      "grad_norm": 1.2780352392355023,
+      "learning_rate": 1.9473720481274347e-05,
+      "loss": 1.6605,
+      "step": 979
+    },
+    {
+      "epoch": 0.13077128369362157,
+      "grad_norm": 1.5045389202002308,
+      "learning_rate": 1.9472336004965387e-05,
+      "loss": 1.7074,
+      "step": 980
+    },
+    {
+      "epoch": 0.1309047237790232,
+      "grad_norm": 1.2421804342522536,
+      "learning_rate": 1.947094975933191e-05,
+      "loss": 1.686,
+      "step": 981
+    },
+    {
+      "epoch": 0.13103816386442488,
+      "grad_norm": 1.1426071394085213,
+      "learning_rate": 1.9469561744632845e-05,
+      "loss": 1.6592,
+      "step": 982
+    },
+    {
+      "epoch": 0.13117160394982652,
+      "grad_norm": 1.6999618031943038,
+      "learning_rate": 1.9468171961127464e-05,
+      "loss": 1.7108,
+      "step": 983
+    },
+    {
+      "epoch": 0.1313050440352282,
+      "grad_norm": 1.2427085932088913,
+      "learning_rate": 1.946678040907536e-05,
+      "loss": 1.657,
+      "step": 984
+    },
+    {
+      "epoch": 0.13143848412062983,
+      "grad_norm": 8.676443998708113,
+      "learning_rate": 1.9465387088736455e-05,
+      "loss": 1.6552,
+      "step": 985
+    },
+    {
+      "epoch": 0.1315719242060315,
+      "grad_norm": 1.6685163662088924,
+      "learning_rate": 1.9463992000371014e-05,
+      "loss": 1.6877,
+      "step": 986
+    },
+    {
+      "epoch": 0.13170536429143315,
+      "grad_norm": 1.4363009160454192,
+      "learning_rate": 1.9462595144239616e-05,
+      "loss": 1.7549,
+      "step": 987
+    },
+    {
+      "epoch": 0.1318388043768348,
+      "grad_norm": 1.4849169479024855,
+      "learning_rate": 1.946119652060318e-05,
+      "loss": 1.6377,
+      "step": 988
+    },
+    {
+      "epoch": 0.13197224446223646,
+      "grad_norm": 1.384439488826471,
+      "learning_rate": 1.9459796129722962e-05,
+      "loss": 1.69,
+      "step": 989
+    },
+    {
+      "epoch": 0.1321056845476381,
+      "grad_norm": 1.486350400208871,
+      "learning_rate": 1.9458393971860522e-05,
+      "loss": 1.6643,
+      "step": 990
+    },
+    {
+      "epoch": 0.13223912463303977,
+      "grad_norm": 1.4827478161059948,
+      "learning_rate": 1.9456990047277777e-05,
+      "loss": 1.6915,
+      "step": 991
+    },
+    {
+      "epoch": 0.13237256471844142,
+      "grad_norm": 1.2049573199404926,
+      "learning_rate": 1.9455584356236965e-05,
+      "loss": 1.6686,
+      "step": 992
+    },
+    {
+      "epoch": 0.13250600480384309,
+      "grad_norm": 1.490030676305235,
+      "learning_rate": 1.9454176899000653e-05,
+      "loss": 1.7358,
+      "step": 993
+    },
+    {
+      "epoch": 0.13263944488924473,
+      "grad_norm": 1.1805251966858894,
+      "learning_rate": 1.9452767675831733e-05,
+      "loss": 1.6745,
+      "step": 994
+    },
+    {
+      "epoch": 0.13277288497464637,
+      "grad_norm": 1.4646671341281234,
+      "learning_rate": 1.9451356686993435e-05,
+      "loss": 1.671,
+      "step": 995
+    },
+    {
+      "epoch": 0.13290632506004804,
+      "grad_norm": 1.2502496308048068,
+      "learning_rate": 1.9449943932749316e-05,
+      "loss": 1.6041,
+      "step": 996
+    },
+    {
+      "epoch": 0.13303976514544968,
+      "grad_norm": 1.3302442965706283,
+      "learning_rate": 1.9448529413363264e-05,
+      "loss": 1.692,
+      "step": 997
+    },
+    {
+      "epoch": 0.13317320523085135,
+      "grad_norm": 1.5733523617366565,
+      "learning_rate": 1.944711312909949e-05,
+      "loss": 1.6779,
+      "step": 998
+    },
+    {
+      "epoch": 0.133306645316253,
+      "grad_norm": 1.2021852120833718,
+      "learning_rate": 1.9445695080222543e-05,
+      "loss": 1.688,
+      "step": 999
+    },
+    {
+      "epoch": 0.13344008540165467,
+      "grad_norm": 1.3704807104442995,
+      "learning_rate": 1.9444275266997302e-05,
+      "loss": 1.6618,
+      "step": 1000
+    },
+    {
+      "epoch": 0.1335735254870563,
+      "grad_norm": 1.1311995894978009,
+      "learning_rate": 1.9442853689688965e-05,
+      "loss": 1.6144,
+      "step": 1001
+    },
+    {
+      "epoch": 0.13370696557245795,
+      "grad_norm": 1.5372526742454895,
+      "learning_rate": 1.9441430348563072e-05,
+      "loss": 1.7336,
+      "step": 1002
+    },
+    {
+      "epoch": 0.13384040565785962,
+      "grad_norm": 1.3090351266009799,
+      "learning_rate": 1.9440005243885482e-05,
+      "loss": 1.6752,
+      "step": 1003
+    },
+    {
+      "epoch": 0.13397384574326127,
+      "grad_norm": 1.313411350120398,
+      "learning_rate": 1.943857837592239e-05,
+      "loss": 1.6838,
+      "step": 1004
+    },
+    {
+      "epoch": 0.13410728582866294,
+      "grad_norm": 1.3955032805375227,
+      "learning_rate": 1.943714974494032e-05,
+      "loss": 1.7377,
+      "step": 1005
+    },
+    {
+      "epoch": 0.13424072591406458,
+      "grad_norm": 1.1613148391890242,
+      "learning_rate": 1.943571935120613e-05,
+      "loss": 1.6595,
+      "step": 1006
+    },
+    {
+      "epoch": 0.13437416599946625,
+      "grad_norm": 1.5018393485227763,
+      "learning_rate": 1.943428719498699e-05,
+      "loss": 1.7533,
+      "step": 1007
+    },
+    {
+      "epoch": 0.1345076060848679,
+      "grad_norm": 1.1758602862709078,
+      "learning_rate": 1.9432853276550412e-05,
+      "loss": 1.6804,
+      "step": 1008
+    },
+    {
+      "epoch": 0.13464104617026956,
+      "grad_norm": 2.1854421038226075,
+      "learning_rate": 1.9431417596164246e-05,
+      "loss": 1.7315,
+      "step": 1009
+    },
+    {
+      "epoch": 0.1347744862556712,
+      "grad_norm": 1.3381407737075066,
+      "learning_rate": 1.942998015409665e-05,
+      "loss": 1.6721,
+      "step": 1010
+    },
+    {
+      "epoch": 0.13490792634107285,
+      "grad_norm": 1.2547528294212573,
+      "learning_rate": 1.9428540950616127e-05,
+      "loss": 1.6782,
+      "step": 1011
+    },
+    {
+      "epoch": 0.13504136642647452,
+      "grad_norm": 5.939952891644821,
+      "learning_rate": 1.94270999859915e-05,
+      "loss": 1.6811,
+      "step": 1012
+    },
+    {
+      "epoch": 0.13517480651187616,
+      "grad_norm": 1.3380178004751162,
+      "learning_rate": 1.942565726049193e-05,
+      "loss": 1.6712,
+      "step": 1013
+    },
+    {
+      "epoch": 0.13530824659727783,
+      "grad_norm": 1.2408694717484239,
+      "learning_rate": 1.9424212774386894e-05,
+      "loss": 1.7259,
+      "step": 1014
+    },
+    {
+      "epoch": 0.13544168668267947,
+      "grad_norm": 1.2756143544585534,
+      "learning_rate": 1.9422766527946217e-05,
+      "loss": 1.6631,
+      "step": 1015
+    },
+    {
+      "epoch": 0.13557512676808114,
+      "grad_norm": 1.4810902782091733,
+      "learning_rate": 1.942131852144003e-05,
+      "loss": 1.6927,
+      "step": 1016
+    },
+    {
+      "epoch": 0.1357085668534828,
+      "grad_norm": 1.2419032087935538,
+      "learning_rate": 1.941986875513881e-05,
+      "loss": 1.6898,
+      "step": 1017
+    },
+    {
+      "epoch": 0.13584200693888443,
+      "grad_norm": 1.2663899524703148,
+      "learning_rate": 1.9418417229313357e-05,
+      "loss": 1.679,
+      "step": 1018
+    },
+    {
+      "epoch": 0.1359754470242861,
+      "grad_norm": 1.6857267169996077,
+      "learning_rate": 1.9416963944234795e-05,
+      "loss": 1.6354,
+      "step": 1019
+    },
+    {
+      "epoch": 0.13610888710968774,
+      "grad_norm": 1.4554351634221496,
+      "learning_rate": 1.9415508900174587e-05,
+      "loss": 1.7462,
+      "step": 1020
+    },
+    {
+      "epoch": 0.1362423271950894,
+      "grad_norm": 1.2229508649241836,
+      "learning_rate": 1.941405209740452e-05,
+      "loss": 1.6975,
+      "step": 1021
+    },
+    {
+      "epoch": 0.13637576728049106,
+      "grad_norm": 1.8919520969087211,
+      "learning_rate": 1.94125935361967e-05,
+      "loss": 1.7258,
+      "step": 1022
+    },
+    {
+      "epoch": 0.13650920736589273,
+      "grad_norm": 1.2669785367995678,
+      "learning_rate": 1.9411133216823573e-05,
+      "loss": 1.7269,
+      "step": 1023
+    },
+    {
+      "epoch": 0.13664264745129437,
+      "grad_norm": 1.394093842588581,
+      "learning_rate": 1.9409671139557913e-05,
+      "loss": 1.6975,
+      "step": 1024
+    },
+    {
+      "epoch": 0.136776087536696,
+      "grad_norm": 1.382684928389381,
+      "learning_rate": 1.9408207304672815e-05,
+      "loss": 1.667,
+      "step": 1025
+    },
+    {
+      "epoch": 0.13690952762209768,
+      "grad_norm": 1.1652111588255187,
+      "learning_rate": 1.940674171244171e-05,
+      "loss": 1.6379,
+      "step": 1026
+    },
+    {
+      "epoch": 0.13704296770749932,
+      "grad_norm": 1.3038078568540548,
+      "learning_rate": 1.9405274363138353e-05,
+      "loss": 1.6447,
+      "step": 1027
+    },
+    {
+      "epoch": 0.137176407792901,
+      "grad_norm": 1.1387426323073633,
+      "learning_rate": 1.940380525703683e-05,
+      "loss": 1.6675,
+      "step": 1028
+    },
+    {
+      "epoch": 0.13730984787830264,
+      "grad_norm": 1.474771378157035,
+      "learning_rate": 1.940233439441155e-05,
+      "loss": 1.7026,
+      "step": 1029
+    },
+    {
+      "epoch": 0.1374432879637043,
+      "grad_norm": 1.3414858835240848,
+      "learning_rate": 1.9400861775537253e-05,
+      "loss": 1.6212,
+      "step": 1030
+    },
+    {
+      "epoch": 0.13757672804910595,
+      "grad_norm": 1.4550778747404767,
+      "learning_rate": 1.939938740068901e-05,
+      "loss": 1.6785,
+      "step": 1031
+    },
+    {
+      "epoch": 0.1377101681345076,
+      "grad_norm": 1.3556680615235186,
+      "learning_rate": 1.939791127014222e-05,
+      "loss": 1.6597,
+      "step": 1032
+    },
+    {
+      "epoch": 0.13784360821990926,
+      "grad_norm": 1.4224278910538175,
+      "learning_rate": 1.93964333841726e-05,
+      "loss": 1.6917,
+      "step": 1033
+    },
+    {
+      "epoch": 0.1379770483053109,
+      "grad_norm": 2.2986502944753275,
+      "learning_rate": 1.9394953743056205e-05,
+      "loss": 1.6871,
+      "step": 1034
+    },
+    {
+      "epoch": 0.13811048839071258,
+      "grad_norm": 2.0365277465356924,
+      "learning_rate": 1.939347234706942e-05,
+      "loss": 1.7383,
+      "step": 1035
+    },
+    {
+      "epoch": 0.13824392847611422,
+      "grad_norm": 1.4908556334356526,
+      "learning_rate": 1.9391989196488947e-05,
+      "loss": 1.6577,
+      "step": 1036
+    },
+    {
+      "epoch": 0.1383773685615159,
+      "grad_norm": 1.2426015406392708,
+      "learning_rate": 1.9390504291591825e-05,
+      "loss": 1.6762,
+      "step": 1037
+    },
+    {
+      "epoch": 0.13851080864691753,
+      "grad_norm": 1.1622345855922083,
+      "learning_rate": 1.9389017632655417e-05,
+      "loss": 1.6452,
+      "step": 1038
+    },
+    {
+      "epoch": 0.13864424873231918,
+      "grad_norm": 1.2036449724168792,
+      "learning_rate": 1.938752921995741e-05,
+      "loss": 1.6652,
+      "step": 1039
+    },
+    {
+      "epoch": 0.13877768881772085,
+      "grad_norm": 1.3193413331431836,
+      "learning_rate": 1.938603905377583e-05,
+      "loss": 1.7153,
+      "step": 1040
+    },
+    {
+      "epoch": 0.1389111289031225,
+      "grad_norm": 3.9652856357976005,
+      "learning_rate": 1.9384547134389017e-05,
+      "loss": 1.6998,
+      "step": 1041
+    },
+    {
+      "epoch": 0.13904456898852416,
+      "grad_norm": 1.2832120473397637,
+      "learning_rate": 1.9383053462075645e-05,
+      "loss": 1.6865,
+      "step": 1042
+    },
+    {
+      "epoch": 0.1391780090739258,
+      "grad_norm": 1.9488157590841544,
+      "learning_rate": 1.938155803711472e-05,
+      "loss": 1.7429,
+      "step": 1043
+    },
+    {
+      "epoch": 0.13931144915932747,
+      "grad_norm": 1.334661146454789,
+      "learning_rate": 1.9380060859785562e-05,
+      "loss": 1.687,
+      "step": 1044
+    },
+    {
+      "epoch": 0.13944488924472911,
+      "grad_norm": 1.254965676826134,
+      "learning_rate": 1.937856193036783e-05,
+      "loss": 1.647,
+      "step": 1045
+    },
+    {
+      "epoch": 0.13957832933013078,
+      "grad_norm": 1.3924358256427258,
+      "learning_rate": 1.9377061249141515e-05,
+      "loss": 1.7171,
+      "step": 1046
+    },
+    {
+      "epoch": 0.13971176941553243,
+      "grad_norm": 1.3956910901280772,
+      "learning_rate": 1.9375558816386915e-05,
+      "loss": 1.743,
+      "step": 1047
+    },
+    {
+      "epoch": 0.13984520950093407,
+      "grad_norm": 1.2701133772552229,
+      "learning_rate": 1.9374054632384677e-05,
+      "loss": 1.7041,
+      "step": 1048
+    },
+    {
+      "epoch": 0.13997864958633574,
+      "grad_norm": 1.3350393841513681,
+      "learning_rate": 1.937254869741576e-05,
+      "loss": 1.6846,
+      "step": 1049
+    },
+    {
+      "epoch": 0.14011208967173738,
+      "grad_norm": 1.4627388869419586,
+      "learning_rate": 1.9371041011761456e-05,
+      "loss": 1.6207,
+      "step": 1050
+    },
+    {
+      "epoch": 0.14024552975713905,
+      "grad_norm": 1.4866992503394356,
+      "learning_rate": 1.9369531575703388e-05,
+      "loss": 1.7129,
+      "step": 1051
+    },
+    {
+      "epoch": 0.1403789698425407,
+      "grad_norm": 1.1685028747662984,
+      "learning_rate": 1.9368020389523493e-05,
+      "loss": 1.668,
+      "step": 1052
+    },
+    {
+      "epoch": 0.14051240992794237,
+      "grad_norm": 1.3215015297643653,
+      "learning_rate": 1.936650745350405e-05,
+      "loss": 1.7145,
+      "step": 1053
+    },
+    {
+      "epoch": 0.140645850013344,
+      "grad_norm": 1.3693292355410813,
+      "learning_rate": 1.9364992767927656e-05,
+      "loss": 1.701,
+      "step": 1054
+    },
+    {
+      "epoch": 0.14077929009874565,
+      "grad_norm": 1.2258997785581536,
+      "learning_rate": 1.9363476333077237e-05,
+      "loss": 1.7085,
+      "step": 1055
+    },
+    {
+      "epoch": 0.14091273018414732,
+      "grad_norm": 1.2455180226057498,
+      "learning_rate": 1.9361958149236045e-05,
+      "loss": 1.7652,
+      "step": 1056
+    },
+    {
+      "epoch": 0.14104617026954896,
+      "grad_norm": 1.1398881313468456,
+      "learning_rate": 1.9360438216687663e-05,
+      "loss": 1.6521,
+      "step": 1057
+    },
+    {
+      "epoch": 0.14117961035495064,
+      "grad_norm": 1.4032373423726727,
+      "learning_rate": 1.9358916535715995e-05,
+      "loss": 1.6822,
+      "step": 1058
+    },
+    {
+      "epoch": 0.14131305044035228,
+      "grad_norm": 1.2386258022282814,
+      "learning_rate": 1.9357393106605273e-05,
+      "loss": 1.6565,
+      "step": 1059
+    },
+    {
+      "epoch": 0.14144649052575395,
+      "grad_norm": 1.2563917204753323,
+      "learning_rate": 1.9355867929640054e-05,
+      "loss": 1.7224,
+      "step": 1060
+    },
+    {
+      "epoch": 0.1415799306111556,
+      "grad_norm": 1.1600878491451077,
+      "learning_rate": 1.9354341005105228e-05,
+      "loss": 1.6831,
+      "step": 1061
+    },
+    {
+      "epoch": 0.14171337069655723,
+      "grad_norm": 1.1898921958138153,
+      "learning_rate": 1.9352812333286006e-05,
+      "loss": 1.7034,
+      "step": 1062
+    },
+    {
+      "epoch": 0.1418468107819589,
+      "grad_norm": 1.3776077175515267,
+      "learning_rate": 1.9351281914467924e-05,
+      "loss": 1.6558,
+      "step": 1063
+    },
+    {
+      "epoch": 0.14198025086736055,
+      "grad_norm": 1.4565706718782248,
+      "learning_rate": 1.934974974893685e-05,
+      "loss": 1.7167,
+      "step": 1064
+    },
+    {
+      "epoch": 0.14211369095276222,
+      "grad_norm": 1.2442621324452396,
+      "learning_rate": 1.9348215836978978e-05,
+      "loss": 1.6459,
+      "step": 1065
+    },
+    {
+      "epoch": 0.14224713103816386,
+      "grad_norm": 1.1267333234718482,
+      "learning_rate": 1.9346680178880813e-05,
+      "loss": 1.7133,
+      "step": 1066
+    },
+    {
+      "epoch": 0.14238057112356553,
+      "grad_norm": 1.1628045406596315,
+      "learning_rate": 1.9345142774929214e-05,
+      "loss": 1.6475,
+      "step": 1067
+    },
+    {
+      "epoch": 0.14251401120896717,
+      "grad_norm": 1.1018713280842254,
+      "learning_rate": 1.9343603625411338e-05,
+      "loss": 1.5936,
+      "step": 1068
+    },
+    {
+      "epoch": 0.14264745129436882,
+      "grad_norm": 1.193003889508144,
+      "learning_rate": 1.9342062730614688e-05,
+      "loss": 1.6788,
+      "step": 1069
+    },
+    {
+      "epoch": 0.14278089137977049,
+      "grad_norm": 1.1343045579344455,
+      "learning_rate": 1.9340520090827086e-05,
+      "loss": 1.6561,
+      "step": 1070
+    },
+    {
+      "epoch": 0.14291433146517213,
+      "grad_norm": 1.1710425209498292,
+      "learning_rate": 1.9338975706336672e-05,
+      "loss": 1.5907,
+      "step": 1071
+    },
+    {
+      "epoch": 0.1430477715505738,
+      "grad_norm": 1.3817597541347761,
+      "learning_rate": 1.9337429577431926e-05,
+      "loss": 1.7147,
+      "step": 1072
+    },
+    {
+      "epoch": 0.14318121163597544,
+      "grad_norm": 1.1573508951337832,
+      "learning_rate": 1.9335881704401642e-05,
+      "loss": 1.698,
+      "step": 1073
+    },
+    {
+      "epoch": 0.1433146517213771,
+      "grad_norm": 1.0912396515600804,
+      "learning_rate": 1.9334332087534956e-05,
+      "loss": 1.6756,
+      "step": 1074
+    },
+    {
+      "epoch": 0.14344809180677875,
+      "grad_norm": 1.1837140980523393,
+      "learning_rate": 1.9332780727121306e-05,
+      "loss": 1.6908,
+      "step": 1075
+    },
+    {
+      "epoch": 0.1435815318921804,
+      "grad_norm": 9.673515773086278,
+      "learning_rate": 1.9331227623450475e-05,
+      "loss": 1.71,
+      "step": 1076
+    },
+    {
+      "epoch": 0.14371497197758207,
+      "grad_norm": 1.3759562691897052,
+      "learning_rate": 1.9329672776812563e-05,
+      "loss": 1.7094,
+      "step": 1077
+    },
+    {
+      "epoch": 0.1438484120629837,
+      "grad_norm": 1.1597684961716581,
+      "learning_rate": 1.9328116187498e-05,
+      "loss": 1.6868,
+      "step": 1078
+    },
+    {
+      "epoch": 0.14398185214838538,
+      "grad_norm": 1.861180144739059,
+      "learning_rate": 1.9326557855797537e-05,
+      "loss": 1.695,
+      "step": 1079
+    },
+    {
+      "epoch": 0.14411529223378702,
+      "grad_norm": 1.3971061309442314,
+      "learning_rate": 1.9324997782002258e-05,
+      "loss": 1.6394,
+      "step": 1080
+    },
+    {
+      "epoch": 0.1442487323191887,
+      "grad_norm": 1.1696642529732444,
+      "learning_rate": 1.9323435966403557e-05,
+      "loss": 1.6628,
+      "step": 1081
+    },
+    {
+      "epoch": 0.14438217240459034,
+      "grad_norm": 1.4363997803769488,
+      "learning_rate": 1.932187240929317e-05,
+      "loss": 1.6358,
+      "step": 1082
+    },
+    {
+      "epoch": 0.144515612489992,
+      "grad_norm": 1.2009780630510838,
+      "learning_rate": 1.932030711096315e-05,
+      "loss": 1.6941,
+      "step": 1083
+    },
+    {
+      "epoch": 0.14464905257539365,
+      "grad_norm": 1.36326083414618,
+      "learning_rate": 1.931874007170588e-05,
+      "loss": 1.7041,
+      "step": 1084
+    },
+    {
+      "epoch": 0.1447824926607953,
+      "grad_norm": 1.1982580377129664,
+      "learning_rate": 1.9317171291814058e-05,
+      "loss": 1.7533,
+      "step": 1085
+    },
+    {
+      "epoch": 0.14491593274619696,
+      "grad_norm": 1.2862961234292616,
+      "learning_rate": 1.931560077158072e-05,
+      "loss": 1.6702,
+      "step": 1086
+    },
+    {
+      "epoch": 0.1450493728315986,
+      "grad_norm": 1.1631924404688276,
+      "learning_rate": 1.9314028511299223e-05,
+      "loss": 1.6285,
+      "step": 1087
+    },
+    {
+      "epoch": 0.14518281291700028,
+      "grad_norm": 1.3994096815725487,
+      "learning_rate": 1.9312454511263242e-05,
+      "loss": 1.6882,
+      "step": 1088
+    },
+    {
+      "epoch": 0.14531625300240192,
+      "grad_norm": 1.2308714130213054,
+      "learning_rate": 1.9310878771766787e-05,
+      "loss": 1.662,
+      "step": 1089
+    },
+    {
+      "epoch": 0.1454496930878036,
+      "grad_norm": 1.2400995841145075,
+      "learning_rate": 1.9309301293104182e-05,
+      "loss": 1.6559,
+      "step": 1090
+    },
+    {
+      "epoch": 0.14558313317320523,
+      "grad_norm": 1.1648127201282061,
+      "learning_rate": 1.930772207557009e-05,
+      "loss": 1.697,
+      "step": 1091
+    },
+    {
+      "epoch": 0.14571657325860687,
+      "grad_norm": 1.1588422983053825,
+      "learning_rate": 1.9306141119459483e-05,
+      "loss": 1.7384,
+      "step": 1092
+    },
+    {
+      "epoch": 0.14585001334400854,
+      "grad_norm": 1.504230986251467,
+      "learning_rate": 1.930455842506767e-05,
+      "loss": 1.6645,
+      "step": 1093
+    },
+    {
+      "epoch": 0.1459834534294102,
+      "grad_norm": 1.4254729141660931,
+      "learning_rate": 1.9302973992690282e-05,
+      "loss": 1.7063,
+      "step": 1094
+    },
+    {
+      "epoch": 0.14611689351481186,
+      "grad_norm": 3.5366147232314296,
+      "learning_rate": 1.9301387822623266e-05,
+      "loss": 1.7281,
+      "step": 1095
+    },
+    {
+      "epoch": 0.1462503336002135,
+      "grad_norm": 1.2361609720127364,
+      "learning_rate": 1.929979991516291e-05,
+      "loss": 1.7209,
+      "step": 1096
+    },
+    {
+      "epoch": 0.14638377368561517,
+      "grad_norm": 1.095253097518337,
+      "learning_rate": 1.929821027060581e-05,
+      "loss": 1.63,
+      "step": 1097
+    },
+    {
+      "epoch": 0.1465172137710168,
+      "grad_norm": 1.2870911756015357,
+      "learning_rate": 1.9296618889248893e-05,
+      "loss": 1.6939,
+      "step": 1098
+    },
+    {
+      "epoch": 0.14665065385641846,
+      "grad_norm": 1.3716292424936098,
+      "learning_rate": 1.9295025771389414e-05,
+      "loss": 1.7245,
+      "step": 1099
+    },
+    {
+      "epoch": 0.14678409394182013,
+      "grad_norm": 1.2427470679061952,
+      "learning_rate": 1.9293430917324955e-05,
+      "loss": 1.666,
+      "step": 1100
+    },
+    {
+      "epoch": 0.14691753402722177,
+      "grad_norm": 1.460158973358874,
+      "learning_rate": 1.9291834327353403e-05,
+      "loss": 1.7147,
+      "step": 1101
+    },
+    {
+      "epoch": 0.14705097411262344,
+      "grad_norm": 1.184155803132356,
+      "learning_rate": 1.9290236001772995e-05,
+      "loss": 1.6566,
+      "step": 1102
+    },
+    {
+      "epoch": 0.14718441419802508,
+      "grad_norm": 1.2799912701156315,
+      "learning_rate": 1.9288635940882273e-05,
+      "loss": 1.6627,
+      "step": 1103
+    },
+    {
+      "epoch": 0.14731785428342675,
+      "grad_norm": 1.4406914311548737,
+      "learning_rate": 1.9287034144980114e-05,
+      "loss": 1.6758,
+      "step": 1104
+    },
+    {
+      "epoch": 0.1474512943688284,
+      "grad_norm": 1.2219976061823485,
+      "learning_rate": 1.928543061436571e-05,
+      "loss": 1.71,
+      "step": 1105
+    },
+    {
+      "epoch": 0.14758473445423004,
+      "grad_norm": 1.2519035119374464,
+      "learning_rate": 1.9283825349338588e-05,
+      "loss": 1.6538,
+      "step": 1106
+    },
+    {
+      "epoch": 0.1477181745396317,
+      "grad_norm": 1.3524232905207045,
+      "learning_rate": 1.9282218350198588e-05,
+      "loss": 1.6776,
+      "step": 1107
+    },
+    {
+      "epoch": 0.14785161462503335,
+      "grad_norm": 1.30724045787273,
+      "learning_rate": 1.928060961724589e-05,
+      "loss": 1.6482,
+      "step": 1108
+    },
+    {
+      "epoch": 0.14798505471043502,
+      "grad_norm": 1.1368808570323825,
+      "learning_rate": 1.9278999150780972e-05,
+      "loss": 1.6194,
+      "step": 1109
+    },
+    {
+      "epoch": 0.14811849479583666,
+      "grad_norm": 1.300916828823837,
+      "learning_rate": 1.927738695110466e-05,
+      "loss": 1.6947,
+      "step": 1110
+    },
+    {
+      "epoch": 0.14825193488123833,
+      "grad_norm": 1.1726520891396903,
+      "learning_rate": 1.9275773018518094e-05,
+      "loss": 1.6945,
+      "step": 1111
+    },
+    {
+      "epoch": 0.14838537496663998,
+      "grad_norm": 1.4312765994187449,
+      "learning_rate": 1.9274157353322736e-05,
+      "loss": 1.7079,
+      "step": 1112
+    },
+    {
+      "epoch": 0.14851881505204162,
+      "grad_norm": 1.3042905231751305,
+      "learning_rate": 1.9272539955820372e-05,
+      "loss": 1.6697,
+      "step": 1113
+    },
+    {
+      "epoch": 0.1486522551374433,
+      "grad_norm": 1.2109176431266062,
+      "learning_rate": 1.9270920826313122e-05,
+      "loss": 1.7066,
+      "step": 1114
+    },
+    {
+      "epoch": 0.14878569522284493,
+      "grad_norm": 1.269814239652846,
+      "learning_rate": 1.926929996510341e-05,
+      "loss": 1.6679,
+      "step": 1115
+    },
+    {
+      "epoch": 0.1489191353082466,
+      "grad_norm": 1.26285461824445,
+      "learning_rate": 1.9267677372494003e-05,
+      "loss": 1.6651,
+      "step": 1116
+    },
+    {
+      "epoch": 0.14905257539364825,
+      "grad_norm": 1.4162229382915708,
+      "learning_rate": 1.9266053048787973e-05,
+      "loss": 1.7093,
+      "step": 1117
+    },
+    {
+      "epoch": 0.14918601547904992,
+      "grad_norm": 1.1695719882927538,
+      "learning_rate": 1.926442699428874e-05,
+      "loss": 1.6775,
+      "step": 1118
+    },
+    {
+      "epoch": 0.14931945556445156,
+      "grad_norm": 1.5241355299411088,
+      "learning_rate": 1.9262799209300017e-05,
+      "loss": 1.7078,
+      "step": 1119
+    },
+    {
+      "epoch": 0.14945289564985323,
+      "grad_norm": 1.3679819271720925,
+      "learning_rate": 1.9261169694125868e-05,
+      "loss": 1.6512,
+      "step": 1120
+    },
+    {
+      "epoch": 0.14958633573525487,
+      "grad_norm": 1.0852517422618726,
+      "learning_rate": 1.925953844907066e-05,
+      "loss": 1.6732,
+      "step": 1121
+    },
+    {
+      "epoch": 0.14971977582065651,
+      "grad_norm": 1.1505718677390935,
+      "learning_rate": 1.9257905474439093e-05,
+      "loss": 1.6627,
+      "step": 1122
+    },
+    {
+      "epoch": 0.14985321590605818,
+      "grad_norm": 1.366145668087986,
+      "learning_rate": 1.9256270770536187e-05,
+      "loss": 1.6224,
+      "step": 1123
+    },
+    {
+      "epoch": 0.14998665599145983,
+      "grad_norm": 1.2009763385012844,
+      "learning_rate": 1.925463433766729e-05,
+      "loss": 1.667,
+      "step": 1124
+    },
+    {
+      "epoch": 0.1501200960768615,
+      "grad_norm": 1.1495050946339627,
+      "learning_rate": 1.9252996176138065e-05,
+      "loss": 1.7039,
+      "step": 1125
+    },
+    {
+      "epoch": 0.15025353616226314,
+      "grad_norm": 1.202004713006955,
+      "learning_rate": 1.9251356286254506e-05,
+      "loss": 1.6953,
+      "step": 1126
+    },
+    {
+      "epoch": 0.1503869762476648,
+      "grad_norm": 1.2049847095982797,
+      "learning_rate": 1.9249714668322922e-05,
+      "loss": 1.7301,
+      "step": 1127
+    },
+    {
+      "epoch": 0.15052041633306645,
+      "grad_norm": 1.3586596709801637,
+      "learning_rate": 1.924807132264995e-05,
+      "loss": 1.6488,
+      "step": 1128
+    },
+    {
+      "epoch": 0.1506538564184681,
+      "grad_norm": 1.304495283882212,
+      "learning_rate": 1.924642624954255e-05,
+      "loss": 1.6784,
+      "step": 1129
+    },
+    {
+      "epoch": 0.15078729650386977,
+      "grad_norm": 1.2254989980261268,
+      "learning_rate": 1.9244779449308e-05,
+      "loss": 1.6712,
+      "step": 1130
+    },
+    {
+      "epoch": 0.1509207365892714,
+      "grad_norm": 1.131251283705494,
+      "learning_rate": 1.9243130922253902e-05,
+      "loss": 1.6639,
+      "step": 1131
+    },
+    {
+      "epoch": 0.15105417667467308,
+      "grad_norm": 1.0999764886521537,
+      "learning_rate": 1.924148066868819e-05,
+      "loss": 1.7247,
+      "step": 1132
+    },
+    {
+      "epoch": 0.15118761676007472,
+      "grad_norm": 1.1537923655335296,
+      "learning_rate": 1.92398286889191e-05,
+      "loss": 1.6669,
+      "step": 1133
+    },
+    {
+      "epoch": 0.1513210568454764,
+      "grad_norm": 1.2811987513376073,
+      "learning_rate": 1.9238174983255216e-05,
+      "loss": 1.6873,
+      "step": 1134
+    },
+    {
+      "epoch": 0.15145449693087804,
+      "grad_norm": 1.189456464225642,
+      "learning_rate": 1.9236519552005425e-05,
+      "loss": 1.7032,
+      "step": 1135
+    },
+    {
+      "epoch": 0.15158793701627968,
+      "grad_norm": 1.226791373769935,
+      "learning_rate": 1.923486239547894e-05,
+      "loss": 1.7184,
+      "step": 1136
+    },
+    {
+      "epoch": 0.15172137710168135,
+      "grad_norm": 1.103146464311287,
+      "learning_rate": 1.9233203513985307e-05,
+      "loss": 1.6491,
+      "step": 1137
+    },
+    {
+      "epoch": 0.151854817187083,
+      "grad_norm": 1.1779372105249897,
+      "learning_rate": 1.923154290783438e-05,
+      "loss": 1.6474,
+      "step": 1138
+    },
+    {
+      "epoch": 0.15198825727248466,
+      "grad_norm": 1.2505895966166218,
+      "learning_rate": 1.922988057733634e-05,
+      "loss": 1.7153,
+      "step": 1139
+    },
+    {
+      "epoch": 0.1521216973578863,
+      "grad_norm": 1.2586045407401536,
+      "learning_rate": 1.92282165228017e-05,
+      "loss": 1.6305,
+      "step": 1140
+    },
+    {
+      "epoch": 0.15225513744328797,
+      "grad_norm": 1.1801506753191604,
+      "learning_rate": 1.9226550744541276e-05,
+      "loss": 1.6803,
+      "step": 1141
+    },
+    {
+      "epoch": 0.15238857752868962,
+      "grad_norm": 1.2155598704280677,
+      "learning_rate": 1.9224883242866223e-05,
+      "loss": 1.6921,
+      "step": 1142
+    },
+    {
+      "epoch": 0.15252201761409126,
+      "grad_norm": 1.1553853036033908,
+      "learning_rate": 1.9223214018088007e-05,
+      "loss": 1.6683,
+      "step": 1143
+    },
+    {
+      "epoch": 0.15265545769949293,
+      "grad_norm": 1.7355840451920468,
+      "learning_rate": 1.9221543070518427e-05,
+      "loss": 1.6839,
+      "step": 1144
+    },
+    {
+      "epoch": 0.15278889778489457,
+      "grad_norm": 10.109110380656796,
+      "learning_rate": 1.9219870400469588e-05,
+      "loss": 1.762,
+      "step": 1145
+    },
+    {
+      "epoch": 0.15292233787029624,
+      "grad_norm": 1.7615641665674961,
+      "learning_rate": 1.9218196008253934e-05,
+      "loss": 1.7248,
+      "step": 1146
+    },
+    {
+      "epoch": 0.15305577795569789,
+      "grad_norm": 15.239890369804119,
+      "learning_rate": 1.921651989418422e-05,
+      "loss": 1.7259,
+      "step": 1147
+    },
+    {
+      "epoch": 0.15318921804109956,
+      "grad_norm": 1.2991891405839884,
+      "learning_rate": 1.9214842058573517e-05,
+      "loss": 1.6952,
+      "step": 1148
+    },
+    {
+      "epoch": 0.1533226581265012,
+      "grad_norm": 1.2912291596013608,
+      "learning_rate": 1.921316250173524e-05,
+      "loss": 1.6993,
+      "step": 1149
+    },
+    {
+      "epoch": 0.15345609821190287,
+      "grad_norm": 1.1663630774969274,
+      "learning_rate": 1.92114812239831e-05,
+      "loss": 1.6719,
+      "step": 1150
+    },
+    {
+      "epoch": 0.1535895382973045,
+      "grad_norm": 1.2285813852625158,
+      "learning_rate": 1.9209798225631144e-05,
+      "loss": 1.7126,
+      "step": 1151
+    },
+    {
+      "epoch": 0.15372297838270615,
+      "grad_norm": 1.1467685561512206,
+      "learning_rate": 1.9208113506993734e-05,
+      "loss": 1.6242,
+      "step": 1152
+    },
+    {
+      "epoch": 0.15385641846810783,
+      "grad_norm": 1.1465340851099293,
+      "learning_rate": 1.920642706838556e-05,
+      "loss": 1.7279,
+      "step": 1153
+    },
+    {
+      "epoch": 0.15398985855350947,
+      "grad_norm": 1.1578599709330035,
+      "learning_rate": 1.9204738910121635e-05,
+      "loss": 1.6549,
+      "step": 1154
+    },
+    {
+      "epoch": 0.15412329863891114,
+      "grad_norm": 1.1683154746708295,
+      "learning_rate": 1.9203049032517276e-05,
+      "loss": 1.6429,
+      "step": 1155
+    },
+    {
+      "epoch": 0.15425673872431278,
+      "grad_norm": 1.20845397374898,
+      "learning_rate": 1.920135743588814e-05,
+      "loss": 1.681,
+      "step": 1156
+    },
+    {
+      "epoch": 0.15439017880971445,
+      "grad_norm": 1.3494049767476677,
+      "learning_rate": 1.9199664120550197e-05,
+      "loss": 1.7222,
+      "step": 1157
+    },
+    {
+      "epoch": 0.1545236188951161,
+      "grad_norm": 1.1544440730765555,
+      "learning_rate": 1.9197969086819735e-05,
+      "loss": 1.6895,
+      "step": 1158
+    },
+    {
+      "epoch": 0.15465705898051774,
+      "grad_norm": 1.1304557747461206,
+      "learning_rate": 1.9196272335013373e-05,
+      "loss": 1.6369,
+      "step": 1159
+    },
+    {
+      "epoch": 0.1547904990659194,
+      "grad_norm": 1.119564016408627,
+      "learning_rate": 1.9194573865448044e-05,
+      "loss": 1.6653,
+      "step": 1160
+    },
+    {
+      "epoch": 0.15492393915132105,
+      "grad_norm": 1.1506348346262587,
+      "learning_rate": 1.9192873678441e-05,
+      "loss": 1.6596,
+      "step": 1161
+    },
+    {
+      "epoch": 0.15505737923672272,
+      "grad_norm": 1.197324233503071,
+      "learning_rate": 1.9191171774309816e-05,
+      "loss": 1.7255,
+      "step": 1162
+    },
+    {
+      "epoch": 0.15519081932212436,
+      "grad_norm": 1.1674782781588378,
+      "learning_rate": 1.9189468153372394e-05,
+      "loss": 1.6745,
+      "step": 1163
+    },
+    {
+      "epoch": 0.15532425940752603,
+      "grad_norm": 1.308070450830591,
+      "learning_rate": 1.9187762815946942e-05,
+      "loss": 1.6307,
+      "step": 1164
+    },
+    {
+      "epoch": 0.15545769949292768,
+      "grad_norm": 1.1249631316465178,
+      "learning_rate": 1.9186055762352003e-05,
+      "loss": 1.6749,
+      "step": 1165
+    },
+    {
+      "epoch": 0.15559113957832932,
+      "grad_norm": 1.22808803451986,
+      "learning_rate": 1.918434699290644e-05,
+      "loss": 1.7167,
+      "step": 1166
+    },
+    {
+      "epoch": 0.155724579663731,
+      "grad_norm": 1.142395151395811,
+      "learning_rate": 1.918263650792942e-05,
+      "loss": 1.7143,
+      "step": 1167
+    },
+    {
+      "epoch": 0.15585801974913263,
+      "grad_norm": 1.1288423274679555,
+      "learning_rate": 1.9180924307740453e-05,
+      "loss": 1.6465,
+      "step": 1168
+    },
+    {
+      "epoch": 0.1559914598345343,
+      "grad_norm": 1.4078641204542017,
+      "learning_rate": 1.9179210392659353e-05,
+      "loss": 1.7182,
+      "step": 1169
+    },
+    {
+      "epoch": 0.15612489991993594,
+      "grad_norm": 1.1847205713407944,
+      "learning_rate": 1.917749476300626e-05,
+      "loss": 1.6681,
+      "step": 1170
+    },
+    {
+      "epoch": 0.15625834000533761,
+      "grad_norm": 1.096423039092125,
+      "learning_rate": 1.9175777419101634e-05,
+      "loss": 1.644,
+      "step": 1171
+    },
+    {
+      "epoch": 0.15639178009073926,
+      "grad_norm": 1.115302048331846,
+      "learning_rate": 1.917405836126626e-05,
+      "loss": 1.7129,
+      "step": 1172
+    },
+    {
+      "epoch": 0.1565252201761409,
+      "grad_norm": 1.1932528993802167,
+      "learning_rate": 1.917233758982123e-05,
+      "loss": 1.6633,
+      "step": 1173
+    },
+    {
+      "epoch": 0.15665866026154257,
+      "grad_norm": 1.4039420662328945,
+      "learning_rate": 1.917061510508797e-05,
+      "loss": 1.7039,
+      "step": 1174
+    },
+    {
+      "epoch": 0.1567921003469442,
+      "grad_norm": 1.3133488883749793,
+      "learning_rate": 1.9168890907388224e-05,
+      "loss": 1.639,
+      "step": 1175
+    },
+    {
+      "epoch": 0.15692554043234588,
+      "grad_norm": 1.1512042689466013,
+      "learning_rate": 1.9167164997044044e-05,
+      "loss": 1.6891,
+      "step": 1176
+    },
+    {
+      "epoch": 0.15705898051774753,
+      "grad_norm": 1.4318667585471574,
+      "learning_rate": 1.9165437374377815e-05,
+      "loss": 1.653,
+      "step": 1177
+    },
+    {
+      "epoch": 0.1571924206031492,
+      "grad_norm": 1.5436421619414216,
+      "learning_rate": 1.9163708039712238e-05,
+      "loss": 1.6979,
+      "step": 1178
+    },
+    {
+      "epoch": 0.15732586068855084,
+      "grad_norm": 1.163504109244437,
+      "learning_rate": 1.916197699337033e-05,
+      "loss": 1.6458,
+      "step": 1179
+    },
+    {
+      "epoch": 0.15745930077395248,
+      "grad_norm": 1.139437842953167,
+      "learning_rate": 1.9160244235675436e-05,
+      "loss": 1.7182,
+      "step": 1180
+    },
+    {
+      "epoch": 0.15759274085935415,
+      "grad_norm": 1.206983035678986,
+      "learning_rate": 1.915850976695121e-05,
+      "loss": 1.6761,
+      "step": 1181
+    },
+    {
+      "epoch": 0.1577261809447558,
+      "grad_norm": 1.1218445645631256,
+      "learning_rate": 1.9156773587521634e-05,
+      "loss": 1.691,
+      "step": 1182
+    },
+    {
+      "epoch": 0.15785962103015747,
+      "grad_norm": 15.435206717839069,
+      "learning_rate": 1.9155035697711008e-05,
+      "loss": 1.6704,
+      "step": 1183
+    },
+    {
+      "epoch": 0.1579930611155591,
+      "grad_norm": 1.253164673663483,
+      "learning_rate": 1.9153296097843944e-05,
+      "loss": 1.6793,
+      "step": 1184
+    },
+    {
+      "epoch": 0.15812650120096078,
+      "grad_norm": 1.2264750200407253,
+      "learning_rate": 1.9151554788245383e-05,
+      "loss": 1.6929,
+      "step": 1185
+    },
+    {
+      "epoch": 0.15825994128636242,
+      "grad_norm": 1.4772217364726286,
+      "learning_rate": 1.9149811769240585e-05,
+      "loss": 1.6452,
+      "step": 1186
+    },
+    {
+      "epoch": 0.1583933813717641,
+      "grad_norm": 1.1489987221515146,
+      "learning_rate": 1.9148067041155122e-05,
+      "loss": 1.6701,
+      "step": 1187
+    },
+    {
+      "epoch": 0.15852682145716573,
+      "grad_norm": 1.2914670100647596,
+      "learning_rate": 1.914632060431489e-05,
+      "loss": 1.7303,
+      "step": 1188
+    },
+    {
+      "epoch": 0.15866026154256738,
+      "grad_norm": 1.299350514801071,
+      "learning_rate": 1.9144572459046104e-05,
+      "loss": 1.681,
+      "step": 1189
+    },
+    {
+      "epoch": 0.15879370162796905,
+      "grad_norm": 1.3119727658820086,
+      "learning_rate": 1.9142822605675296e-05,
+      "loss": 1.696,
+      "step": 1190
+    },
+    {
+      "epoch": 0.1589271417133707,
+      "grad_norm": 1.1081403990699985,
+      "learning_rate": 1.9141071044529318e-05,
+      "loss": 1.6713,
+      "step": 1191
+    },
+    {
+      "epoch": 0.15906058179877236,
+      "grad_norm": 1.1284776489853792,
+      "learning_rate": 1.9139317775935348e-05,
+      "loss": 1.6738,
+      "step": 1192
+    },
+    {
+      "epoch": 0.159194021884174,
+      "grad_norm": 1.1261514805847666,
+      "learning_rate": 1.9137562800220872e-05,
+      "loss": 1.7054,
+      "step": 1193
+    },
+    {
+      "epoch": 0.15932746196957567,
+      "grad_norm": 1.313484879072088,
+      "learning_rate": 1.9135806117713697e-05,
+      "loss": 1.6716,
+      "step": 1194
+    },
+    {
+      "epoch": 0.15946090205497732,
+      "grad_norm": 1.278083810674143,
+      "learning_rate": 1.913404772874196e-05,
+      "loss": 1.6833,
+      "step": 1195
+    },
+    {
+      "epoch": 0.15959434214037896,
+      "grad_norm": 1.4428806272343444,
+      "learning_rate": 1.9132287633634097e-05,
+      "loss": 1.6512,
+      "step": 1196
+    },
+    {
+      "epoch": 0.15972778222578063,
+      "grad_norm": 1.209556177929987,
+      "learning_rate": 1.9130525832718884e-05,
+      "loss": 1.6942,
+      "step": 1197
+    },
+    {
+      "epoch": 0.15986122231118227,
+      "grad_norm": 1.1106358013243507,
+      "learning_rate": 1.9128762326325394e-05,
+      "loss": 1.6982,
+      "step": 1198
+    },
+    {
+      "epoch": 0.15999466239658394,
+      "grad_norm": 1.0994258199335178,
+      "learning_rate": 1.912699711478304e-05,
+      "loss": 1.6675,
+      "step": 1199
+    },
+    {
+      "epoch": 0.16012810248198558,
+      "grad_norm": 1.1185448399034177,
+      "learning_rate": 1.912523019842154e-05,
+      "loss": 1.6926,
+      "step": 1200
+    },
+    {
+      "epoch": 0.16026154256738726,
+      "grad_norm": 1.1152656846403104,
+      "learning_rate": 1.912346157757093e-05,
+      "loss": 1.7078,
+      "step": 1201
+    },
+    {
+      "epoch": 0.1603949826527889,
+      "grad_norm": 1.2924058078691643,
+      "learning_rate": 1.9121691252561578e-05,
+      "loss": 1.7084,
+      "step": 1202
+    },
+    {
+      "epoch": 0.16052842273819054,
+      "grad_norm": 1.0611065471561838,
+      "learning_rate": 1.911991922372415e-05,
+      "loss": 1.6611,
+      "step": 1203
+    },
+    {
+      "epoch": 0.1606618628235922,
+      "grad_norm": 1.3699736830557303,
+      "learning_rate": 1.911814549138965e-05,
+      "loss": 1.6897,
+      "step": 1204
+    },
+    {
+      "epoch": 0.16079530290899385,
+      "grad_norm": 1.4751343067321943,
+      "learning_rate": 1.9116370055889382e-05,
+      "loss": 1.6932,
+      "step": 1205
+    },
+    {
+      "epoch": 0.16092874299439552,
+      "grad_norm": 1.1716253181544283,
+      "learning_rate": 1.911459291755498e-05,
+      "loss": 1.7116,
+      "step": 1206
+    },
+    {
+      "epoch": 0.16106218307979717,
+      "grad_norm": 1.1114747666351001,
+      "learning_rate": 1.91128140767184e-05,
+      "loss": 1.6428,
+      "step": 1207
+    },
+    {
+      "epoch": 0.16119562316519884,
+      "grad_norm": 1.0879838952081482,
+      "learning_rate": 1.9111033533711897e-05,
+      "loss": 1.6373,
+      "step": 1208
+    },
+    {
+      "epoch": 0.16132906325060048,
+      "grad_norm": 1.1094725515932415,
+      "learning_rate": 1.9109251288868064e-05,
+      "loss": 1.6457,
+      "step": 1209
+    },
+    {
+      "epoch": 0.16146250333600212,
+      "grad_norm": 1.1303873737596453,
+      "learning_rate": 1.9107467342519808e-05,
+      "loss": 1.6839,
+      "step": 1210
+    },
+    {
+      "epoch": 0.1615959434214038,
+      "grad_norm": 1.4781044833796941,
+      "learning_rate": 1.9105681695000342e-05,
+      "loss": 1.6406,
+      "step": 1211
+    },
+    {
+      "epoch": 0.16172938350680544,
+      "grad_norm": 1.1657729793459108,
+      "learning_rate": 1.9103894346643204e-05,
+      "loss": 1.6726,
+      "step": 1212
+    },
+    {
+      "epoch": 0.1618628235922071,
+      "grad_norm": 1.2198288103403852,
+      "learning_rate": 1.910210529778226e-05,
+      "loss": 1.6929,
+      "step": 1213
+    },
+    {
+      "epoch": 0.16199626367760875,
+      "grad_norm": 1.1871995635063328,
+      "learning_rate": 1.9100314548751676e-05,
+      "loss": 1.6907,
+      "step": 1214
+    },
+    {
+      "epoch": 0.16212970376301042,
+      "grad_norm": 1.1864807109439603,
+      "learning_rate": 1.9098522099885944e-05,
+      "loss": 1.653,
+      "step": 1215
+    },
+    {
+      "epoch": 0.16226314384841206,
+      "grad_norm": 1.2368576392699981,
+      "learning_rate": 1.9096727951519872e-05,
+      "loss": 1.7201,
+      "step": 1216
+    },
+    {
+      "epoch": 0.1623965839338137,
+      "grad_norm": 1.1260231051664356,
+      "learning_rate": 1.909493210398859e-05,
+      "loss": 1.6123,
+      "step": 1217
+    },
+    {
+      "epoch": 0.16253002401921537,
+      "grad_norm": 1.4256559162950146,
+      "learning_rate": 1.9093134557627544e-05,
+      "loss": 1.666,
+      "step": 1218
+    },
+    {
+      "epoch": 0.16266346410461702,
+      "grad_norm": 1.4887124369987752,
+      "learning_rate": 1.9091335312772493e-05,
+      "loss": 1.7597,
+      "step": 1219
+    },
+    {
+      "epoch": 0.1627969041900187,
+      "grad_norm": 1.1294991712339444,
+      "learning_rate": 1.9089534369759508e-05,
+      "loss": 1.688,
+      "step": 1220
+    },
+    {
+      "epoch": 0.16293034427542033,
+      "grad_norm": 1.177420926926422,
+      "learning_rate": 1.9087731728924996e-05,
+      "loss": 1.7286,
+      "step": 1221
+    },
+    {
+      "epoch": 0.163063784360822,
+      "grad_norm": 1.6913536181421298,
+      "learning_rate": 1.9085927390605663e-05,
+      "loss": 1.6913,
+      "step": 1222
+    },
+    {
+      "epoch": 0.16319722444622364,
+      "grad_norm": 1.2163160869711371,
+      "learning_rate": 1.908412135513854e-05,
+      "loss": 1.6669,
+      "step": 1223
+    },
+    {
+      "epoch": 0.1633306645316253,
+      "grad_norm": 1.3781638643041434,
+      "learning_rate": 1.9082313622860976e-05,
+      "loss": 1.657,
+      "step": 1224
+    },
+    {
+      "epoch": 0.16346410461702696,
+      "grad_norm": 1.121301734458933,
+      "learning_rate": 1.9080504194110633e-05,
+      "loss": 1.7017,
+      "step": 1225
+    },
+    {
+      "epoch": 0.1635975447024286,
+      "grad_norm": 1.2118234782577832,
+      "learning_rate": 1.907869306922549e-05,
+      "loss": 1.7203,
+      "step": 1226
+    },
+    {
+      "epoch": 0.16373098478783027,
+      "grad_norm": 1.2074642428560403,
+      "learning_rate": 1.9076880248543847e-05,
+      "loss": 1.6558,
+      "step": 1227
+    },
+    {
+      "epoch": 0.1638644248732319,
+      "grad_norm": 1.119322476739159,
+      "learning_rate": 1.9075065732404318e-05,
+      "loss": 1.6947,
+      "step": 1228
+    },
+    {
+      "epoch": 0.16399786495863358,
+      "grad_norm": 1.1009776042368575,
+      "learning_rate": 1.9073249521145833e-05,
+      "loss": 1.6984,
+      "step": 1229
+    },
+    {
+      "epoch": 0.16413130504403523,
+      "grad_norm": 1.3636116015946005,
+      "learning_rate": 1.9071431615107642e-05,
+      "loss": 1.6266,
+      "step": 1230
+    },
+    {
+      "epoch": 0.1642647451294369,
+      "grad_norm": 1.2193076150679258,
+      "learning_rate": 1.9069612014629307e-05,
+      "loss": 1.6548,
+      "step": 1231
+    },
+    {
+      "epoch": 0.16439818521483854,
+      "grad_norm": 1.108647716162107,
+      "learning_rate": 1.9067790720050708e-05,
+      "loss": 1.6949,
+      "step": 1232
+    },
+    {
+      "epoch": 0.16453162530024018,
+      "grad_norm": 1.3428440762492813,
+      "learning_rate": 1.9065967731712044e-05,
+      "loss": 1.628,
+      "step": 1233
+    },
+    {
+      "epoch": 0.16466506538564185,
+      "grad_norm": 1.1264569960102961,
+      "learning_rate": 1.9064143049953826e-05,
+      "loss": 1.69,
+      "step": 1234
+    },
+    {
+      "epoch": 0.1647985054710435,
+      "grad_norm": 1.2129394646481007,
+      "learning_rate": 1.9062316675116886e-05,
+      "loss": 1.7039,
+      "step": 1235
+    },
+    {
+      "epoch": 0.16493194555644516,
+      "grad_norm": 1.3168167400509623,
+      "learning_rate": 1.9060488607542373e-05,
+      "loss": 1.6854,
+      "step": 1236
+    },
+    {
+      "epoch": 0.1650653856418468,
+      "grad_norm": 1.172892361541946,
+      "learning_rate": 1.9058658847571744e-05,
+      "loss": 1.6908,
+      "step": 1237
+    },
+    {
+      "epoch": 0.16519882572724848,
+      "grad_norm": 1.8095333564387064,
+      "learning_rate": 1.905682739554678e-05,
+      "loss": 1.6951,
+      "step": 1238
+    },
+    {
+      "epoch": 0.16533226581265012,
+      "grad_norm": 1.3485139476585741,
+      "learning_rate": 1.905499425180957e-05,
+      "loss": 1.6748,
+      "step": 1239
+    },
+    {
+      "epoch": 0.16546570589805176,
+      "grad_norm": 1.2787725949509623,
+      "learning_rate": 1.905315941670253e-05,
+      "loss": 1.6657,
+      "step": 1240
+    },
+    {
+      "epoch": 0.16559914598345343,
+      "grad_norm": 1.2224431058019596,
+      "learning_rate": 1.9051322890568386e-05,
+      "loss": 1.6656,
+      "step": 1241
+    },
+    {
+      "epoch": 0.16573258606885508,
+      "grad_norm": 1.137537755181391,
+      "learning_rate": 1.904948467375018e-05,
+      "loss": 1.6577,
+      "step": 1242
+    },
+    {
+      "epoch": 0.16586602615425675,
+      "grad_norm": 1.0868355979985056,
+      "learning_rate": 1.9047644766591273e-05,
+      "loss": 1.7073,
+      "step": 1243
+    },
+    {
+      "epoch": 0.1659994662396584,
+      "grad_norm": 1.114462768904141,
+      "learning_rate": 1.9045803169435326e-05,
+      "loss": 1.6305,
+      "step": 1244
+    },
+    {
+      "epoch": 0.16613290632506006,
+      "grad_norm": 1.3561430912216557,
+      "learning_rate": 1.9043959882626343e-05,
+      "loss": 1.6719,
+      "step": 1245
+    },
+    {
+      "epoch": 0.1662663464104617,
+      "grad_norm": 1.3197220598884711,
+      "learning_rate": 1.9042114906508623e-05,
+      "loss": 1.6904,
+      "step": 1246
+    },
+    {
+      "epoch": 0.16639978649586334,
+      "grad_norm": 1.1663473878740998,
+      "learning_rate": 1.9040268241426786e-05,
+      "loss": 1.6815,
+      "step": 1247
+    },
+    {
+      "epoch": 0.16653322658126501,
+      "grad_norm": 1.1208171885233782,
+      "learning_rate": 1.9038419887725768e-05,
+      "loss": 1.6888,
+      "step": 1248
+    },
+    {
+      "epoch": 0.16666666666666666,
+      "grad_norm": 1.2317096593237165,
+      "learning_rate": 1.9036569845750822e-05,
+      "loss": 1.6998,
+      "step": 1249
+    },
+    {
+      "epoch": 0.16680010675206833,
+      "grad_norm": 1.2118086361733955,
+      "learning_rate": 1.9034718115847516e-05,
+      "loss": 1.7239,
+      "step": 1250
+    },
+    {
+      "epoch": 0.16693354683746997,
+      "grad_norm": 1.1328099587839497,
+      "learning_rate": 1.9032864698361728e-05,
+      "loss": 1.6664,
+      "step": 1251
+    },
+    {
+      "epoch": 0.16706698692287164,
+      "grad_norm": 1.2351889795437343,
+      "learning_rate": 1.9031009593639656e-05,
+      "loss": 1.7225,
+      "step": 1252
+    },
+    {
+      "epoch": 0.16720042700827328,
+      "grad_norm": 1.1680475805874204,
+      "learning_rate": 1.9029152802027816e-05,
+      "loss": 1.7055,
+      "step": 1253
+    },
+    {
+      "epoch": 0.16733386709367493,
+      "grad_norm": 1.1396480995870708,
+      "learning_rate": 1.9027294323873032e-05,
+      "loss": 1.6754,
+      "step": 1254
+    },
+    {
+      "epoch": 0.1674673071790766,
+      "grad_norm": 1.099951848229152,
+      "learning_rate": 1.902543415952245e-05,
+      "loss": 1.6574,
+      "step": 1255
+    },
+    {
+      "epoch": 0.16760074726447824,
+      "grad_norm": 1.091862726736696,
+      "learning_rate": 1.9023572309323522e-05,
+      "loss": 1.6254,
+      "step": 1256
+    },
+    {
+      "epoch": 0.1677341873498799,
+      "grad_norm": 1.2660462090997906,
+      "learning_rate": 1.9021708773624027e-05,
+      "loss": 1.6797,
+      "step": 1257
+    },
+    {
+      "epoch": 0.16786762743528155,
+      "grad_norm": 1.1200461208933035,
+      "learning_rate": 1.901984355277205e-05,
+      "loss": 1.6885,
+      "step": 1258
+    },
+    {
+      "epoch": 0.16800106752068322,
+      "grad_norm": 1.2050844123558844,
+      "learning_rate": 1.9017976647115988e-05,
+      "loss": 1.6851,
+      "step": 1259
+    },
+    {
+      "epoch": 0.16813450760608487,
+      "grad_norm": 1.142685776004093,
+      "learning_rate": 1.9016108057004566e-05,
+      "loss": 1.625,
+      "step": 1260
+    },
+    {
+      "epoch": 0.16826794769148654,
+      "grad_norm": 1.1684071493919217,
+      "learning_rate": 1.9014237782786806e-05,
+      "loss": 1.6768,
+      "step": 1261
+    },
+    {
+      "epoch": 0.16840138777688818,
+      "grad_norm": 1.1496482446886893,
+      "learning_rate": 1.9012365824812063e-05,
+      "loss": 1.6409,
+      "step": 1262
+    },
+    {
+      "epoch": 0.16853482786228982,
+      "grad_norm": 1.0944121754604215,
+      "learning_rate": 1.9010492183429994e-05,
+      "loss": 1.63,
+      "step": 1263
+    },
+    {
+      "epoch": 0.1686682679476915,
+      "grad_norm": 1.0842462969633766,
+      "learning_rate": 1.9008616858990572e-05,
+      "loss": 1.646,
+      "step": 1264
+    },
+    {
+      "epoch": 0.16880170803309313,
+      "grad_norm": 1.4767767757650743,
+      "learning_rate": 1.9006739851844086e-05,
+      "loss": 1.6454,
+      "step": 1265
+    },
+    {
+      "epoch": 0.1689351481184948,
+      "grad_norm": 1.18032002297524,
+      "learning_rate": 1.9004861162341144e-05,
+      "loss": 1.6861,
+      "step": 1266
+    },
+    {
+      "epoch": 0.16906858820389645,
+      "grad_norm": 1.3372419431896025,
+      "learning_rate": 1.9002980790832663e-05,
+      "loss": 1.6586,
+      "step": 1267
+    },
+    {
+      "epoch": 0.16920202828929812,
+      "grad_norm": 1.172100506483245,
+      "learning_rate": 1.900109873766987e-05,
+      "loss": 1.6657,
+      "step": 1268
+    },
+    {
+      "epoch": 0.16933546837469976,
+      "grad_norm": 1.1145469314883394,
+      "learning_rate": 1.8999215003204316e-05,
+      "loss": 1.7082,
+      "step": 1269
+    },
+    {
+      "epoch": 0.1694689084601014,
+      "grad_norm": 1.1206009643179509,
+      "learning_rate": 1.8997329587787856e-05,
+      "loss": 1.6578,
+      "step": 1270
+    },
+    {
+      "epoch": 0.16960234854550307,
+      "grad_norm": 1.399049311820928,
+      "learning_rate": 1.8995442491772668e-05,
+      "loss": 1.6955,
+      "step": 1271
+    },
+    {
+      "epoch": 0.16973578863090472,
+      "grad_norm": 1.1958580398149483,
+      "learning_rate": 1.899355371551124e-05,
+      "loss": 1.7101,
+      "step": 1272
+    },
+    {
+      "epoch": 0.1698692287163064,
+      "grad_norm": 1.3556638628411706,
+      "learning_rate": 1.8991663259356374e-05,
+      "loss": 1.7108,
+      "step": 1273
+    },
+    {
+      "epoch": 0.17000266880170803,
+      "grad_norm": 1.3451150415787545,
+      "learning_rate": 1.8989771123661186e-05,
+      "loss": 1.7315,
+      "step": 1274
+    },
+    {
+      "epoch": 0.1701361088871097,
+      "grad_norm": 1.3135350609625422,
+      "learning_rate": 1.8987877308779104e-05,
+      "loss": 1.6609,
+      "step": 1275
+    },
+    {
+      "epoch": 0.17026954897251134,
+      "grad_norm": 1.1333530911950962,
+      "learning_rate": 1.8985981815063873e-05,
+      "loss": 1.6684,
+      "step": 1276
+    },
+    {
+      "epoch": 0.17040298905791298,
+      "grad_norm": 1.2791530568330647,
+      "learning_rate": 1.8984084642869546e-05,
+      "loss": 1.6651,
+      "step": 1277
+    },
+    {
+      "epoch": 0.17053642914331466,
+      "grad_norm": 1.3692883154733826,
+      "learning_rate": 1.8982185792550495e-05,
+      "loss": 1.7462,
+      "step": 1278
+    },
+    {
+      "epoch": 0.1706698692287163,
+      "grad_norm": 1.335470306118413,
+      "learning_rate": 1.8980285264461404e-05,
+      "loss": 1.6328,
+      "step": 1279
+    },
+    {
+      "epoch": 0.17080330931411797,
+      "grad_norm": 1.609394077858067,
+      "learning_rate": 1.8978383058957272e-05,
+      "loss": 1.7349,
+      "step": 1280
+    },
+    {
+      "epoch": 0.1709367493995196,
+      "grad_norm": 1.0926180749124967,
+      "learning_rate": 1.8976479176393405e-05,
+      "loss": 1.6251,
+      "step": 1281
+    },
+    {
+      "epoch": 0.17107018948492128,
+      "grad_norm": 1.5796826003726887,
+      "learning_rate": 1.8974573617125433e-05,
+      "loss": 1.7046,
+      "step": 1282
+    },
+    {
+      "epoch": 0.17120362957032292,
+      "grad_norm": 1.2777834800327206,
+      "learning_rate": 1.8972666381509283e-05,
+      "loss": 1.6877,
+      "step": 1283
+    },
+    {
+      "epoch": 0.17133706965572457,
+      "grad_norm": 1.1654445244110958,
+      "learning_rate": 1.8970757469901215e-05,
+      "loss": 1.6931,
+      "step": 1284
+    },
+    {
+      "epoch": 0.17147050974112624,
+      "grad_norm": 1.2695593578179767,
+      "learning_rate": 1.8968846882657784e-05,
+      "loss": 1.6171,
+      "step": 1285
+    },
+    {
+      "epoch": 0.17160394982652788,
+      "grad_norm": 1.1260255267285963,
+      "learning_rate": 1.896693462013587e-05,
+      "loss": 1.6733,
+      "step": 1286
+    },
+    {
+      "epoch": 0.17173738991192955,
+      "grad_norm": 1.1761366558123016,
+      "learning_rate": 1.8965020682692662e-05,
+      "loss": 1.6513,
+      "step": 1287
+    },
+    {
+      "epoch": 0.1718708299973312,
+      "grad_norm": 1.3298600101322868,
+      "learning_rate": 1.8963105070685667e-05,
+      "loss": 1.6684,
+      "step": 1288
+    },
+    {
+      "epoch": 0.17200427008273286,
+      "grad_norm": 1.1292417955776037,
+      "learning_rate": 1.8961187784472685e-05,
+      "loss": 1.6489,
+      "step": 1289
+    },
+    {
+      "epoch": 0.1721377101681345,
+      "grad_norm": 1.2976031015347522,
+      "learning_rate": 1.8959268824411857e-05,
+      "loss": 1.7234,
+      "step": 1290
+    },
+    {
+      "epoch": 0.17227115025353618,
+      "grad_norm": 5.468850373981861,
+      "learning_rate": 1.8957348190861612e-05,
+      "loss": 1.6717,
+      "step": 1291
+    },
+    {
+      "epoch": 0.17240459033893782,
+      "grad_norm": 1.2061510206612853,
+      "learning_rate": 1.8955425884180715e-05,
+      "loss": 1.6682,
+      "step": 1292
+    },
+    {
+      "epoch": 0.17253803042433946,
+      "grad_norm": 1.2340106256554149,
+      "learning_rate": 1.895350190472822e-05,
+      "loss": 1.6452,
+      "step": 1293
+    },
+    {
+      "epoch": 0.17267147050974113,
+      "grad_norm": 1.1295754093637331,
+      "learning_rate": 1.895157625286351e-05,
+      "loss": 1.6993,
+      "step": 1294
+    },
+    {
+      "epoch": 0.17280491059514277,
+      "grad_norm": 1.1791235092887307,
+      "learning_rate": 1.8949648928946275e-05,
+      "loss": 1.6347,
+      "step": 1295
+    },
+    {
+      "epoch": 0.17293835068054444,
+      "grad_norm": 1.0917941158292654,
+      "learning_rate": 1.894771993333651e-05,
+      "loss": 1.6361,
+      "step": 1296
+    },
+    {
+      "epoch": 0.1730717907659461,
+      "grad_norm": 1.243225904269034,
+      "learning_rate": 1.8945789266394543e-05,
+      "loss": 1.6859,
+      "step": 1297
+    },
+    {
+      "epoch": 0.17320523085134776,
+      "grad_norm": 1.4426394116019783,
+      "learning_rate": 1.894385692848099e-05,
+      "loss": 1.6686,
+      "step": 1298
+    },
+    {
+      "epoch": 0.1733386709367494,
+      "grad_norm": 1.1402525280881741,
+      "learning_rate": 1.8941922919956788e-05,
+      "loss": 1.6656,
+      "step": 1299
+    },
+    {
+      "epoch": 0.17347211102215104,
+      "grad_norm": 1.1712942843243668,
+      "learning_rate": 1.8939987241183195e-05,
+      "loss": 1.6622,
+      "step": 1300
+    },
+    {
+      "epoch": 0.1736055511075527,
+      "grad_norm": 1.1420488961794493,
+      "learning_rate": 1.8938049892521772e-05,
+      "loss": 1.6048,
+      "step": 1301
+    },
+    {
+      "epoch": 0.17373899119295436,
+      "grad_norm": 1.4695327916997878,
+      "learning_rate": 1.8936110874334392e-05,
+      "loss": 1.6625,
+      "step": 1302
+    },
+    {
+      "epoch": 0.17387243127835603,
+      "grad_norm": 1.0900641908664435,
+      "learning_rate": 1.8934170186983242e-05,
+      "loss": 1.6396,
+      "step": 1303
+    },
+    {
+      "epoch": 0.17400587136375767,
+      "grad_norm": 1.1307482898544658,
+      "learning_rate": 1.8932227830830822e-05,
+      "loss": 1.6878,
+      "step": 1304
+    },
+    {
+      "epoch": 0.17413931144915934,
+      "grad_norm": 1.0752115709650287,
+      "learning_rate": 1.8930283806239936e-05,
+      "loss": 1.6868,
+      "step": 1305
+    },
+    {
+      "epoch": 0.17427275153456098,
+      "grad_norm": 1.3371575151194106,
+      "learning_rate": 1.8928338113573715e-05,
+      "loss": 1.6374,
+      "step": 1306
+    },
+    {
+      "epoch": 0.17440619161996262,
+      "grad_norm": 1.2309293597249826,
+      "learning_rate": 1.8926390753195583e-05,
+      "loss": 1.6629,
+      "step": 1307
+    },
+    {
+      "epoch": 0.1745396317053643,
+      "grad_norm": 1.1999413493774056,
+      "learning_rate": 1.8924441725469293e-05,
+      "loss": 1.6837,
+      "step": 1308
+    },
+    {
+      "epoch": 0.17467307179076594,
+      "grad_norm": 4.1911788978313735,
+      "learning_rate": 1.8922491030758893e-05,
+      "loss": 1.7471,
+      "step": 1309
+    },
+    {
+      "epoch": 0.1748065118761676,
+      "grad_norm": 1.1779561572349584,
+      "learning_rate": 1.892053866942876e-05,
+      "loss": 1.7246,
+      "step": 1310
+    },
+    {
+      "epoch": 0.17493995196156925,
+      "grad_norm": 1.1617629781133412,
+      "learning_rate": 1.8918584641843564e-05,
+      "loss": 1.6462,
+      "step": 1311
+    },
+    {
+      "epoch": 0.17507339204697092,
+      "grad_norm": 1.2004531219911387,
+      "learning_rate": 1.89166289483683e-05,
+      "loss": 1.6432,
+      "step": 1312
+    },
+    {
+      "epoch": 0.17520683213237256,
+      "grad_norm": 1.1057648722865505,
+      "learning_rate": 1.891467158936827e-05,
+      "loss": 1.6363,
+      "step": 1313
+    },
+    {
+      "epoch": 0.1753402722177742,
+      "grad_norm": 1.0830292426032284,
+      "learning_rate": 1.8912712565209082e-05,
+      "loss": 1.6543,
+      "step": 1314
+    },
+    {
+      "epoch": 0.17547371230317588,
+      "grad_norm": 1.390700802694287,
+      "learning_rate": 1.8910751876256663e-05,
+      "loss": 1.5831,
+      "step": 1315
+    },
+    {
+      "epoch": 0.17560715238857752,
+      "grad_norm": 1.1442545355007538,
+      "learning_rate": 1.890878952287724e-05,
+      "loss": 1.6725,
+      "step": 1316
+    },
+    {
+      "epoch": 0.1757405924739792,
+      "grad_norm": 4.524949596846156,
+      "learning_rate": 1.8906825505437375e-05,
+      "loss": 1.7064,
+      "step": 1317
+    },
+    {
+      "epoch": 0.17587403255938083,
+      "grad_norm": 1.5116207526291794,
+      "learning_rate": 1.8904859824303906e-05,
+      "loss": 1.6625,
+      "step": 1318
+    },
+    {
+      "epoch": 0.1760074726447825,
+      "grad_norm": 1.3022482630711962,
+      "learning_rate": 1.8902892479844012e-05,
+      "loss": 1.6899,
+      "step": 1319
+    },
+    {
+      "epoch": 0.17614091273018415,
+      "grad_norm": 1.3036609447195004,
+      "learning_rate": 1.8900923472425162e-05,
+      "loss": 1.7094,
+      "step": 1320
+    },
+    {
+      "epoch": 0.1762743528155858,
+      "grad_norm": 1.201826945742321,
+      "learning_rate": 1.889895280241515e-05,
+      "loss": 1.6874,
+      "step": 1321
+    },
+    {
+      "epoch": 0.17640779290098746,
+      "grad_norm": 1.331895493241366,
+      "learning_rate": 1.8896980470182074e-05,
+      "loss": 1.6567,
+      "step": 1322
+    },
+    {
+      "epoch": 0.1765412329863891,
+      "grad_norm": 1.2443633859323437,
+      "learning_rate": 1.889500647609434e-05,
+      "loss": 1.667,
+      "step": 1323
+    },
+    {
+      "epoch": 0.17667467307179077,
+      "grad_norm": 1.2258655395860159,
+      "learning_rate": 1.8893030820520676e-05,
+      "loss": 1.6747,
+      "step": 1324
+    },
+    {
+      "epoch": 0.17680811315719241,
+      "grad_norm": 1.1121857279555136,
+      "learning_rate": 1.8891053503830096e-05,
+      "loss": 1.6623,
+      "step": 1325
+    },
+    {
+      "epoch": 0.17694155324259409,
+      "grad_norm": 1.1496733774130343,
+      "learning_rate": 1.8889074526391956e-05,
+      "loss": 1.6505,
+      "step": 1326
+    },
+    {
+      "epoch": 0.17707499332799573,
+      "grad_norm": 1.190622161497496,
+      "learning_rate": 1.8887093888575897e-05,
+      "loss": 1.6946,
+      "step": 1327
+    },
+    {
+      "epoch": 0.1772084334133974,
+      "grad_norm": 1.234769822108329,
+      "learning_rate": 1.8885111590751887e-05,
+      "loss": 1.6822,
+      "step": 1328
+    },
+    {
+      "epoch": 0.17734187349879904,
+      "grad_norm": 1.1961580846084812,
+      "learning_rate": 1.8883127633290193e-05,
+      "loss": 1.6098,
+      "step": 1329
+    },
+    {
+      "epoch": 0.17747531358420068,
+      "grad_norm": 1.1500701811676424,
+      "learning_rate": 1.888114201656139e-05,
+      "loss": 1.6979,
+      "step": 1330
+    },
+    {
+      "epoch": 0.17760875366960235,
+      "grad_norm": 1.068835069330444,
+      "learning_rate": 1.887915474093638e-05,
+      "loss": 1.6443,
+      "step": 1331
+    },
+    {
+      "epoch": 0.177742193755004,
+      "grad_norm": 1.0715312154596754,
+      "learning_rate": 1.8877165806786353e-05,
+      "loss": 1.6178,
+      "step": 1332
+    },
+    {
+      "epoch": 0.17787563384040567,
+      "grad_norm": 1.3293068311768508,
+      "learning_rate": 1.8875175214482823e-05,
+      "loss": 1.6494,
+      "step": 1333
+    },
+    {
+      "epoch": 0.1780090739258073,
+      "grad_norm": 1.2954287617747784,
+      "learning_rate": 1.8873182964397614e-05,
+      "loss": 1.6546,
+      "step": 1334
+    },
+    {
+      "epoch": 0.17814251401120898,
+      "grad_norm": 1.125061002872152,
+      "learning_rate": 1.887118905690285e-05,
+      "loss": 1.6897,
+      "step": 1335
+    },
+    {
+      "epoch": 0.17827595409661062,
+      "grad_norm": 1.0654922527296347,
+      "learning_rate": 1.8869193492370975e-05,
+      "loss": 1.6899,
+      "step": 1336
+    },
+    {
+      "epoch": 0.17840939418201227,
+      "grad_norm": 1.2859684785579713,
+      "learning_rate": 1.8867196271174732e-05,
+      "loss": 1.6489,
+      "step": 1337
+    },
+    {
+      "epoch": 0.17854283426741394,
+      "grad_norm": 1.2913796594054934,
+      "learning_rate": 1.8865197393687184e-05,
+      "loss": 1.6531,
+      "step": 1338
+    },
+    {
+      "epoch": 0.17867627435281558,
+      "grad_norm": 1.1287917089144135,
+      "learning_rate": 1.8863196860281696e-05,
+      "loss": 1.7184,
+      "step": 1339
+    },
+    {
+      "epoch": 0.17880971443821725,
+      "grad_norm": 1.066588369186229,
+      "learning_rate": 1.8861194671331946e-05,
+      "loss": 1.7052,
+      "step": 1340
+    },
+    {
+      "epoch": 0.1789431545236189,
+      "grad_norm": 1.3155491787656075,
+      "learning_rate": 1.8859190827211922e-05,
+      "loss": 1.739,
+      "step": 1341
+    },
+    {
+      "epoch": 0.17907659460902056,
+      "grad_norm": 1.351336369334145,
+      "learning_rate": 1.8857185328295915e-05,
+      "loss": 1.6433,
+      "step": 1342
+    },
+    {
+      "epoch": 0.1792100346944222,
+      "grad_norm": 1.144045347845426,
+      "learning_rate": 1.885517817495853e-05,
+      "loss": 1.6588,
+      "step": 1343
+    },
+    {
+      "epoch": 0.17934347477982385,
+      "grad_norm": 1.0634778809440733,
+      "learning_rate": 1.8853169367574684e-05,
+      "loss": 1.6542,
+      "step": 1344
+    },
+    {
+      "epoch": 0.17947691486522552,
+      "grad_norm": 1.177888372263127,
+      "learning_rate": 1.88511589065196e-05,
+      "loss": 1.6408,
+      "step": 1345
+    },
+    {
+      "epoch": 0.17961035495062716,
+      "grad_norm": 1.2708302318930562,
+      "learning_rate": 1.88491467921688e-05,
+      "loss": 1.6817,
+      "step": 1346
+    },
+    {
+      "epoch": 0.17974379503602883,
+      "grad_norm": 1.170081033949634,
+      "learning_rate": 1.8847133024898135e-05,
+      "loss": 1.6806,
+      "step": 1347
+    },
+    {
+      "epoch": 0.17987723512143047,
+      "grad_norm": 1.0804261336260308,
+      "learning_rate": 1.884511760508375e-05,
+      "loss": 1.6134,
+      "step": 1348
+    },
+    {
+      "epoch": 0.18001067520683214,
+      "grad_norm": 1.087811387530533,
+      "learning_rate": 1.88431005331021e-05,
+      "loss": 1.6393,
+      "step": 1349
+    },
+    {
+      "epoch": 0.1801441152922338,
+      "grad_norm": 1.3853627907695811,
+      "learning_rate": 1.8841081809329952e-05,
+      "loss": 1.7002,
+      "step": 1350
+    },
+    {
+      "epoch": 0.18027755537763543,
+      "grad_norm": 1.1966968339164359,
+      "learning_rate": 1.8839061434144383e-05,
+      "loss": 1.6703,
+      "step": 1351
+    },
+    {
+      "epoch": 0.1804109954630371,
+      "grad_norm": 1.1737840062211229,
+      "learning_rate": 1.8837039407922776e-05,
+      "loss": 1.7185,
+      "step": 1352
+    },
+    {
+      "epoch": 0.18054443554843874,
+      "grad_norm": 1.0896871608203742,
+      "learning_rate": 1.883501573104282e-05,
+      "loss": 1.6583,
+      "step": 1353
+    },
+    {
+      "epoch": 0.1806778756338404,
+      "grad_norm": 1.3179615694791937,
+      "learning_rate": 1.883299040388251e-05,
+      "loss": 1.6311,
+      "step": 1354
+    },
+    {
+      "epoch": 0.18081131571924206,
+      "grad_norm": 1.1110191440501396,
+      "learning_rate": 1.883096342682017e-05,
+      "loss": 1.6461,
+      "step": 1355
+    },
+    {
+      "epoch": 0.18094475580464373,
+      "grad_norm": 1.327469108458132,
+      "learning_rate": 1.88289348002344e-05,
+      "loss": 1.624,
+      "step": 1356
+    },
+    {
+      "epoch": 0.18107819589004537,
+      "grad_norm": 1.1046636131234442,
+      "learning_rate": 1.882690452450413e-05,
+      "loss": 1.6798,
+      "step": 1357
+    },
+    {
+      "epoch": 0.181211635975447,
+      "grad_norm": 1.0969053723393494,
+      "learning_rate": 1.8824872600008595e-05,
+      "loss": 1.6196,
+      "step": 1358
+    },
+    {
+      "epoch": 0.18134507606084868,
+      "grad_norm": 4.9414750848280455,
+      "learning_rate": 1.8822839027127333e-05,
+      "loss": 1.6521,
+      "step": 1359
+    },
+    {
+      "epoch": 0.18147851614625032,
+      "grad_norm": 1.1053740384798774,
+      "learning_rate": 1.8820803806240192e-05,
+      "loss": 1.6409,
+      "step": 1360
+    },
+    {
+      "epoch": 0.181611956231652,
+      "grad_norm": 1.1800356925014113,
+      "learning_rate": 1.881876693772733e-05,
+      "loss": 1.6952,
+      "step": 1361
+    },
+    {
+      "epoch": 0.18174539631705364,
+      "grad_norm": 1.339932629654733,
+      "learning_rate": 1.8816728421969208e-05,
+      "loss": 1.6583,
+      "step": 1362
+    },
+    {
+      "epoch": 0.1818788364024553,
+      "grad_norm": 1.1809439235732044,
+      "learning_rate": 1.88146882593466e-05,
+      "loss": 1.657,
+      "step": 1363
+    },
+    {
+      "epoch": 0.18201227648785695,
+      "grad_norm": 1.1124814368658855,
+      "learning_rate": 1.8812646450240586e-05,
+      "loss": 1.6248,
+      "step": 1364
+    },
+    {
+      "epoch": 0.18214571657325862,
+      "grad_norm": 1.3035826545815201,
+      "learning_rate": 1.881060299503255e-05,
+      "loss": 1.6932,
+      "step": 1365
+    },
+    {
+      "epoch": 0.18227915665866026,
+      "grad_norm": 1.2040318351320083,
+      "learning_rate": 1.8808557894104185e-05,
+      "loss": 1.6214,
+      "step": 1366
+    },
+    {
+      "epoch": 0.1824125967440619,
+      "grad_norm": 1.2476963311275309,
+      "learning_rate": 1.8806511147837492e-05,
+      "loss": 1.6661,
+      "step": 1367
+    },
+    {
+      "epoch": 0.18254603682946358,
+      "grad_norm": 1.1447416720262047,
+      "learning_rate": 1.8804462756614788e-05,
+      "loss": 1.6795,
+      "step": 1368
+    },
+    {
+      "epoch": 0.18267947691486522,
+      "grad_norm": 1.2146315931776808,
+      "learning_rate": 1.8802412720818683e-05,
+      "loss": 1.6683,
+      "step": 1369
+    },
+    {
+      "epoch": 0.1828129170002669,
+      "grad_norm": 1.0961663105222377,
+      "learning_rate": 1.8800361040832098e-05,
+      "loss": 1.6522,
+      "step": 1370
+    },
+    {
+      "epoch": 0.18294635708566853,
+      "grad_norm": 1.4720443132763514,
+      "learning_rate": 1.8798307717038267e-05,
+      "loss": 1.6972,
+      "step": 1371
+    },
+    {
+      "epoch": 0.1830797971710702,
+      "grad_norm": 1.077369472777587,
+      "learning_rate": 1.8796252749820728e-05,
+      "loss": 1.6867,
+      "step": 1372
+    },
+    {
+      "epoch": 0.18321323725647184,
+      "grad_norm": 1.082352978247377,
+      "learning_rate": 1.8794196139563324e-05,
+      "loss": 1.6186,
+      "step": 1373
+    },
+    {
+      "epoch": 0.1833466773418735,
+      "grad_norm": 1.3376025342187408,
+      "learning_rate": 1.8792137886650207e-05,
+      "loss": 1.6242,
+      "step": 1374
+    },
+    {
+      "epoch": 0.18348011742727516,
+      "grad_norm": 1.060000822920305,
+      "learning_rate": 1.8790077991465833e-05,
+      "loss": 1.6694,
+      "step": 1375
+    },
+    {
+      "epoch": 0.1836135575126768,
+      "grad_norm": 1.1946549594821052,
+      "learning_rate": 1.8788016454394972e-05,
+      "loss": 1.7262,
+      "step": 1376
+    },
+    {
+      "epoch": 0.18374699759807847,
+      "grad_norm": 1.1956085836847756,
+      "learning_rate": 1.8785953275822694e-05,
+      "loss": 1.6692,
+      "step": 1377
+    },
+    {
+      "epoch": 0.1838804376834801,
+      "grad_norm": 1.398886222221775,
+      "learning_rate": 1.8783888456134373e-05,
+      "loss": 1.6827,
+      "step": 1378
+    },
+    {
+      "epoch": 0.18401387776888178,
+      "grad_norm": 1.0879716586164903,
+      "learning_rate": 1.8781821995715694e-05,
+      "loss": 1.6799,
+      "step": 1379
+    },
+    {
+      "epoch": 0.18414731785428343,
+      "grad_norm": 1.0905252446482103,
+      "learning_rate": 1.8779753894952658e-05,
+      "loss": 1.6059,
+      "step": 1380
+    },
+    {
+      "epoch": 0.18428075793968507,
+      "grad_norm": 11.800697000575934,
+      "learning_rate": 1.877768415423155e-05,
+      "loss": 1.7131,
+      "step": 1381
+    },
+    {
+      "epoch": 0.18441419802508674,
+      "grad_norm": 1.4694858089280596,
+      "learning_rate": 1.8775612773938984e-05,
+      "loss": 1.6846,
+      "step": 1382
+    },
+    {
+      "epoch": 0.18454763811048838,
+      "grad_norm": 1.4328576424428283,
+      "learning_rate": 1.877353975446186e-05,
+      "loss": 1.6188,
+      "step": 1383
+    },
+    {
+      "epoch": 0.18468107819589005,
+      "grad_norm": 1.3413019372811694,
+      "learning_rate": 1.8771465096187404e-05,
+      "loss": 1.6835,
+      "step": 1384
+    },
+    {
+      "epoch": 0.1848145182812917,
+      "grad_norm": 1.1291865647681716,
+      "learning_rate": 1.8769388799503138e-05,
+      "loss": 1.6709,
+      "step": 1385
+    },
+    {
+      "epoch": 0.18494795836669337,
+      "grad_norm": 1.2530526469172218,
+      "learning_rate": 1.876731086479688e-05,
+      "loss": 1.6951,
+      "step": 1386
+    },
+    {
+      "epoch": 0.185081398452095,
+      "grad_norm": 1.1547439851135595,
+      "learning_rate": 1.876523129245678e-05,
+      "loss": 1.6539,
+      "step": 1387
+    },
+    {
+      "epoch": 0.18521483853749665,
+      "grad_norm": 1.083610486699219,
+      "learning_rate": 1.8763150082871264e-05,
+      "loss": 1.6559,
+      "step": 1388
+    },
+    {
+      "epoch": 0.18534827862289832,
+      "grad_norm": 1.3139745218385874,
+      "learning_rate": 1.876106723642909e-05,
+      "loss": 1.6658,
+      "step": 1389
+    },
+    {
+      "epoch": 0.18548171870829996,
+      "grad_norm": 1.2845524278681613,
+      "learning_rate": 1.8758982753519302e-05,
+      "loss": 1.6613,
+      "step": 1390
+    },
+    {
+      "epoch": 0.18561515879370163,
+      "grad_norm": 1.3886592309613566,
+      "learning_rate": 1.875689663453126e-05,
+      "loss": 1.6973,
+      "step": 1391
+    },
+    {
+      "epoch": 0.18574859887910328,
+      "grad_norm": 1.2127883048481998,
+      "learning_rate": 1.8754808879854628e-05,
+      "loss": 1.6643,
+      "step": 1392
+    },
+    {
+      "epoch": 0.18588203896450495,
+      "grad_norm": 1.1707525451232437,
+      "learning_rate": 1.8752719489879373e-05,
+      "loss": 1.6844,
+      "step": 1393
+    },
+    {
+      "epoch": 0.1860154790499066,
+      "grad_norm": 1.2601795453565978,
+      "learning_rate": 1.8750628464995775e-05,
+      "loss": 1.7148,
+      "step": 1394
+    },
+    {
+      "epoch": 0.18614891913530823,
+      "grad_norm": 1.1313620285698516,
+      "learning_rate": 1.8748535805594406e-05,
+      "loss": 1.6856,
+      "step": 1395
+    },
+    {
+      "epoch": 0.1862823592207099,
+      "grad_norm": 1.2540203774883083,
+      "learning_rate": 1.8746441512066155e-05,
+      "loss": 1.6585,
+      "step": 1396
+    },
+    {
+      "epoch": 0.18641579930611155,
+      "grad_norm": 1.218035613217223,
+      "learning_rate": 1.8744345584802212e-05,
+      "loss": 1.6955,
+      "step": 1397
+    },
+    {
+      "epoch": 0.18654923939151322,
+      "grad_norm": 1.106574827128144,
+      "learning_rate": 1.8742248024194073e-05,
+      "loss": 1.6404,
+      "step": 1398
+    },
+    {
+      "epoch": 0.18668267947691486,
+      "grad_norm": 1.1254058920479488,
+      "learning_rate": 1.8740148830633534e-05,
+      "loss": 1.6456,
+      "step": 1399
+    },
+    {
+      "epoch": 0.18681611956231653,
+      "grad_norm": 1.1988842778577748,
+      "learning_rate": 1.8738048004512707e-05,
+      "loss": 1.6735,
+      "step": 1400
+    },
+    {
+      "epoch": 0.18694955964771817,
+      "grad_norm": 1.1597783939590798,
+      "learning_rate": 1.8735945546224e-05,
+      "loss": 1.6463,
+      "step": 1401
+    },
+    {
+      "epoch": 0.18708299973311984,
+      "grad_norm": 1.1108971920248814,
+      "learning_rate": 1.8733841456160123e-05,
+      "loss": 1.7286,
+      "step": 1402
+    },
+    {
+      "epoch": 0.18721643981852149,
+      "grad_norm": 1.195745553995662,
+      "learning_rate": 1.8731735734714106e-05,
+      "loss": 1.655,
+      "step": 1403
+    },
+    {
+      "epoch": 0.18734987990392313,
+      "grad_norm": 1.1437291616721488,
+      "learning_rate": 1.8729628382279265e-05,
+      "loss": 1.7116,
+      "step": 1404
+    },
+    {
+      "epoch": 0.1874833199893248,
+      "grad_norm": 1.2582582933611897,
+      "learning_rate": 1.872751939924923e-05,
+      "loss": 1.6831,
+      "step": 1405
+    },
+    {
+      "epoch": 0.18761676007472644,
+      "grad_norm": 1.126051643445675,
+      "learning_rate": 1.8725408786017942e-05,
+      "loss": 1.6487,
+      "step": 1406
+    },
+    {
+      "epoch": 0.1877502001601281,
+      "grad_norm": 1.2320240819480008,
+      "learning_rate": 1.8723296542979635e-05,
+      "loss": 1.7051,
+      "step": 1407
+    },
+    {
+      "epoch": 0.18788364024552975,
+      "grad_norm": 1.1614213897490868,
+      "learning_rate": 1.872118267052885e-05,
+      "loss": 1.6758,
+      "step": 1408
+    },
+    {
+      "epoch": 0.18801708033093142,
+      "grad_norm": 1.2384327956894376,
+      "learning_rate": 1.8719067169060436e-05,
+      "loss": 1.6544,
+      "step": 1409
+    },
+    {
+      "epoch": 0.18815052041633307,
+      "grad_norm": 1.4000168772794017,
+      "learning_rate": 1.8716950038969547e-05,
+      "loss": 1.6822,
+      "step": 1410
+    },
+    {
+      "epoch": 0.1882839605017347,
+      "grad_norm": 1.1742590284506205,
+      "learning_rate": 1.871483128065164e-05,
+      "loss": 1.6987,
+      "step": 1411
+    },
+    {
+      "epoch": 0.18841740058713638,
+      "grad_norm": 1.206314558147808,
+      "learning_rate": 1.8712710894502465e-05,
+      "loss": 1.6341,
+      "step": 1412
+    },
+    {
+      "epoch": 0.18855084067253802,
+      "grad_norm": 1.199751013189083,
+      "learning_rate": 1.8710588880918098e-05,
+      "loss": 1.6946,
+      "step": 1413
+    },
+    {
+      "epoch": 0.1886842807579397,
+      "grad_norm": 1.2017855345339992,
+      "learning_rate": 1.8708465240294902e-05,
+      "loss": 1.6726,
+      "step": 1414
+    },
+    {
+      "epoch": 0.18881772084334134,
+      "grad_norm": 1.2227953100114268,
+      "learning_rate": 1.870633997302955e-05,
+      "loss": 1.6407,
+      "step": 1415
+    },
+    {
+      "epoch": 0.188951160928743,
+      "grad_norm": 1.2964391199169882,
+      "learning_rate": 1.8704213079519008e-05,
+      "loss": 1.6367,
+      "step": 1416
+    },
+    {
+      "epoch": 0.18908460101414465,
+      "grad_norm": 1.2207494370258463,
+      "learning_rate": 1.8702084560160572e-05,
+      "loss": 1.7262,
+      "step": 1417
+    },
+    {
+      "epoch": 0.1892180410995463,
+      "grad_norm": 5.905607034298311,
+      "learning_rate": 1.8699954415351813e-05,
+      "loss": 1.5948,
+      "step": 1418
+    },
+    {
+      "epoch": 0.18935148118494796,
+      "grad_norm": 1.1469661893942433,
+      "learning_rate": 1.8697822645490625e-05,
+      "loss": 1.6524,
+      "step": 1419
+    },
+    {
+      "epoch": 0.1894849212703496,
+      "grad_norm": 1.1259620274025783,
+      "learning_rate": 1.8695689250975193e-05,
+      "loss": 1.6331,
+      "step": 1420
+    },
+    {
+      "epoch": 0.18961836135575127,
+      "grad_norm": 1.0763567309071966,
+      "learning_rate": 1.8693554232204014e-05,
+      "loss": 1.7,
+      "step": 1421
+    },
+    {
+      "epoch": 0.18975180144115292,
+      "grad_norm": 1.1307077011187483,
+      "learning_rate": 1.8691417589575885e-05,
+      "loss": 1.6599,
+      "step": 1422
+    },
+    {
+      "epoch": 0.1898852415265546,
+      "grad_norm": 1.2777295336158558,
+      "learning_rate": 1.8689279323489902e-05,
+      "loss": 1.6505,
+      "step": 1423
+    },
+    {
+      "epoch": 0.19001868161195623,
+      "grad_norm": 1.2379608805979987,
+      "learning_rate": 1.8687139434345478e-05,
+      "loss": 1.6474,
+      "step": 1424
+    },
+    {
+      "epoch": 0.19015212169735787,
+      "grad_norm": 1.080970192917941,
+      "learning_rate": 1.8684997922542313e-05,
+      "loss": 1.6836,
+      "step": 1425
+    },
+    {
+      "epoch": 0.19028556178275954,
+      "grad_norm": 1.0712881958144271,
+      "learning_rate": 1.8682854788480417e-05,
+      "loss": 1.6071,
+      "step": 1426
+    },
+    {
+      "epoch": 0.1904190018681612,
+      "grad_norm": 1.26548294983667,
+      "learning_rate": 1.8680710032560102e-05,
+      "loss": 1.6904,
+      "step": 1427
+    },
+    {
+      "epoch": 0.19055244195356286,
+      "grad_norm": 1.0469874519952864,
+      "learning_rate": 1.867856365518199e-05,
+      "loss": 1.6422,
+      "step": 1428
+    },
+    {
+      "epoch": 0.1906858820389645,
+      "grad_norm": 1.0942346487916967,
+      "learning_rate": 1.8676415656746996e-05,
+      "loss": 1.6864,
+      "step": 1429
+    },
+    {
+      "epoch": 0.19081932212436617,
+      "grad_norm": 1.103259467200675,
+      "learning_rate": 1.8674266037656343e-05,
+      "loss": 1.6855,
+      "step": 1430
+    },
+    {
+      "epoch": 0.1909527622097678,
+      "grad_norm": 1.0683367237417463,
+      "learning_rate": 1.867211479831155e-05,
+      "loss": 1.6525,
+      "step": 1431
+    },
+    {
+      "epoch": 0.19108620229516948,
+      "grad_norm": 1.1121615299096883,
+      "learning_rate": 1.8669961939114457e-05,
+      "loss": 1.6904,
+      "step": 1432
+    },
+    {
+      "epoch": 0.19121964238057113,
+      "grad_norm": 1.1617841181260873,
+      "learning_rate": 1.8667807460467178e-05,
+      "loss": 1.6352,
+      "step": 1433
+    },
+    {
+      "epoch": 0.19135308246597277,
+      "grad_norm": 1.084146580072923,
+      "learning_rate": 1.8665651362772152e-05,
+      "loss": 1.6905,
+      "step": 1434
+    },
+    {
+      "epoch": 0.19148652255137444,
+      "grad_norm": 1.1047653403081839,
+      "learning_rate": 1.8663493646432116e-05,
+      "loss": 1.7152,
+      "step": 1435
+    },
+    {
+      "epoch": 0.19161996263677608,
+      "grad_norm": 1.0986414741598525,
+      "learning_rate": 1.8661334311850104e-05,
+      "loss": 1.6874,
+      "step": 1436
+    },
+    {
+      "epoch": 0.19175340272217775,
+      "grad_norm": 1.3689551704009346,
+      "learning_rate": 1.8659173359429455e-05,
+      "loss": 1.6464,
+      "step": 1437
+    },
+    {
+      "epoch": 0.1918868428075794,
+      "grad_norm": 1.2060959441159418,
+      "learning_rate": 1.8657010789573812e-05,
+      "loss": 1.7476,
+      "step": 1438
+    },
+    {
+      "epoch": 0.19202028289298106,
+      "grad_norm": 1.094136327662655,
+      "learning_rate": 1.8654846602687112e-05,
+      "loss": 1.6862,
+      "step": 1439
+    },
+    {
+      "epoch": 0.1921537229783827,
+      "grad_norm": 1.1003792674170316,
+      "learning_rate": 1.865268079917361e-05,
+      "loss": 1.7001,
+      "step": 1440
+    },
+    {
+      "epoch": 0.19228716306378435,
+      "grad_norm": 1.397260975809977,
+      "learning_rate": 1.8650513379437854e-05,
+      "loss": 1.6541,
+      "step": 1441
+    },
+    {
+      "epoch": 0.19242060314918602,
+      "grad_norm": 1.1317586552025098,
+      "learning_rate": 1.864834434388468e-05,
+      "loss": 1.654,
+      "step": 1442
+    },
+    {
+      "epoch": 0.19255404323458766,
+      "grad_norm": 1.2801188417922258,
+      "learning_rate": 1.8646173692919254e-05,
+      "loss": 1.72,
+      "step": 1443
+    },
+    {
+      "epoch": 0.19268748331998933,
+      "grad_norm": 1.1044353428107876,
+      "learning_rate": 1.8644001426947017e-05,
+      "loss": 1.6633,
+      "step": 1444
+    },
+    {
+      "epoch": 0.19282092340539098,
+      "grad_norm": 1.7385377002254025,
+      "learning_rate": 1.8641827546373736e-05,
+      "loss": 1.6934,
+      "step": 1445
+    },
+    {
+      "epoch": 0.19295436349079265,
+      "grad_norm": 1.1251578855313504,
+      "learning_rate": 1.8639652051605454e-05,
+      "loss": 1.6855,
+      "step": 1446
+    },
+    {
+      "epoch": 0.1930878035761943,
+      "grad_norm": 1.0823798317472224,
+      "learning_rate": 1.8637474943048538e-05,
+      "loss": 1.6851,
+      "step": 1447
+    },
+    {
+      "epoch": 0.19322124366159593,
+      "grad_norm": 1.2414924209927352,
+      "learning_rate": 1.8635296221109643e-05,
+      "loss": 1.6342,
+      "step": 1448
+    },
+    {
+      "epoch": 0.1933546837469976,
+      "grad_norm": 1.081073074757343,
+      "learning_rate": 1.8633115886195733e-05,
+      "loss": 1.6326,
+      "step": 1449
+    },
+    {
+      "epoch": 0.19348812383239924,
+      "grad_norm": 1.3276067715392756,
+      "learning_rate": 1.863093393871406e-05,
+      "loss": 1.682,
+      "step": 1450
+    },
+    {
+      "epoch": 0.19362156391780092,
+      "grad_norm": 1.0512206782809177,
+      "learning_rate": 1.86287503790722e-05,
+      "loss": 1.6339,
+      "step": 1451
+    },
+    {
+      "epoch": 0.19375500400320256,
+      "grad_norm": 1.1118643544696887,
+      "learning_rate": 1.8626565207678014e-05,
+      "loss": 1.6412,
+      "step": 1452
+    },
+    {
+      "epoch": 0.19388844408860423,
+      "grad_norm": 1.1881349320604198,
+      "learning_rate": 1.862437842493966e-05,
+      "loss": 1.6535,
+      "step": 1453
+    },
+    {
+      "epoch": 0.19402188417400587,
+      "grad_norm": 1.0650617509844336,
+      "learning_rate": 1.8622190031265608e-05,
+      "loss": 1.6608,
+      "step": 1454
+    },
+    {
+      "epoch": 0.1941553242594075,
+      "grad_norm": 1.2022764202107716,
+      "learning_rate": 1.8620000027064625e-05,
+      "loss": 1.6818,
+      "step": 1455
+    },
+    {
+      "epoch": 0.19428876434480918,
+      "grad_norm": 1.0945586894168828,
+      "learning_rate": 1.861780841274578e-05,
+      "loss": 1.6518,
+      "step": 1456
+    },
+    {
+      "epoch": 0.19442220443021083,
+      "grad_norm": 1.181691965256765,
+      "learning_rate": 1.8615615188718442e-05,
+      "loss": 1.6526,
+      "step": 1457
+    },
+    {
+      "epoch": 0.1945556445156125,
+      "grad_norm": 1.21873002721777,
+      "learning_rate": 1.861342035539228e-05,
+      "loss": 1.6685,
+      "step": 1458
+    },
+    {
+      "epoch": 0.19468908460101414,
+      "grad_norm": 1.1888383727594938,
+      "learning_rate": 1.8611223913177264e-05,
+      "loss": 1.6939,
+      "step": 1459
+    },
+    {
+      "epoch": 0.1948225246864158,
+      "grad_norm": 1.1092188067858701,
+      "learning_rate": 1.8609025862483658e-05,
+      "loss": 1.7001,
+      "step": 1460
+    },
+    {
+      "epoch": 0.19495596477181745,
+      "grad_norm": 1.4312758954493943,
+      "learning_rate": 1.8606826203722043e-05,
+      "loss": 1.6334,
+      "step": 1461
+    },
+    {
+      "epoch": 0.1950894048572191,
+      "grad_norm": 1.2022595063751074,
+      "learning_rate": 1.8604624937303285e-05,
+      "loss": 1.6806,
+      "step": 1462
+    },
+    {
+      "epoch": 0.19522284494262077,
+      "grad_norm": 1.0171052837252916,
+      "learning_rate": 1.8602422063638554e-05,
+      "loss": 1.6488,
+      "step": 1463
+    },
+    {
+      "epoch": 0.1953562850280224,
+      "grad_norm": 1.3510050011424308,
+      "learning_rate": 1.860021758313933e-05,
+      "loss": 1.6833,
+      "step": 1464
+    },
+    {
+      "epoch": 0.19548972511342408,
+      "grad_norm": 1.257684566145725,
+      "learning_rate": 1.859801149621737e-05,
+      "loss": 1.6647,
+      "step": 1465
+    },
+    {
+      "epoch": 0.19562316519882572,
+      "grad_norm": 1.1217536119317055,
+      "learning_rate": 1.8595803803284757e-05,
+      "loss": 1.6662,
+      "step": 1466
+    },
+    {
+      "epoch": 0.1957566052842274,
+      "grad_norm": 1.1668773044915945,
+      "learning_rate": 1.8593594504753863e-05,
+      "loss": 1.711,
+      "step": 1467
+    },
+    {
+      "epoch": 0.19589004536962903,
+      "grad_norm": 1.1958090413202758,
+      "learning_rate": 1.8591383601037357e-05,
+      "loss": 1.709,
+      "step": 1468
+    },
+    {
+      "epoch": 0.1960234854550307,
+      "grad_norm": 1.3120713864562594,
+      "learning_rate": 1.8589171092548208e-05,
+      "loss": 1.6493,
+      "step": 1469
+    },
+    {
+      "epoch": 0.19615692554043235,
+      "grad_norm": 1.1266277557758897,
+      "learning_rate": 1.8586956979699692e-05,
+      "loss": 1.6786,
+      "step": 1470
+    },
+    {
+      "epoch": 0.196290365625834,
+      "grad_norm": 2.464652796809538,
+      "learning_rate": 1.858474126290538e-05,
+      "loss": 1.6538,
+      "step": 1471
+    },
+    {
+      "epoch": 0.19642380571123566,
+      "grad_norm": 1.358175538034263,
+      "learning_rate": 1.8582523942579138e-05,
+      "loss": 1.6298,
+      "step": 1472
+    },
+    {
+      "epoch": 0.1965572457966373,
+      "grad_norm": 1.1218515326836789,
+      "learning_rate": 1.8580305019135144e-05,
+      "loss": 1.6807,
+      "step": 1473
+    },
+    {
+      "epoch": 0.19669068588203897,
+      "grad_norm": 1.1314598132036782,
+      "learning_rate": 1.8578084492987857e-05,
+      "loss": 1.6085,
+      "step": 1474
+    },
+    {
+      "epoch": 0.19682412596744062,
+      "grad_norm": 1.0806632798443008,
+      "learning_rate": 1.8575862364552056e-05,
+      "loss": 1.6943,
+      "step": 1475
+    },
+    {
+      "epoch": 0.1969575660528423,
+      "grad_norm": 1.0456494280144883,
+      "learning_rate": 1.8573638634242806e-05,
+      "loss": 1.6425,
+      "step": 1476
+    },
+    {
+      "epoch": 0.19709100613824393,
+      "grad_norm": 1.4643633002109615,
+      "learning_rate": 1.857141330247547e-05,
+      "loss": 1.717,
+      "step": 1477
+    },
+    {
+      "epoch": 0.19722444622364557,
+      "grad_norm": 1.3907399472017639,
+      "learning_rate": 1.856918636966572e-05,
+      "loss": 1.6938,
+      "step": 1478
+    },
+    {
+      "epoch": 0.19735788630904724,
+      "grad_norm": 1.074837673820265,
+      "learning_rate": 1.856695783622952e-05,
+      "loss": 1.6547,
+      "step": 1479
+    },
+    {
+      "epoch": 0.19749132639444889,
+      "grad_norm": 1.1065228121324435,
+      "learning_rate": 1.8564727702583132e-05,
+      "loss": 1.6468,
+      "step": 1480
+    },
+    {
+      "epoch": 0.19762476647985056,
+      "grad_norm": 1.1931788828028584,
+      "learning_rate": 1.8562495969143125e-05,
+      "loss": 1.658,
+      "step": 1481
+    },
+    {
+      "epoch": 0.1977582065652522,
+      "grad_norm": 1.213866762163482,
+      "learning_rate": 1.8560262636326358e-05,
+      "loss": 1.631,
+      "step": 1482
+    },
+    {
+      "epoch": 0.19789164665065387,
+      "grad_norm": 1.099947809835569,
+      "learning_rate": 1.8558027704549993e-05,
+      "loss": 1.6351,
+      "step": 1483
+    },
+    {
+      "epoch": 0.1980250867360555,
+      "grad_norm": 1.1088893853131983,
+      "learning_rate": 1.8555791174231494e-05,
+      "loss": 1.6723,
+      "step": 1484
+    },
+    {
+      "epoch": 0.19815852682145715,
+      "grad_norm": 1.1011144028074373,
+      "learning_rate": 1.855355304578861e-05,
+      "loss": 1.707,
+      "step": 1485
+    },
+    {
+      "epoch": 0.19829196690685882,
+      "grad_norm": 1.3818385213484043,
+      "learning_rate": 1.85513133196394e-05,
+      "loss": 1.652,
+      "step": 1486
+    },
+    {
+      "epoch": 0.19842540699226047,
+      "grad_norm": 1.1229293638925342,
+      "learning_rate": 1.8549071996202225e-05,
+      "loss": 1.656,
+      "step": 1487
+    },
+    {
+      "epoch": 0.19855884707766214,
+      "grad_norm": 1.080075119649869,
+      "learning_rate": 1.8546829075895737e-05,
+      "loss": 1.7099,
+      "step": 1488
+    },
+    {
+      "epoch": 0.19869228716306378,
+      "grad_norm": 1.2852126608442598,
+      "learning_rate": 1.8544584559138888e-05,
+      "loss": 1.7062,
+      "step": 1489
+    },
+    {
+      "epoch": 0.19882572724846545,
+      "grad_norm": 1.14699198879905,
+      "learning_rate": 1.8542338446350924e-05,
+      "loss": 1.6963,
+      "step": 1490
+    },
+    {
+      "epoch": 0.1989591673338671,
+      "grad_norm": 1.0866913345021045,
+      "learning_rate": 1.85400907379514e-05,
+      "loss": 1.692,
+      "step": 1491
+    },
+    {
+      "epoch": 0.19909260741926874,
+      "grad_norm": 1.1345754144326983,
+      "learning_rate": 1.853784143436016e-05,
+      "loss": 1.6911,
+      "step": 1492
+    },
+    {
+      "epoch": 0.1992260475046704,
+      "grad_norm": 1.0859201677475647,
+      "learning_rate": 1.853559053599734e-05,
+      "loss": 1.7136,
+      "step": 1493
+    },
+    {
+      "epoch": 0.19935948759007205,
+      "grad_norm": 1.114793968801779,
+      "learning_rate": 1.8533338043283392e-05,
+      "loss": 1.6262,
+      "step": 1494
+    },
+    {
+      "epoch": 0.19949292767547372,
+      "grad_norm": 1.0646044544826536,
+      "learning_rate": 1.8531083956639054e-05,
+      "loss": 1.6374,
+      "step": 1495
+    },
+    {
+      "epoch": 0.19962636776087536,
+      "grad_norm": 1.0943515393210304,
+      "learning_rate": 1.8528828276485365e-05,
+      "loss": 1.6565,
+      "step": 1496
+    },
+    {
+      "epoch": 0.19975980784627703,
+      "grad_norm": 1.153219518432924,
+      "learning_rate": 1.852657100324366e-05,
+      "loss": 1.6468,
+      "step": 1497
+    },
+    {
+      "epoch": 0.19989324793167867,
+      "grad_norm": 1.3658364399948308,
+      "learning_rate": 1.8524312137335565e-05,
+      "loss": 1.6546,
+      "step": 1498
+    },
+    {
+      "epoch": 0.20002668801708032,
+      "grad_norm": 1.0830597310395975,
+      "learning_rate": 1.8522051679183017e-05,
+      "loss": 1.6464,
+      "step": 1499
+    },
+    {
+      "epoch": 0.200160128102482,
+      "grad_norm": 1.0897256468244192,
+      "learning_rate": 1.8519789629208245e-05,
+      "loss": 1.6335,
+      "step": 1500
+    },
+    {
+      "epoch": 0.20029356818788363,
+      "grad_norm": 1.1152974632184582,
+      "learning_rate": 1.851752598783377e-05,
+      "loss": 1.6539,
+      "step": 1501
+    },
+    {
+      "epoch": 0.2004270082732853,
+      "grad_norm": 1.070943336830069,
+      "learning_rate": 1.8515260755482414e-05,
+      "loss": 1.663,
+      "step": 1502
+    },
+    {
+      "epoch": 0.20056044835868694,
+      "grad_norm": 1.1301005431558968,
+      "learning_rate": 1.85129939325773e-05,
+      "loss": 1.6386,
+      "step": 1503
+    },
+    {
+      "epoch": 0.20069388844408861,
+      "grad_norm": 1.4376814743628707,
+      "learning_rate": 1.8510725519541845e-05,
+      "loss": 1.6777,
+      "step": 1504
+    },
+    {
+      "epoch": 0.20082732852949026,
+      "grad_norm": 1.1571460766809745,
+      "learning_rate": 1.8508455516799762e-05,
+      "loss": 1.6458,
+      "step": 1505
+    },
+    {
+      "epoch": 0.20096076861489193,
+      "grad_norm": 1.526537352021282,
+      "learning_rate": 1.850618392477506e-05,
+      "loss": 1.6673,
+      "step": 1506
+    },
+    {
+      "epoch": 0.20109420870029357,
+      "grad_norm": 1.171428989010074,
+      "learning_rate": 1.8503910743892044e-05,
+      "loss": 1.6548,
+      "step": 1507
+    },
+    {
+      "epoch": 0.2012276487856952,
+      "grad_norm": 1.0654992362564706,
+      "learning_rate": 1.8501635974575328e-05,
+      "loss": 1.6856,
+      "step": 1508
+    },
+    {
+      "epoch": 0.20136108887109688,
+      "grad_norm": 1.2869430405209437,
+      "learning_rate": 1.84993596172498e-05,
+      "loss": 1.6663,
+      "step": 1509
+    },
+    {
+      "epoch": 0.20149452895649853,
+      "grad_norm": 1.3280855964940679,
+      "learning_rate": 1.849708167234067e-05,
+      "loss": 1.6049,
+      "step": 1510
+    },
+    {
+      "epoch": 0.2016279690419002,
+      "grad_norm": 1.2170509648131613,
+      "learning_rate": 1.8494802140273423e-05,
+      "loss": 1.6645,
+      "step": 1511
+    },
+    {
+      "epoch": 0.20176140912730184,
+      "grad_norm": 1.1693660112296544,
+      "learning_rate": 1.8492521021473854e-05,
+      "loss": 1.6448,
+      "step": 1512
+    },
+    {
+      "epoch": 0.2018948492127035,
+      "grad_norm": 1.1495960407236363,
+      "learning_rate": 1.8490238316368048e-05,
+      "loss": 1.6717,
+      "step": 1513
+    },
+    {
+      "epoch": 0.20202828929810515,
+      "grad_norm": 1.0607525124108388,
+      "learning_rate": 1.848795402538239e-05,
+      "loss": 1.685,
+      "step": 1514
+    },
+    {
+      "epoch": 0.2021617293835068,
+      "grad_norm": 1.1588629194975635,
+      "learning_rate": 1.848566814894356e-05,
+      "loss": 1.6733,
+      "step": 1515
+    },
+    {
+      "epoch": 0.20229516946890846,
+      "grad_norm": 1.0477044055470504,
+      "learning_rate": 1.8483380687478535e-05,
+      "loss": 1.6364,
+      "step": 1516
+    },
+    {
+      "epoch": 0.2024286095543101,
+      "grad_norm": 1.3539461437967024,
+      "learning_rate": 1.848109164141458e-05,
+      "loss": 1.6329,
+      "step": 1517
+    },
+    {
+      "epoch": 0.20256204963971178,
+      "grad_norm": 1.316562445670924,
+      "learning_rate": 1.847880101117927e-05,
+      "loss": 1.7161,
+      "step": 1518
+    },
+    {
+      "epoch": 0.20269548972511342,
+      "grad_norm": 1.1325223006869773,
+      "learning_rate": 1.8476508797200464e-05,
+      "loss": 1.6976,
+      "step": 1519
+    },
+    {
+      "epoch": 0.2028289298105151,
+      "grad_norm": 1.1271240323882026,
+      "learning_rate": 1.847421499990633e-05,
+      "loss": 1.6724,
+      "step": 1520
+    },
+    {
+      "epoch": 0.20296236989591673,
+      "grad_norm": 1.1553839820996277,
+      "learning_rate": 1.8471919619725307e-05,
+      "loss": 1.6322,
+      "step": 1521
+    },
+    {
+      "epoch": 0.20309580998131838,
+      "grad_norm": 1.1027336658945248,
+      "learning_rate": 1.8469622657086158e-05,
+      "loss": 1.6447,
+      "step": 1522
+    },
+    {
+      "epoch": 0.20322925006672005,
+      "grad_norm": 1.183883565807101,
+      "learning_rate": 1.846732411241793e-05,
+      "loss": 1.6999,
+      "step": 1523
+    },
+    {
+      "epoch": 0.2033626901521217,
+      "grad_norm": 1.1642046420989085,
+      "learning_rate": 1.846502398614996e-05,
+      "loss": 1.7149,
+      "step": 1524
+    },
+    {
+      "epoch": 0.20349613023752336,
+      "grad_norm": 1.0755063096784014,
+      "learning_rate": 1.8462722278711887e-05,
+      "loss": 1.5943,
+      "step": 1525
+    },
+    {
+      "epoch": 0.203629570322925,
+      "grad_norm": 1.1419970032576714,
+      "learning_rate": 1.8460418990533647e-05,
+      "loss": 1.7073,
+      "step": 1526
+    },
+    {
+      "epoch": 0.20376301040832667,
+      "grad_norm": 1.302454031541249,
+      "learning_rate": 1.845811412204546e-05,
+      "loss": 1.6758,
+      "step": 1527
+    },
+    {
+      "epoch": 0.20389645049372832,
+      "grad_norm": 1.1668477615464077,
+      "learning_rate": 1.845580767367786e-05,
+      "loss": 1.694,
+      "step": 1528
+    },
+    {
+      "epoch": 0.20402989057912996,
+      "grad_norm": 1.1087107394603948,
+      "learning_rate": 1.8453499645861656e-05,
+      "loss": 1.691,
+      "step": 1529
+    },
+    {
+      "epoch": 0.20416333066453163,
+      "grad_norm": 1.0455825217353827,
+      "learning_rate": 1.8451190039027964e-05,
+      "loss": 1.6684,
+      "step": 1530
+    },
+    {
+      "epoch": 0.20429677074993327,
+      "grad_norm": 1.219445656841785,
+      "learning_rate": 1.8448878853608195e-05,
+      "loss": 1.7155,
+      "step": 1531
+    },
+    {
+      "epoch": 0.20443021083533494,
+      "grad_norm": 1.335960910259822,
+      "learning_rate": 1.8446566090034053e-05,
+      "loss": 1.7326,
+      "step": 1532
+    },
+    {
+      "epoch": 0.20456365092073658,
+      "grad_norm": 1.271132783595969,
+      "learning_rate": 1.8444251748737526e-05,
+      "loss": 1.6686,
+      "step": 1533
+    },
+    {
+      "epoch": 0.20469709100613825,
+      "grad_norm": 1.0514626639526814,
+      "learning_rate": 1.844193583015092e-05,
+      "loss": 1.6366,
+      "step": 1534
+    },
+    {
+      "epoch": 0.2048305310915399,
+      "grad_norm": 1.0454900741987578,
+      "learning_rate": 1.8439618334706812e-05,
+      "loss": 1.6237,
+      "step": 1535
+    },
+    {
+      "epoch": 0.20496397117694154,
+      "grad_norm": 1.073823751577798,
+      "learning_rate": 1.8437299262838092e-05,
+      "loss": 1.6749,
+      "step": 1536
+    },
+    {
+      "epoch": 0.2050974112623432,
+      "grad_norm": 1.2798723643570793,
+      "learning_rate": 1.843497861497793e-05,
+      "loss": 1.6411,
+      "step": 1537
+    },
+    {
+      "epoch": 0.20523085134774485,
+      "grad_norm": 1.1135912832633508,
+      "learning_rate": 1.8432656391559794e-05,
+      "loss": 1.63,
+      "step": 1538
+    },
+    {
+      "epoch": 0.20536429143314652,
+      "grad_norm": 1.0600593382175174,
+      "learning_rate": 1.8430332593017463e-05,
+      "loss": 1.6932,
+      "step": 1539
+    },
+    {
+      "epoch": 0.20549773151854817,
+      "grad_norm": 1.243565535810705,
+      "learning_rate": 1.8428007219784978e-05,
+      "loss": 1.6734,
+      "step": 1540
+    },
+    {
+      "epoch": 0.20563117160394984,
+      "grad_norm": 1.1107029670183746,
+      "learning_rate": 1.8425680272296706e-05,
+      "loss": 1.6954,
+      "step": 1541
+    },
+    {
+      "epoch": 0.20576461168935148,
+      "grad_norm": 1.1611386236626087,
+      "learning_rate": 1.8423351750987285e-05,
+      "loss": 1.6676,
+      "step": 1542
+    },
+    {
+      "epoch": 0.20589805177475315,
+      "grad_norm": 1.0922966177873632,
+      "learning_rate": 1.842102165629166e-05,
+      "loss": 1.6751,
+      "step": 1543
+    },
+    {
+      "epoch": 0.2060314918601548,
+      "grad_norm": 1.1728266705222763,
+      "learning_rate": 1.8418689988645072e-05,
+      "loss": 1.6662,
+      "step": 1544
+    },
+    {
+      "epoch": 0.20616493194555643,
+      "grad_norm": 1.281999128501973,
+      "learning_rate": 1.841635674848304e-05,
+      "loss": 1.6855,
+      "step": 1545
+    },
+    {
+      "epoch": 0.2062983720309581,
+      "grad_norm": 1.1950922902237124,
+      "learning_rate": 1.8414021936241394e-05,
+      "loss": 1.6543,
+      "step": 1546
+    },
+    {
+      "epoch": 0.20643181211635975,
+      "grad_norm": 1.1228445170626977,
+      "learning_rate": 1.8411685552356244e-05,
+      "loss": 1.6518,
+      "step": 1547
+    },
+    {
+      "epoch": 0.20656525220176142,
+      "grad_norm": 1.2061203106698866,
+      "learning_rate": 1.840934759726401e-05,
+      "loss": 1.6469,
+      "step": 1548
+    },
+    {
+      "epoch": 0.20669869228716306,
+      "grad_norm": 1.1059566734641804,
+      "learning_rate": 1.840700807140138e-05,
+      "loss": 1.6259,
+      "step": 1549
+    },
+    {
+      "epoch": 0.20683213237256473,
+      "grad_norm": 1.1551811421928873,
+      "learning_rate": 1.8404666975205367e-05,
+      "loss": 1.6333,
+      "step": 1550
+    },
+    {
+      "epoch": 0.20696557245796637,
+      "grad_norm": 1.1528270603987256,
+      "learning_rate": 1.840232430911325e-05,
+      "loss": 1.7053,
+      "step": 1551
+    },
+    {
+      "epoch": 0.20709901254336802,
+      "grad_norm": 1.0615331899156355,
+      "learning_rate": 1.8399980073562615e-05,
+      "loss": 1.6093,
+      "step": 1552
+    },
+    {
+      "epoch": 0.2072324526287697,
+      "grad_norm": 1.2018501248996023,
+      "learning_rate": 1.8397634268991343e-05,
+      "loss": 1.6831,
+      "step": 1553
+    },
+    {
+      "epoch": 0.20736589271417133,
+      "grad_norm": 1.129260522162621,
+      "learning_rate": 1.8395286895837598e-05,
+      "loss": 1.6725,
+      "step": 1554
+    },
+    {
+      "epoch": 0.207499332799573,
+      "grad_norm": 1.2931052464354142,
+      "learning_rate": 1.8392937954539845e-05,
+      "loss": 1.6261,
+      "step": 1555
+    },
+    {
+      "epoch": 0.20763277288497464,
+      "grad_norm": 1.158760216852382,
+      "learning_rate": 1.8390587445536844e-05,
+      "loss": 1.6595,
+      "step": 1556
+    },
+    {
+      "epoch": 0.2077662129703763,
+      "grad_norm": 1.16264835183651,
+      "learning_rate": 1.8388235369267632e-05,
+      "loss": 1.6538,
+      "step": 1557
+    },
+    {
+      "epoch": 0.20789965305577796,
+      "grad_norm": 1.1390190977362333,
+      "learning_rate": 1.838588172617156e-05,
+      "loss": 1.6174,
+      "step": 1558
+    },
+    {
+      "epoch": 0.2080330931411796,
+      "grad_norm": 1.0281490566989566,
+      "learning_rate": 1.8383526516688257e-05,
+      "loss": 1.6256,
+      "step": 1559
+    },
+    {
+      "epoch": 0.20816653322658127,
+      "grad_norm": 1.3060131907983772,
+      "learning_rate": 1.8381169741257653e-05,
+      "loss": 1.6876,
+      "step": 1560
+    },
+    {
+      "epoch": 0.2082999733119829,
+      "grad_norm": 1.0851729874752505,
+      "learning_rate": 1.8378811400319962e-05,
+      "loss": 1.6448,
+      "step": 1561
+    },
+    {
+      "epoch": 0.20843341339738458,
+      "grad_norm": 1.1173227297527768,
+      "learning_rate": 1.8376451494315697e-05,
+      "loss": 1.6397,
+      "step": 1562
+    },
+    {
+      "epoch": 0.20856685348278622,
+      "grad_norm": 1.1250418404408145,
+      "learning_rate": 1.8374090023685664e-05,
+      "loss": 1.7092,
+      "step": 1563
+    },
+    {
+      "epoch": 0.2087002935681879,
+      "grad_norm": 1.1313533969061775,
+      "learning_rate": 1.8371726988870956e-05,
+      "loss": 1.6625,
+      "step": 1564
+    },
+    {
+      "epoch": 0.20883373365358954,
+      "grad_norm": 1.1578578516746643,
+      "learning_rate": 1.8369362390312965e-05,
+      "loss": 1.6829,
+      "step": 1565
+    },
+    {
+      "epoch": 0.20896717373899118,
+      "grad_norm": 1.0419177694211474,
+      "learning_rate": 1.8366996228453366e-05,
+      "loss": 1.6711,
+      "step": 1566
+    },
+    {
+      "epoch": 0.20910061382439285,
+      "grad_norm": 1.110808392383394,
+      "learning_rate": 1.8364628503734133e-05,
+      "loss": 1.6313,
+      "step": 1567
+    },
+    {
+      "epoch": 0.2092340539097945,
+      "grad_norm": 1.0553327445995773,
+      "learning_rate": 1.8362259216597532e-05,
+      "loss": 1.6769,
+      "step": 1568
+    },
+    {
+      "epoch": 0.20936749399519616,
+      "grad_norm": 1.1605648342199986,
+      "learning_rate": 1.8359888367486116e-05,
+      "loss": 1.6423,
+      "step": 1569
+    },
+    {
+      "epoch": 0.2095009340805978,
+      "grad_norm": 1.3528845887090233,
+      "learning_rate": 1.8357515956842735e-05,
+      "loss": 1.6641,
+      "step": 1570
+    },
+    {
+      "epoch": 0.20963437416599948,
+      "grad_norm": 1.0381834374569774,
+      "learning_rate": 1.8355141985110524e-05,
+      "loss": 1.6645,
+      "step": 1571
+    },
+    {
+      "epoch": 0.20976781425140112,
+      "grad_norm": 1.1268678857914454,
+      "learning_rate": 1.8352766452732922e-05,
+      "loss": 1.6252,
+      "step": 1572
+    },
+    {
+      "epoch": 0.20990125433680276,
+      "grad_norm": 1.07645387262839,
+      "learning_rate": 1.8350389360153645e-05,
+      "loss": 1.6234,
+      "step": 1573
+    },
+    {
+      "epoch": 0.21003469442220443,
+      "grad_norm": 1.1601184907461661,
+      "learning_rate": 1.8348010707816704e-05,
+      "loss": 1.6625,
+      "step": 1574
+    },
+    {
+      "epoch": 0.21016813450760607,
+      "grad_norm": 1.1598070223911565,
+      "learning_rate": 1.8345630496166413e-05,
+      "loss": 1.6072,
+      "step": 1575
+    },
+    {
+      "epoch": 0.21030157459300775,
+      "grad_norm": 1.0478963594602182,
+      "learning_rate": 1.834324872564736e-05,
+      "loss": 1.6713,
+      "step": 1576
+    },
+    {
+      "epoch": 0.2104350146784094,
+      "grad_norm": 1.1864899396092392,
+      "learning_rate": 1.8340865396704442e-05,
+      "loss": 1.6994,
+      "step": 1577
+    },
+    {
+      "epoch": 0.21056845476381106,
+      "grad_norm": 1.2082345518638513,
+      "learning_rate": 1.833848050978283e-05,
+      "loss": 1.6325,
+      "step": 1578
+    },
+    {
+      "epoch": 0.2107018948492127,
+      "grad_norm": 1.0705831509203583,
+      "learning_rate": 1.8336094065327994e-05,
+      "loss": 1.6455,
+      "step": 1579
+    },
+    {
+      "epoch": 0.21083533493461437,
+      "grad_norm": 1.0714080597971536,
+      "learning_rate": 1.8333706063785698e-05,
+      "loss": 1.6415,
+      "step": 1580
+    },
+    {
+      "epoch": 0.21096877502001601,
+      "grad_norm": 1.4855074921163713,
+      "learning_rate": 1.8331316505601995e-05,
+      "loss": 1.6629,
+      "step": 1581
+    },
+    {
+      "epoch": 0.21110221510541766,
+      "grad_norm": 1.0376544838658421,
+      "learning_rate": 1.832892539122322e-05,
+      "loss": 1.6218,
+      "step": 1582
+    },
+    {
+      "epoch": 0.21123565519081933,
+      "grad_norm": 1.0937886957684666,
+      "learning_rate": 1.832653272109601e-05,
+      "loss": 1.6446,
+      "step": 1583
+    },
+    {
+      "epoch": 0.21136909527622097,
+      "grad_norm": 0.9935625042710026,
+      "learning_rate": 1.8324138495667293e-05,
+      "loss": 1.5577,
+      "step": 1584
+    },
+    {
+      "epoch": 0.21150253536162264,
+      "grad_norm": 1.0414931198212998,
+      "learning_rate": 1.8321742715384277e-05,
+      "loss": 1.7021,
+      "step": 1585
+    },
+    {
+      "epoch": 0.21163597544702428,
+      "grad_norm": 1.0888067291778296,
+      "learning_rate": 1.8319345380694464e-05,
+      "loss": 1.6425,
+      "step": 1586
+    },
+    {
+      "epoch": 0.21176941553242595,
+      "grad_norm": 1.1889065641556036,
+      "learning_rate": 1.831694649204566e-05,
+      "loss": 1.6804,
+      "step": 1587
+    },
+    {
+      "epoch": 0.2119028556178276,
+      "grad_norm": 1.2027320435869355,
+      "learning_rate": 1.831454604988594e-05,
+      "loss": 1.6518,
+      "step": 1588
+    },
+    {
+      "epoch": 0.21203629570322924,
+      "grad_norm": 1.1633662750060685,
+      "learning_rate": 1.8312144054663682e-05,
+      "loss": 1.6471,
+      "step": 1589
+    },
+    {
+      "epoch": 0.2121697357886309,
+      "grad_norm": 1.62970080028646,
+      "learning_rate": 1.8309740506827552e-05,
+      "loss": 1.6254,
+      "step": 1590
+    },
+    {
+      "epoch": 0.21230317587403255,
+      "grad_norm": 1.0879307197813337,
+      "learning_rate": 1.8307335406826505e-05,
+      "loss": 1.6575,
+      "step": 1591
+    },
+    {
+      "epoch": 0.21243661595943422,
+      "grad_norm": 1.269123197918674,
+      "learning_rate": 1.830492875510979e-05,
+      "loss": 1.6364,
+      "step": 1592
+    },
+    {
+      "epoch": 0.21257005604483586,
+      "grad_norm": 1.0373985047234797,
+      "learning_rate": 1.8302520552126935e-05,
+      "loss": 1.6571,
+      "step": 1593
+    },
+    {
+      "epoch": 0.21270349613023753,
+      "grad_norm": 1.0542865827845302,
+      "learning_rate": 1.830011079832777e-05,
+      "loss": 1.6695,
+      "step": 1594
+    },
+    {
+      "epoch": 0.21283693621563918,
+      "grad_norm": 1.0402881306171436,
+      "learning_rate": 1.8297699494162406e-05,
+      "loss": 1.6562,
+      "step": 1595
+    },
+    {
+      "epoch": 0.21297037630104082,
+      "grad_norm": 1.1294113481127595,
+      "learning_rate": 1.8295286640081252e-05,
+      "loss": 1.6583,
+      "step": 1596
+    },
+    {
+      "epoch": 0.2131038163864425,
+      "grad_norm": 1.1895908197616782,
+      "learning_rate": 1.8292872236534996e-05,
+      "loss": 1.6487,
+      "step": 1597
+    },
+    {
+      "epoch": 0.21323725647184413,
+      "grad_norm": 1.0379115957592848,
+      "learning_rate": 1.829045628397463e-05,
+      "loss": 1.6322,
+      "step": 1598
+    },
+    {
+      "epoch": 0.2133706965572458,
+      "grad_norm": 1.222610128560367,
+      "learning_rate": 1.8288038782851417e-05,
+      "loss": 1.6657,
+      "step": 1599
+    },
+    {
+      "epoch": 0.21350413664264745,
+      "grad_norm": 1.130257915264502,
+      "learning_rate": 1.828561973361692e-05,
+      "loss": 1.6602,
+      "step": 1600
+    },
+    {
+      "epoch": 0.21363757672804912,
+      "grad_norm": 1.126575657802467,
+      "learning_rate": 1.8283199136722995e-05,
+      "loss": 1.6918,
+      "step": 1601
+    },
+    {
+      "epoch": 0.21377101681345076,
+      "grad_norm": 1.3153982419349068,
+      "learning_rate": 1.828077699262178e-05,
+      "loss": 1.6693,
+      "step": 1602
+    },
+    {
+      "epoch": 0.2139044568988524,
+      "grad_norm": 1.262078471065165,
+      "learning_rate": 1.8278353301765702e-05,
+      "loss": 1.6933,
+      "step": 1603
+    },
+    {
+      "epoch": 0.21403789698425407,
+      "grad_norm": 1.1173878288546095,
+      "learning_rate": 1.8275928064607478e-05,
+      "loss": 1.5781,
+      "step": 1604
+    },
+    {
+      "epoch": 0.21417133706965572,
+      "grad_norm": 2.31232037905848,
+      "learning_rate": 1.8273501281600118e-05,
+      "loss": 1.66,
+      "step": 1605
+    },
+    {
+      "epoch": 0.21430477715505739,
+      "grad_norm": 1.1674269299374407,
+      "learning_rate": 1.8271072953196915e-05,
+      "loss": 1.7023,
+      "step": 1606
+    },
+    {
+      "epoch": 0.21443821724045903,
+      "grad_norm": 1.1591034820385062,
+      "learning_rate": 1.8268643079851457e-05,
+      "loss": 1.706,
+      "step": 1607
+    },
+    {
+      "epoch": 0.2145716573258607,
+      "grad_norm": 1.2450171637095553,
+      "learning_rate": 1.826621166201761e-05,
+      "loss": 1.6084,
+      "step": 1608
+    },
+    {
+      "epoch": 0.21470509741126234,
+      "grad_norm": 1.1159214690409671,
+      "learning_rate": 1.826377870014954e-05,
+      "loss": 1.6834,
+      "step": 1609
+    },
+    {
+      "epoch": 0.214838537496664,
+      "grad_norm": 1.099652892199378,
+      "learning_rate": 1.8261344194701695e-05,
+      "loss": 1.6745,
+      "step": 1610
+    },
+    {
+      "epoch": 0.21497197758206565,
+      "grad_norm": 1.1872360367368382,
+      "learning_rate": 1.8258908146128814e-05,
+      "loss": 1.6332,
+      "step": 1611
+    },
+    {
+      "epoch": 0.2151054176674673,
+      "grad_norm": 1.0948685127381836,
+      "learning_rate": 1.8256470554885922e-05,
+      "loss": 1.631,
+      "step": 1612
+    },
+    {
+      "epoch": 0.21523885775286897,
+      "grad_norm": 1.1075853907414162,
+      "learning_rate": 1.8254031421428334e-05,
+      "loss": 1.7092,
+      "step": 1613
+    },
+    {
+      "epoch": 0.2153722978382706,
+      "grad_norm": 1.1518564227271244,
+      "learning_rate": 1.8251590746211655e-05,
+      "loss": 1.7042,
+      "step": 1614
+    },
+    {
+      "epoch": 0.21550573792367228,
+      "grad_norm": 1.1038926708394592,
+      "learning_rate": 1.8249148529691768e-05,
+      "loss": 1.6277,
+      "step": 1615
+    },
+    {
+      "epoch": 0.21563917800907392,
+      "grad_norm": 1.1063850648205573,
+      "learning_rate": 1.824670477232486e-05,
+      "loss": 1.6506,
+      "step": 1616
+    },
+    {
+      "epoch": 0.2157726180944756,
+      "grad_norm": 1.0603627332408834,
+      "learning_rate": 1.824425947456739e-05,
+      "loss": 1.6636,
+      "step": 1617
+    },
+    {
+      "epoch": 0.21590605817987724,
+      "grad_norm": 1.2546873356938104,
+      "learning_rate": 1.8241812636876113e-05,
+      "loss": 1.6793,
+      "step": 1618
+    },
+    {
+      "epoch": 0.21603949826527888,
+      "grad_norm": 1.181239728860882,
+      "learning_rate": 1.8239364259708076e-05,
+      "loss": 1.6664,
+      "step": 1619
+    },
+    {
+      "epoch": 0.21617293835068055,
+      "grad_norm": 1.180838315282554,
+      "learning_rate": 1.8236914343520598e-05,
+      "loss": 1.6998,
+      "step": 1620
+    },
+    {
+      "epoch": 0.2163063784360822,
+      "grad_norm": 1.1308148567631164,
+      "learning_rate": 1.8234462888771308e-05,
+      "loss": 1.7154,
+      "step": 1621
+    },
+    {
+      "epoch": 0.21643981852148386,
+      "grad_norm": 1.1326281559924998,
+      "learning_rate": 1.82320098959181e-05,
+      "loss": 1.6256,
+      "step": 1622
+    },
+    {
+      "epoch": 0.2165732586068855,
+      "grad_norm": 1.0865149229961686,
+      "learning_rate": 1.822955536541917e-05,
+      "loss": 1.6228,
+      "step": 1623
+    },
+    {
+      "epoch": 0.21670669869228718,
+      "grad_norm": 1.2423233898451607,
+      "learning_rate": 1.8227099297732997e-05,
+      "loss": 1.648,
+      "step": 1624
+    },
+    {
+      "epoch": 0.21684013877768882,
+      "grad_norm": 1.1502479506643792,
+      "learning_rate": 1.8224641693318338e-05,
+      "loss": 1.6578,
+      "step": 1625
+    },
+    {
+      "epoch": 0.21697357886309046,
+      "grad_norm": 1.440887259770447,
+      "learning_rate": 1.8222182552634257e-05,
+      "loss": 1.6826,
+      "step": 1626
+    },
+    {
+      "epoch": 0.21710701894849213,
+      "grad_norm": 1.121187218751373,
+      "learning_rate": 1.8219721876140084e-05,
+      "loss": 1.6264,
+      "step": 1627
+    },
+    {
+      "epoch": 0.21724045903389377,
+      "grad_norm": 4.7724358351335905,
+      "learning_rate": 1.8217259664295452e-05,
+      "loss": 1.6344,
+      "step": 1628
+    },
+    {
+      "epoch": 0.21737389911929544,
+      "grad_norm": 1.2326061740840186,
+      "learning_rate": 1.821479591756027e-05,
+      "loss": 1.6977,
+      "step": 1629
+    },
+    {
+      "epoch": 0.2175073392046971,
+      "grad_norm": 1.1773999953923238,
+      "learning_rate": 1.8212330636394743e-05,
+      "loss": 1.6613,
+      "step": 1630
+    },
+    {
+      "epoch": 0.21764077929009876,
+      "grad_norm": 1.0745824244836635,
+      "learning_rate": 1.820986382125935e-05,
+      "loss": 1.6746,
+      "step": 1631
+    },
+    {
+      "epoch": 0.2177742193755004,
+      "grad_norm": 1.236415394795963,
+      "learning_rate": 1.820739547261487e-05,
+      "loss": 1.6253,
+      "step": 1632
+    },
+    {
+      "epoch": 0.21790765946090204,
+      "grad_norm": 1.1065194060629115,
+      "learning_rate": 1.8204925590922362e-05,
+      "loss": 1.604,
+      "step": 1633
+    },
+    {
+      "epoch": 0.2180410995463037,
+      "grad_norm": 1.0940043580451333,
+      "learning_rate": 1.820245417664317e-05,
+      "loss": 1.6449,
+      "step": 1634
+    },
+    {
+      "epoch": 0.21817453963170536,
+      "grad_norm": 1.1869832191003882,
+      "learning_rate": 1.8199981230238924e-05,
+      "loss": 1.6705,
+      "step": 1635
+    },
+    {
+      "epoch": 0.21830797971710703,
+      "grad_norm": 1.1189817227152579,
+      "learning_rate": 1.8197506752171545e-05,
+      "loss": 1.6218,
+      "step": 1636
+    },
+    {
+      "epoch": 0.21844141980250867,
+      "grad_norm": 1.1473359068259987,
+      "learning_rate": 1.8195030742903236e-05,
+      "loss": 1.6832,
+      "step": 1637
+    },
+    {
+      "epoch": 0.21857485988791034,
+      "grad_norm": 1.2362438437725367,
+      "learning_rate": 1.819255320289649e-05,
+      "loss": 1.6471,
+      "step": 1638
+    },
+    {
+      "epoch": 0.21870829997331198,
+      "grad_norm": 1.1116111042127557,
+      "learning_rate": 1.8190074132614083e-05,
+      "loss": 1.6397,
+      "step": 1639
+    },
+    {
+      "epoch": 0.21884174005871362,
+      "grad_norm": 1.0931240118741188,
+      "learning_rate": 1.818759353251907e-05,
+      "loss": 1.7028,
+      "step": 1640
+    },
+    {
+      "epoch": 0.2189751801441153,
+      "grad_norm": 1.1134694456669874,
+      "learning_rate": 1.818511140307481e-05,
+      "loss": 1.7135,
+      "step": 1641
+    },
+    {
+      "epoch": 0.21910862022951694,
+      "grad_norm": 1.1225439283708083,
+      "learning_rate": 1.8182627744744928e-05,
+      "loss": 1.6863,
+      "step": 1642
+    },
+    {
+      "epoch": 0.2192420603149186,
+      "grad_norm": 1.1226470273814733,
+      "learning_rate": 1.8180142557993346e-05,
+      "loss": 1.6654,
+      "step": 1643
+    },
+    {
+      "epoch": 0.21937550040032025,
+      "grad_norm": 12.702656527886313,
+      "learning_rate": 1.8177655843284268e-05,
+      "loss": 1.6785,
+      "step": 1644
+    },
+    {
+      "epoch": 0.21950894048572192,
+      "grad_norm": 1.1646728669056514,
+      "learning_rate": 1.8175167601082187e-05,
+      "loss": 1.6678,
+      "step": 1645
+    },
+    {
+      "epoch": 0.21964238057112356,
+      "grad_norm": 1.1370827214131438,
+      "learning_rate": 1.8172677831851874e-05,
+      "loss": 1.6678,
+      "step": 1646
+    },
+    {
+      "epoch": 0.21977582065652523,
+      "grad_norm": 1.0805021664243795,
+      "learning_rate": 1.8170186536058393e-05,
+      "loss": 1.6445,
+      "step": 1647
+    },
+    {
+      "epoch": 0.21990926074192688,
+      "grad_norm": 1.068690271627419,
+      "learning_rate": 1.8167693714167088e-05,
+      "loss": 1.6944,
+      "step": 1648
+    },
+    {
+      "epoch": 0.22004270082732852,
+      "grad_norm": 1.2765183302069458,
+      "learning_rate": 1.816519936664359e-05,
+      "loss": 1.6567,
+      "step": 1649
+    },
+    {
+      "epoch": 0.2201761409127302,
+      "grad_norm": 1.1215664466938502,
+      "learning_rate": 1.8162703493953812e-05,
+      "loss": 1.6553,
+      "step": 1650
+    },
+    {
+      "epoch": 0.22030958099813183,
+      "grad_norm": 1.057949579182957,
+      "learning_rate": 1.8160206096563957e-05,
+      "loss": 1.6541,
+      "step": 1651
+    },
+    {
+      "epoch": 0.2204430210835335,
+      "grad_norm": 1.3213535546275152,
+      "learning_rate": 1.8157707174940516e-05,
+      "loss": 1.6824,
+      "step": 1652
+    },
+    {
+      "epoch": 0.22057646116893515,
+      "grad_norm": 1.1669724021565158,
+      "learning_rate": 1.8155206729550248e-05,
+      "loss": 1.6525,
+      "step": 1653
+    },
+    {
+      "epoch": 0.22070990125433682,
+      "grad_norm": 1.3110532840171654,
+      "learning_rate": 1.8152704760860217e-05,
+      "loss": 1.678,
+      "step": 1654
+    },
+    {
+      "epoch": 0.22084334133973846,
+      "grad_norm": 1.1672424586728913,
+      "learning_rate": 1.815020126933776e-05,
+      "loss": 1.6696,
+      "step": 1655
+    },
+    {
+      "epoch": 0.2209767814251401,
+      "grad_norm": 1.3061093334442921,
+      "learning_rate": 1.8147696255450495e-05,
+      "loss": 1.6654,
+      "step": 1656
+    },
+    {
+      "epoch": 0.22111022151054177,
+      "grad_norm": 1.0702194871522392,
+      "learning_rate": 1.8145189719666335e-05,
+      "loss": 1.7042,
+      "step": 1657
+    },
+    {
+      "epoch": 0.2212436615959434,
+      "grad_norm": 1.2742123644262122,
+      "learning_rate": 1.8142681662453473e-05,
+      "loss": 1.6924,
+      "step": 1658
+    },
+    {
+      "epoch": 0.22137710168134508,
+      "grad_norm": 1.090573827396432,
+      "learning_rate": 1.8140172084280384e-05,
+      "loss": 1.7083,
+      "step": 1659
+    },
+    {
+      "epoch": 0.22151054176674673,
+      "grad_norm": 1.0477634184103604,
+      "learning_rate": 1.813766098561583e-05,
+      "loss": 1.6293,
+      "step": 1660
+    },
+    {
+      "epoch": 0.2216439818521484,
+      "grad_norm": 1.1079746682957217,
+      "learning_rate": 1.8135148366928855e-05,
+      "loss": 1.6785,
+      "step": 1661
+    },
+    {
+      "epoch": 0.22177742193755004,
+      "grad_norm": 1.1056303875592357,
+      "learning_rate": 1.8132634228688785e-05,
+      "loss": 1.5936,
+      "step": 1662
+    },
+    {
+      "epoch": 0.22191086202295168,
+      "grad_norm": 1.0872810931934485,
+      "learning_rate": 1.8130118571365235e-05,
+      "loss": 1.6854,
+      "step": 1663
+    },
+    {
+      "epoch": 0.22204430210835335,
+      "grad_norm": 1.180998755115771,
+      "learning_rate": 1.8127601395428104e-05,
+      "loss": 1.5947,
+      "step": 1664
+    },
+    {
+      "epoch": 0.222177742193755,
+      "grad_norm": 1.076365662325189,
+      "learning_rate": 1.8125082701347564e-05,
+      "loss": 1.6827,
+      "step": 1665
+    },
+    {
+      "epoch": 0.22231118227915667,
+      "grad_norm": 1.286948665055857,
+      "learning_rate": 1.8122562489594084e-05,
+      "loss": 1.684,
+      "step": 1666
+    },
+    {
+      "epoch": 0.2224446223645583,
+      "grad_norm": 1.1501953176983266,
+      "learning_rate": 1.812004076063841e-05,
+      "loss": 1.6775,
+      "step": 1667
+    },
+    {
+      "epoch": 0.22257806244995998,
+      "grad_norm": 1.1268435794579326,
+      "learning_rate": 1.8117517514951573e-05,
+      "loss": 1.7057,
+      "step": 1668
+    },
+    {
+      "epoch": 0.22271150253536162,
+      "grad_norm": 1.1620900257687603,
+      "learning_rate": 1.8114992753004887e-05,
+      "loss": 1.7043,
+      "step": 1669
+    },
+    {
+      "epoch": 0.22284494262076326,
+      "grad_norm": 1.1338951625822513,
+      "learning_rate": 1.8112466475269946e-05,
+      "loss": 1.6707,
+      "step": 1670
+    },
+    {
+      "epoch": 0.22297838270616493,
+      "grad_norm": 1.242098630419242,
+      "learning_rate": 1.8109938682218633e-05,
+      "loss": 1.6795,
+      "step": 1671
+    },
+    {
+      "epoch": 0.22311182279156658,
+      "grad_norm": 1.1679404879363304,
+      "learning_rate": 1.8107409374323107e-05,
+      "loss": 1.6397,
+      "step": 1672
+    },
+    {
+      "epoch": 0.22324526287696825,
+      "grad_norm": 1.176042698983641,
+      "learning_rate": 1.8104878552055817e-05,
+      "loss": 1.6591,
+      "step": 1673
+    },
+    {
+      "epoch": 0.2233787029623699,
+      "grad_norm": 1.312264219479212,
+      "learning_rate": 1.810234621588949e-05,
+      "loss": 1.6572,
+      "step": 1674
+    },
+    {
+      "epoch": 0.22351214304777156,
+      "grad_norm": 1.1369656057368374,
+      "learning_rate": 1.809981236629714e-05,
+      "loss": 1.7209,
+      "step": 1675
+    },
+    {
+      "epoch": 0.2236455831331732,
+      "grad_norm": 1.2223686585190343,
+      "learning_rate": 1.809727700375206e-05,
+      "loss": 1.6587,
+      "step": 1676
+    },
+    {
+      "epoch": 0.22377902321857485,
+      "grad_norm": 1.0737175824141045,
+      "learning_rate": 1.8094740128727823e-05,
+      "loss": 1.6123,
+      "step": 1677
+    },
+    {
+      "epoch": 0.22391246330397652,
+      "grad_norm": 1.0487717384337922,
+      "learning_rate": 1.8092201741698297e-05,
+      "loss": 1.603,
+      "step": 1678
+    },
+    {
+      "epoch": 0.22404590338937816,
+      "grad_norm": 1.0154503615440715,
+      "learning_rate": 1.8089661843137616e-05,
+      "loss": 1.6558,
+      "step": 1679
+    },
+    {
+      "epoch": 0.22417934347477983,
+      "grad_norm": 1.0612689044451018,
+      "learning_rate": 1.8087120433520207e-05,
+      "loss": 1.6634,
+      "step": 1680
+    },
+    {
+      "epoch": 0.22431278356018147,
+      "grad_norm": 1.0783694345863775,
+      "learning_rate": 1.8084577513320777e-05,
+      "loss": 1.6973,
+      "step": 1681
+    },
+    {
+      "epoch": 0.22444622364558314,
+      "grad_norm": 1.4059689465691732,
+      "learning_rate": 1.8082033083014315e-05,
+      "loss": 1.6692,
+      "step": 1682
+    },
+    {
+      "epoch": 0.22457966373098479,
+      "grad_norm": 1.0174187245468838,
+      "learning_rate": 1.807948714307609e-05,
+      "loss": 1.6965,
+      "step": 1683
+    },
+    {
+      "epoch": 0.22471310381638646,
+      "grad_norm": 1.2508321455102909,
+      "learning_rate": 1.8076939693981652e-05,
+      "loss": 1.6924,
+      "step": 1684
+    },
+    {
+      "epoch": 0.2248465439017881,
+      "grad_norm": 1.240543795410627,
+      "learning_rate": 1.807439073620684e-05,
+      "loss": 1.6354,
+      "step": 1685
+    },
+    {
+      "epoch": 0.22497998398718974,
+      "grad_norm": 1.5809329681051474,
+      "learning_rate": 1.807184027022777e-05,
+      "loss": 1.6596,
+      "step": 1686
+    },
+    {
+      "epoch": 0.2251134240725914,
+      "grad_norm": 1.1469707941171723,
+      "learning_rate": 1.806928829652084e-05,
+      "loss": 1.7269,
+      "step": 1687
+    },
+    {
+      "epoch": 0.22524686415799305,
+      "grad_norm": 1.0758984989477276,
+      "learning_rate": 1.8066734815562726e-05,
+      "loss": 1.7109,
+      "step": 1688
+    },
+    {
+      "epoch": 0.22538030424339472,
+      "grad_norm": 1.299428663534772,
+      "learning_rate": 1.8064179827830393e-05,
+      "loss": 1.6851,
+      "step": 1689
+    },
+    {
+      "epoch": 0.22551374432879637,
+      "grad_norm": 1.043487684350109,
+      "learning_rate": 1.8061623333801085e-05,
+      "loss": 1.6506,
+      "step": 1690
+    },
+    {
+      "epoch": 0.22564718441419804,
+      "grad_norm": 1.1237380692317867,
+      "learning_rate": 1.805906533395232e-05,
+      "loss": 1.7187,
+      "step": 1691
+    },
+    {
+      "epoch": 0.22578062449959968,
+      "grad_norm": 1.148118031537372,
+      "learning_rate": 1.8056505828761904e-05,
+      "loss": 1.6431,
+      "step": 1692
+    },
+    {
+      "epoch": 0.22591406458500132,
+      "grad_norm": 1.2374398808076512,
+      "learning_rate": 1.8053944818707932e-05,
+      "loss": 1.6198,
+      "step": 1693
+    },
+    {
+      "epoch": 0.226047504670403,
+      "grad_norm": 1.0903178987983715,
+      "learning_rate": 1.8051382304268762e-05,
+      "loss": 1.6473,
+      "step": 1694
+    },
+    {
+      "epoch": 0.22618094475580464,
+      "grad_norm": 1.1205806647882046,
+      "learning_rate": 1.8048818285923047e-05,
+      "loss": 1.6681,
+      "step": 1695
+    },
+    {
+      "epoch": 0.2263143848412063,
+      "grad_norm": 1.150148641527194,
+      "learning_rate": 1.8046252764149715e-05,
+      "loss": 1.6546,
+      "step": 1696
+    },
+    {
+      "epoch": 0.22644782492660795,
+      "grad_norm": 1.1208283417789737,
+      "learning_rate": 1.8043685739427976e-05,
+      "loss": 1.6365,
+      "step": 1697
+    },
+    {
+      "epoch": 0.22658126501200962,
+      "grad_norm": 1.9233641539386115,
+      "learning_rate": 1.804111721223732e-05,
+      "loss": 1.6183,
+      "step": 1698
+    },
+    {
+      "epoch": 0.22671470509741126,
+      "grad_norm": 1.1665972627988115,
+      "learning_rate": 1.8038547183057524e-05,
+      "loss": 1.6547,
+      "step": 1699
+    },
+    {
+      "epoch": 0.2268481451828129,
+      "grad_norm": 1.0590609907387925,
+      "learning_rate": 1.8035975652368635e-05,
+      "loss": 1.6456,
+      "step": 1700
+    },
+    {
+      "epoch": 0.22698158526821458,
+      "grad_norm": 1.0642643971203603,
+      "learning_rate": 1.8033402620650986e-05,
+      "loss": 1.7051,
+      "step": 1701
+    },
+    {
+      "epoch": 0.22711502535361622,
+      "grad_norm": 1.364226790660327,
+      "learning_rate": 1.803082808838519e-05,
+      "loss": 1.6045,
+      "step": 1702
+    },
+    {
+      "epoch": 0.2272484654390179,
+      "grad_norm": 1.1029813182150021,
+      "learning_rate": 1.8028252056052148e-05,
+      "loss": 1.682,
+      "step": 1703
+    },
+    {
+      "epoch": 0.22738190552441953,
+      "grad_norm": 1.1377601146917016,
+      "learning_rate": 1.8025674524133022e-05,
+      "loss": 1.7134,
+      "step": 1704
+    },
+    {
+      "epoch": 0.2275153456098212,
+      "grad_norm": 1.0987200088336204,
+      "learning_rate": 1.8023095493109273e-05,
+      "loss": 1.6286,
+      "step": 1705
+    },
+    {
+      "epoch": 0.22764878569522284,
+      "grad_norm": 1.0648434423713318,
+      "learning_rate": 1.802051496346263e-05,
+      "loss": 1.6687,
+      "step": 1706
+    },
+    {
+      "epoch": 0.2277822257806245,
+      "grad_norm": 1.1058007563564674,
+      "learning_rate": 1.8017932935675108e-05,
+      "loss": 1.6537,
+      "step": 1707
+    },
+    {
+      "epoch": 0.22791566586602616,
+      "grad_norm": 1.1956539306049252,
+      "learning_rate": 1.8015349410229004e-05,
+      "loss": 1.6273,
+      "step": 1708
+    },
+    {
+      "epoch": 0.2280491059514278,
+      "grad_norm": 1.097833315258532,
+      "learning_rate": 1.8012764387606887e-05,
+      "loss": 1.6283,
+      "step": 1709
+    },
+    {
+      "epoch": 0.22818254603682947,
+      "grad_norm": 1.032052358075353,
+      "learning_rate": 1.801017786829161e-05,
+      "loss": 1.6069,
+      "step": 1710
+    },
+    {
+      "epoch": 0.2283159861222311,
+      "grad_norm": 1.0217813572157182,
+      "learning_rate": 1.800758985276631e-05,
+      "loss": 1.6272,
+      "step": 1711
+    },
+    {
+      "epoch": 0.22844942620763278,
+      "grad_norm": 1.1320615088828117,
+      "learning_rate": 1.8005000341514392e-05,
+      "loss": 1.6433,
+      "step": 1712
+    },
+    {
+      "epoch": 0.22858286629303443,
+      "grad_norm": 1.197563433343801,
+      "learning_rate": 1.8002409335019552e-05,
+      "loss": 1.6391,
+      "step": 1713
+    },
+    {
+      "epoch": 0.22871630637843607,
+      "grad_norm": 1.0876681965716914,
+      "learning_rate": 1.799981683376576e-05,
+      "loss": 1.6864,
+      "step": 1714
+    },
+    {
+      "epoch": 0.22884974646383774,
+      "grad_norm": 1.0354932640447034,
+      "learning_rate": 1.7997222838237264e-05,
+      "loss": 1.6708,
+      "step": 1715
+    },
+    {
+      "epoch": 0.22898318654923938,
+      "grad_norm": 2.734360348457865,
+      "learning_rate": 1.7994627348918593e-05,
+      "loss": 1.6584,
+      "step": 1716
+    },
+    {
+      "epoch": 0.22911662663464105,
+      "grad_norm": 1.1394911063346953,
+      "learning_rate": 1.7992030366294555e-05,
+      "loss": 1.5976,
+      "step": 1717
+    },
+    {
+      "epoch": 0.2292500667200427,
+      "grad_norm": 1.1531653881087183,
+      "learning_rate": 1.798943189085024e-05,
+      "loss": 1.705,
+      "step": 1718
+    },
+    {
+      "epoch": 0.22938350680544436,
+      "grad_norm": 1.21758738022185,
+      "learning_rate": 1.7986831923071007e-05,
+      "loss": 1.6886,
+      "step": 1719
+    },
+    {
+      "epoch": 0.229516946890846,
+      "grad_norm": 1.0529056168835895,
+      "learning_rate": 1.7984230463442505e-05,
+      "loss": 1.6642,
+      "step": 1720
+    },
+    {
+      "epoch": 0.22965038697624768,
+      "grad_norm": 1.2125231023635614,
+      "learning_rate": 1.798162751245066e-05,
+      "loss": 1.6182,
+      "step": 1721
+    },
+    {
+      "epoch": 0.22978382706164932,
+      "grad_norm": 1.1933594423844636,
+      "learning_rate": 1.7979023070581664e-05,
+      "loss": 1.609,
+      "step": 1722
+    },
+    {
+      "epoch": 0.22991726714705096,
+      "grad_norm": 1.0700311747867306,
+      "learning_rate": 1.7976417138322008e-05,
+      "loss": 1.6493,
+      "step": 1723
+    },
+    {
+      "epoch": 0.23005070723245263,
+      "grad_norm": 1.2988519356273156,
+      "learning_rate": 1.7973809716158444e-05,
+      "loss": 1.6935,
+      "step": 1724
+    },
+    {
+      "epoch": 0.23018414731785428,
+      "grad_norm": 1.2630821844608227,
+      "learning_rate": 1.7971200804578008e-05,
+      "loss": 1.6877,
+      "step": 1725
+    },
+    {
+      "epoch": 0.23031758740325595,
+      "grad_norm": 1.0760773885115775,
+      "learning_rate": 1.796859040406802e-05,
+      "loss": 1.6648,
+      "step": 1726
+    },
+    {
+      "epoch": 0.2304510274886576,
+      "grad_norm": 1.0659107902842577,
+      "learning_rate": 1.796597851511607e-05,
+      "loss": 1.6765,
+      "step": 1727
+    },
+    {
+      "epoch": 0.23058446757405926,
+      "grad_norm": 1.0640564215933035,
+      "learning_rate": 1.796336513821003e-05,
+      "loss": 1.6519,
+      "step": 1728
+    },
+    {
+      "epoch": 0.2307179076594609,
+      "grad_norm": 1.044766417697617,
+      "learning_rate": 1.7960750273838046e-05,
+      "loss": 1.7201,
+      "step": 1729
+    },
+    {
+      "epoch": 0.23085134774486255,
+      "grad_norm": 1.3620145766553122,
+      "learning_rate": 1.795813392248855e-05,
+      "loss": 1.6141,
+      "step": 1730
+    },
+    {
+      "epoch": 0.23098478783026422,
+      "grad_norm": 1.0460951962418277,
+      "learning_rate": 1.7955516084650245e-05,
+      "loss": 1.6431,
+      "step": 1731
+    },
+    {
+      "epoch": 0.23111822791566586,
+      "grad_norm": 1.0561334650845267,
+      "learning_rate": 1.7952896760812106e-05,
+      "loss": 1.6234,
+      "step": 1732
+    },
+    {
+      "epoch": 0.23125166800106753,
+      "grad_norm": 1.1267791689507525,
+      "learning_rate": 1.7950275951463404e-05,
+      "loss": 1.6478,
+      "step": 1733
+    },
+    {
+      "epoch": 0.23138510808646917,
+      "grad_norm": 1.1014450865423768,
+      "learning_rate": 1.7947653657093672e-05,
+      "loss": 1.6141,
+      "step": 1734
+    },
+    {
+      "epoch": 0.23151854817187084,
+      "grad_norm": 1.089444038239352,
+      "learning_rate": 1.7945029878192722e-05,
+      "loss": 1.6511,
+      "step": 1735
+    },
+    {
+      "epoch": 0.23165198825727248,
+      "grad_norm": 1.1341511933956758,
+      "learning_rate": 1.7942404615250652e-05,
+      "loss": 1.6546,
+      "step": 1736
+    },
+    {
+      "epoch": 0.23178542834267413,
+      "grad_norm": 1.1162152522317423,
+      "learning_rate": 1.7939777868757825e-05,
+      "loss": 1.6413,
+      "step": 1737
+    },
+    {
+      "epoch": 0.2319188684280758,
+      "grad_norm": 1.0922713243144162,
+      "learning_rate": 1.7937149639204888e-05,
+      "loss": 1.6675,
+      "step": 1738
+    },
+    {
+      "epoch": 0.23205230851347744,
+      "grad_norm": 1.1196061707852998,
+      "learning_rate": 1.7934519927082773e-05,
+      "loss": 1.6891,
+      "step": 1739
+    },
+    {
+      "epoch": 0.2321857485988791,
+      "grad_norm": 1.0373630063131618,
+      "learning_rate": 1.7931888732882665e-05,
+      "loss": 1.6752,
+      "step": 1740
+    },
+    {
+      "epoch": 0.23231918868428075,
+      "grad_norm": 1.1715453355241499,
+      "learning_rate": 1.7929256057096056e-05,
+      "loss": 1.6895,
+      "step": 1741
+    },
+    {
+      "epoch": 0.23245262876968242,
+      "grad_norm": 1.1754651897827133,
+      "learning_rate": 1.792662190021469e-05,
+      "loss": 1.6807,
+      "step": 1742
+    },
+    {
+      "epoch": 0.23258606885508407,
+      "grad_norm": 1.089104477056595,
+      "learning_rate": 1.79239862627306e-05,
+      "loss": 1.6262,
+      "step": 1743
+    },
+    {
+      "epoch": 0.2327195089404857,
+      "grad_norm": 1.047704791420425,
+      "learning_rate": 1.79213491451361e-05,
+      "loss": 1.674,
+      "step": 1744
+    },
+    {
+      "epoch": 0.23285294902588738,
+      "grad_norm": 1.0478486718373898,
+      "learning_rate": 1.7918710547923763e-05,
+      "loss": 1.6386,
+      "step": 1745
+    },
+    {
+      "epoch": 0.23298638911128902,
+      "grad_norm": 1.0374153527662222,
+      "learning_rate": 1.7916070471586455e-05,
+      "loss": 1.6535,
+      "step": 1746
+    },
+    {
+      "epoch": 0.2331198291966907,
+      "grad_norm": 1.1734013425233318,
+      "learning_rate": 1.7913428916617307e-05,
+      "loss": 1.6911,
+      "step": 1747
+    },
+    {
+      "epoch": 0.23325326928209233,
+      "grad_norm": 1.0828478675607198,
+      "learning_rate": 1.791078588350974e-05,
+      "loss": 1.6512,
+      "step": 1748
+    },
+    {
+      "epoch": 0.233386709367494,
+      "grad_norm": 1.114480950469954,
+      "learning_rate": 1.7908141372757436e-05,
+      "loss": 1.6491,
+      "step": 1749
+    },
+    {
+      "epoch": 0.23352014945289565,
+      "grad_norm": 1.1436925883505036,
+      "learning_rate": 1.790549538485436e-05,
+      "loss": 1.6854,
+      "step": 1750
+    },
+    {
+      "epoch": 0.23365358953829732,
+      "grad_norm": 10.687146402666658,
+      "learning_rate": 1.7902847920294754e-05,
+      "loss": 1.621,
+      "step": 1751
+    },
+    {
+      "epoch": 0.23378702962369896,
+      "grad_norm": 1.3707877310751082,
+      "learning_rate": 1.7900198979573135e-05,
+      "loss": 1.6246,
+      "step": 1752
+    },
+    {
+      "epoch": 0.2339204697091006,
+      "grad_norm": 1.1877922731508994,
+      "learning_rate": 1.7897548563184296e-05,
+      "loss": 1.6376,
+      "step": 1753
+    },
+    {
+      "epoch": 0.23405390979450227,
+      "grad_norm": 1.36255029255512,
+      "learning_rate": 1.7894896671623296e-05,
+      "loss": 1.6354,
+      "step": 1754
+    },
+    {
+      "epoch": 0.23418734987990392,
+      "grad_norm": 1.4838375564796251,
+      "learning_rate": 1.7892243305385487e-05,
+      "loss": 1.6155,
+      "step": 1755
+    },
+    {
+      "epoch": 0.2343207899653056,
+      "grad_norm": 1.1622308653549671,
+      "learning_rate": 1.7889588464966488e-05,
+      "loss": 1.6665,
+      "step": 1756
+    },
+    {
+      "epoch": 0.23445423005070723,
+      "grad_norm": 1.336261221880211,
+      "learning_rate": 1.788693215086219e-05,
+      "loss": 1.6409,
+      "step": 1757
+    },
+    {
+      "epoch": 0.2345876701361089,
+      "grad_norm": 1.2428755384173737,
+      "learning_rate": 1.788427436356876e-05,
+      "loss": 1.6523,
+      "step": 1758
+    },
+    {
+      "epoch": 0.23472111022151054,
+      "grad_norm": 1.1734745642911637,
+      "learning_rate": 1.7881615103582642e-05,
+      "loss": 1.6823,
+      "step": 1759
+    },
+    {
+      "epoch": 0.23485455030691219,
+      "grad_norm": 1.359796714523189,
+      "learning_rate": 1.7878954371400563e-05,
+      "loss": 1.6766,
+      "step": 1760
+    },
+    {
+      "epoch": 0.23498799039231386,
+      "grad_norm": 1.1543938773322537,
+      "learning_rate": 1.787629216751951e-05,
+      "loss": 1.6808,
+      "step": 1761
+    },
+    {
+      "epoch": 0.2351214304777155,
+      "grad_norm": 1.2231767398910764,
+      "learning_rate": 1.7873628492436757e-05,
+      "loss": 1.6149,
+      "step": 1762
+    },
+    {
+      "epoch": 0.23525487056311717,
+      "grad_norm": 1.3991330507525954,
+      "learning_rate": 1.7870963346649847e-05,
+      "loss": 1.6663,
+      "step": 1763
+    },
+    {
+      "epoch": 0.2353883106485188,
+      "grad_norm": 1.2111519435232232,
+      "learning_rate": 1.7868296730656596e-05,
+      "loss": 1.6832,
+      "step": 1764
+    },
+    {
+      "epoch": 0.23552175073392048,
+      "grad_norm": 1.1858904981204466,
+      "learning_rate": 1.7865628644955098e-05,
+      "loss": 1.6314,
+      "step": 1765
+    },
+    {
+      "epoch": 0.23565519081932212,
+      "grad_norm": 1.2029794734253174,
+      "learning_rate": 1.786295909004373e-05,
+      "loss": 1.5582,
+      "step": 1766
+    },
+    {
+      "epoch": 0.23578863090472377,
+      "grad_norm": 1.319501116381306,
+      "learning_rate": 1.7860288066421118e-05,
+      "loss": 1.6763,
+      "step": 1767
+    },
+    {
+      "epoch": 0.23592207099012544,
+      "grad_norm": 1.4049089786241191,
+      "learning_rate": 1.7857615574586193e-05,
+      "loss": 1.6361,
+      "step": 1768
+    },
+    {
+      "epoch": 0.23605551107552708,
+      "grad_norm": 1.1918917655522046,
+      "learning_rate": 1.785494161503814e-05,
+      "loss": 1.6137,
+      "step": 1769
+    },
+    {
+      "epoch": 0.23618895116092875,
+      "grad_norm": 1.144754220344139,
+      "learning_rate": 1.7852266188276422e-05,
+      "loss": 1.6175,
+      "step": 1770
+    },
+    {
+      "epoch": 0.2363223912463304,
+      "grad_norm": 1.3381800606475023,
+      "learning_rate": 1.7849589294800787e-05,
+      "loss": 1.6195,
+      "step": 1771
+    },
+    {
+      "epoch": 0.23645583133173206,
+      "grad_norm": 1.0458078567231872,
+      "learning_rate": 1.7846910935111242e-05,
+      "loss": 1.6848,
+      "step": 1772
+    },
+    {
+      "epoch": 0.2365892714171337,
+      "grad_norm": 1.4105954493444146,
+      "learning_rate": 1.7844231109708072e-05,
+      "loss": 1.654,
+      "step": 1773
+    },
+    {
+      "epoch": 0.23672271150253535,
+      "grad_norm": 1.289151142605529,
+      "learning_rate": 1.784154981909184e-05,
+      "loss": 1.6533,
+      "step": 1774
+    },
+    {
+      "epoch": 0.23685615158793702,
+      "grad_norm": 1.2707742721373032,
+      "learning_rate": 1.7838867063763383e-05,
+      "loss": 1.581,
+      "step": 1775
+    },
+    {
+      "epoch": 0.23698959167333866,
+      "grad_norm": 1.2053249887223572,
+      "learning_rate": 1.7836182844223808e-05,
+      "loss": 1.6464,
+      "step": 1776
+    },
+    {
+      "epoch": 0.23712303175874033,
+      "grad_norm": 1.06210419581067,
+      "learning_rate": 1.7833497160974495e-05,
+      "loss": 1.5838,
+      "step": 1777
+    },
+    {
+      "epoch": 0.23725647184414198,
+      "grad_norm": 1.3047023677050178,
+      "learning_rate": 1.78308100145171e-05,
+      "loss": 1.5983,
+      "step": 1778
+    },
+    {
+      "epoch": 0.23738991192954365,
+      "grad_norm": 1.29218147582307,
+      "learning_rate": 1.7828121405353553e-05,
+      "loss": 1.6862,
+      "step": 1779
+    },
+    {
+      "epoch": 0.2375233520149453,
+      "grad_norm": 1.0466696307900745,
+      "learning_rate": 1.7825431333986052e-05,
+      "loss": 1.6623,
+      "step": 1780
+    },
+    {
+      "epoch": 0.23765679210034693,
+      "grad_norm": 1.2349477939262232,
+      "learning_rate": 1.7822739800917073e-05,
+      "loss": 1.5709,
+      "step": 1781
+    },
+    {
+      "epoch": 0.2377902321857486,
+      "grad_norm": 1.2147084949319535,
+      "learning_rate": 1.782004680664937e-05,
+      "loss": 1.6726,
+      "step": 1782
+    },
+    {
+      "epoch": 0.23792367227115024,
+      "grad_norm": 1.0232408573291623,
+      "learning_rate": 1.7817352351685954e-05,
+      "loss": 1.6278,
+      "step": 1783
+    },
+    {
+      "epoch": 0.23805711235655191,
+      "grad_norm": 1.13584849589059,
+      "learning_rate": 1.781465643653012e-05,
+      "loss": 1.7017,
+      "step": 1784
+    },
+    {
+      "epoch": 0.23819055244195356,
+      "grad_norm": 1.0754152565333037,
+      "learning_rate": 1.7811959061685438e-05,
+      "loss": 1.6089,
+      "step": 1785
+    },
+    {
+      "epoch": 0.23832399252735523,
+      "grad_norm": 1.1362601818532363,
+      "learning_rate": 1.7809260227655747e-05,
+      "loss": 1.6161,
+      "step": 1786
+    },
+    {
+      "epoch": 0.23845743261275687,
+      "grad_norm": 1.0336376244273875,
+      "learning_rate": 1.7806559934945156e-05,
+      "loss": 1.6126,
+      "step": 1787
+    },
+    {
+      "epoch": 0.23859087269815854,
+      "grad_norm": 1.2284171891499982,
+      "learning_rate": 1.7803858184058053e-05,
+      "loss": 1.609,
+      "step": 1788
+    },
+    {
+      "epoch": 0.23872431278356018,
+      "grad_norm": 1.0993616922411698,
+      "learning_rate": 1.780115497549909e-05,
+      "loss": 1.6893,
+      "step": 1789
+    },
+    {
+      "epoch": 0.23885775286896183,
+      "grad_norm": 1.0345469608413957,
+      "learning_rate": 1.7798450309773195e-05,
+      "loss": 1.6798,
+      "step": 1790
+    },
+    {
+      "epoch": 0.2389911929543635,
+      "grad_norm": 1.2117978955967876,
+      "learning_rate": 1.7795744187385575e-05,
+      "loss": 1.6246,
+      "step": 1791
+    },
+    {
+      "epoch": 0.23912463303976514,
+      "grad_norm": 1.0463913434816907,
+      "learning_rate": 1.7793036608841694e-05,
+      "loss": 1.6595,
+      "step": 1792
+    },
+    {
+      "epoch": 0.2392580731251668,
+      "grad_norm": 1.1218163300477884,
+      "learning_rate": 1.7790327574647306e-05,
+      "loss": 1.6288,
+      "step": 1793
+    },
+    {
+      "epoch": 0.23939151321056845,
+      "grad_norm": 1.3920149154127004,
+      "learning_rate": 1.778761708530842e-05,
+      "loss": 1.638,
+      "step": 1794
+    },
+    {
+      "epoch": 0.23952495329597012,
+      "grad_norm": 1.0453245062677134,
+      "learning_rate": 1.7784905141331334e-05,
+      "loss": 1.7159,
+      "step": 1795
+    },
+    {
+      "epoch": 0.23965839338137176,
+      "grad_norm": 1.0905984393131543,
+      "learning_rate": 1.7782191743222594e-05,
+      "loss": 1.6959,
+      "step": 1796
+    },
+    {
+      "epoch": 0.2397918334667734,
+      "grad_norm": 1.0576313468931025,
+      "learning_rate": 1.7779476891489044e-05,
+      "loss": 1.6626,
+      "step": 1797
+    },
+    {
+      "epoch": 0.23992527355217508,
+      "grad_norm": 1.1004207621683646,
+      "learning_rate": 1.777676058663778e-05,
+      "loss": 1.6204,
+      "step": 1798
+    },
+    {
+      "epoch": 0.24005871363757672,
+      "grad_norm": 1.1873198699247158,
+      "learning_rate": 1.7774042829176186e-05,
+      "loss": 1.7395,
+      "step": 1799
+    },
+    {
+      "epoch": 0.2401921537229784,
+      "grad_norm": 1.0687609878005364,
+      "learning_rate": 1.77713236196119e-05,
+      "loss": 1.6988,
+      "step": 1800
+    },
+    {
+      "epoch": 0.24032559380838003,
+      "grad_norm": 1.0679027457183934,
+      "learning_rate": 1.7768602958452837e-05,
+      "loss": 1.6671,
+      "step": 1801
+    },
+    {
+      "epoch": 0.2404590338937817,
+      "grad_norm": 1.0417473997306737,
+      "learning_rate": 1.7765880846207193e-05,
+      "loss": 1.631,
+      "step": 1802
+    },
+    {
+      "epoch": 0.24059247397918335,
+      "grad_norm": 1.119803585337448,
+      "learning_rate": 1.7763157283383426e-05,
+      "loss": 1.6908,
+      "step": 1803
+    },
+    {
+      "epoch": 0.240725914064585,
+      "grad_norm": 1.0143233262923002,
+      "learning_rate": 1.7760432270490266e-05,
+      "loss": 1.6342,
+      "step": 1804
+    },
+    {
+      "epoch": 0.24085935414998666,
+      "grad_norm": 1.2812766744363617,
+      "learning_rate": 1.7757705808036714e-05,
+      "loss": 1.6356,
+      "step": 1805
+    },
+    {
+      "epoch": 0.2409927942353883,
+      "grad_norm": 1.0926304516768464,
+      "learning_rate": 1.775497789653204e-05,
+      "loss": 1.6898,
+      "step": 1806
+    },
+    {
+      "epoch": 0.24112623432078997,
+      "grad_norm": 1.013517851069127,
+      "learning_rate": 1.7752248536485787e-05,
+      "loss": 1.6464,
+      "step": 1807
+    },
+    {
+      "epoch": 0.24125967440619162,
+      "grad_norm": 1.0833079031525588,
+      "learning_rate": 1.7749517728407775e-05,
+      "loss": 1.6374,
+      "step": 1808
+    },
+    {
+      "epoch": 0.24139311449159329,
+      "grad_norm": 1.031007153005283,
+      "learning_rate": 1.774678547280808e-05,
+      "loss": 1.6838,
+      "step": 1809
+    },
+    {
+      "epoch": 0.24152655457699493,
+      "grad_norm": 1.1069793411968527,
+      "learning_rate": 1.7744051770197058e-05,
+      "loss": 1.6641,
+      "step": 1810
+    },
+    {
+      "epoch": 0.24165999466239657,
+      "grad_norm": 1.143899658709366,
+      "learning_rate": 1.7741316621085336e-05,
+      "loss": 1.7275,
+      "step": 1811
+    },
+    {
+      "epoch": 0.24179343474779824,
+      "grad_norm": 1.0641576337141183,
+      "learning_rate": 1.7738580025983808e-05,
+      "loss": 1.6649,
+      "step": 1812
+    },
+    {
+      "epoch": 0.24192687483319988,
+      "grad_norm": 1.2238048171560199,
+      "learning_rate": 1.773584198540364e-05,
+      "loss": 1.6442,
+      "step": 1813
+    },
+    {
+      "epoch": 0.24206031491860155,
+      "grad_norm": 1.2197399885035782,
+      "learning_rate": 1.773310249985626e-05,
+      "loss": 1.648,
+      "step": 1814
+    },
+    {
+      "epoch": 0.2421937550040032,
+      "grad_norm": 1.0167788681079188,
+      "learning_rate": 1.773036156985338e-05,
+      "loss": 1.6676,
+      "step": 1815
+    },
+    {
+      "epoch": 0.24232719508940487,
+      "grad_norm": 1.2114816489016487,
+      "learning_rate": 1.7727619195906972e-05,
+      "loss": 1.6639,
+      "step": 1816
+    },
+    {
+      "epoch": 0.2424606351748065,
+      "grad_norm": 8.428467351350283,
+      "learning_rate": 1.7724875378529282e-05,
+      "loss": 1.6576,
+      "step": 1817
+    },
+    {
+      "epoch": 0.24259407526020815,
+      "grad_norm": 1.0465542814144335,
+      "learning_rate": 1.7722130118232823e-05,
+      "loss": 1.6399,
+      "step": 1818
+    },
+    {
+      "epoch": 0.24272751534560982,
+      "grad_norm": 1.1988097703131617,
+      "learning_rate": 1.7719383415530375e-05,
+      "loss": 1.626,
+      "step": 1819
+    },
+    {
+      "epoch": 0.24286095543101147,
+      "grad_norm": 1.1423995626336525,
+      "learning_rate": 1.7716635270934996e-05,
+      "loss": 1.6633,
+      "step": 1820
+    },
+    {
+      "epoch": 0.24299439551641314,
+      "grad_norm": 1.0587665476874228,
+      "learning_rate": 1.7713885684960002e-05,
+      "loss": 1.6313,
+      "step": 1821
+    },
+    {
+      "epoch": 0.24312783560181478,
+      "grad_norm": 1.2825083624500806,
+      "learning_rate": 1.771113465811899e-05,
+      "loss": 1.6683,
+      "step": 1822
+    },
+    {
+      "epoch": 0.24326127568721645,
+      "grad_norm": 1.167576975923983,
+      "learning_rate": 1.770838219092582e-05,
+      "loss": 1.6607,
+      "step": 1823
+    },
+    {
+      "epoch": 0.2433947157726181,
+      "grad_norm": 1.0672931246277235,
+      "learning_rate": 1.7705628283894617e-05,
+      "loss": 1.6341,
+      "step": 1824
+    },
+    {
+      "epoch": 0.24352815585801976,
+      "grad_norm": 1.134208618821163,
+      "learning_rate": 1.7702872937539783e-05,
+      "loss": 1.7071,
+      "step": 1825
+    },
+    {
+      "epoch": 0.2436615959434214,
+      "grad_norm": 1.2440133024491684,
+      "learning_rate": 1.7700116152375986e-05,
+      "loss": 1.7007,
+      "step": 1826
+    },
+    {
+      "epoch": 0.24379503602882305,
+      "grad_norm": 1.0256084895859983,
+      "learning_rate": 1.769735792891816e-05,
+      "loss": 1.6557,
+      "step": 1827
+    },
+    {
+      "epoch": 0.24392847611422472,
+      "grad_norm": 1.063245387715446,
+      "learning_rate": 1.769459826768151e-05,
+      "loss": 1.6766,
+      "step": 1828
+    },
+    {
+      "epoch": 0.24406191619962636,
+      "grad_norm": 1.079027783398279,
+      "learning_rate": 1.7691837169181513e-05,
+      "loss": 1.6567,
+      "step": 1829
+    },
+    {
+      "epoch": 0.24419535628502803,
+      "grad_norm": 1.332484756061736,
+      "learning_rate": 1.7689074633933903e-05,
+      "loss": 1.6403,
+      "step": 1830
+    },
+    {
+      "epoch": 0.24432879637042967,
+      "grad_norm": 1.114632167604677,
+      "learning_rate": 1.7686310662454698e-05,
+      "loss": 1.654,
+      "step": 1831
+    },
+    {
+      "epoch": 0.24446223645583134,
+      "grad_norm": 1.0194057684489455,
+      "learning_rate": 1.7683545255260173e-05,
+      "loss": 1.6437,
+      "step": 1832
+    },
+    {
+      "epoch": 0.244595676541233,
+      "grad_norm": 1.1549546103317554,
+      "learning_rate": 1.7680778412866876e-05,
+      "loss": 1.581,
+      "step": 1833
+    },
+    {
+      "epoch": 0.24472911662663463,
+      "grad_norm": 1.2807877506947511,
+      "learning_rate": 1.7678010135791616e-05,
+      "loss": 1.6518,
+      "step": 1834
+    },
+    {
+      "epoch": 0.2448625567120363,
+      "grad_norm": 1.246730596301354,
+      "learning_rate": 1.7675240424551483e-05,
+      "loss": 1.6254,
+      "step": 1835
+    },
+    {
+      "epoch": 0.24499599679743794,
+      "grad_norm": 1.0865214357163941,
+      "learning_rate": 1.7672469279663827e-05,
+      "loss": 1.6957,
+      "step": 1836
+    },
+    {
+      "epoch": 0.2451294368828396,
+      "grad_norm": 1.247069373038027,
+      "learning_rate": 1.766969670164626e-05,
+      "loss": 1.5999,
+      "step": 1837
+    },
+    {
+      "epoch": 0.24526287696824126,
+      "grad_norm": 1.0837926981903307,
+      "learning_rate": 1.7666922691016673e-05,
+      "loss": 1.6915,
+      "step": 1838
+    },
+    {
+      "epoch": 0.24539631705364293,
+      "grad_norm": 1.2095352189589086,
+      "learning_rate": 1.766414724829322e-05,
+      "loss": 1.6801,
+      "step": 1839
+    },
+    {
+      "epoch": 0.24552975713904457,
+      "grad_norm": 1.0527519692262675,
+      "learning_rate": 1.7661370373994318e-05,
+      "loss": 1.7169,
+      "step": 1840
+    },
+    {
+      "epoch": 0.2456631972244462,
+      "grad_norm": 1.06283226700497,
+      "learning_rate": 1.765859206863866e-05,
+      "loss": 1.653,
+      "step": 1841
+    },
+    {
+      "epoch": 0.24579663730984788,
+      "grad_norm": 1.1067902633010984,
+      "learning_rate": 1.7655812332745198e-05,
+      "loss": 1.6609,
+      "step": 1842
+    },
+    {
+      "epoch": 0.24593007739524952,
+      "grad_norm": 1.11251472312465,
+      "learning_rate": 1.7653031166833158e-05,
+      "loss": 1.6144,
+      "step": 1843
+    },
+    {
+      "epoch": 0.2460635174806512,
+      "grad_norm": 1.074234471124788,
+      "learning_rate": 1.765024857142203e-05,
+      "loss": 1.6302,
+      "step": 1844
+    },
+    {
+      "epoch": 0.24619695756605284,
+      "grad_norm": 1.2669376467474798,
+      "learning_rate": 1.764746454703157e-05,
+      "loss": 1.7182,
+      "step": 1845
+    },
+    {
+      "epoch": 0.2463303976514545,
+      "grad_norm": 1.0774090725880676,
+      "learning_rate": 1.7644679094181803e-05,
+      "loss": 1.6356,
+      "step": 1846
+    },
+    {
+      "epoch": 0.24646383773685615,
+      "grad_norm": 1.2670775121665758,
+      "learning_rate": 1.7641892213393017e-05,
+      "loss": 1.6567,
+      "step": 1847
+    },
+    {
+      "epoch": 0.2465972778222578,
+      "grad_norm": 1.2199961678910605,
+      "learning_rate": 1.7639103905185774e-05,
+      "loss": 1.6689,
+      "step": 1848
+    },
+    {
+      "epoch": 0.24673071790765946,
+      "grad_norm": 1.093950746093173,
+      "learning_rate": 1.7636314170080893e-05,
+      "loss": 1.6764,
+      "step": 1849
+    },
+    {
+      "epoch": 0.2468641579930611,
+      "grad_norm": 1.3555020553608637,
+      "learning_rate": 1.7633523008599468e-05,
+      "loss": 1.6779,
+      "step": 1850
+    },
+    {
+      "epoch": 0.24699759807846278,
+      "grad_norm": 1.2770301389385996,
+      "learning_rate": 1.7630730421262857e-05,
+      "loss": 1.6845,
+      "step": 1851
+    },
+    {
+      "epoch": 0.24713103816386442,
+      "grad_norm": 1.04055907190113,
+      "learning_rate": 1.7627936408592684e-05,
+      "loss": 1.71,
+      "step": 1852
+    },
+    {
+      "epoch": 0.2472644782492661,
+      "grad_norm": 1.2660383381787914,
+      "learning_rate": 1.7625140971110834e-05,
+      "loss": 1.6673,
+      "step": 1853
+    },
+    {
+      "epoch": 0.24739791833466773,
+      "grad_norm": 1.1241950492161565,
+      "learning_rate": 1.7622344109339468e-05,
+      "loss": 1.6426,
+      "step": 1854
+    },
+    {
+      "epoch": 0.24753135842006938,
+      "grad_norm": 1.1634703156362944,
+      "learning_rate": 1.7619545823801008e-05,
+      "loss": 1.6395,
+      "step": 1855
+    },
+    {
+      "epoch": 0.24766479850547105,
+      "grad_norm": 1.1887592831201785,
+      "learning_rate": 1.7616746115018136e-05,
+      "loss": 1.6759,
+      "step": 1856
+    },
+    {
+      "epoch": 0.2477982385908727,
+      "grad_norm": 1.04558208144064,
+      "learning_rate": 1.7613944983513812e-05,
+      "loss": 1.6663,
+      "step": 1857
+    },
+    {
+      "epoch": 0.24793167867627436,
+      "grad_norm": 1.2246725069546616,
+      "learning_rate": 1.7611142429811253e-05,
+      "loss": 1.6571,
+      "step": 1858
+    },
+    {
+      "epoch": 0.248065118761676,
+      "grad_norm": 1.064172850739474,
+      "learning_rate": 1.7608338454433945e-05,
+      "loss": 1.6479,
+      "step": 1859
+    },
+    {
+      "epoch": 0.24819855884707767,
+      "grad_norm": 1.1389852926930035,
+      "learning_rate": 1.7605533057905635e-05,
+      "loss": 1.6456,
+      "step": 1860
+    },
+    {
+      "epoch": 0.24833199893247931,
+      "grad_norm": 1.1395310795650293,
+      "learning_rate": 1.7602726240750346e-05,
+      "loss": 1.6769,
+      "step": 1861
+    },
+    {
+      "epoch": 0.24846543901788098,
+      "grad_norm": 1.1049036582128695,
+      "learning_rate": 1.7599918003492354e-05,
+      "loss": 1.6121,
+      "step": 1862
+    },
+    {
+      "epoch": 0.24859887910328263,
+      "grad_norm": 1.2866424724351742,
+      "learning_rate": 1.7597108346656206e-05,
+      "loss": 1.6422,
+      "step": 1863
+    },
+    {
+      "epoch": 0.24873231918868427,
+      "grad_norm": 1.0354908069373776,
+      "learning_rate": 1.7594297270766713e-05,
+      "loss": 1.6581,
+      "step": 1864
+    },
+    {
+      "epoch": 0.24886575927408594,
+      "grad_norm": 1.0312439084550453,
+      "learning_rate": 1.7591484776348958e-05,
+      "loss": 1.6143,
+      "step": 1865
+    },
+    {
+      "epoch": 0.24899919935948758,
+      "grad_norm": 1.0200041648342764,
+      "learning_rate": 1.7588670863928278e-05,
+      "loss": 1.6014,
+      "step": 1866
+    },
+    {
+      "epoch": 0.24913263944488925,
+      "grad_norm": 1.0618120625694842,
+      "learning_rate": 1.758585553403028e-05,
+      "loss": 1.6641,
+      "step": 1867
+    },
+    {
+      "epoch": 0.2492660795302909,
+      "grad_norm": 1.0713518559188207,
+      "learning_rate": 1.758303878718084e-05,
+      "loss": 1.6472,
+      "step": 1868
+    },
+    {
+      "epoch": 0.24939951961569257,
+      "grad_norm": 1.0046786864644603,
+      "learning_rate": 1.7580220623906088e-05,
+      "loss": 1.6321,
+      "step": 1869
+    },
+    {
+      "epoch": 0.2495329597010942,
+      "grad_norm": 1.1134637775554515,
+      "learning_rate": 1.7577401044732428e-05,
+      "loss": 1.6474,
+      "step": 1870
+    },
+    {
+      "epoch": 0.24966639978649585,
+      "grad_norm": 1.27550190258781,
+      "learning_rate": 1.7574580050186524e-05,
+      "loss": 1.6594,
+      "step": 1871
+    },
+    {
+      "epoch": 0.24979983987189752,
+      "grad_norm": 1.0247658171892224,
+      "learning_rate": 1.7571757640795308e-05,
+      "loss": 1.6159,
+      "step": 1872
+    },
+    {
+      "epoch": 0.24993327995729916,
+      "grad_norm": 1.0462382916284119,
+      "learning_rate": 1.7568933817085966e-05,
+      "loss": 1.5982,
+      "step": 1873
+    },
+    {
+      "epoch": 0.2500667200427008,
+      "grad_norm": 1.0804215813368583,
+      "learning_rate": 1.756610857958597e-05,
+      "loss": 1.7014,
+      "step": 1874
+    },
+    {
+      "epoch": 0.2502001601281025,
+      "grad_norm": 1.0738193355707433,
+      "learning_rate": 1.756328192882303e-05,
+      "loss": 1.6642,
+      "step": 1875
+    },
+    {
+      "epoch": 0.25033360021350415,
+      "grad_norm": 1.3455199376826739,
+      "learning_rate": 1.7560453865325143e-05,
+      "loss": 1.6508,
+      "step": 1876
+    },
+    {
+      "epoch": 0.2504670402989058,
+      "grad_norm": 1.084814244415092,
+      "learning_rate": 1.7557624389620548e-05,
+      "loss": 1.6794,
+      "step": 1877
+    },
+    {
+      "epoch": 0.25060048038430743,
+      "grad_norm": 1.061930007641208,
+      "learning_rate": 1.7554793502237765e-05,
+      "loss": 1.6879,
+      "step": 1878
+    },
+    {
+      "epoch": 0.2507339204697091,
+      "grad_norm": 1.0450629751248275,
+      "learning_rate": 1.7551961203705573e-05,
+      "loss": 1.6783,
+      "step": 1879
+    },
+    {
+      "epoch": 0.2508673605551108,
+      "grad_norm": 1.0478570277539474,
+      "learning_rate": 1.7549127494553005e-05,
+      "loss": 1.6616,
+      "step": 1880
+    },
+    {
+      "epoch": 0.2510008006405124,
+      "grad_norm": 1.055824286400764,
+      "learning_rate": 1.7546292375309375e-05,
+      "loss": 1.5953,
+      "step": 1881
+    },
+    {
+      "epoch": 0.25113424072591406,
+      "grad_norm": 1.2242477360612551,
+      "learning_rate": 1.7543455846504245e-05,
+      "loss": 1.6843,
+      "step": 1882
+    },
+    {
+      "epoch": 0.25126768081131573,
+      "grad_norm": 1.0389910946552412,
+      "learning_rate": 1.754061790866745e-05,
+      "loss": 1.6307,
+      "step": 1883
+    },
+    {
+      "epoch": 0.2514011208967174,
+      "grad_norm": 1.054285947670663,
+      "learning_rate": 1.7537778562329083e-05,
+      "loss": 1.6369,
+      "step": 1884
+    },
+    {
+      "epoch": 0.251534560982119,
+      "grad_norm": 1.2248259207278507,
+      "learning_rate": 1.7534937808019502e-05,
+      "loss": 1.6693,
+      "step": 1885
+    },
+    {
+      "epoch": 0.2516680010675207,
+      "grad_norm": 19.66684114165284,
+      "learning_rate": 1.7532095646269324e-05,
+      "loss": 1.7343,
+      "step": 1886
+    },
+    {
+      "epoch": 0.25180144115292236,
+      "grad_norm": 1.1148849102770775,
+      "learning_rate": 1.7529252077609435e-05,
+      "loss": 1.6762,
+      "step": 1887
+    },
+    {
+      "epoch": 0.25193488123832397,
+      "grad_norm": 1.3680236555548535,
+      "learning_rate": 1.7526407102570985e-05,
+      "loss": 1.6241,
+      "step": 1888
+    },
+    {
+      "epoch": 0.25206832132372564,
+      "grad_norm": 1.0884368102610955,
+      "learning_rate": 1.7523560721685376e-05,
+      "loss": 1.6284,
+      "step": 1889
+    },
+    {
+      "epoch": 0.2522017614091273,
+      "grad_norm": 1.0802430716006264,
+      "learning_rate": 1.7520712935484288e-05,
+      "loss": 1.6485,
+      "step": 1890
+    },
+    {
+      "epoch": 0.252335201494529,
+      "grad_norm": 1.0586064611090562,
+      "learning_rate": 1.7517863744499645e-05,
+      "loss": 1.5886,
+      "step": 1891
+    },
+    {
+      "epoch": 0.2524686415799306,
+      "grad_norm": 1.0549630121480713,
+      "learning_rate": 1.7515013149263654e-05,
+      "loss": 1.6656,
+      "step": 1892
+    },
+    {
+      "epoch": 0.25260208166533227,
+      "grad_norm": 1.021796454571563,
+      "learning_rate": 1.7512161150308763e-05,
+      "loss": 1.6442,
+      "step": 1893
+    },
+    {
+      "epoch": 0.25273552175073394,
+      "grad_norm": 1.0820683485752427,
+      "learning_rate": 1.7509307748167703e-05,
+      "loss": 1.7135,
+      "step": 1894
+    },
+    {
+      "epoch": 0.25286896183613555,
+      "grad_norm": 1.1110777101924227,
+      "learning_rate": 1.750645294337345e-05,
+      "loss": 1.6766,
+      "step": 1895
+    },
+    {
+      "epoch": 0.2530024019215372,
+      "grad_norm": 1.380759537151671,
+      "learning_rate": 1.750359673645925e-05,
+      "loss": 1.6784,
+      "step": 1896
+    },
+    {
+      "epoch": 0.2531358420069389,
+      "grad_norm": 1.2299813651145617,
+      "learning_rate": 1.750073912795861e-05,
+      "loss": 1.6679,
+      "step": 1897
+    },
+    {
+      "epoch": 0.25326928209234056,
+      "grad_norm": 13.108704611195062,
+      "learning_rate": 1.7497880118405305e-05,
+      "loss": 1.6712,
+      "step": 1898
+    },
+    {
+      "epoch": 0.2534027221777422,
+      "grad_norm": 8.088838037228543,
+      "learning_rate": 1.7495019708333362e-05,
+      "loss": 1.8226,
+      "step": 1899
+    },
+    {
+      "epoch": 0.25353616226314385,
+      "grad_norm": 1.284583284047351,
+      "learning_rate": 1.7492157898277066e-05,
+      "loss": 1.6856,
+      "step": 1900
+    },
+    {
+      "epoch": 0.2536696023485455,
+      "grad_norm": 1.1185025918470872,
+      "learning_rate": 1.7489294688770976e-05,
+      "loss": 1.6563,
+      "step": 1901
+    },
+    {
+      "epoch": 0.25380304243394713,
+      "grad_norm": 1.310393482796925,
+      "learning_rate": 1.748643008034991e-05,
+      "loss": 1.6655,
+      "step": 1902
+    },
+    {
+      "epoch": 0.2539364825193488,
+      "grad_norm": 1.2823242056261748,
+      "learning_rate": 1.748356407354894e-05,
+      "loss": 1.6294,
+      "step": 1903
+    },
+    {
+      "epoch": 0.2540699226047505,
+      "grad_norm": 1.115456828324836,
+      "learning_rate": 1.7480696668903405e-05,
+      "loss": 1.6311,
+      "step": 1904
+    },
+    {
+      "epoch": 0.25420336269015215,
+      "grad_norm": 1.098358515379306,
+      "learning_rate": 1.7477827866948902e-05,
+      "loss": 1.6474,
+      "step": 1905
+    },
+    {
+      "epoch": 0.25433680277555376,
+      "grad_norm": 1.2499952987679348,
+      "learning_rate": 1.7474957668221294e-05,
+      "loss": 1.6127,
+      "step": 1906
+    },
+    {
+      "epoch": 0.25447024286095543,
+      "grad_norm": 1.0653256471838468,
+      "learning_rate": 1.7472086073256695e-05,
+      "loss": 1.6237,
+      "step": 1907
+    },
+    {
+      "epoch": 0.2546036829463571,
+      "grad_norm": 1.100375109746571,
+      "learning_rate": 1.7469213082591493e-05,
+      "loss": 1.6623,
+      "step": 1908
+    },
+    {
+      "epoch": 0.2547371230317587,
+      "grad_norm": 1.0728865640963992,
+      "learning_rate": 1.746633869676233e-05,
+      "loss": 1.6604,
+      "step": 1909
+    },
+    {
+      "epoch": 0.2548705631171604,
+      "grad_norm": 1.4592548751795749,
+      "learning_rate": 1.7463462916306103e-05,
+      "loss": 1.7118,
+      "step": 1910
+    },
+    {
+      "epoch": 0.25500400320256206,
+      "grad_norm": 1.1932048673196494,
+      "learning_rate": 1.7460585741759978e-05,
+      "loss": 1.6134,
+      "step": 1911
+    },
+    {
+      "epoch": 0.2551374432879637,
+      "grad_norm": 1.2140073379811964,
+      "learning_rate": 1.7457707173661378e-05,
+      "loss": 1.7213,
+      "step": 1912
+    },
+    {
+      "epoch": 0.25527088337336534,
+      "grad_norm": 1.1053613025091942,
+      "learning_rate": 1.7454827212547988e-05,
+      "loss": 1.6205,
+      "step": 1913
+    },
+    {
+      "epoch": 0.255404323458767,
+      "grad_norm": 1.0688175948277936,
+      "learning_rate": 1.745194585895775e-05,
+      "loss": 1.6845,
+      "step": 1914
+    },
+    {
+      "epoch": 0.2555377635441687,
+      "grad_norm": 1.0696512783400127,
+      "learning_rate": 1.744906311342887e-05,
+      "loss": 1.6337,
+      "step": 1915
+    },
+    {
+      "epoch": 0.2556712036295703,
+      "grad_norm": 1.0723419328954982,
+      "learning_rate": 1.744617897649981e-05,
+      "loss": 1.631,
+      "step": 1916
+    },
+    {
+      "epoch": 0.25580464371497197,
+      "grad_norm": 1.0336966371989529,
+      "learning_rate": 1.7443293448709297e-05,
+      "loss": 1.6656,
+      "step": 1917
+    },
+    {
+      "epoch": 0.25593808380037364,
+      "grad_norm": 1.0461578537171803,
+      "learning_rate": 1.7440406530596312e-05,
+      "loss": 1.6317,
+      "step": 1918
+    },
+    {
+      "epoch": 0.2560715238857753,
+      "grad_norm": 1.1097399246146262,
+      "learning_rate": 1.74375182227001e-05,
+      "loss": 1.6383,
+      "step": 1919
+    },
+    {
+      "epoch": 0.2562049639711769,
+      "grad_norm": 1.0337650567788563,
+      "learning_rate": 1.7434628525560163e-05,
+      "loss": 1.6528,
+      "step": 1920
+    },
+    {
+      "epoch": 0.2563384040565786,
+      "grad_norm": 1.018254122534702,
+      "learning_rate": 1.7431737439716262e-05,
+      "loss": 1.6495,
+      "step": 1921
+    },
+    {
+      "epoch": 0.25647184414198027,
+      "grad_norm": 1.0402215429993655,
+      "learning_rate": 1.7428844965708425e-05,
+      "loss": 1.6646,
+      "step": 1922
+    },
+    {
+      "epoch": 0.2566052842273819,
+      "grad_norm": 1.2456549178005256,
+      "learning_rate": 1.7425951104076925e-05,
+      "loss": 1.6625,
+      "step": 1923
+    },
+    {
+      "epoch": 0.25673872431278355,
+      "grad_norm": 1.1766568400761854,
+      "learning_rate": 1.7423055855362306e-05,
+      "loss": 1.6389,
+      "step": 1924
+    },
+    {
+      "epoch": 0.2568721643981852,
+      "grad_norm": 0.9891969725565736,
+      "learning_rate": 1.742015922010537e-05,
+      "loss": 1.6642,
+      "step": 1925
+    },
+    {
+      "epoch": 0.2570056044835869,
+      "grad_norm": 1.0679382580870513,
+      "learning_rate": 1.7417261198847175e-05,
+      "loss": 1.6542,
+      "step": 1926
+    },
+    {
+      "epoch": 0.2571390445689885,
+      "grad_norm": 1.0468412544977608,
+      "learning_rate": 1.7414361792129034e-05,
+      "loss": 1.6376,
+      "step": 1927
+    },
+    {
+      "epoch": 0.2572724846543902,
+      "grad_norm": 1.0350649856988523,
+      "learning_rate": 1.7411461000492527e-05,
+      "loss": 1.6653,
+      "step": 1928
+    },
+    {
+      "epoch": 0.25740592473979185,
+      "grad_norm": 1.036385533715168,
+      "learning_rate": 1.7408558824479485e-05,
+      "loss": 1.6968,
+      "step": 1929
+    },
+    {
+      "epoch": 0.25753936482519346,
+      "grad_norm": 1.0569592050053869,
+      "learning_rate": 1.7405655264632007e-05,
+      "loss": 1.6024,
+      "step": 1930
+    },
+    {
+      "epoch": 0.25767280491059513,
+      "grad_norm": 1.004763067338033,
+      "learning_rate": 1.7402750321492445e-05,
+      "loss": 1.6594,
+      "step": 1931
+    },
+    {
+      "epoch": 0.2578062449959968,
+      "grad_norm": 1.1608084994679726,
+      "learning_rate": 1.7399843995603404e-05,
+      "loss": 1.657,
+      "step": 1932
+    },
+    {
+      "epoch": 0.2579396850813985,
+      "grad_norm": 1.1836864291351534,
+      "learning_rate": 1.739693628750775e-05,
+      "loss": 1.6493,
+      "step": 1933
+    },
+    {
+      "epoch": 0.2580731251668001,
+      "grad_norm": 1.0463963871461142,
+      "learning_rate": 1.739402719774862e-05,
+      "loss": 1.6738,
+      "step": 1934
+    },
+    {
+      "epoch": 0.25820656525220176,
+      "grad_norm": 1.0417140646609107,
+      "learning_rate": 1.7391116726869395e-05,
+      "loss": 1.6386,
+      "step": 1935
+    },
+    {
+      "epoch": 0.25834000533760343,
+      "grad_norm": 1.0230471101003555,
+      "learning_rate": 1.7388204875413716e-05,
+      "loss": 1.6851,
+      "step": 1936
+    },
+    {
+      "epoch": 0.25847344542300504,
+      "grad_norm": 1.175831444603941,
+      "learning_rate": 1.7385291643925478e-05,
+      "loss": 1.631,
+      "step": 1937
+    },
+    {
+      "epoch": 0.2586068855084067,
+      "grad_norm": 1.0005993864292368,
+      "learning_rate": 1.738237703294885e-05,
+      "loss": 1.6293,
+      "step": 1938
+    },
+    {
+      "epoch": 0.2587403255938084,
+      "grad_norm": 1.071233770103943,
+      "learning_rate": 1.7379461043028242e-05,
+      "loss": 1.6477,
+      "step": 1939
+    },
+    {
+      "epoch": 0.25887376567921005,
+      "grad_norm": 1.27924501051401,
+      "learning_rate": 1.7376543674708332e-05,
+      "loss": 1.6601,
+      "step": 1940
+    },
+    {
+      "epoch": 0.25900720576461167,
+      "grad_norm": 1.0592769595491713,
+      "learning_rate": 1.737362492853405e-05,
+      "loss": 1.7318,
+      "step": 1941
+    },
+    {
+      "epoch": 0.25914064585001334,
+      "grad_norm": 1.0043569196961857,
+      "learning_rate": 1.737070480505058e-05,
+      "loss": 1.6029,
+      "step": 1942
+    },
+    {
+      "epoch": 0.259274085935415,
+      "grad_norm": 1.0651608229693987,
+      "learning_rate": 1.7367783304803373e-05,
+      "loss": 1.6302,
+      "step": 1943
+    },
+    {
+      "epoch": 0.2594075260208166,
+      "grad_norm": 1.0274924793106943,
+      "learning_rate": 1.736486042833813e-05,
+      "loss": 1.6225,
+      "step": 1944
+    },
+    {
+      "epoch": 0.2595409661062183,
+      "grad_norm": 1.0973325495792983,
+      "learning_rate": 1.7361936176200806e-05,
+      "loss": 1.637,
+      "step": 1945
+    },
+    {
+      "epoch": 0.25967440619161997,
+      "grad_norm": 1.1480559231420868,
+      "learning_rate": 1.735901054893763e-05,
+      "loss": 1.6579,
+      "step": 1946
+    },
+    {
+      "epoch": 0.25980784627702164,
+      "grad_norm": 9.742353418129637,
+      "learning_rate": 1.7356083547095065e-05,
+      "loss": 1.7036,
+      "step": 1947
+    },
+    {
+      "epoch": 0.25994128636242325,
+      "grad_norm": 1.1930642002916203,
+      "learning_rate": 1.7353155171219845e-05,
+      "loss": 1.686,
+      "step": 1948
+    },
+    {
+      "epoch": 0.2600747264478249,
+      "grad_norm": 1.1344495102081051,
+      "learning_rate": 1.7350225421858963e-05,
+      "loss": 1.6927,
+      "step": 1949
+    },
+    {
+      "epoch": 0.2602081665332266,
+      "grad_norm": 1.160268471890251,
+      "learning_rate": 1.7347294299559656e-05,
+      "loss": 1.6395,
+      "step": 1950
+    },
+    {
+      "epoch": 0.26034160661862826,
+      "grad_norm": 1.0571542473634332,
+      "learning_rate": 1.7344361804869422e-05,
+      "loss": 1.6482,
+      "step": 1951
+    },
+    {
+      "epoch": 0.2604750467040299,
+      "grad_norm": 1.0741618519583103,
+      "learning_rate": 1.734142793833602e-05,
+      "loss": 1.6322,
+      "step": 1952
+    },
+    {
+      "epoch": 0.26060848678943155,
+      "grad_norm": 1.0810620959227508,
+      "learning_rate": 1.733849270050747e-05,
+      "loss": 1.6614,
+      "step": 1953
+    },
+    {
+      "epoch": 0.2607419268748332,
+      "grad_norm": 1.650858745406716,
+      "learning_rate": 1.7335556091932033e-05,
+      "loss": 1.6594,
+      "step": 1954
+    },
+    {
+      "epoch": 0.26087536696023483,
+      "grad_norm": 1.1357173618685,
+      "learning_rate": 1.7332618113158238e-05,
+      "loss": 1.6246,
+      "step": 1955
+    },
+    {
+      "epoch": 0.2610088070456365,
+      "grad_norm": 1.228866928090115,
+      "learning_rate": 1.7329678764734865e-05,
+      "loss": 1.6502,
+      "step": 1956
+    },
+    {
+      "epoch": 0.2611422471310382,
+      "grad_norm": 1.063500974508951,
+      "learning_rate": 1.732673804721095e-05,
+      "loss": 1.6553,
+      "step": 1957
+    },
+    {
+      "epoch": 0.26127568721643984,
+      "grad_norm": 1.1182247677924306,
+      "learning_rate": 1.732379596113578e-05,
+      "loss": 1.5929,
+      "step": 1958
+    },
+    {
+      "epoch": 0.26140912730184146,
+      "grad_norm": 1.0343485691283105,
+      "learning_rate": 1.7320852507058914e-05,
+      "loss": 1.6195,
+      "step": 1959
+    },
+    {
+      "epoch": 0.26154256738724313,
+      "grad_norm": 1.1899371182298315,
+      "learning_rate": 1.7317907685530152e-05,
+      "loss": 1.6208,
+      "step": 1960
+    },
+    {
+      "epoch": 0.2616760074726448,
+      "grad_norm": 1.2655530384071296,
+      "learning_rate": 1.7314961497099546e-05,
+      "loss": 1.6743,
+      "step": 1961
+    },
+    {
+      "epoch": 0.2618094475580464,
+      "grad_norm": 1.1142841192035378,
+      "learning_rate": 1.7312013942317423e-05,
+      "loss": 1.6915,
+      "step": 1962
+    },
+    {
+      "epoch": 0.2619428876434481,
+      "grad_norm": 1.2748126364938344,
+      "learning_rate": 1.7309065021734345e-05,
+      "loss": 1.6708,
+      "step": 1963
+    },
+    {
+      "epoch": 0.26207632772884976,
+      "grad_norm": 1.0763574665626,
+      "learning_rate": 1.7306114735901135e-05,
+      "loss": 1.6867,
+      "step": 1964
+    },
+    {
+      "epoch": 0.2622097678142514,
+      "grad_norm": 1.1840696437423681,
+      "learning_rate": 1.7303163085368876e-05,
+      "loss": 1.6526,
+      "step": 1965
+    },
+    {
+      "epoch": 0.26234320789965304,
+      "grad_norm": 1.0215098328750423,
+      "learning_rate": 1.7300210070688902e-05,
+      "loss": 1.6395,
+      "step": 1966
+    },
+    {
+      "epoch": 0.2624766479850547,
+      "grad_norm": 1.0619335145218514,
+      "learning_rate": 1.7297255692412807e-05,
+      "loss": 1.6554,
+      "step": 1967
+    },
+    {
+      "epoch": 0.2626100880704564,
+      "grad_norm": 1.0512215210347395,
+      "learning_rate": 1.7294299951092427e-05,
+      "loss": 1.6577,
+      "step": 1968
+    },
+    {
+      "epoch": 0.262743528155858,
+      "grad_norm": 1.1231704876083466,
+      "learning_rate": 1.7291342847279864e-05,
+      "loss": 1.61,
+      "step": 1969
+    },
+    {
+      "epoch": 0.26287696824125967,
+      "grad_norm": 1.051508193824374,
+      "learning_rate": 1.7288384381527473e-05,
+      "loss": 1.6351,
+      "step": 1970
+    },
+    {
+      "epoch": 0.26301040832666134,
+      "grad_norm": 1.0623240318595546,
+      "learning_rate": 1.7285424554387863e-05,
+      "loss": 1.6435,
+      "step": 1971
+    },
+    {
+      "epoch": 0.263143848412063,
+      "grad_norm": 1.0134535097893844,
+      "learning_rate": 1.7282463366413895e-05,
+      "loss": 1.6195,
+      "step": 1972
+    },
+    {
+      "epoch": 0.2632772884974646,
+      "grad_norm": 1.0944547998591982,
+      "learning_rate": 1.7279500818158678e-05,
+      "loss": 1.6385,
+      "step": 1973
+    },
+    {
+      "epoch": 0.2634107285828663,
+      "grad_norm": 1.014436566990623,
+      "learning_rate": 1.7276536910175596e-05,
+      "loss": 1.5849,
+      "step": 1974
+    },
+    {
+      "epoch": 0.26354416866826796,
+      "grad_norm": 1.0957069455341975,
+      "learning_rate": 1.727357164301826e-05,
+      "loss": 1.6463,
+      "step": 1975
+    },
+    {
+      "epoch": 0.2636776087536696,
+      "grad_norm": 1.2346921385867662,
+      "learning_rate": 1.7270605017240557e-05,
+      "loss": 1.653,
+      "step": 1976
+    },
+    {
+      "epoch": 0.26381104883907125,
+      "grad_norm": 1.0897637998274192,
+      "learning_rate": 1.7267637033396613e-05,
+      "loss": 1.6921,
+      "step": 1977
+    },
+    {
+      "epoch": 0.2639444889244729,
+      "grad_norm": 1.415961246967308,
+      "learning_rate": 1.7264667692040816e-05,
+      "loss": 1.6308,
+      "step": 1978
+    },
+    {
+      "epoch": 0.2640779290098746,
+      "grad_norm": 14.277629558243925,
+      "learning_rate": 1.726169699372781e-05,
+      "loss": 1.6807,
+      "step": 1979
+    },
+    {
+      "epoch": 0.2642113690952762,
+      "grad_norm": 1.3259197439263068,
+      "learning_rate": 1.725872493901248e-05,
+      "loss": 1.6796,
+      "step": 1980
+    },
+    {
+      "epoch": 0.2643448091806779,
+      "grad_norm": 1.4714297960193359,
+      "learning_rate": 1.7255751528449972e-05,
+      "loss": 1.6084,
+      "step": 1981
+    },
+    {
+      "epoch": 0.26447824926607955,
+      "grad_norm": 1.3783447543972154,
+      "learning_rate": 1.7252776762595695e-05,
+      "loss": 1.674,
+      "step": 1982
+    },
+    {
+      "epoch": 0.26461168935148116,
+      "grad_norm": 1.345089903670124,
+      "learning_rate": 1.724980064200529e-05,
+      "loss": 1.6719,
+      "step": 1983
+    },
+    {
+      "epoch": 0.26474512943688283,
+      "grad_norm": 1.6788138951425957,
+      "learning_rate": 1.724682316723467e-05,
+      "loss": 1.6242,
+      "step": 1984
+    },
+    {
+      "epoch": 0.2648785695222845,
+      "grad_norm": 1.057556035989643,
+      "learning_rate": 1.724384433883999e-05,
+      "loss": 1.6688,
+      "step": 1985
+    },
+    {
+      "epoch": 0.26501200960768617,
+      "grad_norm": 1.935146990397484,
+      "learning_rate": 1.724086415737766e-05,
+      "loss": 1.6685,
+      "step": 1986
+    },
+    {
+      "epoch": 0.2651454496930878,
+      "grad_norm": 1.4029312343106273,
+      "learning_rate": 1.7237882623404347e-05,
+      "loss": 1.69,
+      "step": 1987
+    },
+    {
+      "epoch": 0.26527888977848946,
+      "grad_norm": 1.0640529672227876,
+      "learning_rate": 1.7234899737476968e-05,
+      "loss": 1.5853,
+      "step": 1988
+    },
+    {
+      "epoch": 0.2654123298638911,
+      "grad_norm": 1.0733639436628284,
+      "learning_rate": 1.7231915500152685e-05,
+      "loss": 1.677,
+      "step": 1989
+    },
+    {
+      "epoch": 0.26554576994929274,
+      "grad_norm": 1.0241005822898823,
+      "learning_rate": 1.722892991198893e-05,
+      "loss": 1.6374,
+      "step": 1990
+    },
+    {
+      "epoch": 0.2656792100346944,
+      "grad_norm": 1.0875108906486028,
+      "learning_rate": 1.7225942973543368e-05,
+      "loss": 1.6445,
+      "step": 1991
+    },
+    {
+      "epoch": 0.2658126501200961,
+      "grad_norm": 1.2697214356659474,
+      "learning_rate": 1.7222954685373933e-05,
+      "loss": 1.5832,
+      "step": 1992
+    },
+    {
+      "epoch": 0.26594609020549775,
+      "grad_norm": 1.0830834504640086,
+      "learning_rate": 1.7219965048038795e-05,
+      "loss": 1.6888,
+      "step": 1993
+    },
+    {
+      "epoch": 0.26607953029089937,
+      "grad_norm": 1.3072951866209532,
+      "learning_rate": 1.7216974062096392e-05,
+      "loss": 1.6289,
+      "step": 1994
+    },
+    {
+      "epoch": 0.26621297037630104,
+      "grad_norm": 1.0836421243679286,
+      "learning_rate": 1.72139817281054e-05,
+      "loss": 1.6658,
+      "step": 1995
+    },
+    {
+      "epoch": 0.2663464104617027,
+      "grad_norm": 1.8046043701192065,
+      "learning_rate": 1.7210988046624758e-05,
+      "loss": 1.622,
+      "step": 1996
+    },
+    {
+      "epoch": 0.2664798505471043,
+      "grad_norm": 1.0933300325897486,
+      "learning_rate": 1.720799301821365e-05,
+      "loss": 1.6297,
+      "step": 1997
+    },
+    {
+      "epoch": 0.266613290632506,
+      "grad_norm": 1.0459969375513192,
+      "learning_rate": 1.720499664343151e-05,
+      "loss": 1.6484,
+      "step": 1998
+    },
+    {
+      "epoch": 0.26674673071790767,
+      "grad_norm": 1.0751449290453639,
+      "learning_rate": 1.720199892283803e-05,
+      "loss": 1.625,
+      "step": 1999
+    },
+    {
+      "epoch": 0.26688017080330934,
+      "grad_norm": 1.0928368447753933,
+      "learning_rate": 1.719899985699315e-05,
+      "loss": 1.6806,
+      "step": 2000
+    },
+    {
+      "epoch": 0.26701361088871095,
+      "grad_norm": 1.0349222592130347,
+      "learning_rate": 1.7195999446457053e-05,
+      "loss": 1.595,
+      "step": 2001
+    },
+    {
+      "epoch": 0.2671470509741126,
+      "grad_norm": 1.063465046316729,
+      "learning_rate": 1.7192997691790197e-05,
+      "loss": 1.6143,
+      "step": 2002
+    },
+    {
+      "epoch": 0.2672804910595143,
+      "grad_norm": 1.0284024134886187,
+      "learning_rate": 1.7189994593553266e-05,
+      "loss": 1.6447,
+      "step": 2003
+    },
+    {
+      "epoch": 0.2674139311449159,
+      "grad_norm": 1.307165702348938,
+      "learning_rate": 1.7186990152307203e-05,
+      "loss": 1.6691,
+      "step": 2004
+    },
+    {
+      "epoch": 0.2675473712303176,
+      "grad_norm": 1.119045388640825,
+      "learning_rate": 1.718398436861321e-05,
+      "loss": 1.6466,
+      "step": 2005
+    },
+    {
+      "epoch": 0.26768081131571925,
+      "grad_norm": 1.0226913583499544,
+      "learning_rate": 1.718097724303273e-05,
+      "loss": 1.6593,
+      "step": 2006
+    },
+    {
+      "epoch": 0.2678142514011209,
+      "grad_norm": 1.1645428303381167,
+      "learning_rate": 1.717796877612746e-05,
+      "loss": 1.6316,
+      "step": 2007
+    },
+    {
+      "epoch": 0.26794769148652253,
+      "grad_norm": 1.0290487690958936,
+      "learning_rate": 1.7174958968459344e-05,
+      "loss": 1.6416,
+      "step": 2008
+    },
+    {
+      "epoch": 0.2680811315719242,
+      "grad_norm": 1.113710094399956,
+      "learning_rate": 1.7171947820590584e-05,
+      "loss": 1.6459,
+      "step": 2009
+    },
+    {
+      "epoch": 0.2682145716573259,
+      "grad_norm": 1.0721033027517342,
+      "learning_rate": 1.7168935333083624e-05,
+      "loss": 1.6481,
+      "step": 2010
+    },
+    {
+      "epoch": 0.2683480117427275,
+      "grad_norm": 4.148877971329257,
+      "learning_rate": 1.7165921506501168e-05,
+      "loss": 1.6853,
+      "step": 2011
+    },
+    {
+      "epoch": 0.26848145182812916,
+      "grad_norm": 1.026325617891188,
+      "learning_rate": 1.716290634140616e-05,
+      "loss": 1.6395,
+      "step": 2012
+    },
+    {
+      "epoch": 0.26861489191353083,
+      "grad_norm": 1.0458614779045183,
+      "learning_rate": 1.71598898383618e-05,
+      "loss": 1.6385,
+      "step": 2013
+    },
+    {
+      "epoch": 0.2687483319989325,
+      "grad_norm": 1.3489378298749055,
+      "learning_rate": 1.715687199793154e-05,
+      "loss": 1.6306,
+      "step": 2014
+    },
+    {
+      "epoch": 0.2688817720843341,
+      "grad_norm": 1.066106782654306,
+      "learning_rate": 1.7153852820679073e-05,
+      "loss": 1.6143,
+      "step": 2015
+    },
+    {
+      "epoch": 0.2690152121697358,
+      "grad_norm": 1.0517288617879899,
+      "learning_rate": 1.715083230716835e-05,
+      "loss": 1.6918,
+      "step": 2016
+    },
+    {
+      "epoch": 0.26914865225513745,
+      "grad_norm": 1.0631774833951584,
+      "learning_rate": 1.7147810457963565e-05,
+      "loss": 1.6196,
+      "step": 2017
+    },
+    {
+      "epoch": 0.2692820923405391,
+      "grad_norm": 1.5500958673330705,
+      "learning_rate": 1.714478727362917e-05,
+      "loss": 1.605,
+      "step": 2018
+    },
+    {
+      "epoch": 0.26941553242594074,
+      "grad_norm": 1.2061760172934797,
+      "learning_rate": 1.7141762754729855e-05,
+      "loss": 1.627,
+      "step": 2019
+    },
+    {
+      "epoch": 0.2695489725113424,
+      "grad_norm": 1.063335069996398,
+      "learning_rate": 1.713873690183057e-05,
+      "loss": 1.6546,
+      "step": 2020
+    },
+    {
+      "epoch": 0.2696824125967441,
+      "grad_norm": 1.0345427555272797,
+      "learning_rate": 1.713570971549651e-05,
+      "loss": 1.6332,
+      "step": 2021
+    },
+    {
+      "epoch": 0.2698158526821457,
+      "grad_norm": 1.0387506796892585,
+      "learning_rate": 1.713268119629312e-05,
+      "loss": 1.6208,
+      "step": 2022
+    },
+    {
+      "epoch": 0.26994929276754737,
+      "grad_norm": 1.0451937535961915,
+      "learning_rate": 1.7129651344786088e-05,
+      "loss": 1.6419,
+      "step": 2023
+    },
+    {
+      "epoch": 0.27008273285294904,
+      "grad_norm": 1.0424161111119927,
+      "learning_rate": 1.7126620161541364e-05,
+      "loss": 1.6801,
+      "step": 2024
+    },
+    {
+      "epoch": 0.2702161729383507,
+      "grad_norm": 1.0453586363001601,
+      "learning_rate": 1.712358764712513e-05,
+      "loss": 1.6202,
+      "step": 2025
+    },
+    {
+      "epoch": 0.2703496130237523,
+      "grad_norm": 1.1453862840223692,
+      "learning_rate": 1.7120553802103828e-05,
+      "loss": 1.6623,
+      "step": 2026
+    },
+    {
+      "epoch": 0.270483053109154,
+      "grad_norm": 1.325903344513902,
+      "learning_rate": 1.7117518627044148e-05,
+      "loss": 1.6376,
+      "step": 2027
+    },
+    {
+      "epoch": 0.27061649319455566,
+      "grad_norm": 1.080112760240277,
+      "learning_rate": 1.7114482122513024e-05,
+      "loss": 1.6184,
+      "step": 2028
+    },
+    {
+      "epoch": 0.2707499332799573,
+      "grad_norm": 1.0594663654156977,
+      "learning_rate": 1.711144428907764e-05,
+      "loss": 1.6887,
+      "step": 2029
+    },
+    {
+      "epoch": 0.27088337336535895,
+      "grad_norm": 1.0462695966057693,
+      "learning_rate": 1.710840512730543e-05,
+      "loss": 1.5665,
+      "step": 2030
+    },
+    {
+      "epoch": 0.2710168134507606,
+      "grad_norm": 1.1251591649106787,
+      "learning_rate": 1.7105364637764075e-05,
+      "loss": 1.6294,
+      "step": 2031
+    },
+    {
+      "epoch": 0.2711502535361623,
+      "grad_norm": 1.1329461782399421,
+      "learning_rate": 1.7102322821021505e-05,
+      "loss": 1.6344,
+      "step": 2032
+    },
+    {
+      "epoch": 0.2712836936215639,
+      "grad_norm": 1.0166935218662512,
+      "learning_rate": 1.709927967764589e-05,
+      "loss": 1.6171,
+      "step": 2033
+    },
+    {
+      "epoch": 0.2714171337069656,
+      "grad_norm": 1.2601520388342762,
+      "learning_rate": 1.7096235208205665e-05,
+      "loss": 1.634,
+      "step": 2034
+    },
+    {
+      "epoch": 0.27155057379236724,
+      "grad_norm": 1.076982069439433,
+      "learning_rate": 1.7093189413269497e-05,
+      "loss": 1.6637,
+      "step": 2035
+    },
+    {
+      "epoch": 0.27168401387776886,
+      "grad_norm": 1.0321358819200304,
+      "learning_rate": 1.70901422934063e-05,
+      "loss": 1.6222,
+      "step": 2036
+    },
+    {
+      "epoch": 0.27181745396317053,
+      "grad_norm": 1.0654015458142159,
+      "learning_rate": 1.708709384918525e-05,
+      "loss": 1.586,
+      "step": 2037
+    },
+    {
+      "epoch": 0.2719508940485722,
+      "grad_norm": 0.9875095341342335,
+      "learning_rate": 1.708404408117576e-05,
+      "loss": 1.6419,
+      "step": 2038
+    },
+    {
+      "epoch": 0.27208433413397387,
+      "grad_norm": 1.2163989586437813,
+      "learning_rate": 1.708099298994749e-05,
+      "loss": 1.6356,
+      "step": 2039
+    },
+    {
+      "epoch": 0.2722177742193755,
+      "grad_norm": 1.0424446221236658,
+      "learning_rate": 1.7077940576070347e-05,
+      "loss": 1.6426,
+      "step": 2040
+    },
+    {
+      "epoch": 0.27235121430477716,
+      "grad_norm": 1.1917437972972873,
+      "learning_rate": 1.707488684011449e-05,
+      "loss": 1.6128,
+      "step": 2041
+    },
+    {
+      "epoch": 0.2724846543901788,
+      "grad_norm": 1.0561747251809694,
+      "learning_rate": 1.7071831782650325e-05,
+      "loss": 1.6523,
+      "step": 2042
+    },
+    {
+      "epoch": 0.27261809447558044,
+      "grad_norm": 1.0385624086824277,
+      "learning_rate": 1.7068775404248497e-05,
+      "loss": 1.6634,
+      "step": 2043
+    },
+    {
+      "epoch": 0.2727515345609821,
+      "grad_norm": 1.303828973820152,
+      "learning_rate": 1.7065717705479906e-05,
+      "loss": 1.6508,
+      "step": 2044
+    },
+    {
+      "epoch": 0.2728849746463838,
+      "grad_norm": 1.1910010358615803,
+      "learning_rate": 1.706265868691569e-05,
+      "loss": 1.6699,
+      "step": 2045
+    },
+    {
+      "epoch": 0.27301841473178545,
+      "grad_norm": 1.1784688305684132,
+      "learning_rate": 1.7059598349127245e-05,
+      "loss": 1.645,
+      "step": 2046
+    },
+    {
+      "epoch": 0.27315185481718707,
+      "grad_norm": 0.985356851054121,
+      "learning_rate": 1.7056536692686204e-05,
+      "loss": 1.6146,
+      "step": 2047
+    },
+    {
+      "epoch": 0.27328529490258874,
+      "grad_norm": 1.0255343499954996,
+      "learning_rate": 1.7053473718164455e-05,
+      "loss": 1.6418,
+      "step": 2048
+    },
+    {
+      "epoch": 0.2734187349879904,
+      "grad_norm": 1.0638983947233618,
+      "learning_rate": 1.7050409426134117e-05,
+      "loss": 1.7435,
+      "step": 2049
+    },
+    {
+      "epoch": 0.273552175073392,
+      "grad_norm": 1.0851770979360476,
+      "learning_rate": 1.704734381716757e-05,
+      "loss": 1.658,
+      "step": 2050
+    },
+    {
+      "epoch": 0.2736856151587937,
+      "grad_norm": 1.1763821324448986,
+      "learning_rate": 1.704427689183744e-05,
+      "loss": 1.6535,
+      "step": 2051
+    },
+    {
+      "epoch": 0.27381905524419536,
+      "grad_norm": 1.3104441981575137,
+      "learning_rate": 1.7041208650716586e-05,
+      "loss": 1.6589,
+      "step": 2052
+    },
+    {
+      "epoch": 0.27395249532959703,
+      "grad_norm": 1.0136315973372723,
+      "learning_rate": 1.703813909437812e-05,
+      "loss": 1.6133,
+      "step": 2053
+    },
+    {
+      "epoch": 0.27408593541499865,
+      "grad_norm": 0.9960414593425704,
+      "learning_rate": 1.7035068223395407e-05,
+      "loss": 1.621,
+      "step": 2054
+    },
+    {
+      "epoch": 0.2742193755004003,
+      "grad_norm": 1.2127828477430966,
+      "learning_rate": 1.7031996038342045e-05,
+      "loss": 1.6397,
+      "step": 2055
+    },
+    {
+      "epoch": 0.274352815585802,
+      "grad_norm": 0.9828639066951254,
+      "learning_rate": 1.702892253979189e-05,
+      "loss": 1.6125,
+      "step": 2056
+    },
+    {
+      "epoch": 0.2744862556712036,
+      "grad_norm": 0.9758056512184737,
+      "learning_rate": 1.7025847728319027e-05,
+      "loss": 1.6425,
+      "step": 2057
+    },
+    {
+      "epoch": 0.2746196957566053,
+      "grad_norm": 1.0754633573934569,
+      "learning_rate": 1.7022771604497802e-05,
+      "loss": 1.6622,
+      "step": 2058
+    },
+    {
+      "epoch": 0.27475313584200695,
+      "grad_norm": 1.0344089922720026,
+      "learning_rate": 1.70196941689028e-05,
+      "loss": 1.6888,
+      "step": 2059
+    },
+    {
+      "epoch": 0.2748865759274086,
+      "grad_norm": 1.0736302080536169,
+      "learning_rate": 1.7016615422108847e-05,
+      "loss": 1.658,
+      "step": 2060
+    },
+    {
+      "epoch": 0.27502001601281023,
+      "grad_norm": 1.20409158832475,
+      "learning_rate": 1.7013535364691023e-05,
+      "loss": 1.6427,
+      "step": 2061
+    },
+    {
+      "epoch": 0.2751534560982119,
+      "grad_norm": 1.069163552633532,
+      "learning_rate": 1.7010453997224642e-05,
+      "loss": 1.6237,
+      "step": 2062
+    },
+    {
+      "epoch": 0.27528689618361357,
+      "grad_norm": 1.1894978043506685,
+      "learning_rate": 1.7007371320285275e-05,
+      "loss": 1.6635,
+      "step": 2063
+    },
+    {
+      "epoch": 0.2754203362690152,
+      "grad_norm": 1.1903195019190183,
+      "learning_rate": 1.7004287334448723e-05,
+      "loss": 1.5996,
+      "step": 2064
+    },
+    {
+      "epoch": 0.27555377635441686,
+      "grad_norm": 1.0460875658896527,
+      "learning_rate": 1.7001202040291048e-05,
+      "loss": 1.7078,
+      "step": 2065
+    },
+    {
+      "epoch": 0.2756872164398185,
+      "grad_norm": 1.4369499128117869,
+      "learning_rate": 1.699811543838854e-05,
+      "loss": 1.6444,
+      "step": 2066
+    },
+    {
+      "epoch": 0.2758206565252202,
+      "grad_norm": 1.209341903013088,
+      "learning_rate": 1.6995027529317746e-05,
+      "loss": 1.6181,
+      "step": 2067
+    },
+    {
+      "epoch": 0.2759540966106218,
+      "grad_norm": 1.1036519149996684,
+      "learning_rate": 1.6991938313655453e-05,
+      "loss": 1.6224,
+      "step": 2068
+    },
+    {
+      "epoch": 0.2760875366960235,
+      "grad_norm": 1.146216102199816,
+      "learning_rate": 1.6988847791978687e-05,
+      "loss": 1.6165,
+      "step": 2069
+    },
+    {
+      "epoch": 0.27622097678142515,
+      "grad_norm": 1.1564719422933685,
+      "learning_rate": 1.6985755964864723e-05,
+      "loss": 1.7315,
+      "step": 2070
+    },
+    {
+      "epoch": 0.27635441686682677,
+      "grad_norm": 1.1449774555454721,
+      "learning_rate": 1.698266283289108e-05,
+      "loss": 1.6398,
+      "step": 2071
+    },
+    {
+      "epoch": 0.27648785695222844,
+      "grad_norm": 1.0564082538230344,
+      "learning_rate": 1.6979568396635526e-05,
+      "loss": 1.5539,
+      "step": 2072
+    },
+    {
+      "epoch": 0.2766212970376301,
+      "grad_norm": 1.0646332200728448,
+      "learning_rate": 1.6976472656676058e-05,
+      "loss": 1.6822,
+      "step": 2073
+    },
+    {
+      "epoch": 0.2767547371230318,
+      "grad_norm": 1.2326207337266486,
+      "learning_rate": 1.697337561359093e-05,
+      "loss": 1.5862,
+      "step": 2074
+    },
+    {
+      "epoch": 0.2768881772084334,
+      "grad_norm": 1.2149498397188576,
+      "learning_rate": 1.6970277267958625e-05,
+      "loss": 1.6561,
+      "step": 2075
+    },
+    {
+      "epoch": 0.27702161729383507,
+      "grad_norm": 1.15730574634711,
+      "learning_rate": 1.6967177620357894e-05,
+      "loss": 1.6772,
+      "step": 2076
+    },
+    {
+      "epoch": 0.27715505737923674,
+      "grad_norm": 1.041366398730123,
+      "learning_rate": 1.6964076671367703e-05,
+      "loss": 1.6171,
+      "step": 2077
+    },
+    {
+      "epoch": 0.27728849746463835,
+      "grad_norm": 1.0099579426889422,
+      "learning_rate": 1.696097442156728e-05,
+      "loss": 1.6003,
+      "step": 2078
+    },
+    {
+      "epoch": 0.27742193755004,
+      "grad_norm": 1.1597495545248406,
+      "learning_rate": 1.6957870871536086e-05,
+      "loss": 1.629,
+      "step": 2079
+    },
+    {
+      "epoch": 0.2775553776354417,
+      "grad_norm": 1.2971905519568292,
+      "learning_rate": 1.6954766021853836e-05,
+      "loss": 1.6335,
+      "step": 2080
+    },
+    {
+      "epoch": 0.27768881772084336,
+      "grad_norm": 1.0081447437285673,
+      "learning_rate": 1.6951659873100474e-05,
+      "loss": 1.661,
+      "step": 2081
+    },
+    {
+      "epoch": 0.277822257806245,
+      "grad_norm": 1.155709221150846,
+      "learning_rate": 1.6948552425856197e-05,
+      "loss": 1.6418,
+      "step": 2082
+    },
+    {
+      "epoch": 0.27795569789164665,
+      "grad_norm": 1.3578902806881925,
+      "learning_rate": 1.694544368070144e-05,
+      "loss": 1.7022,
+      "step": 2083
+    },
+    {
+      "epoch": 0.2780891379770483,
+      "grad_norm": 1.0473984071647915,
+      "learning_rate": 1.694233363821688e-05,
+      "loss": 1.6471,
+      "step": 2084
+    },
+    {
+      "epoch": 0.27822257806244993,
+      "grad_norm": 1.0998361204044,
+      "learning_rate": 1.6939222298983432e-05,
+      "loss": 1.6853,
+      "step": 2085
+    },
+    {
+      "epoch": 0.2783560181478516,
+      "grad_norm": 1.0597277511694383,
+      "learning_rate": 1.693610966358227e-05,
+      "loss": 1.6518,
+      "step": 2086
+    },
+    {
+      "epoch": 0.2784894582332533,
+      "grad_norm": 1.0956187785249976,
+      "learning_rate": 1.693299573259479e-05,
+      "loss": 1.6657,
+      "step": 2087
+    },
+    {
+      "epoch": 0.27862289831865494,
+      "grad_norm": 1.136040109917539,
+      "learning_rate": 1.6929880506602644e-05,
+      "loss": 1.6538,
+      "step": 2088
+    },
+    {
+      "epoch": 0.27875633840405656,
+      "grad_norm": 1.13032282747865,
+      "learning_rate": 1.692676398618772e-05,
+      "loss": 1.6329,
+      "step": 2089
+    },
+    {
+      "epoch": 0.27888977848945823,
+      "grad_norm": 1.297996071275606,
+      "learning_rate": 1.6923646171932148e-05,
+      "loss": 1.6816,
+      "step": 2090
+    },
+    {
+      "epoch": 0.2790232185748599,
+      "grad_norm": 1.036055755996991,
+      "learning_rate": 1.6920527064418298e-05,
+      "loss": 1.6716,
+      "step": 2091
+    },
+    {
+      "epoch": 0.27915665866026157,
+      "grad_norm": 1.1008523939624144,
+      "learning_rate": 1.6917406664228785e-05,
+      "loss": 1.6202,
+      "step": 2092
+    },
+    {
+      "epoch": 0.2792900987456632,
+      "grad_norm": 1.1056127215942615,
+      "learning_rate": 1.6914284971946466e-05,
+      "loss": 1.6673,
+      "step": 2093
+    },
+    {
+      "epoch": 0.27942353883106485,
+      "grad_norm": 1.3280112976952287,
+      "learning_rate": 1.6911161988154435e-05,
+      "loss": 1.6588,
+      "step": 2094
+    },
+    {
+      "epoch": 0.2795569789164665,
+      "grad_norm": 1.2052642948323982,
+      "learning_rate": 1.6908037713436037e-05,
+      "loss": 1.6314,
+      "step": 2095
+    },
+    {
+      "epoch": 0.27969041900186814,
+      "grad_norm": 0.9865645913607,
+      "learning_rate": 1.6904912148374837e-05,
+      "loss": 1.6214,
+      "step": 2096
+    },
+    {
+      "epoch": 0.2798238590872698,
+      "grad_norm": 1.059342517625569,
+      "learning_rate": 1.6901785293554667e-05,
+      "loss": 1.6119,
+      "step": 2097
+    },
+    {
+      "epoch": 0.2799572991726715,
+      "grad_norm": 1.0202929873543987,
+      "learning_rate": 1.6898657149559585e-05,
+      "loss": 1.681,
+      "step": 2098
+    },
+    {
+      "epoch": 0.28009073925807315,
+      "grad_norm": 1.0885504607819505,
+      "learning_rate": 1.689552771697389e-05,
+      "loss": 1.641,
+      "step": 2099
+    },
+    {
+      "epoch": 0.28022417934347477,
+      "grad_norm": 1.2387895966845563,
+      "learning_rate": 1.6892396996382125e-05,
+      "loss": 1.6229,
+      "step": 2100
+    },
+    {
+      "epoch": 0.28035761942887644,
+      "grad_norm": 1.091745358204505,
+      "learning_rate": 1.6889264988369074e-05,
+      "loss": 1.5917,
+      "step": 2101
+    },
+    {
+      "epoch": 0.2804910595142781,
+      "grad_norm": 1.003916358618551,
+      "learning_rate": 1.688613169351976e-05,
+      "loss": 1.6095,
+      "step": 2102
+    },
+    {
+      "epoch": 0.2806244995996797,
+      "grad_norm": 0.9972302973585249,
+      "learning_rate": 1.6882997112419452e-05,
+      "loss": 1.6731,
+      "step": 2103
+    },
+    {
+      "epoch": 0.2807579396850814,
+      "grad_norm": 1.0210763608341367,
+      "learning_rate": 1.6879861245653647e-05,
+      "loss": 1.7185,
+      "step": 2104
+    },
+    {
+      "epoch": 0.28089137977048306,
+      "grad_norm": 1.0637023781709813,
+      "learning_rate": 1.687672409380809e-05,
+      "loss": 1.6653,
+      "step": 2105
+    },
+    {
+      "epoch": 0.28102481985588473,
+      "grad_norm": 1.0336190974023656,
+      "learning_rate": 1.687358565746877e-05,
+      "loss": 1.6459,
+      "step": 2106
+    },
+    {
+      "epoch": 0.28115825994128635,
+      "grad_norm": 1.0180661487350322,
+      "learning_rate": 1.68704459372219e-05,
+      "loss": 1.6258,
+      "step": 2107
+    },
+    {
+      "epoch": 0.281291700026688,
+      "grad_norm": 1.1141096916948245,
+      "learning_rate": 1.686730493365396e-05,
+      "loss": 1.658,
+      "step": 2108
+    },
+    {
+      "epoch": 0.2814251401120897,
+      "grad_norm": 3.197208541082059,
+      "learning_rate": 1.686416264735164e-05,
+      "loss": 1.818,
+      "step": 2109
+    },
+    {
+      "epoch": 0.2815585801974913,
+      "grad_norm": 1.221770226887997,
+      "learning_rate": 1.6861019078901894e-05,
+      "loss": 1.558,
+      "step": 2110
+    },
+    {
+      "epoch": 0.281692020282893,
+      "grad_norm": 1.0374977596663437,
+      "learning_rate": 1.6857874228891896e-05,
+      "loss": 1.6517,
+      "step": 2111
+    },
+    {
+      "epoch": 0.28182546036829464,
+      "grad_norm": 1.0335222022664212,
+      "learning_rate": 1.6854728097909073e-05,
+      "loss": 1.6713,
+      "step": 2112
+    },
+    {
+      "epoch": 0.2819589004536963,
+      "grad_norm": 1.3195530821608707,
+      "learning_rate": 1.6851580686541087e-05,
+      "loss": 1.7385,
+      "step": 2113
+    },
+    {
+      "epoch": 0.28209234053909793,
+      "grad_norm": 1.195240974112386,
+      "learning_rate": 1.6848431995375834e-05,
+      "loss": 1.6521,
+      "step": 2114
+    },
+    {
+      "epoch": 0.2822257806244996,
+      "grad_norm": 1.0436378450311068,
+      "learning_rate": 1.684528202500146e-05,
+      "loss": 1.6722,
+      "step": 2115
+    },
+    {
+      "epoch": 0.28235922070990127,
+      "grad_norm": 1.02333315302456,
+      "learning_rate": 1.6842130776006332e-05,
+      "loss": 1.6158,
+      "step": 2116
+    },
+    {
+      "epoch": 0.2824926607953029,
+      "grad_norm": 1.107427692009098,
+      "learning_rate": 1.6838978248979083e-05,
+      "loss": 1.5904,
+      "step": 2117
+    },
+    {
+      "epoch": 0.28262610088070456,
+      "grad_norm": 1.1045697230959286,
+      "learning_rate": 1.683582444450856e-05,
+      "loss": 1.6672,
+      "step": 2118
+    },
+    {
+      "epoch": 0.2827595409661062,
+      "grad_norm": 1.0660148829983587,
+      "learning_rate": 1.6832669363183863e-05,
+      "loss": 1.5848,
+      "step": 2119
+    },
+    {
+      "epoch": 0.2828929810515079,
+      "grad_norm": 0.997746177166706,
+      "learning_rate": 1.6829513005594318e-05,
+      "loss": 1.6168,
+      "step": 2120
+    },
+    {
+      "epoch": 0.2830264211369095,
+      "grad_norm": 1.0851170598026787,
+      "learning_rate": 1.6826355372329502e-05,
+      "loss": 1.6698,
+      "step": 2121
+    },
+    {
+      "epoch": 0.2831598612223112,
+      "grad_norm": 1.157100041505963,
+      "learning_rate": 1.682319646397922e-05,
+      "loss": 1.6833,
+      "step": 2122
+    },
+    {
+      "epoch": 0.28329330130771285,
+      "grad_norm": 1.0494927317075036,
+      "learning_rate": 1.682003628113353e-05,
+      "loss": 1.6263,
+      "step": 2123
+    },
+    {
+      "epoch": 0.28342674139311447,
+      "grad_norm": 1.0386913852887472,
+      "learning_rate": 1.6816874824382704e-05,
+      "loss": 1.6658,
+      "step": 2124
+    },
+    {
+      "epoch": 0.28356018147851614,
+      "grad_norm": 0.9903946170282486,
+      "learning_rate": 1.6813712094317282e-05,
+      "loss": 1.5875,
+      "step": 2125
+    },
+    {
+      "epoch": 0.2836936215639178,
+      "grad_norm": 1.0742398505956183,
+      "learning_rate": 1.681054809152801e-05,
+      "loss": 1.713,
+      "step": 2126
+    },
+    {
+      "epoch": 0.2838270616493195,
+      "grad_norm": 1.0730348396285945,
+      "learning_rate": 1.6807382816605903e-05,
+      "loss": 1.6241,
+      "step": 2127
+    },
+    {
+      "epoch": 0.2839605017347211,
+      "grad_norm": 1.2030549884856265,
+      "learning_rate": 1.6804216270142183e-05,
+      "loss": 1.6708,
+      "step": 2128
+    },
+    {
+      "epoch": 0.28409394182012276,
+      "grad_norm": 1.0495436375707339,
+      "learning_rate": 1.6801048452728338e-05,
+      "loss": 1.6271,
+      "step": 2129
+    },
+    {
+      "epoch": 0.28422738190552443,
+      "grad_norm": 1.0625147413097187,
+      "learning_rate": 1.6797879364956075e-05,
+      "loss": 1.6423,
+      "step": 2130
+    },
+    {
+      "epoch": 0.28436082199092605,
+      "grad_norm": 1.1920231207118999,
+      "learning_rate": 1.679470900741734e-05,
+      "loss": 1.6252,
+      "step": 2131
+    },
+    {
+      "epoch": 0.2844942620763277,
+      "grad_norm": 1.0599528204014201,
+      "learning_rate": 1.6791537380704326e-05,
+      "loss": 1.6599,
+      "step": 2132
+    },
+    {
+      "epoch": 0.2846277021617294,
+      "grad_norm": 1.0231405998686767,
+      "learning_rate": 1.6788364485409454e-05,
+      "loss": 1.6155,
+      "step": 2133
+    },
+    {
+      "epoch": 0.28476114224713106,
+      "grad_norm": 0.9948628308935985,
+      "learning_rate": 1.6785190322125382e-05,
+      "loss": 1.6538,
+      "step": 2134
+    },
+    {
+      "epoch": 0.2848945823325327,
+      "grad_norm": 3.8142735734554662,
+      "learning_rate": 1.678201489144501e-05,
+      "loss": 1.6744,
+      "step": 2135
+    },
+    {
+      "epoch": 0.28502802241793435,
+      "grad_norm": 1.1704524649360155,
+      "learning_rate": 1.6778838193961475e-05,
+      "loss": 1.7045,
+      "step": 2136
+    },
+    {
+      "epoch": 0.285161462503336,
+      "grad_norm": 1.1934124423424233,
+      "learning_rate": 1.6775660230268146e-05,
+      "loss": 1.6685,
+      "step": 2137
+    },
+    {
+      "epoch": 0.28529490258873763,
+      "grad_norm": 1.0625696814614358,
+      "learning_rate": 1.6772481000958628e-05,
+      "loss": 1.6241,
+      "step": 2138
+    },
+    {
+      "epoch": 0.2854283426741393,
+      "grad_norm": 1.0424150836147343,
+      "learning_rate": 1.6769300506626766e-05,
+      "loss": 1.5597,
+      "step": 2139
+    },
+    {
+      "epoch": 0.28556178275954097,
+      "grad_norm": 1.3055292931541427,
+      "learning_rate": 1.676611874786664e-05,
+      "loss": 1.6378,
+      "step": 2140
+    },
+    {
+      "epoch": 0.28569522284494264,
+      "grad_norm": 1.207338132471859,
+      "learning_rate": 1.676293572527257e-05,
+      "loss": 1.6932,
+      "step": 2141
+    },
+    {
+      "epoch": 0.28582866293034426,
+      "grad_norm": 1.3576084701670912,
+      "learning_rate": 1.6759751439439105e-05,
+      "loss": 1.6737,
+      "step": 2142
+    },
+    {
+      "epoch": 0.2859621030157459,
+      "grad_norm": 1.1131581474723593,
+      "learning_rate": 1.675656589096103e-05,
+      "loss": 1.6575,
+      "step": 2143
+    },
+    {
+      "epoch": 0.2860955431011476,
+      "grad_norm": 1.0404856911010416,
+      "learning_rate": 1.6753379080433375e-05,
+      "loss": 1.6199,
+      "step": 2144
+    },
+    {
+      "epoch": 0.2862289831865492,
+      "grad_norm": 1.0326352373682337,
+      "learning_rate": 1.6750191008451403e-05,
+      "loss": 1.6281,
+      "step": 2145
+    },
+    {
+      "epoch": 0.2863624232719509,
+      "grad_norm": 1.0618014692593303,
+      "learning_rate": 1.6747001675610596e-05,
+      "loss": 1.7052,
+      "step": 2146
+    },
+    {
+      "epoch": 0.28649586335735255,
+      "grad_norm": 1.156021303055593,
+      "learning_rate": 1.67438110825067e-05,
+      "loss": 1.6295,
+      "step": 2147
+    },
+    {
+      "epoch": 0.2866293034427542,
+      "grad_norm": 1.0687356689670264,
+      "learning_rate": 1.674061922973567e-05,
+      "loss": 1.6377,
+      "step": 2148
+    },
+    {
+      "epoch": 0.28676274352815584,
+      "grad_norm": 1.0422058937640384,
+      "learning_rate": 1.6737426117893716e-05,
+      "loss": 1.5742,
+      "step": 2149
+    },
+    {
+      "epoch": 0.2868961836135575,
+      "grad_norm": 1.0058798981165387,
+      "learning_rate": 1.673423174757727e-05,
+      "loss": 1.6808,
+      "step": 2150
+    },
+    {
+      "epoch": 0.2870296236989592,
+      "grad_norm": 1.3968858626075848,
+      "learning_rate": 1.6731036119383007e-05,
+      "loss": 1.6662,
+      "step": 2151
+    },
+    {
+      "epoch": 0.2871630637843608,
+      "grad_norm": 1.0783132009504053,
+      "learning_rate": 1.6727839233907833e-05,
+      "loss": 1.62,
+      "step": 2152
+    },
+    {
+      "epoch": 0.28729650386976247,
+      "grad_norm": 1.045602618496203,
+      "learning_rate": 1.672464109174889e-05,
+      "loss": 1.6085,
+      "step": 2153
+    },
+    {
+      "epoch": 0.28742994395516414,
+      "grad_norm": 1.0204746262045936,
+      "learning_rate": 1.672144169350355e-05,
+      "loss": 1.682,
+      "step": 2154
+    },
+    {
+      "epoch": 0.2875633840405658,
+      "grad_norm": 0.9712644138016593,
+      "learning_rate": 1.671824103976943e-05,
+      "loss": 1.591,
+      "step": 2155
+    },
+    {
+      "epoch": 0.2876968241259674,
+      "grad_norm": 1.2273318949317467,
+      "learning_rate": 1.6715039131144375e-05,
+      "loss": 1.6639,
+      "step": 2156
+    },
+    {
+      "epoch": 0.2878302642113691,
+      "grad_norm": 1.241652981677857,
+      "learning_rate": 1.6711835968226463e-05,
+      "loss": 1.656,
+      "step": 2157
+    },
+    {
+      "epoch": 0.28796370429677076,
+      "grad_norm": 1.0646723228459751,
+      "learning_rate": 1.670863155161401e-05,
+      "loss": 1.6509,
+      "step": 2158
+    },
+    {
+      "epoch": 0.28809714438217243,
+      "grad_norm": 0.9640798880564916,
+      "learning_rate": 1.6705425881905564e-05,
+      "loss": 1.6043,
+      "step": 2159
+    },
+    {
+      "epoch": 0.28823058446757405,
+      "grad_norm": 1.0190920068560503,
+      "learning_rate": 1.6702218959699906e-05,
+      "loss": 1.6723,
+      "step": 2160
+    },
+    {
+      "epoch": 0.2883640245529757,
+      "grad_norm": 1.029425832894891,
+      "learning_rate": 1.6699010785596056e-05,
+      "loss": 1.6663,
+      "step": 2161
+    },
+    {
+      "epoch": 0.2884974646383774,
+      "grad_norm": 1.0069920792713327,
+      "learning_rate": 1.669580136019326e-05,
+      "loss": 1.6331,
+      "step": 2162
+    },
+    {
+      "epoch": 0.288630904723779,
+      "grad_norm": 1.1668045734582408,
+      "learning_rate": 1.6692590684091004e-05,
+      "loss": 1.6401,
+      "step": 2163
+    },
+    {
+      "epoch": 0.2887643448091807,
+      "grad_norm": 1.056267375662492,
+      "learning_rate": 1.6689378757889007e-05,
+      "loss": 1.6311,
+      "step": 2164
+    },
+    {
+      "epoch": 0.28889778489458234,
+      "grad_norm": 1.0648778315304113,
+      "learning_rate": 1.6686165582187223e-05,
+      "loss": 1.6636,
+      "step": 2165
+    },
+    {
+      "epoch": 0.289031224979984,
+      "grad_norm": 1.0124410015208933,
+      "learning_rate": 1.668295115758583e-05,
+      "loss": 1.6381,
+      "step": 2166
+    },
+    {
+      "epoch": 0.28916466506538563,
+      "grad_norm": 1.0022830955131297,
+      "learning_rate": 1.6679735484685247e-05,
+      "loss": 1.5945,
+      "step": 2167
+    },
+    {
+      "epoch": 0.2892981051507873,
+      "grad_norm": 12.676569029129164,
+      "learning_rate": 1.6676518564086133e-05,
+      "loss": 1.6591,
+      "step": 2168
+    },
+    {
+      "epoch": 0.28943154523618897,
+      "grad_norm": 1.1813552827187637,
+      "learning_rate": 1.667330039638936e-05,
+      "loss": 1.6071,
+      "step": 2169
+    },
+    {
+      "epoch": 0.2895649853215906,
+      "grad_norm": 1.059369062118976,
+      "learning_rate": 1.6670080982196055e-05,
+      "loss": 1.6634,
+      "step": 2170
+    },
+    {
+      "epoch": 0.28969842540699225,
+      "grad_norm": 12.044345487483524,
+      "learning_rate": 1.6666860322107563e-05,
+      "loss": 1.7567,
+      "step": 2171
+    },
+    {
+      "epoch": 0.2898318654923939,
+      "grad_norm": 1.1174156183857291,
+      "learning_rate": 1.6663638416725466e-05,
+      "loss": 1.6669,
+      "step": 2172
+    },
+    {
+      "epoch": 0.2899653055777956,
+      "grad_norm": 1.422061825117082,
+      "learning_rate": 1.6660415266651588e-05,
+      "loss": 1.6174,
+      "step": 2173
+    },
+    {
+      "epoch": 0.2900987456631972,
+      "grad_norm": 1.09465155740145,
+      "learning_rate": 1.6657190872487964e-05,
+      "loss": 1.6354,
+      "step": 2174
+    },
+    {
+      "epoch": 0.2902321857485989,
+      "grad_norm": 1.6546096916969328,
+      "learning_rate": 1.665396523483688e-05,
+      "loss": 1.5803,
+      "step": 2175
+    },
+    {
+      "epoch": 0.29036562583400055,
+      "grad_norm": 10.336323730521444,
+      "learning_rate": 1.6650738354300848e-05,
+      "loss": 1.6456,
+      "step": 2176
+    },
+    {
+      "epoch": 0.29049906591940217,
+      "grad_norm": 1.1978798944636024,
+      "learning_rate": 1.664751023148262e-05,
+      "loss": 1.6693,
+      "step": 2177
+    },
+    {
+      "epoch": 0.29063250600480384,
+      "grad_norm": 1.1345982910370151,
+      "learning_rate": 1.6644280866985155e-05,
+      "loss": 1.5768,
+      "step": 2178
+    },
+    {
+      "epoch": 0.2907659460902055,
+      "grad_norm": 1.3001746534071794,
+      "learning_rate": 1.6641050261411676e-05,
+      "loss": 1.6761,
+      "step": 2179
+    },
+    {
+      "epoch": 0.2908993861756072,
+      "grad_norm": 1.321671111629523,
+      "learning_rate": 1.6637818415365622e-05,
+      "loss": 1.6474,
+      "step": 2180
+    },
+    {
+      "epoch": 0.2910328262610088,
+      "grad_norm": 1.0570804882194134,
+      "learning_rate": 1.663458532945066e-05,
+      "loss": 1.6042,
+      "step": 2181
+    },
+    {
+      "epoch": 0.29116626634641046,
+      "grad_norm": 1.242601537641294,
+      "learning_rate": 1.66313510042707e-05,
+      "loss": 1.6398,
+      "step": 2182
+    },
+    {
+      "epoch": 0.29129970643181213,
+      "grad_norm": 1.3245077209972027,
+      "learning_rate": 1.6628115440429867e-05,
+      "loss": 1.643,
+      "step": 2183
+    },
+    {
+      "epoch": 0.29143314651721375,
+      "grad_norm": 1.0817940467899572,
+      "learning_rate": 1.662487863853254e-05,
+      "loss": 1.5973,
+      "step": 2184
+    },
+    {
+      "epoch": 0.2915665866026154,
+      "grad_norm": 1.0749074924656212,
+      "learning_rate": 1.6621640599183307e-05,
+      "loss": 1.6395,
+      "step": 2185
+    },
+    {
+      "epoch": 0.2917000266880171,
+      "grad_norm": 1.0699143617678875,
+      "learning_rate": 1.6618401322986998e-05,
+      "loss": 1.6294,
+      "step": 2186
+    },
+    {
+      "epoch": 0.29183346677341876,
+      "grad_norm": 1.2973422398546064,
+      "learning_rate": 1.6615160810548677e-05,
+      "loss": 1.6334,
+      "step": 2187
+    },
+    {
+      "epoch": 0.2919669068588204,
+      "grad_norm": 1.2752225836668032,
+      "learning_rate": 1.6611919062473635e-05,
+      "loss": 1.6568,
+      "step": 2188
+    },
+    {
+      "epoch": 0.29210034694422204,
+      "grad_norm": 1.0820454505520494,
+      "learning_rate": 1.660867607936739e-05,
+      "loss": 1.6779,
+      "step": 2189
+    },
+    {
+      "epoch": 0.2922337870296237,
+      "grad_norm": 1.1086323014280555,
+      "learning_rate": 1.6605431861835695e-05,
+      "loss": 1.6372,
+      "step": 2190
+    },
+    {
+      "epoch": 0.29236722711502533,
+      "grad_norm": 1.1434647125560675,
+      "learning_rate": 1.6602186410484536e-05,
+      "loss": 1.5664,
+      "step": 2191
+    },
+    {
+      "epoch": 0.292500667200427,
+      "grad_norm": 1.6902321777443527,
+      "learning_rate": 1.6598939725920122e-05,
+      "loss": 1.6695,
+      "step": 2192
+    },
+    {
+      "epoch": 0.29263410728582867,
+      "grad_norm": 1.2716880193854676,
+      "learning_rate": 1.65956918087489e-05,
+      "loss": 1.7116,
+      "step": 2193
+    },
+    {
+      "epoch": 0.29276754737123034,
+      "grad_norm": 1.1161202972917799,
+      "learning_rate": 1.6592442659577545e-05,
+      "loss": 1.6309,
+      "step": 2194
+    },
+    {
+      "epoch": 0.29290098745663196,
+      "grad_norm": 1.0699166169760714,
+      "learning_rate": 1.6589192279012956e-05,
+      "loss": 1.6484,
+      "step": 2195
+    },
+    {
+      "epoch": 0.2930344275420336,
+      "grad_norm": 1.0222347107395644,
+      "learning_rate": 1.658594066766227e-05,
+      "loss": 1.6596,
+      "step": 2196
+    },
+    {
+      "epoch": 0.2931678676274353,
+      "grad_norm": 1.0807908428296922,
+      "learning_rate": 1.6582687826132854e-05,
+      "loss": 1.6678,
+      "step": 2197
+    },
+    {
+      "epoch": 0.2933013077128369,
+      "grad_norm": 1.081068785656472,
+      "learning_rate": 1.6579433755032297e-05,
+      "loss": 1.6052,
+      "step": 2198
+    },
+    {
+      "epoch": 0.2934347477982386,
+      "grad_norm": 1.0103124427107717,
+      "learning_rate": 1.6576178454968422e-05,
+      "loss": 1.6409,
+      "step": 2199
+    },
+    {
+      "epoch": 0.29356818788364025,
+      "grad_norm": 0.9811350388342566,
+      "learning_rate": 1.657292192654929e-05,
+      "loss": 1.5593,
+      "step": 2200
+    },
+    {
+      "epoch": 0.2937016279690419,
+      "grad_norm": 1.306039871025814,
+      "learning_rate": 1.656966417038317e-05,
+      "loss": 1.6302,
+      "step": 2201
+    },
+    {
+      "epoch": 0.29383506805444354,
+      "grad_norm": 1.021756773048609,
+      "learning_rate": 1.656640518707859e-05,
+      "loss": 1.6938,
+      "step": 2202
+    },
+    {
+      "epoch": 0.2939685081398452,
+      "grad_norm": 1.0718659276893743,
+      "learning_rate": 1.6563144977244277e-05,
+      "loss": 1.6301,
+      "step": 2203
+    },
+    {
+      "epoch": 0.2941019482252469,
+      "grad_norm": 1.2325944349438038,
+      "learning_rate": 1.6559883541489204e-05,
+      "loss": 1.6064,
+      "step": 2204
+    },
+    {
+      "epoch": 0.2942353883106485,
+      "grad_norm": 1.04367146349336,
+      "learning_rate": 1.6556620880422577e-05,
+      "loss": 1.6461,
+      "step": 2205
+    },
+    {
+      "epoch": 0.29436882839605016,
+      "grad_norm": 1.1064673095517306,
+      "learning_rate": 1.6553356994653818e-05,
+      "loss": 1.6265,
+      "step": 2206
+    },
+    {
+      "epoch": 0.29450226848145183,
+      "grad_norm": 1.0133174183375973,
+      "learning_rate": 1.6550091884792586e-05,
+      "loss": 1.647,
+      "step": 2207
+    },
+    {
+      "epoch": 0.2946357085668535,
+      "grad_norm": 1.0556625341442718,
+      "learning_rate": 1.6546825551448766e-05,
+      "loss": 1.6508,
+      "step": 2208
+    },
+    {
+      "epoch": 0.2947691486522551,
+      "grad_norm": 1.0201247011199586,
+      "learning_rate": 1.6543557995232472e-05,
+      "loss": 1.6311,
+      "step": 2209
+    },
+    {
+      "epoch": 0.2949025887376568,
+      "grad_norm": 1.0341845337571378,
+      "learning_rate": 1.6540289216754042e-05,
+      "loss": 1.6522,
+      "step": 2210
+    },
+    {
+      "epoch": 0.29503602882305846,
+      "grad_norm": 1.0728331970850973,
+      "learning_rate": 1.6537019216624055e-05,
+      "loss": 1.6073,
+      "step": 2211
+    },
+    {
+      "epoch": 0.2951694689084601,
+      "grad_norm": 1.087051064593804,
+      "learning_rate": 1.65337479954533e-05,
+      "loss": 1.6518,
+      "step": 2212
+    },
+    {
+      "epoch": 0.29530290899386175,
+      "grad_norm": 1.1504157623012197,
+      "learning_rate": 1.653047555385281e-05,
+      "loss": 1.6489,
+      "step": 2213
+    },
+    {
+      "epoch": 0.2954363490792634,
+      "grad_norm": 1.6929487032356967,
+      "learning_rate": 1.652720189243384e-05,
+      "loss": 1.6564,
+      "step": 2214
+    },
+    {
+      "epoch": 0.2955697891646651,
+      "grad_norm": 9.547350526694943,
+      "learning_rate": 1.6523927011807873e-05,
+      "loss": 1.6765,
+      "step": 2215
+    },
+    {
+      "epoch": 0.2957032292500667,
+      "grad_norm": 1.166505303552692,
+      "learning_rate": 1.652065091258662e-05,
+      "loss": 1.6389,
+      "step": 2216
+    },
+    {
+      "epoch": 0.29583666933546837,
+      "grad_norm": 1.1611750428510519,
+      "learning_rate": 1.651737359538201e-05,
+      "loss": 1.6382,
+      "step": 2217
+    },
+    {
+      "epoch": 0.29597010942087004,
+      "grad_norm": 1.161944383190572,
+      "learning_rate": 1.651409506080622e-05,
+      "loss": 1.6458,
+      "step": 2218
+    },
+    {
+      "epoch": 0.29610354950627166,
+      "grad_norm": 1.0328377880475517,
+      "learning_rate": 1.6510815309471638e-05,
+      "loss": 1.6698,
+      "step": 2219
+    },
+    {
+      "epoch": 0.2962369895916733,
+      "grad_norm": 1.2670883726284181,
+      "learning_rate": 1.6507534341990884e-05,
+      "loss": 1.6669,
+      "step": 2220
+    },
+    {
+      "epoch": 0.296370429677075,
+      "grad_norm": 1.2369384329163466,
+      "learning_rate": 1.6504252158976804e-05,
+      "loss": 1.6523,
+      "step": 2221
+    },
+    {
+      "epoch": 0.29650386976247667,
+      "grad_norm": 1.0226119191785725,
+      "learning_rate": 1.6500968761042477e-05,
+      "loss": 1.6438,
+      "step": 2222
+    },
+    {
+      "epoch": 0.2966373098478783,
+      "grad_norm": 0.9968700209514012,
+      "learning_rate": 1.6497684148801204e-05,
+      "loss": 1.7007,
+      "step": 2223
+    },
+    {
+      "epoch": 0.29677074993327995,
+      "grad_norm": 1.036350320781629,
+      "learning_rate": 1.6494398322866503e-05,
+      "loss": 1.6678,
+      "step": 2224
+    },
+    {
+      "epoch": 0.2969041900186816,
+      "grad_norm": 0.9847140373170601,
+      "learning_rate": 1.6491111283852147e-05,
+      "loss": 1.6433,
+      "step": 2225
+    },
+    {
+      "epoch": 0.29703763010408324,
+      "grad_norm": 0.9415756130495302,
+      "learning_rate": 1.64878230323721e-05,
+      "loss": 1.6222,
+      "step": 2226
+    },
+    {
+      "epoch": 0.2971710701894849,
+      "grad_norm": 1.4397026580916885,
+      "learning_rate": 1.648453356904058e-05,
+      "loss": 1.777,
+      "step": 2227
+    },
+    {
+      "epoch": 0.2973045102748866,
+      "grad_norm": 1.1618462036671973,
+      "learning_rate": 1.6481242894472016e-05,
+      "loss": 1.6368,
+      "step": 2228
+    },
+    {
+      "epoch": 0.29743795036028825,
+      "grad_norm": 1.0250435345601665,
+      "learning_rate": 1.6477951009281072e-05,
+      "loss": 1.6461,
+      "step": 2229
+    },
+    {
+      "epoch": 0.29757139044568987,
+      "grad_norm": 0.9923678067336387,
+      "learning_rate": 1.6474657914082638e-05,
+      "loss": 1.6785,
+      "step": 2230
+    },
+    {
+      "epoch": 0.29770483053109154,
+      "grad_norm": 1.0267804282384663,
+      "learning_rate": 1.647136360949182e-05,
+      "loss": 1.6343,
+      "step": 2231
+    },
+    {
+      "epoch": 0.2978382706164932,
+      "grad_norm": 1.023674327352044,
+      "learning_rate": 1.6468068096123957e-05,
+      "loss": 1.5966,
+      "step": 2232
+    },
+    {
+      "epoch": 0.2979717107018949,
+      "grad_norm": 1.1944718768952494,
+      "learning_rate": 1.6464771374594615e-05,
+      "loss": 1.6535,
+      "step": 2233
+    },
+    {
+      "epoch": 0.2981051507872965,
+      "grad_norm": 1.152852164366958,
+      "learning_rate": 1.6461473445519585e-05,
+      "loss": 1.6781,
+      "step": 2234
+    },
+    {
+      "epoch": 0.29823859087269816,
+      "grad_norm": 1.0678201700556582,
+      "learning_rate": 1.6458174309514882e-05,
+      "loss": 1.6313,
+      "step": 2235
+    },
+    {
+      "epoch": 0.29837203095809983,
+      "grad_norm": 1.0355313650728948,
+      "learning_rate": 1.6454873967196748e-05,
+      "loss": 1.6683,
+      "step": 2236
+    },
+    {
+      "epoch": 0.29850547104350145,
+      "grad_norm": 1.0884233154304546,
+      "learning_rate": 1.645157241918165e-05,
+      "loss": 1.5859,
+      "step": 2237
+    },
+    {
+      "epoch": 0.2986389111289031,
+      "grad_norm": 1.0472190401927721,
+      "learning_rate": 1.6448269666086278e-05,
+      "loss": 1.6596,
+      "step": 2238
+    },
+    {
+      "epoch": 0.2987723512143048,
+      "grad_norm": 1.0311680152934846,
+      "learning_rate": 1.6444965708527546e-05,
+      "loss": 1.6525,
+      "step": 2239
+    },
+    {
+      "epoch": 0.29890579129970646,
+      "grad_norm": 2.10561403694868,
+      "learning_rate": 1.64416605471226e-05,
+      "loss": 1.6469,
+      "step": 2240
+    },
+    {
+      "epoch": 0.2990392313851081,
+      "grad_norm": 1.0497996204276114,
+      "learning_rate": 1.6438354182488802e-05,
+      "loss": 1.607,
+      "step": 2241
+    },
+    {
+      "epoch": 0.29917267147050974,
+      "grad_norm": 1.043507533207965,
+      "learning_rate": 1.6435046615243747e-05,
+      "loss": 1.6621,
+      "step": 2242
+    },
+    {
+      "epoch": 0.2993061115559114,
+      "grad_norm": 1.042723485887548,
+      "learning_rate": 1.6431737846005254e-05,
+      "loss": 1.6912,
+      "step": 2243
+    },
+    {
+      "epoch": 0.29943955164131303,
+      "grad_norm": 1.0051152191792925,
+      "learning_rate": 1.6428427875391353e-05,
+      "loss": 1.6349,
+      "step": 2244
+    },
+    {
+      "epoch": 0.2995729917267147,
+      "grad_norm": 1.128346274800114,
+      "learning_rate": 1.642511670402032e-05,
+      "loss": 1.7053,
+      "step": 2245
+    },
+    {
+      "epoch": 0.29970643181211637,
+      "grad_norm": 1.0722777257603044,
+      "learning_rate": 1.642180433251064e-05,
+      "loss": 1.6878,
+      "step": 2246
+    },
+    {
+      "epoch": 0.29983987189751804,
+      "grad_norm": 1.017261398089863,
+      "learning_rate": 1.6418490761481022e-05,
+      "loss": 1.6521,
+      "step": 2247
+    },
+    {
+      "epoch": 0.29997331198291965,
+      "grad_norm": 1.0292260761586245,
+      "learning_rate": 1.6415175991550412e-05,
+      "loss": 1.6755,
+      "step": 2248
+    },
+    {
+      "epoch": 0.3001067520683213,
+      "grad_norm": 1.0931207786222135,
+      "learning_rate": 1.6411860023337962e-05,
+      "loss": 1.6749,
+      "step": 2249
+    },
+    {
+      "epoch": 0.300240192153723,
+      "grad_norm": 1.0831524536934993,
+      "learning_rate": 1.6408542857463062e-05,
+      "loss": 1.587,
+      "step": 2250
+    },
+    {
+      "epoch": 0.3003736322391246,
+      "grad_norm": 1.0454308838172524,
+      "learning_rate": 1.6405224494545322e-05,
+      "loss": 1.6651,
+      "step": 2251
+    },
+    {
+      "epoch": 0.3005070723245263,
+      "grad_norm": 1.0806258557902406,
+      "learning_rate": 1.640190493520457e-05,
+      "loss": 1.6699,
+      "step": 2252
+    },
+    {
+      "epoch": 0.30064051240992795,
+      "grad_norm": 12.400419377894314,
+      "learning_rate": 1.6398584180060867e-05,
+      "loss": 1.6377,
+      "step": 2253
+    },
+    {
+      "epoch": 0.3007739524953296,
+      "grad_norm": 1.2948113785574655,
+      "learning_rate": 1.6395262229734486e-05,
+      "loss": 1.607,
+      "step": 2254
+    },
+    {
+      "epoch": 0.30090739258073124,
+      "grad_norm": 1.198607665528997,
+      "learning_rate": 1.639193908484594e-05,
+      "loss": 1.7028,
+      "step": 2255
+    },
+    {
+      "epoch": 0.3010408326661329,
+      "grad_norm": 1.1317395785557423,
+      "learning_rate": 1.6388614746015946e-05,
+      "loss": 1.6525,
+      "step": 2256
+    },
+    {
+      "epoch": 0.3011742727515346,
+      "grad_norm": 1.1476872660243271,
+      "learning_rate": 1.6385289213865452e-05,
+      "loss": 1.64,
+      "step": 2257
+    },
+    {
+      "epoch": 0.3013077128369362,
+      "grad_norm": 1.480875813813185,
+      "learning_rate": 1.6381962489015633e-05,
+      "loss": 1.6505,
+      "step": 2258
+    },
+    {
+      "epoch": 0.30144115292233786,
+      "grad_norm": 1.0975918871872568,
+      "learning_rate": 1.637863457208788e-05,
+      "loss": 1.6105,
+      "step": 2259
+    },
+    {
+      "epoch": 0.30157459300773953,
+      "grad_norm": 1.147793532979415,
+      "learning_rate": 1.6375305463703816e-05,
+      "loss": 1.6267,
+      "step": 2260
+    },
+    {
+      "epoch": 0.3017080330931412,
+      "grad_norm": 1.2879635429689233,
+      "learning_rate": 1.6371975164485277e-05,
+      "loss": 1.7015,
+      "step": 2261
+    },
+    {
+      "epoch": 0.3018414731785428,
+      "grad_norm": 1.3283545115204671,
+      "learning_rate": 1.636864367505432e-05,
+      "loss": 1.6196,
+      "step": 2262
+    },
+    {
+      "epoch": 0.3019749132639445,
+      "grad_norm": 1.0479344997197932,
+      "learning_rate": 1.636531099603324e-05,
+      "loss": 1.6243,
+      "step": 2263
+    },
+    {
+      "epoch": 0.30210835334934616,
+      "grad_norm": 1.0765762306122926,
+      "learning_rate": 1.6361977128044535e-05,
+      "loss": 1.6196,
+      "step": 2264
+    },
+    {
+      "epoch": 0.3022417934347478,
+      "grad_norm": 1.0309118980744778,
+      "learning_rate": 1.6358642071710935e-05,
+      "loss": 1.6135,
+      "step": 2265
+    },
+    {
+      "epoch": 0.30237523352014944,
+      "grad_norm": 0.9953925772810231,
+      "learning_rate": 1.6355305827655398e-05,
+      "loss": 1.6429,
+      "step": 2266
+    },
+    {
+      "epoch": 0.3025086736055511,
+      "grad_norm": 1.2130624721172607,
+      "learning_rate": 1.6351968396501084e-05,
+      "loss": 1.6343,
+      "step": 2267
+    },
+    {
+      "epoch": 0.3026421136909528,
+      "grad_norm": 1.0691255501843615,
+      "learning_rate": 1.6348629778871393e-05,
+      "loss": 1.6658,
+      "step": 2268
+    },
+    {
+      "epoch": 0.3027755537763544,
+      "grad_norm": 1.0154122944841089,
+      "learning_rate": 1.6345289975389946e-05,
+      "loss": 1.6048,
+      "step": 2269
+    },
+    {
+      "epoch": 0.30290899386175607,
+      "grad_norm": 1.0303963960703182,
+      "learning_rate": 1.6341948986680574e-05,
+      "loss": 1.5902,
+      "step": 2270
+    },
+    {
+      "epoch": 0.30304243394715774,
+      "grad_norm": 1.103571288321129,
+      "learning_rate": 1.6338606813367334e-05,
+      "loss": 1.6284,
+      "step": 2271
+    },
+    {
+      "epoch": 0.30317587403255936,
+      "grad_norm": 1.0340663960820138,
+      "learning_rate": 1.633526345607451e-05,
+      "loss": 1.6625,
+      "step": 2272
+    },
+    {
+      "epoch": 0.303309314117961,
+      "grad_norm": 1.0868515192551496,
+      "learning_rate": 1.6331918915426607e-05,
+      "loss": 1.6508,
+      "step": 2273
+    },
+    {
+      "epoch": 0.3034427542033627,
+      "grad_norm": 1.0806584970772868,
+      "learning_rate": 1.6328573192048337e-05,
+      "loss": 1.6793,
+      "step": 2274
+    },
+    {
+      "epoch": 0.30357619428876437,
+      "grad_norm": 1.1131107600277885,
+      "learning_rate": 1.632522628656465e-05,
+      "loss": 1.6524,
+      "step": 2275
+    },
+    {
+      "epoch": 0.303709634374166,
+      "grad_norm": 0.970695670852881,
+      "learning_rate": 1.6321878199600705e-05,
+      "loss": 1.6254,
+      "step": 2276
+    },
+    {
+      "epoch": 0.30384307445956765,
+      "grad_norm": 1.0449698885856158,
+      "learning_rate": 1.6318528931781893e-05,
+      "loss": 1.6488,
+      "step": 2277
+    },
+    {
+      "epoch": 0.3039765145449693,
+      "grad_norm": 2.3976140688143754,
+      "learning_rate": 1.6315178483733817e-05,
+      "loss": 1.6349,
+      "step": 2278
+    },
+    {
+      "epoch": 0.30410995463037094,
+      "grad_norm": 1.0721815977781701,
+      "learning_rate": 1.6311826856082298e-05,
+      "loss": 1.6444,
+      "step": 2279
+    },
+    {
+      "epoch": 0.3042433947157726,
+      "grad_norm": 1.0877886525914027,
+      "learning_rate": 1.6308474049453385e-05,
+      "loss": 1.6218,
+      "step": 2280
+    },
+    {
+      "epoch": 0.3043768348011743,
+      "grad_norm": 1.2215891275833313,
+      "learning_rate": 1.6305120064473348e-05,
+      "loss": 1.6729,
+      "step": 2281
+    },
+    {
+      "epoch": 0.30451027488657595,
+      "grad_norm": 1.0066658286438341,
+      "learning_rate": 1.6301764901768665e-05,
+      "loss": 1.6514,
+      "step": 2282
+    },
+    {
+      "epoch": 0.30464371497197756,
+      "grad_norm": 5.18175722987443,
+      "learning_rate": 1.629840856196605e-05,
+      "loss": 1.7849,
+      "step": 2283
+    },
+    {
+      "epoch": 0.30477715505737923,
+      "grad_norm": 1.0661836522987367,
+      "learning_rate": 1.6295051045692422e-05,
+      "loss": 1.6515,
+      "step": 2284
+    },
+    {
+      "epoch": 0.3049105951427809,
+      "grad_norm": 0.9948954364094378,
+      "learning_rate": 1.6291692353574934e-05,
+      "loss": 1.6653,
+      "step": 2285
+    },
+    {
+      "epoch": 0.3050440352281825,
+      "grad_norm": 1.1776280843724025,
+      "learning_rate": 1.6288332486240946e-05,
+      "loss": 1.6249,
+      "step": 2286
+    },
+    {
+      "epoch": 0.3051774753135842,
+      "grad_norm": 0.9803581714490958,
+      "learning_rate": 1.6284971444318046e-05,
+      "loss": 1.6381,
+      "step": 2287
+    },
+    {
+      "epoch": 0.30531091539898586,
+      "grad_norm": 1.0334824917780492,
+      "learning_rate": 1.628160922843404e-05,
+      "loss": 1.7033,
+      "step": 2288
+    },
+    {
+      "epoch": 0.30544435548438753,
+      "grad_norm": 1.1825464876579848,
+      "learning_rate": 1.6278245839216947e-05,
+      "loss": 1.6364,
+      "step": 2289
+    },
+    {
+      "epoch": 0.30557779556978915,
+      "grad_norm": 1.089932123868756,
+      "learning_rate": 1.627488127729501e-05,
+      "loss": 1.6358,
+      "step": 2290
+    },
+    {
+      "epoch": 0.3057112356551908,
+      "grad_norm": 1.0063881587721086,
+      "learning_rate": 1.6271515543296693e-05,
+      "loss": 1.6245,
+      "step": 2291
+    },
+    {
+      "epoch": 0.3058446757405925,
+      "grad_norm": 0.9557576711140595,
+      "learning_rate": 1.6268148637850675e-05,
+      "loss": 1.6148,
+      "step": 2292
+    },
+    {
+      "epoch": 0.3059781158259941,
+      "grad_norm": 1.155386941840756,
+      "learning_rate": 1.626478056158586e-05,
+      "loss": 1.6679,
+      "step": 2293
+    },
+    {
+      "epoch": 0.30611155591139577,
+      "grad_norm": 1.002724278928917,
+      "learning_rate": 1.626141131513136e-05,
+      "loss": 1.6717,
+      "step": 2294
+    },
+    {
+      "epoch": 0.30624499599679744,
+      "grad_norm": 1.0181937079803456,
+      "learning_rate": 1.6258040899116517e-05,
+      "loss": 1.6276,
+      "step": 2295
+    },
+    {
+      "epoch": 0.3063784360821991,
+      "grad_norm": 1.0190676733905637,
+      "learning_rate": 1.6254669314170887e-05,
+      "loss": 1.6428,
+      "step": 2296
+    },
+    {
+      "epoch": 0.3065118761676007,
+      "grad_norm": 1.0126611742120983,
+      "learning_rate": 1.625129656092424e-05,
+      "loss": 1.6943,
+      "step": 2297
+    },
+    {
+      "epoch": 0.3066453162530024,
+      "grad_norm": 0.9702747048248926,
+      "learning_rate": 1.624792264000657e-05,
+      "loss": 1.6091,
+      "step": 2298
+    },
+    {
+      "epoch": 0.30677875633840407,
+      "grad_norm": 1.0176969387216936,
+      "learning_rate": 1.624454755204808e-05,
+      "loss": 1.6432,
+      "step": 2299
+    },
+    {
+      "epoch": 0.30691219642380574,
+      "grad_norm": 0.9905054594963236,
+      "learning_rate": 1.624117129767921e-05,
+      "loss": 1.6908,
+      "step": 2300
+    },
+    {
+      "epoch": 0.30704563650920735,
+      "grad_norm": 1.0990745253836063,
+      "learning_rate": 1.62377938775306e-05,
+      "loss": 1.6545,
+      "step": 2301
+    },
+    {
+      "epoch": 0.307179076594609,
+      "grad_norm": 0.9872989722764705,
+      "learning_rate": 1.623441529223311e-05,
+      "loss": 1.6699,
+      "step": 2302
+    },
+    {
+      "epoch": 0.3073125166800107,
+      "grad_norm": 1.2705211460028154,
+      "learning_rate": 1.6231035542417826e-05,
+      "loss": 1.6582,
+      "step": 2303
+    },
+    {
+      "epoch": 0.3074459567654123,
+      "grad_norm": 1.0150840493548405,
+      "learning_rate": 1.6227654628716044e-05,
+      "loss": 1.665,
+      "step": 2304
+    },
+    {
+      "epoch": 0.307579396850814,
+      "grad_norm": 1.1379930735351889,
+      "learning_rate": 1.6224272551759288e-05,
+      "loss": 1.676,
+      "step": 2305
+    },
+    {
+      "epoch": 0.30771283693621565,
+      "grad_norm": 1.2444071746179561,
+      "learning_rate": 1.622088931217928e-05,
+      "loss": 1.6459,
+      "step": 2306
+    },
+    {
+      "epoch": 0.3078462770216173,
+      "grad_norm": 0.9976784854166414,
+      "learning_rate": 1.621750491060798e-05,
+      "loss": 1.6636,
+      "step": 2307
+    },
+    {
+      "epoch": 0.30797971710701894,
+      "grad_norm": 1.2682449189830896,
+      "learning_rate": 1.6214119347677548e-05,
+      "loss": 1.6616,
+      "step": 2308
+    },
+    {
+      "epoch": 0.3081131571924206,
+      "grad_norm": 1.0829225349432583,
+      "learning_rate": 1.621073262402037e-05,
+      "loss": 1.644,
+      "step": 2309
+    },
+    {
+      "epoch": 0.3082465972778223,
+      "grad_norm": 1.0838885803573566,
+      "learning_rate": 1.6207344740269056e-05,
+      "loss": 1.6494,
+      "step": 2310
+    },
+    {
+      "epoch": 0.3083800373632239,
+      "grad_norm": 1.0708462365866025,
+      "learning_rate": 1.6203955697056416e-05,
+      "loss": 1.6671,
+      "step": 2311
+    },
+    {
+      "epoch": 0.30851347744862556,
+      "grad_norm": 0.9823970370572715,
+      "learning_rate": 1.6200565495015487e-05,
+      "loss": 1.6396,
+      "step": 2312
+    },
+    {
+      "epoch": 0.30864691753402723,
+      "grad_norm": 1.0031602532602553,
+      "learning_rate": 1.6197174134779517e-05,
+      "loss": 1.6102,
+      "step": 2313
+    },
+    {
+      "epoch": 0.3087803576194289,
+      "grad_norm": 1.2206609039872867,
+      "learning_rate": 1.619378161698198e-05,
+      "loss": 1.648,
+      "step": 2314
+    },
+    {
+      "epoch": 0.3089137977048305,
+      "grad_norm": 1.3166269143643101,
+      "learning_rate": 1.6190387942256548e-05,
+      "loss": 1.6009,
+      "step": 2315
+    },
+    {
+      "epoch": 0.3090472377902322,
+      "grad_norm": 1.2675618240810085,
+      "learning_rate": 1.6186993111237134e-05,
+      "loss": 1.6224,
+      "step": 2316
+    },
+    {
+      "epoch": 0.30918067787563386,
+      "grad_norm": 1.0640732745150545,
+      "learning_rate": 1.618359712455785e-05,
+      "loss": 1.6662,
+      "step": 2317
+    },
+    {
+      "epoch": 0.3093141179610355,
+      "grad_norm": 1.019032967258582,
+      "learning_rate": 1.6180199982853026e-05,
+      "loss": 1.6237,
+      "step": 2318
+    },
+    {
+      "epoch": 0.30944755804643714,
+      "grad_norm": 1.0944634109792915,
+      "learning_rate": 1.6176801686757207e-05,
+      "loss": 1.6415,
+      "step": 2319
+    },
+    {
+      "epoch": 0.3095809981318388,
+      "grad_norm": 1.24072168667135,
+      "learning_rate": 1.6173402236905156e-05,
+      "loss": 1.5915,
+      "step": 2320
+    },
+    {
+      "epoch": 0.3097144382172405,
+      "grad_norm": 0.9746140716316298,
+      "learning_rate": 1.6170001633931857e-05,
+      "loss": 1.6443,
+      "step": 2321
+    },
+    {
+      "epoch": 0.3098478783026421,
+      "grad_norm": 1.1844321844657195,
+      "learning_rate": 1.6166599878472502e-05,
+      "loss": 1.6174,
+      "step": 2322
+    },
+    {
+      "epoch": 0.30998131838804377,
+      "grad_norm": 1.001876907977257,
+      "learning_rate": 1.61631969711625e-05,
+      "loss": 1.6162,
+      "step": 2323
+    },
+    {
+      "epoch": 0.31011475847344544,
+      "grad_norm": 0.9886662045991245,
+      "learning_rate": 1.6159792912637467e-05,
+      "loss": 1.6354,
+      "step": 2324
+    },
+    {
+      "epoch": 0.31024819855884705,
+      "grad_norm": 0.9775352505641491,
+      "learning_rate": 1.6156387703533258e-05,
+      "loss": 1.6204,
+      "step": 2325
+    },
+    {
+      "epoch": 0.3103816386442487,
+      "grad_norm": 0.9826413139650488,
+      "learning_rate": 1.6152981344485915e-05,
+      "loss": 1.6398,
+      "step": 2326
+    },
+    {
+      "epoch": 0.3105150787296504,
+      "grad_norm": 1.0049022279607043,
+      "learning_rate": 1.6149573836131713e-05,
+      "loss": 1.6421,
+      "step": 2327
+    },
+    {
+      "epoch": 0.31064851881505207,
+      "grad_norm": 1.151819306544648,
+      "learning_rate": 1.6146165179107135e-05,
+      "loss": 1.5993,
+      "step": 2328
+    },
+    {
+      "epoch": 0.3107819589004537,
+      "grad_norm": 1.144986255564853,
+      "learning_rate": 1.6142755374048876e-05,
+      "loss": 1.6405,
+      "step": 2329
+    },
+    {
+      "epoch": 0.31091539898585535,
+      "grad_norm": 1.0389802092801697,
+      "learning_rate": 1.613934442159385e-05,
+      "loss": 1.6334,
+      "step": 2330
+    },
+    {
+      "epoch": 0.311048839071257,
+      "grad_norm": 1.016003258094143,
+      "learning_rate": 1.6135932322379188e-05,
+      "loss": 1.6236,
+      "step": 2331
+    },
+    {
+      "epoch": 0.31118227915665864,
+      "grad_norm": 0.966534466876772,
+      "learning_rate": 1.6132519077042224e-05,
+      "loss": 1.6443,
+      "step": 2332
+    },
+    {
+      "epoch": 0.3113157192420603,
+      "grad_norm": 1.0190490159617054,
+      "learning_rate": 1.6129104686220522e-05,
+      "loss": 1.6327,
+      "step": 2333
+    },
+    {
+      "epoch": 0.311449159327462,
+      "grad_norm": 0.9508676466180186,
+      "learning_rate": 1.6125689150551846e-05,
+      "loss": 1.5877,
+      "step": 2334
+    },
+    {
+      "epoch": 0.31158259941286365,
+      "grad_norm": 1.0322824165183992,
+      "learning_rate": 1.612227247067418e-05,
+      "loss": 1.6722,
+      "step": 2335
+    },
+    {
+      "epoch": 0.31171603949826526,
+      "grad_norm": 1.1605340980731977,
+      "learning_rate": 1.6118854647225722e-05,
+      "loss": 1.6116,
+      "step": 2336
+    },
+    {
+      "epoch": 0.31184947958366693,
+      "grad_norm": 1.0040963152787474,
+      "learning_rate": 1.611543568084488e-05,
+      "loss": 1.6418,
+      "step": 2337
+    },
+    {
+      "epoch": 0.3119829196690686,
+      "grad_norm": 1.0426733287211412,
+      "learning_rate": 1.6112015572170283e-05,
+      "loss": 1.6439,
+      "step": 2338
+    },
+    {
+      "epoch": 0.3121163597544702,
+      "grad_norm": 1.0528462917004136,
+      "learning_rate": 1.6108594321840763e-05,
+      "loss": 1.6341,
+      "step": 2339
+    },
+    {
+      "epoch": 0.3122497998398719,
+      "grad_norm": 1.005308410868297,
+      "learning_rate": 1.6105171930495373e-05,
+      "loss": 1.6207,
+      "step": 2340
+    },
+    {
+      "epoch": 0.31238323992527356,
+      "grad_norm": 1.0757636731132663,
+      "learning_rate": 1.6101748398773374e-05,
+      "loss": 1.651,
+      "step": 2341
+    },
+    {
+      "epoch": 0.31251668001067523,
+      "grad_norm": 1.157330419763036,
+      "learning_rate": 1.609832372731425e-05,
+      "loss": 1.5569,
+      "step": 2342
+    },
+    {
+      "epoch": 0.31265012009607684,
+      "grad_norm": 1.1655530310462219,
+      "learning_rate": 1.609489791675768e-05,
+      "loss": 1.622,
+      "step": 2343
+    },
+    {
+      "epoch": 0.3127835601814785,
+      "grad_norm": 1.0163847934254429,
+      "learning_rate": 1.6091470967743577e-05,
+      "loss": 1.6283,
+      "step": 2344
+    },
+    {
+      "epoch": 0.3129170002668802,
+      "grad_norm": 0.9913652557421357,
+      "learning_rate": 1.608804288091205e-05,
+      "loss": 1.6029,
+      "step": 2345
+    },
+    {
+      "epoch": 0.3130504403522818,
+      "grad_norm": 1.0250978527017118,
+      "learning_rate": 1.6084613656903426e-05,
+      "loss": 1.6701,
+      "step": 2346
+    },
+    {
+      "epoch": 0.31318388043768347,
+      "grad_norm": 1.0036453857759162,
+      "learning_rate": 1.6081183296358246e-05,
+      "loss": 1.6656,
+      "step": 2347
+    },
+    {
+      "epoch": 0.31331732052308514,
+      "grad_norm": 1.0093725549829087,
+      "learning_rate": 1.607775179991726e-05,
+      "loss": 1.62,
+      "step": 2348
+    },
+    {
+      "epoch": 0.3134507606084868,
+      "grad_norm": 1.0549793980242639,
+      "learning_rate": 1.6074319168221446e-05,
+      "loss": 1.6474,
+      "step": 2349
+    },
+    {
+      "epoch": 0.3135842006938884,
+      "grad_norm": 1.0014461970290505,
+      "learning_rate": 1.6070885401911964e-05,
+      "loss": 1.6402,
+      "step": 2350
+    },
+    {
+      "epoch": 0.3137176407792901,
+      "grad_norm": 1.2302428695760383,
+      "learning_rate": 1.606745050163021e-05,
+      "loss": 1.6615,
+      "step": 2351
+    },
+    {
+      "epoch": 0.31385108086469177,
+      "grad_norm": 1.278062546442448,
+      "learning_rate": 1.6064014468017778e-05,
+      "loss": 1.6443,
+      "step": 2352
+    },
+    {
+      "epoch": 0.3139845209500934,
+      "grad_norm": 1.014494022872304,
+      "learning_rate": 1.606057730171649e-05,
+      "loss": 1.6771,
+      "step": 2353
+    },
+    {
+      "epoch": 0.31411796103549505,
+      "grad_norm": 0.9992975913870881,
+      "learning_rate": 1.6057139003368365e-05,
+      "loss": 1.6673,
+      "step": 2354
+    },
+    {
+      "epoch": 0.3142514011208967,
+      "grad_norm": 1.0264389162628824,
+      "learning_rate": 1.6053699573615634e-05,
+      "loss": 1.645,
+      "step": 2355
+    },
+    {
+      "epoch": 0.3143848412062984,
+      "grad_norm": 1.2998311967872789,
+      "learning_rate": 1.605025901310075e-05,
+      "loss": 1.653,
+      "step": 2356
+    },
+    {
+      "epoch": 0.3145182812917,
+      "grad_norm": 1.1129013572248108,
+      "learning_rate": 1.604681732246636e-05,
+      "loss": 1.5694,
+      "step": 2357
+    },
+    {
+      "epoch": 0.3146517213771017,
+      "grad_norm": 0.998120908236157,
+      "learning_rate": 1.6043374502355348e-05,
+      "loss": 1.6577,
+      "step": 2358
+    },
+    {
+      "epoch": 0.31478516146250335,
+      "grad_norm": 1.1224918644835182,
+      "learning_rate": 1.603993055341078e-05,
+      "loss": 1.6469,
+      "step": 2359
+    },
+    {
+      "epoch": 0.31491860154790496,
+      "grad_norm": 1.0269545547781742,
+      "learning_rate": 1.6036485476275957e-05,
+      "loss": 1.6068,
+      "step": 2360
+    },
+    {
+      "epoch": 0.31505204163330663,
+      "grad_norm": 1.136514585808053,
+      "learning_rate": 1.603303927159437e-05,
+      "loss": 1.6382,
+      "step": 2361
+    },
+    {
+      "epoch": 0.3151854817187083,
+      "grad_norm": 1.2494184496056624,
+      "learning_rate": 1.602959194000974e-05,
+      "loss": 1.7152,
+      "step": 2362
+    },
+    {
+      "epoch": 0.31531892180411,
+      "grad_norm": 1.0140755418750587,
+      "learning_rate": 1.602614348216598e-05,
+      "loss": 1.591,
+      "step": 2363
+    },
+    {
+      "epoch": 0.3154523618895116,
+      "grad_norm": 1.0007366397635697,
+      "learning_rate": 1.602269389870723e-05,
+      "loss": 1.6349,
+      "step": 2364
+    },
+    {
+      "epoch": 0.31558580197491326,
+      "grad_norm": 1.0379624969316805,
+      "learning_rate": 1.6019243190277832e-05,
+      "loss": 1.583,
+      "step": 2365
+    },
+    {
+      "epoch": 0.31571924206031493,
+      "grad_norm": 1.122118459336519,
+      "learning_rate": 1.6015791357522336e-05,
+      "loss": 1.5789,
+      "step": 2366
+    },
+    {
+      "epoch": 0.31585268214571655,
+      "grad_norm": 1.218518313111331,
+      "learning_rate": 1.601233840108551e-05,
+      "loss": 1.6483,
+      "step": 2367
+    },
+    {
+      "epoch": 0.3159861222311182,
+      "grad_norm": 1.0325320279857855,
+      "learning_rate": 1.600888432161232e-05,
+      "loss": 1.6786,
+      "step": 2368
+    },
+    {
+      "epoch": 0.3161195623165199,
+      "grad_norm": 1.0105488940747966,
+      "learning_rate": 1.600542911974796e-05,
+      "loss": 1.6214,
+      "step": 2369
+    },
+    {
+      "epoch": 0.31625300240192156,
+      "grad_norm": 1.0174955025149985,
+      "learning_rate": 1.6001972796137804e-05,
+      "loss": 1.6188,
+      "step": 2370
+    },
+    {
+      "epoch": 0.31638644248732317,
+      "grad_norm": 0.9686366099778118,
+      "learning_rate": 1.5998515351427472e-05,
+      "loss": 1.6052,
+      "step": 2371
+    },
+    {
+      "epoch": 0.31651988257272484,
+      "grad_norm": 1.0943501073343973,
+      "learning_rate": 1.5995056786262763e-05,
+      "loss": 1.6652,
+      "step": 2372
+    },
+    {
+      "epoch": 0.3166533226581265,
+      "grad_norm": 1.1056669145759417,
+      "learning_rate": 1.59915971012897e-05,
+      "loss": 1.5992,
+      "step": 2373
+    },
+    {
+      "epoch": 0.3167867627435282,
+      "grad_norm": 1.0153669225410895,
+      "learning_rate": 1.5988136297154525e-05,
+      "loss": 1.6659,
+      "step": 2374
+    },
+    {
+      "epoch": 0.3169202028289298,
+      "grad_norm": 0.9485851246206498,
+      "learning_rate": 1.5984674374503663e-05,
+      "loss": 1.638,
+      "step": 2375
+    },
+    {
+      "epoch": 0.31705364291433147,
+      "grad_norm": 1.2485290662581918,
+      "learning_rate": 1.5981211333983765e-05,
+      "loss": 1.6167,
+      "step": 2376
+    },
+    {
+      "epoch": 0.31718708299973314,
+      "grad_norm": 1.0025032236221025,
+      "learning_rate": 1.5977747176241688e-05,
+      "loss": 1.6485,
+      "step": 2377
+    },
+    {
+      "epoch": 0.31732052308513475,
+      "grad_norm": 1.027037150167266,
+      "learning_rate": 1.59742819019245e-05,
+      "loss": 1.635,
+      "step": 2378
+    },
+    {
+      "epoch": 0.3174539631705364,
+      "grad_norm": 1.0390794115599424,
+      "learning_rate": 1.5970815511679472e-05,
+      "loss": 1.6242,
+      "step": 2379
+    },
+    {
+      "epoch": 0.3175874032559381,
+      "grad_norm": 1.178314062565435,
+      "learning_rate": 1.5967348006154084e-05,
+      "loss": 1.6109,
+      "step": 2380
+    },
+    {
+      "epoch": 0.31772084334133976,
+      "grad_norm": 1.1670387061276462,
+      "learning_rate": 1.5963879385996032e-05,
+      "loss": 1.6196,
+      "step": 2381
+    },
+    {
+      "epoch": 0.3178542834267414,
+      "grad_norm": 1.0774835661150515,
+      "learning_rate": 1.596040965185321e-05,
+      "loss": 1.6497,
+      "step": 2382
+    },
+    {
+      "epoch": 0.31798772351214305,
+      "grad_norm": 1.1109464505151274,
+      "learning_rate": 1.5956938804373726e-05,
+      "loss": 1.6395,
+      "step": 2383
+    },
+    {
+      "epoch": 0.3181211635975447,
+      "grad_norm": 1.0759976671368048,
+      "learning_rate": 1.59534668442059e-05,
+      "loss": 1.6022,
+      "step": 2384
+    },
+    {
+      "epoch": 0.31825460368294634,
+      "grad_norm": 1.0363642572438576,
+      "learning_rate": 1.5949993771998248e-05,
+      "loss": 1.6947,
+      "step": 2385
+    },
+    {
+      "epoch": 0.318388043768348,
+      "grad_norm": 1.1364653119855215,
+      "learning_rate": 1.59465195883995e-05,
+      "loss": 1.6464,
+      "step": 2386
+    },
+    {
+      "epoch": 0.3185214838537497,
+      "grad_norm": 1.1363035834435125,
+      "learning_rate": 1.59430442940586e-05,
+      "loss": 1.6516,
+      "step": 2387
+    },
+    {
+      "epoch": 0.31865492393915135,
+      "grad_norm": 1.0660321057250608,
+      "learning_rate": 1.5939567889624693e-05,
+      "loss": 1.6402,
+      "step": 2388
+    },
+    {
+      "epoch": 0.31878836402455296,
+      "grad_norm": 1.0643000544064976,
+      "learning_rate": 1.5936090375747123e-05,
+      "loss": 1.6054,
+      "step": 2389
+    },
+    {
+      "epoch": 0.31892180410995463,
+      "grad_norm": 1.234549999729085,
+      "learning_rate": 1.593261175307546e-05,
+      "loss": 1.6499,
+      "step": 2390
+    },
+    {
+      "epoch": 0.3190552441953563,
+      "grad_norm": 1.1660252709285277,
+      "learning_rate": 1.592913202225946e-05,
+      "loss": 1.6534,
+      "step": 2391
+    },
+    {
+      "epoch": 0.3191886842807579,
+      "grad_norm": 1.0300136024667785,
+      "learning_rate": 1.592565118394911e-05,
+      "loss": 1.6732,
+      "step": 2392
+    },
+    {
+      "epoch": 0.3193221243661596,
+      "grad_norm": 1.0430395771115888,
+      "learning_rate": 1.592216923879458e-05,
+      "loss": 1.6707,
+      "step": 2393
+    },
+    {
+      "epoch": 0.31945556445156126,
+      "grad_norm": 1.056183541401937,
+      "learning_rate": 1.591868618744627e-05,
+      "loss": 1.6245,
+      "step": 2394
+    },
+    {
+      "epoch": 0.31958900453696293,
+      "grad_norm": 1.0460496279175218,
+      "learning_rate": 1.591520203055476e-05,
+      "loss": 1.6525,
+      "step": 2395
+    },
+    {
+      "epoch": 0.31972244462236454,
+      "grad_norm": 1.168050816657518,
+      "learning_rate": 1.5911716768770863e-05,
+      "loss": 1.692,
+      "step": 2396
+    },
+    {
+      "epoch": 0.3198558847077662,
+      "grad_norm": 1.014257361343356,
+      "learning_rate": 1.5908230402745576e-05,
+      "loss": 1.639,
+      "step": 2397
+    },
+    {
+      "epoch": 0.3199893247931679,
+      "grad_norm": 0.9882262303186528,
+      "learning_rate": 1.590474293313012e-05,
+      "loss": 1.6204,
+      "step": 2398
+    },
+    {
+      "epoch": 0.3201227648785695,
+      "grad_norm": 1.0647363395044251,
+      "learning_rate": 1.5901254360575912e-05,
+      "loss": 1.6276,
+      "step": 2399
+    },
+    {
+      "epoch": 0.32025620496397117,
+      "grad_norm": 1.205341934103895,
+      "learning_rate": 1.589776468573458e-05,
+      "loss": 1.6697,
+      "step": 2400
+    },
+    {
+      "epoch": 0.32038964504937284,
+      "grad_norm": 1.1796009355838823,
+      "learning_rate": 1.589427390925795e-05,
+      "loss": 1.6451,
+      "step": 2401
+    },
+    {
+      "epoch": 0.3205230851347745,
+      "grad_norm": 1.238674085054047,
+      "learning_rate": 1.5890782031798065e-05,
+      "loss": 1.6161,
+      "step": 2402
+    },
+    {
+      "epoch": 0.3206565252201761,
+      "grad_norm": 1.0448493530439331,
+      "learning_rate": 1.5887289054007166e-05,
+      "loss": 1.6487,
+      "step": 2403
+    },
+    {
+      "epoch": 0.3207899653055778,
+      "grad_norm": 1.196197573510979,
+      "learning_rate": 1.5883794976537697e-05,
+      "loss": 1.6033,
+      "step": 2404
+    },
+    {
+      "epoch": 0.32092340539097947,
+      "grad_norm": 1.0194877794457862,
+      "learning_rate": 1.588029980004232e-05,
+      "loss": 1.6694,
+      "step": 2405
+    },
+    {
+      "epoch": 0.3210568454763811,
+      "grad_norm": 1.0318768025671652,
+      "learning_rate": 1.5876803525173883e-05,
+      "loss": 1.6659,
+      "step": 2406
+    },
+    {
+      "epoch": 0.32119028556178275,
+      "grad_norm": 0.9865820243660075,
+      "learning_rate": 1.5873306152585466e-05,
+      "loss": 1.6008,
+      "step": 2407
+    },
+    {
+      "epoch": 0.3213237256471844,
+      "grad_norm": 1.016542806768825,
+      "learning_rate": 1.5869807682930327e-05,
+      "loss": 1.5922,
+      "step": 2408
+    },
+    {
+      "epoch": 0.3214571657325861,
+      "grad_norm": 6.608911252971496,
+      "learning_rate": 1.5866308116861944e-05,
+      "loss": 1.6503,
+      "step": 2409
+    },
+    {
+      "epoch": 0.3215906058179877,
+      "grad_norm": 1.2676970243988273,
+      "learning_rate": 1.586280745503399e-05,
+      "loss": 1.6671,
+      "step": 2410
+    },
+    {
+      "epoch": 0.3217240459033894,
+      "grad_norm": 1.1773866986931305,
+      "learning_rate": 1.585930569810036e-05,
+      "loss": 1.6256,
+      "step": 2411
+    },
+    {
+      "epoch": 0.32185748598879105,
+      "grad_norm": 1.0205646117216987,
+      "learning_rate": 1.5855802846715134e-05,
+      "loss": 1.6661,
+      "step": 2412
+    },
+    {
+      "epoch": 0.32199092607419266,
+      "grad_norm": 1.1132522973470225,
+      "learning_rate": 1.5852298901532604e-05,
+      "loss": 1.615,
+      "step": 2413
+    },
+    {
+      "epoch": 0.32212436615959433,
+      "grad_norm": 1.0003679945512642,
+      "learning_rate": 1.5848793863207276e-05,
+      "loss": 1.6667,
+      "step": 2414
+    },
+    {
+      "epoch": 0.322257806244996,
+      "grad_norm": 1.0445124912690658,
+      "learning_rate": 1.5845287732393845e-05,
+      "loss": 1.6084,
+      "step": 2415
+    },
+    {
+      "epoch": 0.3223912463303977,
+      "grad_norm": 1.341548510183378,
+      "learning_rate": 1.5841780509747213e-05,
+      "loss": 1.5919,
+      "step": 2416
+    },
+    {
+      "epoch": 0.3225246864157993,
+      "grad_norm": 0.9783216242963252,
+      "learning_rate": 1.5838272195922492e-05,
+      "loss": 1.6474,
+      "step": 2417
+    },
+    {
+      "epoch": 0.32265812650120096,
+      "grad_norm": 1.0079878098801114,
+      "learning_rate": 1.5834762791575e-05,
+      "loss": 1.6694,
+      "step": 2418
+    },
+    {
+      "epoch": 0.32279156658660263,
+      "grad_norm": 0.9777003006073635,
+      "learning_rate": 1.5831252297360248e-05,
+      "loss": 1.6087,
+      "step": 2419
+    },
+    {
+      "epoch": 0.32292500667200424,
+      "grad_norm": 1.0210674211146504,
+      "learning_rate": 1.5827740713933958e-05,
+      "loss": 1.6225,
+      "step": 2420
+    },
+    {
+      "epoch": 0.3230584467574059,
+      "grad_norm": 0.9767096638361283,
+      "learning_rate": 1.5824228041952054e-05,
+      "loss": 1.6115,
+      "step": 2421
+    },
+    {
+      "epoch": 0.3231918868428076,
+      "grad_norm": 1.049229142974769,
+      "learning_rate": 1.5820714282070663e-05,
+      "loss": 1.586,
+      "step": 2422
+    },
+    {
+      "epoch": 0.32332532692820926,
+      "grad_norm": 0.9905778604640099,
+      "learning_rate": 1.5817199434946115e-05,
+      "loss": 1.5858,
+      "step": 2423
+    },
+    {
+      "epoch": 0.32345876701361087,
+      "grad_norm": 1.0227388841710876,
+      "learning_rate": 1.5813683501234946e-05,
+      "loss": 1.6233,
+      "step": 2424
+    },
+    {
+      "epoch": 0.32359220709901254,
+      "grad_norm": 1.0857008064049787,
+      "learning_rate": 1.581016648159389e-05,
+      "loss": 1.6423,
+      "step": 2425
+    },
+    {
+      "epoch": 0.3237256471844142,
+      "grad_norm": 1.1213206892388514,
+      "learning_rate": 1.5806648376679885e-05,
+      "loss": 1.6308,
+      "step": 2426
+    },
+    {
+      "epoch": 0.3238590872698158,
+      "grad_norm": 1.0371863646742951,
+      "learning_rate": 1.5803129187150074e-05,
+      "loss": 1.613,
+      "step": 2427
+    },
+    {
+      "epoch": 0.3239925273552175,
+      "grad_norm": 1.0285238928995435,
+      "learning_rate": 1.57996089136618e-05,
+      "loss": 1.632,
+      "step": 2428
+    },
+    {
+      "epoch": 0.32412596744061917,
+      "grad_norm": 1.0053527521421413,
+      "learning_rate": 1.5796087556872616e-05,
+      "loss": 1.6295,
+      "step": 2429
+    },
+    {
+      "epoch": 0.32425940752602084,
+      "grad_norm": 1.0593515172965136,
+      "learning_rate": 1.5792565117440267e-05,
+      "loss": 1.6714,
+      "step": 2430
+    },
+    {
+      "epoch": 0.32439284761142245,
+      "grad_norm": 0.9962509724536518,
+      "learning_rate": 1.5789041596022702e-05,
+      "loss": 1.6014,
+      "step": 2431
+    },
+    {
+      "epoch": 0.3245262876968241,
+      "grad_norm": 1.150755264452728,
+      "learning_rate": 1.5785516993278082e-05,
+      "loss": 1.568,
+      "step": 2432
+    },
+    {
+      "epoch": 0.3246597277822258,
+      "grad_norm": 1.1482871392650946,
+      "learning_rate": 1.578199130986476e-05,
+      "loss": 1.6891,
+      "step": 2433
+    },
+    {
+      "epoch": 0.3247931678676274,
+      "grad_norm": 1.0524081924283053,
+      "learning_rate": 1.5778464546441283e-05,
+      "loss": 1.6186,
+      "step": 2434
+    },
+    {
+      "epoch": 0.3249266079530291,
+      "grad_norm": 0.9963103718861039,
+      "learning_rate": 1.5774936703666423e-05,
+      "loss": 1.6579,
+      "step": 2435
+    },
+    {
+      "epoch": 0.32506004803843075,
+      "grad_norm": 1.034416639979626,
+      "learning_rate": 1.5771407782199137e-05,
+      "loss": 1.6743,
+      "step": 2436
+    },
+    {
+      "epoch": 0.3251934881238324,
+      "grad_norm": 1.0233115564518966,
+      "learning_rate": 1.5767877782698588e-05,
+      "loss": 1.6388,
+      "step": 2437
+    },
+    {
+      "epoch": 0.32532692820923403,
+      "grad_norm": 1.0203611371448205,
+      "learning_rate": 1.576434670582414e-05,
+      "loss": 1.6796,
+      "step": 2438
+    },
+    {
+      "epoch": 0.3254603682946357,
+      "grad_norm": 1.1316243969954793,
+      "learning_rate": 1.5760814552235353e-05,
+      "loss": 1.6287,
+      "step": 2439
+    },
+    {
+      "epoch": 0.3255938083800374,
+      "grad_norm": 2.773420421285832,
+      "learning_rate": 1.5757281322592e-05,
+      "loss": 1.7335,
+      "step": 2440
+    },
+    {
+      "epoch": 0.32572724846543905,
+      "grad_norm": 0.9752268956267421,
+      "learning_rate": 1.5753747017554043e-05,
+      "loss": 1.6287,
+      "step": 2441
+    },
+    {
+      "epoch": 0.32586068855084066,
+      "grad_norm": 1.1865730627498443,
+      "learning_rate": 1.5750211637781654e-05,
+      "loss": 1.6662,
+      "step": 2442
+    },
+    {
+      "epoch": 0.32599412863624233,
+      "grad_norm": 1.066883641739535,
+      "learning_rate": 1.57466751839352e-05,
+      "loss": 1.6201,
+      "step": 2443
+    },
+    {
+      "epoch": 0.326127568721644,
+      "grad_norm": 1.012400294015082,
+      "learning_rate": 1.5743137656675248e-05,
+      "loss": 1.6834,
+      "step": 2444
+    },
+    {
+      "epoch": 0.3262610088070456,
+      "grad_norm": 1.2919053431818395,
+      "learning_rate": 1.5739599056662572e-05,
+      "loss": 1.63,
+      "step": 2445
+    },
+    {
+      "epoch": 0.3263944488924473,
+      "grad_norm": 1.1008048203769303,
+      "learning_rate": 1.5736059384558136e-05,
+      "loss": 1.6535,
+      "step": 2446
+    },
+    {
+      "epoch": 0.32652788897784896,
+      "grad_norm": 1.0198928027043783,
+      "learning_rate": 1.573251864102312e-05,
+      "loss": 1.6043,
+      "step": 2447
+    },
+    {
+      "epoch": 0.3266613290632506,
+      "grad_norm": 1.255841958086851,
+      "learning_rate": 1.5728976826718887e-05,
+      "loss": 1.613,
+      "step": 2448
+    },
+    {
+      "epoch": 0.32679476914865224,
+      "grad_norm": 1.1079219569085927,
+      "learning_rate": 1.572543394230701e-05,
+      "loss": 1.5904,
+      "step": 2449
+    },
+    {
+      "epoch": 0.3269282092340539,
+      "grad_norm": 1.0252017128719715,
+      "learning_rate": 1.5721889988449258e-05,
+      "loss": 1.6363,
+      "step": 2450
+    },
+    {
+      "epoch": 0.3270616493194556,
+      "grad_norm": 1.051784185647374,
+      "learning_rate": 1.5718344965807602e-05,
+      "loss": 1.611,
+      "step": 2451
+    },
+    {
+      "epoch": 0.3271950894048572,
+      "grad_norm": 0.9834253613840619,
+      "learning_rate": 1.5714798875044214e-05,
+      "loss": 1.6072,
+      "step": 2452
+    },
+    {
+      "epoch": 0.32732852949025887,
+      "grad_norm": 1.0314033695212004,
+      "learning_rate": 1.5711251716821458e-05,
+      "loss": 1.6627,
+      "step": 2453
+    },
+    {
+      "epoch": 0.32746196957566054,
+      "grad_norm": 0.9745160928801103,
+      "learning_rate": 1.570770349180191e-05,
+      "loss": 1.6387,
+      "step": 2454
+    },
+    {
+      "epoch": 0.3275954096610622,
+      "grad_norm": 1.0004048445912241,
+      "learning_rate": 1.5704154200648328e-05,
+      "loss": 1.616,
+      "step": 2455
+    },
+    {
+      "epoch": 0.3277288497464638,
+      "grad_norm": 1.0107807331175838,
+      "learning_rate": 1.5700603844023695e-05,
+      "loss": 1.6356,
+      "step": 2456
+    },
+    {
+      "epoch": 0.3278622898318655,
+      "grad_norm": 0.9520529144617134,
+      "learning_rate": 1.5697052422591163e-05,
+      "loss": 1.5897,
+      "step": 2457
+    },
+    {
+      "epoch": 0.32799572991726716,
+      "grad_norm": 1.0058106372712747,
+      "learning_rate": 1.56934999370141e-05,
+      "loss": 1.6069,
+      "step": 2458
+    },
+    {
+      "epoch": 0.3281291700026688,
+      "grad_norm": 1.0663435525855554,
+      "learning_rate": 1.5689946387956075e-05,
+      "loss": 1.7066,
+      "step": 2459
+    },
+    {
+      "epoch": 0.32826261008807045,
+      "grad_norm": 0.9831844195678802,
+      "learning_rate": 1.5686391776080844e-05,
+      "loss": 1.5786,
+      "step": 2460
+    },
+    {
+      "epoch": 0.3283960501734721,
+      "grad_norm": 1.0642774090185225,
+      "learning_rate": 1.5682836102052376e-05,
+      "loss": 1.6452,
+      "step": 2461
+    },
+    {
+      "epoch": 0.3285294902588738,
+      "grad_norm": 1.037534479854548,
+      "learning_rate": 1.5679279366534822e-05,
+      "loss": 1.6082,
+      "step": 2462
+    },
+    {
+      "epoch": 0.3286629303442754,
+      "grad_norm": 1.173992093363848,
+      "learning_rate": 1.5675721570192545e-05,
+      "loss": 1.6464,
+      "step": 2463
+    },
+    {
+      "epoch": 0.3287963704296771,
+      "grad_norm": 1.0051707385512147,
+      "learning_rate": 1.5672162713690098e-05,
+      "loss": 1.6256,
+      "step": 2464
+    },
+    {
+      "epoch": 0.32892981051507875,
+      "grad_norm": 1.0045689191856597,
+      "learning_rate": 1.5668602797692238e-05,
+      "loss": 1.629,
+      "step": 2465
+    },
+    {
+      "epoch": 0.32906325060048036,
+      "grad_norm": 1.0240816684727387,
+      "learning_rate": 1.566504182286391e-05,
+      "loss": 1.6125,
+      "step": 2466
+    },
+    {
+      "epoch": 0.32919669068588203,
+      "grad_norm": 0.9893049585917324,
+      "learning_rate": 1.5661479789870277e-05,
+      "loss": 1.6334,
+      "step": 2467
+    },
+    {
+      "epoch": 0.3293301307712837,
+      "grad_norm": 1.0764320535215586,
+      "learning_rate": 1.5657916699376668e-05,
+      "loss": 1.6781,
+      "step": 2468
+    },
+    {
+      "epoch": 0.3294635708566854,
+      "grad_norm": 1.2439578576993158,
+      "learning_rate": 1.565435255204864e-05,
+      "loss": 1.6416,
+      "step": 2469
+    },
+    {
+      "epoch": 0.329597010942087,
+      "grad_norm": 1.006873672025213,
+      "learning_rate": 1.5650787348551934e-05,
+      "loss": 1.6524,
+      "step": 2470
+    },
+    {
+      "epoch": 0.32973045102748866,
+      "grad_norm": 1.0804959807305459,
+      "learning_rate": 1.564722108955249e-05,
+      "loss": 1.6437,
+      "step": 2471
+    },
+    {
+      "epoch": 0.32986389111289033,
+      "grad_norm": 1.1619045794374623,
+      "learning_rate": 1.564365377571644e-05,
+      "loss": 1.5983,
+      "step": 2472
+    },
+    {
+      "epoch": 0.32999733119829194,
+      "grad_norm": 1.1240948384895437,
+      "learning_rate": 1.5640085407710122e-05,
+      "loss": 1.6569,
+      "step": 2473
+    },
+    {
+      "epoch": 0.3301307712836936,
+      "grad_norm": 1.0470044546188255,
+      "learning_rate": 1.5636515986200064e-05,
+      "loss": 1.6384,
+      "step": 2474
+    },
+    {
+      "epoch": 0.3302642113690953,
+      "grad_norm": 0.9851348342236802,
+      "learning_rate": 1.563294551185299e-05,
+      "loss": 1.6211,
+      "step": 2475
+    },
+    {
+      "epoch": 0.33039765145449695,
+      "grad_norm": 1.0566689739549326,
+      "learning_rate": 1.5629373985335836e-05,
+      "loss": 1.6107,
+      "step": 2476
+    },
+    {
+      "epoch": 0.33053109153989857,
+      "grad_norm": 1.2307836242301233,
+      "learning_rate": 1.5625801407315706e-05,
+      "loss": 1.6671,
+      "step": 2477
+    },
+    {
+      "epoch": 0.33066453162530024,
+      "grad_norm": 1.1555471488035003,
+      "learning_rate": 1.562222777845993e-05,
+      "loss": 1.6066,
+      "step": 2478
+    },
+    {
+      "epoch": 0.3307979717107019,
+      "grad_norm": 1.132011537007146,
+      "learning_rate": 1.5618653099436015e-05,
+      "loss": 1.6749,
+      "step": 2479
+    },
+    {
+      "epoch": 0.3309314117961035,
+      "grad_norm": 1.0591256350110159,
+      "learning_rate": 1.561507737091167e-05,
+      "loss": 1.6399,
+      "step": 2480
+    },
+    {
+      "epoch": 0.3310648518815052,
+      "grad_norm": 1.006730799707782,
+      "learning_rate": 1.561150059355481e-05,
+      "loss": 1.641,
+      "step": 2481
+    },
+    {
+      "epoch": 0.33119829196690687,
+      "grad_norm": 1.114208556739606,
+      "learning_rate": 1.560792276803352e-05,
+      "loss": 1.645,
+      "step": 2482
+    },
+    {
+      "epoch": 0.33133173205230854,
+      "grad_norm": 1.0266741767972964,
+      "learning_rate": 1.5604343895016107e-05,
+      "loss": 1.6305,
+      "step": 2483
+    },
+    {
+      "epoch": 0.33146517213771015,
+      "grad_norm": 1.0172781397984123,
+      "learning_rate": 1.5600763975171065e-05,
+      "loss": 1.6619,
+      "step": 2484
+    },
+    {
+      "epoch": 0.3315986122231118,
+      "grad_norm": 1.0013978329841358,
+      "learning_rate": 1.5597183009167076e-05,
+      "loss": 1.6364,
+      "step": 2485
+    },
+    {
+      "epoch": 0.3317320523085135,
+      "grad_norm": 1.023294574063603,
+      "learning_rate": 1.559360099767303e-05,
+      "loss": 1.6665,
+      "step": 2486
+    },
+    {
+      "epoch": 0.3318654923939151,
+      "grad_norm": 1.0000206147873272,
+      "learning_rate": 1.5590017941358002e-05,
+      "loss": 1.5405,
+      "step": 2487
+    },
+    {
+      "epoch": 0.3319989324793168,
+      "grad_norm": 1.0226703188487969,
+      "learning_rate": 1.5586433840891268e-05,
+      "loss": 1.6402,
+      "step": 2488
+    },
+    {
+      "epoch": 0.33213237256471845,
+      "grad_norm": 1.1484976146272843,
+      "learning_rate": 1.5582848696942298e-05,
+      "loss": 1.6451,
+      "step": 2489
+    },
+    {
+      "epoch": 0.3322658126501201,
+      "grad_norm": 1.0151865246487197,
+      "learning_rate": 1.557926251018075e-05,
+      "loss": 1.6442,
+      "step": 2490
+    },
+    {
+      "epoch": 0.33239925273552173,
+      "grad_norm": 1.0315722760396437,
+      "learning_rate": 1.557567528127649e-05,
+      "loss": 1.6478,
+      "step": 2491
+    },
+    {
+      "epoch": 0.3325326928209234,
+      "grad_norm": 0.9943912244470389,
+      "learning_rate": 1.557208701089957e-05,
+      "loss": 1.622,
+      "step": 2492
+    },
+    {
+      "epoch": 0.3326661329063251,
+      "grad_norm": 1.27101225719953,
+      "learning_rate": 1.5568497699720237e-05,
+      "loss": 1.619,
+      "step": 2493
+    },
+    {
+      "epoch": 0.3327995729917267,
+      "grad_norm": 1.0729429345846575,
+      "learning_rate": 1.556490734840893e-05,
+      "loss": 1.6505,
+      "step": 2494
+    },
+    {
+      "epoch": 0.33293301307712836,
+      "grad_norm": 1.1996428230299965,
+      "learning_rate": 1.5561315957636294e-05,
+      "loss": 1.6437,
+      "step": 2495
+    },
+    {
+      "epoch": 0.33306645316253003,
+      "grad_norm": 1.019452040571454,
+      "learning_rate": 1.555772352807315e-05,
+      "loss": 1.5922,
+      "step": 2496
+    },
+    {
+      "epoch": 0.3331998932479317,
+      "grad_norm": 0.9966197268098483,
+      "learning_rate": 1.5554130060390535e-05,
+      "loss": 1.6469,
+      "step": 2497
+    },
+    {
+      "epoch": 0.3333333333333333,
+      "grad_norm": 0.9924997296677041,
+      "learning_rate": 1.5550535555259654e-05,
+      "loss": 1.6905,
+      "step": 2498
+    },
+    {
+      "epoch": 0.333466773418735,
+      "grad_norm": 1.0222682047094578,
+      "learning_rate": 1.554694001335193e-05,
+      "loss": 1.6463,
+      "step": 2499
+    },
+    {
+      "epoch": 0.33360021350413666,
+      "grad_norm": 0.9803234389294245,
+      "learning_rate": 1.5543343435338965e-05,
+      "loss": 1.6299,
+      "step": 2500
+    },
+    {
+      "epoch": 0.33373365358953827,
+      "grad_norm": 1.1178595006838485,
+      "learning_rate": 1.5539745821892558e-05,
+      "loss": 1.5962,
+      "step": 2501
+    },
+    {
+      "epoch": 0.33386709367493994,
+      "grad_norm": 1.2903970617621783,
+      "learning_rate": 1.5536147173684706e-05,
+      "loss": 1.6377,
+      "step": 2502
+    },
+    {
+      "epoch": 0.3340005337603416,
+      "grad_norm": 0.9991404310713827,
+      "learning_rate": 1.553254749138759e-05,
+      "loss": 1.676,
+      "step": 2503
+    },
+    {
+      "epoch": 0.3341339738457433,
+      "grad_norm": 1.0017240659943722,
+      "learning_rate": 1.55289467756736e-05,
+      "loss": 1.6382,
+      "step": 2504
+    },
+    {
+      "epoch": 0.3342674139311449,
+      "grad_norm": 1.0389840759263576,
+      "learning_rate": 1.5525345027215296e-05,
+      "loss": 1.6566,
+      "step": 2505
+    },
+    {
+      "epoch": 0.33440085401654657,
+      "grad_norm": 1.1995901930820982,
+      "learning_rate": 1.552174224668545e-05,
+      "loss": 1.586,
+      "step": 2506
+    },
+    {
+      "epoch": 0.33453429410194824,
+      "grad_norm": 1.0567962769597756,
+      "learning_rate": 1.551813843475702e-05,
+      "loss": 1.6554,
+      "step": 2507
+    },
+    {
+      "epoch": 0.33466773418734985,
+      "grad_norm": 1.005841753947666,
+      "learning_rate": 1.5514533592103154e-05,
+      "loss": 1.625,
+      "step": 2508
+    },
+    {
+      "epoch": 0.3348011742727515,
+      "grad_norm": 0.9857509429122189,
+      "learning_rate": 1.5510927719397203e-05,
+      "loss": 1.6568,
+      "step": 2509
+    },
+    {
+      "epoch": 0.3349346143581532,
+      "grad_norm": 1.0001185366594785,
+      "learning_rate": 1.55073208173127e-05,
+      "loss": 1.6372,
+      "step": 2510
+    },
+    {
+      "epoch": 0.33506805444355486,
+      "grad_norm": 1.015290456272018,
+      "learning_rate": 1.5503712886523366e-05,
+      "loss": 1.6635,
+      "step": 2511
+    },
+    {
+      "epoch": 0.3352014945289565,
+      "grad_norm": 1.010515707552436,
+      "learning_rate": 1.5500103927703128e-05,
+      "loss": 1.5674,
+      "step": 2512
+    },
+    {
+      "epoch": 0.33533493461435815,
+      "grad_norm": 1.2221140435977949,
+      "learning_rate": 1.54964939415261e-05,
+      "loss": 1.6435,
+      "step": 2513
+    },
+    {
+      "epoch": 0.3354683746997598,
+      "grad_norm": 1.0752443571797692,
+      "learning_rate": 1.549288292866658e-05,
+      "loss": 1.6192,
+      "step": 2514
+    },
+    {
+      "epoch": 0.3356018147851615,
+      "grad_norm": 0.9988521065908124,
+      "learning_rate": 1.5489270889799074e-05,
+      "loss": 1.6454,
+      "step": 2515
+    },
+    {
+      "epoch": 0.3357352548705631,
+      "grad_norm": 1.386021157462588,
+      "learning_rate": 1.548565782559826e-05,
+      "loss": 1.6567,
+      "step": 2516
+    },
+    {
+      "epoch": 0.3358686949559648,
+      "grad_norm": 1.2499411923947341,
+      "learning_rate": 1.5482043736739018e-05,
+      "loss": 1.65,
+      "step": 2517
+    },
+    {
+      "epoch": 0.33600213504136645,
+      "grad_norm": 1.040398945045627,
+      "learning_rate": 1.5478428623896423e-05,
+      "loss": 1.6505,
+      "step": 2518
+    },
+    {
+      "epoch": 0.33613557512676806,
+      "grad_norm": 1.0467166314907603,
+      "learning_rate": 1.5474812487745738e-05,
+      "loss": 1.6145,
+      "step": 2519
+    },
+    {
+      "epoch": 0.33626901521216973,
+      "grad_norm": 0.9735753640039505,
+      "learning_rate": 1.547119532896241e-05,
+      "loss": 1.5752,
+      "step": 2520
+    },
+    {
+      "epoch": 0.3364024552975714,
+      "grad_norm": 1.1331650065762788,
+      "learning_rate": 1.5467577148222087e-05,
+      "loss": 1.5787,
+      "step": 2521
+    },
+    {
+      "epoch": 0.33653589538297307,
+      "grad_norm": 1.0303627504262778,
+      "learning_rate": 1.5463957946200606e-05,
+      "loss": 1.5802,
+      "step": 2522
+    },
+    {
+      "epoch": 0.3366693354683747,
+      "grad_norm": 1.0155983478918824,
+      "learning_rate": 1.5460337723573985e-05,
+      "loss": 1.6243,
+      "step": 2523
+    },
+    {
+      "epoch": 0.33680277555377636,
+      "grad_norm": 1.0289284948725979,
+      "learning_rate": 1.545671648101845e-05,
+      "loss": 1.658,
+      "step": 2524
+    },
+    {
+      "epoch": 0.336936215639178,
+      "grad_norm": 1.2059550726829327,
+      "learning_rate": 1.5453094219210402e-05,
+      "loss": 1.6115,
+      "step": 2525
+    },
+    {
+      "epoch": 0.33706965572457964,
+      "grad_norm": 1.0560606404340085,
+      "learning_rate": 1.544947093882644e-05,
+      "loss": 1.6066,
+      "step": 2526
+    },
+    {
+      "epoch": 0.3372030958099813,
+      "grad_norm": 1.1249026083140772,
+      "learning_rate": 1.5445846640543353e-05,
+      "loss": 1.6824,
+      "step": 2527
+    },
+    {
+      "epoch": 0.337336535895383,
+      "grad_norm": 1.0524987361668996,
+      "learning_rate": 1.5442221325038113e-05,
+      "loss": 1.6226,
+      "step": 2528
+    },
+    {
+      "epoch": 0.33746997598078465,
+      "grad_norm": 1.1006613854032747,
+      "learning_rate": 1.5438594992987893e-05,
+      "loss": 1.6257,
+      "step": 2529
+    },
+    {
+      "epoch": 0.33760341606618627,
+      "grad_norm": 1.1312957489575692,
+      "learning_rate": 1.543496764507005e-05,
+      "loss": 1.6335,
+      "step": 2530
+    },
+    {
+      "epoch": 0.33773685615158794,
+      "grad_norm": 1.0553318381697188,
+      "learning_rate": 1.543133928196213e-05,
+      "loss": 1.6227,
+      "step": 2531
+    },
+    {
+      "epoch": 0.3378702962369896,
+      "grad_norm": 1.1640026808016215,
+      "learning_rate": 1.542770990434187e-05,
+      "loss": 1.615,
+      "step": 2532
+    },
+    {
+      "epoch": 0.3380037363223912,
+      "grad_norm": 1.1330496348678385,
+      "learning_rate": 1.5424079512887197e-05,
+      "loss": 1.6289,
+      "step": 2533
+    },
+    {
+      "epoch": 0.3381371764077929,
+      "grad_norm": 1.235500382588085,
+      "learning_rate": 1.5420448108276225e-05,
+      "loss": 1.6293,
+      "step": 2534
+    },
+    {
+      "epoch": 0.33827061649319456,
+      "grad_norm": 0.9892442737683309,
+      "learning_rate": 1.5416815691187263e-05,
+      "loss": 1.6451,
+      "step": 2535
+    },
+    {
+      "epoch": 0.33840405657859624,
+      "grad_norm": 1.0714941246892575,
+      "learning_rate": 1.54131822622988e-05,
+      "loss": 1.6383,
+      "step": 2536
+    },
+    {
+      "epoch": 0.33853749666399785,
+      "grad_norm": 0.9765304919004115,
+      "learning_rate": 1.5409547822289523e-05,
+      "loss": 1.6106,
+      "step": 2537
+    },
+    {
+      "epoch": 0.3386709367493995,
+      "grad_norm": 1.1397242313512768,
+      "learning_rate": 1.5405912371838303e-05,
+      "loss": 1.6759,
+      "step": 2538
+    },
+    {
+      "epoch": 0.3388043768348012,
+      "grad_norm": 1.04733183839702,
+      "learning_rate": 1.5402275911624202e-05,
+      "loss": 1.6065,
+      "step": 2539
+    },
+    {
+      "epoch": 0.3389378169202028,
+      "grad_norm": 1.0923000163302528,
+      "learning_rate": 1.5398638442326466e-05,
+      "loss": 1.5728,
+      "step": 2540
+    },
+    {
+      "epoch": 0.3390712570056045,
+      "grad_norm": 1.012586965117265,
+      "learning_rate": 1.5394999964624534e-05,
+      "loss": 1.6695,
+      "step": 2541
+    },
+    {
+      "epoch": 0.33920469709100615,
+      "grad_norm": 1.2400537139148242,
+      "learning_rate": 1.5391360479198035e-05,
+      "loss": 1.6388,
+      "step": 2542
+    },
+    {
+      "epoch": 0.3393381371764078,
+      "grad_norm": 1.2509378456829192,
+      "learning_rate": 1.538771998672678e-05,
+      "loss": 1.596,
+      "step": 2543
+    },
+    {
+      "epoch": 0.33947157726180943,
+      "grad_norm": 1.075544011557188,
+      "learning_rate": 1.538407848789077e-05,
+      "loss": 1.5977,
+      "step": 2544
+    },
+    {
+      "epoch": 0.3396050173472111,
+      "grad_norm": 1.2176294488296542,
+      "learning_rate": 1.5380435983370207e-05,
+      "loss": 1.6392,
+      "step": 2545
+    },
+    {
+      "epoch": 0.3397384574326128,
+      "grad_norm": 1.3397975236283604,
+      "learning_rate": 1.5376792473845453e-05,
+      "loss": 1.6818,
+      "step": 2546
+    },
+    {
+      "epoch": 0.3398718975180144,
+      "grad_norm": 1.046773813890146,
+      "learning_rate": 1.5373147959997085e-05,
+      "loss": 1.6099,
+      "step": 2547
+    },
+    {
+      "epoch": 0.34000533760341606,
+      "grad_norm": 1.0648283430543095,
+      "learning_rate": 1.5369502442505852e-05,
+      "loss": 1.6449,
+      "step": 2548
+    },
+    {
+      "epoch": 0.34013877768881773,
+      "grad_norm": 1.1528358481625516,
+      "learning_rate": 1.53658559220527e-05,
+      "loss": 1.6178,
+      "step": 2549
+    },
+    {
+      "epoch": 0.3402722177742194,
+      "grad_norm": 1.230719491103403,
+      "learning_rate": 1.5362208399318753e-05,
+      "loss": 1.6431,
+      "step": 2550
+    },
+    {
+      "epoch": 0.340405657859621,
+      "grad_norm": 0.9911045985776983,
+      "learning_rate": 1.5358559874985326e-05,
+      "loss": 1.6593,
+      "step": 2551
+    },
+    {
+      "epoch": 0.3405390979450227,
+      "grad_norm": 1.0148399749816694,
+      "learning_rate": 1.5354910349733926e-05,
+      "loss": 1.6305,
+      "step": 2552
+    },
+    {
+      "epoch": 0.34067253803042435,
+      "grad_norm": 0.9757587532893943,
+      "learning_rate": 1.5351259824246244e-05,
+      "loss": 1.6463,
+      "step": 2553
+    },
+    {
+      "epoch": 0.34080597811582597,
+      "grad_norm": 1.0377383659545665,
+      "learning_rate": 1.5347608299204147e-05,
+      "loss": 1.5969,
+      "step": 2554
+    },
+    {
+      "epoch": 0.34093941820122764,
+      "grad_norm": 1.1327301101792895,
+      "learning_rate": 1.5343955775289708e-05,
+      "loss": 1.5886,
+      "step": 2555
+    },
+    {
+      "epoch": 0.3410728582866293,
+      "grad_norm": 1.1199829427800625,
+      "learning_rate": 1.534030225318517e-05,
+      "loss": 1.6568,
+      "step": 2556
+    },
+    {
+      "epoch": 0.341206298372031,
+      "grad_norm": 1.1255501040237739,
+      "learning_rate": 1.533664773357298e-05,
+      "loss": 1.6195,
+      "step": 2557
+    },
+    {
+      "epoch": 0.3413397384574326,
+      "grad_norm": 0.9751153526994987,
+      "learning_rate": 1.5332992217135746e-05,
+      "loss": 1.6075,
+      "step": 2558
+    },
+    {
+      "epoch": 0.34147317854283427,
+      "grad_norm": 1.0250661162211099,
+      "learning_rate": 1.5329335704556284e-05,
+      "loss": 1.6398,
+      "step": 2559
+    },
+    {
+      "epoch": 0.34160661862823594,
+      "grad_norm": 1.1963912340246656,
+      "learning_rate": 1.532567819651759e-05,
+      "loss": 1.6364,
+      "step": 2560
+    },
+    {
+      "epoch": 0.34174005871363755,
+      "grad_norm": 1.0270905478988817,
+      "learning_rate": 1.532201969370285e-05,
+      "loss": 1.6027,
+      "step": 2561
+    },
+    {
+      "epoch": 0.3418734987990392,
+      "grad_norm": 0.9716309113456719,
+      "learning_rate": 1.531836019679542e-05,
+      "loss": 1.6109,
+      "step": 2562
+    },
+    {
+      "epoch": 0.3420069388844409,
+      "grad_norm": 1.0974203593959186,
+      "learning_rate": 1.5314699706478856e-05,
+      "loss": 1.6531,
+      "step": 2563
+    },
+    {
+      "epoch": 0.34214037896984256,
+      "grad_norm": 1.1280911638633904,
+      "learning_rate": 1.53110382234369e-05,
+      "loss": 1.6526,
+      "step": 2564
+    },
+    {
+      "epoch": 0.3422738190552442,
+      "grad_norm": 1.0352666364430376,
+      "learning_rate": 1.530737574835347e-05,
+      "loss": 1.638,
+      "step": 2565
+    },
+    {
+      "epoch": 0.34240725914064585,
+      "grad_norm": 0.9852011126418923,
+      "learning_rate": 1.530371228191268e-05,
+      "loss": 1.629,
+      "step": 2566
+    },
+    {
+      "epoch": 0.3425406992260475,
+      "grad_norm": 1.7874095738872313,
+      "learning_rate": 1.5300047824798815e-05,
+      "loss": 1.6322,
+      "step": 2567
+    },
+    {
+      "epoch": 0.34267413931144913,
+      "grad_norm": 1.068401710484991,
+      "learning_rate": 1.5296382377696364e-05,
+      "loss": 1.6951,
+      "step": 2568
+    },
+    {
+      "epoch": 0.3428075793968508,
+      "grad_norm": 0.9922432887897237,
+      "learning_rate": 1.5292715941289985e-05,
+      "loss": 1.6335,
+      "step": 2569
+    },
+    {
+      "epoch": 0.3429410194822525,
+      "grad_norm": 0.9469476131303167,
+      "learning_rate": 1.528904851626453e-05,
+      "loss": 1.5556,
+      "step": 2570
+    },
+    {
+      "epoch": 0.34307445956765414,
+      "grad_norm": 1.0252957676256522,
+      "learning_rate": 1.528538010330503e-05,
+      "loss": 1.6207,
+      "step": 2571
+    },
+    {
+      "epoch": 0.34320789965305576,
+      "grad_norm": 1.256901118189651,
+      "learning_rate": 1.52817107030967e-05,
+      "loss": 1.5499,
+      "step": 2572
+    },
+    {
+      "epoch": 0.34334133973845743,
+      "grad_norm": 1.1760660704565096,
+      "learning_rate": 1.5278040316324947e-05,
+      "loss": 1.5907,
+      "step": 2573
+    },
+    {
+      "epoch": 0.3434747798238591,
+      "grad_norm": 1.056335970772178,
+      "learning_rate": 1.5274368943675358e-05,
+      "loss": 1.6585,
+      "step": 2574
+    },
+    {
+      "epoch": 0.3436082199092607,
+      "grad_norm": 1.0440163675723362,
+      "learning_rate": 1.5270696585833697e-05,
+      "loss": 1.6956,
+      "step": 2575
+    },
+    {
+      "epoch": 0.3437416599946624,
+      "grad_norm": 1.1496273753007886,
+      "learning_rate": 1.5267023243485927e-05,
+      "loss": 1.626,
+      "step": 2576
+    },
+    {
+      "epoch": 0.34387510008006406,
+      "grad_norm": 0.989190837912653,
+      "learning_rate": 1.526334891731818e-05,
+      "loss": 1.5821,
+      "step": 2577
+    },
+    {
+      "epoch": 0.3440085401654657,
+      "grad_norm": 0.975382429954711,
+      "learning_rate": 1.5259673608016782e-05,
+      "loss": 1.6546,
+      "step": 2578
+    },
+    {
+      "epoch": 0.34414198025086734,
+      "grad_norm": 1.2380666712332904,
+      "learning_rate": 1.5255997316268234e-05,
+      "loss": 1.6596,
+      "step": 2579
+    },
+    {
+      "epoch": 0.344275420336269,
+      "grad_norm": 1.0226269070252,
+      "learning_rate": 1.5252320042759232e-05,
+      "loss": 1.6045,
+      "step": 2580
+    },
+    {
+      "epoch": 0.3444088604216707,
+      "grad_norm": 0.9919122022483479,
+      "learning_rate": 1.5248641788176647e-05,
+      "loss": 1.5316,
+      "step": 2581
+    },
+    {
+      "epoch": 0.34454230050707235,
+      "grad_norm": 1.1347269775006643,
+      "learning_rate": 1.5244962553207531e-05,
+      "loss": 1.6384,
+      "step": 2582
+    },
+    {
+      "epoch": 0.34467574059247397,
+      "grad_norm": 0.9828544036878673,
+      "learning_rate": 1.524128233853913e-05,
+      "loss": 1.6399,
+      "step": 2583
+    },
+    {
+      "epoch": 0.34480918067787564,
+      "grad_norm": 1.3101691050309692,
+      "learning_rate": 1.523760114485886e-05,
+      "loss": 1.6285,
+      "step": 2584
+    },
+    {
+      "epoch": 0.3449426207632773,
+      "grad_norm": 1.0927621397780132,
+      "learning_rate": 1.5233918972854329e-05,
+      "loss": 1.5724,
+      "step": 2585
+    },
+    {
+      "epoch": 0.3450760608486789,
+      "grad_norm": 1.0744584089585294,
+      "learning_rate": 1.5230235823213325e-05,
+      "loss": 1.5861,
+      "step": 2586
+    },
+    {
+      "epoch": 0.3452095009340806,
+      "grad_norm": 1.1708529579248028,
+      "learning_rate": 1.5226551696623815e-05,
+      "loss": 1.6419,
+      "step": 2587
+    },
+    {
+      "epoch": 0.34534294101948226,
+      "grad_norm": 1.170484702302893,
+      "learning_rate": 1.5222866593773955e-05,
+      "loss": 1.6793,
+      "step": 2588
+    },
+    {
+      "epoch": 0.34547638110488393,
+      "grad_norm": 0.9940131617224933,
+      "learning_rate": 1.521918051535208e-05,
+      "loss": 1.6196,
+      "step": 2589
+    },
+    {
+      "epoch": 0.34560982119028555,
+      "grad_norm": 1.0291494485832964,
+      "learning_rate": 1.5215493462046711e-05,
+      "loss": 1.6436,
+      "step": 2590
+    },
+    {
+      "epoch": 0.3457432612756872,
+      "grad_norm": 1.1772938675721651,
+      "learning_rate": 1.5211805434546541e-05,
+      "loss": 1.5946,
+      "step": 2591
+    },
+    {
+      "epoch": 0.3458767013610889,
+      "grad_norm": 1.3230747393201525,
+      "learning_rate": 1.5208116433540454e-05,
+      "loss": 1.6298,
+      "step": 2592
+    },
+    {
+      "epoch": 0.3460101414464905,
+      "grad_norm": 1.0683806522297445,
+      "learning_rate": 1.5204426459717516e-05,
+      "loss": 1.6174,
+      "step": 2593
+    },
+    {
+      "epoch": 0.3461435815318922,
+      "grad_norm": 1.1046905551827517,
+      "learning_rate": 1.520073551376697e-05,
+      "loss": 1.6434,
+      "step": 2594
+    },
+    {
+      "epoch": 0.34627702161729385,
+      "grad_norm": 1.070450277819012,
+      "learning_rate": 1.519704359637824e-05,
+      "loss": 1.6019,
+      "step": 2595
+    },
+    {
+      "epoch": 0.3464104617026955,
+      "grad_norm": 1.3058537795961895,
+      "learning_rate": 1.519335070824094e-05,
+      "loss": 1.6049,
+      "step": 2596
+    },
+    {
+      "epoch": 0.34654390178809713,
+      "grad_norm": 1.0080854625304823,
+      "learning_rate": 1.5189656850044853e-05,
+      "loss": 1.6254,
+      "step": 2597
+    },
+    {
+      "epoch": 0.3466773418734988,
+      "grad_norm": 1.2017908924230127,
+      "learning_rate": 1.5185962022479954e-05,
+      "loss": 1.6323,
+      "step": 2598
+    },
+    {
+      "epoch": 0.34681078195890047,
+      "grad_norm": 1.0124260951271844,
+      "learning_rate": 1.5182266226236394e-05,
+      "loss": 1.6621,
+      "step": 2599
+    },
+    {
+      "epoch": 0.3469442220443021,
+      "grad_norm": 1.2316810096284962,
+      "learning_rate": 1.5178569462004506e-05,
+      "loss": 1.5827,
+      "step": 2600
+    },
+    {
+      "epoch": 0.34707766212970376,
+      "grad_norm": 1.0242976500030199,
+      "learning_rate": 1.51748717304748e-05,
+      "loss": 1.6034,
+      "step": 2601
+    },
+    {
+      "epoch": 0.3472111022151054,
+      "grad_norm": 1.0030046362345064,
+      "learning_rate": 1.5171173032337976e-05,
+      "loss": 1.654,
+      "step": 2602
+    },
+    {
+      "epoch": 0.3473445423005071,
+      "grad_norm": 1.0713133664664767,
+      "learning_rate": 1.5167473368284902e-05,
+      "loss": 1.6352,
+      "step": 2603
+    },
+    {
+      "epoch": 0.3474779823859087,
+      "grad_norm": 1.0969747552844828,
+      "learning_rate": 1.516377273900664e-05,
+      "loss": 1.6475,
+      "step": 2604
+    },
+    {
+      "epoch": 0.3476114224713104,
+      "grad_norm": 0.9899374388035455,
+      "learning_rate": 1.5160071145194422e-05,
+      "loss": 1.6286,
+      "step": 2605
+    },
+    {
+      "epoch": 0.34774486255671205,
+      "grad_norm": 0.9679955249669937,
+      "learning_rate": 1.515636858753966e-05,
+      "loss": 1.6383,
+      "step": 2606
+    },
+    {
+      "epoch": 0.34787830264211367,
+      "grad_norm": 1.0249461107418476,
+      "learning_rate": 1.5152665066733957e-05,
+      "loss": 1.6083,
+      "step": 2607
+    },
+    {
+      "epoch": 0.34801174272751534,
+      "grad_norm": 1.303220194256495,
+      "learning_rate": 1.5148960583469085e-05,
+      "loss": 1.6688,
+      "step": 2608
+    },
+    {
+      "epoch": 0.348145182812917,
+      "grad_norm": 1.0188653868990143,
+      "learning_rate": 1.5145255138436999e-05,
+      "loss": 1.6288,
+      "step": 2609
+    },
+    {
+      "epoch": 0.3482786228983187,
+      "grad_norm": 1.2584157965822476,
+      "learning_rate": 1.5141548732329831e-05,
+      "loss": 1.6492,
+      "step": 2610
+    },
+    {
+      "epoch": 0.3484120629837203,
+      "grad_norm": 1.0306474809508248,
+      "learning_rate": 1.5137841365839901e-05,
+      "loss": 1.6854,
+      "step": 2611
+    },
+    {
+      "epoch": 0.34854550306912196,
+      "grad_norm": 1.056381467093589,
+      "learning_rate": 1.51341330396597e-05,
+      "loss": 1.6271,
+      "step": 2612
+    },
+    {
+      "epoch": 0.34867894315452364,
+      "grad_norm": 1.0295530745478738,
+      "learning_rate": 1.51304237544819e-05,
+      "loss": 1.6633,
+      "step": 2613
+    },
+    {
+      "epoch": 0.34881238323992525,
+      "grad_norm": 1.0398915950232068,
+      "learning_rate": 1.5126713510999359e-05,
+      "loss": 1.627,
+      "step": 2614
+    },
+    {
+      "epoch": 0.3489458233253269,
+      "grad_norm": 1.0146998246386771,
+      "learning_rate": 1.51230023099051e-05,
+      "loss": 1.627,
+      "step": 2615
+    },
+    {
+      "epoch": 0.3490792634107286,
+      "grad_norm": 1.0172109821324007,
+      "learning_rate": 1.5119290151892339e-05,
+      "loss": 1.5836,
+      "step": 2616
+    },
+    {
+      "epoch": 0.34921270349613026,
+      "grad_norm": 1.0267520201741256,
+      "learning_rate": 1.5115577037654464e-05,
+      "loss": 1.6439,
+      "step": 2617
+    },
+    {
+      "epoch": 0.3493461435815319,
+      "grad_norm": 1.0457857231337446,
+      "learning_rate": 1.5111862967885036e-05,
+      "loss": 1.6157,
+      "step": 2618
+    },
+    {
+      "epoch": 0.34947958366693355,
+      "grad_norm": 1.0555824876624866,
+      "learning_rate": 1.5108147943277813e-05,
+      "loss": 1.5927,
+      "step": 2619
+    },
+    {
+      "epoch": 0.3496130237523352,
+      "grad_norm": 0.9824068117065244,
+      "learning_rate": 1.5104431964526707e-05,
+      "loss": 1.6322,
+      "step": 2620
+    },
+    {
+      "epoch": 0.34974646383773683,
+      "grad_norm": 1.0882477650288422,
+      "learning_rate": 1.510071503232583e-05,
+      "loss": 1.5678,
+      "step": 2621
+    },
+    {
+      "epoch": 0.3498799039231385,
+      "grad_norm": 0.966848710204604,
+      "learning_rate": 1.5096997147369457e-05,
+      "loss": 1.6656,
+      "step": 2622
+    },
+    {
+      "epoch": 0.3500133440085402,
+      "grad_norm": 0.9955616430245504,
+      "learning_rate": 1.509327831035205e-05,
+      "loss": 1.6429,
+      "step": 2623
+    },
+    {
+      "epoch": 0.35014678409394184,
+      "grad_norm": 1.5091711649173347,
+      "learning_rate": 1.5089558521968242e-05,
+      "loss": 1.7588,
+      "step": 2624
+    },
+    {
+      "epoch": 0.35028022417934346,
+      "grad_norm": 0.9707332773761684,
+      "learning_rate": 1.5085837782912851e-05,
+      "loss": 1.6116,
+      "step": 2625
+    },
+    {
+      "epoch": 0.35041366426474513,
+      "grad_norm": 0.9642328468057494,
+      "learning_rate": 1.5082116093880862e-05,
+      "loss": 1.6222,
+      "step": 2626
+    },
+    {
+      "epoch": 0.3505471043501468,
+      "grad_norm": 1.0194989516497073,
+      "learning_rate": 1.5078393455567452e-05,
+      "loss": 1.6335,
+      "step": 2627
+    },
+    {
+      "epoch": 0.3506805444355484,
+      "grad_norm": 0.9709628625829926,
+      "learning_rate": 1.5074669868667964e-05,
+      "loss": 1.6335,
+      "step": 2628
+    },
+    {
+      "epoch": 0.3508139845209501,
+      "grad_norm": 1.0095163214896312,
+      "learning_rate": 1.5070945333877923e-05,
+      "loss": 1.6442,
+      "step": 2629
+    },
+    {
+      "epoch": 0.35094742460635175,
+      "grad_norm": 1.1931846380684248,
+      "learning_rate": 1.506721985189303e-05,
+      "loss": 1.638,
+      "step": 2630
+    },
+    {
+      "epoch": 0.3510808646917534,
+      "grad_norm": 0.9839265090005194,
+      "learning_rate": 1.5063493423409154e-05,
+      "loss": 1.6196,
+      "step": 2631
+    },
+    {
+      "epoch": 0.35121430477715504,
+      "grad_norm": 1.1186506149798139,
+      "learning_rate": 1.5059766049122364e-05,
+      "loss": 1.6225,
+      "step": 2632
+    },
+    {
+      "epoch": 0.3513477448625567,
+      "grad_norm": 1.0864720980811629,
+      "learning_rate": 1.5056037729728878e-05,
+      "loss": 1.5935,
+      "step": 2633
+    },
+    {
+      "epoch": 0.3514811849479584,
+      "grad_norm": 1.2246349786822723,
+      "learning_rate": 1.5052308465925115e-05,
+      "loss": 1.5767,
+      "step": 2634
+    },
+    {
+      "epoch": 0.35161462503336,
+      "grad_norm": 1.0333745712542706,
+      "learning_rate": 1.5048578258407646e-05,
+      "loss": 1.6637,
+      "step": 2635
+    },
+    {
+      "epoch": 0.35174806511876167,
+      "grad_norm": 1.0018496493629654,
+      "learning_rate": 1.5044847107873246e-05,
+      "loss": 1.6278,
+      "step": 2636
+    },
+    {
+      "epoch": 0.35188150520416334,
+      "grad_norm": 1.1758224822257188,
+      "learning_rate": 1.5041115015018841e-05,
+      "loss": 1.6303,
+      "step": 2637
+    },
+    {
+      "epoch": 0.352014945289565,
+      "grad_norm": 1.250987136440272,
+      "learning_rate": 1.5037381980541546e-05,
+      "loss": 1.6571,
+      "step": 2638
+    },
+    {
+      "epoch": 0.3521483853749666,
+      "grad_norm": 1.2181097297996704,
+      "learning_rate": 1.503364800513865e-05,
+      "loss": 1.704,
+      "step": 2639
+    },
+    {
+      "epoch": 0.3522818254603683,
+      "grad_norm": 1.0786935661772743,
+      "learning_rate": 1.5029913089507617e-05,
+      "loss": 1.5949,
+      "step": 2640
+    },
+    {
+      "epoch": 0.35241526554576996,
+      "grad_norm": 1.0531753164421735,
+      "learning_rate": 1.5026177234346087e-05,
+      "loss": 1.6373,
+      "step": 2641
+    },
+    {
+      "epoch": 0.3525487056311716,
+      "grad_norm": 1.3629803662802447,
+      "learning_rate": 1.5022440440351873e-05,
+      "loss": 1.6762,
+      "step": 2642
+    },
+    {
+      "epoch": 0.35268214571657325,
+      "grad_norm": 1.1620679753807144,
+      "learning_rate": 1.501870270822297e-05,
+      "loss": 1.6133,
+      "step": 2643
+    },
+    {
+      "epoch": 0.3528155858019749,
+      "grad_norm": 1.0348126543518374,
+      "learning_rate": 1.5014964038657538e-05,
+      "loss": 1.654,
+      "step": 2644
+    },
+    {
+      "epoch": 0.3529490258873766,
+      "grad_norm": 1.1304213137830432,
+      "learning_rate": 1.5011224432353924e-05,
+      "loss": 1.6386,
+      "step": 2645
+    },
+    {
+      "epoch": 0.3530824659727782,
+      "grad_norm": 1.1277308613925912,
+      "learning_rate": 1.5007483890010642e-05,
+      "loss": 1.6194,
+      "step": 2646
+    },
+    {
+      "epoch": 0.3532159060581799,
+      "grad_norm": 1.0163090940302686,
+      "learning_rate": 1.5003742412326377e-05,
+      "loss": 1.6575,
+      "step": 2647
+    },
+    {
+      "epoch": 0.35334934614358154,
+      "grad_norm": 1.0125879795963375,
+      "learning_rate": 1.5000000000000002e-05,
+      "loss": 1.6453,
+      "step": 2648
+    },
+    {
+      "epoch": 0.35348278622898316,
+      "grad_norm": 1.0098185677548541,
+      "learning_rate": 1.4996256653730554e-05,
+      "loss": 1.6068,
+      "step": 2649
+    },
+    {
+      "epoch": 0.35361622631438483,
+      "grad_norm": 1.2735854212695745,
+      "learning_rate": 1.4992512374217247e-05,
+      "loss": 1.6487,
+      "step": 2650
+    },
+    {
+      "epoch": 0.3537496663997865,
+      "grad_norm": 1.0932430437817762,
+      "learning_rate": 1.498876716215947e-05,
+      "loss": 1.6284,
+      "step": 2651
+    },
+    {
+      "epoch": 0.35388310648518817,
+      "grad_norm": 0.9777781439946768,
+      "learning_rate": 1.4985021018256786e-05,
+      "loss": 1.625,
+      "step": 2652
+    },
+    {
+      "epoch": 0.3540165465705898,
+      "grad_norm": 1.1136916517793387,
+      "learning_rate": 1.4981273943208937e-05,
+      "loss": 1.6708,
+      "step": 2653
+    },
+    {
+      "epoch": 0.35414998665599146,
+      "grad_norm": 1.373624233594401,
+      "learning_rate": 1.4977525937715824e-05,
+      "loss": 1.6216,
+      "step": 2654
+    },
+    {
+      "epoch": 0.3542834267413931,
+      "grad_norm": 1.0595448685922348,
+      "learning_rate": 1.4973777002477538e-05,
+      "loss": 1.5722,
+      "step": 2655
+    },
+    {
+      "epoch": 0.3544168668267948,
+      "grad_norm": 1.2378751910850914,
+      "learning_rate": 1.4970027138194336e-05,
+      "loss": 1.6311,
+      "step": 2656
+    },
+    {
+      "epoch": 0.3545503069121964,
+      "grad_norm": 1.1786205067232547,
+      "learning_rate": 1.4966276345566654e-05,
+      "loss": 1.6456,
+      "step": 2657
+    },
+    {
+      "epoch": 0.3546837469975981,
+      "grad_norm": 1.0572423131869906,
+      "learning_rate": 1.4962524625295092e-05,
+      "loss": 1.6392,
+      "step": 2658
+    },
+    {
+      "epoch": 0.35481718708299975,
+      "grad_norm": 1.0148567443614354,
+      "learning_rate": 1.495877197808043e-05,
+      "loss": 1.6173,
+      "step": 2659
+    },
+    {
+      "epoch": 0.35495062716840137,
+      "grad_norm": 0.9578348594491616,
+      "learning_rate": 1.4955018404623623e-05,
+      "loss": 1.628,
+      "step": 2660
+    },
+    {
+      "epoch": 0.35508406725380304,
+      "grad_norm": 1.1181766176664125,
+      "learning_rate": 1.4951263905625788e-05,
+      "loss": 1.5787,
+      "step": 2661
+    },
+    {
+      "epoch": 0.3552175073392047,
+      "grad_norm": 1.1415697297087068,
+      "learning_rate": 1.4947508481788231e-05,
+      "loss": 1.5782,
+      "step": 2662
+    },
+    {
+      "epoch": 0.3553509474246064,
+      "grad_norm": 0.9866478663748992,
+      "learning_rate": 1.494375213381242e-05,
+      "loss": 1.5652,
+      "step": 2663
+    },
+    {
+      "epoch": 0.355484387510008,
+      "grad_norm": 1.2168790215873122,
+      "learning_rate": 1.4939994862399996e-05,
+      "loss": 1.6349,
+      "step": 2664
+    },
+    {
+      "epoch": 0.35561782759540966,
+      "grad_norm": 1.0419160584710712,
+      "learning_rate": 1.4936236668252772e-05,
+      "loss": 1.625,
+      "step": 2665
+    },
+    {
+      "epoch": 0.35575126768081133,
+      "grad_norm": 1.0033295412035523,
+      "learning_rate": 1.4932477552072745e-05,
+      "loss": 1.6232,
+      "step": 2666
+    },
+    {
+      "epoch": 0.35588470776621295,
+      "grad_norm": 0.9869871754447705,
+      "learning_rate": 1.4928717514562066e-05,
+      "loss": 1.5835,
+      "step": 2667
+    },
+    {
+      "epoch": 0.3560181478516146,
+      "grad_norm": 0.992408467300245,
+      "learning_rate": 1.4924956556423072e-05,
+      "loss": 1.603,
+      "step": 2668
+    },
+    {
+      "epoch": 0.3561515879370163,
+      "grad_norm": 0.9824615329438933,
+      "learning_rate": 1.4921194678358266e-05,
+      "loss": 1.5976,
+      "step": 2669
+    },
+    {
+      "epoch": 0.35628502802241796,
+      "grad_norm": 0.9934973089076394,
+      "learning_rate": 1.4917431881070323e-05,
+      "loss": 1.595,
+      "step": 2670
+    },
+    {
+      "epoch": 0.3564184681078196,
+      "grad_norm": 0.9868126452811748,
+      "learning_rate": 1.4913668165262095e-05,
+      "loss": 1.6287,
+      "step": 2671
+    },
+    {
+      "epoch": 0.35655190819322125,
+      "grad_norm": 1.0146369534155588,
+      "learning_rate": 1.4909903531636593e-05,
+      "loss": 1.5831,
+      "step": 2672
+    },
+    {
+      "epoch": 0.3566853482786229,
+      "grad_norm": 1.0128166092521045,
+      "learning_rate": 1.4906137980897017e-05,
+      "loss": 1.6241,
+      "step": 2673
+    },
+    {
+      "epoch": 0.35681878836402453,
+      "grad_norm": 1.0086684442228309,
+      "learning_rate": 1.4902371513746723e-05,
+      "loss": 1.6162,
+      "step": 2674
+    },
+    {
+      "epoch": 0.3569522284494262,
+      "grad_norm": 1.0430890218619906,
+      "learning_rate": 1.4898604130889243e-05,
+      "loss": 1.6444,
+      "step": 2675
+    },
+    {
+      "epoch": 0.35708566853482787,
+      "grad_norm": 0.9873271173986554,
+      "learning_rate": 1.489483583302829e-05,
+      "loss": 1.6058,
+      "step": 2676
+    },
+    {
+      "epoch": 0.35721910862022954,
+      "grad_norm": 0.9633834898755523,
+      "learning_rate": 1.4891066620867726e-05,
+      "loss": 1.5722,
+      "step": 2677
+    },
+    {
+      "epoch": 0.35735254870563116,
+      "grad_norm": 1.0004128851077128,
+      "learning_rate": 1.4887296495111611e-05,
+      "loss": 1.6337,
+      "step": 2678
+    },
+    {
+      "epoch": 0.3574859887910328,
+      "grad_norm": 1.1889045283167872,
+      "learning_rate": 1.4883525456464152e-05,
+      "loss": 1.5775,
+      "step": 2679
+    },
+    {
+      "epoch": 0.3576194288764345,
+      "grad_norm": 2.5683847190418736,
+      "learning_rate": 1.4879753505629739e-05,
+      "loss": 1.6822,
+      "step": 2680
+    },
+    {
+      "epoch": 0.3577528689618361,
+      "grad_norm": 1.0263603075923868,
+      "learning_rate": 1.4875980643312931e-05,
+      "loss": 1.6556,
+      "step": 2681
+    },
+    {
+      "epoch": 0.3578863090472378,
+      "grad_norm": 1.0275465262931753,
+      "learning_rate": 1.4872206870218451e-05,
+      "loss": 1.6586,
+      "step": 2682
+    },
+    {
+      "epoch": 0.35801974913263945,
+      "grad_norm": 1.0578558588998055,
+      "learning_rate": 1.4868432187051201e-05,
+      "loss": 1.6126,
+      "step": 2683
+    },
+    {
+      "epoch": 0.3581531892180411,
+      "grad_norm": 1.0209797685600523,
+      "learning_rate": 1.4864656594516245e-05,
+      "loss": 1.6051,
+      "step": 2684
+    },
+    {
+      "epoch": 0.35828662930344274,
+      "grad_norm": 1.0698279049277513,
+      "learning_rate": 1.4860880093318827e-05,
+      "loss": 1.6162,
+      "step": 2685
+    },
+    {
+      "epoch": 0.3584200693888444,
+      "grad_norm": 1.0175812497772847,
+      "learning_rate": 1.4857102684164349e-05,
+      "loss": 1.6055,
+      "step": 2686
+    },
+    {
+      "epoch": 0.3585535094742461,
+      "grad_norm": 1.0104871621067544,
+      "learning_rate": 1.485332436775839e-05,
+      "loss": 1.6041,
+      "step": 2687
+    },
+    {
+      "epoch": 0.3586869495596477,
+      "grad_norm": 1.1287495893065134,
+      "learning_rate": 1.4849545144806697e-05,
+      "loss": 1.603,
+      "step": 2688
+    },
+    {
+      "epoch": 0.35882038964504936,
+      "grad_norm": 1.0368374324423373,
+      "learning_rate": 1.4845765016015183e-05,
+      "loss": 1.6295,
+      "step": 2689
+    },
+    {
+      "epoch": 0.35895382973045104,
+      "grad_norm": 1.0590887497081614,
+      "learning_rate": 1.4841983982089936e-05,
+      "loss": 1.6473,
+      "step": 2690
+    },
+    {
+      "epoch": 0.3590872698158527,
+      "grad_norm": 1.1066849585601135,
+      "learning_rate": 1.4838202043737209e-05,
+      "loss": 1.6567,
+      "step": 2691
+    },
+    {
+      "epoch": 0.3592207099012543,
+      "grad_norm": 1.018947820472309,
+      "learning_rate": 1.4834419201663429e-05,
+      "loss": 1.6562,
+      "step": 2692
+    },
+    {
+      "epoch": 0.359354149986656,
+      "grad_norm": 1.128393294375381,
+      "learning_rate": 1.483063545657518e-05,
+      "loss": 1.6634,
+      "step": 2693
+    },
+    {
+      "epoch": 0.35948759007205766,
+      "grad_norm": 1.0975481758375052,
+      "learning_rate": 1.4826850809179228e-05,
+      "loss": 1.6274,
+      "step": 2694
+    },
+    {
+      "epoch": 0.3596210301574593,
+      "grad_norm": 1.1035342439097737,
+      "learning_rate": 1.4823065260182499e-05,
+      "loss": 1.6199,
+      "step": 2695
+    },
+    {
+      "epoch": 0.35975447024286095,
+      "grad_norm": 1.0823826339805787,
+      "learning_rate": 1.4819278810292096e-05,
+      "loss": 1.5614,
+      "step": 2696
+    },
+    {
+      "epoch": 0.3598879103282626,
+      "grad_norm": 1.0810219041790743,
+      "learning_rate": 1.4815491460215277e-05,
+      "loss": 1.6457,
+      "step": 2697
+    },
+    {
+      "epoch": 0.3600213504136643,
+      "grad_norm": 1.008665774442881,
+      "learning_rate": 1.481170321065948e-05,
+      "loss": 1.6252,
+      "step": 2698
+    },
+    {
+      "epoch": 0.3601547904990659,
+      "grad_norm": 1.0280588719731665,
+      "learning_rate": 1.4807914062332307e-05,
+      "loss": 1.6583,
+      "step": 2699
+    },
+    {
+      "epoch": 0.3602882305844676,
+      "grad_norm": 1.0944712865857813,
+      "learning_rate": 1.4804124015941528e-05,
+      "loss": 1.6016,
+      "step": 2700
+    },
+    {
+      "epoch": 0.36042167066986924,
+      "grad_norm": 1.0990375701438984,
+      "learning_rate": 1.480033307219508e-05,
+      "loss": 1.5977,
+      "step": 2701
+    },
+    {
+      "epoch": 0.36055511075527086,
+      "grad_norm": 1.0149776394646597,
+      "learning_rate": 1.4796541231801068e-05,
+      "loss": 1.6027,
+      "step": 2702
+    },
+    {
+      "epoch": 0.36068855084067253,
+      "grad_norm": 1.1355704225231722,
+      "learning_rate": 1.4792748495467763e-05,
+      "loss": 1.5975,
+      "step": 2703
+    },
+    {
+      "epoch": 0.3608219909260742,
+      "grad_norm": 2.0682004186698784,
+      "learning_rate": 1.4788954863903608e-05,
+      "loss": 1.5899,
+      "step": 2704
+    },
+    {
+      "epoch": 0.36095543101147587,
+      "grad_norm": 1.0954874367416438,
+      "learning_rate": 1.478516033781721e-05,
+      "loss": 1.5893,
+      "step": 2705
+    },
+    {
+      "epoch": 0.3610888710968775,
+      "grad_norm": 1.09549046341114,
+      "learning_rate": 1.4781364917917339e-05,
+      "loss": 1.6175,
+      "step": 2706
+    },
+    {
+      "epoch": 0.36122231118227915,
+      "grad_norm": 1.0665598499326367,
+      "learning_rate": 1.477756860491294e-05,
+      "loss": 1.6201,
+      "step": 2707
+    },
+    {
+      "epoch": 0.3613557512676808,
+      "grad_norm": 1.0431226101624222,
+      "learning_rate": 1.4773771399513122e-05,
+      "loss": 1.5921,
+      "step": 2708
+    },
+    {
+      "epoch": 0.36148919135308244,
+      "grad_norm": 1.0347916658313776,
+      "learning_rate": 1.4769973302427154e-05,
+      "loss": 1.5922,
+      "step": 2709
+    },
+    {
+      "epoch": 0.3616226314384841,
+      "grad_norm": 1.188173513439476,
+      "learning_rate": 1.4766174314364486e-05,
+      "loss": 1.6227,
+      "step": 2710
+    },
+    {
+      "epoch": 0.3617560715238858,
+      "grad_norm": 1.1816851578943308,
+      "learning_rate": 1.476237443603472e-05,
+      "loss": 1.6272,
+      "step": 2711
+    },
+    {
+      "epoch": 0.36188951160928745,
+      "grad_norm": 1.047966632364183,
+      "learning_rate": 1.475857366814763e-05,
+      "loss": 1.6298,
+      "step": 2712
+    },
+    {
+      "epoch": 0.36202295169468907,
+      "grad_norm": 1.134317205650707,
+      "learning_rate": 1.4754772011413154e-05,
+      "loss": 1.5883,
+      "step": 2713
+    },
+    {
+      "epoch": 0.36215639178009074,
+      "grad_norm": 1.0021173494097093,
+      "learning_rate": 1.4750969466541407e-05,
+      "loss": 1.6369,
+      "step": 2714
+    },
+    {
+      "epoch": 0.3622898318654924,
+      "grad_norm": 1.1709267393721203,
+      "learning_rate": 1.4747166034242652e-05,
+      "loss": 1.5887,
+      "step": 2715
+    },
+    {
+      "epoch": 0.362423271950894,
+      "grad_norm": 1.1258130657900838,
+      "learning_rate": 1.4743361715227331e-05,
+      "loss": 1.6778,
+      "step": 2716
+    },
+    {
+      "epoch": 0.3625567120362957,
+      "grad_norm": 0.9885486941912983,
+      "learning_rate": 1.4739556510206047e-05,
+      "loss": 1.6048,
+      "step": 2717
+    },
+    {
+      "epoch": 0.36269015212169736,
+      "grad_norm": 1.0268383631231532,
+      "learning_rate": 1.473575041988957e-05,
+      "loss": 1.594,
+      "step": 2718
+    },
+    {
+      "epoch": 0.36282359220709903,
+      "grad_norm": 1.190424029103695,
+      "learning_rate": 1.4731943444988831e-05,
+      "loss": 1.63,
+      "step": 2719
+    },
+    {
+      "epoch": 0.36295703229250065,
+      "grad_norm": 1.0167446687335664,
+      "learning_rate": 1.4728135586214933e-05,
+      "loss": 1.6216,
+      "step": 2720
+    },
+    {
+      "epoch": 0.3630904723779023,
+      "grad_norm": 1.0972474419145468,
+      "learning_rate": 1.472432684427914e-05,
+      "loss": 1.6385,
+      "step": 2721
+    },
+    {
+      "epoch": 0.363223912463304,
+      "grad_norm": 1.1518104547155832,
+      "learning_rate": 1.4720517219892881e-05,
+      "loss": 1.6232,
+      "step": 2722
+    },
+    {
+      "epoch": 0.36335735254870566,
+      "grad_norm": 7.034867431932981,
+      "learning_rate": 1.4716706713767748e-05,
+      "loss": 1.5632,
+      "step": 2723
+    },
+    {
+      "epoch": 0.3634907926341073,
+      "grad_norm": 1.166832787708479,
+      "learning_rate": 1.4712895326615505e-05,
+      "loss": 1.607,
+      "step": 2724
+    },
+    {
+      "epoch": 0.36362423271950894,
+      "grad_norm": 1.0638457063513032,
+      "learning_rate": 1.4709083059148073e-05,
+      "loss": 1.6328,
+      "step": 2725
+    },
+    {
+      "epoch": 0.3637576728049106,
+      "grad_norm": 1.0079889035962148,
+      "learning_rate": 1.470526991207754e-05,
+      "loss": 1.6063,
+      "step": 2726
+    },
+    {
+      "epoch": 0.36389111289031223,
+      "grad_norm": 1.0596287476120059,
+      "learning_rate": 1.4701455886116159e-05,
+      "loss": 1.6502,
+      "step": 2727
+    },
+    {
+      "epoch": 0.3640245529757139,
+      "grad_norm": 0.9990070809309942,
+      "learning_rate": 1.4697640981976347e-05,
+      "loss": 1.6093,
+      "step": 2728
+    },
+    {
+      "epoch": 0.36415799306111557,
+      "grad_norm": 1.1369821739015482,
+      "learning_rate": 1.4693825200370682e-05,
+      "loss": 1.6204,
+      "step": 2729
+    },
+    {
+      "epoch": 0.36429143314651724,
+      "grad_norm": 1.1486606111537232,
+      "learning_rate": 1.4690008542011912e-05,
+      "loss": 1.656,
+      "step": 2730
+    },
+    {
+      "epoch": 0.36442487323191886,
+      "grad_norm": 1.038173101624409,
+      "learning_rate": 1.4686191007612945e-05,
+      "loss": 1.5957,
+      "step": 2731
+    },
+    {
+      "epoch": 0.3645583133173205,
+      "grad_norm": 1.3383555292824745,
+      "learning_rate": 1.4682372597886851e-05,
+      "loss": 1.6269,
+      "step": 2732
+    },
+    {
+      "epoch": 0.3646917534027222,
+      "grad_norm": 1.1725746335257135,
+      "learning_rate": 1.4678553313546869e-05,
+      "loss": 1.6646,
+      "step": 2733
+    },
+    {
+      "epoch": 0.3648251934881238,
+      "grad_norm": 1.0117764785870433,
+      "learning_rate": 1.4674733155306392e-05,
+      "loss": 1.5871,
+      "step": 2734
+    },
+    {
+      "epoch": 0.3649586335735255,
+      "grad_norm": 1.139180055289969,
+      "learning_rate": 1.4670912123878987e-05,
+      "loss": 1.666,
+      "step": 2735
+    },
+    {
+      "epoch": 0.36509207365892715,
+      "grad_norm": 1.160600503319934,
+      "learning_rate": 1.4667090219978376e-05,
+      "loss": 1.6344,
+      "step": 2736
+    },
+    {
+      "epoch": 0.3652255137443288,
+      "grad_norm": 0.9804628531594025,
+      "learning_rate": 1.4663267444318451e-05,
+      "loss": 1.6504,
+      "step": 2737
+    },
+    {
+      "epoch": 0.36535895382973044,
+      "grad_norm": 1.2244168997082292,
+      "learning_rate": 1.465944379761326e-05,
+      "loss": 1.6267,
+      "step": 2738
+    },
+    {
+      "epoch": 0.3654923939151321,
+      "grad_norm": 1.142784430348776,
+      "learning_rate": 1.4655619280577015e-05,
+      "loss": 1.6029,
+      "step": 2739
+    },
+    {
+      "epoch": 0.3656258340005338,
+      "grad_norm": 1.1856642918086469,
+      "learning_rate": 1.46517938939241e-05,
+      "loss": 1.6084,
+      "step": 2740
+    },
+    {
+      "epoch": 0.3657592740859354,
+      "grad_norm": 1.1684003107608707,
+      "learning_rate": 1.4647967638369043e-05,
+      "loss": 1.673,
+      "step": 2741
+    },
+    {
+      "epoch": 0.36589271417133706,
+      "grad_norm": 0.9808643566324687,
+      "learning_rate": 1.4644140514626553e-05,
+      "loss": 1.6068,
+      "step": 2742
+    },
+    {
+      "epoch": 0.36602615425673873,
+      "grad_norm": 1.0096442361014633,
+      "learning_rate": 1.4640312523411491e-05,
+      "loss": 1.6395,
+      "step": 2743
+    },
+    {
+      "epoch": 0.3661595943421404,
+      "grad_norm": 1.0898350041048794,
+      "learning_rate": 1.4636483665438878e-05,
+      "loss": 1.7007,
+      "step": 2744
+    },
+    {
+      "epoch": 0.366293034427542,
+      "grad_norm": 0.9501423520521707,
+      "learning_rate": 1.4632653941423911e-05,
+      "loss": 1.5732,
+      "step": 2745
+    },
+    {
+      "epoch": 0.3664264745129437,
+      "grad_norm": 1.001779649980241,
+      "learning_rate": 1.4628823352081929e-05,
+      "loss": 1.6178,
+      "step": 2746
+    },
+    {
+      "epoch": 0.36655991459834536,
+      "grad_norm": 1.0032387536870389,
+      "learning_rate": 1.4624991898128445e-05,
+      "loss": 1.6319,
+      "step": 2747
+    },
+    {
+      "epoch": 0.366693354683747,
+      "grad_norm": 0.9476089503337908,
+      "learning_rate": 1.4621159580279129e-05,
+      "loss": 1.5986,
+      "step": 2748
+    },
+    {
+      "epoch": 0.36682679476914865,
+      "grad_norm": 0.9672551178709096,
+      "learning_rate": 1.4617326399249823e-05,
+      "loss": 1.6172,
+      "step": 2749
+    },
+    {
+      "epoch": 0.3669602348545503,
+      "grad_norm": 0.9999506532379091,
+      "learning_rate": 1.4613492355756514e-05,
+      "loss": 1.6411,
+      "step": 2750
+    },
+    {
+      "epoch": 0.367093674939952,
+      "grad_norm": 1.0229999120124786,
+      "learning_rate": 1.460965745051536e-05,
+      "loss": 1.6105,
+      "step": 2751
+    },
+    {
+      "epoch": 0.3672271150253536,
+      "grad_norm": 1.008641986491956,
+      "learning_rate": 1.4605821684242674e-05,
+      "loss": 1.6731,
+      "step": 2752
+    },
+    {
+      "epoch": 0.36736055511075527,
+      "grad_norm": 1.042758475874373,
+      "learning_rate": 1.460198505765494e-05,
+      "loss": 1.554,
+      "step": 2753
+    },
+    {
+      "epoch": 0.36749399519615694,
+      "grad_norm": 1.1472564831402514,
+      "learning_rate": 1.4598147571468791e-05,
+      "loss": 1.6019,
+      "step": 2754
+    },
+    {
+      "epoch": 0.36762743528155856,
+      "grad_norm": 0.9737879253423936,
+      "learning_rate": 1.4594309226401027e-05,
+      "loss": 1.6319,
+      "step": 2755
+    },
+    {
+      "epoch": 0.3677608753669602,
+      "grad_norm": 0.9632772149963865,
+      "learning_rate": 1.4590470023168611e-05,
+      "loss": 1.5865,
+      "step": 2756
+    },
+    {
+      "epoch": 0.3678943154523619,
+      "grad_norm": 1.1241458546313505,
+      "learning_rate": 1.4586629962488654e-05,
+      "loss": 1.6882,
+      "step": 2757
+    },
+    {
+      "epoch": 0.36802775553776357,
+      "grad_norm": 1.0034203806903836,
+      "learning_rate": 1.4582789045078445e-05,
+      "loss": 1.6177,
+      "step": 2758
+    },
+    {
+      "epoch": 0.3681611956231652,
+      "grad_norm": 0.9619488800895231,
+      "learning_rate": 1.4578947271655417e-05,
+      "loss": 1.6104,
+      "step": 2759
+    },
+    {
+      "epoch": 0.36829463570856685,
+      "grad_norm": 1.0707252512115273,
+      "learning_rate": 1.4575104642937175e-05,
+      "loss": 1.5266,
+      "step": 2760
+    },
+    {
+      "epoch": 0.3684280757939685,
+      "grad_norm": 1.0221303660590428,
+      "learning_rate": 1.4571261159641472e-05,
+      "loss": 1.6157,
+      "step": 2761
+    },
+    {
+      "epoch": 0.36856151587937014,
+      "grad_norm": 1.1856740576654383,
+      "learning_rate": 1.4567416822486233e-05,
+      "loss": 1.5843,
+      "step": 2762
+    },
+    {
+      "epoch": 0.3686949559647718,
+      "grad_norm": 1.057753625247735,
+      "learning_rate": 1.456357163218953e-05,
+      "loss": 1.6437,
+      "step": 2763
+    },
+    {
+      "epoch": 0.3688283960501735,
+      "grad_norm": 1.167606648878092,
+      "learning_rate": 1.4559725589469609e-05,
+      "loss": 1.62,
+      "step": 2764
+    },
+    {
+      "epoch": 0.36896183613557515,
+      "grad_norm": 1.0332270116796345,
+      "learning_rate": 1.4555878695044859e-05,
+      "loss": 1.5791,
+      "step": 2765
+    },
+    {
+      "epoch": 0.36909527622097676,
+      "grad_norm": 1.0273288838594588,
+      "learning_rate": 1.4552030949633839e-05,
+      "loss": 1.6575,
+      "step": 2766
+    },
+    {
+      "epoch": 0.36922871630637843,
+      "grad_norm": 1.0275198082813934,
+      "learning_rate": 1.4548182353955267e-05,
+      "loss": 1.626,
+      "step": 2767
+    },
+    {
+      "epoch": 0.3693621563917801,
+      "grad_norm": 1.009577517124903,
+      "learning_rate": 1.4544332908728011e-05,
+      "loss": 1.6139,
+      "step": 2768
+    },
+    {
+      "epoch": 0.3694955964771817,
+      "grad_norm": 0.9669422230954393,
+      "learning_rate": 1.454048261467111e-05,
+      "loss": 1.6449,
+      "step": 2769
+    },
+    {
+      "epoch": 0.3696290365625834,
+      "grad_norm": 1.2874584717461675,
+      "learning_rate": 1.453663147250375e-05,
+      "loss": 1.5875,
+      "step": 2770
+    },
+    {
+      "epoch": 0.36976247664798506,
+      "grad_norm": 1.025068484975409,
+      "learning_rate": 1.4532779482945283e-05,
+      "loss": 1.6252,
+      "step": 2771
+    },
+    {
+      "epoch": 0.36989591673338673,
+      "grad_norm": 1.005806168229964,
+      "learning_rate": 1.4528926646715215e-05,
+      "loss": 1.6093,
+      "step": 2772
+    },
+    {
+      "epoch": 0.37002935681878835,
+      "grad_norm": 1.208605409648711,
+      "learning_rate": 1.4525072964533213e-05,
+      "loss": 1.6543,
+      "step": 2773
+    },
+    {
+      "epoch": 0.37016279690419,
+      "grad_norm": 1.0895321534730344,
+      "learning_rate": 1.4521218437119105e-05,
+      "loss": 1.6162,
+      "step": 2774
+    },
+    {
+      "epoch": 0.3702962369895917,
+      "grad_norm": 0.9832287315998608,
+      "learning_rate": 1.4517363065192865e-05,
+      "loss": 1.5716,
+      "step": 2775
+    },
+    {
+      "epoch": 0.3704296770749933,
+      "grad_norm": 1.2767029129562657,
+      "learning_rate": 1.4513506849474639e-05,
+      "loss": 1.6252,
+      "step": 2776
+    },
+    {
+      "epoch": 0.370563117160395,
+      "grad_norm": 1.0660572092116536,
+      "learning_rate": 1.450964979068472e-05,
+      "loss": 1.5639,
+      "step": 2777
+    },
+    {
+      "epoch": 0.37069655724579664,
+      "grad_norm": 1.0431072621084843,
+      "learning_rate": 1.4505791889543565e-05,
+      "loss": 1.6989,
+      "step": 2778
+    },
+    {
+      "epoch": 0.3708299973311983,
+      "grad_norm": 1.1890314142284257,
+      "learning_rate": 1.4501933146771785e-05,
+      "loss": 1.6471,
+      "step": 2779
+    },
+    {
+      "epoch": 0.37096343741659993,
+      "grad_norm": 1.1319054159207953,
+      "learning_rate": 1.449807356309015e-05,
+      "loss": 1.5851,
+      "step": 2780
+    },
+    {
+      "epoch": 0.3710968775020016,
+      "grad_norm": 1.0145901459250226,
+      "learning_rate": 1.4494213139219587e-05,
+      "loss": 1.6279,
+      "step": 2781
+    },
+    {
+      "epoch": 0.37123031758740327,
+      "grad_norm": 0.9600581548472692,
+      "learning_rate": 1.4490351875881174e-05,
+      "loss": 1.6292,
+      "step": 2782
+    },
+    {
+      "epoch": 0.3713637576728049,
+      "grad_norm": 1.1450325048073626,
+      "learning_rate": 1.448648977379616e-05,
+      "loss": 1.5946,
+      "step": 2783
+    },
+    {
+      "epoch": 0.37149719775820655,
+      "grad_norm": 0.945594443926737,
+      "learning_rate": 1.4482626833685936e-05,
+      "loss": 1.6552,
+      "step": 2784
+    },
+    {
+      "epoch": 0.3716306378436082,
+      "grad_norm": 1.168278859108492,
+      "learning_rate": 1.4478763056272054e-05,
+      "loss": 1.6365,
+      "step": 2785
+    },
+    {
+      "epoch": 0.3717640779290099,
+      "grad_norm": 1.108937200609111,
+      "learning_rate": 1.4474898442276226e-05,
+      "loss": 1.5822,
+      "step": 2786
+    },
+    {
+      "epoch": 0.3718975180144115,
+      "grad_norm": 0.9571836543010801,
+      "learning_rate": 1.447103299242032e-05,
+      "loss": 1.5651,
+      "step": 2787
+    },
+    {
+      "epoch": 0.3720309580998132,
+      "grad_norm": 10.157439322526868,
+      "learning_rate": 1.4467166707426357e-05,
+      "loss": 1.6417,
+      "step": 2788
+    },
+    {
+      "epoch": 0.37216439818521485,
+      "grad_norm": 1.0795077239195907,
+      "learning_rate": 1.4463299588016508e-05,
+      "loss": 1.6007,
+      "step": 2789
+    },
+    {
+      "epoch": 0.37229783827061647,
+      "grad_norm": 1.02760083176124,
+      "learning_rate": 1.4459431634913118e-05,
+      "loss": 1.5834,
+      "step": 2790
+    },
+    {
+      "epoch": 0.37243127835601814,
+      "grad_norm": 1.052624541464461,
+      "learning_rate": 1.4455562848838666e-05,
+      "loss": 1.5611,
+      "step": 2791
+    },
+    {
+      "epoch": 0.3725647184414198,
+      "grad_norm": 1.0230420491562466,
+      "learning_rate": 1.4451693230515807e-05,
+      "loss": 1.6217,
+      "step": 2792
+    },
+    {
+      "epoch": 0.3726981585268215,
+      "grad_norm": 0.9764342262244122,
+      "learning_rate": 1.4447822780667332e-05,
+      "loss": 1.6139,
+      "step": 2793
+    },
+    {
+      "epoch": 0.3728315986122231,
+      "grad_norm": 1.128603754261654,
+      "learning_rate": 1.4443951500016207e-05,
+      "loss": 1.6407,
+      "step": 2794
+    },
+    {
+      "epoch": 0.37296503869762476,
+      "grad_norm": 1.1192050512936182,
+      "learning_rate": 1.4440079389285535e-05,
+      "loss": 1.6054,
+      "step": 2795
+    },
+    {
+      "epoch": 0.37309847878302643,
+      "grad_norm": 1.0094200838680423,
+      "learning_rate": 1.4436206449198584e-05,
+      "loss": 1.584,
+      "step": 2796
+    },
+    {
+      "epoch": 0.3732319188684281,
+      "grad_norm": 1.0820268405383773,
+      "learning_rate": 1.443233268047878e-05,
+      "loss": 1.6154,
+      "step": 2797
+    },
+    {
+      "epoch": 0.3733653589538297,
+      "grad_norm": 1.0454550915071419,
+      "learning_rate": 1.4428458083849693e-05,
+      "loss": 1.5913,
+      "step": 2798
+    },
+    {
+      "epoch": 0.3734987990392314,
+      "grad_norm": 1.0114425617387492,
+      "learning_rate": 1.4424582660035059e-05,
+      "loss": 1.6053,
+      "step": 2799
+    },
+    {
+      "epoch": 0.37363223912463306,
+      "grad_norm": 1.0867361007007368,
+      "learning_rate": 1.4420706409758753e-05,
+      "loss": 1.6393,
+      "step": 2800
+    },
+    {
+      "epoch": 0.3737656792100347,
+      "grad_norm": 1.1845819474365904,
+      "learning_rate": 1.4416829333744828e-05,
+      "loss": 1.6132,
+      "step": 2801
+    },
+    {
+      "epoch": 0.37389911929543634,
+      "grad_norm": 1.0205258166513704,
+      "learning_rate": 1.4412951432717467e-05,
+      "loss": 1.6227,
+      "step": 2802
+    },
+    {
+      "epoch": 0.374032559380838,
+      "grad_norm": 1.0438780709981974,
+      "learning_rate": 1.4409072707401024e-05,
+      "loss": 1.636,
+      "step": 2803
+    },
+    {
+      "epoch": 0.3741659994662397,
+      "grad_norm": 1.0589699296179769,
+      "learning_rate": 1.4405193158519998e-05,
+      "loss": 1.6119,
+      "step": 2804
+    },
+    {
+      "epoch": 0.3742994395516413,
+      "grad_norm": 0.9619237433348188,
+      "learning_rate": 1.4401312786799044e-05,
+      "loss": 1.6066,
+      "step": 2805
+    },
+    {
+      "epoch": 0.37443287963704297,
+      "grad_norm": 1.0797086642870346,
+      "learning_rate": 1.4397431592962974e-05,
+      "loss": 1.5915,
+      "step": 2806
+    },
+    {
+      "epoch": 0.37456631972244464,
+      "grad_norm": 0.9662518764136966,
+      "learning_rate": 1.439354957773675e-05,
+      "loss": 1.6207,
+      "step": 2807
+    },
+    {
+      "epoch": 0.37469975980784626,
+      "grad_norm": 1.1237578053297503,
+      "learning_rate": 1.4389666741845485e-05,
+      "loss": 1.6194,
+      "step": 2808
+    },
+    {
+      "epoch": 0.3748331998932479,
+      "grad_norm": 0.9840735377034793,
+      "learning_rate": 1.4385783086014451e-05,
+      "loss": 1.6028,
+      "step": 2809
+    },
+    {
+      "epoch": 0.3749666399786496,
+      "grad_norm": 1.0118695314963424,
+      "learning_rate": 1.4381898610969071e-05,
+      "loss": 1.5789,
+      "step": 2810
+    },
+    {
+      "epoch": 0.37510008006405127,
+      "grad_norm": 1.2580488826742735,
+      "learning_rate": 1.437801331743492e-05,
+      "loss": 1.6479,
+      "step": 2811
+    },
+    {
+      "epoch": 0.3752335201494529,
+      "grad_norm": 1.0726283441727469,
+      "learning_rate": 1.4374127206137727e-05,
+      "loss": 1.6275,
+      "step": 2812
+    },
+    {
+      "epoch": 0.37536696023485455,
+      "grad_norm": 1.0383966783564913,
+      "learning_rate": 1.4370240277803374e-05,
+      "loss": 1.6389,
+      "step": 2813
+    },
+    {
+      "epoch": 0.3755004003202562,
+      "grad_norm": 1.323435758142834,
+      "learning_rate": 1.4366352533157893e-05,
+      "loss": 1.5959,
+      "step": 2814
+    },
+    {
+      "epoch": 0.37563384040565784,
+      "grad_norm": 1.1012025284439741,
+      "learning_rate": 1.4362463972927472e-05,
+      "loss": 1.5828,
+      "step": 2815
+    },
+    {
+      "epoch": 0.3757672804910595,
+      "grad_norm": 1.1171657662365868,
+      "learning_rate": 1.4358574597838448e-05,
+      "loss": 1.6421,
+      "step": 2816
+    },
+    {
+      "epoch": 0.3759007205764612,
+      "grad_norm": 1.0544176694187513,
+      "learning_rate": 1.4354684408617316e-05,
+      "loss": 1.6215,
+      "step": 2817
+    },
+    {
+      "epoch": 0.37603416066186285,
+      "grad_norm": 1.0414279460927407,
+      "learning_rate": 1.4350793405990716e-05,
+      "loss": 1.5807,
+      "step": 2818
+    },
+    {
+      "epoch": 0.37616760074726446,
+      "grad_norm": 1.1376881831993566,
+      "learning_rate": 1.434690159068544e-05,
+      "loss": 1.6808,
+      "step": 2819
+    },
+    {
+      "epoch": 0.37630104083266613,
+      "grad_norm": 1.0250101070023419,
+      "learning_rate": 1.4343008963428442e-05,
+      "loss": 1.6021,
+      "step": 2820
+    },
+    {
+      "epoch": 0.3764344809180678,
+      "grad_norm": 1.1002348920804779,
+      "learning_rate": 1.4339115524946815e-05,
+      "loss": 1.6074,
+      "step": 2821
+    },
+    {
+      "epoch": 0.3765679210034694,
+      "grad_norm": 1.0123930431013084,
+      "learning_rate": 1.4335221275967812e-05,
+      "loss": 1.6426,
+      "step": 2822
+    },
+    {
+      "epoch": 0.3767013610888711,
+      "grad_norm": 1.2209497008197363,
+      "learning_rate": 1.433132621721883e-05,
+      "loss": 1.6471,
+      "step": 2823
+    },
+    {
+      "epoch": 0.37683480117427276,
+      "grad_norm": 1.0189131304238113,
+      "learning_rate": 1.432743034942743e-05,
+      "loss": 1.6218,
+      "step": 2824
+    },
+    {
+      "epoch": 0.37696824125967443,
+      "grad_norm": 1.0178791123141788,
+      "learning_rate": 1.4323533673321304e-05,
+      "loss": 1.6242,
+      "step": 2825
+    },
+    {
+      "epoch": 0.37710168134507605,
+      "grad_norm": 0.9991762214209088,
+      "learning_rate": 1.4319636189628316e-05,
+      "loss": 1.6042,
+      "step": 2826
+    },
+    {
+      "epoch": 0.3772351214304777,
+      "grad_norm": 1.016918051574137,
+      "learning_rate": 1.431573789907647e-05,
+      "loss": 1.6255,
+      "step": 2827
+    },
+    {
+      "epoch": 0.3773685615158794,
+      "grad_norm": 0.968894964593842,
+      "learning_rate": 1.4311838802393921e-05,
+      "loss": 1.5977,
+      "step": 2828
+    },
+    {
+      "epoch": 0.377502001601281,
+      "grad_norm": 0.9868572481236255,
+      "learning_rate": 1.4307938900308971e-05,
+      "loss": 1.6564,
+      "step": 2829
+    },
+    {
+      "epoch": 0.37763544168668267,
+      "grad_norm": 1.6711098134309899,
+      "learning_rate": 1.4304038193550087e-05,
+      "loss": 1.6036,
+      "step": 2830
+    },
+    {
+      "epoch": 0.37776888177208434,
+      "grad_norm": 1.2720681532176408,
+      "learning_rate": 1.4300136682845874e-05,
+      "loss": 1.6172,
+      "step": 2831
+    },
+    {
+      "epoch": 0.377902321857486,
+      "grad_norm": 0.9972549450196833,
+      "learning_rate": 1.4296234368925082e-05,
+      "loss": 1.6582,
+      "step": 2832
+    },
+    {
+      "epoch": 0.3780357619428876,
+      "grad_norm": 1.033670572949725,
+      "learning_rate": 1.4292331252516627e-05,
+      "loss": 1.598,
+      "step": 2833
+    },
+    {
+      "epoch": 0.3781692020282893,
+      "grad_norm": 1.3209056443094034,
+      "learning_rate": 1.4288427334349562e-05,
+      "loss": 1.5872,
+      "step": 2834
+    },
+    {
+      "epoch": 0.37830264211369097,
+      "grad_norm": 1.00727971762687,
+      "learning_rate": 1.4284522615153098e-05,
+      "loss": 1.6444,
+      "step": 2835
+    },
+    {
+      "epoch": 0.3784360821990926,
+      "grad_norm": 1.1341291826363633,
+      "learning_rate": 1.4280617095656591e-05,
+      "loss": 1.6396,
+      "step": 2836
+    },
+    {
+      "epoch": 0.37856952228449425,
+      "grad_norm": 1.088875710596739,
+      "learning_rate": 1.4276710776589546e-05,
+      "loss": 1.5717,
+      "step": 2837
+    },
+    {
+      "epoch": 0.3787029623698959,
+      "grad_norm": 1.5266027482348503,
+      "learning_rate": 1.4272803658681622e-05,
+      "loss": 1.6019,
+      "step": 2838
+    },
+    {
+      "epoch": 0.3788364024552976,
+      "grad_norm": 1.1012146674109902,
+      "learning_rate": 1.4268895742662618e-05,
+      "loss": 1.6891,
+      "step": 2839
+    },
+    {
+      "epoch": 0.3789698425406992,
+      "grad_norm": 1.2571923095061812,
+      "learning_rate": 1.4264987029262497e-05,
+      "loss": 1.637,
+      "step": 2840
+    },
+    {
+      "epoch": 0.3791032826261009,
+      "grad_norm": 1.0068568438867096,
+      "learning_rate": 1.4261077519211358e-05,
+      "loss": 1.5796,
+      "step": 2841
+    },
+    {
+      "epoch": 0.37923672271150255,
+      "grad_norm": 1.008009465786665,
+      "learning_rate": 1.4257167213239451e-05,
+      "loss": 1.636,
+      "step": 2842
+    },
+    {
+      "epoch": 0.37937016279690416,
+      "grad_norm": 1.0065398200515008,
+      "learning_rate": 1.4253256112077176e-05,
+      "loss": 1.6736,
+      "step": 2843
+    },
+    {
+      "epoch": 0.37950360288230583,
+      "grad_norm": 0.9492387619872094,
+      "learning_rate": 1.4249344216455085e-05,
+      "loss": 1.632,
+      "step": 2844
+    },
+    {
+      "epoch": 0.3796370429677075,
+      "grad_norm": 0.9873644452435473,
+      "learning_rate": 1.4245431527103879e-05,
+      "loss": 1.5646,
+      "step": 2845
+    },
+    {
+      "epoch": 0.3797704830531092,
+      "grad_norm": 0.9940129192913516,
+      "learning_rate": 1.4241518044754397e-05,
+      "loss": 1.6068,
+      "step": 2846
+    },
+    {
+      "epoch": 0.3799039231385108,
+      "grad_norm": 0.9786673958035172,
+      "learning_rate": 1.4237603770137635e-05,
+      "loss": 1.6508,
+      "step": 2847
+    },
+    {
+      "epoch": 0.38003736322391246,
+      "grad_norm": 1.0669464495332486,
+      "learning_rate": 1.4233688703984737e-05,
+      "loss": 1.5973,
+      "step": 2848
+    },
+    {
+      "epoch": 0.38017080330931413,
+      "grad_norm": 0.9699464474368997,
+      "learning_rate": 1.4229772847026993e-05,
+      "loss": 1.5928,
+      "step": 2849
+    },
+    {
+      "epoch": 0.38030424339471575,
+      "grad_norm": 0.9602245627811605,
+      "learning_rate": 1.4225856199995836e-05,
+      "loss": 1.6669,
+      "step": 2850
+    },
+    {
+      "epoch": 0.3804376834801174,
+      "grad_norm": 1.0000207901051725,
+      "learning_rate": 1.4221938763622856e-05,
+      "loss": 1.5871,
+      "step": 2851
+    },
+    {
+      "epoch": 0.3805711235655191,
+      "grad_norm": 1.1884986045536179,
+      "learning_rate": 1.4218020538639782e-05,
+      "loss": 1.6184,
+      "step": 2852
+    },
+    {
+      "epoch": 0.38070456365092076,
+      "grad_norm": 1.0365928654144692,
+      "learning_rate": 1.4214101525778495e-05,
+      "loss": 1.665,
+      "step": 2853
+    },
+    {
+      "epoch": 0.3808380037363224,
+      "grad_norm": 0.9983952327474825,
+      "learning_rate": 1.4210181725771026e-05,
+      "loss": 1.6284,
+      "step": 2854
+    },
+    {
+      "epoch": 0.38097144382172404,
+      "grad_norm": 1.0013533383795308,
+      "learning_rate": 1.420626113934954e-05,
+      "loss": 1.6141,
+      "step": 2855
+    },
+    {
+      "epoch": 0.3811048839071257,
+      "grad_norm": 1.1423554042414863,
+      "learning_rate": 1.4202339767246367e-05,
+      "loss": 1.6336,
+      "step": 2856
+    },
+    {
+      "epoch": 0.38123832399252733,
+      "grad_norm": 0.9703881563875031,
+      "learning_rate": 1.419841761019397e-05,
+      "loss": 1.6393,
+      "step": 2857
+    },
+    {
+      "epoch": 0.381371764077929,
+      "grad_norm": 0.9642652272603265,
+      "learning_rate": 1.4194494668924966e-05,
+      "loss": 1.6384,
+      "step": 2858
+    },
+    {
+      "epoch": 0.38150520416333067,
+      "grad_norm": 1.131226033110534,
+      "learning_rate": 1.419057094417211e-05,
+      "loss": 1.6435,
+      "step": 2859
+    },
+    {
+      "epoch": 0.38163864424873234,
+      "grad_norm": 0.9917335521919108,
+      "learning_rate": 1.4186646436668312e-05,
+      "loss": 1.6202,
+      "step": 2860
+    },
+    {
+      "epoch": 0.38177208433413395,
+      "grad_norm": 1.097375636507755,
+      "learning_rate": 1.4182721147146633e-05,
+      "loss": 1.7061,
+      "step": 2861
+    },
+    {
+      "epoch": 0.3819055244195356,
+      "grad_norm": 0.9975696907135774,
+      "learning_rate": 1.4178795076340259e-05,
+      "loss": 1.6118,
+      "step": 2862
+    },
+    {
+      "epoch": 0.3820389645049373,
+      "grad_norm": 0.9870041753998313,
+      "learning_rate": 1.4174868224982547e-05,
+      "loss": 1.6137,
+      "step": 2863
+    },
+    {
+      "epoch": 0.38217240459033897,
+      "grad_norm": 1.165292789080929,
+      "learning_rate": 1.417094059380698e-05,
+      "loss": 1.6004,
+      "step": 2864
+    },
+    {
+      "epoch": 0.3823058446757406,
+      "grad_norm": 1.145639325521145,
+      "learning_rate": 1.4167012183547203e-05,
+      "loss": 1.6066,
+      "step": 2865
+    },
+    {
+      "epoch": 0.38243928476114225,
+      "grad_norm": 1.1503077171125726,
+      "learning_rate": 1.4163082994936988e-05,
+      "loss": 1.615,
+      "step": 2866
+    },
+    {
+      "epoch": 0.3825727248465439,
+      "grad_norm": 1.0404542214973824,
+      "learning_rate": 1.4159153028710268e-05,
+      "loss": 1.6169,
+      "step": 2867
+    },
+    {
+      "epoch": 0.38270616493194554,
+      "grad_norm": 0.996004145579702,
+      "learning_rate": 1.415522228560112e-05,
+      "loss": 1.6029,
+      "step": 2868
+    },
+    {
+      "epoch": 0.3828396050173472,
+      "grad_norm": 1.2254062428716475,
+      "learning_rate": 1.4151290766343752e-05,
+      "loss": 1.6211,
+      "step": 2869
+    },
+    {
+      "epoch": 0.3829730451027489,
+      "grad_norm": 0.992077891417151,
+      "learning_rate": 1.4147358471672541e-05,
+      "loss": 1.5512,
+      "step": 2870
+    },
+    {
+      "epoch": 0.38310648518815055,
+      "grad_norm": 0.952569959666935,
+      "learning_rate": 1.4143425402321982e-05,
+      "loss": 1.6052,
+      "step": 2871
+    },
+    {
+      "epoch": 0.38323992527355216,
+      "grad_norm": 1.2306890442084908,
+      "learning_rate": 1.4139491559026732e-05,
+      "loss": 1.5909,
+      "step": 2872
+    },
+    {
+      "epoch": 0.38337336535895383,
+      "grad_norm": 0.9759054239574055,
+      "learning_rate": 1.4135556942521594e-05,
+      "loss": 1.6061,
+      "step": 2873
+    },
+    {
+      "epoch": 0.3835068054443555,
+      "grad_norm": 0.9913047688419279,
+      "learning_rate": 1.4131621553541499e-05,
+      "loss": 1.6104,
+      "step": 2874
+    },
+    {
+      "epoch": 0.3836402455297571,
+      "grad_norm": 1.246242626432296,
+      "learning_rate": 1.412768539282154e-05,
+      "loss": 1.6091,
+      "step": 2875
+    },
+    {
+      "epoch": 0.3837736856151588,
+      "grad_norm": 0.9433807734418067,
+      "learning_rate": 1.4123748461096942e-05,
+      "loss": 1.6305,
+      "step": 2876
+    },
+    {
+      "epoch": 0.38390712570056046,
+      "grad_norm": 0.982212397205281,
+      "learning_rate": 1.4119810759103088e-05,
+      "loss": 1.6221,
+      "step": 2877
+    },
+    {
+      "epoch": 0.38404056578596213,
+      "grad_norm": 1.0575127644185438,
+      "learning_rate": 1.4115872287575485e-05,
+      "loss": 1.602,
+      "step": 2878
+    },
+    {
+      "epoch": 0.38417400587136374,
+      "grad_norm": 1.1374999128456018,
+      "learning_rate": 1.4111933047249802e-05,
+      "loss": 1.6081,
+      "step": 2879
+    },
+    {
+      "epoch": 0.3843074459567654,
+      "grad_norm": 9.571339510483169,
+      "learning_rate": 1.410799303886184e-05,
+      "loss": 1.6055,
+      "step": 2880
+    },
+    {
+      "epoch": 0.3844408860421671,
+      "grad_norm": 1.145304347974641,
+      "learning_rate": 1.4104052263147553e-05,
+      "loss": 1.6612,
+      "step": 2881
+    },
+    {
+      "epoch": 0.3845743261275687,
+      "grad_norm": 1.099466074032018,
+      "learning_rate": 1.4100110720843025e-05,
+      "loss": 1.5704,
+      "step": 2882
+    },
+    {
+      "epoch": 0.38470776621297037,
+      "grad_norm": 1.0509418929003724,
+      "learning_rate": 1.4096168412684497e-05,
+      "loss": 1.648,
+      "step": 2883
+    },
+    {
+      "epoch": 0.38484120629837204,
+      "grad_norm": 1.1230867470007317,
+      "learning_rate": 1.409222533940835e-05,
+      "loss": 1.6401,
+      "step": 2884
+    },
+    {
+      "epoch": 0.3849746463837737,
+      "grad_norm": 1.007807394086856,
+      "learning_rate": 1.4088281501751095e-05,
+      "loss": 1.6452,
+      "step": 2885
+    },
+    {
+      "epoch": 0.3851080864691753,
+      "grad_norm": 1.061247642675541,
+      "learning_rate": 1.408433690044941e-05,
+      "loss": 1.6049,
+      "step": 2886
+    },
+    {
+      "epoch": 0.385241526554577,
+      "grad_norm": 1.0601502148961404,
+      "learning_rate": 1.4080391536240088e-05,
+      "loss": 1.6375,
+      "step": 2887
+    },
+    {
+      "epoch": 0.38537496663997867,
+      "grad_norm": 0.9817836229390783,
+      "learning_rate": 1.4076445409860086e-05,
+      "loss": 1.6106,
+      "step": 2888
+    },
+    {
+      "epoch": 0.3855084067253803,
+      "grad_norm": 1.2826160394488593,
+      "learning_rate": 1.4072498522046494e-05,
+      "loss": 1.6531,
+      "step": 2889
+    },
+    {
+      "epoch": 0.38564184681078195,
+      "grad_norm": 1.201976231365931,
+      "learning_rate": 1.4068550873536543e-05,
+      "loss": 1.6206,
+      "step": 2890
+    },
+    {
+      "epoch": 0.3857752868961836,
+      "grad_norm": 1.026733475170239,
+      "learning_rate": 1.4064602465067613e-05,
+      "loss": 1.6505,
+      "step": 2891
+    },
+    {
+      "epoch": 0.3859087269815853,
+      "grad_norm": 1.236834192825807,
+      "learning_rate": 1.406065329737722e-05,
+      "loss": 1.6342,
+      "step": 2892
+    },
+    {
+      "epoch": 0.3860421670669869,
+      "grad_norm": 1.150715310952092,
+      "learning_rate": 1.4056703371203023e-05,
+      "loss": 1.6205,
+      "step": 2893
+    },
+    {
+      "epoch": 0.3861756071523886,
+      "grad_norm": 1.0642046173241948,
+      "learning_rate": 1.4052752687282824e-05,
+      "loss": 1.6052,
+      "step": 2894
+    },
+    {
+      "epoch": 0.38630904723779025,
+      "grad_norm": 1.0978042763206004,
+      "learning_rate": 1.4048801246354567e-05,
+      "loss": 1.6058,
+      "step": 2895
+    },
+    {
+      "epoch": 0.38644248732319186,
+      "grad_norm": 1.177880101416564,
+      "learning_rate": 1.4044849049156328e-05,
+      "loss": 1.6127,
+      "step": 2896
+    },
+    {
+      "epoch": 0.38657592740859353,
+      "grad_norm": 1.017132075267852,
+      "learning_rate": 1.4040896096426346e-05,
+      "loss": 1.6143,
+      "step": 2897
+    },
+    {
+      "epoch": 0.3867093674939952,
+      "grad_norm": 1.0501232630820656,
+      "learning_rate": 1.4036942388902976e-05,
+      "loss": 1.5509,
+      "step": 2898
+    },
+    {
+      "epoch": 0.3868428075793969,
+      "grad_norm": 1.2571254734715342,
+      "learning_rate": 1.403298792732473e-05,
+      "loss": 1.6453,
+      "step": 2899
+    },
+    {
+      "epoch": 0.3869762476647985,
+      "grad_norm": 0.9875221484898435,
+      "learning_rate": 1.4029032712430262e-05,
+      "loss": 1.596,
+      "step": 2900
+    },
+    {
+      "epoch": 0.38710968775020016,
+      "grad_norm": 1.1285508086710352,
+      "learning_rate": 1.4025076744958348e-05,
+      "loss": 1.6236,
+      "step": 2901
+    },
+    {
+      "epoch": 0.38724312783560183,
+      "grad_norm": 1.2197035948818218,
+      "learning_rate": 1.4021120025647932e-05,
+      "loss": 1.5706,
+      "step": 2902
+    },
+    {
+      "epoch": 0.38737656792100345,
+      "grad_norm": 1.0890256309645825,
+      "learning_rate": 1.4017162555238072e-05,
+      "loss": 1.5571,
+      "step": 2903
+    },
+    {
+      "epoch": 0.3875100080064051,
+      "grad_norm": 1.080695243503158,
+      "learning_rate": 1.4013204334467987e-05,
+      "loss": 1.6454,
+      "step": 2904
+    },
+    {
+      "epoch": 0.3876434480918068,
+      "grad_norm": 0.9814101368260878,
+      "learning_rate": 1.4009245364077024e-05,
+      "loss": 1.6153,
+      "step": 2905
+    },
+    {
+      "epoch": 0.38777688817720846,
+      "grad_norm": 0.9904485921133632,
+      "learning_rate": 1.4005285644804673e-05,
+      "loss": 1.6276,
+      "step": 2906
+    },
+    {
+      "epoch": 0.38791032826261007,
+      "grad_norm": 0.9972089960040826,
+      "learning_rate": 1.4001325177390565e-05,
+      "loss": 1.6205,
+      "step": 2907
+    },
+    {
+      "epoch": 0.38804376834801174,
+      "grad_norm": 1.2730345239020744,
+      "learning_rate": 1.3997363962574473e-05,
+      "loss": 1.6412,
+      "step": 2908
+    },
+    {
+      "epoch": 0.3881772084334134,
+      "grad_norm": 0.9675278590077582,
+      "learning_rate": 1.3993402001096304e-05,
+      "loss": 1.6351,
+      "step": 2909
+    },
+    {
+      "epoch": 0.388310648518815,
+      "grad_norm": 1.0292534070839274,
+      "learning_rate": 1.3989439293696105e-05,
+      "loss": 1.6089,
+      "step": 2910
+    },
+    {
+      "epoch": 0.3884440886042167,
+      "grad_norm": 1.019759996442744,
+      "learning_rate": 1.3985475841114071e-05,
+      "loss": 1.63,
+      "step": 2911
+    },
+    {
+      "epoch": 0.38857752868961837,
+      "grad_norm": 0.9807744844897855,
+      "learning_rate": 1.3981511644090523e-05,
+      "loss": 1.6103,
+      "step": 2912
+    },
+    {
+      "epoch": 0.38871096877502004,
+      "grad_norm": 1.025209057661364,
+      "learning_rate": 1.3977546703365934e-05,
+      "loss": 1.5834,
+      "step": 2913
+    },
+    {
+      "epoch": 0.38884440886042165,
+      "grad_norm": 1.089409574851773,
+      "learning_rate": 1.3973581019680906e-05,
+      "loss": 1.707,
+      "step": 2914
+    },
+    {
+      "epoch": 0.3889778489458233,
+      "grad_norm": 0.9724916772280182,
+      "learning_rate": 1.3969614593776184e-05,
+      "loss": 1.6204,
+      "step": 2915
+    },
+    {
+      "epoch": 0.389111289031225,
+      "grad_norm": 1.007702700819211,
+      "learning_rate": 1.3965647426392653e-05,
+      "loss": 1.6122,
+      "step": 2916
+    },
+    {
+      "epoch": 0.3892447291166266,
+      "grad_norm": 0.9419094811442577,
+      "learning_rate": 1.396167951827133e-05,
+      "loss": 1.5903,
+      "step": 2917
+    },
+    {
+      "epoch": 0.3893781692020283,
+      "grad_norm": 1.0374726700116226,
+      "learning_rate": 1.395771087015338e-05,
+      "loss": 1.6066,
+      "step": 2918
+    },
+    {
+      "epoch": 0.38951160928742995,
+      "grad_norm": 1.0328706825533176,
+      "learning_rate": 1.3953741482780102e-05,
+      "loss": 1.5566,
+      "step": 2919
+    },
+    {
+      "epoch": 0.3896450493728316,
+      "grad_norm": 1.2728567621408549,
+      "learning_rate": 1.3949771356892927e-05,
+      "loss": 1.6384,
+      "step": 2920
+    },
+    {
+      "epoch": 0.38977848945823323,
+      "grad_norm": 1.0123347440431487,
+      "learning_rate": 1.3945800493233432e-05,
+      "loss": 1.5981,
+      "step": 2921
+    },
+    {
+      "epoch": 0.3899119295436349,
+      "grad_norm": 0.991927710941825,
+      "learning_rate": 1.3941828892543332e-05,
+      "loss": 1.6323,
+      "step": 2922
+    },
+    {
+      "epoch": 0.3900453696290366,
+      "grad_norm": 0.9860434334934471,
+      "learning_rate": 1.3937856555564472e-05,
+      "loss": 1.5993,
+      "step": 2923
+    },
+    {
+      "epoch": 0.3901788097144382,
+      "grad_norm": 1.1272050680874104,
+      "learning_rate": 1.3933883483038843e-05,
+      "loss": 1.6889,
+      "step": 2924
+    },
+    {
+      "epoch": 0.39031224979983986,
+      "grad_norm": 1.013248475393018,
+      "learning_rate": 1.392990967570857e-05,
+      "loss": 1.5832,
+      "step": 2925
+    },
+    {
+      "epoch": 0.39044568988524153,
+      "grad_norm": 0.9972907037837129,
+      "learning_rate": 1.392593513431591e-05,
+      "loss": 1.6312,
+      "step": 2926
+    },
+    {
+      "epoch": 0.3905791299706432,
+      "grad_norm": 1.0368488326806515,
+      "learning_rate": 1.392195985960327e-05,
+      "loss": 1.593,
+      "step": 2927
+    },
+    {
+      "epoch": 0.3907125700560448,
+      "grad_norm": 0.9818098293390851,
+      "learning_rate": 1.3917983852313174e-05,
+      "loss": 1.6771,
+      "step": 2928
+    },
+    {
+      "epoch": 0.3908460101414465,
+      "grad_norm": 0.9843368644401832,
+      "learning_rate": 1.391400711318831e-05,
+      "loss": 1.6287,
+      "step": 2929
+    },
+    {
+      "epoch": 0.39097945022684816,
+      "grad_norm": 1.1575289949746161,
+      "learning_rate": 1.3910029642971473e-05,
+      "loss": 1.5908,
+      "step": 2930
+    },
+    {
+      "epoch": 0.3911128903122498,
+      "grad_norm": 1.0086519987992577,
+      "learning_rate": 1.390605144240562e-05,
+      "loss": 1.5931,
+      "step": 2931
+    },
+    {
+      "epoch": 0.39124633039765144,
+      "grad_norm": 0.9903875800619729,
+      "learning_rate": 1.390207251223383e-05,
+      "loss": 1.5842,
+      "step": 2932
+    },
+    {
+      "epoch": 0.3913797704830531,
+      "grad_norm": 0.9577055345082642,
+      "learning_rate": 1.3898092853199318e-05,
+      "loss": 1.6194,
+      "step": 2933
+    },
+    {
+      "epoch": 0.3915132105684548,
+      "grad_norm": 1.2357946602483,
+      "learning_rate": 1.3894112466045448e-05,
+      "loss": 1.6087,
+      "step": 2934
+    },
+    {
+      "epoch": 0.3916466506538564,
+      "grad_norm": 1.0769210602839947,
+      "learning_rate": 1.3890131351515703e-05,
+      "loss": 1.6153,
+      "step": 2935
+    },
+    {
+      "epoch": 0.39178009073925807,
+      "grad_norm": 1.10534271606038,
+      "learning_rate": 1.388614951035371e-05,
+      "loss": 1.6172,
+      "step": 2936
+    },
+    {
+      "epoch": 0.39191353082465974,
+      "grad_norm": 1.0791728229209796,
+      "learning_rate": 1.3882166943303239e-05,
+      "loss": 1.5885,
+      "step": 2937
+    },
+    {
+      "epoch": 0.3920469709100614,
+      "grad_norm": 1.048943382922718,
+      "learning_rate": 1.387818365110818e-05,
+      "loss": 1.6466,
+      "step": 2938
+    },
+    {
+      "epoch": 0.392180410995463,
+      "grad_norm": 0.9775623637977353,
+      "learning_rate": 1.3874199634512568e-05,
+      "loss": 1.6842,
+      "step": 2939
+    },
+    {
+      "epoch": 0.3923138510808647,
+      "grad_norm": 1.2823226827585823,
+      "learning_rate": 1.3870214894260577e-05,
+      "loss": 1.6346,
+      "step": 2940
+    },
+    {
+      "epoch": 0.39244729116626637,
+      "grad_norm": 2.9554542149643392,
+      "learning_rate": 1.3866229431096506e-05,
+      "loss": 1.631,
+      "step": 2941
+    },
+    {
+      "epoch": 0.392580731251668,
+      "grad_norm": 1.4313180292149659,
+      "learning_rate": 1.3862243245764795e-05,
+      "loss": 1.5588,
+      "step": 2942
+    },
+    {
+      "epoch": 0.39271417133706965,
+      "grad_norm": 1.0226581878161995,
+      "learning_rate": 1.3858256339010021e-05,
+      "loss": 1.5696,
+      "step": 2943
+    },
+    {
+      "epoch": 0.3928476114224713,
+      "grad_norm": 1.203432785432225,
+      "learning_rate": 1.385426871157689e-05,
+      "loss": 1.5932,
+      "step": 2944
+    },
+    {
+      "epoch": 0.392981051507873,
+      "grad_norm": 1.2211552689992726,
+      "learning_rate": 1.3850280364210246e-05,
+      "loss": 1.6112,
+      "step": 2945
+    },
+    {
+      "epoch": 0.3931144915932746,
+      "grad_norm": 1.0952032810522179,
+      "learning_rate": 1.3846291297655066e-05,
+      "loss": 1.6057,
+      "step": 2946
+    },
+    {
+      "epoch": 0.3932479316786763,
+      "grad_norm": 1.1552204218064228,
+      "learning_rate": 1.3842301512656465e-05,
+      "loss": 1.6422,
+      "step": 2947
+    },
+    {
+      "epoch": 0.39338137176407795,
+      "grad_norm": 1.0183412684546411,
+      "learning_rate": 1.3838311009959686e-05,
+      "loss": 1.5689,
+      "step": 2948
+    },
+    {
+      "epoch": 0.39351481184947956,
+      "grad_norm": 1.047132191195357,
+      "learning_rate": 1.3834319790310113e-05,
+      "loss": 1.6462,
+      "step": 2949
+    },
+    {
+      "epoch": 0.39364825193488123,
+      "grad_norm": 0.9755539464048916,
+      "learning_rate": 1.3830327854453258e-05,
+      "loss": 1.6675,
+      "step": 2950
+    },
+    {
+      "epoch": 0.3937816920202829,
+      "grad_norm": 1.035910398586731,
+      "learning_rate": 1.3826335203134768e-05,
+      "loss": 1.6491,
+      "step": 2951
+    },
+    {
+      "epoch": 0.3939151321056846,
+      "grad_norm": 1.0231501045177207,
+      "learning_rate": 1.382234183710043e-05,
+      "loss": 1.644,
+      "step": 2952
+    },
+    {
+      "epoch": 0.3940485721910862,
+      "grad_norm": 1.0110762267018458,
+      "learning_rate": 1.3818347757096152e-05,
+      "loss": 1.6056,
+      "step": 2953
+    },
+    {
+      "epoch": 0.39418201227648786,
+      "grad_norm": 1.9202250806838923,
+      "learning_rate": 1.381435296386799e-05,
+      "loss": 1.5915,
+      "step": 2954
+    },
+    {
+      "epoch": 0.39431545236188953,
+      "grad_norm": 1.06141099974795,
+      "learning_rate": 1.381035745816212e-05,
+      "loss": 1.6049,
+      "step": 2955
+    },
+    {
+      "epoch": 0.39444889244729114,
+      "grad_norm": 1.0770580734734054,
+      "learning_rate": 1.3806361240724862e-05,
+      "loss": 1.6341,
+      "step": 2956
+    },
+    {
+      "epoch": 0.3945823325326928,
+      "grad_norm": 1.0439072213146445,
+      "learning_rate": 1.3802364312302659e-05,
+      "loss": 1.6512,
+      "step": 2957
+    },
+    {
+      "epoch": 0.3947157726180945,
+      "grad_norm": 1.0133974448550223,
+      "learning_rate": 1.3798366673642095e-05,
+      "loss": 1.6098,
+      "step": 2958
+    },
+    {
+      "epoch": 0.39484921270349616,
+      "grad_norm": 1.0228026592925512,
+      "learning_rate": 1.3794368325489881e-05,
+      "loss": 1.6292,
+      "step": 2959
+    },
+    {
+      "epoch": 0.39498265278889777,
+      "grad_norm": 1.2383549797377225,
+      "learning_rate": 1.3790369268592865e-05,
+      "loss": 1.6379,
+      "step": 2960
+    },
+    {
+      "epoch": 0.39511609287429944,
+      "grad_norm": 1.0527485547831446,
+      "learning_rate": 1.3786369503698024e-05,
+      "loss": 1.6094,
+      "step": 2961
+    },
+    {
+      "epoch": 0.3952495329597011,
+      "grad_norm": 1.2613117991115332,
+      "learning_rate": 1.3782369031552469e-05,
+      "loss": 1.6241,
+      "step": 2962
+    },
+    {
+      "epoch": 0.3953829730451027,
+      "grad_norm": 1.0530326147165556,
+      "learning_rate": 1.3778367852903443e-05,
+      "loss": 1.6379,
+      "step": 2963
+    },
+    {
+      "epoch": 0.3955164131305044,
+      "grad_norm": 2.2697531891999043,
+      "learning_rate": 1.3774365968498323e-05,
+      "loss": 1.6122,
+      "step": 2964
+    },
+    {
+      "epoch": 0.39564985321590607,
+      "grad_norm": 1.0088488119923664,
+      "learning_rate": 1.377036337908461e-05,
+      "loss": 1.6069,
+      "step": 2965
+    },
+    {
+      "epoch": 0.39578329330130774,
+      "grad_norm": 1.1919777045176008,
+      "learning_rate": 1.3766360085409947e-05,
+      "loss": 1.6282,
+      "step": 2966
+    },
+    {
+      "epoch": 0.39591673338670935,
+      "grad_norm": 1.0394509374153975,
+      "learning_rate": 1.37623560882221e-05,
+      "loss": 1.6261,
+      "step": 2967
+    },
+    {
+      "epoch": 0.396050173472111,
+      "grad_norm": 1.4778314816259879,
+      "learning_rate": 1.375835138826897e-05,
+      "loss": 1.5999,
+      "step": 2968
+    },
+    {
+      "epoch": 0.3961836135575127,
+      "grad_norm": 1.076372619162072,
+      "learning_rate": 1.3754345986298594e-05,
+      "loss": 1.63,
+      "step": 2969
+    },
+    {
+      "epoch": 0.3963170536429143,
+      "grad_norm": 1.0029057794554,
+      "learning_rate": 1.3750339883059132e-05,
+      "loss": 1.5545,
+      "step": 2970
+    },
+    {
+      "epoch": 0.396450493728316,
+      "grad_norm": 0.9763005134113368,
+      "learning_rate": 1.3746333079298882e-05,
+      "loss": 1.6233,
+      "step": 2971
+    },
+    {
+      "epoch": 0.39658393381371765,
+      "grad_norm": 0.9997766940306253,
+      "learning_rate": 1.374232557576626e-05,
+      "loss": 1.6042,
+      "step": 2972
+    },
+    {
+      "epoch": 0.3967173738991193,
+      "grad_norm": 1.081981802944108,
+      "learning_rate": 1.3738317373209833e-05,
+      "loss": 1.6644,
+      "step": 2973
+    },
+    {
+      "epoch": 0.39685081398452093,
+      "grad_norm": 0.9784210671800242,
+      "learning_rate": 1.3734308472378281e-05,
+      "loss": 1.6678,
+      "step": 2974
+    },
+    {
+      "epoch": 0.3969842540699226,
+      "grad_norm": 1.0089465304994234,
+      "learning_rate": 1.3730298874020424e-05,
+      "loss": 1.5872,
+      "step": 2975
+    },
+    {
+      "epoch": 0.3971176941553243,
+      "grad_norm": 0.9897744114028344,
+      "learning_rate": 1.372628857888521e-05,
+      "loss": 1.6752,
+      "step": 2976
+    },
+    {
+      "epoch": 0.3972511342407259,
+      "grad_norm": 1.0113015313370548,
+      "learning_rate": 1.3722277587721713e-05,
+      "loss": 1.6196,
+      "step": 2977
+    },
+    {
+      "epoch": 0.39738457432612756,
+      "grad_norm": 1.248353362601478,
+      "learning_rate": 1.3718265901279144e-05,
+      "loss": 1.6446,
+      "step": 2978
+    },
+    {
+      "epoch": 0.39751801441152923,
+      "grad_norm": 0.9976910628273125,
+      "learning_rate": 1.3714253520306835e-05,
+      "loss": 1.6559,
+      "step": 2979
+    },
+    {
+      "epoch": 0.3976514544969309,
+      "grad_norm": 0.9395020657372218,
+      "learning_rate": 1.371024044555426e-05,
+      "loss": 1.5507,
+      "step": 2980
+    },
+    {
+      "epoch": 0.3977848945823325,
+      "grad_norm": 1.059218486991219,
+      "learning_rate": 1.3706226677771011e-05,
+      "loss": 1.6096,
+      "step": 2981
+    },
+    {
+      "epoch": 0.3979183346677342,
+      "grad_norm": 1.1913088141870016,
+      "learning_rate": 1.3702212217706819e-05,
+      "loss": 1.605,
+      "step": 2982
+    },
+    {
+      "epoch": 0.39805177475313586,
+      "grad_norm": 1.0112436026820137,
+      "learning_rate": 1.3698197066111529e-05,
+      "loss": 1.6008,
+      "step": 2983
+    },
+    {
+      "epoch": 0.39818521483853747,
+      "grad_norm": 1.2284860321693414,
+      "learning_rate": 1.3694181223735138e-05,
+      "loss": 1.6154,
+      "step": 2984
+    },
+    {
+      "epoch": 0.39831865492393914,
+      "grad_norm": 1.0098640773316505,
+      "learning_rate": 1.3690164691327752e-05,
+      "loss": 1.6321,
+      "step": 2985
+    },
+    {
+      "epoch": 0.3984520950093408,
+      "grad_norm": 0.9717411694382783,
+      "learning_rate": 1.3686147469639616e-05,
+      "loss": 1.6847,
+      "step": 2986
+    },
+    {
+      "epoch": 0.3985855350947425,
+      "grad_norm": 0.9284579466010013,
+      "learning_rate": 1.3682129559421102e-05,
+      "loss": 1.5653,
+      "step": 2987
+    },
+    {
+      "epoch": 0.3987189751801441,
+      "grad_norm": 1.0645254671871938,
+      "learning_rate": 1.3678110961422708e-05,
+      "loss": 1.5928,
+      "step": 2988
+    },
+    {
+      "epoch": 0.39885241526554577,
+      "grad_norm": 1.1099047312189834,
+      "learning_rate": 1.3674091676395067e-05,
+      "loss": 1.5968,
+      "step": 2989
+    },
+    {
+      "epoch": 0.39898585535094744,
+      "grad_norm": 1.0039809381615434,
+      "learning_rate": 1.3670071705088925e-05,
+      "loss": 1.5968,
+      "step": 2990
+    },
+    {
+      "epoch": 0.39911929543634905,
+      "grad_norm": 1.1183712241615666,
+      "learning_rate": 1.366605104825518e-05,
+      "loss": 1.5859,
+      "step": 2991
+    },
+    {
+      "epoch": 0.3992527355217507,
+      "grad_norm": 1.2317478828162558,
+      "learning_rate": 1.3662029706644834e-05,
+      "loss": 1.633,
+      "step": 2992
+    },
+    {
+      "epoch": 0.3993861756071524,
+      "grad_norm": 1.0317396851316578,
+      "learning_rate": 1.3658007681009038e-05,
+      "loss": 1.604,
+      "step": 2993
+    },
+    {
+      "epoch": 0.39951961569255406,
+      "grad_norm": 1.0355968618370723,
+      "learning_rate": 1.365398497209905e-05,
+      "loss": 1.6271,
+      "step": 2994
+    },
+    {
+      "epoch": 0.3996530557779557,
+      "grad_norm": 0.9930994963511364,
+      "learning_rate": 1.3649961580666274e-05,
+      "loss": 1.6196,
+      "step": 2995
+    },
+    {
+      "epoch": 0.39978649586335735,
+      "grad_norm": 1.2269862375772342,
+      "learning_rate": 1.364593750746223e-05,
+      "loss": 1.6219,
+      "step": 2996
+    },
+    {
+      "epoch": 0.399919935948759,
+      "grad_norm": 0.9700028478438667,
+      "learning_rate": 1.3641912753238572e-05,
+      "loss": 1.6564,
+      "step": 2997
+    },
+    {
+      "epoch": 0.40005337603416063,
+      "grad_norm": 0.9545800135062432,
+      "learning_rate": 1.3637887318747077e-05,
+      "loss": 1.6116,
+      "step": 2998
+    },
+    {
+      "epoch": 0.4001868161195623,
+      "grad_norm": 0.9570154971534128,
+      "learning_rate": 1.3633861204739647e-05,
+      "loss": 1.6712,
+      "step": 2999
+    },
+    {
+      "epoch": 0.400320256204964,
+      "grad_norm": 1.2079133599715546,
+      "learning_rate": 1.3629834411968317e-05,
+      "loss": 1.6293,
+      "step": 3000
+    },
+    {
+      "epoch": 0.40045369629036565,
+      "grad_norm": 0.9836480600801262,
+      "learning_rate": 1.362580694118525e-05,
+      "loss": 1.5987,
+      "step": 3001
+    },
+    {
+      "epoch": 0.40058713637576726,
+      "grad_norm": 0.9910032459354412,
+      "learning_rate": 1.3621778793142721e-05,
+      "loss": 1.6064,
+      "step": 3002
+    },
+    {
+      "epoch": 0.40072057646116893,
+      "grad_norm": 1.0967755414857783,
+      "learning_rate": 1.3617749968593155e-05,
+      "loss": 1.6292,
+      "step": 3003
+    },
+    {
+      "epoch": 0.4008540165465706,
+      "grad_norm": 1.0003163865850948,
+      "learning_rate": 1.3613720468289083e-05,
+      "loss": 1.6362,
+      "step": 3004
+    },
+    {
+      "epoch": 0.40098745663197227,
+      "grad_norm": 1.0315705480717272,
+      "learning_rate": 1.3609690292983171e-05,
+      "loss": 1.5829,
+      "step": 3005
+    },
+    {
+      "epoch": 0.4011208967173739,
+      "grad_norm": 0.976139700381007,
+      "learning_rate": 1.3605659443428208e-05,
+      "loss": 1.6224,
+      "step": 3006
+    },
+    {
+      "epoch": 0.40125433680277556,
+      "grad_norm": 0.9736571196522579,
+      "learning_rate": 1.3601627920377114e-05,
+      "loss": 1.5872,
+      "step": 3007
+    },
+    {
+      "epoch": 0.40138777688817723,
+      "grad_norm": 1.0002994090288835,
+      "learning_rate": 1.359759572458293e-05,
+      "loss": 1.6186,
+      "step": 3008
+    },
+    {
+      "epoch": 0.40152121697357884,
+      "grad_norm": 1.045079222117039,
+      "learning_rate": 1.3593562856798828e-05,
+      "loss": 1.5786,
+      "step": 3009
+    },
+    {
+      "epoch": 0.4016546570589805,
+      "grad_norm": 0.9688502047059884,
+      "learning_rate": 1.3589529317778097e-05,
+      "loss": 1.6257,
+      "step": 3010
+    },
+    {
+      "epoch": 0.4017880971443822,
+      "grad_norm": 1.0897993308963512,
+      "learning_rate": 1.3585495108274155e-05,
+      "loss": 1.6415,
+      "step": 3011
+    },
+    {
+      "epoch": 0.40192153722978385,
+      "grad_norm": 1.024133361111547,
+      "learning_rate": 1.3581460229040552e-05,
+      "loss": 1.6256,
+      "step": 3012
+    },
+    {
+      "epoch": 0.40205497731518547,
+      "grad_norm": 1.0901551656265192,
+      "learning_rate": 1.3577424680830953e-05,
+      "loss": 1.6085,
+      "step": 3013
+    },
+    {
+      "epoch": 0.40218841740058714,
+      "grad_norm": 1.0074753288531135,
+      "learning_rate": 1.3573388464399158e-05,
+      "loss": 1.6159,
+      "step": 3014
+    },
+    {
+      "epoch": 0.4023218574859888,
+      "grad_norm": 1.0808434730186502,
+      "learning_rate": 1.3569351580499077e-05,
+      "loss": 1.6081,
+      "step": 3015
+    },
+    {
+      "epoch": 0.4024552975713904,
+      "grad_norm": 1.3390539579090508,
+      "learning_rate": 1.3565314029884764e-05,
+      "loss": 1.6476,
+      "step": 3016
+    },
+    {
+      "epoch": 0.4025887376567921,
+      "grad_norm": 1.1325494027036593,
+      "learning_rate": 1.356127581331038e-05,
+      "loss": 1.6408,
+      "step": 3017
+    },
+    {
+      "epoch": 0.40272217774219377,
+      "grad_norm": 1.1561856323405204,
+      "learning_rate": 1.3557236931530223e-05,
+      "loss": 1.6112,
+      "step": 3018
+    },
+    {
+      "epoch": 0.40285561782759544,
+      "grad_norm": 1.0096598053584744,
+      "learning_rate": 1.3553197385298704e-05,
+      "loss": 1.6228,
+      "step": 3019
+    },
+    {
+      "epoch": 0.40298905791299705,
+      "grad_norm": 0.9619782469364923,
+      "learning_rate": 1.3549157175370374e-05,
+      "loss": 1.6378,
+      "step": 3020
+    },
+    {
+      "epoch": 0.4031224979983987,
+      "grad_norm": 1.0828538858146166,
+      "learning_rate": 1.3545116302499888e-05,
+      "loss": 1.6662,
+      "step": 3021
+    },
+    {
+      "epoch": 0.4032559380838004,
+      "grad_norm": 1.1612286498560225,
+      "learning_rate": 1.3541074767442039e-05,
+      "loss": 1.6038,
+      "step": 3022
+    },
+    {
+      "epoch": 0.403389378169202,
+      "grad_norm": 1.0469641391213618,
+      "learning_rate": 1.3537032570951742e-05,
+      "loss": 1.5947,
+      "step": 3023
+    },
+    {
+      "epoch": 0.4035228182546037,
+      "grad_norm": 1.1061755741793078,
+      "learning_rate": 1.353298971378403e-05,
+      "loss": 1.541,
+      "step": 3024
+    },
+    {
+      "epoch": 0.40365625834000535,
+      "grad_norm": 1.255991398258131,
+      "learning_rate": 1.3528946196694067e-05,
+      "loss": 1.6509,
+      "step": 3025
+    },
+    {
+      "epoch": 0.403789698425407,
+      "grad_norm": 1.2128774118932317,
+      "learning_rate": 1.352490202043713e-05,
+      "loss": 1.6228,
+      "step": 3026
+    },
+    {
+      "epoch": 0.40392313851080863,
+      "grad_norm": 0.9952887623145266,
+      "learning_rate": 1.3520857185768627e-05,
+      "loss": 1.647,
+      "step": 3027
+    },
+    {
+      "epoch": 0.4040565785962103,
+      "grad_norm": 0.9867845852620436,
+      "learning_rate": 1.3516811693444092e-05,
+      "loss": 1.5929,
+      "step": 3028
+    },
+    {
+      "epoch": 0.404190018681612,
+      "grad_norm": 1.0634687182480442,
+      "learning_rate": 1.3512765544219168e-05,
+      "loss": 1.6571,
+      "step": 3029
+    },
+    {
+      "epoch": 0.4043234587670136,
+      "grad_norm": 0.9923107764534385,
+      "learning_rate": 1.350871873884964e-05,
+      "loss": 1.6509,
+      "step": 3030
+    },
+    {
+      "epoch": 0.40445689885241526,
+      "grad_norm": 0.9418612013435159,
+      "learning_rate": 1.3504671278091396e-05,
+      "loss": 1.6459,
+      "step": 3031
+    },
+    {
+      "epoch": 0.40459033893781693,
+      "grad_norm": 0.9743708699323149,
+      "learning_rate": 1.3500623162700464e-05,
+      "loss": 1.5773,
+      "step": 3032
+    },
+    {
+      "epoch": 0.4047237790232186,
+      "grad_norm": 1.0021337250013098,
+      "learning_rate": 1.3496574393432978e-05,
+      "loss": 1.6072,
+      "step": 3033
+    },
+    {
+      "epoch": 0.4048572191086202,
+      "grad_norm": 0.9905407512325389,
+      "learning_rate": 1.3492524971045202e-05,
+      "loss": 1.5955,
+      "step": 3034
+    },
+    {
+      "epoch": 0.4049906591940219,
+      "grad_norm": 0.9829445518512612,
+      "learning_rate": 1.3488474896293531e-05,
+      "loss": 1.6473,
+      "step": 3035
+    },
+    {
+      "epoch": 0.40512409927942356,
+      "grad_norm": 1.071551407117045,
+      "learning_rate": 1.3484424169934465e-05,
+      "loss": 1.6365,
+      "step": 3036
+    },
+    {
+      "epoch": 0.40525753936482517,
+      "grad_norm": 1.0858770844558168,
+      "learning_rate": 1.3480372792724636e-05,
+      "loss": 1.5908,
+      "step": 3037
+    },
+    {
+      "epoch": 0.40539097945022684,
+      "grad_norm": 1.097690642719979,
+      "learning_rate": 1.3476320765420794e-05,
+      "loss": 1.6341,
+      "step": 3038
+    },
+    {
+      "epoch": 0.4055244195356285,
+      "grad_norm": 0.9550194281013512,
+      "learning_rate": 1.3472268088779814e-05,
+      "loss": 1.6558,
+      "step": 3039
+    },
+    {
+      "epoch": 0.4056578596210302,
+      "grad_norm": 0.9831862531483857,
+      "learning_rate": 1.3468214763558686e-05,
+      "loss": 1.5536,
+      "step": 3040
+    },
+    {
+      "epoch": 0.4057912997064318,
+      "grad_norm": 1.1319350865381137,
+      "learning_rate": 1.3464160790514531e-05,
+      "loss": 1.6586,
+      "step": 3041
+    },
+    {
+      "epoch": 0.40592473979183347,
+      "grad_norm": 0.988139658682674,
+      "learning_rate": 1.3460106170404579e-05,
+      "loss": 1.5955,
+      "step": 3042
+    },
+    {
+      "epoch": 0.40605817987723514,
+      "grad_norm": 1.0084348562604524,
+      "learning_rate": 1.3456050903986189e-05,
+      "loss": 1.6381,
+      "step": 3043
+    },
+    {
+      "epoch": 0.40619161996263675,
+      "grad_norm": 1.0391432220005692,
+      "learning_rate": 1.3451994992016839e-05,
+      "loss": 1.5705,
+      "step": 3044
+    },
+    {
+      "epoch": 0.4063250600480384,
+      "grad_norm": 0.9843503035286154,
+      "learning_rate": 1.3447938435254127e-05,
+      "loss": 1.5664,
+      "step": 3045
+    },
+    {
+      "epoch": 0.4064585001334401,
+      "grad_norm": 0.9833405027126961,
+      "learning_rate": 1.3443881234455772e-05,
+      "loss": 1.6586,
+      "step": 3046
+    },
+    {
+      "epoch": 0.40659194021884176,
+      "grad_norm": 0.9898970019998985,
+      "learning_rate": 1.3439823390379609e-05,
+      "loss": 1.6347,
+      "step": 3047
+    },
+    {
+      "epoch": 0.4067253803042434,
+      "grad_norm": 1.0000597042602213,
+      "learning_rate": 1.3435764903783605e-05,
+      "loss": 1.6239,
+      "step": 3048
+    },
+    {
+      "epoch": 0.40685882038964505,
+      "grad_norm": 0.9928896355998765,
+      "learning_rate": 1.3431705775425835e-05,
+      "loss": 1.5948,
+      "step": 3049
+    },
+    {
+      "epoch": 0.4069922604750467,
+      "grad_norm": 1.2336188070965246,
+      "learning_rate": 1.3427646006064492e-05,
+      "loss": 1.5953,
+      "step": 3050
+    },
+    {
+      "epoch": 0.40712570056044833,
+      "grad_norm": 1.0034026633786906,
+      "learning_rate": 1.3423585596457906e-05,
+      "loss": 1.6316,
+      "step": 3051
+    },
+    {
+      "epoch": 0.40725914064585,
+      "grad_norm": 2.1481053784351336,
+      "learning_rate": 1.3419524547364506e-05,
+      "loss": 1.6295,
+      "step": 3052
+    },
+    {
+      "epoch": 0.4073925807312517,
+      "grad_norm": 1.1897086059631425,
+      "learning_rate": 1.3415462859542856e-05,
+      "loss": 1.5635,
+      "step": 3053
+    },
+    {
+      "epoch": 0.40752602081665334,
+      "grad_norm": 1.023418328395939,
+      "learning_rate": 1.3411400533751628e-05,
+      "loss": 1.6415,
+      "step": 3054
+    },
+    {
+      "epoch": 0.40765946090205496,
+      "grad_norm": 0.98213814874442,
+      "learning_rate": 1.3407337570749622e-05,
+      "loss": 1.6428,
+      "step": 3055
+    },
+    {
+      "epoch": 0.40779290098745663,
+      "grad_norm": 1.283169457369164,
+      "learning_rate": 1.3403273971295749e-05,
+      "loss": 1.6235,
+      "step": 3056
+    },
+    {
+      "epoch": 0.4079263410728583,
+      "grad_norm": 0.9925266060021142,
+      "learning_rate": 1.339920973614905e-05,
+      "loss": 1.621,
+      "step": 3057
+    },
+    {
+      "epoch": 0.4080597811582599,
+      "grad_norm": 0.9614672523532065,
+      "learning_rate": 1.3395144866068673e-05,
+      "loss": 1.5774,
+      "step": 3058
+    },
+    {
+      "epoch": 0.4081932212436616,
+      "grad_norm": 0.9815628673629643,
+      "learning_rate": 1.3391079361813888e-05,
+      "loss": 1.5849,
+      "step": 3059
+    },
+    {
+      "epoch": 0.40832666132906326,
+      "grad_norm": 1.0926187808935628,
+      "learning_rate": 1.338701322414409e-05,
+      "loss": 1.6139,
+      "step": 3060
+    },
+    {
+      "epoch": 0.4084601014144649,
+      "grad_norm": 0.9687836831287908,
+      "learning_rate": 1.3382946453818782e-05,
+      "loss": 1.5821,
+      "step": 3061
+    },
+    {
+      "epoch": 0.40859354149986654,
+      "grad_norm": 1.2615662453906389,
+      "learning_rate": 1.3378879051597594e-05,
+      "loss": 1.6187,
+      "step": 3062
+    },
+    {
+      "epoch": 0.4087269815852682,
+      "grad_norm": 1.143934867223233,
+      "learning_rate": 1.3374811018240268e-05,
+      "loss": 1.6114,
+      "step": 3063
+    },
+    {
+      "epoch": 0.4088604216706699,
+      "grad_norm": 0.9474085642193837,
+      "learning_rate": 1.3370742354506667e-05,
+      "loss": 1.6128,
+      "step": 3064
+    },
+    {
+      "epoch": 0.4089938617560715,
+      "grad_norm": 0.9872797898071304,
+      "learning_rate": 1.3366673061156775e-05,
+      "loss": 1.6521,
+      "step": 3065
+    },
+    {
+      "epoch": 0.40912730184147317,
+      "grad_norm": 1.3660459817420605,
+      "learning_rate": 1.3362603138950681e-05,
+      "loss": 1.5566,
+      "step": 3066
+    },
+    {
+      "epoch": 0.40926074192687484,
+      "grad_norm": 1.0188587037661874,
+      "learning_rate": 1.335853258864861e-05,
+      "loss": 1.6181,
+      "step": 3067
+    },
+    {
+      "epoch": 0.4093941820122765,
+      "grad_norm": 0.9730036168336966,
+      "learning_rate": 1.3354461411010887e-05,
+      "loss": 1.6294,
+      "step": 3068
+    },
+    {
+      "epoch": 0.4095276220976781,
+      "grad_norm": 1.0088183488719165,
+      "learning_rate": 1.3350389606797966e-05,
+      "loss": 1.599,
+      "step": 3069
+    },
+    {
+      "epoch": 0.4096610621830798,
+      "grad_norm": 0.9662028481495397,
+      "learning_rate": 1.3346317176770409e-05,
+      "loss": 1.6385,
+      "step": 3070
+    },
+    {
+      "epoch": 0.40979450226848146,
+      "grad_norm": 0.9917563809796904,
+      "learning_rate": 1.3342244121688905e-05,
+      "loss": 1.6225,
+      "step": 3071
+    },
+    {
+      "epoch": 0.4099279423538831,
+      "grad_norm": 0.9830125253461315,
+      "learning_rate": 1.3338170442314254e-05,
+      "loss": 1.6107,
+      "step": 3072
+    },
+    {
+      "epoch": 0.41006138243928475,
+      "grad_norm": 0.9826012171089086,
+      "learning_rate": 1.3334096139407368e-05,
+      "loss": 1.5969,
+      "step": 3073
+    },
+    {
+      "epoch": 0.4101948225246864,
+      "grad_norm": 1.3174009396986306,
+      "learning_rate": 1.3330021213729283e-05,
+      "loss": 1.5944,
+      "step": 3074
+    },
+    {
+      "epoch": 0.4103282626100881,
+      "grad_norm": 0.971753779897976,
+      "learning_rate": 1.332594566604115e-05,
+      "loss": 1.6118,
+      "step": 3075
+    },
+    {
+      "epoch": 0.4104617026954897,
+      "grad_norm": 0.9913292187671285,
+      "learning_rate": 1.3321869497104233e-05,
+      "loss": 1.6234,
+      "step": 3076
+    },
+    {
+      "epoch": 0.4105951427808914,
+      "grad_norm": 1.1276395807046564,
+      "learning_rate": 1.3317792707679915e-05,
+      "loss": 1.6527,
+      "step": 3077
+    },
+    {
+      "epoch": 0.41072858286629305,
+      "grad_norm": 0.9668291003462994,
+      "learning_rate": 1.3313715298529697e-05,
+      "loss": 1.6186,
+      "step": 3078
+    },
+    {
+      "epoch": 0.4108620229516947,
+      "grad_norm": 0.995376867257694,
+      "learning_rate": 1.3309637270415185e-05,
+      "loss": 1.6667,
+      "step": 3079
+    },
+    {
+      "epoch": 0.41099546303709633,
+      "grad_norm": 1.0129402139207564,
+      "learning_rate": 1.3305558624098116e-05,
+      "loss": 1.6623,
+      "step": 3080
+    },
+    {
+      "epoch": 0.411128903122498,
+      "grad_norm": 1.1770451338395234,
+      "learning_rate": 1.3301479360340329e-05,
+      "loss": 1.6662,
+      "step": 3081
+    },
+    {
+      "epoch": 0.41126234320789967,
+      "grad_norm": 1.0658383068003052,
+      "learning_rate": 1.3297399479903787e-05,
+      "loss": 1.6057,
+      "step": 3082
+    },
+    {
+      "epoch": 0.4113957832933013,
+      "grad_norm": 1.1694647133067295,
+      "learning_rate": 1.3293318983550563e-05,
+      "loss": 1.6122,
+      "step": 3083
+    },
+    {
+      "epoch": 0.41152922337870296,
+      "grad_norm": 0.942161713707869,
+      "learning_rate": 1.3289237872042851e-05,
+      "loss": 1.606,
+      "step": 3084
+    },
+    {
+      "epoch": 0.41166266346410463,
+      "grad_norm": 0.9757692839225567,
+      "learning_rate": 1.3285156146142954e-05,
+      "loss": 1.6343,
+      "step": 3085
+    },
+    {
+      "epoch": 0.4117961035495063,
+      "grad_norm": 0.9486089863227384,
+      "learning_rate": 1.3281073806613289e-05,
+      "loss": 1.5431,
+      "step": 3086
+    },
+    {
+      "epoch": 0.4119295436349079,
+      "grad_norm": 0.9732494030371487,
+      "learning_rate": 1.3276990854216396e-05,
+      "loss": 1.6314,
+      "step": 3087
+    },
+    {
+      "epoch": 0.4120629837203096,
+      "grad_norm": 1.1111387987861114,
+      "learning_rate": 1.3272907289714918e-05,
+      "loss": 1.6326,
+      "step": 3088
+    },
+    {
+      "epoch": 0.41219642380571125,
+      "grad_norm": 0.9652044439453622,
+      "learning_rate": 1.3268823113871627e-05,
+      "loss": 1.6681,
+      "step": 3089
+    },
+    {
+      "epoch": 0.41232986389111287,
+      "grad_norm": 0.9561565798940396,
+      "learning_rate": 1.3264738327449389e-05,
+      "loss": 1.5918,
+      "step": 3090
+    },
+    {
+      "epoch": 0.41246330397651454,
+      "grad_norm": 1.3152845877647086,
+      "learning_rate": 1.3260652931211207e-05,
+      "loss": 1.607,
+      "step": 3091
+    },
+    {
+      "epoch": 0.4125967440619162,
+      "grad_norm": 0.989537074682015,
+      "learning_rate": 1.325656692592018e-05,
+      "loss": 1.6011,
+      "step": 3092
+    },
+    {
+      "epoch": 0.4127301841473179,
+      "grad_norm": 0.9734140085573458,
+      "learning_rate": 1.3252480312339526e-05,
+      "loss": 1.648,
+      "step": 3093
+    },
+    {
+      "epoch": 0.4128636242327195,
+      "grad_norm": 0.9558460700959762,
+      "learning_rate": 1.3248393091232583e-05,
+      "loss": 1.5901,
+      "step": 3094
+    },
+    {
+      "epoch": 0.41299706431812117,
+      "grad_norm": 0.9913733635852103,
+      "learning_rate": 1.3244305263362796e-05,
+      "loss": 1.6253,
+      "step": 3095
+    },
+    {
+      "epoch": 0.41313050440352284,
+      "grad_norm": 1.1740036558110662,
+      "learning_rate": 1.3240216829493722e-05,
+      "loss": 1.5879,
+      "step": 3096
+    },
+    {
+      "epoch": 0.41326394448892445,
+      "grad_norm": 1.1257399618437767,
+      "learning_rate": 1.3236127790389036e-05,
+      "loss": 1.617,
+      "step": 3097
+    },
+    {
+      "epoch": 0.4133973845743261,
+      "grad_norm": 0.9756326130850466,
+      "learning_rate": 1.323203814681252e-05,
+      "loss": 1.596,
+      "step": 3098
+    },
+    {
+      "epoch": 0.4135308246597278,
+      "grad_norm": 0.91654844990653,
+      "learning_rate": 1.3227947899528081e-05,
+      "loss": 1.5737,
+      "step": 3099
+    },
+    {
+      "epoch": 0.41366426474512946,
+      "grad_norm": 1.0075079473682016,
+      "learning_rate": 1.3223857049299724e-05,
+      "loss": 1.5862,
+      "step": 3100
+    },
+    {
+      "epoch": 0.4137977048305311,
+      "grad_norm": 1.0128738725898727,
+      "learning_rate": 1.3219765596891576e-05,
+      "loss": 1.6742,
+      "step": 3101
+    },
+    {
+      "epoch": 0.41393114491593275,
+      "grad_norm": 4.5247192473377655,
+      "learning_rate": 1.321567354306787e-05,
+      "loss": 1.6274,
+      "step": 3102
+    },
+    {
+      "epoch": 0.4140645850013344,
+      "grad_norm": 1.2025566547291768,
+      "learning_rate": 1.3211580888592964e-05,
+      "loss": 1.6561,
+      "step": 3103
+    },
+    {
+      "epoch": 0.41419802508673603,
+      "grad_norm": 1.2057864440073822,
+      "learning_rate": 1.3207487634231308e-05,
+      "loss": 1.6046,
+      "step": 3104
+    },
+    {
+      "epoch": 0.4143314651721377,
+      "grad_norm": 1.032603380515401,
+      "learning_rate": 1.3203393780747482e-05,
+      "loss": 1.6494,
+      "step": 3105
+    },
+    {
+      "epoch": 0.4144649052575394,
+      "grad_norm": 0.984221055123714,
+      "learning_rate": 1.3199299328906173e-05,
+      "loss": 1.5627,
+      "step": 3106
+    },
+    {
+      "epoch": 0.41459834534294104,
+      "grad_norm": 1.0080633070266374,
+      "learning_rate": 1.3195204279472171e-05,
+      "loss": 1.6455,
+      "step": 3107
+    },
+    {
+      "epoch": 0.41473178542834266,
+      "grad_norm": 1.0172031612333243,
+      "learning_rate": 1.319110863321039e-05,
+      "loss": 1.6523,
+      "step": 3108
+    },
+    {
+      "epoch": 0.41486522551374433,
+      "grad_norm": 0.9819246274039858,
+      "learning_rate": 1.3187012390885844e-05,
+      "loss": 1.6192,
+      "step": 3109
+    },
+    {
+      "epoch": 0.414998665599146,
+      "grad_norm": 1.0218273012540824,
+      "learning_rate": 1.3182915553263676e-05,
+      "loss": 1.5381,
+      "step": 3110
+    },
+    {
+      "epoch": 0.4151321056845476,
+      "grad_norm": 1.1870404877073484,
+      "learning_rate": 1.3178818121109116e-05,
+      "loss": 1.5987,
+      "step": 3111
+    },
+    {
+      "epoch": 0.4152655457699493,
+      "grad_norm": 0.9665697428225051,
+      "learning_rate": 1.3174720095187527e-05,
+      "loss": 1.5904,
+      "step": 3112
+    },
+    {
+      "epoch": 0.41539898585535096,
+      "grad_norm": 1.0024983274761543,
+      "learning_rate": 1.3170621476264368e-05,
+      "loss": 1.5948,
+      "step": 3113
+    },
+    {
+      "epoch": 0.4155324259407526,
+      "grad_norm": 1.3283967395783622,
+      "learning_rate": 1.3166522265105216e-05,
+      "loss": 1.6199,
+      "step": 3114
+    },
+    {
+      "epoch": 0.41566586602615424,
+      "grad_norm": 1.0606323988381896,
+      "learning_rate": 1.3162422462475757e-05,
+      "loss": 1.5634,
+      "step": 3115
+    },
+    {
+      "epoch": 0.4157993061115559,
+      "grad_norm": 1.0009105274760155,
+      "learning_rate": 1.3158322069141788e-05,
+      "loss": 1.6536,
+      "step": 3116
+    },
+    {
+      "epoch": 0.4159327461969576,
+      "grad_norm": 0.998531188875352,
+      "learning_rate": 1.3154221085869215e-05,
+      "loss": 1.569,
+      "step": 3117
+    },
+    {
+      "epoch": 0.4160661862823592,
+      "grad_norm": 1.0318215900037766,
+      "learning_rate": 1.3150119513424054e-05,
+      "loss": 1.5927,
+      "step": 3118
+    },
+    {
+      "epoch": 0.41619962636776087,
+      "grad_norm": 1.09023864908779,
+      "learning_rate": 1.3146017352572435e-05,
+      "loss": 1.588,
+      "step": 3119
+    },
+    {
+      "epoch": 0.41633306645316254,
+      "grad_norm": 1.1300901695201984,
+      "learning_rate": 1.3141914604080593e-05,
+      "loss": 1.5848,
+      "step": 3120
+    },
+    {
+      "epoch": 0.4164665065385642,
+      "grad_norm": 0.9765411604695532,
+      "learning_rate": 1.3137811268714875e-05,
+      "loss": 1.5884,
+      "step": 3121
+    },
+    {
+      "epoch": 0.4165999466239658,
+      "grad_norm": 0.9917987982620767,
+      "learning_rate": 1.3133707347241735e-05,
+      "loss": 1.6489,
+      "step": 3122
+    },
+    {
+      "epoch": 0.4167333867093675,
+      "grad_norm": 1.0549323113529163,
+      "learning_rate": 1.3129602840427741e-05,
+      "loss": 1.6262,
+      "step": 3123
+    },
+    {
+      "epoch": 0.41686682679476916,
+      "grad_norm": 1.0110367235245075,
+      "learning_rate": 1.3125497749039574e-05,
+      "loss": 1.6089,
+      "step": 3124
+    },
+    {
+      "epoch": 0.4170002668801708,
+      "grad_norm": 1.1439338907133554,
+      "learning_rate": 1.312139207384401e-05,
+      "loss": 1.5526,
+      "step": 3125
+    },
+    {
+      "epoch": 0.41713370696557245,
+      "grad_norm": 1.2855762643940758,
+      "learning_rate": 1.3117285815607943e-05,
+      "loss": 1.6033,
+      "step": 3126
+    },
+    {
+      "epoch": 0.4172671470509741,
+      "grad_norm": 1.1679452250656743,
+      "learning_rate": 1.311317897509838e-05,
+      "loss": 1.6454,
+      "step": 3127
+    },
+    {
+      "epoch": 0.4174005871363758,
+      "grad_norm": 1.0069767600727495,
+      "learning_rate": 1.3109071553082426e-05,
+      "loss": 1.6761,
+      "step": 3128
+    },
+    {
+      "epoch": 0.4175340272217774,
+      "grad_norm": 0.9913179933467611,
+      "learning_rate": 1.3104963550327307e-05,
+      "loss": 1.5818,
+      "step": 3129
+    },
+    {
+      "epoch": 0.4176674673071791,
+      "grad_norm": 1.008746692632273,
+      "learning_rate": 1.3100854967600346e-05,
+      "loss": 1.5777,
+      "step": 3130
+    },
+    {
+      "epoch": 0.41780090739258074,
+      "grad_norm": 1.1253948107175336,
+      "learning_rate": 1.3096745805668985e-05,
+      "loss": 1.6108,
+      "step": 3131
+    },
+    {
+      "epoch": 0.41793434747798236,
+      "grad_norm": 1.0118014744129107,
+      "learning_rate": 1.309263606530076e-05,
+      "loss": 1.5923,
+      "step": 3132
+    },
+    {
+      "epoch": 0.41806778756338403,
+      "grad_norm": 1.0052085925079868,
+      "learning_rate": 1.3088525747263334e-05,
+      "loss": 1.617,
+      "step": 3133
+    },
+    {
+      "epoch": 0.4182012276487857,
+      "grad_norm": 1.028648913612512,
+      "learning_rate": 1.308441485232446e-05,
+      "loss": 1.5913,
+      "step": 3134
+    },
+    {
+      "epoch": 0.41833466773418737,
+      "grad_norm": 0.9714092683461435,
+      "learning_rate": 1.308030338125201e-05,
+      "loss": 1.5615,
+      "step": 3135
+    },
+    {
+      "epoch": 0.418468107819589,
+      "grad_norm": 0.9640330982395612,
+      "learning_rate": 1.307619133481396e-05,
+      "loss": 1.5793,
+      "step": 3136
+    },
+    {
+      "epoch": 0.41860154790499066,
+      "grad_norm": 0.9654481412303503,
+      "learning_rate": 1.3072078713778391e-05,
+      "loss": 1.5752,
+      "step": 3137
+    },
+    {
+      "epoch": 0.4187349879903923,
+      "grad_norm": 1.0550083483656565,
+      "learning_rate": 1.3067965518913495e-05,
+      "loss": 1.6224,
+      "step": 3138
+    },
+    {
+      "epoch": 0.41886842807579394,
+      "grad_norm": 1.0212917389559222,
+      "learning_rate": 1.3063851750987566e-05,
+      "loss": 1.657,
+      "step": 3139
+    },
+    {
+      "epoch": 0.4190018681611956,
+      "grad_norm": 7.254823593563437,
+      "learning_rate": 1.305973741076902e-05,
+      "loss": 1.6833,
+      "step": 3140
+    },
+    {
+      "epoch": 0.4191353082465973,
+      "grad_norm": 1.192322814644137,
+      "learning_rate": 1.3055622499026358e-05,
+      "loss": 1.6178,
+      "step": 3141
+    },
+    {
+      "epoch": 0.41926874833199895,
+      "grad_norm": 1.0727747759744528,
+      "learning_rate": 1.3051507016528206e-05,
+      "loss": 1.6121,
+      "step": 3142
+    },
+    {
+      "epoch": 0.41940218841740057,
+      "grad_norm": 1.090188535676899,
+      "learning_rate": 1.3047390964043282e-05,
+      "loss": 1.6112,
+      "step": 3143
+    },
+    {
+      "epoch": 0.41953562850280224,
+      "grad_norm": 1.0358843384129968,
+      "learning_rate": 1.3043274342340426e-05,
+      "loss": 1.6308,
+      "step": 3144
+    },
+    {
+      "epoch": 0.4196690685882039,
+      "grad_norm": 1.0321088750194614,
+      "learning_rate": 1.3039157152188569e-05,
+      "loss": 1.6028,
+      "step": 3145
+    },
+    {
+      "epoch": 0.4198025086736055,
+      "grad_norm": 0.9448257397915428,
+      "learning_rate": 1.3035039394356761e-05,
+      "loss": 1.5707,
+      "step": 3146
+    },
+    {
+      "epoch": 0.4199359487590072,
+      "grad_norm": 1.1440735516549028,
+      "learning_rate": 1.3030921069614145e-05,
+      "loss": 1.6633,
+      "step": 3147
+    },
+    {
+      "epoch": 0.42006938884440886,
+      "grad_norm": 1.690173809325389,
+      "learning_rate": 1.3026802178729985e-05,
+      "loss": 1.643,
+      "step": 3148
+    },
+    {
+      "epoch": 0.42020282892981053,
+      "grad_norm": 0.9819001583539148,
+      "learning_rate": 1.302268272247364e-05,
+      "loss": 1.613,
+      "step": 3149
+    },
+    {
+      "epoch": 0.42033626901521215,
+      "grad_norm": 1.2486781524480748,
+      "learning_rate": 1.3018562701614572e-05,
+      "loss": 1.6,
+      "step": 3150
+    },
+    {
+      "epoch": 0.4204697091006138,
+      "grad_norm": 1.075695894476628,
+      "learning_rate": 1.3014442116922363e-05,
+      "loss": 1.6246,
+      "step": 3151
+    },
+    {
+      "epoch": 0.4206031491860155,
+      "grad_norm": 1.0010043297382096,
+      "learning_rate": 1.3010320969166688e-05,
+      "loss": 1.6071,
+      "step": 3152
+    },
+    {
+      "epoch": 0.42073658927141716,
+      "grad_norm": 1.0686067267812616,
+      "learning_rate": 1.300619925911733e-05,
+      "loss": 1.5974,
+      "step": 3153
+    },
+    {
+      "epoch": 0.4208700293568188,
+      "grad_norm": 1.5326383589995518,
+      "learning_rate": 1.3002076987544173e-05,
+      "loss": 1.616,
+      "step": 3154
+    },
+    {
+      "epoch": 0.42100346944222045,
+      "grad_norm": 1.1448204497310523,
+      "learning_rate": 1.2997954155217216e-05,
+      "loss": 1.6295,
+      "step": 3155
+    },
+    {
+      "epoch": 0.4211369095276221,
+      "grad_norm": 1.0093147440921253,
+      "learning_rate": 1.2993830762906558e-05,
+      "loss": 1.5872,
+      "step": 3156
+    },
+    {
+      "epoch": 0.42127034961302373,
+      "grad_norm": 1.1083677568054542,
+      "learning_rate": 1.29897068113824e-05,
+      "loss": 1.6149,
+      "step": 3157
+    },
+    {
+      "epoch": 0.4214037896984254,
+      "grad_norm": 1.0172339428080068,
+      "learning_rate": 1.2985582301415045e-05,
+      "loss": 1.6115,
+      "step": 3158
+    },
+    {
+      "epoch": 0.42153722978382707,
+      "grad_norm": 1.122382756050523,
+      "learning_rate": 1.298145723377491e-05,
+      "loss": 1.5786,
+      "step": 3159
+    },
+    {
+      "epoch": 0.42167066986922874,
+      "grad_norm": 1.0389932694756445,
+      "learning_rate": 1.2977331609232511e-05,
+      "loss": 1.634,
+      "step": 3160
+    },
+    {
+      "epoch": 0.42180410995463036,
+      "grad_norm": 1.0581673343897444,
+      "learning_rate": 1.2973205428558461e-05,
+      "loss": 1.5624,
+      "step": 3161
+    },
+    {
+      "epoch": 0.42193755004003203,
+      "grad_norm": 1.7234834631306422,
+      "learning_rate": 1.2969078692523491e-05,
+      "loss": 1.6275,
+      "step": 3162
+    },
+    {
+      "epoch": 0.4220709901254337,
+      "grad_norm": 0.9813583900737928,
+      "learning_rate": 1.2964951401898427e-05,
+      "loss": 1.6535,
+      "step": 3163
+    },
+    {
+      "epoch": 0.4222044302108353,
+      "grad_norm": 0.9898306998533509,
+      "learning_rate": 1.2960823557454196e-05,
+      "loss": 1.6149,
+      "step": 3164
+    },
+    {
+      "epoch": 0.422337870296237,
+      "grad_norm": 1.0720992088919126,
+      "learning_rate": 1.2956695159961835e-05,
+      "loss": 1.6335,
+      "step": 3165
+    },
+    {
+      "epoch": 0.42247131038163865,
+      "grad_norm": 0.9836128501217848,
+      "learning_rate": 1.2952566210192483e-05,
+      "loss": 1.6103,
+      "step": 3166
+    },
+    {
+      "epoch": 0.4226047504670403,
+      "grad_norm": 1.0276047259882335,
+      "learning_rate": 1.2948436708917377e-05,
+      "loss": 1.529,
+      "step": 3167
+    },
+    {
+      "epoch": 0.42273819055244194,
+      "grad_norm": 0.9856021354742732,
+      "learning_rate": 1.2944306656907863e-05,
+      "loss": 1.6331,
+      "step": 3168
+    },
+    {
+      "epoch": 0.4228716306378436,
+      "grad_norm": 1.102227555293798,
+      "learning_rate": 1.2940176054935392e-05,
+      "loss": 1.6351,
+      "step": 3169
+    },
+    {
+      "epoch": 0.4230050707232453,
+      "grad_norm": 1.163021632169652,
+      "learning_rate": 1.2936044903771507e-05,
+      "loss": 1.6315,
+      "step": 3170
+    },
+    {
+      "epoch": 0.4231385108086469,
+      "grad_norm": 0.9872219387832729,
+      "learning_rate": 1.293191320418786e-05,
+      "loss": 1.5893,
+      "step": 3171
+    },
+    {
+      "epoch": 0.42327195089404857,
+      "grad_norm": 0.9582985814406623,
+      "learning_rate": 1.2927780956956208e-05,
+      "loss": 1.5824,
+      "step": 3172
+    },
+    {
+      "epoch": 0.42340539097945024,
+      "grad_norm": 1.1512050311999436,
+      "learning_rate": 1.2923648162848407e-05,
+      "loss": 1.5659,
+      "step": 3173
+    },
+    {
+      "epoch": 0.4235388310648519,
+      "grad_norm": 0.9359070452550783,
+      "learning_rate": 1.2919514822636419e-05,
+      "loss": 1.5615,
+      "step": 3174
+    },
+    {
+      "epoch": 0.4236722711502535,
+      "grad_norm": 1.095401798316347,
+      "learning_rate": 1.29153809370923e-05,
+      "loss": 1.5976,
+      "step": 3175
+    },
+    {
+      "epoch": 0.4238057112356552,
+      "grad_norm": 1.148887324591379,
+      "learning_rate": 1.2911246506988215e-05,
+      "loss": 1.6529,
+      "step": 3176
+    },
+    {
+      "epoch": 0.42393915132105686,
+      "grad_norm": 1.045994466470146,
+      "learning_rate": 1.2907111533096429e-05,
+      "loss": 1.6055,
+      "step": 3177
+    },
+    {
+      "epoch": 0.4240725914064585,
+      "grad_norm": 1.0087477247067256,
+      "learning_rate": 1.2902976016189304e-05,
+      "loss": 1.6546,
+      "step": 3178
+    },
+    {
+      "epoch": 0.42420603149186015,
+      "grad_norm": 1.016393015662281,
+      "learning_rate": 1.2898839957039313e-05,
+      "loss": 1.6213,
+      "step": 3179
+    },
+    {
+      "epoch": 0.4243394715772618,
+      "grad_norm": 1.0626699808769553,
+      "learning_rate": 1.2894703356419023e-05,
+      "loss": 1.5953,
+      "step": 3180
+    },
+    {
+      "epoch": 0.4244729116626635,
+      "grad_norm": 1.055001415948082,
+      "learning_rate": 1.2890566215101103e-05,
+      "loss": 1.6129,
+      "step": 3181
+    },
+    {
+      "epoch": 0.4246063517480651,
+      "grad_norm": 1.289178506303072,
+      "learning_rate": 1.2886428533858323e-05,
+      "loss": 1.5734,
+      "step": 3182
+    },
+    {
+      "epoch": 0.4247397918334668,
+      "grad_norm": 0.9875166066103552,
+      "learning_rate": 1.2882290313463561e-05,
+      "loss": 1.5684,
+      "step": 3183
+    },
+    {
+      "epoch": 0.42487323191886844,
+      "grad_norm": 0.9865955139060382,
+      "learning_rate": 1.2878151554689779e-05,
+      "loss": 1.611,
+      "step": 3184
+    },
+    {
+      "epoch": 0.42500667200427006,
+      "grad_norm": 0.9518489443348546,
+      "learning_rate": 1.287401225831006e-05,
+      "loss": 1.6425,
+      "step": 3185
+    },
+    {
+      "epoch": 0.42514011208967173,
+      "grad_norm": 0.9998924842921293,
+      "learning_rate": 1.286987242509757e-05,
+      "loss": 1.5908,
+      "step": 3186
+    },
+    {
+      "epoch": 0.4252735521750734,
+      "grad_norm": 0.9623933406546137,
+      "learning_rate": 1.2865732055825584e-05,
+      "loss": 1.6013,
+      "step": 3187
+    },
+    {
+      "epoch": 0.42540699226047507,
+      "grad_norm": 1.014360008466468,
+      "learning_rate": 1.2861591151267483e-05,
+      "loss": 1.6092,
+      "step": 3188
+    },
+    {
+      "epoch": 0.4255404323458767,
+      "grad_norm": 1.0008036520235628,
+      "learning_rate": 1.2857449712196733e-05,
+      "loss": 1.6417,
+      "step": 3189
+    },
+    {
+      "epoch": 0.42567387243127836,
+      "grad_norm": 1.0695928415515632,
+      "learning_rate": 1.2853307739386908e-05,
+      "loss": 1.6225,
+      "step": 3190
+    },
+    {
+      "epoch": 0.42580731251668,
+      "grad_norm": 0.9505901957555256,
+      "learning_rate": 1.2849165233611687e-05,
+      "loss": 1.5678,
+      "step": 3191
+    },
+    {
+      "epoch": 0.42594075260208164,
+      "grad_norm": 1.0097739379690966,
+      "learning_rate": 1.2845022195644838e-05,
+      "loss": 1.6279,
+      "step": 3192
+    },
+    {
+      "epoch": 0.4260741926874833,
+      "grad_norm": 0.9745534806645423,
+      "learning_rate": 1.2840878626260231e-05,
+      "loss": 1.6465,
+      "step": 3193
+    },
+    {
+      "epoch": 0.426207632772885,
+      "grad_norm": 0.9315201229811604,
+      "learning_rate": 1.2836734526231844e-05,
+      "loss": 1.518,
+      "step": 3194
+    },
+    {
+      "epoch": 0.42634107285828665,
+      "grad_norm": 0.9736368620203836,
+      "learning_rate": 1.2832589896333747e-05,
+      "loss": 1.5971,
+      "step": 3195
+    },
+    {
+      "epoch": 0.42647451294368827,
+      "grad_norm": 1.1002775590993574,
+      "learning_rate": 1.2828444737340105e-05,
+      "loss": 1.6011,
+      "step": 3196
+    },
+    {
+      "epoch": 0.42660795302908994,
+      "grad_norm": 1.0076905774463565,
+      "learning_rate": 1.282429905002519e-05,
+      "loss": 1.6084,
+      "step": 3197
+    },
+    {
+      "epoch": 0.4267413931144916,
+      "grad_norm": 0.9760408811448265,
+      "learning_rate": 1.2820152835163366e-05,
+      "loss": 1.5717,
+      "step": 3198
+    },
+    {
+      "epoch": 0.4268748331998932,
+      "grad_norm": 0.9853900063619548,
+      "learning_rate": 1.2816006093529106e-05,
+      "loss": 1.6631,
+      "step": 3199
+    },
+    {
+      "epoch": 0.4270082732852949,
+      "grad_norm": 1.2084382692384266,
+      "learning_rate": 1.2811858825896965e-05,
+      "loss": 1.6046,
+      "step": 3200
+    },
+    {
+      "epoch": 0.42714171337069656,
+      "grad_norm": 0.9676385504121742,
+      "learning_rate": 1.2807711033041613e-05,
+      "loss": 1.6724,
+      "step": 3201
+    },
+    {
+      "epoch": 0.42727515345609823,
+      "grad_norm": 0.9778031832163321,
+      "learning_rate": 1.2803562715737802e-05,
+      "loss": 1.5844,
+      "step": 3202
+    },
+    {
+      "epoch": 0.42740859354149985,
+      "grad_norm": 0.954310934487862,
+      "learning_rate": 1.2799413874760398e-05,
+      "loss": 1.5669,
+      "step": 3203
+    },
+    {
+      "epoch": 0.4275420336269015,
+      "grad_norm": 1.102491655832009,
+      "learning_rate": 1.2795264510884357e-05,
+      "loss": 1.5134,
+      "step": 3204
+    },
+    {
+      "epoch": 0.4276754737123032,
+      "grad_norm": 1.1945300325637898,
+      "learning_rate": 1.2791114624884728e-05,
+      "loss": 1.6208,
+      "step": 3205
+    },
+    {
+      "epoch": 0.4278089137977048,
+      "grad_norm": 1.0204751169748503,
+      "learning_rate": 1.2786964217536666e-05,
+      "loss": 1.5952,
+      "step": 3206
+    },
+    {
+      "epoch": 0.4279423538831065,
+      "grad_norm": 1.2582481061790463,
+      "learning_rate": 1.278281328961542e-05,
+      "loss": 1.5961,
+      "step": 3207
+    },
+    {
+      "epoch": 0.42807579396850814,
+      "grad_norm": 1.0191124850439586,
+      "learning_rate": 1.2778661841896333e-05,
+      "loss": 1.6123,
+      "step": 3208
+    },
+    {
+      "epoch": 0.4282092340539098,
+      "grad_norm": 1.135529578537383,
+      "learning_rate": 1.277450987515485e-05,
+      "loss": 1.6442,
+      "step": 3209
+    },
+    {
+      "epoch": 0.42834267413931143,
+      "grad_norm": 1.099285147964999,
+      "learning_rate": 1.2770357390166513e-05,
+      "loss": 1.6706,
+      "step": 3210
+    },
+    {
+      "epoch": 0.4284761142247131,
+      "grad_norm": 1.017546212676164,
+      "learning_rate": 1.2766204387706955e-05,
+      "loss": 1.582,
+      "step": 3211
+    },
+    {
+      "epoch": 0.42860955431011477,
+      "grad_norm": 1.1530348114411721,
+      "learning_rate": 1.2762050868551913e-05,
+      "loss": 1.6239,
+      "step": 3212
+    },
+    {
+      "epoch": 0.4287429943955164,
+      "grad_norm": 0.9715880360042976,
+      "learning_rate": 1.275789683347722e-05,
+      "loss": 1.578,
+      "step": 3213
+    },
+    {
+      "epoch": 0.42887643448091806,
+      "grad_norm": 0.9866979071658567,
+      "learning_rate": 1.2753742283258793e-05,
+      "loss": 1.6167,
+      "step": 3214
+    },
+    {
+      "epoch": 0.4290098745663197,
+      "grad_norm": 0.9922836437350845,
+      "learning_rate": 1.2749587218672663e-05,
+      "loss": 1.6215,
+      "step": 3215
+    },
+    {
+      "epoch": 0.4291433146517214,
+      "grad_norm": 1.2015902362721134,
+      "learning_rate": 1.2745431640494944e-05,
+      "loss": 1.6423,
+      "step": 3216
+    },
+    {
+      "epoch": 0.429276754737123,
+      "grad_norm": 1.0161108583369185,
+      "learning_rate": 1.2741275549501853e-05,
+      "loss": 1.6025,
+      "step": 3217
+    },
+    {
+      "epoch": 0.4294101948225247,
+      "grad_norm": 0.9630893697181409,
+      "learning_rate": 1.2737118946469697e-05,
+      "loss": 1.5888,
+      "step": 3218
+    },
+    {
+      "epoch": 0.42954363490792635,
+      "grad_norm": 0.9590114370085036,
+      "learning_rate": 1.2732961832174888e-05,
+      "loss": 1.5735,
+      "step": 3219
+    },
+    {
+      "epoch": 0.429677074993328,
+      "grad_norm": 0.9698052625279515,
+      "learning_rate": 1.2728804207393925e-05,
+      "loss": 1.5758,
+      "step": 3220
+    },
+    {
+      "epoch": 0.42981051507872964,
+      "grad_norm": 1.1113879856734203,
+      "learning_rate": 1.2724646072903403e-05,
+      "loss": 1.5833,
+      "step": 3221
+    },
+    {
+      "epoch": 0.4299439551641313,
+      "grad_norm": 0.9741250433041043,
+      "learning_rate": 1.2720487429480017e-05,
+      "loss": 1.604,
+      "step": 3222
+    },
+    {
+      "epoch": 0.430077395249533,
+      "grad_norm": 1.0434937344586053,
+      "learning_rate": 1.2716328277900553e-05,
+      "loss": 1.6282,
+      "step": 3223
+    },
+    {
+      "epoch": 0.4302108353349346,
+      "grad_norm": 1.218598876665266,
+      "learning_rate": 1.2712168618941895e-05,
+      "loss": 1.6405,
+      "step": 3224
+    },
+    {
+      "epoch": 0.43034427542033626,
+      "grad_norm": 0.9274834745968594,
+      "learning_rate": 1.2708008453381015e-05,
+      "loss": 1.5594,
+      "step": 3225
+    },
+    {
+      "epoch": 0.43047771550573793,
+      "grad_norm": 0.9572245478353469,
+      "learning_rate": 1.2703847781994988e-05,
+      "loss": 1.5985,
+      "step": 3226
+    },
+    {
+      "epoch": 0.4306111555911396,
+      "grad_norm": 0.9804393710597245,
+      "learning_rate": 1.2699686605560984e-05,
+      "loss": 1.6098,
+      "step": 3227
+    },
+    {
+      "epoch": 0.4307445956765412,
+      "grad_norm": 0.9465366010602443,
+      "learning_rate": 1.2695524924856252e-05,
+      "loss": 1.6262,
+      "step": 3228
+    },
+    {
+      "epoch": 0.4308780357619429,
+      "grad_norm": 1.1417073930345842,
+      "learning_rate": 1.2691362740658162e-05,
+      "loss": 1.6277,
+      "step": 3229
+    },
+    {
+      "epoch": 0.43101147584734456,
+      "grad_norm": 1.078579390525697,
+      "learning_rate": 1.2687200053744148e-05,
+      "loss": 1.5878,
+      "step": 3230
+    },
+    {
+      "epoch": 0.4311449159327462,
+      "grad_norm": 0.9835105604076124,
+      "learning_rate": 1.2683036864891762e-05,
+      "loss": 1.585,
+      "step": 3231
+    },
+    {
+      "epoch": 0.43127835601814785,
+      "grad_norm": 0.9942457861542611,
+      "learning_rate": 1.2678873174878637e-05,
+      "loss": 1.6226,
+      "step": 3232
+    },
+    {
+      "epoch": 0.4314117961035495,
+      "grad_norm": 1.077188400186942,
+      "learning_rate": 1.2674708984482503e-05,
+      "loss": 1.617,
+      "step": 3233
+    },
+    {
+      "epoch": 0.4315452361889512,
+      "grad_norm": 0.9450353505746655,
+      "learning_rate": 1.2670544294481184e-05,
+      "loss": 1.5642,
+      "step": 3234
+    },
+    {
+      "epoch": 0.4316786762743528,
+      "grad_norm": 1.0397917929273337,
+      "learning_rate": 1.2666379105652593e-05,
+      "loss": 1.5236,
+      "step": 3235
+    },
+    {
+      "epoch": 0.43181211635975447,
+      "grad_norm": 1.2967775820644207,
+      "learning_rate": 1.2662213418774747e-05,
+      "loss": 1.6222,
+      "step": 3236
+    },
+    {
+      "epoch": 0.43194555644515614,
+      "grad_norm": 0.9953395285218007,
+      "learning_rate": 1.2658047234625741e-05,
+      "loss": 1.6417,
+      "step": 3237
+    },
+    {
+      "epoch": 0.43207899653055776,
+      "grad_norm": 1.0148344019514997,
+      "learning_rate": 1.2653880553983777e-05,
+      "loss": 1.6523,
+      "step": 3238
+    },
+    {
+      "epoch": 0.4322124366159594,
+      "grad_norm": 0.919347172987927,
+      "learning_rate": 1.264971337762714e-05,
+      "loss": 1.5869,
+      "step": 3239
+    },
+    {
+      "epoch": 0.4323458767013611,
+      "grad_norm": 0.9647756161069322,
+      "learning_rate": 1.2645545706334213e-05,
+      "loss": 1.6183,
+      "step": 3240
+    },
+    {
+      "epoch": 0.43247931678676277,
+      "grad_norm": 0.9857220961118699,
+      "learning_rate": 1.2641377540883469e-05,
+      "loss": 1.5573,
+      "step": 3241
+    },
+    {
+      "epoch": 0.4326127568721644,
+      "grad_norm": 1.004715345407096,
+      "learning_rate": 1.2637208882053469e-05,
+      "loss": 1.5961,
+      "step": 3242
+    },
+    {
+      "epoch": 0.43274619695756605,
+      "grad_norm": 0.951945556321612,
+      "learning_rate": 1.2633039730622883e-05,
+      "loss": 1.5411,
+      "step": 3243
+    },
+    {
+      "epoch": 0.4328796370429677,
+      "grad_norm": 0.9907591493074744,
+      "learning_rate": 1.2628870087370446e-05,
+      "loss": 1.5953,
+      "step": 3244
+    },
+    {
+      "epoch": 0.43301307712836934,
+      "grad_norm": 0.9807352408063607,
+      "learning_rate": 1.2624699953075015e-05,
+      "loss": 1.5794,
+      "step": 3245
+    },
+    {
+      "epoch": 0.433146517213771,
+      "grad_norm": 1.3426423736461461,
+      "learning_rate": 1.262052932851551e-05,
+      "loss": 1.6409,
+      "step": 3246
+    },
+    {
+      "epoch": 0.4332799572991727,
+      "grad_norm": 0.9714093568492226,
+      "learning_rate": 1.2616358214470967e-05,
+      "loss": 1.5626,
+      "step": 3247
+    },
+    {
+      "epoch": 0.43341339738457435,
+      "grad_norm": 1.2115238526991772,
+      "learning_rate": 1.2612186611720494e-05,
+      "loss": 1.5797,
+      "step": 3248
+    },
+    {
+      "epoch": 0.43354683746997597,
+      "grad_norm": 1.1075061582793597,
+      "learning_rate": 1.2608014521043305e-05,
+      "loss": 1.6192,
+      "step": 3249
+    },
+    {
+      "epoch": 0.43368027755537764,
+      "grad_norm": 0.9318065689517002,
+      "learning_rate": 1.2603841943218695e-05,
+      "loss": 1.5508,
+      "step": 3250
+    },
+    {
+      "epoch": 0.4338137176407793,
+      "grad_norm": 0.9342106979423349,
+      "learning_rate": 1.2599668879026057e-05,
+      "loss": 1.5932,
+      "step": 3251
+    },
+    {
+      "epoch": 0.4339471577261809,
+      "grad_norm": 0.9881294610347703,
+      "learning_rate": 1.259549532924487e-05,
+      "loss": 1.6374,
+      "step": 3252
+    },
+    {
+      "epoch": 0.4340805978115826,
+      "grad_norm": 0.9842232409667695,
+      "learning_rate": 1.259132129465471e-05,
+      "loss": 1.5709,
+      "step": 3253
+    },
+    {
+      "epoch": 0.43421403789698426,
+      "grad_norm": 1.0325015498789856,
+      "learning_rate": 1.2587146776035233e-05,
+      "loss": 1.6679,
+      "step": 3254
+    },
+    {
+      "epoch": 0.43434747798238593,
+      "grad_norm": 0.9483677954324838,
+      "learning_rate": 1.2582971774166195e-05,
+      "loss": 1.6129,
+      "step": 3255
+    },
+    {
+      "epoch": 0.43448091806778755,
+      "grad_norm": 1.0038948510108203,
+      "learning_rate": 1.2578796289827437e-05,
+      "loss": 1.639,
+      "step": 3256
+    },
+    {
+      "epoch": 0.4346143581531892,
+      "grad_norm": 1.0870982465609627,
+      "learning_rate": 1.2574620323798891e-05,
+      "loss": 1.559,
+      "step": 3257
+    },
+    {
+      "epoch": 0.4347477982385909,
+      "grad_norm": 1.1919615794250933,
+      "learning_rate": 1.257044387686058e-05,
+      "loss": 1.5857,
+      "step": 3258
+    },
+    {
+      "epoch": 0.4348812383239925,
+      "grad_norm": 2.641937471598561,
+      "learning_rate": 1.2566266949792625e-05,
+      "loss": 1.5931,
+      "step": 3259
+    },
+    {
+      "epoch": 0.4350146784093942,
+      "grad_norm": 0.9743834744800941,
+      "learning_rate": 1.2562089543375215e-05,
+      "loss": 1.6235,
+      "step": 3260
+    },
+    {
+      "epoch": 0.43514811849479584,
+      "grad_norm": 1.1785303574207608,
+      "learning_rate": 1.2557911658388655e-05,
+      "loss": 1.6243,
+      "step": 3261
+    },
+    {
+      "epoch": 0.4352815585801975,
+      "grad_norm": 1.3537174009151587,
+      "learning_rate": 1.2553733295613314e-05,
+      "loss": 1.5904,
+      "step": 3262
+    },
+    {
+      "epoch": 0.43541499866559913,
+      "grad_norm": 1.0054913667026213,
+      "learning_rate": 1.2549554455829676e-05,
+      "loss": 1.5623,
+      "step": 3263
+    },
+    {
+      "epoch": 0.4355484387510008,
+      "grad_norm": 1.1821789472170066,
+      "learning_rate": 1.2545375139818287e-05,
+      "loss": 1.5667,
+      "step": 3264
+    },
+    {
+      "epoch": 0.43568187883640247,
+      "grad_norm": 1.0060542893441342,
+      "learning_rate": 1.2541195348359805e-05,
+      "loss": 1.6021,
+      "step": 3265
+    },
+    {
+      "epoch": 0.4358153189218041,
+      "grad_norm": 0.9774430570620395,
+      "learning_rate": 1.2537015082234963e-05,
+      "loss": 1.6073,
+      "step": 3266
+    },
+    {
+      "epoch": 0.43594875900720576,
+      "grad_norm": 1.031179584529375,
+      "learning_rate": 1.253283434222459e-05,
+      "loss": 1.6927,
+      "step": 3267
+    },
+    {
+      "epoch": 0.4360821990926074,
+      "grad_norm": 1.66539793901367,
+      "learning_rate": 1.2528653129109597e-05,
+      "loss": 1.6507,
+      "step": 3268
+    },
+    {
+      "epoch": 0.4362156391780091,
+      "grad_norm": 1.0734574677292559,
+      "learning_rate": 1.2524471443670992e-05,
+      "loss": 1.6163,
+      "step": 3269
+    },
+    {
+      "epoch": 0.4363490792634107,
+      "grad_norm": 1.0252180431657283,
+      "learning_rate": 1.2520289286689864e-05,
+      "loss": 1.6268,
+      "step": 3270
+    },
+    {
+      "epoch": 0.4364825193488124,
+      "grad_norm": 1.0638123845106042,
+      "learning_rate": 1.2516106658947389e-05,
+      "loss": 1.5639,
+      "step": 3271
+    },
+    {
+      "epoch": 0.43661595943421405,
+      "grad_norm": 0.9457091191118885,
+      "learning_rate": 1.251192356122484e-05,
+      "loss": 1.6036,
+      "step": 3272
+    },
+    {
+      "epoch": 0.43674939951961567,
+      "grad_norm": 0.9937359629999882,
+      "learning_rate": 1.2507739994303564e-05,
+      "loss": 1.6365,
+      "step": 3273
+    },
+    {
+      "epoch": 0.43688283960501734,
+      "grad_norm": 1.1942032412216412,
+      "learning_rate": 1.2503555958965014e-05,
+      "loss": 1.6593,
+      "step": 3274
+    },
+    {
+      "epoch": 0.437016279690419,
+      "grad_norm": 1.0002882913147546,
+      "learning_rate": 1.2499371455990714e-05,
+      "loss": 1.5772,
+      "step": 3275
+    },
+    {
+      "epoch": 0.4371497197758207,
+      "grad_norm": 0.9629918020750007,
+      "learning_rate": 1.2495186486162284e-05,
+      "loss": 1.5488,
+      "step": 3276
+    },
+    {
+      "epoch": 0.4372831598612223,
+      "grad_norm": 0.9857190069244037,
+      "learning_rate": 1.2491001050261425e-05,
+      "loss": 1.6248,
+      "step": 3277
+    },
+    {
+      "epoch": 0.43741659994662396,
+      "grad_norm": 1.099517425168157,
+      "learning_rate": 1.2486815149069928e-05,
+      "loss": 1.5877,
+      "step": 3278
+    },
+    {
+      "epoch": 0.43755004003202563,
+      "grad_norm": 1.1165501715980635,
+      "learning_rate": 1.248262878336968e-05,
+      "loss": 1.6246,
+      "step": 3279
+    },
+    {
+      "epoch": 0.43768348011742725,
+      "grad_norm": 1.2759947966412557,
+      "learning_rate": 1.2478441953942637e-05,
+      "loss": 1.579,
+      "step": 3280
+    },
+    {
+      "epoch": 0.4378169202028289,
+      "grad_norm": 1.042737881072034,
+      "learning_rate": 1.2474254661570858e-05,
+      "loss": 1.6238,
+      "step": 3281
+    },
+    {
+      "epoch": 0.4379503602882306,
+      "grad_norm": 0.9997308223160407,
+      "learning_rate": 1.2470066907036475e-05,
+      "loss": 1.6247,
+      "step": 3282
+    },
+    {
+      "epoch": 0.43808380037363226,
+      "grad_norm": 0.9873336853234015,
+      "learning_rate": 1.2465878691121717e-05,
+      "loss": 1.594,
+      "step": 3283
+    },
+    {
+      "epoch": 0.4382172404590339,
+      "grad_norm": 1.073937141396183,
+      "learning_rate": 1.2461690014608898e-05,
+      "loss": 1.5442,
+      "step": 3284
+    },
+    {
+      "epoch": 0.43835068054443554,
+      "grad_norm": 1.066440340364437,
+      "learning_rate": 1.2457500878280408e-05,
+      "loss": 1.5947,
+      "step": 3285
+    },
+    {
+      "epoch": 0.4384841206298372,
+      "grad_norm": 0.9823058406854864,
+      "learning_rate": 1.2453311282918738e-05,
+      "loss": 1.5988,
+      "step": 3286
+    },
+    {
+      "epoch": 0.43861756071523883,
+      "grad_norm": 0.9529342446435208,
+      "learning_rate": 1.2449121229306449e-05,
+      "loss": 1.5879,
+      "step": 3287
+    },
+    {
+      "epoch": 0.4387510008006405,
+      "grad_norm": 1.2995317730664497,
+      "learning_rate": 1.2444930718226201e-05,
+      "loss": 1.6078,
+      "step": 3288
+    },
+    {
+      "epoch": 0.43888444088604217,
+      "grad_norm": 1.0273219898248251,
+      "learning_rate": 1.2440739750460728e-05,
+      "loss": 1.5873,
+      "step": 3289
+    },
+    {
+      "epoch": 0.43901788097144384,
+      "grad_norm": 0.9613704608116244,
+      "learning_rate": 1.2436548326792858e-05,
+      "loss": 1.5971,
+      "step": 3290
+    },
+    {
+      "epoch": 0.43915132105684546,
+      "grad_norm": 0.9741621574671587,
+      "learning_rate": 1.2432356448005507e-05,
+      "loss": 1.5664,
+      "step": 3291
+    },
+    {
+      "epoch": 0.4392847611422471,
+      "grad_norm": 0.9421233395790419,
+      "learning_rate": 1.2428164114881663e-05,
+      "loss": 1.6159,
+      "step": 3292
+    },
+    {
+      "epoch": 0.4394182012276488,
+      "grad_norm": 0.9714375843061244,
+      "learning_rate": 1.2423971328204407e-05,
+      "loss": 1.5948,
+      "step": 3293
+    },
+    {
+      "epoch": 0.43955164131305047,
+      "grad_norm": 0.9853929138462288,
+      "learning_rate": 1.2419778088756904e-05,
+      "loss": 1.6295,
+      "step": 3294
+    },
+    {
+      "epoch": 0.4396850813984521,
+      "grad_norm": 0.9658869477722872,
+      "learning_rate": 1.2415584397322406e-05,
+      "loss": 1.6026,
+      "step": 3295
+    },
+    {
+      "epoch": 0.43981852148385375,
+      "grad_norm": 0.9824798838337809,
+      "learning_rate": 1.2411390254684246e-05,
+      "loss": 1.6511,
+      "step": 3296
+    },
+    {
+      "epoch": 0.4399519615692554,
+      "grad_norm": 1.1148315528827775,
+      "learning_rate": 1.2407195661625838e-05,
+      "loss": 1.6555,
+      "step": 3297
+    },
+    {
+      "epoch": 0.44008540165465704,
+      "grad_norm": 0.9935944482329716,
+      "learning_rate": 1.240300061893069e-05,
+      "loss": 1.5694,
+      "step": 3298
+    },
+    {
+      "epoch": 0.4402188417400587,
+      "grad_norm": 1.0104396213439037,
+      "learning_rate": 1.2398805127382382e-05,
+      "loss": 1.6266,
+      "step": 3299
+    },
+    {
+      "epoch": 0.4403522818254604,
+      "grad_norm": 0.9406209717223495,
+      "learning_rate": 1.2394609187764593e-05,
+      "loss": 1.6403,
+      "step": 3300
+    },
+    {
+      "epoch": 0.44048572191086205,
+      "grad_norm": 0.9669078454658077,
+      "learning_rate": 1.2390412800861066e-05,
+      "loss": 1.614,
+      "step": 3301
+    },
+    {
+      "epoch": 0.44061916199626366,
+      "grad_norm": 0.9556156862945018,
+      "learning_rate": 1.2386215967455648e-05,
+      "loss": 1.5857,
+      "step": 3302
+    },
+    {
+      "epoch": 0.44075260208166533,
+      "grad_norm": 0.9777122703402282,
+      "learning_rate": 1.2382018688332251e-05,
+      "loss": 1.6382,
+      "step": 3303
+    },
+    {
+      "epoch": 0.440886042167067,
+      "grad_norm": 0.9515365211823855,
+      "learning_rate": 1.2377820964274887e-05,
+      "loss": 1.5879,
+      "step": 3304
+    },
+    {
+      "epoch": 0.4410194822524686,
+      "grad_norm": 0.9807886665234027,
+      "learning_rate": 1.2373622796067637e-05,
+      "loss": 1.6029,
+      "step": 3305
+    },
+    {
+      "epoch": 0.4411529223378703,
+      "grad_norm": 0.9635105279990112,
+      "learning_rate": 1.2369424184494673e-05,
+      "loss": 1.5992,
+      "step": 3306
+    },
+    {
+      "epoch": 0.44128636242327196,
+      "grad_norm": 0.9665429167972764,
+      "learning_rate": 1.236522513034025e-05,
+      "loss": 1.5908,
+      "step": 3307
+    },
+    {
+      "epoch": 0.44141980250867363,
+      "grad_norm": 1.0554959304610874,
+      "learning_rate": 1.2361025634388701e-05,
+      "loss": 1.6039,
+      "step": 3308
+    },
+    {
+      "epoch": 0.44155324259407525,
+      "grad_norm": 1.206338234294267,
+      "learning_rate": 1.2356825697424449e-05,
+      "loss": 1.6125,
+      "step": 3309
+    },
+    {
+      "epoch": 0.4416866826794769,
+      "grad_norm": 0.9701936211446273,
+      "learning_rate": 1.2352625320231984e-05,
+      "loss": 1.5516,
+      "step": 3310
+    },
+    {
+      "epoch": 0.4418201227648786,
+      "grad_norm": 1.011429585012499,
+      "learning_rate": 1.2348424503595898e-05,
+      "loss": 1.6186,
+      "step": 3311
+    },
+    {
+      "epoch": 0.4419535628502802,
+      "grad_norm": 1.1168012771237403,
+      "learning_rate": 1.234422324830085e-05,
+      "loss": 1.565,
+      "step": 3312
+    },
+    {
+      "epoch": 0.44208700293568187,
+      "grad_norm": 1.0738983437774423,
+      "learning_rate": 1.2340021555131592e-05,
+      "loss": 1.6279,
+      "step": 3313
+    },
+    {
+      "epoch": 0.44222044302108354,
+      "grad_norm": 0.9987855843220456,
+      "learning_rate": 1.2335819424872948e-05,
+      "loss": 1.5884,
+      "step": 3314
+    },
+    {
+      "epoch": 0.4423538831064852,
+      "grad_norm": 1.1047058977235424,
+      "learning_rate": 1.233161685830983e-05,
+      "loss": 1.6259,
+      "step": 3315
+    },
+    {
+      "epoch": 0.4424873231918868,
+      "grad_norm": 12.94394608036084,
+      "learning_rate": 1.2327413856227231e-05,
+      "loss": 1.6551,
+      "step": 3316
+    },
+    {
+      "epoch": 0.4426207632772885,
+      "grad_norm": 0.9922991075854602,
+      "learning_rate": 1.232321041941022e-05,
+      "loss": 1.6057,
+      "step": 3317
+    },
+    {
+      "epoch": 0.44275420336269017,
+      "grad_norm": 1.5832757455395445,
+      "learning_rate": 1.2319006548643955e-05,
+      "loss": 1.5688,
+      "step": 3318
+    },
+    {
+      "epoch": 0.4428876434480918,
+      "grad_norm": 1.0990617479523308,
+      "learning_rate": 1.2314802244713671e-05,
+      "loss": 1.6046,
+      "step": 3319
+    },
+    {
+      "epoch": 0.44302108353349345,
+      "grad_norm": 1.1847275698751005,
+      "learning_rate": 1.2310597508404683e-05,
+      "loss": 1.5834,
+      "step": 3320
+    },
+    {
+      "epoch": 0.4431545236188951,
+      "grad_norm": 1.044830968363967,
+      "learning_rate": 1.2306392340502382e-05,
+      "loss": 1.5843,
+      "step": 3321
+    },
+    {
+      "epoch": 0.4432879637042968,
+      "grad_norm": 1.2604211119820607,
+      "learning_rate": 1.2302186741792255e-05,
+      "loss": 1.6076,
+      "step": 3322
+    },
+    {
+      "epoch": 0.4434214037896984,
+      "grad_norm": 1.041133538088134,
+      "learning_rate": 1.2297980713059857e-05,
+      "loss": 1.6237,
+      "step": 3323
+    },
+    {
+      "epoch": 0.4435548438751001,
+      "grad_norm": 1.4649823119870107,
+      "learning_rate": 1.229377425509082e-05,
+      "loss": 1.6073,
+      "step": 3324
+    },
+    {
+      "epoch": 0.44368828396050175,
+      "grad_norm": 0.9934245058649951,
+      "learning_rate": 1.2289567368670873e-05,
+      "loss": 1.5818,
+      "step": 3325
+    },
+    {
+      "epoch": 0.44382172404590337,
+      "grad_norm": 0.9664114510742975,
+      "learning_rate": 1.2285360054585807e-05,
+      "loss": 1.644,
+      "step": 3326
+    },
+    {
+      "epoch": 0.44395516413130504,
+      "grad_norm": 1.015993424077106,
+      "learning_rate": 1.2281152313621505e-05,
+      "loss": 1.5808,
+      "step": 3327
+    },
+    {
+      "epoch": 0.4440886042167067,
+      "grad_norm": 1.0461329594907725,
+      "learning_rate": 1.2276944146563918e-05,
+      "loss": 1.6381,
+      "step": 3328
+    },
+    {
+      "epoch": 0.4442220443021084,
+      "grad_norm": 1.0135143248090914,
+      "learning_rate": 1.2272735554199091e-05,
+      "loss": 1.577,
+      "step": 3329
+    },
+    {
+      "epoch": 0.44435548438751,
+      "grad_norm": 1.0070141228550389,
+      "learning_rate": 1.2268526537313142e-05,
+      "loss": 1.5906,
+      "step": 3330
+    },
+    {
+      "epoch": 0.44448892447291166,
+      "grad_norm": 0.9761989502121017,
+      "learning_rate": 1.2264317096692257e-05,
+      "loss": 1.5424,
+      "step": 3331
+    },
+    {
+      "epoch": 0.44462236455831333,
+      "grad_norm": 0.9755643448669162,
+      "learning_rate": 1.2260107233122724e-05,
+      "loss": 1.621,
+      "step": 3332
+    },
+    {
+      "epoch": 0.44475580464371495,
+      "grad_norm": 1.1020152124691254,
+      "learning_rate": 1.2255896947390891e-05,
+      "loss": 1.5605,
+      "step": 3333
+    },
+    {
+      "epoch": 0.4448892447291166,
+      "grad_norm": 1.3925175705055546,
+      "learning_rate": 1.2251686240283191e-05,
+      "loss": 1.6224,
+      "step": 3334
+    },
+    {
+      "epoch": 0.4450226848145183,
+      "grad_norm": 1.0243447818889067,
+      "learning_rate": 1.224747511258614e-05,
+      "loss": 1.6326,
+      "step": 3335
+    },
+    {
+      "epoch": 0.44515612489991996,
+      "grad_norm": 1.064152850059643,
+      "learning_rate": 1.2243263565086325e-05,
+      "loss": 1.5844,
+      "step": 3336
+    },
+    {
+      "epoch": 0.4452895649853216,
+      "grad_norm": 1.0995677570425262,
+      "learning_rate": 1.2239051598570417e-05,
+      "loss": 1.6008,
+      "step": 3337
+    },
+    {
+      "epoch": 0.44542300507072324,
+      "grad_norm": 0.9935950777647963,
+      "learning_rate": 1.2234839213825163e-05,
+      "loss": 1.6462,
+      "step": 3338
+    },
+    {
+      "epoch": 0.4455564451561249,
+      "grad_norm": 0.9919635923562781,
+      "learning_rate": 1.2230626411637388e-05,
+      "loss": 1.6029,
+      "step": 3339
+    },
+    {
+      "epoch": 0.44568988524152653,
+      "grad_norm": 1.3321534487303814,
+      "learning_rate": 1.2226413192793998e-05,
+      "loss": 1.6439,
+      "step": 3340
+    },
+    {
+      "epoch": 0.4458233253269282,
+      "grad_norm": 1.0388019535817785,
+      "learning_rate": 1.222219955808197e-05,
+      "loss": 1.6172,
+      "step": 3341
+    },
+    {
+      "epoch": 0.44595676541232987,
+      "grad_norm": 0.9948371593169276,
+      "learning_rate": 1.2217985508288366e-05,
+      "loss": 1.5684,
+      "step": 3342
+    },
+    {
+      "epoch": 0.44609020549773154,
+      "grad_norm": 0.9566696947998643,
+      "learning_rate": 1.2213771044200323e-05,
+      "loss": 1.5894,
+      "step": 3343
+    },
+    {
+      "epoch": 0.44622364558313315,
+      "grad_norm": 1.0144760346896669,
+      "learning_rate": 1.220955616660505e-05,
+      "loss": 1.6187,
+      "step": 3344
+    },
+    {
+      "epoch": 0.4463570856685348,
+      "grad_norm": 0.9783013624017898,
+      "learning_rate": 1.2205340876289842e-05,
+      "loss": 1.5798,
+      "step": 3345
+    },
+    {
+      "epoch": 0.4464905257539365,
+      "grad_norm": 1.1775470120965925,
+      "learning_rate": 1.220112517404207e-05,
+      "loss": 1.6027,
+      "step": 3346
+    },
+    {
+      "epoch": 0.4466239658393381,
+      "grad_norm": 1.0832460341808767,
+      "learning_rate": 1.2196909060649173e-05,
+      "loss": 1.6384,
+      "step": 3347
+    },
+    {
+      "epoch": 0.4467574059247398,
+      "grad_norm": 1.5868021023209082,
+      "learning_rate": 1.219269253689868e-05,
+      "loss": 1.5758,
+      "step": 3348
+    },
+    {
+      "epoch": 0.44689084601014145,
+      "grad_norm": 1.1241637872673438,
+      "learning_rate": 1.2188475603578186e-05,
+      "loss": 1.5948,
+      "step": 3349
+    },
+    {
+      "epoch": 0.4470242860955431,
+      "grad_norm": 0.9779424708497318,
+      "learning_rate": 1.2184258261475364e-05,
+      "loss": 1.6002,
+      "step": 3350
+    },
+    {
+      "epoch": 0.44715772618094474,
+      "grad_norm": 0.986774564760355,
+      "learning_rate": 1.2180040511377966e-05,
+      "loss": 1.5651,
+      "step": 3351
+    },
+    {
+      "epoch": 0.4472911662663464,
+      "grad_norm": 0.9605234817512258,
+      "learning_rate": 1.2175822354073826e-05,
+      "loss": 1.5525,
+      "step": 3352
+    },
+    {
+      "epoch": 0.4474246063517481,
+      "grad_norm": 0.9492182807666106,
+      "learning_rate": 1.2171603790350836e-05,
+      "loss": 1.5657,
+      "step": 3353
+    },
+    {
+      "epoch": 0.4475580464371497,
+      "grad_norm": 1.004022400441443,
+      "learning_rate": 1.2167384820996988e-05,
+      "loss": 1.5807,
+      "step": 3354
+    },
+    {
+      "epoch": 0.44769148652255136,
+      "grad_norm": 0.9732333755288013,
+      "learning_rate": 1.2163165446800332e-05,
+      "loss": 1.6453,
+      "step": 3355
+    },
+    {
+      "epoch": 0.44782492660795303,
+      "grad_norm": 1.0036906910118235,
+      "learning_rate": 1.2158945668548997e-05,
+      "loss": 1.6128,
+      "step": 3356
+    },
+    {
+      "epoch": 0.4479583666933547,
+      "grad_norm": 1.0085064642711343,
+      "learning_rate": 1.215472548703119e-05,
+      "loss": 1.613,
+      "step": 3357
+    },
+    {
+      "epoch": 0.4480918067787563,
+      "grad_norm": 1.2764967944347796,
+      "learning_rate": 1.2150504903035196e-05,
+      "loss": 1.5606,
+      "step": 3358
+    },
+    {
+      "epoch": 0.448225246864158,
+      "grad_norm": 0.995504684910624,
+      "learning_rate": 1.2146283917349373e-05,
+      "loss": 1.6166,
+      "step": 3359
+    },
+    {
+      "epoch": 0.44835868694955966,
+      "grad_norm": 0.9568604455855414,
+      "learning_rate": 1.214206253076215e-05,
+      "loss": 1.5966,
+      "step": 3360
+    },
+    {
+      "epoch": 0.44849212703496133,
+      "grad_norm": 1.1039131846221917,
+      "learning_rate": 1.2137840744062032e-05,
+      "loss": 1.5601,
+      "step": 3361
+    },
+    {
+      "epoch": 0.44862556712036294,
+      "grad_norm": 1.005103856197314,
+      "learning_rate": 1.2133618558037607e-05,
+      "loss": 1.5586,
+      "step": 3362
+    },
+    {
+      "epoch": 0.4487590072057646,
+      "grad_norm": 0.9893281059878032,
+      "learning_rate": 1.2129395973477522e-05,
+      "loss": 1.664,
+      "step": 3363
+    },
+    {
+      "epoch": 0.4488924472911663,
+      "grad_norm": 1.3408090451344328,
+      "learning_rate": 1.212517299117052e-05,
+      "loss": 1.5822,
+      "step": 3364
+    },
+    {
+      "epoch": 0.4490258873765679,
+      "grad_norm": 1.1012723616514362,
+      "learning_rate": 1.2120949611905393e-05,
+      "loss": 1.5357,
+      "step": 3365
+    },
+    {
+      "epoch": 0.44915932746196957,
+      "grad_norm": 1.1116159284883955,
+      "learning_rate": 1.2116725836471031e-05,
+      "loss": 1.5744,
+      "step": 3366
+    },
+    {
+      "epoch": 0.44929276754737124,
+      "grad_norm": 0.9861652273388498,
+      "learning_rate": 1.211250166565638e-05,
+      "loss": 1.6247,
+      "step": 3367
+    },
+    {
+      "epoch": 0.4494262076327729,
+      "grad_norm": 1.1947078047934485,
+      "learning_rate": 1.2108277100250472e-05,
+      "loss": 1.6308,
+      "step": 3368
+    },
+    {
+      "epoch": 0.4495596477181745,
+      "grad_norm": 0.946296734521657,
+      "learning_rate": 1.2104052141042402e-05,
+      "loss": 1.5815,
+      "step": 3369
+    },
+    {
+      "epoch": 0.4496930878035762,
+      "grad_norm": 0.9521803217831275,
+      "learning_rate": 1.2099826788821347e-05,
+      "loss": 1.6103,
+      "step": 3370
+    },
+    {
+      "epoch": 0.44982652788897787,
+      "grad_norm": 0.9822551188681323,
+      "learning_rate": 1.2095601044376558e-05,
+      "loss": 1.5977,
+      "step": 3371
+    },
+    {
+      "epoch": 0.4499599679743795,
+      "grad_norm": 0.9614883569465209,
+      "learning_rate": 1.2091374908497352e-05,
+      "loss": 1.5943,
+      "step": 3372
+    },
+    {
+      "epoch": 0.45009340805978115,
+      "grad_norm": 1.0694103742791419,
+      "learning_rate": 1.2087148381973126e-05,
+      "loss": 1.6066,
+      "step": 3373
+    },
+    {
+      "epoch": 0.4502268481451828,
+      "grad_norm": 1.1709430242560819,
+      "learning_rate": 1.2082921465593345e-05,
+      "loss": 1.6409,
+      "step": 3374
+    },
+    {
+      "epoch": 0.4503602882305845,
+      "grad_norm": 0.9918589860941452,
+      "learning_rate": 1.2078694160147549e-05,
+      "loss": 1.5869,
+      "step": 3375
+    },
+    {
+      "epoch": 0.4504937283159861,
+      "grad_norm": 1.189016080561096,
+      "learning_rate": 1.2074466466425348e-05,
+      "loss": 1.6317,
+      "step": 3376
+    },
+    {
+      "epoch": 0.4506271684013878,
+      "grad_norm": 1.0693280010712716,
+      "learning_rate": 1.2070238385216431e-05,
+      "loss": 1.5298,
+      "step": 3377
+    },
+    {
+      "epoch": 0.45076060848678945,
+      "grad_norm": 0.9802287723386353,
+      "learning_rate": 1.2066009917310557e-05,
+      "loss": 1.5671,
+      "step": 3378
+    },
+    {
+      "epoch": 0.45089404857219106,
+      "grad_norm": 1.0001980093487308,
+      "learning_rate": 1.2061781063497549e-05,
+      "loss": 1.57,
+      "step": 3379
+    },
+    {
+      "epoch": 0.45102748865759273,
+      "grad_norm": 1.1240347451070862,
+      "learning_rate": 1.2057551824567315e-05,
+      "loss": 1.6214,
+      "step": 3380
+    },
+    {
+      "epoch": 0.4511609287429944,
+      "grad_norm": 1.016259709659121,
+      "learning_rate": 1.2053322201309827e-05,
+      "loss": 1.5418,
+      "step": 3381
+    },
+    {
+      "epoch": 0.4512943688283961,
+      "grad_norm": 0.9810480790681763,
+      "learning_rate": 1.2049092194515129e-05,
+      "loss": 1.5886,
+      "step": 3382
+    },
+    {
+      "epoch": 0.4514278089137977,
+      "grad_norm": 1.0999439267701696,
+      "learning_rate": 1.2044861804973339e-05,
+      "loss": 1.6381,
+      "step": 3383
+    },
+    {
+      "epoch": 0.45156124899919936,
+      "grad_norm": 0.9599654396714106,
+      "learning_rate": 1.2040631033474645e-05,
+      "loss": 1.5533,
+      "step": 3384
+    },
+    {
+      "epoch": 0.45169468908460103,
+      "grad_norm": 0.9466851026826987,
+      "learning_rate": 1.2036399880809307e-05,
+      "loss": 1.5944,
+      "step": 3385
+    },
+    {
+      "epoch": 0.45182812917000265,
+      "grad_norm": 0.9678219133269332,
+      "learning_rate": 1.2032168347767656e-05,
+      "loss": 1.6415,
+      "step": 3386
+    },
+    {
+      "epoch": 0.4519615692554043,
+      "grad_norm": 1.0940832917339585,
+      "learning_rate": 1.2027936435140097e-05,
+      "loss": 1.5896,
+      "step": 3387
+    },
+    {
+      "epoch": 0.452095009340806,
+      "grad_norm": 1.352412592314921,
+      "learning_rate": 1.2023704143717099e-05,
+      "loss": 1.5859,
+      "step": 3388
+    },
+    {
+      "epoch": 0.45222844942620766,
+      "grad_norm": 1.0442749639629683,
+      "learning_rate": 1.2019471474289209e-05,
+      "loss": 1.603,
+      "step": 3389
+    },
+    {
+      "epoch": 0.45236188951160927,
+      "grad_norm": 1.1089040085027504,
+      "learning_rate": 1.2015238427647039e-05,
+      "loss": 1.5855,
+      "step": 3390
+    },
+    {
+      "epoch": 0.45249532959701094,
+      "grad_norm": 1.172896894033313,
+      "learning_rate": 1.2011005004581275e-05,
+      "loss": 1.5967,
+      "step": 3391
+    },
+    {
+      "epoch": 0.4526287696824126,
+      "grad_norm": 1.0422201350270612,
+      "learning_rate": 1.2006771205882673e-05,
+      "loss": 1.588,
+      "step": 3392
+    },
+    {
+      "epoch": 0.4527622097678142,
+      "grad_norm": 0.9833998073404502,
+      "learning_rate": 1.2002537032342054e-05,
+      "loss": 1.6304,
+      "step": 3393
+    },
+    {
+      "epoch": 0.4528956498532159,
+      "grad_norm": 1.1092253938233356,
+      "learning_rate": 1.1998302484750322e-05,
+      "loss": 1.6304,
+      "step": 3394
+    },
+    {
+      "epoch": 0.45302908993861757,
+      "grad_norm": 1.0958444605435562,
+      "learning_rate": 1.1994067563898435e-05,
+      "loss": 1.5521,
+      "step": 3395
+    },
+    {
+      "epoch": 0.45316253002401924,
+      "grad_norm": 0.9941289195859265,
+      "learning_rate": 1.1989832270577432e-05,
+      "loss": 1.5917,
+      "step": 3396
+    },
+    {
+      "epoch": 0.45329597010942085,
+      "grad_norm": 1.0144543411061597,
+      "learning_rate": 1.1985596605578413e-05,
+      "loss": 1.5596,
+      "step": 3397
+    },
+    {
+      "epoch": 0.4534294101948225,
+      "grad_norm": 0.9660753780508365,
+      "learning_rate": 1.198136056969256e-05,
+      "loss": 1.5673,
+      "step": 3398
+    },
+    {
+      "epoch": 0.4535628502802242,
+      "grad_norm": 1.1410534595473867,
+      "learning_rate": 1.1977124163711108e-05,
+      "loss": 1.6167,
+      "step": 3399
+    },
+    {
+      "epoch": 0.4536962903656258,
+      "grad_norm": 0.9906940152019744,
+      "learning_rate": 1.1972887388425374e-05,
+      "loss": 1.5943,
+      "step": 3400
+    },
+    {
+      "epoch": 0.4538297304510275,
+      "grad_norm": 1.1280922031572023,
+      "learning_rate": 1.1968650244626733e-05,
+      "loss": 1.6018,
+      "step": 3401
+    },
+    {
+      "epoch": 0.45396317053642915,
+      "grad_norm": 1.2472440687519395,
+      "learning_rate": 1.1964412733106648e-05,
+      "loss": 1.5388,
+      "step": 3402
+    },
+    {
+      "epoch": 0.4540966106218308,
+      "grad_norm": 1.0569580420732267,
+      "learning_rate": 1.1960174854656623e-05,
+      "loss": 1.5699,
+      "step": 3403
+    },
+    {
+      "epoch": 0.45423005070723244,
+      "grad_norm": 1.0325493795551215,
+      "learning_rate": 1.1955936610068257e-05,
+      "loss": 1.6212,
+      "step": 3404
+    },
+    {
+      "epoch": 0.4543634907926341,
+      "grad_norm": 0.9811755437676621,
+      "learning_rate": 1.1951698000133203e-05,
+      "loss": 1.6113,
+      "step": 3405
+    },
+    {
+      "epoch": 0.4544969308780358,
+      "grad_norm": 1.2389436209658955,
+      "learning_rate": 1.1947459025643177e-05,
+      "loss": 1.6252,
+      "step": 3406
+    },
+    {
+      "epoch": 0.4546303709634374,
+      "grad_norm": 1.0656924772924845,
+      "learning_rate": 1.1943219687389984e-05,
+      "loss": 1.6018,
+      "step": 3407
+    },
+    {
+      "epoch": 0.45476381104883906,
+      "grad_norm": 1.1894263785585688,
+      "learning_rate": 1.1938979986165476e-05,
+      "loss": 1.6683,
+      "step": 3408
+    },
+    {
+      "epoch": 0.45489725113424073,
+      "grad_norm": 0.9518835451824228,
+      "learning_rate": 1.193473992276158e-05,
+      "loss": 1.6263,
+      "step": 3409
+    },
+    {
+      "epoch": 0.4550306912196424,
+      "grad_norm": 0.9790868312599547,
+      "learning_rate": 1.1930499497970296e-05,
+      "loss": 1.6023,
+      "step": 3410
+    },
+    {
+      "epoch": 0.455164131305044,
+      "grad_norm": 1.0311273154214786,
+      "learning_rate": 1.1926258712583685e-05,
+      "loss": 1.6393,
+      "step": 3411
+    },
+    {
+      "epoch": 0.4552975713904457,
+      "grad_norm": 0.9778706613876423,
+      "learning_rate": 1.192201756739388e-05,
+      "loss": 1.5959,
+      "step": 3412
+    },
+    {
+      "epoch": 0.45543101147584736,
+      "grad_norm": 1.0265470549844626,
+      "learning_rate": 1.1917776063193073e-05,
+      "loss": 1.5936,
+      "step": 3413
+    },
+    {
+      "epoch": 0.455564451561249,
+      "grad_norm": 1.0845896709498193,
+      "learning_rate": 1.1913534200773536e-05,
+      "loss": 1.5618,
+      "step": 3414
+    },
+    {
+      "epoch": 0.45569789164665064,
+      "grad_norm": 0.9625869411364787,
+      "learning_rate": 1.1909291980927592e-05,
+      "loss": 1.5905,
+      "step": 3415
+    },
+    {
+      "epoch": 0.4558313317320523,
+      "grad_norm": 1.072697736476755,
+      "learning_rate": 1.1905049404447649e-05,
+      "loss": 1.5894,
+      "step": 3416
+    },
+    {
+      "epoch": 0.455964771817454,
+      "grad_norm": 1.142515736949019,
+      "learning_rate": 1.1900806472126162e-05,
+      "loss": 1.5963,
+      "step": 3417
+    },
+    {
+      "epoch": 0.4560982119028556,
+      "grad_norm": 0.9816650958835474,
+      "learning_rate": 1.189656318475567e-05,
+      "loss": 1.6364,
+      "step": 3418
+    },
+    {
+      "epoch": 0.45623165198825727,
+      "grad_norm": 0.9293454502463179,
+      "learning_rate": 1.189231954312877e-05,
+      "loss": 1.5843,
+      "step": 3419
+    },
+    {
+      "epoch": 0.45636509207365894,
+      "grad_norm": 1.2000199775285119,
+      "learning_rate": 1.188807554803812e-05,
+      "loss": 1.6047,
+      "step": 3420
+    },
+    {
+      "epoch": 0.45649853215906055,
+      "grad_norm": 0.9742674279366978,
+      "learning_rate": 1.1883831200276459e-05,
+      "loss": 1.6254,
+      "step": 3421
+    },
+    {
+      "epoch": 0.4566319722444622,
+      "grad_norm": 1.1378422279774119,
+      "learning_rate": 1.1879586500636574e-05,
+      "loss": 1.5892,
+      "step": 3422
+    },
+    {
+      "epoch": 0.4567654123298639,
+      "grad_norm": 1.1369928812081609,
+      "learning_rate": 1.1875341449911333e-05,
+      "loss": 1.5573,
+      "step": 3423
+    },
+    {
+      "epoch": 0.45689885241526557,
+      "grad_norm": 0.9878064111116921,
+      "learning_rate": 1.1871096048893662e-05,
+      "loss": 1.6357,
+      "step": 3424
+    },
+    {
+      "epoch": 0.4570322925006672,
+      "grad_norm": 1.0138596892355174,
+      "learning_rate": 1.1866850298376549e-05,
+      "loss": 1.6267,
+      "step": 3425
+    },
+    {
+      "epoch": 0.45716573258606885,
+      "grad_norm": 1.03109296011423,
+      "learning_rate": 1.1862604199153058e-05,
+      "loss": 1.6326,
+      "step": 3426
+    },
+    {
+      "epoch": 0.4572991726714705,
+      "grad_norm": 0.983321243406952,
+      "learning_rate": 1.1858357752016307e-05,
+      "loss": 1.6337,
+      "step": 3427
+    },
+    {
+      "epoch": 0.45743261275687214,
+      "grad_norm": 1.069437667120727,
+      "learning_rate": 1.1854110957759487e-05,
+      "loss": 1.5948,
+      "step": 3428
+    },
+    {
+      "epoch": 0.4575660528422738,
+      "grad_norm": 0.9630689249149076,
+      "learning_rate": 1.1849863817175848e-05,
+      "loss": 1.6009,
+      "step": 3429
+    },
+    {
+      "epoch": 0.4576994929276755,
+      "grad_norm": 1.195697896821968,
+      "learning_rate": 1.1845616331058714e-05,
+      "loss": 1.6401,
+      "step": 3430
+    },
+    {
+      "epoch": 0.45783293301307715,
+      "grad_norm": 1.0373880035420473,
+      "learning_rate": 1.1841368500201457e-05,
+      "loss": 1.5717,
+      "step": 3431
+    },
+    {
+      "epoch": 0.45796637309847876,
+      "grad_norm": 1.0152163282536957,
+      "learning_rate": 1.1837120325397533e-05,
+      "loss": 1.6412,
+      "step": 3432
+    },
+    {
+      "epoch": 0.45809981318388043,
+      "grad_norm": 0.9465550717350613,
+      "learning_rate": 1.1832871807440448e-05,
+      "loss": 1.6038,
+      "step": 3433
+    },
+    {
+      "epoch": 0.4582332532692821,
+      "grad_norm": 0.9393643634807096,
+      "learning_rate": 1.1828622947123774e-05,
+      "loss": 1.542,
+      "step": 3434
+    },
+    {
+      "epoch": 0.4583666933546838,
+      "grad_norm": 0.9998488486015676,
+      "learning_rate": 1.1824373745241159e-05,
+      "loss": 1.6714,
+      "step": 3435
+    },
+    {
+      "epoch": 0.4585001334400854,
+      "grad_norm": 1.0463357172364034,
+      "learning_rate": 1.1820124202586294e-05,
+      "loss": 1.6897,
+      "step": 3436
+    },
+    {
+      "epoch": 0.45863357352548706,
+      "grad_norm": 0.9977756148556246,
+      "learning_rate": 1.1815874319952954e-05,
+      "loss": 1.5722,
+      "step": 3437
+    },
+    {
+      "epoch": 0.45876701361088873,
+      "grad_norm": 0.9781228945231666,
+      "learning_rate": 1.1811624098134963e-05,
+      "loss": 1.5516,
+      "step": 3438
+    },
+    {
+      "epoch": 0.45890045369629034,
+      "grad_norm": 0.9515422839609833,
+      "learning_rate": 1.180737353792622e-05,
+      "loss": 1.6011,
+      "step": 3439
+    },
+    {
+      "epoch": 0.459033893781692,
+      "grad_norm": 1.099348617644577,
+      "learning_rate": 1.1803122640120675e-05,
+      "loss": 1.5827,
+      "step": 3440
+    },
+    {
+      "epoch": 0.4591673338670937,
+      "grad_norm": 1.1211811802963583,
+      "learning_rate": 1.1798871405512352e-05,
+      "loss": 1.6251,
+      "step": 3441
+    },
+    {
+      "epoch": 0.45930077395249536,
+      "grad_norm": 0.9456734985827367,
+      "learning_rate": 1.1794619834895329e-05,
+      "loss": 1.5872,
+      "step": 3442
+    },
+    {
+      "epoch": 0.45943421403789697,
+      "grad_norm": 1.0345174270836286,
+      "learning_rate": 1.1790367929063756e-05,
+      "loss": 1.5859,
+      "step": 3443
+    },
+    {
+      "epoch": 0.45956765412329864,
+      "grad_norm": 1.1988418117321362,
+      "learning_rate": 1.1786115688811836e-05,
+      "loss": 1.6008,
+      "step": 3444
+    },
+    {
+      "epoch": 0.4597010942087003,
+      "grad_norm": 0.9823921150047465,
+      "learning_rate": 1.1781863114933845e-05,
+      "loss": 1.5628,
+      "step": 3445
+    },
+    {
+      "epoch": 0.4598345342941019,
+      "grad_norm": 0.9709054414773352,
+      "learning_rate": 1.1777610208224107e-05,
+      "loss": 1.592,
+      "step": 3446
+    },
+    {
+      "epoch": 0.4599679743795036,
+      "grad_norm": 0.913819442194726,
+      "learning_rate": 1.1773356969477023e-05,
+      "loss": 1.5924,
+      "step": 3447
+    },
+    {
+      "epoch": 0.46010141446490527,
+      "grad_norm": 0.9314527155118181,
+      "learning_rate": 1.1769103399487047e-05,
+      "loss": 1.5904,
+      "step": 3448
+    },
+    {
+      "epoch": 0.46023485455030694,
+      "grad_norm": 1.132160179087773,
+      "learning_rate": 1.1764849499048699e-05,
+      "loss": 1.6422,
+      "step": 3449
+    },
+    {
+      "epoch": 0.46036829463570855,
+      "grad_norm": 1.2291620905625844,
+      "learning_rate": 1.1760595268956556e-05,
+      "loss": 1.5999,
+      "step": 3450
+    },
+    {
+      "epoch": 0.4605017347211102,
+      "grad_norm": 0.990821816789439,
+      "learning_rate": 1.1756340710005264e-05,
+      "loss": 1.6038,
+      "step": 3451
+    },
+    {
+      "epoch": 0.4606351748065119,
+      "grad_norm": 1.1734233511272216,
+      "learning_rate": 1.175208582298952e-05,
+      "loss": 1.5486,
+      "step": 3452
+    },
+    {
+      "epoch": 0.4607686148919135,
+      "grad_norm": 1.0603222014465778,
+      "learning_rate": 1.1747830608704098e-05,
+      "loss": 1.5833,
+      "step": 3453
+    },
+    {
+      "epoch": 0.4609020549773152,
+      "grad_norm": 0.9788386077828661,
+      "learning_rate": 1.1743575067943813e-05,
+      "loss": 1.5973,
+      "step": 3454
+    },
+    {
+      "epoch": 0.46103549506271685,
+      "grad_norm": 1.1654930199086528,
+      "learning_rate": 1.1739319201503561e-05,
+      "loss": 1.6585,
+      "step": 3455
+    },
+    {
+      "epoch": 0.4611689351481185,
+      "grad_norm": 0.975258594253613,
+      "learning_rate": 1.1735063010178283e-05,
+      "loss": 1.5713,
+      "step": 3456
+    },
+    {
+      "epoch": 0.46130237523352013,
+      "grad_norm": 1.2641032198352828,
+      "learning_rate": 1.1730806494762987e-05,
+      "loss": 1.5982,
+      "step": 3457
+    },
+    {
+      "epoch": 0.4614358153189218,
+      "grad_norm": 1.242994584150048,
+      "learning_rate": 1.1726549656052748e-05,
+      "loss": 1.6169,
+      "step": 3458
+    },
+    {
+      "epoch": 0.4615692554043235,
+      "grad_norm": 0.9635164521389502,
+      "learning_rate": 1.1722292494842688e-05,
+      "loss": 1.598,
+      "step": 3459
+    },
+    {
+      "epoch": 0.4617026954897251,
+      "grad_norm": 0.9513310854478186,
+      "learning_rate": 1.1718035011928002e-05,
+      "loss": 1.5811,
+      "step": 3460
+    },
+    {
+      "epoch": 0.46183613557512676,
+      "grad_norm": 0.946112789103914,
+      "learning_rate": 1.1713777208103933e-05,
+      "loss": 1.5948,
+      "step": 3461
+    },
+    {
+      "epoch": 0.46196957566052843,
+      "grad_norm": 0.9946176475506616,
+      "learning_rate": 1.1709519084165797e-05,
+      "loss": 1.6119,
+      "step": 3462
+    },
+    {
+      "epoch": 0.4621030157459301,
+      "grad_norm": 0.9874079907996892,
+      "learning_rate": 1.1705260640908955e-05,
+      "loss": 1.6237,
+      "step": 3463
+    },
+    {
+      "epoch": 0.4622364558313317,
+      "grad_norm": 1.7891192519617156,
+      "learning_rate": 1.1701001879128843e-05,
+      "loss": 1.5662,
+      "step": 3464
+    },
+    {
+      "epoch": 0.4623698959167334,
+      "grad_norm": 1.0421651256059734,
+      "learning_rate": 1.1696742799620946e-05,
+      "loss": 1.6307,
+      "step": 3465
+    },
+    {
+      "epoch": 0.46250333600213506,
+      "grad_norm": 0.979356707410168,
+      "learning_rate": 1.1692483403180814e-05,
+      "loss": 1.607,
+      "step": 3466
+    },
+    {
+      "epoch": 0.46263677608753667,
+      "grad_norm": 1.1226620423723173,
+      "learning_rate": 1.1688223690604052e-05,
+      "loss": 1.5689,
+      "step": 3467
+    },
+    {
+      "epoch": 0.46277021617293834,
+      "grad_norm": 1.0965605702509604,
+      "learning_rate": 1.1683963662686324e-05,
+      "loss": 1.5973,
+      "step": 3468
+    },
+    {
+      "epoch": 0.46290365625834,
+      "grad_norm": 1.0921547887155878,
+      "learning_rate": 1.167970332022336e-05,
+      "loss": 1.5853,
+      "step": 3469
+    },
+    {
+      "epoch": 0.4630370963437417,
+      "grad_norm": 1.1736803003382774,
+      "learning_rate": 1.1675442664010935e-05,
+      "loss": 1.5885,
+      "step": 3470
+    },
+    {
+      "epoch": 0.4631705364291433,
+      "grad_norm": 1.082987731836864,
+      "learning_rate": 1.1671181694844897e-05,
+      "loss": 1.6115,
+      "step": 3471
+    },
+    {
+      "epoch": 0.46330397651454497,
+      "grad_norm": 0.9708478904094577,
+      "learning_rate": 1.1666920413521146e-05,
+      "loss": 1.597,
+      "step": 3472
+    },
+    {
+      "epoch": 0.46343741659994664,
+      "grad_norm": 1.0126029044158875,
+      "learning_rate": 1.1662658820835639e-05,
+      "loss": 1.6249,
+      "step": 3473
+    },
+    {
+      "epoch": 0.46357085668534825,
+      "grad_norm": 1.0207349433202457,
+      "learning_rate": 1.1658396917584397e-05,
+      "loss": 1.5908,
+      "step": 3474
+    },
+    {
+      "epoch": 0.4637042967707499,
+      "grad_norm": 0.9372775427319033,
+      "learning_rate": 1.1654134704563492e-05,
+      "loss": 1.6199,
+      "step": 3475
+    },
+    {
+      "epoch": 0.4638377368561516,
+      "grad_norm": 1.1272128181343017,
+      "learning_rate": 1.1649872182569058e-05,
+      "loss": 1.6617,
+      "step": 3476
+    },
+    {
+      "epoch": 0.46397117694155326,
+      "grad_norm": 1.110304791632166,
+      "learning_rate": 1.1645609352397282e-05,
+      "loss": 1.6041,
+      "step": 3477
+    },
+    {
+      "epoch": 0.4641046170269549,
+      "grad_norm": 1.0017766303846791,
+      "learning_rate": 1.1641346214844417e-05,
+      "loss": 1.6007,
+      "step": 3478
+    },
+    {
+      "epoch": 0.46423805711235655,
+      "grad_norm": 0.9880107584390453,
+      "learning_rate": 1.1637082770706764e-05,
+      "loss": 1.6511,
+      "step": 3479
+    },
+    {
+      "epoch": 0.4643714971977582,
+      "grad_norm": 1.2922751760310007,
+      "learning_rate": 1.1632819020780693e-05,
+      "loss": 1.6072,
+      "step": 3480
+    },
+    {
+      "epoch": 0.46450493728315984,
+      "grad_norm": 0.9883475103271387,
+      "learning_rate": 1.1628554965862615e-05,
+      "loss": 1.6685,
+      "step": 3481
+    },
+    {
+      "epoch": 0.4646383773685615,
+      "grad_norm": 1.1928187505375785,
+      "learning_rate": 1.1624290606749012e-05,
+      "loss": 1.6114,
+      "step": 3482
+    },
+    {
+      "epoch": 0.4647718174539632,
+      "grad_norm": 1.0066345866358597,
+      "learning_rate": 1.1620025944236418e-05,
+      "loss": 1.6761,
+      "step": 3483
+    },
+    {
+      "epoch": 0.46490525753936485,
+      "grad_norm": 1.0953896264697256,
+      "learning_rate": 1.161576097912142e-05,
+      "loss": 1.5629,
+      "step": 3484
+    },
+    {
+      "epoch": 0.46503869762476646,
+      "grad_norm": 0.984883775282697,
+      "learning_rate": 1.161149571220067e-05,
+      "loss": 1.5929,
+      "step": 3485
+    },
+    {
+      "epoch": 0.46517213771016813,
+      "grad_norm": 0.9445601434963037,
+      "learning_rate": 1.1607230144270866e-05,
+      "loss": 1.6208,
+      "step": 3486
+    },
+    {
+      "epoch": 0.4653055777955698,
+      "grad_norm": 0.9240480810034248,
+      "learning_rate": 1.1602964276128774e-05,
+      "loss": 1.5202,
+      "step": 3487
+    },
+    {
+      "epoch": 0.4654390178809714,
+      "grad_norm": 1.2349918808254974,
+      "learning_rate": 1.1598698108571205e-05,
+      "loss": 1.5377,
+      "step": 3488
+    },
+    {
+      "epoch": 0.4655724579663731,
+      "grad_norm": 1.0096021895282354,
+      "learning_rate": 1.1594431642395027e-05,
+      "loss": 1.593,
+      "step": 3489
+    },
+    {
+      "epoch": 0.46570589805177476,
+      "grad_norm": 1.0714154627752588,
+      "learning_rate": 1.1590164878397174e-05,
+      "loss": 1.5892,
+      "step": 3490
+    },
+    {
+      "epoch": 0.46583933813717643,
+      "grad_norm": 0.9628486041384523,
+      "learning_rate": 1.1585897817374628e-05,
+      "loss": 1.5884,
+      "step": 3491
+    },
+    {
+      "epoch": 0.46597277822257804,
+      "grad_norm": 0.9752330456513079,
+      "learning_rate": 1.1581630460124424e-05,
+      "loss": 1.576,
+      "step": 3492
+    },
+    {
+      "epoch": 0.4661062183079797,
+      "grad_norm": 0.9802951196533614,
+      "learning_rate": 1.1577362807443657e-05,
+      "loss": 1.5754,
+      "step": 3493
+    },
+    {
+      "epoch": 0.4662396583933814,
+      "grad_norm": 1.1032933740856454,
+      "learning_rate": 1.1573094860129479e-05,
+      "loss": 1.5705,
+      "step": 3494
+    },
+    {
+      "epoch": 0.466373098478783,
+      "grad_norm": 0.9835441668655777,
+      "learning_rate": 1.1568826618979087e-05,
+      "loss": 1.6197,
+      "step": 3495
+    },
+    {
+      "epoch": 0.46650653856418467,
+      "grad_norm": 0.9873490655801218,
+      "learning_rate": 1.1564558084789749e-05,
+      "loss": 1.6491,
+      "step": 3496
+    },
+    {
+      "epoch": 0.46663997864958634,
+      "grad_norm": 1.1431057876156487,
+      "learning_rate": 1.1560289258358773e-05,
+      "loss": 1.6006,
+      "step": 3497
+    },
+    {
+      "epoch": 0.466773418734988,
+      "grad_norm": 1.1765306955508206,
+      "learning_rate": 1.1556020140483523e-05,
+      "loss": 1.6422,
+      "step": 3498
+    },
+    {
+      "epoch": 0.4669068588203896,
+      "grad_norm": 0.900400853861777,
+      "learning_rate": 1.1551750731961433e-05,
+      "loss": 1.575,
+      "step": 3499
+    },
+    {
+      "epoch": 0.4670402989057913,
+      "grad_norm": 0.957183643780711,
+      "learning_rate": 1.1547481033589971e-05,
+      "loss": 1.5993,
+      "step": 3500
+    },
+    {
+      "epoch": 0.46717373899119297,
+      "grad_norm": 1.1014233549217831,
+      "learning_rate": 1.1543211046166672e-05,
+      "loss": 1.5655,
+      "step": 3501
+    },
+    {
+      "epoch": 0.46730717907659464,
+      "grad_norm": 0.9656596890342347,
+      "learning_rate": 1.1538940770489118e-05,
+      "loss": 1.5974,
+      "step": 3502
+    },
+    {
+      "epoch": 0.46744061916199625,
+      "grad_norm": 1.1217113813358535,
+      "learning_rate": 1.1534670207354952e-05,
+      "loss": 1.6228,
+      "step": 3503
+    },
+    {
+      "epoch": 0.4675740592473979,
+      "grad_norm": 1.0088247160067236,
+      "learning_rate": 1.1530399357561861e-05,
+      "loss": 1.6261,
+      "step": 3504
+    },
+    {
+      "epoch": 0.4677074993327996,
+      "grad_norm": 1.0056946912456848,
+      "learning_rate": 1.1526128221907595e-05,
+      "loss": 1.572,
+      "step": 3505
+    },
+    {
+      "epoch": 0.4678409394182012,
+      "grad_norm": 0.984689525859442,
+      "learning_rate": 1.1521856801189954e-05,
+      "loss": 1.5713,
+      "step": 3506
+    },
+    {
+      "epoch": 0.4679743795036029,
+      "grad_norm": 0.9741585900265354,
+      "learning_rate": 1.1517585096206788e-05,
+      "loss": 1.587,
+      "step": 3507
+    },
+    {
+      "epoch": 0.46810781958900455,
+      "grad_norm": 0.9889481536506061,
+      "learning_rate": 1.1513313107756007e-05,
+      "loss": 1.5535,
+      "step": 3508
+    },
+    {
+      "epoch": 0.4682412596744062,
+      "grad_norm": 0.9996665079177771,
+      "learning_rate": 1.1509040836635568e-05,
+      "loss": 1.5746,
+      "step": 3509
+    },
+    {
+      "epoch": 0.46837469975980783,
+      "grad_norm": 1.1930390389598884,
+      "learning_rate": 1.1504768283643476e-05,
+      "loss": 1.5963,
+      "step": 3510
+    },
+    {
+      "epoch": 0.4685081398452095,
+      "grad_norm": 1.015228011873478,
+      "learning_rate": 1.1500495449577806e-05,
+      "loss": 1.5463,
+      "step": 3511
+    },
+    {
+      "epoch": 0.4686415799306112,
+      "grad_norm": 0.9872673249391366,
+      "learning_rate": 1.149622233523667e-05,
+      "loss": 1.5713,
+      "step": 3512
+    },
+    {
+      "epoch": 0.4687750200160128,
+      "grad_norm": 1.1412360147938814,
+      "learning_rate": 1.1491948941418234e-05,
+      "loss": 1.6021,
+      "step": 3513
+    },
+    {
+      "epoch": 0.46890846010141446,
+      "grad_norm": 1.0170872902758004,
+      "learning_rate": 1.1487675268920721e-05,
+      "loss": 1.6049,
+      "step": 3514
+    },
+    {
+      "epoch": 0.46904190018681613,
+      "grad_norm": 0.9653090890079009,
+      "learning_rate": 1.148340131854241e-05,
+      "loss": 1.5985,
+      "step": 3515
+    },
+    {
+      "epoch": 0.4691753402722178,
+      "grad_norm": 1.0182190080319995,
+      "learning_rate": 1.147912709108162e-05,
+      "loss": 1.6062,
+      "step": 3516
+    },
+    {
+      "epoch": 0.4693087803576194,
+      "grad_norm": 1.1931214012383156,
+      "learning_rate": 1.1474852587336731e-05,
+      "loss": 1.6188,
+      "step": 3517
+    },
+    {
+      "epoch": 0.4694422204430211,
+      "grad_norm": 0.9521461072306361,
+      "learning_rate": 1.147057780810617e-05,
+      "loss": 1.546,
+      "step": 3518
+    },
+    {
+      "epoch": 0.46957566052842276,
+      "grad_norm": 1.0407743299860555,
+      "learning_rate": 1.1466302754188417e-05,
+      "loss": 1.6001,
+      "step": 3519
+    },
+    {
+      "epoch": 0.46970910061382437,
+      "grad_norm": 1.0007245963172906,
+      "learning_rate": 1.1462027426382002e-05,
+      "loss": 1.5495,
+      "step": 3520
+    },
+    {
+      "epoch": 0.46984254069922604,
+      "grad_norm": 1.00485415407387,
+      "learning_rate": 1.145775182548551e-05,
+      "loss": 1.5786,
+      "step": 3521
+    },
+    {
+      "epoch": 0.4699759807846277,
+      "grad_norm": 0.9919189415028729,
+      "learning_rate": 1.1453475952297577e-05,
+      "loss": 1.6291,
+      "step": 3522
+    },
+    {
+      "epoch": 0.4701094208700294,
+      "grad_norm": 1.1020789294943216,
+      "learning_rate": 1.1449199807616882e-05,
+      "loss": 1.533,
+      "step": 3523
+    },
+    {
+      "epoch": 0.470242860955431,
+      "grad_norm": 1.3158857310226424,
+      "learning_rate": 1.1444923392242165e-05,
+      "loss": 1.6232,
+      "step": 3524
+    },
+    {
+      "epoch": 0.47037630104083267,
+      "grad_norm": 1.1748271233312169,
+      "learning_rate": 1.1440646706972207e-05,
+      "loss": 1.5839,
+      "step": 3525
+    },
+    {
+      "epoch": 0.47050974112623434,
+      "grad_norm": 0.9926376613394198,
+      "learning_rate": 1.143636975260585e-05,
+      "loss": 1.6421,
+      "step": 3526
+    },
+    {
+      "epoch": 0.47064318121163595,
+      "grad_norm": 0.9974971729341541,
+      "learning_rate": 1.1432092529941972e-05,
+      "loss": 1.6288,
+      "step": 3527
+    },
+    {
+      "epoch": 0.4707766212970376,
+      "grad_norm": 1.07546662467877,
+      "learning_rate": 1.142781503977952e-05,
+      "loss": 1.597,
+      "step": 3528
+    },
+    {
+      "epoch": 0.4709100613824393,
+      "grad_norm": 1.0889300149702186,
+      "learning_rate": 1.1423537282917469e-05,
+      "loss": 1.5977,
+      "step": 3529
+    },
+    {
+      "epoch": 0.47104350146784096,
+      "grad_norm": 1.081280310987519,
+      "learning_rate": 1.1419259260154864e-05,
+      "loss": 1.5912,
+      "step": 3530
+    },
+    {
+      "epoch": 0.4711769415532426,
+      "grad_norm": 1.0372628884781476,
+      "learning_rate": 1.141498097229079e-05,
+      "loss": 1.6334,
+      "step": 3531
+    },
+    {
+      "epoch": 0.47131038163864425,
+      "grad_norm": 1.1883125180451641,
+      "learning_rate": 1.1410702420124377e-05,
+      "loss": 1.6159,
+      "step": 3532
+    },
+    {
+      "epoch": 0.4714438217240459,
+      "grad_norm": 0.9457223363765193,
+      "learning_rate": 1.1406423604454816e-05,
+      "loss": 1.6048,
+      "step": 3533
+    },
+    {
+      "epoch": 0.47157726180944753,
+      "grad_norm": 1.1775437736757077,
+      "learning_rate": 1.1402144526081338e-05,
+      "loss": 1.5696,
+      "step": 3534
+    },
+    {
+      "epoch": 0.4717107018948492,
+      "grad_norm": 0.9405003660186544,
+      "learning_rate": 1.1397865185803227e-05,
+      "loss": 1.6001,
+      "step": 3535
+    },
+    {
+      "epoch": 0.4718441419802509,
+      "grad_norm": 0.9593945256058269,
+      "learning_rate": 1.1393585584419812e-05,
+      "loss": 1.5745,
+      "step": 3536
+    },
+    {
+      "epoch": 0.47197758206565255,
+      "grad_norm": 1.037264876154079,
+      "learning_rate": 1.1389305722730478e-05,
+      "loss": 1.5845,
+      "step": 3537
+    },
+    {
+      "epoch": 0.47211102215105416,
+      "grad_norm": 1.0164422235337136,
+      "learning_rate": 1.1385025601534654e-05,
+      "loss": 1.5994,
+      "step": 3538
+    },
+    {
+      "epoch": 0.47224446223645583,
+      "grad_norm": 1.0015580982419405,
+      "learning_rate": 1.1380745221631813e-05,
+      "loss": 1.5474,
+      "step": 3539
+    },
+    {
+      "epoch": 0.4723779023218575,
+      "grad_norm": 0.9922512359099007,
+      "learning_rate": 1.137646458382149e-05,
+      "loss": 1.6085,
+      "step": 3540
+    },
+    {
+      "epoch": 0.4725113424072591,
+      "grad_norm": 0.9513030545535126,
+      "learning_rate": 1.137218368890325e-05,
+      "loss": 1.583,
+      "step": 3541
+    },
+    {
+      "epoch": 0.4726447824926608,
+      "grad_norm": 1.0137343494231201,
+      "learning_rate": 1.1367902537676722e-05,
+      "loss": 1.6333,
+      "step": 3542
+    },
+    {
+      "epoch": 0.47277822257806246,
+      "grad_norm": 1.053043019819065,
+      "learning_rate": 1.1363621130941573e-05,
+      "loss": 1.5458,
+      "step": 3543
+    },
+    {
+      "epoch": 0.4729116626634641,
+      "grad_norm": 0.991708813143814,
+      "learning_rate": 1.1359339469497525e-05,
+      "loss": 1.5617,
+      "step": 3544
+    },
+    {
+      "epoch": 0.47304510274886574,
+      "grad_norm": 0.9881479349845488,
+      "learning_rate": 1.1355057554144338e-05,
+      "loss": 1.5599,
+      "step": 3545
+    },
+    {
+      "epoch": 0.4731785428342674,
+      "grad_norm": 0.9677599502311968,
+      "learning_rate": 1.1350775385681827e-05,
+      "loss": 1.661,
+      "step": 3546
+    },
+    {
+      "epoch": 0.4733119829196691,
+      "grad_norm": 0.9637140530562013,
+      "learning_rate": 1.1346492964909856e-05,
+      "loss": 1.5783,
+      "step": 3547
+    },
+    {
+      "epoch": 0.4734454230050707,
+      "grad_norm": 1.0174979764772296,
+      "learning_rate": 1.1342210292628327e-05,
+      "loss": 1.6682,
+      "step": 3548
+    },
+    {
+      "epoch": 0.47357886309047237,
+      "grad_norm": 0.956767731266162,
+      "learning_rate": 1.1337927369637198e-05,
+      "loss": 1.6158,
+      "step": 3549
+    },
+    {
+      "epoch": 0.47371230317587404,
+      "grad_norm": 1.1908661498029431,
+      "learning_rate": 1.1333644196736468e-05,
+      "loss": 1.5445,
+      "step": 3550
+    },
+    {
+      "epoch": 0.4738457432612757,
+      "grad_norm": 1.0665948861651355,
+      "learning_rate": 1.132936077472619e-05,
+      "loss": 1.6007,
+      "step": 3551
+    },
+    {
+      "epoch": 0.4739791833466773,
+      "grad_norm": 0.9949979990909358,
+      "learning_rate": 1.1325077104406455e-05,
+      "loss": 1.6371,
+      "step": 3552
+    },
+    {
+      "epoch": 0.474112623432079,
+      "grad_norm": 1.1119040943652865,
+      "learning_rate": 1.1320793186577398e-05,
+      "loss": 1.5935,
+      "step": 3553
+    },
+    {
+      "epoch": 0.47424606351748066,
+      "grad_norm": 1.2024374524083807,
+      "learning_rate": 1.1316509022039215e-05,
+      "loss": 1.5906,
+      "step": 3554
+    },
+    {
+      "epoch": 0.4743795036028823,
+      "grad_norm": 0.9908007603433848,
+      "learning_rate": 1.1312224611592132e-05,
+      "loss": 1.6032,
+      "step": 3555
+    },
+    {
+      "epoch": 0.47451294368828395,
+      "grad_norm": 1.0257647339374822,
+      "learning_rate": 1.1307939956036437e-05,
+      "loss": 1.6118,
+      "step": 3556
+    },
+    {
+      "epoch": 0.4746463837736856,
+      "grad_norm": 0.9890146650400216,
+      "learning_rate": 1.1303655056172447e-05,
+      "loss": 1.6011,
+      "step": 3557
+    },
+    {
+      "epoch": 0.4747798238590873,
+      "grad_norm": 8.167075213418878,
+      "learning_rate": 1.1299369912800537e-05,
+      "loss": 1.5942,
+      "step": 3558
+    },
+    {
+      "epoch": 0.4749132639444889,
+      "grad_norm": 0.9783430062632217,
+      "learning_rate": 1.1295084526721119e-05,
+      "loss": 1.5433,
+      "step": 3559
+    },
+    {
+      "epoch": 0.4750467040298906,
+      "grad_norm": 1.0096065716017133,
+      "learning_rate": 1.129079889873466e-05,
+      "loss": 1.6455,
+      "step": 3560
+    },
+    {
+      "epoch": 0.47518014411529225,
+      "grad_norm": 1.0537017716025912,
+      "learning_rate": 1.1286513029641657e-05,
+      "loss": 1.5721,
+      "step": 3561
+    },
+    {
+      "epoch": 0.47531358420069386,
+      "grad_norm": 1.1698259882179534,
+      "learning_rate": 1.1282226920242669e-05,
+      "loss": 1.5693,
+      "step": 3562
+    },
+    {
+      "epoch": 0.47544702428609553,
+      "grad_norm": 1.096536277070531,
+      "learning_rate": 1.1277940571338296e-05,
+      "loss": 1.651,
+      "step": 3563
+    },
+    {
+      "epoch": 0.4755804643714972,
+      "grad_norm": 0.9870886196000643,
+      "learning_rate": 1.1273653983729169e-05,
+      "loss": 1.5927,
+      "step": 3564
+    },
+    {
+      "epoch": 0.4757139044568989,
+      "grad_norm": 1.0227690927337232,
+      "learning_rate": 1.1269367158215982e-05,
+      "loss": 1.5799,
+      "step": 3565
+    },
+    {
+      "epoch": 0.4758473445423005,
+      "grad_norm": 0.981433243463809,
+      "learning_rate": 1.1265080095599459e-05,
+      "loss": 1.6033,
+      "step": 3566
+    },
+    {
+      "epoch": 0.47598078462770216,
+      "grad_norm": 0.9241540078980625,
+      "learning_rate": 1.126079279668038e-05,
+      "loss": 1.6059,
+      "step": 3567
+    },
+    {
+      "epoch": 0.47611422471310383,
+      "grad_norm": 0.9966040467022464,
+      "learning_rate": 1.1256505262259561e-05,
+      "loss": 1.6017,
+      "step": 3568
+    },
+    {
+      "epoch": 0.47624766479850544,
+      "grad_norm": 1.00788906341642,
+      "learning_rate": 1.1252217493137863e-05,
+      "loss": 1.6031,
+      "step": 3569
+    },
+    {
+      "epoch": 0.4763811048839071,
+      "grad_norm": 1.1866583647363547,
+      "learning_rate": 1.1247929490116198e-05,
+      "loss": 1.5722,
+      "step": 3570
+    },
+    {
+      "epoch": 0.4765145449693088,
+      "grad_norm": 1.0142892640611485,
+      "learning_rate": 1.124364125399551e-05,
+      "loss": 1.6128,
+      "step": 3571
+    },
+    {
+      "epoch": 0.47664798505471045,
+      "grad_norm": 0.9882028516457761,
+      "learning_rate": 1.1239352785576795e-05,
+      "loss": 1.6019,
+      "step": 3572
+    },
+    {
+      "epoch": 0.47678142514011207,
+      "grad_norm": 0.9485363827235442,
+      "learning_rate": 1.1235064085661094e-05,
+      "loss": 1.5451,
+      "step": 3573
+    },
+    {
+      "epoch": 0.47691486522551374,
+      "grad_norm": 1.0423832728014304,
+      "learning_rate": 1.1230775155049478e-05,
+      "loss": 1.6573,
+      "step": 3574
+    },
+    {
+      "epoch": 0.4770483053109154,
+      "grad_norm": 0.9522899518242469,
+      "learning_rate": 1.122648599454308e-05,
+      "loss": 1.5389,
+      "step": 3575
+    },
+    {
+      "epoch": 0.4771817453963171,
+      "grad_norm": 1.0795989027553348,
+      "learning_rate": 1.122219660494306e-05,
+      "loss": 1.5245,
+      "step": 3576
+    },
+    {
+      "epoch": 0.4773151854817187,
+      "grad_norm": 0.9703295676591189,
+      "learning_rate": 1.121790698705063e-05,
+      "loss": 1.6262,
+      "step": 3577
+    },
+    {
+      "epoch": 0.47744862556712037,
+      "grad_norm": 0.9466306660145242,
+      "learning_rate": 1.1213617141667042e-05,
+      "loss": 1.5771,
+      "step": 3578
+    },
+    {
+      "epoch": 0.47758206565252204,
+      "grad_norm": 1.1652671591625094,
+      "learning_rate": 1.1209327069593587e-05,
+      "loss": 1.6108,
+      "step": 3579
+    },
+    {
+      "epoch": 0.47771550573792365,
+      "grad_norm": 0.9637891663914724,
+      "learning_rate": 1.1205036771631606e-05,
+      "loss": 1.6177,
+      "step": 3580
+    },
+    {
+      "epoch": 0.4778489458233253,
+      "grad_norm": 1.0591406701546173,
+      "learning_rate": 1.1200746248582478e-05,
+      "loss": 1.6277,
+      "step": 3581
+    },
+    {
+      "epoch": 0.477982385908727,
+      "grad_norm": 1.075013785861123,
+      "learning_rate": 1.1196455501247619e-05,
+      "loss": 1.6509,
+      "step": 3582
+    },
+    {
+      "epoch": 0.47811582599412866,
+      "grad_norm": 1.0099560088443318,
+      "learning_rate": 1.1192164530428495e-05,
+      "loss": 1.5818,
+      "step": 3583
+    },
+    {
+      "epoch": 0.4782492660795303,
+      "grad_norm": 1.2309318054357847,
+      "learning_rate": 1.1187873336926609e-05,
+      "loss": 1.5818,
+      "step": 3584
+    },
+    {
+      "epoch": 0.47838270616493195,
+      "grad_norm": 1.0494190297239068,
+      "learning_rate": 1.1183581921543507e-05,
+      "loss": 1.5966,
+      "step": 3585
+    },
+    {
+      "epoch": 0.4785161462503336,
+      "grad_norm": 0.9551043069869481,
+      "learning_rate": 1.1179290285080782e-05,
+      "loss": 1.5893,
+      "step": 3586
+    },
+    {
+      "epoch": 0.47864958633573523,
+      "grad_norm": 0.9368775120300123,
+      "learning_rate": 1.1174998428340055e-05,
+      "loss": 1.6137,
+      "step": 3587
+    },
+    {
+      "epoch": 0.4787830264211369,
+      "grad_norm": 1.1506412501949679,
+      "learning_rate": 1.1170706352123002e-05,
+      "loss": 1.6004,
+      "step": 3588
+    },
+    {
+      "epoch": 0.4789164665065386,
+      "grad_norm": 1.0842690119366971,
+      "learning_rate": 1.116641405723133e-05,
+      "loss": 1.6812,
+      "step": 3589
+    },
+    {
+      "epoch": 0.47904990659194024,
+      "grad_norm": 1.1775677070138726,
+      "learning_rate": 1.1162121544466794e-05,
+      "loss": 1.6091,
+      "step": 3590
+    },
+    {
+      "epoch": 0.47918334667734186,
+      "grad_norm": 0.9567709845629377,
+      "learning_rate": 1.1157828814631179e-05,
+      "loss": 1.596,
+      "step": 3591
+    },
+    {
+      "epoch": 0.47931678676274353,
+      "grad_norm": 0.970022267558387,
+      "learning_rate": 1.115353586852633e-05,
+      "loss": 1.657,
+      "step": 3592
+    },
+    {
+      "epoch": 0.4794502268481452,
+      "grad_norm": 0.9647653037200512,
+      "learning_rate": 1.1149242706954111e-05,
+      "loss": 1.5747,
+      "step": 3593
+    },
+    {
+      "epoch": 0.4795836669335468,
+      "grad_norm": 1.0040826902190914,
+      "learning_rate": 1.1144949330716441e-05,
+      "loss": 1.624,
+      "step": 3594
+    },
+    {
+      "epoch": 0.4797171070189485,
+      "grad_norm": 7.924396978950378,
+      "learning_rate": 1.1140655740615274e-05,
+      "loss": 1.6323,
+      "step": 3595
+    },
+    {
+      "epoch": 0.47985054710435016,
+      "grad_norm": 1.2546815828202986,
+      "learning_rate": 1.1136361937452595e-05,
+      "loss": 1.5693,
+      "step": 3596
+    },
+    {
+      "epoch": 0.4799839871897518,
+      "grad_norm": 1.0142710081731812,
+      "learning_rate": 1.113206792203045e-05,
+      "loss": 1.5772,
+      "step": 3597
+    },
+    {
+      "epoch": 0.48011742727515344,
+      "grad_norm": 0.9910079104514572,
+      "learning_rate": 1.1127773695150904e-05,
+      "loss": 1.6178,
+      "step": 3598
+    },
+    {
+      "epoch": 0.4802508673605551,
+      "grad_norm": 0.9866567619524242,
+      "learning_rate": 1.1123479257616072e-05,
+      "loss": 1.5508,
+      "step": 3599
+    },
+    {
+      "epoch": 0.4803843074459568,
+      "grad_norm": 1.0128213856045458,
+      "learning_rate": 1.1119184610228103e-05,
+      "loss": 1.5817,
+      "step": 3600
+    },
+    {
+      "epoch": 0.4805177475313584,
+      "grad_norm": 1.07285721769306,
+      "learning_rate": 1.1114889753789193e-05,
+      "loss": 1.5987,
+      "step": 3601
+    },
+    {
+      "epoch": 0.48065118761676007,
+      "grad_norm": 1.1570972062497342,
+      "learning_rate": 1.1110594689101572e-05,
+      "loss": 1.5853,
+      "step": 3602
+    },
+    {
+      "epoch": 0.48078462770216174,
+      "grad_norm": 0.9546422737264942,
+      "learning_rate": 1.1106299416967508e-05,
+      "loss": 1.5883,
+      "step": 3603
+    },
+    {
+      "epoch": 0.4809180677875634,
+      "grad_norm": 1.2179319826085693,
+      "learning_rate": 1.1102003938189308e-05,
+      "loss": 1.6357,
+      "step": 3604
+    },
+    {
+      "epoch": 0.481051507872965,
+      "grad_norm": 1.016415996934519,
+      "learning_rate": 1.1097708253569317e-05,
+      "loss": 1.5237,
+      "step": 3605
+    },
+    {
+      "epoch": 0.4811849479583667,
+      "grad_norm": 1.0641665463344543,
+      "learning_rate": 1.1093412363909926e-05,
+      "loss": 1.5878,
+      "step": 3606
+    },
+    {
+      "epoch": 0.48131838804376836,
+      "grad_norm": 0.9711099512141003,
+      "learning_rate": 1.1089116270013552e-05,
+      "loss": 1.5518,
+      "step": 3607
+    },
+    {
+      "epoch": 0.48145182812917,
+      "grad_norm": 1.2233276218733482,
+      "learning_rate": 1.108481997268266e-05,
+      "loss": 1.5926,
+      "step": 3608
+    },
+    {
+      "epoch": 0.48158526821457165,
+      "grad_norm": 1.1743091714878038,
+      "learning_rate": 1.1080523472719745e-05,
+      "loss": 1.5618,
+      "step": 3609
+    },
+    {
+      "epoch": 0.4817187082999733,
+      "grad_norm": 1.226731129476479,
+      "learning_rate": 1.1076226770927349e-05,
+      "loss": 1.5961,
+      "step": 3610
+    },
+    {
+      "epoch": 0.481852148385375,
+      "grad_norm": 1.061342840791177,
+      "learning_rate": 1.1071929868108046e-05,
+      "loss": 1.6121,
+      "step": 3611
+    },
+    {
+      "epoch": 0.4819855884707766,
+      "grad_norm": 1.0043889707036777,
+      "learning_rate": 1.1067632765064449e-05,
+      "loss": 1.559,
+      "step": 3612
+    },
+    {
+      "epoch": 0.4821190285561783,
+      "grad_norm": 0.9334874178935394,
+      "learning_rate": 1.1063335462599208e-05,
+      "loss": 1.6363,
+      "step": 3613
+    },
+    {
+      "epoch": 0.48225246864157995,
+      "grad_norm": 0.9541127453212594,
+      "learning_rate": 1.1059037961515005e-05,
+      "loss": 1.5859,
+      "step": 3614
+    },
+    {
+      "epoch": 0.48238590872698156,
+      "grad_norm": 1.2659844329901926,
+      "learning_rate": 1.1054740262614571e-05,
+      "loss": 1.6083,
+      "step": 3615
+    },
+    {
+      "epoch": 0.48251934881238323,
+      "grad_norm": 1.009394331400396,
+      "learning_rate": 1.1050442366700666e-05,
+      "loss": 1.5752,
+      "step": 3616
+    },
+    {
+      "epoch": 0.4826527888977849,
+      "grad_norm": 0.9588149443609268,
+      "learning_rate": 1.1046144274576085e-05,
+      "loss": 1.5623,
+      "step": 3617
+    },
+    {
+      "epoch": 0.48278622898318657,
+      "grad_norm": 1.1186126571235782,
+      "learning_rate": 1.1041845987043664e-05,
+      "loss": 1.5748,
+      "step": 3618
+    },
+    {
+      "epoch": 0.4829196690685882,
+      "grad_norm": 1.1820677624653801,
+      "learning_rate": 1.1037547504906275e-05,
+      "loss": 1.5901,
+      "step": 3619
+    },
+    {
+      "epoch": 0.48305310915398986,
+      "grad_norm": 0.9674980847745737,
+      "learning_rate": 1.1033248828966825e-05,
+      "loss": 1.5353,
+      "step": 3620
+    },
+    {
+      "epoch": 0.4831865492393915,
+      "grad_norm": 1.037410201754664,
+      "learning_rate": 1.1028949960028257e-05,
+      "loss": 1.5884,
+      "step": 3621
+    },
+    {
+      "epoch": 0.48331998932479314,
+      "grad_norm": 1.0179633311651575,
+      "learning_rate": 1.1024650898893554e-05,
+      "loss": 1.6185,
+      "step": 3622
+    },
+    {
+      "epoch": 0.4834534294101948,
+      "grad_norm": 1.024857123528208,
+      "learning_rate": 1.1020351646365726e-05,
+      "loss": 1.6043,
+      "step": 3623
+    },
+    {
+      "epoch": 0.4835868694955965,
+      "grad_norm": 1.0553654696487733,
+      "learning_rate": 1.1016052203247829e-05,
+      "loss": 1.5471,
+      "step": 3624
+    },
+    {
+      "epoch": 0.48372030958099815,
+      "grad_norm": 0.9703280830869259,
+      "learning_rate": 1.1011752570342949e-05,
+      "loss": 1.4886,
+      "step": 3625
+    },
+    {
+      "epoch": 0.48385374966639977,
+      "grad_norm": 0.9916678812963029,
+      "learning_rate": 1.1007452748454206e-05,
+      "loss": 1.6389,
+      "step": 3626
+    },
+    {
+      "epoch": 0.48398718975180144,
+      "grad_norm": 1.1877794657449894,
+      "learning_rate": 1.1003152738384762e-05,
+      "loss": 1.6355,
+      "step": 3627
+    },
+    {
+      "epoch": 0.4841206298372031,
+      "grad_norm": 1.1020608920211157,
+      "learning_rate": 1.0998852540937806e-05,
+      "loss": 1.6111,
+      "step": 3628
+    },
+    {
+      "epoch": 0.4842540699226047,
+      "grad_norm": 0.9894785991525039,
+      "learning_rate": 1.0994552156916569e-05,
+      "loss": 1.5638,
+      "step": 3629
+    },
+    {
+      "epoch": 0.4843875100080064,
+      "grad_norm": 1.3735777211891464,
+      "learning_rate": 1.0990251587124313e-05,
+      "loss": 1.5996,
+      "step": 3630
+    },
+    {
+      "epoch": 0.48452095009340806,
+      "grad_norm": 1.2295882874045658,
+      "learning_rate": 1.0985950832364333e-05,
+      "loss": 1.6332,
+      "step": 3631
+    },
+    {
+      "epoch": 0.48465439017880974,
+      "grad_norm": 1.0441878854812512,
+      "learning_rate": 1.0981649893439965e-05,
+      "loss": 1.6525,
+      "step": 3632
+    },
+    {
+      "epoch": 0.48478783026421135,
+      "grad_norm": 1.0469977037741003,
+      "learning_rate": 1.0977348771154572e-05,
+      "loss": 1.5817,
+      "step": 3633
+    },
+    {
+      "epoch": 0.484921270349613,
+      "grad_norm": 1.0948912787896075,
+      "learning_rate": 1.0973047466311556e-05,
+      "loss": 1.5934,
+      "step": 3634
+    },
+    {
+      "epoch": 0.4850547104350147,
+      "grad_norm": 1.2238531316763264,
+      "learning_rate": 1.0968745979714355e-05,
+      "loss": 1.6424,
+      "step": 3635
+    },
+    {
+      "epoch": 0.4851881505204163,
+      "grad_norm": 0.9811824398502238,
+      "learning_rate": 1.0964444312166432e-05,
+      "loss": 1.6459,
+      "step": 3636
+    },
+    {
+      "epoch": 0.485321590605818,
+      "grad_norm": 1.011203803623103,
+      "learning_rate": 1.0960142464471293e-05,
+      "loss": 1.5845,
+      "step": 3637
+    },
+    {
+      "epoch": 0.48545503069121965,
+      "grad_norm": 1.1695581236457981,
+      "learning_rate": 1.0955840437432472e-05,
+      "loss": 1.5702,
+      "step": 3638
+    },
+    {
+      "epoch": 0.4855884707766213,
+      "grad_norm": 1.0248849738040775,
+      "learning_rate": 1.095153823185354e-05,
+      "loss": 1.5914,
+      "step": 3639
+    },
+    {
+      "epoch": 0.48572191086202293,
+      "grad_norm": 1.028338773698459,
+      "learning_rate": 1.0947235848538103e-05,
+      "loss": 1.6658,
+      "step": 3640
+    },
+    {
+      "epoch": 0.4858553509474246,
+      "grad_norm": 0.9565872039882299,
+      "learning_rate": 1.094293328828979e-05,
+      "loss": 1.6082,
+      "step": 3641
+    },
+    {
+      "epoch": 0.4859887910328263,
+      "grad_norm": 0.9438603164321578,
+      "learning_rate": 1.0938630551912275e-05,
+      "loss": 1.6025,
+      "step": 3642
+    },
+    {
+      "epoch": 0.48612223111822794,
+      "grad_norm": 1.1670924197530739,
+      "learning_rate": 1.0934327640209264e-05,
+      "loss": 1.5918,
+      "step": 3643
+    },
+    {
+      "epoch": 0.48625567120362956,
+      "grad_norm": 0.990930041258616,
+      "learning_rate": 1.0930024553984482e-05,
+      "loss": 1.5648,
+      "step": 3644
+    },
+    {
+      "epoch": 0.48638911128903123,
+      "grad_norm": 0.9674206698473387,
+      "learning_rate": 1.0925721294041704e-05,
+      "loss": 1.5957,
+      "step": 3645
+    },
+    {
+      "epoch": 0.4865225513744329,
+      "grad_norm": 1.0275864571669882,
+      "learning_rate": 1.0921417861184728e-05,
+      "loss": 1.6259,
+      "step": 3646
+    },
+    {
+      "epoch": 0.4866559914598345,
+      "grad_norm": 1.0030357183260572,
+      "learning_rate": 1.091711425621739e-05,
+      "loss": 1.6057,
+      "step": 3647
+    },
+    {
+      "epoch": 0.4867894315452362,
+      "grad_norm": 1.1702999599033863,
+      "learning_rate": 1.0912810479943546e-05,
+      "loss": 1.6031,
+      "step": 3648
+    },
+    {
+      "epoch": 0.48692287163063785,
+      "grad_norm": 1.09322465585248,
+      "learning_rate": 1.0908506533167096e-05,
+      "loss": 1.5765,
+      "step": 3649
+    },
+    {
+      "epoch": 0.4870563117160395,
+      "grad_norm": 4.935499852254243,
+      "learning_rate": 1.0904202416691973e-05,
+      "loss": 1.6665,
+      "step": 3650
+    },
+    {
+      "epoch": 0.48718975180144114,
+      "grad_norm": 1.0758505123866504,
+      "learning_rate": 1.0899898131322131e-05,
+      "loss": 1.5891,
+      "step": 3651
+    },
+    {
+      "epoch": 0.4873231918868428,
+      "grad_norm": 0.9638129734070673,
+      "learning_rate": 1.0895593677861564e-05,
+      "loss": 1.6171,
+      "step": 3652
+    },
+    {
+      "epoch": 0.4874566319722445,
+      "grad_norm": 0.9472603260400863,
+      "learning_rate": 1.0891289057114297e-05,
+      "loss": 1.5797,
+      "step": 3653
+    },
+    {
+      "epoch": 0.4875900720576461,
+      "grad_norm": 1.266655084226231,
+      "learning_rate": 1.088698426988438e-05,
+      "loss": 1.5955,
+      "step": 3654
+    },
+    {
+      "epoch": 0.48772351214304777,
+      "grad_norm": 1.0120153785951085,
+      "learning_rate": 1.08826793169759e-05,
+      "loss": 1.5924,
+      "step": 3655
+    },
+    {
+      "epoch": 0.48785695222844944,
+      "grad_norm": 1.2704389458734962,
+      "learning_rate": 1.0878374199192974e-05,
+      "loss": 1.5805,
+      "step": 3656
+    },
+    {
+      "epoch": 0.4879903923138511,
+      "grad_norm": 1.0968646982460992,
+      "learning_rate": 1.0874068917339749e-05,
+      "loss": 1.5939,
+      "step": 3657
+    },
+    {
+      "epoch": 0.4881238323992527,
+      "grad_norm": 0.9666879260201952,
+      "learning_rate": 1.08697634722204e-05,
+      "loss": 1.5857,
+      "step": 3658
+    },
+    {
+      "epoch": 0.4882572724846544,
+      "grad_norm": 1.0192846733338263,
+      "learning_rate": 1.0865457864639139e-05,
+      "loss": 1.5674,
+      "step": 3659
+    },
+    {
+      "epoch": 0.48839071257005606,
+      "grad_norm": 0.9524173836018979,
+      "learning_rate": 1.08611520954002e-05,
+      "loss": 1.5765,
+      "step": 3660
+    },
+    {
+      "epoch": 0.4885241526554577,
+      "grad_norm": 1.0159163718461748,
+      "learning_rate": 1.0856846165307858e-05,
+      "loss": 1.5984,
+      "step": 3661
+    },
+    {
+      "epoch": 0.48865759274085935,
+      "grad_norm": 1.0174232310565294,
+      "learning_rate": 1.0852540075166404e-05,
+      "loss": 1.6555,
+      "step": 3662
+    },
+    {
+      "epoch": 0.488791032826261,
+      "grad_norm": 1.1547851642140878,
+      "learning_rate": 1.0848233825780171e-05,
+      "loss": 1.6071,
+      "step": 3663
+    },
+    {
+      "epoch": 0.4889244729116627,
+      "grad_norm": 1.0165575340976178,
+      "learning_rate": 1.0843927417953517e-05,
+      "loss": 1.6243,
+      "step": 3664
+    },
+    {
+      "epoch": 0.4890579129970643,
+      "grad_norm": 1.0034096707629767,
+      "learning_rate": 1.0839620852490831e-05,
+      "loss": 1.5922,
+      "step": 3665
+    },
+    {
+      "epoch": 0.489191353082466,
+      "grad_norm": 0.9459155508614493,
+      "learning_rate": 1.083531413019653e-05,
+      "loss": 1.595,
+      "step": 3666
+    },
+    {
+      "epoch": 0.48932479316786764,
+      "grad_norm": 0.9796138002566875,
+      "learning_rate": 1.0831007251875056e-05,
+      "loss": 1.5732,
+      "step": 3667
+    },
+    {
+      "epoch": 0.48945823325326926,
+      "grad_norm": 0.9370636379889555,
+      "learning_rate": 1.0826700218330895e-05,
+      "loss": 1.6168,
+      "step": 3668
+    },
+    {
+      "epoch": 0.48959167333867093,
+      "grad_norm": 1.032334040488697,
+      "learning_rate": 1.082239303036854e-05,
+      "loss": 1.5999,
+      "step": 3669
+    },
+    {
+      "epoch": 0.4897251134240726,
+      "grad_norm": 1.2833243357988338,
+      "learning_rate": 1.0818085688792532e-05,
+      "loss": 1.5513,
+      "step": 3670
+    },
+    {
+      "epoch": 0.48985855350947427,
+      "grad_norm": 0.9893354575062938,
+      "learning_rate": 1.0813778194407432e-05,
+      "loss": 1.6065,
+      "step": 3671
+    },
+    {
+      "epoch": 0.4899919935948759,
+      "grad_norm": 0.9731145363428029,
+      "learning_rate": 1.0809470548017828e-05,
+      "loss": 1.6603,
+      "step": 3672
+    },
+    {
+      "epoch": 0.49012543368027756,
+      "grad_norm": 0.966161570856358,
+      "learning_rate": 1.0805162750428345e-05,
+      "loss": 1.5408,
+      "step": 3673
+    },
+    {
+      "epoch": 0.4902588737656792,
+      "grad_norm": 1.1087789778186123,
+      "learning_rate": 1.0800854802443626e-05,
+      "loss": 1.5869,
+      "step": 3674
+    },
+    {
+      "epoch": 0.49039231385108084,
+      "grad_norm": 1.012321782174784,
+      "learning_rate": 1.0796546704868348e-05,
+      "loss": 1.6429,
+      "step": 3675
+    },
+    {
+      "epoch": 0.4905257539364825,
+      "grad_norm": 0.9858444455189594,
+      "learning_rate": 1.0792238458507215e-05,
+      "loss": 1.5603,
+      "step": 3676
+    },
+    {
+      "epoch": 0.4906591940218842,
+      "grad_norm": 1.130805582793479,
+      "learning_rate": 1.0787930064164959e-05,
+      "loss": 1.6041,
+      "step": 3677
+    },
+    {
+      "epoch": 0.49079263410728585,
+      "grad_norm": 0.9578865164171134,
+      "learning_rate": 1.0783621522646336e-05,
+      "loss": 1.5791,
+      "step": 3678
+    },
+    {
+      "epoch": 0.49092607419268747,
+      "grad_norm": 0.9913348610814515,
+      "learning_rate": 1.0779312834756134e-05,
+      "loss": 1.6555,
+      "step": 3679
+    },
+    {
+      "epoch": 0.49105951427808914,
+      "grad_norm": 0.9482971209808349,
+      "learning_rate": 1.0775004001299173e-05,
+      "loss": 1.5926,
+      "step": 3680
+    },
+    {
+      "epoch": 0.4911929543634908,
+      "grad_norm": 0.9705291068336154,
+      "learning_rate": 1.0770695023080282e-05,
+      "loss": 1.6281,
+      "step": 3681
+    },
+    {
+      "epoch": 0.4913263944488924,
+      "grad_norm": 1.1282798108079959,
+      "learning_rate": 1.0766385900904337e-05,
+      "loss": 1.5623,
+      "step": 3682
+    },
+    {
+      "epoch": 0.4914598345342941,
+      "grad_norm": 0.994148436807919,
+      "learning_rate": 1.0762076635576231e-05,
+      "loss": 1.6486,
+      "step": 3683
+    },
+    {
+      "epoch": 0.49159327461969576,
+      "grad_norm": 0.9583103245372037,
+      "learning_rate": 1.0757767227900888e-05,
+      "loss": 1.5759,
+      "step": 3684
+    },
+    {
+      "epoch": 0.49172671470509743,
+      "grad_norm": 0.9434367640143331,
+      "learning_rate": 1.075345767868325e-05,
+      "loss": 1.6,
+      "step": 3685
+    },
+    {
+      "epoch": 0.49186015479049905,
+      "grad_norm": 1.054046209894045,
+      "learning_rate": 1.0749147988728302e-05,
+      "loss": 1.6308,
+      "step": 3686
+    },
+    {
+      "epoch": 0.4919935948759007,
+      "grad_norm": 0.9921895864830362,
+      "learning_rate": 1.0744838158841034e-05,
+      "loss": 1.6343,
+      "step": 3687
+    },
+    {
+      "epoch": 0.4921270349613024,
+      "grad_norm": 1.2122795180169637,
+      "learning_rate": 1.074052818982648e-05,
+      "loss": 1.5849,
+      "step": 3688
+    },
+    {
+      "epoch": 0.492260475046704,
+      "grad_norm": 0.9788260702681275,
+      "learning_rate": 1.0736218082489691e-05,
+      "loss": 1.6075,
+      "step": 3689
+    },
+    {
+      "epoch": 0.4923939151321057,
+      "grad_norm": 1.3297573232605928,
+      "learning_rate": 1.0731907837635747e-05,
+      "loss": 1.6099,
+      "step": 3690
+    },
+    {
+      "epoch": 0.49252735521750735,
+      "grad_norm": 0.9550880821538821,
+      "learning_rate": 1.0727597456069755e-05,
+      "loss": 1.6125,
+      "step": 3691
+    },
+    {
+      "epoch": 0.492660795302909,
+      "grad_norm": 0.9619871054912161,
+      "learning_rate": 1.0723286938596836e-05,
+      "loss": 1.6079,
+      "step": 3692
+    },
+    {
+      "epoch": 0.49279423538831063,
+      "grad_norm": 1.1849634487362217,
+      "learning_rate": 1.0718976286022157e-05,
+      "loss": 1.6274,
+      "step": 3693
+    },
+    {
+      "epoch": 0.4929276754737123,
+      "grad_norm": 1.2585836751661978,
+      "learning_rate": 1.0714665499150888e-05,
+      "loss": 1.5792,
+      "step": 3694
+    },
+    {
+      "epoch": 0.49306111555911397,
+      "grad_norm": 0.9682726552538469,
+      "learning_rate": 1.0710354578788247e-05,
+      "loss": 1.5719,
+      "step": 3695
+    },
+    {
+      "epoch": 0.4931945556445156,
+      "grad_norm": 1.0111852850310221,
+      "learning_rate": 1.0706043525739454e-05,
+      "loss": 1.5881,
+      "step": 3696
+    },
+    {
+      "epoch": 0.49332799572991726,
+      "grad_norm": 1.0158485435950086,
+      "learning_rate": 1.070173234080977e-05,
+      "loss": 1.557,
+      "step": 3697
+    },
+    {
+      "epoch": 0.4934614358153189,
+      "grad_norm": 0.9601365629223316,
+      "learning_rate": 1.0697421024804475e-05,
+      "loss": 1.5992,
+      "step": 3698
+    },
+    {
+      "epoch": 0.4935948759007206,
+      "grad_norm": 0.9711819226222418,
+      "learning_rate": 1.0693109578528875e-05,
+      "loss": 1.5762,
+      "step": 3699
+    },
+    {
+      "epoch": 0.4937283159861222,
+      "grad_norm": 0.9747571113243313,
+      "learning_rate": 1.0688798002788295e-05,
+      "loss": 1.627,
+      "step": 3700
+    },
+    {
+      "epoch": 0.4938617560715239,
+      "grad_norm": 0.9155209925749146,
+      "learning_rate": 1.068448629838809e-05,
+      "loss": 1.5424,
+      "step": 3701
+    },
+    {
+      "epoch": 0.49399519615692555,
+      "grad_norm": 0.9680580273820439,
+      "learning_rate": 1.0680174466133639e-05,
+      "loss": 1.5782,
+      "step": 3702
+    },
+    {
+      "epoch": 0.49412863624232717,
+      "grad_norm": 0.9685689662721528,
+      "learning_rate": 1.067586250683034e-05,
+      "loss": 1.5927,
+      "step": 3703
+    },
+    {
+      "epoch": 0.49426207632772884,
+      "grad_norm": 0.9349199396735328,
+      "learning_rate": 1.0671550421283618e-05,
+      "loss": 1.6169,
+      "step": 3704
+    },
+    {
+      "epoch": 0.4943955164131305,
+      "grad_norm": 0.9766752432473302,
+      "learning_rate": 1.0667238210298927e-05,
+      "loss": 1.6029,
+      "step": 3705
+    },
+    {
+      "epoch": 0.4945289564985322,
+      "grad_norm": 1.1337519998185137,
+      "learning_rate": 1.0662925874681733e-05,
+      "loss": 1.5993,
+      "step": 3706
+    },
+    {
+      "epoch": 0.4946623965839338,
+      "grad_norm": 0.9756943770828267,
+      "learning_rate": 1.0658613415237535e-05,
+      "loss": 1.6645,
+      "step": 3707
+    },
+    {
+      "epoch": 0.49479583666933546,
+      "grad_norm": 0.9602283668761221,
+      "learning_rate": 1.0654300832771847e-05,
+      "loss": 1.6323,
+      "step": 3708
+    },
+    {
+      "epoch": 0.49492927675473714,
+      "grad_norm": 1.2356580385421276,
+      "learning_rate": 1.0649988128090216e-05,
+      "loss": 1.5881,
+      "step": 3709
+    },
+    {
+      "epoch": 0.49506271684013875,
+      "grad_norm": 0.9804354142593109,
+      "learning_rate": 1.06456753019982e-05,
+      "loss": 1.5797,
+      "step": 3710
+    },
+    {
+      "epoch": 0.4951961569255404,
+      "grad_norm": 0.9413974484117791,
+      "learning_rate": 1.0641362355301392e-05,
+      "loss": 1.5959,
+      "step": 3711
+    },
+    {
+      "epoch": 0.4953295970109421,
+      "grad_norm": 1.0173430778833028,
+      "learning_rate": 1.0637049288805395e-05,
+      "loss": 1.6835,
+      "step": 3712
+    },
+    {
+      "epoch": 0.49546303709634376,
+      "grad_norm": 1.0230987750185572,
+      "learning_rate": 1.0632736103315843e-05,
+      "loss": 1.5686,
+      "step": 3713
+    },
+    {
+      "epoch": 0.4955964771817454,
+      "grad_norm": 0.9776487226730937,
+      "learning_rate": 1.0628422799638396e-05,
+      "loss": 1.5768,
+      "step": 3714
+    },
+    {
+      "epoch": 0.49572991726714705,
+      "grad_norm": 1.1209401946738562,
+      "learning_rate": 1.0624109378578721e-05,
+      "loss": 1.5674,
+      "step": 3715
+    },
+    {
+      "epoch": 0.4958633573525487,
+      "grad_norm": 1.1825505150445272,
+      "learning_rate": 1.0619795840942524e-05,
+      "loss": 1.5712,
+      "step": 3716
+    },
+    {
+      "epoch": 0.4959967974379504,
+      "grad_norm": 0.939437949764124,
+      "learning_rate": 1.0615482187535515e-05,
+      "loss": 1.5544,
+      "step": 3717
+    },
+    {
+      "epoch": 0.496130237523352,
+      "grad_norm": 1.0394749330983024,
+      "learning_rate": 1.0611168419163444e-05,
+      "loss": 1.6918,
+      "step": 3718
+    },
+    {
+      "epoch": 0.4962636776087537,
+      "grad_norm": 1.0963483530806055,
+      "learning_rate": 1.060685453663207e-05,
+      "loss": 1.5935,
+      "step": 3719
+    },
+    {
+      "epoch": 0.49639711769415534,
+      "grad_norm": 0.9214028923141944,
+      "learning_rate": 1.0602540540747179e-05,
+      "loss": 1.605,
+      "step": 3720
+    },
+    {
+      "epoch": 0.49653055777955696,
+      "grad_norm": 0.9811475135324195,
+      "learning_rate": 1.0598226432314573e-05,
+      "loss": 1.5914,
+      "step": 3721
+    },
+    {
+      "epoch": 0.49666399786495863,
+      "grad_norm": 0.9866310168569355,
+      "learning_rate": 1.0593912212140086e-05,
+      "loss": 1.6091,
+      "step": 3722
+    },
+    {
+      "epoch": 0.4967974379503603,
+      "grad_norm": 0.9398392471927407,
+      "learning_rate": 1.0589597881029554e-05,
+      "loss": 1.5977,
+      "step": 3723
+    },
+    {
+      "epoch": 0.49693087803576197,
+      "grad_norm": 0.9946024082527848,
+      "learning_rate": 1.0585283439788851e-05,
+      "loss": 1.532,
+      "step": 3724
+    },
+    {
+      "epoch": 0.4970643181211636,
+      "grad_norm": 0.9747426365894848,
+      "learning_rate": 1.0580968889223868e-05,
+      "loss": 1.5502,
+      "step": 3725
+    },
+    {
+      "epoch": 0.49719775820656525,
+      "grad_norm": 1.0170841216880055,
+      "learning_rate": 1.0576654230140508e-05,
+      "loss": 1.5805,
+      "step": 3726
+    },
+    {
+      "epoch": 0.4973311982919669,
+      "grad_norm": 0.9642570769540154,
+      "learning_rate": 1.0572339463344707e-05,
+      "loss": 1.5801,
+      "step": 3727
+    },
+    {
+      "epoch": 0.49746463837736854,
+      "grad_norm": 0.9628913796090154,
+      "learning_rate": 1.0568024589642408e-05,
+      "loss": 1.605,
+      "step": 3728
+    },
+    {
+      "epoch": 0.4975980784627702,
+      "grad_norm": 0.9946599596096027,
+      "learning_rate": 1.0563709609839581e-05,
+      "loss": 1.5589,
+      "step": 3729
+    },
+    {
+      "epoch": 0.4977315185481719,
+      "grad_norm": 1.1095821154904126,
+      "learning_rate": 1.055939452474222e-05,
+      "loss": 1.5844,
+      "step": 3730
+    },
+    {
+      "epoch": 0.49786495863357355,
+      "grad_norm": 1.1727773862931867,
+      "learning_rate": 1.0555079335156328e-05,
+      "loss": 1.5687,
+      "step": 3731
+    },
+    {
+      "epoch": 0.49799839871897517,
+      "grad_norm": 1.0061224792693193,
+      "learning_rate": 1.055076404188794e-05,
+      "loss": 1.5969,
+      "step": 3732
+    },
+    {
+      "epoch": 0.49813183880437684,
+      "grad_norm": 0.9895293534993739,
+      "learning_rate": 1.0546448645743097e-05,
+      "loss": 1.6145,
+      "step": 3733
+    },
+    {
+      "epoch": 0.4982652788897785,
+      "grad_norm": 1.0574963661665566,
+      "learning_rate": 1.054213314752787e-05,
+      "loss": 1.6219,
+      "step": 3734
+    },
+    {
+      "epoch": 0.4983987189751801,
+      "grad_norm": 1.0560776793903364,
+      "learning_rate": 1.0537817548048341e-05,
+      "loss": 1.577,
+      "step": 3735
+    },
+    {
+      "epoch": 0.4985321590605818,
+      "grad_norm": 1.0083197939133182,
+      "learning_rate": 1.0533501848110617e-05,
+      "loss": 1.6201,
+      "step": 3736
+    },
+    {
+      "epoch": 0.49866559914598346,
+      "grad_norm": 0.9934023783176211,
+      "learning_rate": 1.0529186048520825e-05,
+      "loss": 1.5933,
+      "step": 3737
+    },
+    {
+      "epoch": 0.49879903923138513,
+      "grad_norm": 1.243448461460969,
+      "learning_rate": 1.0524870150085103e-05,
+      "loss": 1.6314,
+      "step": 3738
+    },
+    {
+      "epoch": 0.49893247931678675,
+      "grad_norm": 1.1745919731057533,
+      "learning_rate": 1.0520554153609613e-05,
+      "loss": 1.6081,
+      "step": 3739
+    },
+    {
+      "epoch": 0.4990659194021884,
+      "grad_norm": 0.9690359243672029,
+      "learning_rate": 1.0516238059900532e-05,
+      "loss": 1.6033,
+      "step": 3740
+    },
+    {
+      "epoch": 0.4991993594875901,
+      "grad_norm": 0.9910492376976103,
+      "learning_rate": 1.0511921869764062e-05,
+      "loss": 1.6147,
+      "step": 3741
+    },
+    {
+      "epoch": 0.4993327995729917,
+      "grad_norm": 0.9841002474852324,
+      "learning_rate": 1.0507605584006413e-05,
+      "loss": 1.5855,
+      "step": 3742
+    },
+    {
+      "epoch": 0.4994662396583934,
+      "grad_norm": 0.9835960197152689,
+      "learning_rate": 1.0503289203433822e-05,
+      "loss": 1.5912,
+      "step": 3743
+    },
+    {
+      "epoch": 0.49959967974379504,
+      "grad_norm": 0.9331205335074193,
+      "learning_rate": 1.049897272885254e-05,
+      "loss": 1.5957,
+      "step": 3744
+    },
+    {
+      "epoch": 0.4997331198291967,
+      "grad_norm": 1.0670315368640506,
+      "learning_rate": 1.0494656161068828e-05,
+      "loss": 1.5724,
+      "step": 3745
+    },
+    {
+      "epoch": 0.49986655991459833,
+      "grad_norm": 0.9964186881122445,
+      "learning_rate": 1.049033950088898e-05,
+      "loss": 1.6113,
+      "step": 3746
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.9622713227335437,
+      "learning_rate": 1.0486022749119294e-05,
+      "loss": 1.6103,
+      "step": 3747
+    },
+    {
+      "epoch": 0.5001334400854016,
+      "grad_norm": 1.5585035104556884,
+      "learning_rate": 1.0481705906566092e-05,
+      "loss": 1.5407,
+      "step": 3748
+    },
+    {
+      "epoch": 0.5002668801708033,
+      "grad_norm": 0.9525397308234141,
+      "learning_rate": 1.0477388974035713e-05,
+      "loss": 1.56,
+      "step": 3749
+    },
+    {
+      "epoch": 0.500400320256205,
+      "grad_norm": 1.2183987988540275,
+      "learning_rate": 1.0473071952334508e-05,
+      "loss": 1.5559,
+      "step": 3750
+    },
+    {
+      "epoch": 0.5005337603416066,
+      "grad_norm": 1.0009923680441615,
+      "learning_rate": 1.0468754842268849e-05,
+      "loss": 1.6096,
+      "step": 3751
+    },
+    {
+      "epoch": 0.5006672004270083,
+      "grad_norm": 0.9445861127024189,
+      "learning_rate": 1.046443764464512e-05,
+      "loss": 1.6056,
+      "step": 3752
+    },
+    {
+      "epoch": 0.5008006405124099,
+      "grad_norm": 0.9742127353900115,
+      "learning_rate": 1.046012036026973e-05,
+      "loss": 1.6246,
+      "step": 3753
+    },
+    {
+      "epoch": 0.5009340805978116,
+      "grad_norm": 0.9520930850005337,
+      "learning_rate": 1.0455802989949092e-05,
+      "loss": 1.5292,
+      "step": 3754
+    },
+    {
+      "epoch": 0.5010675206832133,
+      "grad_norm": 0.935444191000388,
+      "learning_rate": 1.0451485534489649e-05,
+      "loss": 1.5537,
+      "step": 3755
+    },
+    {
+      "epoch": 0.5012009607686149,
+      "grad_norm": 0.9491673736063649,
+      "learning_rate": 1.0447167994697846e-05,
+      "loss": 1.5746,
+      "step": 3756
+    },
+    {
+      "epoch": 0.5013344008540166,
+      "grad_norm": 0.9554087936976446,
+      "learning_rate": 1.0442850371380155e-05,
+      "loss": 1.6095,
+      "step": 3757
+    },
+    {
+      "epoch": 0.5014678409394182,
+      "grad_norm": 1.2103781704529109,
+      "learning_rate": 1.0438532665343053e-05,
+      "loss": 1.5771,
+      "step": 3758
+    },
+    {
+      "epoch": 0.5016012810248198,
+      "grad_norm": 1.2128718749710947,
+      "learning_rate": 1.0434214877393045e-05,
+      "loss": 1.5952,
+      "step": 3759
+    },
+    {
+      "epoch": 0.5017347211102215,
+      "grad_norm": 0.9603242533460603,
+      "learning_rate": 1.042989700833664e-05,
+      "loss": 1.5571,
+      "step": 3760
+    },
+    {
+      "epoch": 0.5018681611956232,
+      "grad_norm": 1.1013519805324967,
+      "learning_rate": 1.042557905898037e-05,
+      "loss": 1.5555,
+      "step": 3761
+    },
+    {
+      "epoch": 0.5020016012810248,
+      "grad_norm": 1.113312267406771,
+      "learning_rate": 1.0421261030130776e-05,
+      "loss": 1.5986,
+      "step": 3762
+    },
+    {
+      "epoch": 0.5021350413664265,
+      "grad_norm": 0.9630320933350655,
+      "learning_rate": 1.041694292259442e-05,
+      "loss": 1.5864,
+      "step": 3763
+    },
+    {
+      "epoch": 0.5022684814518281,
+      "grad_norm": 0.9790488821683675,
+      "learning_rate": 1.041262473717787e-05,
+      "loss": 1.6064,
+      "step": 3764
+    },
+    {
+      "epoch": 0.5024019215372297,
+      "grad_norm": 1.0221627458345286,
+      "learning_rate": 1.0408306474687719e-05,
+      "loss": 1.6228,
+      "step": 3765
+    },
+    {
+      "epoch": 0.5025353616226315,
+      "grad_norm": 0.9477242589703042,
+      "learning_rate": 1.0403988135930568e-05,
+      "loss": 1.6004,
+      "step": 3766
+    },
+    {
+      "epoch": 0.5026688017080331,
+      "grad_norm": 0.9934183349840355,
+      "learning_rate": 1.039966972171303e-05,
+      "loss": 1.554,
+      "step": 3767
+    },
+    {
+      "epoch": 0.5028022417934348,
+      "grad_norm": 0.9582444414220985,
+      "learning_rate": 1.0395351232841739e-05,
+      "loss": 1.595,
+      "step": 3768
+    },
+    {
+      "epoch": 0.5029356818788364,
+      "grad_norm": 1.0891630480558472,
+      "learning_rate": 1.039103267012334e-05,
+      "loss": 1.555,
+      "step": 3769
+    },
+    {
+      "epoch": 0.503069121964238,
+      "grad_norm": 0.9957005065240585,
+      "learning_rate": 1.038671403436449e-05,
+      "loss": 1.6035,
+      "step": 3770
+    },
+    {
+      "epoch": 0.5032025620496398,
+      "grad_norm": 1.2830993460776894,
+      "learning_rate": 1.0382395326371861e-05,
+      "loss": 1.6037,
+      "step": 3771
+    },
+    {
+      "epoch": 0.5033360021350414,
+      "grad_norm": 0.9767721406060091,
+      "learning_rate": 1.0378076546952138e-05,
+      "loss": 1.5673,
+      "step": 3772
+    },
+    {
+      "epoch": 0.503469442220443,
+      "grad_norm": 0.9734968119255416,
+      "learning_rate": 1.0373757696912024e-05,
+      "loss": 1.5924,
+      "step": 3773
+    },
+    {
+      "epoch": 0.5036028823058447,
+      "grad_norm": 1.1161953113263952,
+      "learning_rate": 1.0369438777058226e-05,
+      "loss": 1.6249,
+      "step": 3774
+    },
+    {
+      "epoch": 0.5037363223912463,
+      "grad_norm": 0.9572612259574967,
+      "learning_rate": 1.0365119788197468e-05,
+      "loss": 1.5998,
+      "step": 3775
+    },
+    {
+      "epoch": 0.5038697624766479,
+      "grad_norm": 0.9432193177518614,
+      "learning_rate": 1.0360800731136493e-05,
+      "loss": 1.5321,
+      "step": 3776
+    },
+    {
+      "epoch": 0.5040032025620497,
+      "grad_norm": 1.0074347003449116,
+      "learning_rate": 1.0356481606682047e-05,
+      "loss": 1.6185,
+      "step": 3777
+    },
+    {
+      "epoch": 0.5041366426474513,
+      "grad_norm": 0.9454406764250083,
+      "learning_rate": 1.0352162415640898e-05,
+      "loss": 1.5624,
+      "step": 3778
+    },
+    {
+      "epoch": 0.5042700827328529,
+      "grad_norm": 1.004743047178849,
+      "learning_rate": 1.034784315881982e-05,
+      "loss": 1.6052,
+      "step": 3779
+    },
+    {
+      "epoch": 0.5044035228182546,
+      "grad_norm": 1.0081054951554982,
+      "learning_rate": 1.0343523837025598e-05,
+      "loss": 1.5967,
+      "step": 3780
+    },
+    {
+      "epoch": 0.5045369629036562,
+      "grad_norm": 1.1995648160845978,
+      "learning_rate": 1.0339204451065035e-05,
+      "loss": 1.5653,
+      "step": 3781
+    },
+    {
+      "epoch": 0.504670402989058,
+      "grad_norm": 1.0397558826608888,
+      "learning_rate": 1.0334885001744943e-05,
+      "loss": 1.6437,
+      "step": 3782
+    },
+    {
+      "epoch": 0.5048038430744596,
+      "grad_norm": 0.9903660282060232,
+      "learning_rate": 1.0330565489872144e-05,
+      "loss": 1.5705,
+      "step": 3783
+    },
+    {
+      "epoch": 0.5049372831598612,
+      "grad_norm": 0.9972288626686299,
+      "learning_rate": 1.0326245916253478e-05,
+      "loss": 1.5884,
+      "step": 3784
+    },
+    {
+      "epoch": 0.5050707232452629,
+      "grad_norm": 0.9476343942885561,
+      "learning_rate": 1.0321926281695787e-05,
+      "loss": 1.6023,
+      "step": 3785
+    },
+    {
+      "epoch": 0.5052041633306645,
+      "grad_norm": 0.9738449763592262,
+      "learning_rate": 1.0317606587005936e-05,
+      "loss": 1.5609,
+      "step": 3786
+    },
+    {
+      "epoch": 0.5053376034160662,
+      "grad_norm": 1.039829392353301,
+      "learning_rate": 1.0313286832990788e-05,
+      "loss": 1.6218,
+      "step": 3787
+    },
+    {
+      "epoch": 0.5054710435014679,
+      "grad_norm": 1.0381227642185753,
+      "learning_rate": 1.0308967020457223e-05,
+      "loss": 1.6251,
+      "step": 3788
+    },
+    {
+      "epoch": 0.5056044835868695,
+      "grad_norm": 0.9867023264523529,
+      "learning_rate": 1.0304647150212142e-05,
+      "loss": 1.5399,
+      "step": 3789
+    },
+    {
+      "epoch": 0.5057379236722711,
+      "grad_norm": 0.9867055454700514,
+      "learning_rate": 1.0300327223062436e-05,
+      "loss": 1.6344,
+      "step": 3790
+    },
+    {
+      "epoch": 0.5058713637576728,
+      "grad_norm": 1.0315163736938113,
+      "learning_rate": 1.029600723981503e-05,
+      "loss": 1.6119,
+      "step": 3791
+    },
+    {
+      "epoch": 0.5060048038430744,
+      "grad_norm": 0.9670113161953792,
+      "learning_rate": 1.0291687201276837e-05,
+      "loss": 1.5639,
+      "step": 3792
+    },
+    {
+      "epoch": 0.5061382439284761,
+      "grad_norm": 0.9921317429814296,
+      "learning_rate": 1.0287367108254796e-05,
+      "loss": 1.5514,
+      "step": 3793
+    },
+    {
+      "epoch": 0.5062716840138778,
+      "grad_norm": 1.0082597213286801,
+      "learning_rate": 1.0283046961555854e-05,
+      "loss": 1.5689,
+      "step": 3794
+    },
+    {
+      "epoch": 0.5064051240992794,
+      "grad_norm": 1.394709462594669,
+      "learning_rate": 1.027872676198696e-05,
+      "loss": 1.6046,
+      "step": 3795
+    },
+    {
+      "epoch": 0.5065385641846811,
+      "grad_norm": 1.0174333180253856,
+      "learning_rate": 1.0274406510355082e-05,
+      "loss": 1.5815,
+      "step": 3796
+    },
+    {
+      "epoch": 0.5066720042700827,
+      "grad_norm": 1.0733270672727413,
+      "learning_rate": 1.0270086207467188e-05,
+      "loss": 1.5879,
+      "step": 3797
+    },
+    {
+      "epoch": 0.5068054443554844,
+      "grad_norm": 1.0234017065005216,
+      "learning_rate": 1.0265765854130272e-05,
+      "loss": 1.6075,
+      "step": 3798
+    },
+    {
+      "epoch": 0.5069388844408861,
+      "grad_norm": 1.08499646847553,
+      "learning_rate": 1.0261445451151314e-05,
+      "loss": 1.6305,
+      "step": 3799
+    },
+    {
+      "epoch": 0.5070723245262877,
+      "grad_norm": 0.9704549685581058,
+      "learning_rate": 1.0257124999337324e-05,
+      "loss": 1.5948,
+      "step": 3800
+    },
+    {
+      "epoch": 0.5072057646116893,
+      "grad_norm": 7.9206996649868,
+      "learning_rate": 1.0252804499495314e-05,
+      "loss": 1.6509,
+      "step": 3801
+    },
+    {
+      "epoch": 0.507339204697091,
+      "grad_norm": 1.0387910709315784,
+      "learning_rate": 1.02484839524323e-05,
+      "loss": 1.5677,
+      "step": 3802
+    },
+    {
+      "epoch": 0.5074726447824927,
+      "grad_norm": 1.050554110371522,
+      "learning_rate": 1.0244163358955315e-05,
+      "loss": 1.6384,
+      "step": 3803
+    },
+    {
+      "epoch": 0.5076060848678943,
+      "grad_norm": 1.1559737239306573,
+      "learning_rate": 1.023984271987139e-05,
+      "loss": 1.6467,
+      "step": 3804
+    },
+    {
+      "epoch": 0.507739524953296,
+      "grad_norm": 1.0907297235523432,
+      "learning_rate": 1.0235522035987581e-05,
+      "loss": 1.632,
+      "step": 3805
+    },
+    {
+      "epoch": 0.5078729650386976,
+      "grad_norm": 1.0831891794610449,
+      "learning_rate": 1.0231201308110936e-05,
+      "loss": 1.6218,
+      "step": 3806
+    },
+    {
+      "epoch": 0.5080064051240992,
+      "grad_norm": 0.9955232658741183,
+      "learning_rate": 1.0226880537048518e-05,
+      "loss": 1.6157,
+      "step": 3807
+    },
+    {
+      "epoch": 0.508139845209501,
+      "grad_norm": 1.0436607505823476,
+      "learning_rate": 1.02225597236074e-05,
+      "loss": 1.5852,
+      "step": 3808
+    },
+    {
+      "epoch": 0.5082732852949026,
+      "grad_norm": 1.0380692870921138,
+      "learning_rate": 1.0218238868594656e-05,
+      "loss": 1.4999,
+      "step": 3809
+    },
+    {
+      "epoch": 0.5084067253803043,
+      "grad_norm": 0.95584778107239,
+      "learning_rate": 1.021391797281738e-05,
+      "loss": 1.5926,
+      "step": 3810
+    },
+    {
+      "epoch": 0.5085401654657059,
+      "grad_norm": 1.13056350804676,
+      "learning_rate": 1.0209597037082658e-05,
+      "loss": 1.599,
+      "step": 3811
+    },
+    {
+      "epoch": 0.5086736055511075,
+      "grad_norm": 1.0358679977491891,
+      "learning_rate": 1.02052760621976e-05,
+      "loss": 1.6291,
+      "step": 3812
+    },
+    {
+      "epoch": 0.5088070456365092,
+      "grad_norm": 0.9752350063012711,
+      "learning_rate": 1.0200955048969307e-05,
+      "loss": 1.5868,
+      "step": 3813
+    },
+    {
+      "epoch": 0.5089404857219109,
+      "grad_norm": 2.6417596103398284,
+      "learning_rate": 1.0196633998204903e-05,
+      "loss": 1.5943,
+      "step": 3814
+    },
+    {
+      "epoch": 0.5090739258073125,
+      "grad_norm": 1.1501400580375445,
+      "learning_rate": 1.0192312910711504e-05,
+      "loss": 1.5856,
+      "step": 3815
+    },
+    {
+      "epoch": 0.5092073658927142,
+      "grad_norm": 1.08515119526117,
+      "learning_rate": 1.0187991787296243e-05,
+      "loss": 1.6207,
+      "step": 3816
+    },
+    {
+      "epoch": 0.5093408059781158,
+      "grad_norm": 0.9540437758466517,
+      "learning_rate": 1.0183670628766258e-05,
+      "loss": 1.643,
+      "step": 3817
+    },
+    {
+      "epoch": 0.5094742460635174,
+      "grad_norm": 1.0299232805390495,
+      "learning_rate": 1.017934943592869e-05,
+      "loss": 1.6673,
+      "step": 3818
+    },
+    {
+      "epoch": 0.5096076861489192,
+      "grad_norm": 1.0039257031904898,
+      "learning_rate": 1.017502820959069e-05,
+      "loss": 1.5761,
+      "step": 3819
+    },
+    {
+      "epoch": 0.5097411262343208,
+      "grad_norm": 0.9346765769936345,
+      "learning_rate": 1.017070695055941e-05,
+      "loss": 1.5594,
+      "step": 3820
+    },
+    {
+      "epoch": 0.5098745663197225,
+      "grad_norm": 0.9858372448777835,
+      "learning_rate": 1.0166385659642017e-05,
+      "loss": 1.5746,
+      "step": 3821
+    },
+    {
+      "epoch": 0.5100080064051241,
+      "grad_norm": 1.0006528357635736,
+      "learning_rate": 1.0162064337645678e-05,
+      "loss": 1.6062,
+      "step": 3822
+    },
+    {
+      "epoch": 0.5101414464905257,
+      "grad_norm": 0.9404463154241544,
+      "learning_rate": 1.0157742985377567e-05,
+      "loss": 1.6299,
+      "step": 3823
+    },
+    {
+      "epoch": 0.5102748865759275,
+      "grad_norm": 1.0401934215904935,
+      "learning_rate": 1.015342160364486e-05,
+      "loss": 1.61,
+      "step": 3824
+    },
+    {
+      "epoch": 0.5104083266613291,
+      "grad_norm": 0.9544524367700885,
+      "learning_rate": 1.0149100193254744e-05,
+      "loss": 1.5981,
+      "step": 3825
+    },
+    {
+      "epoch": 0.5105417667467307,
+      "grad_norm": 1.1258447776031195,
+      "learning_rate": 1.0144778755014411e-05,
+      "loss": 1.5924,
+      "step": 3826
+    },
+    {
+      "epoch": 0.5106752068321324,
+      "grad_norm": 1.048900303879943,
+      "learning_rate": 1.0140457289731056e-05,
+      "loss": 1.5822,
+      "step": 3827
+    },
+    {
+      "epoch": 0.510808646917534,
+      "grad_norm": 0.9978329789351215,
+      "learning_rate": 1.0136135798211874e-05,
+      "loss": 1.5754,
+      "step": 3828
+    },
+    {
+      "epoch": 0.5109420870029356,
+      "grad_norm": 1.0026431461629155,
+      "learning_rate": 1.013181428126408e-05,
+      "loss": 1.5149,
+      "step": 3829
+    },
+    {
+      "epoch": 0.5110755270883374,
+      "grad_norm": 0.9910636307338526,
+      "learning_rate": 1.012749273969488e-05,
+      "loss": 1.6098,
+      "step": 3830
+    },
+    {
+      "epoch": 0.511208967173739,
+      "grad_norm": 0.9607665572703091,
+      "learning_rate": 1.0123171174311482e-05,
+      "loss": 1.596,
+      "step": 3831
+    },
+    {
+      "epoch": 0.5113424072591406,
+      "grad_norm": 0.9629876623056436,
+      "learning_rate": 1.0118849585921114e-05,
+      "loss": 1.5757,
+      "step": 3832
+    },
+    {
+      "epoch": 0.5114758473445423,
+      "grad_norm": 0.9574400509477552,
+      "learning_rate": 1.0114527975330997e-05,
+      "loss": 1.5883,
+      "step": 3833
+    },
+    {
+      "epoch": 0.5116092874299439,
+      "grad_norm": 1.0875584418211826,
+      "learning_rate": 1.0110206343348354e-05,
+      "loss": 1.5646,
+      "step": 3834
+    },
+    {
+      "epoch": 0.5117427275153457,
+      "grad_norm": 1.169489430147777,
+      "learning_rate": 1.0105884690780426e-05,
+      "loss": 1.5489,
+      "step": 3835
+    },
+    {
+      "epoch": 0.5118761676007473,
+      "grad_norm": 1.4633224716352873,
+      "learning_rate": 1.0101563018434441e-05,
+      "loss": 1.612,
+      "step": 3836
+    },
+    {
+      "epoch": 0.5120096076861489,
+      "grad_norm": 0.9544832295384696,
+      "learning_rate": 1.0097241327117642e-05,
+      "loss": 1.5445,
+      "step": 3837
+    },
+    {
+      "epoch": 0.5121430477715506,
+      "grad_norm": 1.0099497190113178,
+      "learning_rate": 1.0092919617637267e-05,
+      "loss": 1.6371,
+      "step": 3838
+    },
+    {
+      "epoch": 0.5122764878569522,
+      "grad_norm": 0.9869374341847159,
+      "learning_rate": 1.0088597890800568e-05,
+      "loss": 1.5956,
+      "step": 3839
+    },
+    {
+      "epoch": 0.5124099279423538,
+      "grad_norm": 0.9740701564399505,
+      "learning_rate": 1.008427614741479e-05,
+      "loss": 1.6369,
+      "step": 3840
+    },
+    {
+      "epoch": 0.5125433680277556,
+      "grad_norm": 1.0435801092284192,
+      "learning_rate": 1.0079954388287187e-05,
+      "loss": 1.6168,
+      "step": 3841
+    },
+    {
+      "epoch": 0.5126768081131572,
+      "grad_norm": 1.0717965001586232,
+      "learning_rate": 1.007563261422502e-05,
+      "loss": 1.5815,
+      "step": 3842
+    },
+    {
+      "epoch": 0.5128102481985588,
+      "grad_norm": 0.951753888240989,
+      "learning_rate": 1.0071310826035536e-05,
+      "loss": 1.5753,
+      "step": 3843
+    },
+    {
+      "epoch": 0.5129436882839605,
+      "grad_norm": 1.0427073825832405,
+      "learning_rate": 1.0066989024526004e-05,
+      "loss": 1.571,
+      "step": 3844
+    },
+    {
+      "epoch": 0.5130771283693621,
+      "grad_norm": 0.9665144556917987,
+      "learning_rate": 1.0062667210503682e-05,
+      "loss": 1.5675,
+      "step": 3845
+    },
+    {
+      "epoch": 0.5132105684547638,
+      "grad_norm": 0.9841126172568884,
+      "learning_rate": 1.0058345384775843e-05,
+      "loss": 1.566,
+      "step": 3846
+    },
+    {
+      "epoch": 0.5133440085401655,
+      "grad_norm": 1.0231266781644845,
+      "learning_rate": 1.0054023548149747e-05,
+      "loss": 1.6019,
+      "step": 3847
+    },
+    {
+      "epoch": 0.5134774486255671,
+      "grad_norm": 1.013759415570729,
+      "learning_rate": 1.004970170143267e-05,
+      "loss": 1.5977,
+      "step": 3848
+    },
+    {
+      "epoch": 0.5136108887109688,
+      "grad_norm": 0.9596994251252524,
+      "learning_rate": 1.0045379845431877e-05,
+      "loss": 1.6183,
+      "step": 3849
+    },
+    {
+      "epoch": 0.5137443287963704,
+      "grad_norm": 0.9562275485239146,
+      "learning_rate": 1.004105798095465e-05,
+      "loss": 1.5953,
+      "step": 3850
+    },
+    {
+      "epoch": 0.5138777688817721,
+      "grad_norm": 1.0389139499008198,
+      "learning_rate": 1.0036736108808258e-05,
+      "loss": 1.6092,
+      "step": 3851
+    },
+    {
+      "epoch": 0.5140112089671738,
+      "grad_norm": 1.1170799172414176,
+      "learning_rate": 1.0032414229799978e-05,
+      "loss": 1.6152,
+      "step": 3852
+    },
+    {
+      "epoch": 0.5141446490525754,
+      "grad_norm": 1.05178593292583,
+      "learning_rate": 1.0028092344737093e-05,
+      "loss": 1.5869,
+      "step": 3853
+    },
+    {
+      "epoch": 0.514278089137977,
+      "grad_norm": 0.9406322853227639,
+      "learning_rate": 1.0023770454426873e-05,
+      "loss": 1.5934,
+      "step": 3854
+    },
+    {
+      "epoch": 0.5144115292233787,
+      "grad_norm": 1.0236251156656235,
+      "learning_rate": 1.0019448559676605e-05,
+      "loss": 1.5402,
+      "step": 3855
+    },
+    {
+      "epoch": 0.5145449693087804,
+      "grad_norm": 0.9301092670030722,
+      "learning_rate": 1.0015126661293566e-05,
+      "loss": 1.5849,
+      "step": 3856
+    },
+    {
+      "epoch": 0.514678409394182,
+      "grad_norm": 0.9761352830656781,
+      "learning_rate": 1.0010804760085037e-05,
+      "loss": 1.5598,
+      "step": 3857
+    },
+    {
+      "epoch": 0.5148118494795837,
+      "grad_norm": 0.9784181556059924,
+      "learning_rate": 1.0006482856858306e-05,
+      "loss": 1.6337,
+      "step": 3858
+    },
+    {
+      "epoch": 0.5149452895649853,
+      "grad_norm": 1.0310199190875282,
+      "learning_rate": 1.000216095242065e-05,
+      "loss": 1.5553,
+      "step": 3859
+    },
+    {
+      "epoch": 0.5150787296503869,
+      "grad_norm": 0.9413946989014104,
+      "learning_rate": 9.997839047579351e-06,
+      "loss": 1.5955,
+      "step": 3860
+    },
+    {
+      "epoch": 0.5152121697357886,
+      "grad_norm": 1.0310793127717695,
+      "learning_rate": 9.993517143141695e-06,
+      "loss": 1.5891,
+      "step": 3861
+    },
+    {
+      "epoch": 0.5153456098211903,
+      "grad_norm": 1.1466677120724456,
+      "learning_rate": 9.989195239914964e-06,
+      "loss": 1.5896,
+      "step": 3862
+    },
+    {
+      "epoch": 0.515479049906592,
+      "grad_norm": 2.0571376714105436,
+      "learning_rate": 9.984873338706439e-06,
+      "loss": 1.6166,
+      "step": 3863
+    },
+    {
+      "epoch": 0.5156124899919936,
+      "grad_norm": 0.9772929955397268,
+      "learning_rate": 9.980551440323398e-06,
+      "loss": 1.6124,
+      "step": 3864
+    },
+    {
+      "epoch": 0.5157459300773952,
+      "grad_norm": 0.995449303499809,
+      "learning_rate": 9.97622954557313e-06,
+      "loss": 1.55,
+      "step": 3865
+    },
+    {
+      "epoch": 0.515879370162797,
+      "grad_norm": 0.9947089431138232,
+      "learning_rate": 9.971907655262914e-06,
+      "loss": 1.6228,
+      "step": 3866
+    },
+    {
+      "epoch": 0.5160128102481986,
+      "grad_norm": 1.0584016266307699,
+      "learning_rate": 9.967585770200023e-06,
+      "loss": 1.5816,
+      "step": 3867
+    },
+    {
+      "epoch": 0.5161462503336002,
+      "grad_norm": 3.1959955421300643,
+      "learning_rate": 9.963263891191743e-06,
+      "loss": 1.584,
+      "step": 3868
+    },
+    {
+      "epoch": 0.5162796904190019,
+      "grad_norm": 1.0146744361794484,
+      "learning_rate": 9.958942019045352e-06,
+      "loss": 1.5777,
+      "step": 3869
+    },
+    {
+      "epoch": 0.5164131305044035,
+      "grad_norm": 1.0815439584288364,
+      "learning_rate": 9.954620154568125e-06,
+      "loss": 1.5771,
+      "step": 3870
+    },
+    {
+      "epoch": 0.5165465705898051,
+      "grad_norm": 0.9947735001779133,
+      "learning_rate": 9.950298298567333e-06,
+      "loss": 1.6323,
+      "step": 3871
+    },
+    {
+      "epoch": 0.5166800106752069,
+      "grad_norm": 1.025350666289356,
+      "learning_rate": 9.945976451850256e-06,
+      "loss": 1.5726,
+      "step": 3872
+    },
+    {
+      "epoch": 0.5168134507606085,
+      "grad_norm": 1.0309153481695548,
+      "learning_rate": 9.941654615224162e-06,
+      "loss": 1.6411,
+      "step": 3873
+    },
+    {
+      "epoch": 0.5169468908460101,
+      "grad_norm": 0.9706291026832317,
+      "learning_rate": 9.937332789496318e-06,
+      "loss": 1.5344,
+      "step": 3874
+    },
+    {
+      "epoch": 0.5170803309314118,
+      "grad_norm": 1.0442149306572364,
+      "learning_rate": 9.933010975474e-06,
+      "loss": 1.5868,
+      "step": 3875
+    },
+    {
+      "epoch": 0.5172137710168134,
+      "grad_norm": 0.9801227554963917,
+      "learning_rate": 9.928689173964467e-06,
+      "loss": 1.584,
+      "step": 3876
+    },
+    {
+      "epoch": 0.5173472111022152,
+      "grad_norm": 0.9612770789559671,
+      "learning_rate": 9.924367385774987e-06,
+      "loss": 1.5624,
+      "step": 3877
+    },
+    {
+      "epoch": 0.5174806511876168,
+      "grad_norm": 12.372251835234461,
+      "learning_rate": 9.920045611712812e-06,
+      "loss": 1.6331,
+      "step": 3878
+    },
+    {
+      "epoch": 0.5176140912730184,
+      "grad_norm": 1.0323071031650972,
+      "learning_rate": 9.915723852585211e-06,
+      "loss": 1.5644,
+      "step": 3879
+    },
+    {
+      "epoch": 0.5177475313584201,
+      "grad_norm": 1.0939706773461586,
+      "learning_rate": 9.911402109199437e-06,
+      "loss": 1.5847,
+      "step": 3880
+    },
+    {
+      "epoch": 0.5178809714438217,
+      "grad_norm": 1.1539521887238684,
+      "learning_rate": 9.907080382362733e-06,
+      "loss": 1.6092,
+      "step": 3881
+    },
+    {
+      "epoch": 0.5180144115292233,
+      "grad_norm": 1.4844965472462381,
+      "learning_rate": 9.902758672882361e-06,
+      "loss": 1.6156,
+      "step": 3882
+    },
+    {
+      "epoch": 0.5181478516146251,
+      "grad_norm": 1.0477571393778307,
+      "learning_rate": 9.898436981565562e-06,
+      "loss": 1.5609,
+      "step": 3883
+    },
+    {
+      "epoch": 0.5182812917000267,
+      "grad_norm": 0.9919178224095909,
+      "learning_rate": 9.894115309219579e-06,
+      "loss": 1.6325,
+      "step": 3884
+    },
+    {
+      "epoch": 0.5184147317854283,
+      "grad_norm": 1.039487001315658,
+      "learning_rate": 9.889793656651646e-06,
+      "loss": 1.6016,
+      "step": 3885
+    },
+    {
+      "epoch": 0.51854817187083,
+      "grad_norm": 0.9999173249401138,
+      "learning_rate": 9.885472024669007e-06,
+      "loss": 1.5301,
+      "step": 3886
+    },
+    {
+      "epoch": 0.5186816119562316,
+      "grad_norm": 1.0016577339792758,
+      "learning_rate": 9.881150414078891e-06,
+      "loss": 1.6022,
+      "step": 3887
+    },
+    {
+      "epoch": 0.5188150520416333,
+      "grad_norm": 1.022731064253213,
+      "learning_rate": 9.87682882568852e-06,
+      "loss": 1.6181,
+      "step": 3888
+    },
+    {
+      "epoch": 0.518948492127035,
+      "grad_norm": 1.053953730841017,
+      "learning_rate": 9.872507260305124e-06,
+      "loss": 1.5722,
+      "step": 3889
+    },
+    {
+      "epoch": 0.5190819322124366,
+      "grad_norm": 0.9483103314838104,
+      "learning_rate": 9.868185718735922e-06,
+      "loss": 1.5721,
+      "step": 3890
+    },
+    {
+      "epoch": 0.5192153722978383,
+      "grad_norm": 0.9716784609015309,
+      "learning_rate": 9.863864201788127e-06,
+      "loss": 1.5958,
+      "step": 3891
+    },
+    {
+      "epoch": 0.5193488123832399,
+      "grad_norm": 7.41992607929927,
+      "learning_rate": 9.859542710268947e-06,
+      "loss": 1.601,
+      "step": 3892
+    },
+    {
+      "epoch": 0.5194822524686415,
+      "grad_norm": 1.0842848553275373,
+      "learning_rate": 9.85522124498559e-06,
+      "loss": 1.6163,
+      "step": 3893
+    },
+    {
+      "epoch": 0.5196156925540433,
+      "grad_norm": 1.1825348756660392,
+      "learning_rate": 9.850899806745258e-06,
+      "loss": 1.5681,
+      "step": 3894
+    },
+    {
+      "epoch": 0.5197491326394449,
+      "grad_norm": 1.219221323073541,
+      "learning_rate": 9.846578396355144e-06,
+      "loss": 1.6045,
+      "step": 3895
+    },
+    {
+      "epoch": 0.5198825727248465,
+      "grad_norm": 0.9727027071204921,
+      "learning_rate": 9.842257014622436e-06,
+      "loss": 1.6032,
+      "step": 3896
+    },
+    {
+      "epoch": 0.5200160128102482,
+      "grad_norm": 0.9822928747027722,
+      "learning_rate": 9.837935662354325e-06,
+      "loss": 1.5696,
+      "step": 3897
+    },
+    {
+      "epoch": 0.5201494528956498,
+      "grad_norm": 0.9616838656896538,
+      "learning_rate": 9.833614340357986e-06,
+      "loss": 1.5782,
+      "step": 3898
+    },
+    {
+      "epoch": 0.5202828929810515,
+      "grad_norm": 1.0479795740977809,
+      "learning_rate": 9.829293049440592e-06,
+      "loss": 1.5405,
+      "step": 3899
+    },
+    {
+      "epoch": 0.5204163330664532,
+      "grad_norm": 1.02813031490662,
+      "learning_rate": 9.824971790409314e-06,
+      "loss": 1.5542,
+      "step": 3900
+    },
+    {
+      "epoch": 0.5205497731518548,
+      "grad_norm": 1.1751181903794317,
+      "learning_rate": 9.820650564071314e-06,
+      "loss": 1.6051,
+      "step": 3901
+    },
+    {
+      "epoch": 0.5206832132372565,
+      "grad_norm": 1.01933272576621,
+      "learning_rate": 9.816329371233747e-06,
+      "loss": 1.6075,
+      "step": 3902
+    },
+    {
+      "epoch": 0.5208166533226581,
+      "grad_norm": 1.065928748207789,
+      "learning_rate": 9.812008212703758e-06,
+      "loss": 1.6226,
+      "step": 3903
+    },
+    {
+      "epoch": 0.5209500934080598,
+      "grad_norm": 1.0491540793462275,
+      "learning_rate": 9.8076870892885e-06,
+      "loss": 1.62,
+      "step": 3904
+    },
+    {
+      "epoch": 0.5210835334934615,
+      "grad_norm": 0.9904597628301255,
+      "learning_rate": 9.803366001795102e-06,
+      "loss": 1.5997,
+      "step": 3905
+    },
+    {
+      "epoch": 0.5212169735788631,
+      "grad_norm": 0.9273698585055911,
+      "learning_rate": 9.799044951030693e-06,
+      "loss": 1.5687,
+      "step": 3906
+    },
+    {
+      "epoch": 0.5213504136642647,
+      "grad_norm": 1.051717478475509,
+      "learning_rate": 9.794723937802403e-06,
+      "loss": 1.5735,
+      "step": 3907
+    },
+    {
+      "epoch": 0.5214838537496664,
+      "grad_norm": 0.9428209123723238,
+      "learning_rate": 9.790402962917343e-06,
+      "loss": 1.5791,
+      "step": 3908
+    },
+    {
+      "epoch": 0.521617293835068,
+      "grad_norm": 0.9740214957204884,
+      "learning_rate": 9.786082027182626e-06,
+      "loss": 1.6194,
+      "step": 3909
+    },
+    {
+      "epoch": 0.5217507339204697,
+      "grad_norm": 0.9865488024149134,
+      "learning_rate": 9.781761131405346e-06,
+      "loss": 1.5704,
+      "step": 3910
+    },
+    {
+      "epoch": 0.5218841740058714,
+      "grad_norm": 0.9376345788702144,
+      "learning_rate": 9.777440276392604e-06,
+      "loss": 1.5978,
+      "step": 3911
+    },
+    {
+      "epoch": 0.522017614091273,
+      "grad_norm": 0.9550562216569386,
+      "learning_rate": 9.773119462951487e-06,
+      "loss": 1.6584,
+      "step": 3912
+    },
+    {
+      "epoch": 0.5221510541766746,
+      "grad_norm": 0.9603118976190848,
+      "learning_rate": 9.768798691889066e-06,
+      "loss": 1.5987,
+      "step": 3913
+    },
+    {
+      "epoch": 0.5222844942620763,
+      "grad_norm": 1.1573687573433378,
+      "learning_rate": 9.76447796401242e-06,
+      "loss": 1.5802,
+      "step": 3914
+    },
+    {
+      "epoch": 0.522417934347478,
+      "grad_norm": 1.0874659242828772,
+      "learning_rate": 9.760157280128611e-06,
+      "loss": 1.629,
+      "step": 3915
+    },
+    {
+      "epoch": 0.5225513744328797,
+      "grad_norm": 1.2673180853706751,
+      "learning_rate": 9.755836641044686e-06,
+      "loss": 1.5725,
+      "step": 3916
+    },
+    {
+      "epoch": 0.5226848145182813,
+      "grad_norm": 0.933164844674267,
+      "learning_rate": 9.7515160475677e-06,
+      "loss": 1.6217,
+      "step": 3917
+    },
+    {
+      "epoch": 0.5228182546036829,
+      "grad_norm": 0.951797289455154,
+      "learning_rate": 9.747195500504687e-06,
+      "loss": 1.5448,
+      "step": 3918
+    },
+    {
+      "epoch": 0.5229516946890846,
+      "grad_norm": 0.9440107265148531,
+      "learning_rate": 9.742875000662679e-06,
+      "loss": 1.5796,
+      "step": 3919
+    },
+    {
+      "epoch": 0.5230851347744863,
+      "grad_norm": 0.9563610291259368,
+      "learning_rate": 9.738554548848686e-06,
+      "loss": 1.5944,
+      "step": 3920
+    },
+    {
+      "epoch": 0.5232185748598879,
+      "grad_norm": 0.9662588914444328,
+      "learning_rate": 9.734234145869731e-06,
+      "loss": 1.5565,
+      "step": 3921
+    },
+    {
+      "epoch": 0.5233520149452896,
+      "grad_norm": 0.9669810427348244,
+      "learning_rate": 9.729913792532813e-06,
+      "loss": 1.584,
+      "step": 3922
+    },
+    {
+      "epoch": 0.5234854550306912,
+      "grad_norm": 0.9499418557964407,
+      "learning_rate": 9.725593489644925e-06,
+      "loss": 1.5611,
+      "step": 3923
+    },
+    {
+      "epoch": 0.5236188951160928,
+      "grad_norm": 0.9825228300383785,
+      "learning_rate": 9.721273238013042e-06,
+      "loss": 1.5699,
+      "step": 3924
+    },
+    {
+      "epoch": 0.5237523352014946,
+      "grad_norm": 1.0180904933312087,
+      "learning_rate": 9.716953038444149e-06,
+      "loss": 1.6399,
+      "step": 3925
+    },
+    {
+      "epoch": 0.5238857752868962,
+      "grad_norm": 0.9427216935978336,
+      "learning_rate": 9.712632891745208e-06,
+      "loss": 1.6125,
+      "step": 3926
+    },
+    {
+      "epoch": 0.5240192153722978,
+      "grad_norm": 0.9985494748443655,
+      "learning_rate": 9.708312798723168e-06,
+      "loss": 1.607,
+      "step": 3927
+    },
+    {
+      "epoch": 0.5241526554576995,
+      "grad_norm": 0.9775720421880318,
+      "learning_rate": 9.703992760184976e-06,
+      "loss": 1.6487,
+      "step": 3928
+    },
+    {
+      "epoch": 0.5242860955431011,
+      "grad_norm": 1.0440323447007265,
+      "learning_rate": 9.699672776937567e-06,
+      "loss": 1.6203,
+      "step": 3929
+    },
+    {
+      "epoch": 0.5244195356285029,
+      "grad_norm": 1.073896330976579,
+      "learning_rate": 9.695352849787865e-06,
+      "loss": 1.6193,
+      "step": 3930
+    },
+    {
+      "epoch": 0.5245529757139045,
+      "grad_norm": 0.9666654017851334,
+      "learning_rate": 9.69103297954278e-06,
+      "loss": 1.6498,
+      "step": 3931
+    },
+    {
+      "epoch": 0.5246864157993061,
+      "grad_norm": 1.1178863635267453,
+      "learning_rate": 9.686713167009216e-06,
+      "loss": 1.6103,
+      "step": 3932
+    },
+    {
+      "epoch": 0.5248198558847078,
+      "grad_norm": 1.0066761509572055,
+      "learning_rate": 9.68239341299407e-06,
+      "loss": 1.6333,
+      "step": 3933
+    },
+    {
+      "epoch": 0.5249532959701094,
+      "grad_norm": 0.9297439952534905,
+      "learning_rate": 9.678073718304215e-06,
+      "loss": 1.5461,
+      "step": 3934
+    },
+    {
+      "epoch": 0.525086736055511,
+      "grad_norm": 0.9488135408582773,
+      "learning_rate": 9.673754083746523e-06,
+      "loss": 1.6361,
+      "step": 3935
+    },
+    {
+      "epoch": 0.5252201761409128,
+      "grad_norm": 0.9845229518536815,
+      "learning_rate": 9.669434510127858e-06,
+      "loss": 1.6412,
+      "step": 3936
+    },
+    {
+      "epoch": 0.5253536162263144,
+      "grad_norm": 0.9802847942341244,
+      "learning_rate": 9.66511499825506e-06,
+      "loss": 1.5989,
+      "step": 3937
+    },
+    {
+      "epoch": 0.525487056311716,
+      "grad_norm": 0.9632382904131538,
+      "learning_rate": 9.660795548934965e-06,
+      "loss": 1.6071,
+      "step": 3938
+    },
+    {
+      "epoch": 0.5256204963971177,
+      "grad_norm": 1.0699178531684905,
+      "learning_rate": 9.656476162974403e-06,
+      "loss": 1.6323,
+      "step": 3939
+    },
+    {
+      "epoch": 0.5257539364825193,
+      "grad_norm": 0.9511806151181409,
+      "learning_rate": 9.652156841180185e-06,
+      "loss": 1.6131,
+      "step": 3940
+    },
+    {
+      "epoch": 0.525887376567921,
+      "grad_norm": 0.9838327792979288,
+      "learning_rate": 9.647837584359107e-06,
+      "loss": 1.5408,
+      "step": 3941
+    },
+    {
+      "epoch": 0.5260208166533227,
+      "grad_norm": 1.1196962403177173,
+      "learning_rate": 9.643518393317953e-06,
+      "loss": 1.583,
+      "step": 3942
+    },
+    {
+      "epoch": 0.5261542567387243,
+      "grad_norm": 0.9499526650383459,
+      "learning_rate": 9.63919926886351e-06,
+      "loss": 1.567,
+      "step": 3943
+    },
+    {
+      "epoch": 0.526287696824126,
+      "grad_norm": 0.9520997264212869,
+      "learning_rate": 9.634880211802535e-06,
+      "loss": 1.6393,
+      "step": 3944
+    },
+    {
+      "epoch": 0.5264211369095276,
+      "grad_norm": 1.2483925528165147,
+      "learning_rate": 9.630561222941776e-06,
+      "loss": 1.6007,
+      "step": 3945
+    },
+    {
+      "epoch": 0.5265545769949292,
+      "grad_norm": 0.9647842321723614,
+      "learning_rate": 9.62624230308798e-06,
+      "loss": 1.6036,
+      "step": 3946
+    },
+    {
+      "epoch": 0.526688017080331,
+      "grad_norm": 1.051538449954908,
+      "learning_rate": 9.621923453047864e-06,
+      "loss": 1.6024,
+      "step": 3947
+    },
+    {
+      "epoch": 0.5268214571657326,
+      "grad_norm": 1.0353802905237608,
+      "learning_rate": 9.617604673628142e-06,
+      "loss": 1.5987,
+      "step": 3948
+    },
+    {
+      "epoch": 0.5269548972511342,
+      "grad_norm": 0.9689586440798221,
+      "learning_rate": 9.61328596563551e-06,
+      "loss": 1.5775,
+      "step": 3949
+    },
+    {
+      "epoch": 0.5270883373365359,
+      "grad_norm": 1.110258725459353,
+      "learning_rate": 9.608967329876662e-06,
+      "loss": 1.5446,
+      "step": 3950
+    },
+    {
+      "epoch": 0.5272217774219375,
+      "grad_norm": 0.9779790351795254,
+      "learning_rate": 9.604648767158263e-06,
+      "loss": 1.5587,
+      "step": 3951
+    },
+    {
+      "epoch": 0.5273552175073392,
+      "grad_norm": 0.9136985660924678,
+      "learning_rate": 9.600330278286972e-06,
+      "loss": 1.5626,
+      "step": 3952
+    },
+    {
+      "epoch": 0.5274886575927409,
+      "grad_norm": 0.9427079281067329,
+      "learning_rate": 9.596011864069434e-06,
+      "loss": 1.597,
+      "step": 3953
+    },
+    {
+      "epoch": 0.5276220976781425,
+      "grad_norm": 0.9576255235969523,
+      "learning_rate": 9.591693525312283e-06,
+      "loss": 1.5782,
+      "step": 3954
+    },
+    {
+      "epoch": 0.5277555377635441,
+      "grad_norm": 0.9802126130169267,
+      "learning_rate": 9.587375262822132e-06,
+      "loss": 1.572,
+      "step": 3955
+    },
+    {
+      "epoch": 0.5278889778489458,
+      "grad_norm": 0.967258469940652,
+      "learning_rate": 9.583057077405583e-06,
+      "loss": 1.5762,
+      "step": 3956
+    },
+    {
+      "epoch": 0.5280224179343475,
+      "grad_norm": 1.034695214577326,
+      "learning_rate": 9.578738969869227e-06,
+      "loss": 1.5677,
+      "step": 3957
+    },
+    {
+      "epoch": 0.5281558580197492,
+      "grad_norm": 0.9403505616977419,
+      "learning_rate": 9.574420941019634e-06,
+      "loss": 1.5619,
+      "step": 3958
+    },
+    {
+      "epoch": 0.5282892981051508,
+      "grad_norm": 0.9592411207184994,
+      "learning_rate": 9.570102991663361e-06,
+      "loss": 1.6072,
+      "step": 3959
+    },
+    {
+      "epoch": 0.5284227381905524,
+      "grad_norm": 0.9743736463148753,
+      "learning_rate": 9.565785122606957e-06,
+      "loss": 1.5505,
+      "step": 3960
+    },
+    {
+      "epoch": 0.5285561782759541,
+      "grad_norm": 1.0090249364331059,
+      "learning_rate": 9.561467334656949e-06,
+      "loss": 1.5881,
+      "step": 3961
+    },
+    {
+      "epoch": 0.5286896183613558,
+      "grad_norm": 0.9420075410877025,
+      "learning_rate": 9.55714962861985e-06,
+      "loss": 1.618,
+      "step": 3962
+    },
+    {
+      "epoch": 0.5288230584467574,
+      "grad_norm": 0.9692223836450211,
+      "learning_rate": 9.552832005302154e-06,
+      "loss": 1.5772,
+      "step": 3963
+    },
+    {
+      "epoch": 0.5289564985321591,
+      "grad_norm": 0.9489004175073099,
+      "learning_rate": 9.548514465510353e-06,
+      "loss": 1.6164,
+      "step": 3964
+    },
+    {
+      "epoch": 0.5290899386175607,
+      "grad_norm": 0.9421154885492968,
+      "learning_rate": 9.54419701005091e-06,
+      "loss": 1.5717,
+      "step": 3965
+    },
+    {
+      "epoch": 0.5292233787029623,
+      "grad_norm": 0.9424115466890982,
+      "learning_rate": 9.539879639730276e-06,
+      "loss": 1.5631,
+      "step": 3966
+    },
+    {
+      "epoch": 0.529356818788364,
+      "grad_norm": 2.7276724914362083,
+      "learning_rate": 9.53556235535488e-06,
+      "loss": 1.5696,
+      "step": 3967
+    },
+    {
+      "epoch": 0.5294902588737657,
+      "grad_norm": 0.9386033684279529,
+      "learning_rate": 9.531245157731154e-06,
+      "loss": 1.5285,
+      "step": 3968
+    },
+    {
+      "epoch": 0.5296236989591674,
+      "grad_norm": 0.8899585596068399,
+      "learning_rate": 9.526928047665495e-06,
+      "loss": 1.5118,
+      "step": 3969
+    },
+    {
+      "epoch": 0.529757139044569,
+      "grad_norm": 0.9826968773740292,
+      "learning_rate": 9.522611025964288e-06,
+      "loss": 1.5888,
+      "step": 3970
+    },
+    {
+      "epoch": 0.5298905791299706,
+      "grad_norm": 1.0400188151367677,
+      "learning_rate": 9.51829409343391e-06,
+      "loss": 1.6159,
+      "step": 3971
+    },
+    {
+      "epoch": 0.5300240192153723,
+      "grad_norm": 0.9561987813115096,
+      "learning_rate": 9.51397725088071e-06,
+      "loss": 1.5299,
+      "step": 3972
+    },
+    {
+      "epoch": 0.530157459300774,
+      "grad_norm": 0.9612234452056033,
+      "learning_rate": 9.509660499111027e-06,
+      "loss": 1.5819,
+      "step": 3973
+    },
+    {
+      "epoch": 0.5302908993861756,
+      "grad_norm": 0.9139976394076528,
+      "learning_rate": 9.505343838931177e-06,
+      "loss": 1.5485,
+      "step": 3974
+    },
+    {
+      "epoch": 0.5304243394715773,
+      "grad_norm": 1.0692106111452742,
+      "learning_rate": 9.501027271147466e-06,
+      "loss": 1.581,
+      "step": 3975
+    },
+    {
+      "epoch": 0.5305577795569789,
+      "grad_norm": 1.0131134517933047,
+      "learning_rate": 9.496710796566181e-06,
+      "loss": 1.6409,
+      "step": 3976
+    },
+    {
+      "epoch": 0.5306912196423805,
+      "grad_norm": 0.9508397877941868,
+      "learning_rate": 9.492394415993587e-06,
+      "loss": 1.5288,
+      "step": 3977
+    },
+    {
+      "epoch": 0.5308246597277823,
+      "grad_norm": 1.0043689220791059,
+      "learning_rate": 9.48807813023594e-06,
+      "loss": 1.5687,
+      "step": 3978
+    },
+    {
+      "epoch": 0.5309580998131839,
+      "grad_norm": 1.0338708219457289,
+      "learning_rate": 9.483761940099471e-06,
+      "loss": 1.6097,
+      "step": 3979
+    },
+    {
+      "epoch": 0.5310915398985855,
+      "grad_norm": 0.9661798704066481,
+      "learning_rate": 9.479445846390389e-06,
+      "loss": 1.6003,
+      "step": 3980
+    },
+    {
+      "epoch": 0.5312249799839872,
+      "grad_norm": 1.1436314427778058,
+      "learning_rate": 9.475129849914898e-06,
+      "loss": 1.6077,
+      "step": 3981
+    },
+    {
+      "epoch": 0.5313584200693888,
+      "grad_norm": 1.1001896581114239,
+      "learning_rate": 9.470813951479177e-06,
+      "loss": 1.5785,
+      "step": 3982
+    },
+    {
+      "epoch": 0.5314918601547906,
+      "grad_norm": 0.9585361064875515,
+      "learning_rate": 9.466498151889385e-06,
+      "loss": 1.5498,
+      "step": 3983
+    },
+    {
+      "epoch": 0.5316253002401922,
+      "grad_norm": 1.0620960267902282,
+      "learning_rate": 9.46218245195166e-06,
+      "loss": 1.5932,
+      "step": 3984
+    },
+    {
+      "epoch": 0.5317587403255938,
+      "grad_norm": 1.0499763780853832,
+      "learning_rate": 9.457866852472132e-06,
+      "loss": 1.6056,
+      "step": 3985
+    },
+    {
+      "epoch": 0.5318921804109955,
+      "grad_norm": 0.973771229695574,
+      "learning_rate": 9.453551354256907e-06,
+      "loss": 1.5651,
+      "step": 3986
+    },
+    {
+      "epoch": 0.5320256204963971,
+      "grad_norm": 0.9807116282877302,
+      "learning_rate": 9.449235958112065e-06,
+      "loss": 1.5977,
+      "step": 3987
+    },
+    {
+      "epoch": 0.5321590605817987,
+      "grad_norm": 0.9631392166085614,
+      "learning_rate": 9.444920664843671e-06,
+      "loss": 1.6167,
+      "step": 3988
+    },
+    {
+      "epoch": 0.5322925006672005,
+      "grad_norm": 0.993702665081086,
+      "learning_rate": 9.440605475257782e-06,
+      "loss": 1.5629,
+      "step": 3989
+    },
+    {
+      "epoch": 0.5324259407526021,
+      "grad_norm": 1.107669942509192,
+      "learning_rate": 9.436290390160422e-06,
+      "loss": 1.5701,
+      "step": 3990
+    },
+    {
+      "epoch": 0.5325593808380037,
+      "grad_norm": 1.0557051966228765,
+      "learning_rate": 9.431975410357597e-06,
+      "loss": 1.5419,
+      "step": 3991
+    },
+    {
+      "epoch": 0.5326928209234054,
+      "grad_norm": 1.0739759302965814,
+      "learning_rate": 9.427660536655296e-06,
+      "loss": 1.6363,
+      "step": 3992
+    },
+    {
+      "epoch": 0.532826261008807,
+      "grad_norm": 0.9274766595100078,
+      "learning_rate": 9.423345769859494e-06,
+      "loss": 1.5837,
+      "step": 3993
+    },
+    {
+      "epoch": 0.5329597010942086,
+      "grad_norm": 0.9774667853108802,
+      "learning_rate": 9.419031110776137e-06,
+      "loss": 1.6363,
+      "step": 3994
+    },
+    {
+      "epoch": 0.5330931411796104,
+      "grad_norm": 0.9665074812404877,
+      "learning_rate": 9.41471656021115e-06,
+      "loss": 1.5858,
+      "step": 3995
+    },
+    {
+      "epoch": 0.533226581265012,
+      "grad_norm": 0.9516009165945285,
+      "learning_rate": 9.410402118970447e-06,
+      "loss": 1.6294,
+      "step": 3996
+    },
+    {
+      "epoch": 0.5333600213504137,
+      "grad_norm": 0.9668981552357112,
+      "learning_rate": 9.40608778785992e-06,
+      "loss": 1.5588,
+      "step": 3997
+    },
+    {
+      "epoch": 0.5334934614358153,
+      "grad_norm": 0.9784909518756532,
+      "learning_rate": 9.401773567685428e-06,
+      "loss": 1.5723,
+      "step": 3998
+    },
+    {
+      "epoch": 0.533626901521217,
+      "grad_norm": 0.9517428419308205,
+      "learning_rate": 9.397459459252823e-06,
+      "loss": 1.6369,
+      "step": 3999
+    },
+    {
+      "epoch": 0.5337603416066187,
+      "grad_norm": 0.9506059122541574,
+      "learning_rate": 9.393145463367932e-06,
+      "loss": 1.6134,
+      "step": 4000
+    },
+    {
+      "epoch": 0.5338937816920203,
+      "grad_norm": 0.9352905785580472,
+      "learning_rate": 9.388831580836559e-06,
+      "loss": 1.596,
+      "step": 4001
+    },
+    {
+      "epoch": 0.5340272217774219,
+      "grad_norm": 0.9230020903319824,
+      "learning_rate": 9.384517812464485e-06,
+      "loss": 1.542,
+      "step": 4002
+    },
+    {
+      "epoch": 0.5341606618628236,
+      "grad_norm": 1.0665813163596254,
+      "learning_rate": 9.38020415905748e-06,
+      "loss": 1.5823,
+      "step": 4003
+    },
+    {
+      "epoch": 0.5342941019482252,
+      "grad_norm": 0.9983547945807545,
+      "learning_rate": 9.37589062142128e-06,
+      "loss": 1.6159,
+      "step": 4004
+    },
+    {
+      "epoch": 0.5344275420336269,
+      "grad_norm": 1.0317932796668363,
+      "learning_rate": 9.371577200361609e-06,
+      "loss": 1.5753,
+      "step": 4005
+    },
+    {
+      "epoch": 0.5345609821190286,
+      "grad_norm": 0.9463210756193553,
+      "learning_rate": 9.367263896684155e-06,
+      "loss": 1.5713,
+      "step": 4006
+    },
+    {
+      "epoch": 0.5346944222044302,
+      "grad_norm": 0.9578078666890666,
+      "learning_rate": 9.362950711194607e-06,
+      "loss": 1.5438,
+      "step": 4007
+    },
+    {
+      "epoch": 0.5348278622898318,
+      "grad_norm": 0.9984488850157227,
+      "learning_rate": 9.358637644698614e-06,
+      "loss": 1.5737,
+      "step": 4008
+    },
+    {
+      "epoch": 0.5349613023752335,
+      "grad_norm": 0.9572054111804951,
+      "learning_rate": 9.354324698001802e-06,
+      "loss": 1.5819,
+      "step": 4009
+    },
+    {
+      "epoch": 0.5350947424606352,
+      "grad_norm": 0.9538875645675433,
+      "learning_rate": 9.350011871909787e-06,
+      "loss": 1.5913,
+      "step": 4010
+    },
+    {
+      "epoch": 0.5352281825460369,
+      "grad_norm": 1.0188818733645557,
+      "learning_rate": 9.345699167228156e-06,
+      "loss": 1.5497,
+      "step": 4011
+    },
+    {
+      "epoch": 0.5353616226314385,
+      "grad_norm": 0.9520922440061501,
+      "learning_rate": 9.34138658476247e-06,
+      "loss": 1.6327,
+      "step": 4012
+    },
+    {
+      "epoch": 0.5354950627168401,
+      "grad_norm": 0.9539885539724596,
+      "learning_rate": 9.337074125318268e-06,
+      "loss": 1.579,
+      "step": 4013
+    },
+    {
+      "epoch": 0.5356285028022418,
+      "grad_norm": 1.2835828749577056,
+      "learning_rate": 9.332761789701076e-06,
+      "loss": 1.5306,
+      "step": 4014
+    },
+    {
+      "epoch": 0.5357619428876434,
+      "grad_norm": 1.023893323462547,
+      "learning_rate": 9.328449578716384e-06,
+      "loss": 1.5653,
+      "step": 4015
+    },
+    {
+      "epoch": 0.5358953829730451,
+      "grad_norm": 1.2706402832851262,
+      "learning_rate": 9.324137493169664e-06,
+      "loss": 1.561,
+      "step": 4016
+    },
+    {
+      "epoch": 0.5360288230584468,
+      "grad_norm": 0.9500733970079291,
+      "learning_rate": 9.319825533866364e-06,
+      "loss": 1.5684,
+      "step": 4017
+    },
+    {
+      "epoch": 0.5361622631438484,
+      "grad_norm": 0.9654037579664738,
+      "learning_rate": 9.315513701611913e-06,
+      "loss": 1.5545,
+      "step": 4018
+    },
+    {
+      "epoch": 0.53629570322925,
+      "grad_norm": 0.9637541303641998,
+      "learning_rate": 9.31120199721171e-06,
+      "loss": 1.6,
+      "step": 4019
+    },
+    {
+      "epoch": 0.5364291433146517,
+      "grad_norm": 0.9449747779438818,
+      "learning_rate": 9.306890421471128e-06,
+      "loss": 1.5729,
+      "step": 4020
+    },
+    {
+      "epoch": 0.5365625834000534,
+      "grad_norm": 1.015400427968874,
+      "learning_rate": 9.302578975195527e-06,
+      "loss": 1.5912,
+      "step": 4021
+    },
+    {
+      "epoch": 0.536696023485455,
+      "grad_norm": 1.2000776620053406,
+      "learning_rate": 9.298267659190234e-06,
+      "loss": 1.6051,
+      "step": 4022
+    },
+    {
+      "epoch": 0.5368294635708567,
+      "grad_norm": 0.9947410519922049,
+      "learning_rate": 9.293956474260548e-06,
+      "loss": 1.584,
+      "step": 4023
+    },
+    {
+      "epoch": 0.5369629036562583,
+      "grad_norm": 0.9569770795832728,
+      "learning_rate": 9.289645421211756e-06,
+      "loss": 1.5406,
+      "step": 4024
+    },
+    {
+      "epoch": 0.53709634374166,
+      "grad_norm": 1.0002749981588506,
+      "learning_rate": 9.285334500849113e-06,
+      "loss": 1.6485,
+      "step": 4025
+    },
+    {
+      "epoch": 0.5372297838270617,
+      "grad_norm": 0.9254684190371253,
+      "learning_rate": 9.281023713977848e-06,
+      "loss": 1.6221,
+      "step": 4026
+    },
+    {
+      "epoch": 0.5373632239124633,
+      "grad_norm": 1.0069500001650689,
+      "learning_rate": 9.276713061403164e-06,
+      "loss": 1.6204,
+      "step": 4027
+    },
+    {
+      "epoch": 0.537496663997865,
+      "grad_norm": 1.0859626601584031,
+      "learning_rate": 9.27240254393025e-06,
+      "loss": 1.5635,
+      "step": 4028
+    },
+    {
+      "epoch": 0.5376301040832666,
+      "grad_norm": 0.9601269217825726,
+      "learning_rate": 9.268092162364256e-06,
+      "loss": 1.6201,
+      "step": 4029
+    },
+    {
+      "epoch": 0.5377635441686682,
+      "grad_norm": 1.005400728893139,
+      "learning_rate": 9.263781917510312e-06,
+      "loss": 1.559,
+      "step": 4030
+    },
+    {
+      "epoch": 0.53789698425407,
+      "grad_norm": 0.9879107408018795,
+      "learning_rate": 9.259471810173522e-06,
+      "loss": 1.5615,
+      "step": 4031
+    },
+    {
+      "epoch": 0.5380304243394716,
+      "grad_norm": 1.035330873394029,
+      "learning_rate": 9.255161841158968e-06,
+      "loss": 1.6013,
+      "step": 4032
+    },
+    {
+      "epoch": 0.5381638644248732,
+      "grad_norm": 1.092318326623765,
+      "learning_rate": 9.250852011271703e-06,
+      "loss": 1.6165,
+      "step": 4033
+    },
+    {
+      "epoch": 0.5382973045102749,
+      "grad_norm": 1.018361237141597,
+      "learning_rate": 9.24654232131675e-06,
+      "loss": 1.5758,
+      "step": 4034
+    },
+    {
+      "epoch": 0.5384307445956765,
+      "grad_norm": 0.9345797395155986,
+      "learning_rate": 9.242232772099116e-06,
+      "loss": 1.5766,
+      "step": 4035
+    },
+    {
+      "epoch": 0.5385641846810783,
+      "grad_norm": 0.9679231135179047,
+      "learning_rate": 9.237923364423772e-06,
+      "loss": 1.5835,
+      "step": 4036
+    },
+    {
+      "epoch": 0.5386976247664799,
+      "grad_norm": 1.0031021300962772,
+      "learning_rate": 9.233614099095668e-06,
+      "loss": 1.6144,
+      "step": 4037
+    },
+    {
+      "epoch": 0.5388310648518815,
+      "grad_norm": 0.9915425586784125,
+      "learning_rate": 9.229304976919721e-06,
+      "loss": 1.6167,
+      "step": 4038
+    },
+    {
+      "epoch": 0.5389645049372832,
+      "grad_norm": 0.9956070431123422,
+      "learning_rate": 9.224995998700832e-06,
+      "loss": 1.6112,
+      "step": 4039
+    },
+    {
+      "epoch": 0.5390979450226848,
+      "grad_norm": 0.9175317898321539,
+      "learning_rate": 9.220687165243867e-06,
+      "loss": 1.5324,
+      "step": 4040
+    },
+    {
+      "epoch": 0.5392313851080864,
+      "grad_norm": 0.9964567601709668,
+      "learning_rate": 9.216378477353664e-06,
+      "loss": 1.5438,
+      "step": 4041
+    },
+    {
+      "epoch": 0.5393648251934882,
+      "grad_norm": 1.2134149238548315,
+      "learning_rate": 9.212069935835044e-06,
+      "loss": 1.5954,
+      "step": 4042
+    },
+    {
+      "epoch": 0.5394982652788898,
+      "grad_norm": 0.9893376951501015,
+      "learning_rate": 9.207761541492788e-06,
+      "loss": 1.635,
+      "step": 4043
+    },
+    {
+      "epoch": 0.5396317053642914,
+      "grad_norm": 0.9752793339516473,
+      "learning_rate": 9.203453295131657e-06,
+      "loss": 1.5758,
+      "step": 4044
+    },
+    {
+      "epoch": 0.5397651454496931,
+      "grad_norm": 0.9849050003080004,
+      "learning_rate": 9.199145197556375e-06,
+      "loss": 1.59,
+      "step": 4045
+    },
+    {
+      "epoch": 0.5398985855350947,
+      "grad_norm": 1.3161223332489436,
+      "learning_rate": 9.194837249571658e-06,
+      "loss": 1.6444,
+      "step": 4046
+    },
+    {
+      "epoch": 0.5400320256204963,
+      "grad_norm": 1.0146112609602544,
+      "learning_rate": 9.190529451982173e-06,
+      "loss": 1.615,
+      "step": 4047
+    },
+    {
+      "epoch": 0.5401654657058981,
+      "grad_norm": 0.9544932036464441,
+      "learning_rate": 9.18622180559257e-06,
+      "loss": 1.6292,
+      "step": 4048
+    },
+    {
+      "epoch": 0.5402989057912997,
+      "grad_norm": 0.9541877435381779,
+      "learning_rate": 9.18191431120747e-06,
+      "loss": 1.5881,
+      "step": 4049
+    },
+    {
+      "epoch": 0.5404323458767014,
+      "grad_norm": 0.9729000021252879,
+      "learning_rate": 9.177606969631464e-06,
+      "loss": 1.577,
+      "step": 4050
+    },
+    {
+      "epoch": 0.540565785962103,
+      "grad_norm": 0.9802692030861584,
+      "learning_rate": 9.173299781669112e-06,
+      "loss": 1.5656,
+      "step": 4051
+    },
+    {
+      "epoch": 0.5406992260475046,
+      "grad_norm": 0.9490128443326544,
+      "learning_rate": 9.168992748124943e-06,
+      "loss": 1.5705,
+      "step": 4052
+    },
+    {
+      "epoch": 0.5408326661329064,
+      "grad_norm": 0.9644978819486809,
+      "learning_rate": 9.164685869803474e-06,
+      "loss": 1.5745,
+      "step": 4053
+    },
+    {
+      "epoch": 0.540966106218308,
+      "grad_norm": 0.9346805372231108,
+      "learning_rate": 9.160379147509172e-06,
+      "loss": 1.5811,
+      "step": 4054
+    },
+    {
+      "epoch": 0.5410995463037096,
+      "grad_norm": 1.0592259248945302,
+      "learning_rate": 9.156072582046483e-06,
+      "loss": 1.5821,
+      "step": 4055
+    },
+    {
+      "epoch": 0.5412329863891113,
+      "grad_norm": 0.9912214570473821,
+      "learning_rate": 9.15176617421983e-06,
+      "loss": 1.61,
+      "step": 4056
+    },
+    {
+      "epoch": 0.5413664264745129,
+      "grad_norm": 0.9898284421098535,
+      "learning_rate": 9.147459924833599e-06,
+      "loss": 1.574,
+      "step": 4057
+    },
+    {
+      "epoch": 0.5414998665599146,
+      "grad_norm": 0.9503201276694209,
+      "learning_rate": 9.143153834692147e-06,
+      "loss": 1.5615,
+      "step": 4058
+    },
+    {
+      "epoch": 0.5416333066453163,
+      "grad_norm": 0.933992350962357,
+      "learning_rate": 9.138847904599802e-06,
+      "loss": 1.5795,
+      "step": 4059
+    },
+    {
+      "epoch": 0.5417667467307179,
+      "grad_norm": 1.188754739203571,
+      "learning_rate": 9.134542135360863e-06,
+      "loss": 1.5382,
+      "step": 4060
+    },
+    {
+      "epoch": 0.5419001868161195,
+      "grad_norm": 0.9697813847683896,
+      "learning_rate": 9.130236527779602e-06,
+      "loss": 1.5861,
+      "step": 4061
+    },
+    {
+      "epoch": 0.5420336269015212,
+      "grad_norm": 0.9800137082741205,
+      "learning_rate": 9.125931082660255e-06,
+      "loss": 1.5988,
+      "step": 4062
+    },
+    {
+      "epoch": 0.5421670669869229,
+      "grad_norm": 1.0320301308567623,
+      "learning_rate": 9.121625800807028e-06,
+      "loss": 1.5667,
+      "step": 4063
+    },
+    {
+      "epoch": 0.5423005070723246,
+      "grad_norm": 0.9333833148734103,
+      "learning_rate": 9.117320683024104e-06,
+      "loss": 1.5643,
+      "step": 4064
+    },
+    {
+      "epoch": 0.5424339471577262,
+      "grad_norm": 0.9278004209324023,
+      "learning_rate": 9.113015730115626e-06,
+      "loss": 1.5886,
+      "step": 4065
+    },
+    {
+      "epoch": 0.5425673872431278,
+      "grad_norm": 0.9639727825463563,
+      "learning_rate": 9.108710942885705e-06,
+      "loss": 1.5911,
+      "step": 4066
+    },
+    {
+      "epoch": 0.5427008273285295,
+      "grad_norm": 0.9307792098772837,
+      "learning_rate": 9.104406322138438e-06,
+      "loss": 1.5635,
+      "step": 4067
+    },
+    {
+      "epoch": 0.5428342674139311,
+      "grad_norm": 1.0667447561969865,
+      "learning_rate": 9.100101868677872e-06,
+      "loss": 1.5563,
+      "step": 4068
+    },
+    {
+      "epoch": 0.5429677074993328,
+      "grad_norm": 0.912096392390571,
+      "learning_rate": 9.095797583308033e-06,
+      "loss": 1.566,
+      "step": 4069
+    },
+    {
+      "epoch": 0.5431011475847345,
+      "grad_norm": 1.3022530608595129,
+      "learning_rate": 9.091493466832905e-06,
+      "loss": 1.5883,
+      "step": 4070
+    },
+    {
+      "epoch": 0.5432345876701361,
+      "grad_norm": 0.9697016564423682,
+      "learning_rate": 9.087189520056459e-06,
+      "loss": 1.6272,
+      "step": 4071
+    },
+    {
+      "epoch": 0.5433680277555377,
+      "grad_norm": 1.1713917719657274,
+      "learning_rate": 9.082885743782617e-06,
+      "loss": 1.539,
+      "step": 4072
+    },
+    {
+      "epoch": 0.5435014678409394,
+      "grad_norm": 0.9945992018009926,
+      "learning_rate": 9.078582138815272e-06,
+      "loss": 1.5641,
+      "step": 4073
+    },
+    {
+      "epoch": 0.5436349079263411,
+      "grad_norm": 1.09987126567883,
+      "learning_rate": 9.074278705958297e-06,
+      "loss": 1.5987,
+      "step": 4074
+    },
+    {
+      "epoch": 0.5437683480117427,
+      "grad_norm": 0.9528284537379281,
+      "learning_rate": 9.069975446015522e-06,
+      "loss": 1.5384,
+      "step": 4075
+    },
+    {
+      "epoch": 0.5439017880971444,
+      "grad_norm": 5.663201568713656,
+      "learning_rate": 9.065672359790743e-06,
+      "loss": 1.4918,
+      "step": 4076
+    },
+    {
+      "epoch": 0.544035228182546,
+      "grad_norm": 0.9679094263779862,
+      "learning_rate": 9.061369448087725e-06,
+      "loss": 1.5623,
+      "step": 4077
+    },
+    {
+      "epoch": 0.5441686682679477,
+      "grad_norm": 2.567953353801192,
+      "learning_rate": 9.057066711710212e-06,
+      "loss": 1.5437,
+      "step": 4078
+    },
+    {
+      "epoch": 0.5443021083533494,
+      "grad_norm": 1.1402343926159393,
+      "learning_rate": 9.052764151461902e-06,
+      "loss": 1.6053,
+      "step": 4079
+    },
+    {
+      "epoch": 0.544435548438751,
+      "grad_norm": 1.0524535435440139,
+      "learning_rate": 9.048461768146461e-06,
+      "loss": 1.678,
+      "step": 4080
+    },
+    {
+      "epoch": 0.5445689885241527,
+      "grad_norm": 0.9636960807728361,
+      "learning_rate": 9.044159562567532e-06,
+      "loss": 1.6102,
+      "step": 4081
+    },
+    {
+      "epoch": 0.5447024286095543,
+      "grad_norm": 1.0002275536656045,
+      "learning_rate": 9.03985753552871e-06,
+      "loss": 1.6551,
+      "step": 4082
+    },
+    {
+      "epoch": 0.5448358686949559,
+      "grad_norm": 0.993667451230517,
+      "learning_rate": 9.035555687833571e-06,
+      "loss": 1.5973,
+      "step": 4083
+    },
+    {
+      "epoch": 0.5449693087803577,
+      "grad_norm": 1.4929672107006777,
+      "learning_rate": 9.031254020285648e-06,
+      "loss": 1.5943,
+      "step": 4084
+    },
+    {
+      "epoch": 0.5451027488657593,
+      "grad_norm": 0.9664973212133721,
+      "learning_rate": 9.026952533688445e-06,
+      "loss": 1.5989,
+      "step": 4085
+    },
+    {
+      "epoch": 0.5452361889511609,
+      "grad_norm": 0.9850092658759886,
+      "learning_rate": 9.022651228845431e-06,
+      "loss": 1.6171,
+      "step": 4086
+    },
+    {
+      "epoch": 0.5453696290365626,
+      "grad_norm": 0.971030931040022,
+      "learning_rate": 9.018350106560036e-06,
+      "loss": 1.5712,
+      "step": 4087
+    },
+    {
+      "epoch": 0.5455030691219642,
+      "grad_norm": 1.2898471611348323,
+      "learning_rate": 9.014049167635668e-06,
+      "loss": 1.5786,
+      "step": 4088
+    },
+    {
+      "epoch": 0.5456365092073658,
+      "grad_norm": 0.9722886044790409,
+      "learning_rate": 9.00974841287569e-06,
+      "loss": 1.5522,
+      "step": 4089
+    },
+    {
+      "epoch": 0.5457699492927676,
+      "grad_norm": 1.274819296685493,
+      "learning_rate": 9.005447843083436e-06,
+      "loss": 1.6179,
+      "step": 4090
+    },
+    {
+      "epoch": 0.5459033893781692,
+      "grad_norm": 0.9802333166916215,
+      "learning_rate": 9.001147459062196e-06,
+      "loss": 1.5479,
+      "step": 4091
+    },
+    {
+      "epoch": 0.5460368294635709,
+      "grad_norm": 0.9612343249635548,
+      "learning_rate": 8.996847261615242e-06,
+      "loss": 1.5777,
+      "step": 4092
+    },
+    {
+      "epoch": 0.5461702695489725,
+      "grad_norm": 1.0934912868937452,
+      "learning_rate": 8.992547251545798e-06,
+      "loss": 1.5733,
+      "step": 4093
+    },
+    {
+      "epoch": 0.5463037096343741,
+      "grad_norm": 1.051460553342782,
+      "learning_rate": 8.988247429657058e-06,
+      "loss": 1.5806,
+      "step": 4094
+    },
+    {
+      "epoch": 0.5464371497197759,
+      "grad_norm": 1.0365839786798572,
+      "learning_rate": 8.983947796752174e-06,
+      "loss": 1.5982,
+      "step": 4095
+    },
+    {
+      "epoch": 0.5465705898051775,
+      "grad_norm": 1.1028438221141459,
+      "learning_rate": 8.979648353634278e-06,
+      "loss": 1.6002,
+      "step": 4096
+    },
+    {
+      "epoch": 0.5467040298905791,
+      "grad_norm": 0.9267409999435865,
+      "learning_rate": 8.975349101106451e-06,
+      "loss": 1.5865,
+      "step": 4097
+    },
+    {
+      "epoch": 0.5468374699759808,
+      "grad_norm": 0.9232565563556838,
+      "learning_rate": 8.971050039971743e-06,
+      "loss": 1.5637,
+      "step": 4098
+    },
+    {
+      "epoch": 0.5469709100613824,
+      "grad_norm": 1.0548309109185285,
+      "learning_rate": 8.966751171033177e-06,
+      "loss": 1.5598,
+      "step": 4099
+    },
+    {
+      "epoch": 0.547104350146784,
+      "grad_norm": 0.9446200469309572,
+      "learning_rate": 8.962452495093729e-06,
+      "loss": 1.5853,
+      "step": 4100
+    },
+    {
+      "epoch": 0.5472377902321858,
+      "grad_norm": 0.9503336907790122,
+      "learning_rate": 8.95815401295634e-06,
+      "loss": 1.5885,
+      "step": 4101
+    },
+    {
+      "epoch": 0.5473712303175874,
+      "grad_norm": 1.1631702714115804,
+      "learning_rate": 8.953855725423918e-06,
+      "loss": 1.5864,
+      "step": 4102
+    },
+    {
+      "epoch": 0.5475046704029891,
+      "grad_norm": 0.928539952699467,
+      "learning_rate": 8.949557633299335e-06,
+      "loss": 1.5443,
+      "step": 4103
+    },
+    {
+      "epoch": 0.5476381104883907,
+      "grad_norm": 0.9789123315297109,
+      "learning_rate": 8.945259737385432e-06,
+      "loss": 1.6024,
+      "step": 4104
+    },
+    {
+      "epoch": 0.5477715505737923,
+      "grad_norm": 0.9695913713026796,
+      "learning_rate": 8.940962038484997e-06,
+      "loss": 1.5721,
+      "step": 4105
+    },
+    {
+      "epoch": 0.5479049906591941,
+      "grad_norm": 1.2560495210786051,
+      "learning_rate": 8.936664537400797e-06,
+      "loss": 1.5756,
+      "step": 4106
+    },
+    {
+      "epoch": 0.5480384307445957,
+      "grad_norm": 1.0698213851452432,
+      "learning_rate": 8.932367234935554e-06,
+      "loss": 1.5682,
+      "step": 4107
+    },
+    {
+      "epoch": 0.5481718708299973,
+      "grad_norm": 1.0468767253083986,
+      "learning_rate": 8.928070131891959e-06,
+      "loss": 1.5945,
+      "step": 4108
+    },
+    {
+      "epoch": 0.548305310915399,
+      "grad_norm": 0.9589490681881839,
+      "learning_rate": 8.923773229072653e-06,
+      "loss": 1.6103,
+      "step": 4109
+    },
+    {
+      "epoch": 0.5484387510008006,
+      "grad_norm": 0.9736679031008811,
+      "learning_rate": 8.919476527280257e-06,
+      "loss": 1.543,
+      "step": 4110
+    },
+    {
+      "epoch": 0.5485721910862023,
+      "grad_norm": 0.9528406682287175,
+      "learning_rate": 8.915180027317345e-06,
+      "loss": 1.58,
+      "step": 4111
+    },
+    {
+      "epoch": 0.548705631171604,
+      "grad_norm": 1.0097173511452968,
+      "learning_rate": 8.91088372998645e-06,
+      "loss": 1.616,
+      "step": 4112
+    },
+    {
+      "epoch": 0.5488390712570056,
+      "grad_norm": 0.9881054764092163,
+      "learning_rate": 8.906587636090078e-06,
+      "loss": 1.5138,
+      "step": 4113
+    },
+    {
+      "epoch": 0.5489725113424072,
+      "grad_norm": 0.9333735783197412,
+      "learning_rate": 8.902291746430686e-06,
+      "loss": 1.5852,
+      "step": 4114
+    },
+    {
+      "epoch": 0.5491059514278089,
+      "grad_norm": 0.9489550127540991,
+      "learning_rate": 8.897996061810697e-06,
+      "loss": 1.563,
+      "step": 4115
+    },
+    {
+      "epoch": 0.5492393915132106,
+      "grad_norm": 0.962585687561712,
+      "learning_rate": 8.893700583032494e-06,
+      "loss": 1.5634,
+      "step": 4116
+    },
+    {
+      "epoch": 0.5493728315986123,
+      "grad_norm": 1.0345096599792367,
+      "learning_rate": 8.88940531089843e-06,
+      "loss": 1.5799,
+      "step": 4117
+    },
+    {
+      "epoch": 0.5495062716840139,
+      "grad_norm": 0.9472708442632688,
+      "learning_rate": 8.885110246210809e-06,
+      "loss": 1.5588,
+      "step": 4118
+    },
+    {
+      "epoch": 0.5496397117694155,
+      "grad_norm": 0.9677807823462035,
+      "learning_rate": 8.880815389771896e-06,
+      "loss": 1.5661,
+      "step": 4119
+    },
+    {
+      "epoch": 0.5497731518548172,
+      "grad_norm": 0.9496679955804133,
+      "learning_rate": 8.876520742383931e-06,
+      "loss": 1.5943,
+      "step": 4120
+    },
+    {
+      "epoch": 0.5499065919402188,
+      "grad_norm": 0.9783924610253193,
+      "learning_rate": 8.8722263048491e-06,
+      "loss": 1.566,
+      "step": 4121
+    },
+    {
+      "epoch": 0.5500400320256205,
+      "grad_norm": 1.1878684772471404,
+      "learning_rate": 8.867932077969555e-06,
+      "loss": 1.6396,
+      "step": 4122
+    },
+    {
+      "epoch": 0.5501734721110222,
+      "grad_norm": 0.9611842305243232,
+      "learning_rate": 8.863638062547406e-06,
+      "loss": 1.5659,
+      "step": 4123
+    },
+    {
+      "epoch": 0.5503069121964238,
+      "grad_norm": 0.9306514165663081,
+      "learning_rate": 8.85934425938473e-06,
+      "loss": 1.606,
+      "step": 4124
+    },
+    {
+      "epoch": 0.5504403522818254,
+      "grad_norm": 0.9465699027935999,
+      "learning_rate": 8.85505066928356e-06,
+      "loss": 1.6102,
+      "step": 4125
+    },
+    {
+      "epoch": 0.5505737923672271,
+      "grad_norm": 0.9749768439700673,
+      "learning_rate": 8.85075729304589e-06,
+      "loss": 1.5877,
+      "step": 4126
+    },
+    {
+      "epoch": 0.5507072324526288,
+      "grad_norm": 0.9815904995653956,
+      "learning_rate": 8.846464131473671e-06,
+      "loss": 1.538,
+      "step": 4127
+    },
+    {
+      "epoch": 0.5508406725380304,
+      "grad_norm": 0.9804198563493697,
+      "learning_rate": 8.842171185368822e-06,
+      "loss": 1.5889,
+      "step": 4128
+    },
+    {
+      "epoch": 0.5509741126234321,
+      "grad_norm": 1.0046191555459012,
+      "learning_rate": 8.837878455533213e-06,
+      "loss": 1.563,
+      "step": 4129
+    },
+    {
+      "epoch": 0.5511075527088337,
+      "grad_norm": 0.9815357324359852,
+      "learning_rate": 8.833585942768672e-06,
+      "loss": 1.605,
+      "step": 4130
+    },
+    {
+      "epoch": 0.5512409927942354,
+      "grad_norm": 1.0030901091320008,
+      "learning_rate": 8.829293647877002e-06,
+      "loss": 1.5989,
+      "step": 4131
+    },
+    {
+      "epoch": 0.551374432879637,
+      "grad_norm": 1.0070184181295139,
+      "learning_rate": 8.825001571659948e-06,
+      "loss": 1.6026,
+      "step": 4132
+    },
+    {
+      "epoch": 0.5515078729650387,
+      "grad_norm": 0.9393916243147451,
+      "learning_rate": 8.820709714919223e-06,
+      "loss": 1.5979,
+      "step": 4133
+    },
+    {
+      "epoch": 0.5516413130504404,
+      "grad_norm": 0.9663628069756637,
+      "learning_rate": 8.816418078456493e-06,
+      "loss": 1.6353,
+      "step": 4134
+    },
+    {
+      "epoch": 0.551774753135842,
+      "grad_norm": 1.0337970146828375,
+      "learning_rate": 8.812126663073394e-06,
+      "loss": 1.591,
+      "step": 4135
+    },
+    {
+      "epoch": 0.5519081932212436,
+      "grad_norm": 0.944300804007319,
+      "learning_rate": 8.807835469571511e-06,
+      "loss": 1.5938,
+      "step": 4136
+    },
+    {
+      "epoch": 0.5520416333066454,
+      "grad_norm": 0.9431341497957093,
+      "learning_rate": 8.803544498752383e-06,
+      "loss": 1.5962,
+      "step": 4137
+    },
+    {
+      "epoch": 0.552175073392047,
+      "grad_norm": 0.932632959879075,
+      "learning_rate": 8.799253751417526e-06,
+      "loss": 1.5937,
+      "step": 4138
+    },
+    {
+      "epoch": 0.5523085134774486,
+      "grad_norm": 0.9584857084573524,
+      "learning_rate": 8.794963228368397e-06,
+      "loss": 1.5368,
+      "step": 4139
+    },
+    {
+      "epoch": 0.5524419535628503,
+      "grad_norm": 0.959115772237164,
+      "learning_rate": 8.790672930406416e-06,
+      "loss": 1.5858,
+      "step": 4140
+    },
+    {
+      "epoch": 0.5525753936482519,
+      "grad_norm": 1.1797548555444335,
+      "learning_rate": 8.78638285833296e-06,
+      "loss": 1.5726,
+      "step": 4141
+    },
+    {
+      "epoch": 0.5527088337336535,
+      "grad_norm": 0.9424673897338993,
+      "learning_rate": 8.782093012949373e-06,
+      "loss": 1.5949,
+      "step": 4142
+    },
+    {
+      "epoch": 0.5528422738190553,
+      "grad_norm": 0.9713375108175255,
+      "learning_rate": 8.777803395056945e-06,
+      "loss": 1.6092,
+      "step": 4143
+    },
+    {
+      "epoch": 0.5529757139044569,
+      "grad_norm": 0.9774551665234783,
+      "learning_rate": 8.773514005456923e-06,
+      "loss": 1.5954,
+      "step": 4144
+    },
+    {
+      "epoch": 0.5531091539898586,
+      "grad_norm": 1.0793312067897565,
+      "learning_rate": 8.769224844950525e-06,
+      "loss": 1.5821,
+      "step": 4145
+    },
+    {
+      "epoch": 0.5532425940752602,
+      "grad_norm": 1.8746469240528907,
+      "learning_rate": 8.76493591433891e-06,
+      "loss": 1.5594,
+      "step": 4146
+    },
+    {
+      "epoch": 0.5533760341606618,
+      "grad_norm": 0.990387748915661,
+      "learning_rate": 8.760647214423206e-06,
+      "loss": 1.5758,
+      "step": 4147
+    },
+    {
+      "epoch": 0.5535094742460636,
+      "grad_norm": 0.9682888207800222,
+      "learning_rate": 8.756358746004492e-06,
+      "loss": 1.574,
+      "step": 4148
+    },
+    {
+      "epoch": 0.5536429143314652,
+      "grad_norm": 1.10384040970137,
+      "learning_rate": 8.752070509883805e-06,
+      "loss": 1.5835,
+      "step": 4149
+    },
+    {
+      "epoch": 0.5537763544168668,
+      "grad_norm": 0.9794947979779128,
+      "learning_rate": 8.747782506862139e-06,
+      "loss": 1.5744,
+      "step": 4150
+    },
+    {
+      "epoch": 0.5539097945022685,
+      "grad_norm": 1.0987054858543068,
+      "learning_rate": 8.74349473774044e-06,
+      "loss": 1.6253,
+      "step": 4151
+    },
+    {
+      "epoch": 0.5540432345876701,
+      "grad_norm": 0.9490644187289331,
+      "learning_rate": 8.739207203319622e-06,
+      "loss": 1.5796,
+      "step": 4152
+    },
+    {
+      "epoch": 0.5541766746730717,
+      "grad_norm": 1.0378405608665844,
+      "learning_rate": 8.734919904400544e-06,
+      "loss": 1.5961,
+      "step": 4153
+    },
+    {
+      "epoch": 0.5543101147584735,
+      "grad_norm": 0.9322492359210561,
+      "learning_rate": 8.730632841784023e-06,
+      "loss": 1.5715,
+      "step": 4154
+    },
+    {
+      "epoch": 0.5544435548438751,
+      "grad_norm": 0.9781596956125678,
+      "learning_rate": 8.726346016270833e-06,
+      "loss": 1.5597,
+      "step": 4155
+    },
+    {
+      "epoch": 0.5545769949292767,
+      "grad_norm": 0.9527266527507525,
+      "learning_rate": 8.72205942866171e-06,
+      "loss": 1.6075,
+      "step": 4156
+    },
+    {
+      "epoch": 0.5547104350146784,
+      "grad_norm": 0.9106864869517425,
+      "learning_rate": 8.717773079757333e-06,
+      "loss": 1.5677,
+      "step": 4157
+    },
+    {
+      "epoch": 0.55484387510008,
+      "grad_norm": 0.9851888382485386,
+      "learning_rate": 8.713486970358348e-06,
+      "loss": 1.5512,
+      "step": 4158
+    },
+    {
+      "epoch": 0.5549773151854818,
+      "grad_norm": 0.986066626003795,
+      "learning_rate": 8.709201101265346e-06,
+      "loss": 1.6065,
+      "step": 4159
+    },
+    {
+      "epoch": 0.5551107552708834,
+      "grad_norm": 0.9632922780457756,
+      "learning_rate": 8.704915473278885e-06,
+      "loss": 1.5635,
+      "step": 4160
+    },
+    {
+      "epoch": 0.555244195356285,
+      "grad_norm": 0.9616053211021358,
+      "learning_rate": 8.700630087199468e-06,
+      "loss": 1.5781,
+      "step": 4161
+    },
+    {
+      "epoch": 0.5553776354416867,
+      "grad_norm": 0.9536548628997336,
+      "learning_rate": 8.696344943827553e-06,
+      "loss": 1.5276,
+      "step": 4162
+    },
+    {
+      "epoch": 0.5555110755270883,
+      "grad_norm": 1.0768239441946357,
+      "learning_rate": 8.692060043963565e-06,
+      "loss": 1.6121,
+      "step": 4163
+    },
+    {
+      "epoch": 0.55564451561249,
+      "grad_norm": 1.214532797811325,
+      "learning_rate": 8.68777538840787e-06,
+      "loss": 1.6292,
+      "step": 4164
+    },
+    {
+      "epoch": 0.5557779556978917,
+      "grad_norm": 0.9333872970846957,
+      "learning_rate": 8.68349097796079e-06,
+      "loss": 1.5566,
+      "step": 4165
+    },
+    {
+      "epoch": 0.5559113957832933,
+      "grad_norm": 0.9800519597605887,
+      "learning_rate": 8.679206813422605e-06,
+      "loss": 1.5816,
+      "step": 4166
+    },
+    {
+      "epoch": 0.5560448358686949,
+      "grad_norm": 1.0132442396753862,
+      "learning_rate": 8.67492289559355e-06,
+      "loss": 1.5314,
+      "step": 4167
+    },
+    {
+      "epoch": 0.5561782759540966,
+      "grad_norm": 0.961576475954957,
+      "learning_rate": 8.670639225273813e-06,
+      "loss": 1.6011,
+      "step": 4168
+    },
+    {
+      "epoch": 0.5563117160394982,
+      "grad_norm": 0.9788757658928525,
+      "learning_rate": 8.66635580326353e-06,
+      "loss": 1.5529,
+      "step": 4169
+    },
+    {
+      "epoch": 0.5564451561248999,
+      "grad_norm": 1.0085557507010017,
+      "learning_rate": 8.662072630362803e-06,
+      "loss": 1.5839,
+      "step": 4170
+    },
+    {
+      "epoch": 0.5565785962103016,
+      "grad_norm": 1.0390389689513875,
+      "learning_rate": 8.657789707371675e-06,
+      "loss": 1.6039,
+      "step": 4171
+    },
+    {
+      "epoch": 0.5567120362957032,
+      "grad_norm": 8.058031207442108,
+      "learning_rate": 8.65350703509015e-06,
+      "loss": 1.6436,
+      "step": 4172
+    },
+    {
+      "epoch": 0.5568454763811049,
+      "grad_norm": 0.9705636062848059,
+      "learning_rate": 8.649224614318174e-06,
+      "loss": 1.5705,
+      "step": 4173
+    },
+    {
+      "epoch": 0.5569789164665065,
+      "grad_norm": 0.9395606140307743,
+      "learning_rate": 8.644942445855666e-06,
+      "loss": 1.619,
+      "step": 4174
+    },
+    {
+      "epoch": 0.5571123565519082,
+      "grad_norm": 0.9746333291733118,
+      "learning_rate": 8.64066053050248e-06,
+      "loss": 1.5455,
+      "step": 4175
+    },
+    {
+      "epoch": 0.5572457966373099,
+      "grad_norm": 0.9673212477085391,
+      "learning_rate": 8.636378869058427e-06,
+      "loss": 1.5958,
+      "step": 4176
+    },
+    {
+      "epoch": 0.5573792367227115,
+      "grad_norm": 0.9991053615117632,
+      "learning_rate": 8.63209746232328e-06,
+      "loss": 1.6009,
+      "step": 4177
+    },
+    {
+      "epoch": 0.5575126768081131,
+      "grad_norm": 0.9645490800883675,
+      "learning_rate": 8.627816311096753e-06,
+      "loss": 1.5588,
+      "step": 4178
+    },
+    {
+      "epoch": 0.5576461168935148,
+      "grad_norm": 1.0862567601365494,
+      "learning_rate": 8.623535416178516e-06,
+      "loss": 1.6125,
+      "step": 4179
+    },
+    {
+      "epoch": 0.5577795569789165,
+      "grad_norm": 0.9565579659399656,
+      "learning_rate": 8.619254778368187e-06,
+      "loss": 1.5513,
+      "step": 4180
+    },
+    {
+      "epoch": 0.5579129970643181,
+      "grad_norm": 1.0678592227935304,
+      "learning_rate": 8.61497439846535e-06,
+      "loss": 1.6321,
+      "step": 4181
+    },
+    {
+      "epoch": 0.5580464371497198,
+      "grad_norm": 0.96056585056449,
+      "learning_rate": 8.610694277269526e-06,
+      "loss": 1.6026,
+      "step": 4182
+    },
+    {
+      "epoch": 0.5581798772351214,
+      "grad_norm": 1.011056257284224,
+      "learning_rate": 8.60641441558019e-06,
+      "loss": 1.5451,
+      "step": 4183
+    },
+    {
+      "epoch": 0.5583133173205231,
+      "grad_norm": 0.9413247204545455,
+      "learning_rate": 8.602134814196776e-06,
+      "loss": 1.5566,
+      "step": 4184
+    },
+    {
+      "epoch": 0.5584467574059248,
+      "grad_norm": 1.0445816035815227,
+      "learning_rate": 8.597855473918664e-06,
+      "loss": 1.5626,
+      "step": 4185
+    },
+    {
+      "epoch": 0.5585801974913264,
+      "grad_norm": 0.9575339480339196,
+      "learning_rate": 8.593576395545187e-06,
+      "loss": 1.6168,
+      "step": 4186
+    },
+    {
+      "epoch": 0.5587136375767281,
+      "grad_norm": 0.9282496732846511,
+      "learning_rate": 8.589297579875624e-06,
+      "loss": 1.5874,
+      "step": 4187
+    },
+    {
+      "epoch": 0.5588470776621297,
+      "grad_norm": 0.9840758991935276,
+      "learning_rate": 8.585019027709211e-06,
+      "loss": 1.6467,
+      "step": 4188
+    },
+    {
+      "epoch": 0.5589805177475313,
+      "grad_norm": 1.0328945704121253,
+      "learning_rate": 8.580740739845138e-06,
+      "loss": 1.5509,
+      "step": 4189
+    },
+    {
+      "epoch": 0.559113957832933,
+      "grad_norm": 1.0478866163329408,
+      "learning_rate": 8.576462717082532e-06,
+      "loss": 1.5587,
+      "step": 4190
+    },
+    {
+      "epoch": 0.5592473979183347,
+      "grad_norm": 1.1273918515526908,
+      "learning_rate": 8.572184960220485e-06,
+      "loss": 1.6093,
+      "step": 4191
+    },
+    {
+      "epoch": 0.5593808380037363,
+      "grad_norm": 0.9626145010766713,
+      "learning_rate": 8.56790747005803e-06,
+      "loss": 1.5516,
+      "step": 4192
+    },
+    {
+      "epoch": 0.559514278089138,
+      "grad_norm": 2.3623694177009043,
+      "learning_rate": 8.563630247394157e-06,
+      "loss": 1.6187,
+      "step": 4193
+    },
+    {
+      "epoch": 0.5596477181745396,
+      "grad_norm": 1.056578864134312,
+      "learning_rate": 8.559353293027793e-06,
+      "loss": 1.5665,
+      "step": 4194
+    },
+    {
+      "epoch": 0.5597811582599412,
+      "grad_norm": 0.9458997799992117,
+      "learning_rate": 8.555076607757837e-06,
+      "loss": 1.5715,
+      "step": 4195
+    },
+    {
+      "epoch": 0.559914598345343,
+      "grad_norm": 0.9618299176959383,
+      "learning_rate": 8.55080019238312e-06,
+      "loss": 1.5743,
+      "step": 4196
+    },
+    {
+      "epoch": 0.5600480384307446,
+      "grad_norm": 0.9743576637342902,
+      "learning_rate": 8.546524047702428e-06,
+      "loss": 1.6132,
+      "step": 4197
+    },
+    {
+      "epoch": 0.5601814785161463,
+      "grad_norm": 0.954261768012623,
+      "learning_rate": 8.54224817451449e-06,
+      "loss": 1.5947,
+      "step": 4198
+    },
+    {
+      "epoch": 0.5603149186015479,
+      "grad_norm": 0.967928239432757,
+      "learning_rate": 8.537972573618e-06,
+      "loss": 1.5806,
+      "step": 4199
+    },
+    {
+      "epoch": 0.5604483586869495,
+      "grad_norm": 0.9509794485601105,
+      "learning_rate": 8.53369724581159e-06,
+      "loss": 1.5455,
+      "step": 4200
+    },
+    {
+      "epoch": 0.5605817987723513,
+      "grad_norm": 0.948976660250996,
+      "learning_rate": 8.529422191893832e-06,
+      "loss": 1.6024,
+      "step": 4201
+    },
+    {
+      "epoch": 0.5607152388577529,
+      "grad_norm": 0.9018937165977122,
+      "learning_rate": 8.525147412663272e-06,
+      "loss": 1.5771,
+      "step": 4202
+    },
+    {
+      "epoch": 0.5608486789431545,
+      "grad_norm": 0.9881572931305829,
+      "learning_rate": 8.520872908918382e-06,
+      "loss": 1.5898,
+      "step": 4203
+    },
+    {
+      "epoch": 0.5609821190285562,
+      "grad_norm": 1.124313244736089,
+      "learning_rate": 8.516598681457595e-06,
+      "loss": 1.6095,
+      "step": 4204
+    },
+    {
+      "epoch": 0.5611155591139578,
+      "grad_norm": 1.1089045566415832,
+      "learning_rate": 8.512324731079277e-06,
+      "loss": 1.5805,
+      "step": 4205
+    },
+    {
+      "epoch": 0.5612489991993594,
+      "grad_norm": 0.9797580686965339,
+      "learning_rate": 8.508051058581768e-06,
+      "loss": 1.5679,
+      "step": 4206
+    },
+    {
+      "epoch": 0.5613824392847612,
+      "grad_norm": 0.9722387996589369,
+      "learning_rate": 8.503777664763336e-06,
+      "loss": 1.5833,
+      "step": 4207
+    },
+    {
+      "epoch": 0.5615158793701628,
+      "grad_norm": 1.07631421997575,
+      "learning_rate": 8.499504550422195e-06,
+      "loss": 1.562,
+      "step": 4208
+    },
+    {
+      "epoch": 0.5616493194555644,
+      "grad_norm": 1.0147686908909404,
+      "learning_rate": 8.495231716356525e-06,
+      "loss": 1.5465,
+      "step": 4209
+    },
+    {
+      "epoch": 0.5617827595409661,
+      "grad_norm": 0.9427076649487832,
+      "learning_rate": 8.490959163364436e-06,
+      "loss": 1.5497,
+      "step": 4210
+    },
+    {
+      "epoch": 0.5619161996263677,
+      "grad_norm": 1.1230893036871608,
+      "learning_rate": 8.486686892243996e-06,
+      "loss": 1.5773,
+      "step": 4211
+    },
+    {
+      "epoch": 0.5620496397117695,
+      "grad_norm": 0.9289630374962218,
+      "learning_rate": 8.482414903793213e-06,
+      "loss": 1.6441,
+      "step": 4212
+    },
+    {
+      "epoch": 0.5621830797971711,
+      "grad_norm": 0.9520481957952125,
+      "learning_rate": 8.478143198810048e-06,
+      "loss": 1.6163,
+      "step": 4213
+    },
+    {
+      "epoch": 0.5623165198825727,
+      "grad_norm": 1.0545873114232873,
+      "learning_rate": 8.473871778092408e-06,
+      "loss": 1.5663,
+      "step": 4214
+    },
+    {
+      "epoch": 0.5624499599679744,
+      "grad_norm": 0.9532253546176354,
+      "learning_rate": 8.469600642438139e-06,
+      "loss": 1.5472,
+      "step": 4215
+    },
+    {
+      "epoch": 0.562583400053376,
+      "grad_norm": 0.9979955697123374,
+      "learning_rate": 8.465329792645051e-06,
+      "loss": 1.5256,
+      "step": 4216
+    },
+    {
+      "epoch": 0.5627168401387777,
+      "grad_norm": 0.9702416076361973,
+      "learning_rate": 8.461059229510886e-06,
+      "loss": 1.541,
+      "step": 4217
+    },
+    {
+      "epoch": 0.5628502802241794,
+      "grad_norm": 0.998163150298418,
+      "learning_rate": 8.456788953833333e-06,
+      "loss": 1.586,
+      "step": 4218
+    },
+    {
+      "epoch": 0.562983720309581,
+      "grad_norm": 0.9618357742711305,
+      "learning_rate": 8.45251896641003e-06,
+      "loss": 1.5574,
+      "step": 4219
+    },
+    {
+      "epoch": 0.5631171603949826,
+      "grad_norm": 0.9617959238619088,
+      "learning_rate": 8.448249268038569e-06,
+      "loss": 1.5801,
+      "step": 4220
+    },
+    {
+      "epoch": 0.5632506004803843,
+      "grad_norm": 0.9388770256513779,
+      "learning_rate": 8.443979859516478e-06,
+      "loss": 1.5987,
+      "step": 4221
+    },
+    {
+      "epoch": 0.563384040565786,
+      "grad_norm": 1.0984016692407286,
+      "learning_rate": 8.439710741641234e-06,
+      "loss": 1.6375,
+      "step": 4222
+    },
+    {
+      "epoch": 0.5635174806511876,
+      "grad_norm": 0.9883405451834961,
+      "learning_rate": 8.435441915210254e-06,
+      "loss": 1.5521,
+      "step": 4223
+    },
+    {
+      "epoch": 0.5636509207365893,
+      "grad_norm": 1.216557539530969,
+      "learning_rate": 8.431173381020915e-06,
+      "loss": 1.5352,
+      "step": 4224
+    },
+    {
+      "epoch": 0.5637843608219909,
+      "grad_norm": 0.9774137933177777,
+      "learning_rate": 8.426905139870528e-06,
+      "loss": 1.555,
+      "step": 4225
+    },
+    {
+      "epoch": 0.5639178009073926,
+      "grad_norm": 1.2860621275048014,
+      "learning_rate": 8.422637192556345e-06,
+      "loss": 1.5603,
+      "step": 4226
+    },
+    {
+      "epoch": 0.5640512409927942,
+      "grad_norm": 1.0444121247666944,
+      "learning_rate": 8.418369539875579e-06,
+      "loss": 1.6277,
+      "step": 4227
+    },
+    {
+      "epoch": 0.5641846810781959,
+      "grad_norm": 0.9276944621654523,
+      "learning_rate": 8.414102182625377e-06,
+      "loss": 1.6169,
+      "step": 4228
+    },
+    {
+      "epoch": 0.5643181211635976,
+      "grad_norm": 0.9785924417239961,
+      "learning_rate": 8.40983512160283e-06,
+      "loss": 1.5704,
+      "step": 4229
+    },
+    {
+      "epoch": 0.5644515612489992,
+      "grad_norm": 0.9784553941398924,
+      "learning_rate": 8.405568357604975e-06,
+      "loss": 1.5979,
+      "step": 4230
+    },
+    {
+      "epoch": 0.5645850013344008,
+      "grad_norm": 1.0150051857133884,
+      "learning_rate": 8.4013018914288e-06,
+      "loss": 1.5686,
+      "step": 4231
+    },
+    {
+      "epoch": 0.5647184414198025,
+      "grad_norm": 1.0133584620768177,
+      "learning_rate": 8.39703572387123e-06,
+      "loss": 1.5938,
+      "step": 4232
+    },
+    {
+      "epoch": 0.5648518815052042,
+      "grad_norm": 0.9864373077024665,
+      "learning_rate": 8.392769855729134e-06,
+      "loss": 1.6306,
+      "step": 4233
+    },
+    {
+      "epoch": 0.5649853215906058,
+      "grad_norm": 0.9740231720507322,
+      "learning_rate": 8.388504287799333e-06,
+      "loss": 1.6303,
+      "step": 4234
+    },
+    {
+      "epoch": 0.5651187616760075,
+      "grad_norm": 1.0375657601783386,
+      "learning_rate": 8.384239020878583e-06,
+      "loss": 1.6057,
+      "step": 4235
+    },
+    {
+      "epoch": 0.5652522017614091,
+      "grad_norm": 0.9612076808860169,
+      "learning_rate": 8.379974055763587e-06,
+      "loss": 1.5814,
+      "step": 4236
+    },
+    {
+      "epoch": 0.5653856418468107,
+      "grad_norm": 0.9641347727662392,
+      "learning_rate": 8.37570939325099e-06,
+      "loss": 1.5954,
+      "step": 4237
+    },
+    {
+      "epoch": 0.5655190819322125,
+      "grad_norm": 1.1246728358254432,
+      "learning_rate": 8.371445034137388e-06,
+      "loss": 1.5883,
+      "step": 4238
+    },
+    {
+      "epoch": 0.5656525220176141,
+      "grad_norm": 0.9658697763988295,
+      "learning_rate": 8.367180979219314e-06,
+      "loss": 1.5931,
+      "step": 4239
+    },
+    {
+      "epoch": 0.5657859621030158,
+      "grad_norm": 1.1394715701292135,
+      "learning_rate": 8.362917229293236e-06,
+      "loss": 1.5455,
+      "step": 4240
+    },
+    {
+      "epoch": 0.5659194021884174,
+      "grad_norm": 1.0026059848550364,
+      "learning_rate": 8.358653785155586e-06,
+      "loss": 1.5607,
+      "step": 4241
+    },
+    {
+      "epoch": 0.566052842273819,
+      "grad_norm": 0.9721001045323454,
+      "learning_rate": 8.354390647602721e-06,
+      "loss": 1.6145,
+      "step": 4242
+    },
+    {
+      "epoch": 0.5661862823592207,
+      "grad_norm": 0.9081483328670256,
+      "learning_rate": 8.350127817430949e-06,
+      "loss": 1.5278,
+      "step": 4243
+    },
+    {
+      "epoch": 0.5663197224446224,
+      "grad_norm": 1.0420600474407877,
+      "learning_rate": 8.34586529543651e-06,
+      "loss": 1.5287,
+      "step": 4244
+    },
+    {
+      "epoch": 0.566453162530024,
+      "grad_norm": 1.0148442225077963,
+      "learning_rate": 8.341603082415604e-06,
+      "loss": 1.5427,
+      "step": 4245
+    },
+    {
+      "epoch": 0.5665866026154257,
+      "grad_norm": 0.9893277215121336,
+      "learning_rate": 8.337341179164363e-06,
+      "loss": 1.6118,
+      "step": 4246
+    },
+    {
+      "epoch": 0.5667200427008273,
+      "grad_norm": 1.2148103117653153,
+      "learning_rate": 8.333079586478854e-06,
+      "loss": 1.582,
+      "step": 4247
+    },
+    {
+      "epoch": 0.5668534827862289,
+      "grad_norm": 0.9457405540546562,
+      "learning_rate": 8.328818305155105e-06,
+      "loss": 1.6044,
+      "step": 4248
+    },
+    {
+      "epoch": 0.5669869228716307,
+      "grad_norm": 0.990303038169507,
+      "learning_rate": 8.324557335989068e-06,
+      "loss": 1.5833,
+      "step": 4249
+    },
+    {
+      "epoch": 0.5671203629570323,
+      "grad_norm": 0.9371628277905556,
+      "learning_rate": 8.320296679776647e-06,
+      "loss": 1.5408,
+      "step": 4250
+    },
+    {
+      "epoch": 0.567253803042434,
+      "grad_norm": 0.989229363405912,
+      "learning_rate": 8.316036337313678e-06,
+      "loss": 1.5705,
+      "step": 4251
+    },
+    {
+      "epoch": 0.5673872431278356,
+      "grad_norm": 0.9163649237832772,
+      "learning_rate": 8.31177630939595e-06,
+      "loss": 1.577,
+      "step": 4252
+    },
+    {
+      "epoch": 0.5675206832132372,
+      "grad_norm": 0.9865296079255822,
+      "learning_rate": 8.307516596819188e-06,
+      "loss": 1.5335,
+      "step": 4253
+    },
+    {
+      "epoch": 0.567654123298639,
+      "grad_norm": 0.9327295789695722,
+      "learning_rate": 8.303257200379055e-06,
+      "loss": 1.6193,
+      "step": 4254
+    },
+    {
+      "epoch": 0.5677875633840406,
+      "grad_norm": 0.996608302154684,
+      "learning_rate": 8.298998120871159e-06,
+      "loss": 1.5586,
+      "step": 4255
+    },
+    {
+      "epoch": 0.5679210034694422,
+      "grad_norm": 1.0163058132224498,
+      "learning_rate": 8.294739359091048e-06,
+      "loss": 1.5789,
+      "step": 4256
+    },
+    {
+      "epoch": 0.5680544435548439,
+      "grad_norm": 0.9545226447240991,
+      "learning_rate": 8.29048091583421e-06,
+      "loss": 1.5417,
+      "step": 4257
+    },
+    {
+      "epoch": 0.5681878836402455,
+      "grad_norm": 0.9899206964927434,
+      "learning_rate": 8.286222791896068e-06,
+      "loss": 1.6199,
+      "step": 4258
+    },
+    {
+      "epoch": 0.5683213237256471,
+      "grad_norm": 1.0349824476929077,
+      "learning_rate": 8.281964988072001e-06,
+      "loss": 1.5948,
+      "step": 4259
+    },
+    {
+      "epoch": 0.5684547638110489,
+      "grad_norm": 0.9700116138822578,
+      "learning_rate": 8.277707505157313e-06,
+      "loss": 1.5548,
+      "step": 4260
+    },
+    {
+      "epoch": 0.5685882038964505,
+      "grad_norm": 0.9509425096012817,
+      "learning_rate": 8.273450343947255e-06,
+      "loss": 1.6276,
+      "step": 4261
+    },
+    {
+      "epoch": 0.5687216439818521,
+      "grad_norm": 1.2640017963359513,
+      "learning_rate": 8.269193505237013e-06,
+      "loss": 1.5947,
+      "step": 4262
+    },
+    {
+      "epoch": 0.5688550840672538,
+      "grad_norm": 0.9467480150131761,
+      "learning_rate": 8.26493698982172e-06,
+      "loss": 1.6069,
+      "step": 4263
+    },
+    {
+      "epoch": 0.5689885241526554,
+      "grad_norm": 0.9907646708465446,
+      "learning_rate": 8.260680798496444e-06,
+      "loss": 1.578,
+      "step": 4264
+    },
+    {
+      "epoch": 0.5691219642380572,
+      "grad_norm": 1.1061065525912817,
+      "learning_rate": 8.256424932056187e-06,
+      "loss": 1.579,
+      "step": 4265
+    },
+    {
+      "epoch": 0.5692554043234588,
+      "grad_norm": 1.325438533023951,
+      "learning_rate": 8.252169391295905e-06,
+      "loss": 1.5684,
+      "step": 4266
+    },
+    {
+      "epoch": 0.5693888444088604,
+      "grad_norm": 0.9261579791060407,
+      "learning_rate": 8.247914177010482e-06,
+      "loss": 1.5429,
+      "step": 4267
+    },
+    {
+      "epoch": 0.5695222844942621,
+      "grad_norm": 0.9433004236090312,
+      "learning_rate": 8.243659289994741e-06,
+      "loss": 1.5836,
+      "step": 4268
+    },
+    {
+      "epoch": 0.5696557245796637,
+      "grad_norm": 0.9493027808271886,
+      "learning_rate": 8.239404731043446e-06,
+      "loss": 1.5597,
+      "step": 4269
+    },
+    {
+      "epoch": 0.5697891646650654,
+      "grad_norm": 0.972218396115009,
+      "learning_rate": 8.235150500951304e-06,
+      "loss": 1.6362,
+      "step": 4270
+    },
+    {
+      "epoch": 0.5699226047504671,
+      "grad_norm": 0.9262069004868365,
+      "learning_rate": 8.230896600512956e-06,
+      "loss": 1.6098,
+      "step": 4271
+    },
+    {
+      "epoch": 0.5700560448358687,
+      "grad_norm": 1.0129980611981801,
+      "learning_rate": 8.226643030522979e-06,
+      "loss": 1.6419,
+      "step": 4272
+    },
+    {
+      "epoch": 0.5701894849212703,
+      "grad_norm": 0.9574543891512742,
+      "learning_rate": 8.222389791775895e-06,
+      "loss": 1.5302,
+      "step": 4273
+    },
+    {
+      "epoch": 0.570322925006672,
+      "grad_norm": 2.7284176094811077,
+      "learning_rate": 8.218136885066158e-06,
+      "loss": 1.6217,
+      "step": 4274
+    },
+    {
+      "epoch": 0.5704563650920736,
+      "grad_norm": 0.9819347041563397,
+      "learning_rate": 8.213884311188166e-06,
+      "loss": 1.5888,
+      "step": 4275
+    },
+    {
+      "epoch": 0.5705898051774753,
+      "grad_norm": 0.9895366406849194,
+      "learning_rate": 8.209632070936246e-06,
+      "loss": 1.5732,
+      "step": 4276
+    },
+    {
+      "epoch": 0.570723245262877,
+      "grad_norm": 0.9348464052451259,
+      "learning_rate": 8.205380165104673e-06,
+      "loss": 1.5906,
+      "step": 4277
+    },
+    {
+      "epoch": 0.5708566853482786,
+      "grad_norm": 0.9528816676643417,
+      "learning_rate": 8.201128594487653e-06,
+      "loss": 1.6113,
+      "step": 4278
+    },
+    {
+      "epoch": 0.5709901254336803,
+      "grad_norm": 0.9776229291280035,
+      "learning_rate": 8.196877359879327e-06,
+      "loss": 1.5856,
+      "step": 4279
+    },
+    {
+      "epoch": 0.5711235655190819,
+      "grad_norm": 0.9414417277905998,
+      "learning_rate": 8.192626462073782e-06,
+      "loss": 1.564,
+      "step": 4280
+    },
+    {
+      "epoch": 0.5712570056044836,
+      "grad_norm": 0.9808001957769379,
+      "learning_rate": 8.18837590186504e-06,
+      "loss": 1.5853,
+      "step": 4281
+    },
+    {
+      "epoch": 0.5713904456898853,
+      "grad_norm": 1.1219868857209958,
+      "learning_rate": 8.184125680047051e-06,
+      "loss": 1.5385,
+      "step": 4282
+    },
+    {
+      "epoch": 0.5715238857752869,
+      "grad_norm": 0.9792466209394581,
+      "learning_rate": 8.179875797413707e-06,
+      "loss": 1.6143,
+      "step": 4283
+    },
+    {
+      "epoch": 0.5716573258606885,
+      "grad_norm": 0.9655150327845816,
+      "learning_rate": 8.175626254758847e-06,
+      "loss": 1.5814,
+      "step": 4284
+    },
+    {
+      "epoch": 0.5717907659460902,
+      "grad_norm": 1.0677323984435168,
+      "learning_rate": 8.171377052876228e-06,
+      "loss": 1.5212,
+      "step": 4285
+    },
+    {
+      "epoch": 0.5719242060314919,
+      "grad_norm": 1.0478556969877482,
+      "learning_rate": 8.167128192559557e-06,
+      "loss": 1.5864,
+      "step": 4286
+    },
+    {
+      "epoch": 0.5720576461168935,
+      "grad_norm": 0.9806397349097133,
+      "learning_rate": 8.162879674602469e-06,
+      "loss": 1.5688,
+      "step": 4287
+    },
+    {
+      "epoch": 0.5721910862022952,
+      "grad_norm": 0.957753049608993,
+      "learning_rate": 8.158631499798545e-06,
+      "loss": 1.5769,
+      "step": 4288
+    },
+    {
+      "epoch": 0.5723245262876968,
+      "grad_norm": 0.9714614263607235,
+      "learning_rate": 8.154383668941293e-06,
+      "loss": 1.5687,
+      "step": 4289
+    },
+    {
+      "epoch": 0.5724579663730984,
+      "grad_norm": 1.1716096470089616,
+      "learning_rate": 8.150136182824152e-06,
+      "loss": 1.5396,
+      "step": 4290
+    },
+    {
+      "epoch": 0.5725914064585002,
+      "grad_norm": 0.992260967601937,
+      "learning_rate": 8.145889042240516e-06,
+      "loss": 1.5828,
+      "step": 4291
+    },
+    {
+      "epoch": 0.5727248465439018,
+      "grad_norm": 0.965708034327343,
+      "learning_rate": 8.141642247983696e-06,
+      "loss": 1.5623,
+      "step": 4292
+    },
+    {
+      "epoch": 0.5728582866293035,
+      "grad_norm": 0.9907802430036895,
+      "learning_rate": 8.137395800846948e-06,
+      "loss": 1.6284,
+      "step": 4293
+    },
+    {
+      "epoch": 0.5729917267147051,
+      "grad_norm": 0.9396948795159468,
+      "learning_rate": 8.133149701623454e-06,
+      "loss": 1.6403,
+      "step": 4294
+    },
+    {
+      "epoch": 0.5731251668001067,
+      "grad_norm": 0.9597661967095336,
+      "learning_rate": 8.12890395110634e-06,
+      "loss": 1.5808,
+      "step": 4295
+    },
+    {
+      "epoch": 0.5732586068855084,
+      "grad_norm": 9.3684749649107,
+      "learning_rate": 8.124658550088668e-06,
+      "loss": 1.6139,
+      "step": 4296
+    },
+    {
+      "epoch": 0.5733920469709101,
+      "grad_norm": 0.9286997041501741,
+      "learning_rate": 8.120413499363427e-06,
+      "loss": 1.5668,
+      "step": 4297
+    },
+    {
+      "epoch": 0.5735254870563117,
+      "grad_norm": 0.9796234552599323,
+      "learning_rate": 8.116168799723544e-06,
+      "loss": 1.5597,
+      "step": 4298
+    },
+    {
+      "epoch": 0.5736589271417134,
+      "grad_norm": 1.0267380425578665,
+      "learning_rate": 8.111924451961883e-06,
+      "loss": 1.6226,
+      "step": 4299
+    },
+    {
+      "epoch": 0.573792367227115,
+      "grad_norm": 1.125467638614621,
+      "learning_rate": 8.107680456871236e-06,
+      "loss": 1.6001,
+      "step": 4300
+    },
+    {
+      "epoch": 0.5739258073125166,
+      "grad_norm": 0.9428006871087928,
+      "learning_rate": 8.10343681524433e-06,
+      "loss": 1.6035,
+      "step": 4301
+    },
+    {
+      "epoch": 0.5740592473979184,
+      "grad_norm": 0.9693180970207167,
+      "learning_rate": 8.09919352787384e-06,
+      "loss": 1.5463,
+      "step": 4302
+    },
+    {
+      "epoch": 0.57419268748332,
+      "grad_norm": 1.0176972242986655,
+      "learning_rate": 8.094950595552356e-06,
+      "loss": 1.6065,
+      "step": 4303
+    },
+    {
+      "epoch": 0.5743261275687216,
+      "grad_norm": 1.1248089412402027,
+      "learning_rate": 8.09070801907241e-06,
+      "loss": 1.6226,
+      "step": 4304
+    },
+    {
+      "epoch": 0.5744595676541233,
+      "grad_norm": 0.9608200644959385,
+      "learning_rate": 8.086465799226468e-06,
+      "loss": 1.5211,
+      "step": 4305
+    },
+    {
+      "epoch": 0.5745930077395249,
+      "grad_norm": 0.9756758699322823,
+      "learning_rate": 8.08222393680693e-06,
+      "loss": 1.5682,
+      "step": 4306
+    },
+    {
+      "epoch": 0.5747264478249267,
+      "grad_norm": 1.0688838840416501,
+      "learning_rate": 8.077982432606125e-06,
+      "loss": 1.5725,
+      "step": 4307
+    },
+    {
+      "epoch": 0.5748598879103283,
+      "grad_norm": 1.1309638787833816,
+      "learning_rate": 8.073741287416317e-06,
+      "loss": 1.6059,
+      "step": 4308
+    },
+    {
+      "epoch": 0.5749933279957299,
+      "grad_norm": 1.0556322400323577,
+      "learning_rate": 8.069500502029705e-06,
+      "loss": 1.5937,
+      "step": 4309
+    },
+    {
+      "epoch": 0.5751267680811316,
+      "grad_norm": 0.9506475357460462,
+      "learning_rate": 8.065260077238423e-06,
+      "loss": 1.5802,
+      "step": 4310
+    },
+    {
+      "epoch": 0.5752602081665332,
+      "grad_norm": 1.077301912471358,
+      "learning_rate": 8.061020013834526e-06,
+      "loss": 1.594,
+      "step": 4311
+    },
+    {
+      "epoch": 0.5753936482519348,
+      "grad_norm": 0.9475246370591415,
+      "learning_rate": 8.056780312610019e-06,
+      "loss": 1.5518,
+      "step": 4312
+    },
+    {
+      "epoch": 0.5755270883373366,
+      "grad_norm": 1.1642637584365982,
+      "learning_rate": 8.052540974356825e-06,
+      "loss": 1.5499,
+      "step": 4313
+    },
+    {
+      "epoch": 0.5756605284227382,
+      "grad_norm": 0.962853058046866,
+      "learning_rate": 8.048301999866803e-06,
+      "loss": 1.625,
+      "step": 4314
+    },
+    {
+      "epoch": 0.5757939685081398,
+      "grad_norm": 0.951938089036065,
+      "learning_rate": 8.044063389931745e-06,
+      "loss": 1.6001,
+      "step": 4315
+    },
+    {
+      "epoch": 0.5759274085935415,
+      "grad_norm": 0.9850586486463077,
+      "learning_rate": 8.039825145343378e-06,
+      "loss": 1.5917,
+      "step": 4316
+    },
+    {
+      "epoch": 0.5760608486789431,
+      "grad_norm": 0.9996996255343955,
+      "learning_rate": 8.035587266893357e-06,
+      "loss": 1.59,
+      "step": 4317
+    },
+    {
+      "epoch": 0.5761942887643449,
+      "grad_norm": 0.9361786850612416,
+      "learning_rate": 8.031349755373269e-06,
+      "loss": 1.5217,
+      "step": 4318
+    },
+    {
+      "epoch": 0.5763277288497465,
+      "grad_norm": 0.9534574866115021,
+      "learning_rate": 8.02711261157463e-06,
+      "loss": 1.6341,
+      "step": 4319
+    },
+    {
+      "epoch": 0.5764611689351481,
+      "grad_norm": 0.9607054461212049,
+      "learning_rate": 8.022875836288896e-06,
+      "loss": 1.5407,
+      "step": 4320
+    },
+    {
+      "epoch": 0.5765946090205498,
+      "grad_norm": 1.035113379371202,
+      "learning_rate": 8.018639430307445e-06,
+      "loss": 1.546,
+      "step": 4321
+    },
+    {
+      "epoch": 0.5767280491059514,
+      "grad_norm": 0.9424964205669921,
+      "learning_rate": 8.014403394421585e-06,
+      "loss": 1.5562,
+      "step": 4322
+    },
+    {
+      "epoch": 0.576861489191353,
+      "grad_norm": 0.9423821513514596,
+      "learning_rate": 8.010167729422571e-06,
+      "loss": 1.5583,
+      "step": 4323
+    },
+    {
+      "epoch": 0.5769949292767548,
+      "grad_norm": 0.9340881718791657,
+      "learning_rate": 8.005932436101567e-06,
+      "loss": 1.5567,
+      "step": 4324
+    },
+    {
+      "epoch": 0.5771283693621564,
+      "grad_norm": 0.9680824040104852,
+      "learning_rate": 8.001697515249683e-06,
+      "loss": 1.5881,
+      "step": 4325
+    },
+    {
+      "epoch": 0.577261809447558,
+      "grad_norm": 0.9638573991235978,
+      "learning_rate": 7.997462967657946e-06,
+      "loss": 1.5619,
+      "step": 4326
+    },
+    {
+      "epoch": 0.5773952495329597,
+      "grad_norm": 0.9479429964037883,
+      "learning_rate": 7.993228794117332e-06,
+      "loss": 1.573,
+      "step": 4327
+    },
+    {
+      "epoch": 0.5775286896183613,
+      "grad_norm": 0.9744883318609773,
+      "learning_rate": 7.988994995418731e-06,
+      "loss": 1.5964,
+      "step": 4328
+    },
+    {
+      "epoch": 0.577662129703763,
+      "grad_norm": 0.9469778260011422,
+      "learning_rate": 7.984761572352963e-06,
+      "loss": 1.5858,
+      "step": 4329
+    },
+    {
+      "epoch": 0.5777955697891647,
+      "grad_norm": 0.9464430279636084,
+      "learning_rate": 7.980528525710795e-06,
+      "loss": 1.6171,
+      "step": 4330
+    },
+    {
+      "epoch": 0.5779290098745663,
+      "grad_norm": 7.978193421669048,
+      "learning_rate": 7.976295856282904e-06,
+      "loss": 1.5857,
+      "step": 4331
+    },
+    {
+      "epoch": 0.578062449959968,
+      "grad_norm": 1.2761028754638115,
+      "learning_rate": 7.97206356485991e-06,
+      "loss": 1.6031,
+      "step": 4332
+    },
+    {
+      "epoch": 0.5781958900453696,
+      "grad_norm": 0.993124845092899,
+      "learning_rate": 7.967831652232346e-06,
+      "loss": 1.5551,
+      "step": 4333
+    },
+    {
+      "epoch": 0.5783293301307713,
+      "grad_norm": 0.962042815286027,
+      "learning_rate": 7.963600119190695e-06,
+      "loss": 1.5865,
+      "step": 4334
+    },
+    {
+      "epoch": 0.578462770216173,
+      "grad_norm": 0.9541286812670964,
+      "learning_rate": 7.959368966525358e-06,
+      "loss": 1.5067,
+      "step": 4335
+    },
+    {
+      "epoch": 0.5785962103015746,
+      "grad_norm": 0.9197038368832774,
+      "learning_rate": 7.955138195026663e-06,
+      "loss": 1.5261,
+      "step": 4336
+    },
+    {
+      "epoch": 0.5787296503869762,
+      "grad_norm": 0.9334310052059513,
+      "learning_rate": 7.950907805484874e-06,
+      "loss": 1.5611,
+      "step": 4337
+    },
+    {
+      "epoch": 0.5788630904723779,
+      "grad_norm": 0.9637654944676048,
+      "learning_rate": 7.946677798690175e-06,
+      "loss": 1.5307,
+      "step": 4338
+    },
+    {
+      "epoch": 0.5789965305577796,
+      "grad_norm": 0.9357279064845219,
+      "learning_rate": 7.942448175432687e-06,
+      "loss": 1.5625,
+      "step": 4339
+    },
+    {
+      "epoch": 0.5791299706431812,
+      "grad_norm": 1.0168364623595538,
+      "learning_rate": 7.938218936502451e-06,
+      "loss": 1.6031,
+      "step": 4340
+    },
+    {
+      "epoch": 0.5792634107285829,
+      "grad_norm": 0.9662217980950242,
+      "learning_rate": 7.933990082689447e-06,
+      "loss": 1.5805,
+      "step": 4341
+    },
+    {
+      "epoch": 0.5793968508139845,
+      "grad_norm": 0.9669704491262874,
+      "learning_rate": 7.92976161478357e-06,
+      "loss": 1.5878,
+      "step": 4342
+    },
+    {
+      "epoch": 0.5795302908993861,
+      "grad_norm": 1.00799200737011,
+      "learning_rate": 7.925533533574652e-06,
+      "loss": 1.5348,
+      "step": 4343
+    },
+    {
+      "epoch": 0.5796637309847879,
+      "grad_norm": 0.9856175919162833,
+      "learning_rate": 7.921305839852454e-06,
+      "loss": 1.5661,
+      "step": 4344
+    },
+    {
+      "epoch": 0.5797971710701895,
+      "grad_norm": 1.0344494447784194,
+      "learning_rate": 7.91707853440666e-06,
+      "loss": 1.5692,
+      "step": 4345
+    },
+    {
+      "epoch": 0.5799306111555912,
+      "grad_norm": 0.9964855077609798,
+      "learning_rate": 7.912851618026878e-06,
+      "loss": 1.6038,
+      "step": 4346
+    },
+    {
+      "epoch": 0.5800640512409928,
+      "grad_norm": 1.0409910183629953,
+      "learning_rate": 7.90862509150265e-06,
+      "loss": 1.5509,
+      "step": 4347
+    },
+    {
+      "epoch": 0.5801974913263944,
+      "grad_norm": 0.9473445719334401,
+      "learning_rate": 7.904398955623443e-06,
+      "loss": 1.6044,
+      "step": 4348
+    },
+    {
+      "epoch": 0.5803309314117961,
+      "grad_norm": 1.0011241389147565,
+      "learning_rate": 7.900173211178655e-06,
+      "loss": 1.5819,
+      "step": 4349
+    },
+    {
+      "epoch": 0.5804643714971978,
+      "grad_norm": 0.9616239459082261,
+      "learning_rate": 7.895947858957603e-06,
+      "loss": 1.5791,
+      "step": 4350
+    },
+    {
+      "epoch": 0.5805978115825994,
+      "grad_norm": 0.9896389964933032,
+      "learning_rate": 7.891722899749531e-06,
+      "loss": 1.5909,
+      "step": 4351
+    },
+    {
+      "epoch": 0.5807312516680011,
+      "grad_norm": 1.1806253383689567,
+      "learning_rate": 7.887498334343625e-06,
+      "loss": 1.5656,
+      "step": 4352
+    },
+    {
+      "epoch": 0.5808646917534027,
+      "grad_norm": 0.9467806964207979,
+      "learning_rate": 7.883274163528974e-06,
+      "loss": 1.5911,
+      "step": 4353
+    },
+    {
+      "epoch": 0.5809981318388043,
+      "grad_norm": 0.9523674420210726,
+      "learning_rate": 7.879050388094606e-06,
+      "loss": 1.5564,
+      "step": 4354
+    },
+    {
+      "epoch": 0.5811315719242061,
+      "grad_norm": 0.9826286319469207,
+      "learning_rate": 7.874827008829485e-06,
+      "loss": 1.6093,
+      "step": 4355
+    },
+    {
+      "epoch": 0.5812650120096077,
+      "grad_norm": 0.9353553399457585,
+      "learning_rate": 7.87060402652248e-06,
+      "loss": 1.5576,
+      "step": 4356
+    },
+    {
+      "epoch": 0.5813984520950093,
+      "grad_norm": 0.9890860146426698,
+      "learning_rate": 7.866381441962398e-06,
+      "loss": 1.6012,
+      "step": 4357
+    },
+    {
+      "epoch": 0.581531892180411,
+      "grad_norm": 1.3238728327798368,
+      "learning_rate": 7.86215925593797e-06,
+      "loss": 1.5197,
+      "step": 4358
+    },
+    {
+      "epoch": 0.5816653322658126,
+      "grad_norm": 0.9735487703890018,
+      "learning_rate": 7.857937469237853e-06,
+      "loss": 1.577,
+      "step": 4359
+    },
+    {
+      "epoch": 0.5817987723512144,
+      "grad_norm": 1.0582339375938654,
+      "learning_rate": 7.85371608265063e-06,
+      "loss": 1.5682,
+      "step": 4360
+    },
+    {
+      "epoch": 0.581932212436616,
+      "grad_norm": 1.1375161685312762,
+      "learning_rate": 7.849495096964803e-06,
+      "loss": 1.5383,
+      "step": 4361
+    },
+    {
+      "epoch": 0.5820656525220176,
+      "grad_norm": 1.026443022449838,
+      "learning_rate": 7.84527451296881e-06,
+      "loss": 1.5996,
+      "step": 4362
+    },
+    {
+      "epoch": 0.5821990926074193,
+      "grad_norm": 0.97828369672077,
+      "learning_rate": 7.841054331451008e-06,
+      "loss": 1.5338,
+      "step": 4363
+    },
+    {
+      "epoch": 0.5823325326928209,
+      "grad_norm": 1.0179518604332118,
+      "learning_rate": 7.836834553199675e-06,
+      "loss": 1.568,
+      "step": 4364
+    },
+    {
+      "epoch": 0.5824659727782225,
+      "grad_norm": 0.9804583013213366,
+      "learning_rate": 7.832615179003013e-06,
+      "loss": 1.5702,
+      "step": 4365
+    },
+    {
+      "epoch": 0.5825994128636243,
+      "grad_norm": 1.1561708940014976,
+      "learning_rate": 7.828396209649166e-06,
+      "loss": 1.5852,
+      "step": 4366
+    },
+    {
+      "epoch": 0.5827328529490259,
+      "grad_norm": 0.9703494752720414,
+      "learning_rate": 7.824177645926181e-06,
+      "loss": 1.6168,
+      "step": 4367
+    },
+    {
+      "epoch": 0.5828662930344275,
+      "grad_norm": 0.975607666820386,
+      "learning_rate": 7.819959488622034e-06,
+      "loss": 1.5722,
+      "step": 4368
+    },
+    {
+      "epoch": 0.5829997331198292,
+      "grad_norm": 1.330373136193926,
+      "learning_rate": 7.81574173852464e-06,
+      "loss": 1.5116,
+      "step": 4369
+    },
+    {
+      "epoch": 0.5831331732052308,
+      "grad_norm": 0.9750167029171607,
+      "learning_rate": 7.811524396421818e-06,
+      "loss": 1.5523,
+      "step": 4370
+    },
+    {
+      "epoch": 0.5832666132906325,
+      "grad_norm": 0.977713970035295,
+      "learning_rate": 7.807307463101323e-06,
+      "loss": 1.5085,
+      "step": 4371
+    },
+    {
+      "epoch": 0.5834000533760342,
+      "grad_norm": 0.9684749707247869,
+      "learning_rate": 7.803090939350825e-06,
+      "loss": 1.5986,
+      "step": 4372
+    },
+    {
+      "epoch": 0.5835334934614358,
+      "grad_norm": 0.9894691014054225,
+      "learning_rate": 7.798874825957932e-06,
+      "loss": 1.5637,
+      "step": 4373
+    },
+    {
+      "epoch": 0.5836669335468375,
+      "grad_norm": 0.9811750958682591,
+      "learning_rate": 7.79465912371016e-06,
+      "loss": 1.5513,
+      "step": 4374
+    },
+    {
+      "epoch": 0.5838003736322391,
+      "grad_norm": 0.9268597986006433,
+      "learning_rate": 7.790443833394951e-06,
+      "loss": 1.5789,
+      "step": 4375
+    },
+    {
+      "epoch": 0.5839338137176407,
+      "grad_norm": 0.9589691936010886,
+      "learning_rate": 7.786228955799682e-06,
+      "loss": 1.6262,
+      "step": 4376
+    },
+    {
+      "epoch": 0.5840672538030425,
+      "grad_norm": 1.0121372632908987,
+      "learning_rate": 7.782014491711638e-06,
+      "loss": 1.5929,
+      "step": 4377
+    },
+    {
+      "epoch": 0.5842006938884441,
+      "grad_norm": 0.9838055243486561,
+      "learning_rate": 7.777800441918036e-06,
+      "loss": 1.6404,
+      "step": 4378
+    },
+    {
+      "epoch": 0.5843341339738457,
+      "grad_norm": 0.9753282451165607,
+      "learning_rate": 7.773586807206007e-06,
+      "loss": 1.614,
+      "step": 4379
+    },
+    {
+      "epoch": 0.5844675740592474,
+      "grad_norm": 0.9364592889583468,
+      "learning_rate": 7.769373588362616e-06,
+      "loss": 1.6258,
+      "step": 4380
+    },
+    {
+      "epoch": 0.584601014144649,
+      "grad_norm": 0.9733374026966305,
+      "learning_rate": 7.765160786174838e-06,
+      "loss": 1.5896,
+      "step": 4381
+    },
+    {
+      "epoch": 0.5847344542300507,
+      "grad_norm": 0.9583014709288828,
+      "learning_rate": 7.760948401429587e-06,
+      "loss": 1.569,
+      "step": 4382
+    },
+    {
+      "epoch": 0.5848678943154524,
+      "grad_norm": 1.1438908541726007,
+      "learning_rate": 7.756736434913678e-06,
+      "loss": 1.5556,
+      "step": 4383
+    },
+    {
+      "epoch": 0.585001334400854,
+      "grad_norm": 0.9585046094942674,
+      "learning_rate": 7.752524887413864e-06,
+      "loss": 1.5598,
+      "step": 4384
+    },
+    {
+      "epoch": 0.5851347744862557,
+      "grad_norm": 1.0484117367016692,
+      "learning_rate": 7.748313759716812e-06,
+      "loss": 1.6091,
+      "step": 4385
+    },
+    {
+      "epoch": 0.5852682145716573,
+      "grad_norm": 0.9196030055638242,
+      "learning_rate": 7.74410305260911e-06,
+      "loss": 1.5314,
+      "step": 4386
+    },
+    {
+      "epoch": 0.585401654657059,
+      "grad_norm": 0.9757449887475365,
+      "learning_rate": 7.739892766877278e-06,
+      "loss": 1.577,
+      "step": 4387
+    },
+    {
+      "epoch": 0.5855350947424607,
+      "grad_norm": 1.2029427959464503,
+      "learning_rate": 7.735682903307745e-06,
+      "loss": 1.6089,
+      "step": 4388
+    },
+    {
+      "epoch": 0.5856685348278623,
+      "grad_norm": 0.9931185472448057,
+      "learning_rate": 7.731473462686865e-06,
+      "loss": 1.566,
+      "step": 4389
+    },
+    {
+      "epoch": 0.5858019749132639,
+      "grad_norm": 0.9306099137325194,
+      "learning_rate": 7.727264445800909e-06,
+      "loss": 1.559,
+      "step": 4390
+    },
+    {
+      "epoch": 0.5859354149986656,
+      "grad_norm": 0.9714231921122422,
+      "learning_rate": 7.723055853436084e-06,
+      "loss": 1.6069,
+      "step": 4391
+    },
+    {
+      "epoch": 0.5860688550840673,
+      "grad_norm": 0.9857980626951369,
+      "learning_rate": 7.718847686378502e-06,
+      "loss": 1.5685,
+      "step": 4392
+    },
+    {
+      "epoch": 0.5862022951694689,
+      "grad_norm": 1.0373287455803648,
+      "learning_rate": 7.714639945414193e-06,
+      "loss": 1.6062,
+      "step": 4393
+    },
+    {
+      "epoch": 0.5863357352548706,
+      "grad_norm": 1.0138880235597887,
+      "learning_rate": 7.71043263132913e-06,
+      "loss": 1.5886,
+      "step": 4394
+    },
+    {
+      "epoch": 0.5864691753402722,
+      "grad_norm": 0.9451198980798459,
+      "learning_rate": 7.70622574490918e-06,
+      "loss": 1.6238,
+      "step": 4395
+    },
+    {
+      "epoch": 0.5866026154256738,
+      "grad_norm": 0.9760854352122434,
+      "learning_rate": 7.70201928694015e-06,
+      "loss": 1.5641,
+      "step": 4396
+    },
+    {
+      "epoch": 0.5867360555110755,
+      "grad_norm": 0.9771605508734201,
+      "learning_rate": 7.697813258207747e-06,
+      "loss": 1.5616,
+      "step": 4397
+    },
+    {
+      "epoch": 0.5868694955964772,
+      "grad_norm": 1.016442594503307,
+      "learning_rate": 7.693607659497621e-06,
+      "loss": 1.572,
+      "step": 4398
+    },
+    {
+      "epoch": 0.5870029356818789,
+      "grad_norm": 1.1109077879279698,
+      "learning_rate": 7.689402491595324e-06,
+      "loss": 1.5504,
+      "step": 4399
+    },
+    {
+      "epoch": 0.5871363757672805,
+      "grad_norm": 7.23681120527737,
+      "learning_rate": 7.685197755286332e-06,
+      "loss": 1.6183,
+      "step": 4400
+    },
+    {
+      "epoch": 0.5872698158526821,
+      "grad_norm": 1.0014151692392677,
+      "learning_rate": 7.680993451356047e-06,
+      "loss": 1.6214,
+      "step": 4401
+    },
+    {
+      "epoch": 0.5874032559380838,
+      "grad_norm": 1.0492817400153631,
+      "learning_rate": 7.676789580589781e-06,
+      "loss": 1.5989,
+      "step": 4402
+    },
+    {
+      "epoch": 0.5875366960234855,
+      "grad_norm": 0.9443703490938348,
+      "learning_rate": 7.672586143772772e-06,
+      "loss": 1.5646,
+      "step": 4403
+    },
+    {
+      "epoch": 0.5876701361088871,
+      "grad_norm": 0.9410293011189405,
+      "learning_rate": 7.66838314169017e-06,
+      "loss": 1.5367,
+      "step": 4404
+    },
+    {
+      "epoch": 0.5878035761942888,
+      "grad_norm": 0.9504920239170566,
+      "learning_rate": 7.664180575127054e-06,
+      "loss": 1.5748,
+      "step": 4405
+    },
+    {
+      "epoch": 0.5879370162796904,
+      "grad_norm": 1.1257241646301057,
+      "learning_rate": 7.659978444868412e-06,
+      "loss": 1.5842,
+      "step": 4406
+    },
+    {
+      "epoch": 0.588070456365092,
+      "grad_norm": 1.0186274137996902,
+      "learning_rate": 7.65577675169915e-06,
+      "loss": 1.591,
+      "step": 4407
+    },
+    {
+      "epoch": 0.5882038964504938,
+      "grad_norm": 1.1327767010645338,
+      "learning_rate": 7.651575496404104e-06,
+      "loss": 1.654,
+      "step": 4408
+    },
+    {
+      "epoch": 0.5883373365358954,
+      "grad_norm": 0.9109735683849596,
+      "learning_rate": 7.64737467976802e-06,
+      "loss": 1.5817,
+      "step": 4409
+    },
+    {
+      "epoch": 0.588470776621297,
+      "grad_norm": 0.9553652260357995,
+      "learning_rate": 7.643174302575558e-06,
+      "loss": 1.5662,
+      "step": 4410
+    },
+    {
+      "epoch": 0.5886042167066987,
+      "grad_norm": 0.9727279713423016,
+      "learning_rate": 7.638974365611299e-06,
+      "loss": 1.6463,
+      "step": 4411
+    },
+    {
+      "epoch": 0.5887376567921003,
+      "grad_norm": 0.9442402335099999,
+      "learning_rate": 7.634774869659751e-06,
+      "loss": 1.5588,
+      "step": 4412
+    },
+    {
+      "epoch": 0.588871096877502,
+      "grad_norm": 0.9022310204474651,
+      "learning_rate": 7.630575815505328e-06,
+      "loss": 1.5363,
+      "step": 4413
+    },
+    {
+      "epoch": 0.5890045369629037,
+      "grad_norm": 0.9386827766287047,
+      "learning_rate": 7.626377203932368e-06,
+      "loss": 1.5746,
+      "step": 4414
+    },
+    {
+      "epoch": 0.5891379770483053,
+      "grad_norm": 1.086942823690136,
+      "learning_rate": 7.622179035725116e-06,
+      "loss": 1.6061,
+      "step": 4415
+    },
+    {
+      "epoch": 0.589271417133707,
+      "grad_norm": 1.0205058353350165,
+      "learning_rate": 7.617981311667751e-06,
+      "loss": 1.5322,
+      "step": 4416
+    },
+    {
+      "epoch": 0.5894048572191086,
+      "grad_norm": 0.9823401514999283,
+      "learning_rate": 7.613784032544358e-06,
+      "loss": 1.5882,
+      "step": 4417
+    },
+    {
+      "epoch": 0.5895382973045102,
+      "grad_norm": 0.9433142149165429,
+      "learning_rate": 7.609587199138934e-06,
+      "loss": 1.5114,
+      "step": 4418
+    },
+    {
+      "epoch": 0.589671737389912,
+      "grad_norm": 0.9904727920933845,
+      "learning_rate": 7.605390812235412e-06,
+      "loss": 1.595,
+      "step": 4419
+    },
+    {
+      "epoch": 0.5898051774753136,
+      "grad_norm": 0.9453239515038033,
+      "learning_rate": 7.60119487261762e-06,
+      "loss": 1.5716,
+      "step": 4420
+    },
+    {
+      "epoch": 0.5899386175607152,
+      "grad_norm": 0.9710304464960416,
+      "learning_rate": 7.596999381069316e-06,
+      "loss": 1.6152,
+      "step": 4421
+    },
+    {
+      "epoch": 0.5900720576461169,
+      "grad_norm": 0.9149997314948785,
+      "learning_rate": 7.592804338374166e-06,
+      "loss": 1.5745,
+      "step": 4422
+    },
+    {
+      "epoch": 0.5902054977315185,
+      "grad_norm": 0.9611341617563518,
+      "learning_rate": 7.588609745315758e-06,
+      "loss": 1.5301,
+      "step": 4423
+    },
+    {
+      "epoch": 0.5903389378169202,
+      "grad_norm": 1.0320270411702828,
+      "learning_rate": 7.584415602677597e-06,
+      "loss": 1.6336,
+      "step": 4424
+    },
+    {
+      "epoch": 0.5904723779023219,
+      "grad_norm": 0.9578784490487312,
+      "learning_rate": 7.580221911243098e-06,
+      "loss": 1.612,
+      "step": 4425
+    },
+    {
+      "epoch": 0.5906058179877235,
+      "grad_norm": 1.0925180577920106,
+      "learning_rate": 7.576028671795596e-06,
+      "loss": 1.6268,
+      "step": 4426
+    },
+    {
+      "epoch": 0.5907392580731252,
+      "grad_norm": 1.0026716186226203,
+      "learning_rate": 7.571835885118341e-06,
+      "loss": 1.545,
+      "step": 4427
+    },
+    {
+      "epoch": 0.5908726981585268,
+      "grad_norm": 1.1376864232127364,
+      "learning_rate": 7.567643551994498e-06,
+      "loss": 1.579,
+      "step": 4428
+    },
+    {
+      "epoch": 0.5910061382439284,
+      "grad_norm": 0.99803761360832,
+      "learning_rate": 7.56345167320714e-06,
+      "loss": 1.5658,
+      "step": 4429
+    },
+    {
+      "epoch": 0.5911395783293302,
+      "grad_norm": 0.9610248802222642,
+      "learning_rate": 7.5592602495392744e-06,
+      "loss": 1.6057,
+      "step": 4430
+    },
+    {
+      "epoch": 0.5912730184147318,
+      "grad_norm": 0.9418422477809483,
+      "learning_rate": 7.555069281773805e-06,
+      "loss": 1.5575,
+      "step": 4431
+    },
+    {
+      "epoch": 0.5914064585001334,
+      "grad_norm": 0.9421321981921353,
+      "learning_rate": 7.550878770693551e-06,
+      "loss": 1.5754,
+      "step": 4432
+    },
+    {
+      "epoch": 0.5915398985855351,
+      "grad_norm": 1.035596217537491,
+      "learning_rate": 7.546688717081265e-06,
+      "loss": 1.5615,
+      "step": 4433
+    },
+    {
+      "epoch": 0.5916733386709367,
+      "grad_norm": 0.9571740732982612,
+      "learning_rate": 7.542499121719593e-06,
+      "loss": 1.5694,
+      "step": 4434
+    },
+    {
+      "epoch": 0.5918067787563384,
+      "grad_norm": 0.9590760123078373,
+      "learning_rate": 7.538309985391107e-06,
+      "loss": 1.5727,
+      "step": 4435
+    },
+    {
+      "epoch": 0.5919402188417401,
+      "grad_norm": 0.9626015322928804,
+      "learning_rate": 7.534121308878283e-06,
+      "loss": 1.5456,
+      "step": 4436
+    },
+    {
+      "epoch": 0.5920736589271417,
+      "grad_norm": 0.9448338727741227,
+      "learning_rate": 7.529933092963527e-06,
+      "loss": 1.5902,
+      "step": 4437
+    },
+    {
+      "epoch": 0.5922070990125433,
+      "grad_norm": 1.0098216915844183,
+      "learning_rate": 7.525745338429148e-06,
+      "loss": 1.5554,
+      "step": 4438
+    },
+    {
+      "epoch": 0.592340539097945,
+      "grad_norm": 0.9387499498237363,
+      "learning_rate": 7.521558046057364e-06,
+      "loss": 1.5319,
+      "step": 4439
+    },
+    {
+      "epoch": 0.5924739791833467,
+      "grad_norm": 0.9367552163787548,
+      "learning_rate": 7.517371216630324e-06,
+      "loss": 1.5604,
+      "step": 4440
+    },
+    {
+      "epoch": 0.5926074192687484,
+      "grad_norm": 15.945356768463004,
+      "learning_rate": 7.513184850930075e-06,
+      "loss": 1.5926,
+      "step": 4441
+    },
+    {
+      "epoch": 0.59274085935415,
+      "grad_norm": 1.0201759619596091,
+      "learning_rate": 7.5089989497385805e-06,
+      "loss": 1.5564,
+      "step": 4442
+    },
+    {
+      "epoch": 0.5928742994395516,
+      "grad_norm": 1.0381537793869493,
+      "learning_rate": 7.504813513837721e-06,
+      "loss": 1.5926,
+      "step": 4443
+    },
+    {
+      "epoch": 0.5930077395249533,
+      "grad_norm": 1.066841128052637,
+      "learning_rate": 7.5006285440092894e-06,
+      "loss": 1.5384,
+      "step": 4444
+    },
+    {
+      "epoch": 0.593141179610355,
+      "grad_norm": 1.060786355395775,
+      "learning_rate": 7.496444041034987e-06,
+      "loss": 1.5944,
+      "step": 4445
+    },
+    {
+      "epoch": 0.5932746196957566,
+      "grad_norm": 0.9860687095888007,
+      "learning_rate": 7.492260005696435e-06,
+      "loss": 1.5632,
+      "step": 4446
+    },
+    {
+      "epoch": 0.5934080597811583,
+      "grad_norm": 0.9858498336031716,
+      "learning_rate": 7.488076438775164e-06,
+      "loss": 1.5638,
+      "step": 4447
+    },
+    {
+      "epoch": 0.5935414998665599,
+      "grad_norm": 0.9833114803634568,
+      "learning_rate": 7.483893341052613e-06,
+      "loss": 1.5688,
+      "step": 4448
+    },
+    {
+      "epoch": 0.5936749399519615,
+      "grad_norm": 0.9443909139265638,
+      "learning_rate": 7.479710713310142e-06,
+      "loss": 1.6111,
+      "step": 4449
+    },
+    {
+      "epoch": 0.5938083800373632,
+      "grad_norm": 1.754532656615733,
+      "learning_rate": 7.4755285563290084e-06,
+      "loss": 1.5721,
+      "step": 4450
+    },
+    {
+      "epoch": 0.5939418201227649,
+      "grad_norm": 0.9696978925421736,
+      "learning_rate": 7.4713468708904035e-06,
+      "loss": 1.583,
+      "step": 4451
+    },
+    {
+      "epoch": 0.5940752602081665,
+      "grad_norm": 0.9849603466349487,
+      "learning_rate": 7.4671656577754134e-06,
+      "loss": 1.5552,
+      "step": 4452
+    },
+    {
+      "epoch": 0.5942087002935682,
+      "grad_norm": 0.9804321553897487,
+      "learning_rate": 7.462984917765042e-06,
+      "loss": 1.5638,
+      "step": 4453
+    },
+    {
+      "epoch": 0.5943421403789698,
+      "grad_norm": 0.987700482819893,
+      "learning_rate": 7.4588046516401965e-06,
+      "loss": 1.605,
+      "step": 4454
+    },
+    {
+      "epoch": 0.5944755804643715,
+      "grad_norm": 0.9960716798209621,
+      "learning_rate": 7.454624860181716e-06,
+      "loss": 1.5791,
+      "step": 4455
+    },
+    {
+      "epoch": 0.5946090205497732,
+      "grad_norm": 1.0747136053703605,
+      "learning_rate": 7.450445544170331e-06,
+      "loss": 1.5723,
+      "step": 4456
+    },
+    {
+      "epoch": 0.5947424606351748,
+      "grad_norm": 0.9519414871231925,
+      "learning_rate": 7.446266704386685e-06,
+      "loss": 1.5776,
+      "step": 4457
+    },
+    {
+      "epoch": 0.5948759007205765,
+      "grad_norm": 1.0093105843986834,
+      "learning_rate": 7.442088341611349e-06,
+      "loss": 1.5719,
+      "step": 4458
+    },
+    {
+      "epoch": 0.5950093408059781,
+      "grad_norm": 0.9633876822009539,
+      "learning_rate": 7.437910456624786e-06,
+      "loss": 1.5895,
+      "step": 4459
+    },
+    {
+      "epoch": 0.5951427808913797,
+      "grad_norm": 1.1771212337662196,
+      "learning_rate": 7.4337330502073815e-06,
+      "loss": 1.5678,
+      "step": 4460
+    },
+    {
+      "epoch": 0.5952762209767815,
+      "grad_norm": 1.027523600819857,
+      "learning_rate": 7.429556123139418e-06,
+      "loss": 1.5628,
+      "step": 4461
+    },
+    {
+      "epoch": 0.5954096610621831,
+      "grad_norm": 1.0233162890520189,
+      "learning_rate": 7.425379676201112e-06,
+      "loss": 1.6438,
+      "step": 4462
+    },
+    {
+      "epoch": 0.5955431011475847,
+      "grad_norm": 0.982674022132527,
+      "learning_rate": 7.421203710172569e-06,
+      "loss": 1.5377,
+      "step": 4463
+    },
+    {
+      "epoch": 0.5956765412329864,
+      "grad_norm": 0.928927279763224,
+      "learning_rate": 7.417028225833809e-06,
+      "loss": 1.5629,
+      "step": 4464
+    },
+    {
+      "epoch": 0.595809981318388,
+      "grad_norm": 1.003275613884901,
+      "learning_rate": 7.412853223964771e-06,
+      "loss": 1.5441,
+      "step": 4465
+    },
+    {
+      "epoch": 0.5959434214037898,
+      "grad_norm": 1.0353673874246174,
+      "learning_rate": 7.408678705345292e-06,
+      "loss": 1.6081,
+      "step": 4466
+    },
+    {
+      "epoch": 0.5960768614891914,
+      "grad_norm": 0.9778720926332395,
+      "learning_rate": 7.40450467075513e-06,
+      "loss": 1.5633,
+      "step": 4467
+    },
+    {
+      "epoch": 0.596210301574593,
+      "grad_norm": 1.0932199693315858,
+      "learning_rate": 7.400331120973943e-06,
+      "loss": 1.5514,
+      "step": 4468
+    },
+    {
+      "epoch": 0.5963437416599947,
+      "grad_norm": 0.9995117901765698,
+      "learning_rate": 7.3961580567813065e-06,
+      "loss": 1.6022,
+      "step": 4469
+    },
+    {
+      "epoch": 0.5964771817453963,
+      "grad_norm": 1.04545722147908,
+      "learning_rate": 7.391985478956699e-06,
+      "loss": 1.5811,
+      "step": 4470
+    },
+    {
+      "epoch": 0.5966106218307979,
+      "grad_norm": 0.9491415931893125,
+      "learning_rate": 7.387813388279507e-06,
+      "loss": 1.5659,
+      "step": 4471
+    },
+    {
+      "epoch": 0.5967440619161997,
+      "grad_norm": 0.9962391326790979,
+      "learning_rate": 7.383641785529037e-06,
+      "loss": 1.5627,
+      "step": 4472
+    },
+    {
+      "epoch": 0.5968775020016013,
+      "grad_norm": 0.9766553612308028,
+      "learning_rate": 7.379470671484492e-06,
+      "loss": 1.5631,
+      "step": 4473
+    },
+    {
+      "epoch": 0.5970109420870029,
+      "grad_norm": 0.9060488686225177,
+      "learning_rate": 7.375300046924991e-06,
+      "loss": 1.5714,
+      "step": 4474
+    },
+    {
+      "epoch": 0.5971443821724046,
+      "grad_norm": 0.9993988316021899,
+      "learning_rate": 7.3711299126295535e-06,
+      "loss": 1.6233,
+      "step": 4475
+    },
+    {
+      "epoch": 0.5972778222578062,
+      "grad_norm": 0.9685553543361992,
+      "learning_rate": 7.366960269377122e-06,
+      "loss": 1.5233,
+      "step": 4476
+    },
+    {
+      "epoch": 0.5974112623432078,
+      "grad_norm": 0.9992661533210414,
+      "learning_rate": 7.362791117946533e-06,
+      "loss": 1.5369,
+      "step": 4477
+    },
+    {
+      "epoch": 0.5975447024286096,
+      "grad_norm": 0.9445669078870059,
+      "learning_rate": 7.3586224591165335e-06,
+      "loss": 1.5614,
+      "step": 4478
+    },
+    {
+      "epoch": 0.5976781425140112,
+      "grad_norm": 1.1684515472112844,
+      "learning_rate": 7.354454293665789e-06,
+      "loss": 1.5789,
+      "step": 4479
+    },
+    {
+      "epoch": 0.5978115825994129,
+      "grad_norm": 0.9866975359267203,
+      "learning_rate": 7.350286622372863e-06,
+      "loss": 1.5559,
+      "step": 4480
+    },
+    {
+      "epoch": 0.5979450226848145,
+      "grad_norm": 0.948924170206914,
+      "learning_rate": 7.346119446016228e-06,
+      "loss": 1.6123,
+      "step": 4481
+    },
+    {
+      "epoch": 0.5980784627702161,
+      "grad_norm": 6.114143464113286,
+      "learning_rate": 7.34195276537426e-06,
+      "loss": 1.5768,
+      "step": 4482
+    },
+    {
+      "epoch": 0.5982119028556179,
+      "grad_norm": 1.0955825105939028,
+      "learning_rate": 7.337786581225257e-06,
+      "loss": 1.5832,
+      "step": 4483
+    },
+    {
+      "epoch": 0.5983453429410195,
+      "grad_norm": 1.1743034044346985,
+      "learning_rate": 7.333620894347408e-06,
+      "loss": 1.6208,
+      "step": 4484
+    },
+    {
+      "epoch": 0.5984787830264211,
+      "grad_norm": 0.9813455603327857,
+      "learning_rate": 7.329455705518822e-06,
+      "loss": 1.5635,
+      "step": 4485
+    },
+    {
+      "epoch": 0.5986122231118228,
+      "grad_norm": 0.97080471220344,
+      "learning_rate": 7.325291015517499e-06,
+      "loss": 1.5297,
+      "step": 4486
+    },
+    {
+      "epoch": 0.5987456631972244,
+      "grad_norm": 1.0737172814869553,
+      "learning_rate": 7.321126825121367e-06,
+      "loss": 1.5659,
+      "step": 4487
+    },
+    {
+      "epoch": 0.5988791032826261,
+      "grad_norm": 0.9965236541353721,
+      "learning_rate": 7.316963135108239e-06,
+      "loss": 1.5796,
+      "step": 4488
+    },
+    {
+      "epoch": 0.5990125433680278,
+      "grad_norm": 0.9817919653306049,
+      "learning_rate": 7.3127999462558515e-06,
+      "loss": 1.6373,
+      "step": 4489
+    },
+    {
+      "epoch": 0.5991459834534294,
+      "grad_norm": 1.050110329669737,
+      "learning_rate": 7.308637259341842e-06,
+      "loss": 1.5925,
+      "step": 4490
+    },
+    {
+      "epoch": 0.599279423538831,
+      "grad_norm": 0.967042532570296,
+      "learning_rate": 7.304475075143749e-06,
+      "loss": 1.6259,
+      "step": 4491
+    },
+    {
+      "epoch": 0.5994128636242327,
+      "grad_norm": 4.112038683707192,
+      "learning_rate": 7.3003133944390226e-06,
+      "loss": 1.5667,
+      "step": 4492
+    },
+    {
+      "epoch": 0.5995463037096344,
+      "grad_norm": 1.0543783444772516,
+      "learning_rate": 7.296152218005012e-06,
+      "loss": 1.547,
+      "step": 4493
+    },
+    {
+      "epoch": 0.5996797437950361,
+      "grad_norm": 1.0059093670666783,
+      "learning_rate": 7.291991546618987e-06,
+      "loss": 1.5244,
+      "step": 4494
+    },
+    {
+      "epoch": 0.5998131838804377,
+      "grad_norm": 0.9544888819843551,
+      "learning_rate": 7.28783138105811e-06,
+      "loss": 1.5661,
+      "step": 4495
+    },
+    {
+      "epoch": 0.5999466239658393,
+      "grad_norm": 0.99919602721375,
+      "learning_rate": 7.283671722099447e-06,
+      "loss": 1.5961,
+      "step": 4496
+    },
+    {
+      "epoch": 0.600080064051241,
+      "grad_norm": 0.9615117915366642,
+      "learning_rate": 7.279512570519984e-06,
+      "loss": 1.5151,
+      "step": 4497
+    },
+    {
+      "epoch": 0.6002135041366427,
+      "grad_norm": 0.9810089776515174,
+      "learning_rate": 7.275353927096599e-06,
+      "loss": 1.5632,
+      "step": 4498
+    },
+    {
+      "epoch": 0.6003469442220443,
+      "grad_norm": 0.930480896476526,
+      "learning_rate": 7.271195792606079e-06,
+      "loss": 1.5504,
+      "step": 4499
+    },
+    {
+      "epoch": 0.600480384307446,
+      "grad_norm": 1.1160803987845584,
+      "learning_rate": 7.2670381678251135e-06,
+      "loss": 1.5948,
+      "step": 4500
+    },
+    {
+      "epoch": 0.6006138243928476,
+      "grad_norm": 0.9662586762512358,
+      "learning_rate": 7.262881053530304e-06,
+      "loss": 1.5188,
+      "step": 4501
+    },
+    {
+      "epoch": 0.6007472644782492,
+      "grad_norm": 0.9546482055535903,
+      "learning_rate": 7.258724450498153e-06,
+      "loss": 1.5879,
+      "step": 4502
+    },
+    {
+      "epoch": 0.600880704563651,
+      "grad_norm": 0.9262806341117151,
+      "learning_rate": 7.254568359505059e-06,
+      "loss": 1.5666,
+      "step": 4503
+    },
+    {
+      "epoch": 0.6010141446490526,
+      "grad_norm": 0.9344199271306789,
+      "learning_rate": 7.250412781327341e-06,
+      "loss": 1.584,
+      "step": 4504
+    },
+    {
+      "epoch": 0.6011475847344542,
+      "grad_norm": 1.006266539272583,
+      "learning_rate": 7.24625771674121e-06,
+      "loss": 1.6019,
+      "step": 4505
+    },
+    {
+      "epoch": 0.6012810248198559,
+      "grad_norm": 0.9707157824495067,
+      "learning_rate": 7.242103166522786e-06,
+      "loss": 1.5485,
+      "step": 4506
+    },
+    {
+      "epoch": 0.6014144649052575,
+      "grad_norm": 0.9330255414867166,
+      "learning_rate": 7.2379491314480874e-06,
+      "loss": 1.6067,
+      "step": 4507
+    },
+    {
+      "epoch": 0.6015479049906592,
+      "grad_norm": 1.1082901632889555,
+      "learning_rate": 7.233795612293048e-06,
+      "loss": 1.6019,
+      "step": 4508
+    },
+    {
+      "epoch": 0.6016813450760609,
+      "grad_norm": 0.9629851142489154,
+      "learning_rate": 7.229642609833489e-06,
+      "loss": 1.6062,
+      "step": 4509
+    },
+    {
+      "epoch": 0.6018147851614625,
+      "grad_norm": 0.9416715246155154,
+      "learning_rate": 7.2254901248451515e-06,
+      "loss": 1.5347,
+      "step": 4510
+    },
+    {
+      "epoch": 0.6019482252468642,
+      "grad_norm": 1.0187472160513316,
+      "learning_rate": 7.22133815810367e-06,
+      "loss": 1.5216,
+      "step": 4511
+    },
+    {
+      "epoch": 0.6020816653322658,
+      "grad_norm": 1.2896554547647145,
+      "learning_rate": 7.217186710384585e-06,
+      "loss": 1.588,
+      "step": 4512
+    },
+    {
+      "epoch": 0.6022151054176674,
+      "grad_norm": 0.9584433847605133,
+      "learning_rate": 7.213035782463339e-06,
+      "loss": 1.5431,
+      "step": 4513
+    },
+    {
+      "epoch": 0.6023485455030692,
+      "grad_norm": 0.948451999723668,
+      "learning_rate": 7.208885375115273e-06,
+      "loss": 1.5385,
+      "step": 4514
+    },
+    {
+      "epoch": 0.6024819855884708,
+      "grad_norm": 0.9144383102664776,
+      "learning_rate": 7.204735489115646e-06,
+      "loss": 1.6077,
+      "step": 4515
+    },
+    {
+      "epoch": 0.6026154256738724,
+      "grad_norm": 1.0021475280100742,
+      "learning_rate": 7.200586125239605e-06,
+      "loss": 1.5783,
+      "step": 4516
+    },
+    {
+      "epoch": 0.6027488657592741,
+      "grad_norm": 1.0074542834053597,
+      "learning_rate": 7.196437284262202e-06,
+      "loss": 1.6181,
+      "step": 4517
+    },
+    {
+      "epoch": 0.6028823058446757,
+      "grad_norm": 0.9555123773303952,
+      "learning_rate": 7.19228896695839e-06,
+      "loss": 1.5896,
+      "step": 4518
+    },
+    {
+      "epoch": 0.6030157459300773,
+      "grad_norm": 0.9474460595871118,
+      "learning_rate": 7.1881411741030385e-06,
+      "loss": 1.525,
+      "step": 4519
+    },
+    {
+      "epoch": 0.6031491860154791,
+      "grad_norm": 0.9503729257356935,
+      "learning_rate": 7.1839939064708985e-06,
+      "loss": 1.6017,
+      "step": 4520
+    },
+    {
+      "epoch": 0.6032826261008807,
+      "grad_norm": 0.9907523282551414,
+      "learning_rate": 7.179847164836633e-06,
+      "loss": 1.585,
+      "step": 4521
+    },
+    {
+      "epoch": 0.6034160661862824,
+      "grad_norm": 0.9768177889092633,
+      "learning_rate": 7.1757009499748135e-06,
+      "loss": 1.5499,
+      "step": 4522
+    },
+    {
+      "epoch": 0.603549506271684,
+      "grad_norm": 0.9674311604833397,
+      "learning_rate": 7.171555262659899e-06,
+      "loss": 1.5691,
+      "step": 4523
+    },
+    {
+      "epoch": 0.6036829463570856,
+      "grad_norm": 1.045115554264623,
+      "learning_rate": 7.167410103666258e-06,
+      "loss": 1.5396,
+      "step": 4524
+    },
+    {
+      "epoch": 0.6038163864424874,
+      "grad_norm": 1.008009383355456,
+      "learning_rate": 7.1632654737681565e-06,
+      "loss": 1.5004,
+      "step": 4525
+    },
+    {
+      "epoch": 0.603949826527889,
+      "grad_norm": 0.9418379068039596,
+      "learning_rate": 7.15912137373977e-06,
+      "loss": 1.5583,
+      "step": 4526
+    },
+    {
+      "epoch": 0.6040832666132906,
+      "grad_norm": 0.9601231973109439,
+      "learning_rate": 7.154977804355167e-06,
+      "loss": 1.5904,
+      "step": 4527
+    },
+    {
+      "epoch": 0.6042167066986923,
+      "grad_norm": 1.091493380346413,
+      "learning_rate": 7.1508347663883175e-06,
+      "loss": 1.5876,
+      "step": 4528
+    },
+    {
+      "epoch": 0.6043501467840939,
+      "grad_norm": 0.9773122566221465,
+      "learning_rate": 7.146692260613095e-06,
+      "loss": 1.5567,
+      "step": 4529
+    },
+    {
+      "epoch": 0.6044835868694955,
+      "grad_norm": 0.9744498878123181,
+      "learning_rate": 7.142550287803271e-06,
+      "loss": 1.6091,
+      "step": 4530
+    },
+    {
+      "epoch": 0.6046170269548973,
+      "grad_norm": 0.9623687799737277,
+      "learning_rate": 7.138408848732521e-06,
+      "loss": 1.5327,
+      "step": 4531
+    },
+    {
+      "epoch": 0.6047504670402989,
+      "grad_norm": 1.0233729179790898,
+      "learning_rate": 7.134267944174415e-06,
+      "loss": 1.5748,
+      "step": 4532
+    },
+    {
+      "epoch": 0.6048839071257006,
+      "grad_norm": 1.0743557870760962,
+      "learning_rate": 7.130127574902433e-06,
+      "loss": 1.5972,
+      "step": 4533
+    },
+    {
+      "epoch": 0.6050173472111022,
+      "grad_norm": 0.9535415958368197,
+      "learning_rate": 7.125987741689946e-06,
+      "loss": 1.6278,
+      "step": 4534
+    },
+    {
+      "epoch": 0.6051507872965038,
+      "grad_norm": 0.9506929714552727,
+      "learning_rate": 7.121848445310221e-06,
+      "loss": 1.5897,
+      "step": 4535
+    },
+    {
+      "epoch": 0.6052842273819056,
+      "grad_norm": 1.0206134352483975,
+      "learning_rate": 7.117709686536442e-06,
+      "loss": 1.5732,
+      "step": 4536
+    },
+    {
+      "epoch": 0.6054176674673072,
+      "grad_norm": 0.9419673451960351,
+      "learning_rate": 7.113571466141678e-06,
+      "loss": 1.5487,
+      "step": 4537
+    },
+    {
+      "epoch": 0.6055511075527088,
+      "grad_norm": 0.9968498706511424,
+      "learning_rate": 7.1094337848989e-06,
+      "loss": 1.6093,
+      "step": 4538
+    },
+    {
+      "epoch": 0.6056845476381105,
+      "grad_norm": 0.9476217156397696,
+      "learning_rate": 7.105296643580979e-06,
+      "loss": 1.564,
+      "step": 4539
+    },
+    {
+      "epoch": 0.6058179877235121,
+      "grad_norm": 0.936868206502605,
+      "learning_rate": 7.101160042960688e-06,
+      "loss": 1.5533,
+      "step": 4540
+    },
+    {
+      "epoch": 0.6059514278089138,
+      "grad_norm": 1.0845350953229937,
+      "learning_rate": 7.097023983810699e-06,
+      "loss": 1.553,
+      "step": 4541
+    },
+    {
+      "epoch": 0.6060848678943155,
+      "grad_norm": 1.6937000276166494,
+      "learning_rate": 7.092888466903574e-06,
+      "loss": 1.602,
+      "step": 4542
+    },
+    {
+      "epoch": 0.6062183079797171,
+      "grad_norm": 0.9553209112139722,
+      "learning_rate": 7.0887534930117885e-06,
+      "loss": 1.6083,
+      "step": 4543
+    },
+    {
+      "epoch": 0.6063517480651187,
+      "grad_norm": 0.9702968584345638,
+      "learning_rate": 7.084619062907704e-06,
+      "loss": 1.6188,
+      "step": 4544
+    },
+    {
+      "epoch": 0.6064851881505204,
+      "grad_norm": 0.9483887892908953,
+      "learning_rate": 7.080485177363585e-06,
+      "loss": 1.5665,
+      "step": 4545
+    },
+    {
+      "epoch": 0.606618628235922,
+      "grad_norm": 0.9832425329814938,
+      "learning_rate": 7.076351837151593e-06,
+      "loss": 1.582,
+      "step": 4546
+    },
+    {
+      "epoch": 0.6067520683213238,
+      "grad_norm": 0.9685038627751058,
+      "learning_rate": 7.072219043043793e-06,
+      "loss": 1.6042,
+      "step": 4547
+    },
+    {
+      "epoch": 0.6068855084067254,
+      "grad_norm": 0.93835156044391,
+      "learning_rate": 7.0680867958121434e-06,
+      "loss": 1.5932,
+      "step": 4548
+    },
+    {
+      "epoch": 0.607018948492127,
+      "grad_norm": 0.9471054104804755,
+      "learning_rate": 7.063955096228498e-06,
+      "loss": 1.6032,
+      "step": 4549
+    },
+    {
+      "epoch": 0.6071523885775287,
+      "grad_norm": 0.9886055085335903,
+      "learning_rate": 7.059823945064611e-06,
+      "loss": 1.5914,
+      "step": 4550
+    },
+    {
+      "epoch": 0.6072858286629303,
+      "grad_norm": 1.1513593773945008,
+      "learning_rate": 7.055693343092138e-06,
+      "loss": 1.5919,
+      "step": 4551
+    },
+    {
+      "epoch": 0.607419268748332,
+      "grad_norm": 0.9399419338670711,
+      "learning_rate": 7.051563291082624e-06,
+      "loss": 1.6072,
+      "step": 4552
+    },
+    {
+      "epoch": 0.6075527088337337,
+      "grad_norm": 0.968419038546263,
+      "learning_rate": 7.047433789807518e-06,
+      "loss": 1.5867,
+      "step": 4553
+    },
+    {
+      "epoch": 0.6076861489191353,
+      "grad_norm": 1.159641461420781,
+      "learning_rate": 7.043304840038166e-06,
+      "loss": 1.6149,
+      "step": 4554
+    },
+    {
+      "epoch": 0.6078195890045369,
+      "grad_norm": 0.9281119870608585,
+      "learning_rate": 7.039176442545808e-06,
+      "loss": 1.5342,
+      "step": 4555
+    },
+    {
+      "epoch": 0.6079530290899386,
+      "grad_norm": 0.9235387184302374,
+      "learning_rate": 7.035048598101578e-06,
+      "loss": 1.5764,
+      "step": 4556
+    },
+    {
+      "epoch": 0.6080864691753403,
+      "grad_norm": 0.9721122200439459,
+      "learning_rate": 7.03092130747651e-06,
+      "loss": 1.549,
+      "step": 4557
+    },
+    {
+      "epoch": 0.6082199092607419,
+      "grad_norm": 0.9229902607898596,
+      "learning_rate": 7.02679457144154e-06,
+      "loss": 1.5277,
+      "step": 4558
+    },
+    {
+      "epoch": 0.6083533493461436,
+      "grad_norm": 1.0561032370889225,
+      "learning_rate": 7.022668390767495e-06,
+      "loss": 1.5849,
+      "step": 4559
+    },
+    {
+      "epoch": 0.6084867894315452,
+      "grad_norm": 0.910655450515761,
+      "learning_rate": 7.018542766225091e-06,
+      "loss": 1.5658,
+      "step": 4560
+    },
+    {
+      "epoch": 0.6086202295169469,
+      "grad_norm": 0.9741339058858085,
+      "learning_rate": 7.0144176985849565e-06,
+      "loss": 1.5601,
+      "step": 4561
+    },
+    {
+      "epoch": 0.6087536696023486,
+      "grad_norm": 0.9130348746535919,
+      "learning_rate": 7.0102931886176055e-06,
+      "loss": 1.5731,
+      "step": 4562
+    },
+    {
+      "epoch": 0.6088871096877502,
+      "grad_norm": 0.9215074638471374,
+      "learning_rate": 7.006169237093447e-06,
+      "loss": 1.5319,
+      "step": 4563
+    },
+    {
+      "epoch": 0.6090205497731519,
+      "grad_norm": 0.920657107314274,
+      "learning_rate": 7.002045844782785e-06,
+      "loss": 1.5355,
+      "step": 4564
+    },
+    {
+      "epoch": 0.6091539898585535,
+      "grad_norm": 0.9316084219094781,
+      "learning_rate": 6.9979230124558295e-06,
+      "loss": 1.5888,
+      "step": 4565
+    },
+    {
+      "epoch": 0.6092874299439551,
+      "grad_norm": 1.0136087188967267,
+      "learning_rate": 6.9938007408826765e-06,
+      "loss": 1.6226,
+      "step": 4566
+    },
+    {
+      "epoch": 0.6094208700293569,
+      "grad_norm": 0.9419684119712161,
+      "learning_rate": 6.989679030833314e-06,
+      "loss": 1.5506,
+      "step": 4567
+    },
+    {
+      "epoch": 0.6095543101147585,
+      "grad_norm": 0.9475350185524345,
+      "learning_rate": 6.9855578830776385e-06,
+      "loss": 1.5925,
+      "step": 4568
+    },
+    {
+      "epoch": 0.6096877502001601,
+      "grad_norm": 1.069989064607908,
+      "learning_rate": 6.98143729838543e-06,
+      "loss": 1.6082,
+      "step": 4569
+    },
+    {
+      "epoch": 0.6098211902855618,
+      "grad_norm": 0.9321140114017321,
+      "learning_rate": 6.977317277526366e-06,
+      "loss": 1.5179,
+      "step": 4570
+    },
+    {
+      "epoch": 0.6099546303709634,
+      "grad_norm": 1.0848876865520107,
+      "learning_rate": 6.973197821270018e-06,
+      "loss": 1.6046,
+      "step": 4571
+    },
+    {
+      "epoch": 0.610088070456365,
+      "grad_norm": 0.9683930596028457,
+      "learning_rate": 6.969078930385858e-06,
+      "loss": 1.5473,
+      "step": 4572
+    },
+    {
+      "epoch": 0.6102215105417668,
+      "grad_norm": 0.9338234054464237,
+      "learning_rate": 6.964960605643243e-06,
+      "loss": 1.5351,
+      "step": 4573
+    },
+    {
+      "epoch": 0.6103549506271684,
+      "grad_norm": 0.9528182144070801,
+      "learning_rate": 6.960842847811432e-06,
+      "loss": 1.5674,
+      "step": 4574
+    },
+    {
+      "epoch": 0.6104883907125701,
+      "grad_norm": 0.950808534998642,
+      "learning_rate": 6.956725657659578e-06,
+      "loss": 1.5983,
+      "step": 4575
+    },
+    {
+      "epoch": 0.6106218307979717,
+      "grad_norm": 1.0116221503224754,
+      "learning_rate": 6.95260903595672e-06,
+      "loss": 1.5796,
+      "step": 4576
+    },
+    {
+      "epoch": 0.6107552708833733,
+      "grad_norm": 0.9762343267133087,
+      "learning_rate": 6.948492983471799e-06,
+      "loss": 1.6156,
+      "step": 4577
+    },
+    {
+      "epoch": 0.6108887109687751,
+      "grad_norm": 0.9069025915216041,
+      "learning_rate": 6.944377500973642e-06,
+      "loss": 1.5628,
+      "step": 4578
+    },
+    {
+      "epoch": 0.6110221510541767,
+      "grad_norm": 0.9272734873764399,
+      "learning_rate": 6.9402625892309825e-06,
+      "loss": 1.5324,
+      "step": 4579
+    },
+    {
+      "epoch": 0.6111555911395783,
+      "grad_norm": 0.9668585275369073,
+      "learning_rate": 6.936148249012436e-06,
+      "loss": 1.5459,
+      "step": 4580
+    },
+    {
+      "epoch": 0.61128903122498,
+      "grad_norm": 0.9728627382480686,
+      "learning_rate": 6.932034481086512e-06,
+      "loss": 1.647,
+      "step": 4581
+    },
+    {
+      "epoch": 0.6114224713103816,
+      "grad_norm": 0.950645414784866,
+      "learning_rate": 6.927921286221613e-06,
+      "loss": 1.545,
+      "step": 4582
+    },
+    {
+      "epoch": 0.6115559113957832,
+      "grad_norm": 1.0447314600826156,
+      "learning_rate": 6.923808665186045e-06,
+      "loss": 1.6108,
+      "step": 4583
+    },
+    {
+      "epoch": 0.611689351481185,
+      "grad_norm": 1.0871835933095921,
+      "learning_rate": 6.919696618747994e-06,
+      "loss": 1.5817,
+      "step": 4584
+    },
+    {
+      "epoch": 0.6118227915665866,
+      "grad_norm": 1.041327828019737,
+      "learning_rate": 6.915585147675541e-06,
+      "loss": 1.5483,
+      "step": 4585
+    },
+    {
+      "epoch": 0.6119562316519882,
+      "grad_norm": 0.9762414215476425,
+      "learning_rate": 6.911474252736667e-06,
+      "loss": 1.507,
+      "step": 4586
+    },
+    {
+      "epoch": 0.6120896717373899,
+      "grad_norm": 1.030786578810606,
+      "learning_rate": 6.907363934699241e-06,
+      "loss": 1.6357,
+      "step": 4587
+    },
+    {
+      "epoch": 0.6122231118227915,
+      "grad_norm": 0.9814383825607843,
+      "learning_rate": 6.90325419433102e-06,
+      "loss": 1.5921,
+      "step": 4588
+    },
+    {
+      "epoch": 0.6123565519081933,
+      "grad_norm": 0.9363299777601847,
+      "learning_rate": 6.8991450323996535e-06,
+      "loss": 1.5827,
+      "step": 4589
+    },
+    {
+      "epoch": 0.6124899919935949,
+      "grad_norm": 0.9538862237104196,
+      "learning_rate": 6.895036449672694e-06,
+      "loss": 1.5372,
+      "step": 4590
+    },
+    {
+      "epoch": 0.6126234320789965,
+      "grad_norm": 0.97494009355015,
+      "learning_rate": 6.890928446917575e-06,
+      "loss": 1.561,
+      "step": 4591
+    },
+    {
+      "epoch": 0.6127568721643982,
+      "grad_norm": 0.9647040856357274,
+      "learning_rate": 6.886821024901622e-06,
+      "loss": 1.5882,
+      "step": 4592
+    },
+    {
+      "epoch": 0.6128903122497998,
+      "grad_norm": 1.091307209481251,
+      "learning_rate": 6.8827141843920585e-06,
+      "loss": 1.6223,
+      "step": 4593
+    },
+    {
+      "epoch": 0.6130237523352015,
+      "grad_norm": 0.9582456852774733,
+      "learning_rate": 6.878607926155992e-06,
+      "loss": 1.5471,
+      "step": 4594
+    },
+    {
+      "epoch": 0.6131571924206032,
+      "grad_norm": 0.9618755168259653,
+      "learning_rate": 6.874502250960429e-06,
+      "loss": 1.5851,
+      "step": 4595
+    },
+    {
+      "epoch": 0.6132906325060048,
+      "grad_norm": 0.9960213913651693,
+      "learning_rate": 6.870397159572257e-06,
+      "loss": 1.5711,
+      "step": 4596
+    },
+    {
+      "epoch": 0.6134240725914064,
+      "grad_norm": 0.9675989710680151,
+      "learning_rate": 6.866292652758266e-06,
+      "loss": 1.5615,
+      "step": 4597
+    },
+    {
+      "epoch": 0.6135575126768081,
+      "grad_norm": 1.1290180972157486,
+      "learning_rate": 6.862188731285131e-06,
+      "loss": 1.5547,
+      "step": 4598
+    },
+    {
+      "epoch": 0.6136909527622098,
+      "grad_norm": 1.0884500014392584,
+      "learning_rate": 6.8580853959194095e-06,
+      "loss": 1.5555,
+      "step": 4599
+    },
+    {
+      "epoch": 0.6138243928476115,
+      "grad_norm": 0.9752550116948083,
+      "learning_rate": 6.853982647427568e-06,
+      "loss": 1.5554,
+      "step": 4600
+    },
+    {
+      "epoch": 0.6139578329330131,
+      "grad_norm": 0.9556126452390394,
+      "learning_rate": 6.84988048657595e-06,
+      "loss": 1.6092,
+      "step": 4601
+    },
+    {
+      "epoch": 0.6140912730184147,
+      "grad_norm": 0.9563824103011194,
+      "learning_rate": 6.845778914130792e-06,
+      "loss": 1.5739,
+      "step": 4602
+    },
+    {
+      "epoch": 0.6142247131038164,
+      "grad_norm": 0.9841983764198419,
+      "learning_rate": 6.841677930858215e-06,
+      "loss": 1.5626,
+      "step": 4603
+    },
+    {
+      "epoch": 0.614358153189218,
+      "grad_norm": 0.945890058393798,
+      "learning_rate": 6.837577537524247e-06,
+      "loss": 1.5651,
+      "step": 4604
+    },
+    {
+      "epoch": 0.6144915932746197,
+      "grad_norm": 0.9637365996725469,
+      "learning_rate": 6.833477734894789e-06,
+      "loss": 1.5531,
+      "step": 4605
+    },
+    {
+      "epoch": 0.6146250333600214,
+      "grad_norm": 14.074871733099643,
+      "learning_rate": 6.829378523735635e-06,
+      "loss": 1.4761,
+      "step": 4606
+    },
+    {
+      "epoch": 0.614758473445423,
+      "grad_norm": 1.0353012877423016,
+      "learning_rate": 6.825279904812476e-06,
+      "loss": 1.5987,
+      "step": 4607
+    },
+    {
+      "epoch": 0.6148919135308246,
+      "grad_norm": 1.069043606434025,
+      "learning_rate": 6.821181878890886e-06,
+      "loss": 1.6326,
+      "step": 4608
+    },
+    {
+      "epoch": 0.6150253536162263,
+      "grad_norm": 0.9544980613861165,
+      "learning_rate": 6.817084446736329e-06,
+      "loss": 1.5769,
+      "step": 4609
+    },
+    {
+      "epoch": 0.615158793701628,
+      "grad_norm": 1.0198488235290817,
+      "learning_rate": 6.812987609114155e-06,
+      "loss": 1.5888,
+      "step": 4610
+    },
+    {
+      "epoch": 0.6152922337870296,
+      "grad_norm": 1.0114306343190909,
+      "learning_rate": 6.808891366789614e-06,
+      "loss": 1.5866,
+      "step": 4611
+    },
+    {
+      "epoch": 0.6154256738724313,
+      "grad_norm": 0.9683323450632398,
+      "learning_rate": 6.804795720527832e-06,
+      "loss": 1.6103,
+      "step": 4612
+    },
+    {
+      "epoch": 0.6155591139578329,
+      "grad_norm": 0.9651109935189197,
+      "learning_rate": 6.800700671093831e-06,
+      "loss": 1.5516,
+      "step": 4613
+    },
+    {
+      "epoch": 0.6156925540432346,
+      "grad_norm": 1.02345582166681,
+      "learning_rate": 6.796606219252519e-06,
+      "loss": 1.5824,
+      "step": 4614
+    },
+    {
+      "epoch": 0.6158259941286363,
+      "grad_norm": 0.9451195015480363,
+      "learning_rate": 6.7925123657686956e-06,
+      "loss": 1.5651,
+      "step": 4615
+    },
+    {
+      "epoch": 0.6159594342140379,
+      "grad_norm": 0.9485310120839627,
+      "learning_rate": 6.78841911140704e-06,
+      "loss": 1.5163,
+      "step": 4616
+    },
+    {
+      "epoch": 0.6160928742994396,
+      "grad_norm": 0.9543391087151242,
+      "learning_rate": 6.784326456932129e-06,
+      "loss": 1.6248,
+      "step": 4617
+    },
+    {
+      "epoch": 0.6162263143848412,
+      "grad_norm": 1.1822886103916135,
+      "learning_rate": 6.7802344031084264e-06,
+      "loss": 1.589,
+      "step": 4618
+    },
+    {
+      "epoch": 0.6163597544702428,
+      "grad_norm": 1.0774092766967276,
+      "learning_rate": 6.77614295070028e-06,
+      "loss": 1.5877,
+      "step": 4619
+    },
+    {
+      "epoch": 0.6164931945556446,
+      "grad_norm": 1.0639094196701442,
+      "learning_rate": 6.772052100471924e-06,
+      "loss": 1.6265,
+      "step": 4620
+    },
+    {
+      "epoch": 0.6166266346410462,
+      "grad_norm": 0.9726691460017839,
+      "learning_rate": 6.76796185318748e-06,
+      "loss": 1.5791,
+      "step": 4621
+    },
+    {
+      "epoch": 0.6167600747264478,
+      "grad_norm": 0.8990169867198253,
+      "learning_rate": 6.763872209610969e-06,
+      "loss": 1.5604,
+      "step": 4622
+    },
+    {
+      "epoch": 0.6168935148118495,
+      "grad_norm": 0.9730961102585249,
+      "learning_rate": 6.759783170506283e-06,
+      "loss": 1.5419,
+      "step": 4623
+    },
+    {
+      "epoch": 0.6170269548972511,
+      "grad_norm": 1.2338560915834138,
+      "learning_rate": 6.755694736637206e-06,
+      "loss": 1.5565,
+      "step": 4624
+    },
+    {
+      "epoch": 0.6171603949826527,
+      "grad_norm": 0.9206856089021481,
+      "learning_rate": 6.7516069087674186e-06,
+      "loss": 1.5654,
+      "step": 4625
+    },
+    {
+      "epoch": 0.6172938350680545,
+      "grad_norm": 0.9814257642355421,
+      "learning_rate": 6.747519687660477e-06,
+      "loss": 1.5811,
+      "step": 4626
+    },
+    {
+      "epoch": 0.6174272751534561,
+      "grad_norm": 0.9803535643820207,
+      "learning_rate": 6.743433074079826e-06,
+      "loss": 1.6115,
+      "step": 4627
+    },
+    {
+      "epoch": 0.6175607152388578,
+      "grad_norm": 0.9340160005856276,
+      "learning_rate": 6.739347068788795e-06,
+      "loss": 1.5486,
+      "step": 4628
+    },
+    {
+      "epoch": 0.6176941553242594,
+      "grad_norm": 3.36171539060685,
+      "learning_rate": 6.7352616725506125e-06,
+      "loss": 1.5931,
+      "step": 4629
+    },
+    {
+      "epoch": 0.617827595409661,
+      "grad_norm": 1.4276244458599252,
+      "learning_rate": 6.731176886128379e-06,
+      "loss": 1.5847,
+      "step": 4630
+    },
+    {
+      "epoch": 0.6179610354950628,
+      "grad_norm": 1.0518515488300342,
+      "learning_rate": 6.727092710285081e-06,
+      "loss": 1.5419,
+      "step": 4631
+    },
+    {
+      "epoch": 0.6180944755804644,
+      "grad_norm": 0.9733143124272681,
+      "learning_rate": 6.723009145783607e-06,
+      "loss": 1.5984,
+      "step": 4632
+    },
+    {
+      "epoch": 0.618227915665866,
+      "grad_norm": 0.9431452106066329,
+      "learning_rate": 6.718926193386714e-06,
+      "loss": 1.5877,
+      "step": 4633
+    },
+    {
+      "epoch": 0.6183613557512677,
+      "grad_norm": 1.0629787651629599,
+      "learning_rate": 6.714843853857052e-06,
+      "loss": 1.5851,
+      "step": 4634
+    },
+    {
+      "epoch": 0.6184947958366693,
+      "grad_norm": 1.110188030750085,
+      "learning_rate": 6.710762127957152e-06,
+      "loss": 1.5875,
+      "step": 4635
+    },
+    {
+      "epoch": 0.618628235922071,
+      "grad_norm": 1.1946206918422089,
+      "learning_rate": 6.706681016449441e-06,
+      "loss": 1.6129,
+      "step": 4636
+    },
+    {
+      "epoch": 0.6187616760074727,
+      "grad_norm": 0.996759831026245,
+      "learning_rate": 6.702600520096216e-06,
+      "loss": 1.5783,
+      "step": 4637
+    },
+    {
+      "epoch": 0.6188951160928743,
+      "grad_norm": 0.945997143369553,
+      "learning_rate": 6.698520639659674e-06,
+      "loss": 1.5278,
+      "step": 4638
+    },
+    {
+      "epoch": 0.6190285561782759,
+      "grad_norm": 0.9733621613054791,
+      "learning_rate": 6.694441375901888e-06,
+      "loss": 1.5809,
+      "step": 4639
+    },
+    {
+      "epoch": 0.6191619962636776,
+      "grad_norm": 0.9735974397589957,
+      "learning_rate": 6.690362729584818e-06,
+      "loss": 1.6003,
+      "step": 4640
+    },
+    {
+      "epoch": 0.6192954363490792,
+      "grad_norm": 0.9864220968652933,
+      "learning_rate": 6.686284701470309e-06,
+      "loss": 1.5841,
+      "step": 4641
+    },
+    {
+      "epoch": 0.619428876434481,
+      "grad_norm": 0.9772535789032403,
+      "learning_rate": 6.682207292320084e-06,
+      "loss": 1.5611,
+      "step": 4642
+    },
+    {
+      "epoch": 0.6195623165198826,
+      "grad_norm": 1.0280279728242607,
+      "learning_rate": 6.678130502895769e-06,
+      "loss": 1.5529,
+      "step": 4643
+    },
+    {
+      "epoch": 0.6196957566052842,
+      "grad_norm": 0.9248902395905066,
+      "learning_rate": 6.674054333958854e-06,
+      "loss": 1.5616,
+      "step": 4644
+    },
+    {
+      "epoch": 0.6198291966906859,
+      "grad_norm": 0.9523872687282976,
+      "learning_rate": 6.669978786270721e-06,
+      "loss": 1.5617,
+      "step": 4645
+    },
+    {
+      "epoch": 0.6199626367760875,
+      "grad_norm": 0.9776370256998808,
+      "learning_rate": 6.665903860592635e-06,
+      "loss": 1.5794,
+      "step": 4646
+    },
+    {
+      "epoch": 0.6200960768614892,
+      "grad_norm": 0.952043520590941,
+      "learning_rate": 6.661829557685751e-06,
+      "loss": 1.4796,
+      "step": 4647
+    },
+    {
+      "epoch": 0.6202295169468909,
+      "grad_norm": 1.1835704731983463,
+      "learning_rate": 6.6577558783110985e-06,
+      "loss": 1.5927,
+      "step": 4648
+    },
+    {
+      "epoch": 0.6203629570322925,
+      "grad_norm": 0.9658315781942995,
+      "learning_rate": 6.65368282322959e-06,
+      "loss": 1.5559,
+      "step": 4649
+    },
+    {
+      "epoch": 0.6204963971176941,
+      "grad_norm": 0.9972805039885417,
+      "learning_rate": 6.649610393202037e-06,
+      "loss": 1.6225,
+      "step": 4650
+    },
+    {
+      "epoch": 0.6206298372030958,
+      "grad_norm": 1.0977055810151135,
+      "learning_rate": 6.645538588989117e-06,
+      "loss": 1.5555,
+      "step": 4651
+    },
+    {
+      "epoch": 0.6207632772884975,
+      "grad_norm": 1.25686084150492,
+      "learning_rate": 6.641467411351395e-06,
+      "loss": 1.6138,
+      "step": 4652
+    },
+    {
+      "epoch": 0.6208967173738991,
+      "grad_norm": 0.9606097211322975,
+      "learning_rate": 6.637396861049319e-06,
+      "loss": 1.5605,
+      "step": 4653
+    },
+    {
+      "epoch": 0.6210301574593008,
+      "grad_norm": 0.9543190537739319,
+      "learning_rate": 6.6333269388432295e-06,
+      "loss": 1.5894,
+      "step": 4654
+    },
+    {
+      "epoch": 0.6211635975447024,
+      "grad_norm": 0.9663911647320823,
+      "learning_rate": 6.6292576454933355e-06,
+      "loss": 1.5667,
+      "step": 4655
+    },
+    {
+      "epoch": 0.6212970376301041,
+      "grad_norm": 1.0490534949737367,
+      "learning_rate": 6.625188981759734e-06,
+      "loss": 1.6118,
+      "step": 4656
+    },
+    {
+      "epoch": 0.6214304777155057,
+      "grad_norm": 1.053734926356376,
+      "learning_rate": 6.621120948402411e-06,
+      "loss": 1.595,
+      "step": 4657
+    },
+    {
+      "epoch": 0.6215639178009074,
+      "grad_norm": 0.9503515200342192,
+      "learning_rate": 6.617053546181222e-06,
+      "loss": 1.5537,
+      "step": 4658
+    },
+    {
+      "epoch": 0.6216973578863091,
+      "grad_norm": 0.9603504106356093,
+      "learning_rate": 6.612986775855914e-06,
+      "loss": 1.5798,
+      "step": 4659
+    },
+    {
+      "epoch": 0.6218307979717107,
+      "grad_norm": 0.9347443417349168,
+      "learning_rate": 6.6089206381861135e-06,
+      "loss": 1.5463,
+      "step": 4660
+    },
+    {
+      "epoch": 0.6219642380571123,
+      "grad_norm": 0.9559788324052123,
+      "learning_rate": 6.60485513393133e-06,
+      "loss": 1.5512,
+      "step": 4661
+    },
+    {
+      "epoch": 0.622097678142514,
+      "grad_norm": 0.9350187591640845,
+      "learning_rate": 6.600790263850953e-06,
+      "loss": 1.5178,
+      "step": 4662
+    },
+    {
+      "epoch": 0.6222311182279157,
+      "grad_norm": 0.9721750502110509,
+      "learning_rate": 6.59672602870425e-06,
+      "loss": 1.6067,
+      "step": 4663
+    },
+    {
+      "epoch": 0.6223645583133173,
+      "grad_norm": 0.9681735106486332,
+      "learning_rate": 6.592662429250381e-06,
+      "loss": 1.5639,
+      "step": 4664
+    },
+    {
+      "epoch": 0.622497998398719,
+      "grad_norm": 1.0145576572286186,
+      "learning_rate": 6.588599466248376e-06,
+      "loss": 1.5534,
+      "step": 4665
+    },
+    {
+      "epoch": 0.6226314384841206,
+      "grad_norm": 1.0483912253591907,
+      "learning_rate": 6.58453714045715e-06,
+      "loss": 1.5743,
+      "step": 4666
+    },
+    {
+      "epoch": 0.6227648785695223,
+      "grad_norm": 0.9862160876054513,
+      "learning_rate": 6.580475452635495e-06,
+      "loss": 1.5696,
+      "step": 4667
+    },
+    {
+      "epoch": 0.622898318654924,
+      "grad_norm": 0.9554839586876005,
+      "learning_rate": 6.576414403542098e-06,
+      "loss": 1.5909,
+      "step": 4668
+    },
+    {
+      "epoch": 0.6230317587403256,
+      "grad_norm": 1.0820406864016012,
+      "learning_rate": 6.57235399393551e-06,
+      "loss": 1.5565,
+      "step": 4669
+    },
+    {
+      "epoch": 0.6231651988257273,
+      "grad_norm": 0.98515013329777,
+      "learning_rate": 6.568294224574168e-06,
+      "loss": 1.5059,
+      "step": 4670
+    },
+    {
+      "epoch": 0.6232986389111289,
+      "grad_norm": 1.0239649377684183,
+      "learning_rate": 6.564235096216397e-06,
+      "loss": 1.6116,
+      "step": 4671
+    },
+    {
+      "epoch": 0.6234320789965305,
+      "grad_norm": 0.9784224342622083,
+      "learning_rate": 6.560176609620392e-06,
+      "loss": 1.6186,
+      "step": 4672
+    },
+    {
+      "epoch": 0.6235655190819323,
+      "grad_norm": 0.9678732574412918,
+      "learning_rate": 6.556118765544233e-06,
+      "loss": 1.5708,
+      "step": 4673
+    },
+    {
+      "epoch": 0.6236989591673339,
+      "grad_norm": 0.9800954857490598,
+      "learning_rate": 6.5520615647458754e-06,
+      "loss": 1.5868,
+      "step": 4674
+    },
+    {
+      "epoch": 0.6238323992527355,
+      "grad_norm": 1.052059487698131,
+      "learning_rate": 6.548005007983163e-06,
+      "loss": 1.5298,
+      "step": 4675
+    },
+    {
+      "epoch": 0.6239658393381372,
+      "grad_norm": 0.9774410359007661,
+      "learning_rate": 6.543949096013814e-06,
+      "loss": 1.6378,
+      "step": 4676
+    },
+    {
+      "epoch": 0.6240992794235388,
+      "grad_norm": 1.0154666358789204,
+      "learning_rate": 6.539893829595425e-06,
+      "loss": 1.5616,
+      "step": 4677
+    },
+    {
+      "epoch": 0.6242327195089404,
+      "grad_norm": 1.0875488553183978,
+      "learning_rate": 6.535839209485473e-06,
+      "loss": 1.6052,
+      "step": 4678
+    },
+    {
+      "epoch": 0.6243661595943422,
+      "grad_norm": 0.996755457440969,
+      "learning_rate": 6.531785236441316e-06,
+      "loss": 1.5472,
+      "step": 4679
+    },
+    {
+      "epoch": 0.6244995996797438,
+      "grad_norm": 0.9597796664804172,
+      "learning_rate": 6.527731911220188e-06,
+      "loss": 1.5385,
+      "step": 4680
+    },
+    {
+      "epoch": 0.6246330397651455,
+      "grad_norm": 0.9868274657841508,
+      "learning_rate": 6.523679234579207e-06,
+      "loss": 1.5555,
+      "step": 4681
+    },
+    {
+      "epoch": 0.6247664798505471,
+      "grad_norm": 0.9610538520046898,
+      "learning_rate": 6.5196272072753665e-06,
+      "loss": 1.6318,
+      "step": 4682
+    },
+    {
+      "epoch": 0.6248999199359487,
+      "grad_norm": 0.9714178990062254,
+      "learning_rate": 6.515575830065538e-06,
+      "loss": 1.5617,
+      "step": 4683
+    },
+    {
+      "epoch": 0.6250333600213505,
+      "grad_norm": 1.108314947127384,
+      "learning_rate": 6.511525103706473e-06,
+      "loss": 1.6024,
+      "step": 4684
+    },
+    {
+      "epoch": 0.6251668001067521,
+      "grad_norm": 0.9704286966755745,
+      "learning_rate": 6.507475028954797e-06,
+      "loss": 1.5746,
+      "step": 4685
+    },
+    {
+      "epoch": 0.6253002401921537,
+      "grad_norm": 0.9303993833203598,
+      "learning_rate": 6.503425606567026e-06,
+      "loss": 1.5919,
+      "step": 4686
+    },
+    {
+      "epoch": 0.6254336802775554,
+      "grad_norm": 0.9605139751499199,
+      "learning_rate": 6.499376837299541e-06,
+      "loss": 1.5261,
+      "step": 4687
+    },
+    {
+      "epoch": 0.625567120362957,
+      "grad_norm": 0.9509770084954142,
+      "learning_rate": 6.4953287219086035e-06,
+      "loss": 1.5105,
+      "step": 4688
+    },
+    {
+      "epoch": 0.6257005604483586,
+      "grad_norm": 0.9729887663740894,
+      "learning_rate": 6.491281261150362e-06,
+      "loss": 1.6217,
+      "step": 4689
+    },
+    {
+      "epoch": 0.6258340005337604,
+      "grad_norm": 0.9413725926216433,
+      "learning_rate": 6.487234455780833e-06,
+      "loss": 1.5847,
+      "step": 4690
+    },
+    {
+      "epoch": 0.625967440619162,
+      "grad_norm": 0.9874467867153237,
+      "learning_rate": 6.483188306555913e-06,
+      "loss": 1.5823,
+      "step": 4691
+    },
+    {
+      "epoch": 0.6261008807045636,
+      "grad_norm": 0.9489046613674973,
+      "learning_rate": 6.479142814231374e-06,
+      "loss": 1.5656,
+      "step": 4692
+    },
+    {
+      "epoch": 0.6262343207899653,
+      "grad_norm": 0.9603586164165097,
+      "learning_rate": 6.475097979562873e-06,
+      "loss": 1.5801,
+      "step": 4693
+    },
+    {
+      "epoch": 0.6263677608753669,
+      "grad_norm": 0.9391371736350217,
+      "learning_rate": 6.471053803305938e-06,
+      "loss": 1.6123,
+      "step": 4694
+    },
+    {
+      "epoch": 0.6265012009607687,
+      "grad_norm": 0.9897178331583285,
+      "learning_rate": 6.467010286215971e-06,
+      "loss": 1.5621,
+      "step": 4695
+    },
+    {
+      "epoch": 0.6266346410461703,
+      "grad_norm": 1.3397968048069753,
+      "learning_rate": 6.462967429048259e-06,
+      "loss": 1.5448,
+      "step": 4696
+    },
+    {
+      "epoch": 0.6267680811315719,
+      "grad_norm": 0.9969569105203754,
+      "learning_rate": 6.458925232557964e-06,
+      "loss": 1.5672,
+      "step": 4697
+    },
+    {
+      "epoch": 0.6269015212169736,
+      "grad_norm": 1.4282904976235595,
+      "learning_rate": 6.4548836975001165e-06,
+      "loss": 1.5755,
+      "step": 4698
+    },
+    {
+      "epoch": 0.6270349613023752,
+      "grad_norm": 0.9304292340727252,
+      "learning_rate": 6.4508428246296306e-06,
+      "loss": 1.5704,
+      "step": 4699
+    },
+    {
+      "epoch": 0.6271684013877769,
+      "grad_norm": 1.0676289094313383,
+      "learning_rate": 6.446802614701298e-06,
+      "loss": 1.6326,
+      "step": 4700
+    },
+    {
+      "epoch": 0.6273018414731786,
+      "grad_norm": 0.9536617515956951,
+      "learning_rate": 6.44276306846978e-06,
+      "loss": 1.5803,
+      "step": 4701
+    },
+    {
+      "epoch": 0.6274352815585802,
+      "grad_norm": 1.0650052391897409,
+      "learning_rate": 6.438724186689621e-06,
+      "loss": 1.5345,
+      "step": 4702
+    },
+    {
+      "epoch": 0.6275687216439818,
+      "grad_norm": 0.928912755262742,
+      "learning_rate": 6.43468597011524e-06,
+      "loss": 1.6111,
+      "step": 4703
+    },
+    {
+      "epoch": 0.6277021617293835,
+      "grad_norm": 1.0609739791096682,
+      "learning_rate": 6.4306484195009246e-06,
+      "loss": 1.5605,
+      "step": 4704
+    },
+    {
+      "epoch": 0.6278356018147851,
+      "grad_norm": 1.0242522017660989,
+      "learning_rate": 6.426611535600848e-06,
+      "loss": 1.5373,
+      "step": 4705
+    },
+    {
+      "epoch": 0.6279690419001868,
+      "grad_norm": 1.1731834357230497,
+      "learning_rate": 6.422575319169047e-06,
+      "loss": 1.5407,
+      "step": 4706
+    },
+    {
+      "epoch": 0.6281024819855885,
+      "grad_norm": 0.9787653325551281,
+      "learning_rate": 6.418539770959451e-06,
+      "loss": 1.6097,
+      "step": 4707
+    },
+    {
+      "epoch": 0.6282359220709901,
+      "grad_norm": 1.0177554635966886,
+      "learning_rate": 6.414504891725848e-06,
+      "loss": 1.5137,
+      "step": 4708
+    },
+    {
+      "epoch": 0.6283693621563918,
+      "grad_norm": 0.9446114640014235,
+      "learning_rate": 6.41047068222191e-06,
+      "loss": 1.6302,
+      "step": 4709
+    },
+    {
+      "epoch": 0.6285028022417934,
+      "grad_norm": 4.459094665047282,
+      "learning_rate": 6.406437143201174e-06,
+      "loss": 1.6297,
+      "step": 4710
+    },
+    {
+      "epoch": 0.6286362423271951,
+      "grad_norm": 0.9485026202281593,
+      "learning_rate": 6.402404275417071e-06,
+      "loss": 1.532,
+      "step": 4711
+    },
+    {
+      "epoch": 0.6287696824125968,
+      "grad_norm": 0.9321489717238166,
+      "learning_rate": 6.39837207962289e-06,
+      "loss": 1.5491,
+      "step": 4712
+    },
+    {
+      "epoch": 0.6289031224979984,
+      "grad_norm": 1.013553078508425,
+      "learning_rate": 6.394340556571794e-06,
+      "loss": 1.5433,
+      "step": 4713
+    },
+    {
+      "epoch": 0.6290365625834,
+      "grad_norm": 0.9238968592943705,
+      "learning_rate": 6.390309707016833e-06,
+      "loss": 1.5176,
+      "step": 4714
+    },
+    {
+      "epoch": 0.6291700026688017,
+      "grad_norm": 0.9769175672259005,
+      "learning_rate": 6.386279531710921e-06,
+      "loss": 1.5942,
+      "step": 4715
+    },
+    {
+      "epoch": 0.6293034427542034,
+      "grad_norm": 0.9697292609419739,
+      "learning_rate": 6.382250031406851e-06,
+      "loss": 1.5822,
+      "step": 4716
+    },
+    {
+      "epoch": 0.629436882839605,
+      "grad_norm": 1.0258012834838501,
+      "learning_rate": 6.378221206857278e-06,
+      "loss": 1.5787,
+      "step": 4717
+    },
+    {
+      "epoch": 0.6295703229250067,
+      "grad_norm": 0.9633940936233738,
+      "learning_rate": 6.374193058814755e-06,
+      "loss": 1.557,
+      "step": 4718
+    },
+    {
+      "epoch": 0.6297037630104083,
+      "grad_norm": 0.9921666668432783,
+      "learning_rate": 6.370165588031686e-06,
+      "loss": 1.5687,
+      "step": 4719
+    },
+    {
+      "epoch": 0.6298372030958099,
+      "grad_norm": 0.9501732810691954,
+      "learning_rate": 6.366138795260356e-06,
+      "loss": 1.4943,
+      "step": 4720
+    },
+    {
+      "epoch": 0.6299706431812117,
+      "grad_norm": 0.952051603638449,
+      "learning_rate": 6.362112681252928e-06,
+      "loss": 1.5632,
+      "step": 4721
+    },
+    {
+      "epoch": 0.6301040832666133,
+      "grad_norm": 1.0075164901271165,
+      "learning_rate": 6.358087246761432e-06,
+      "loss": 1.5648,
+      "step": 4722
+    },
+    {
+      "epoch": 0.630237523352015,
+      "grad_norm": 0.9580671930174274,
+      "learning_rate": 6.354062492537772e-06,
+      "loss": 1.5348,
+      "step": 4723
+    },
+    {
+      "epoch": 0.6303709634374166,
+      "grad_norm": 0.9723415922337499,
+      "learning_rate": 6.3500384193337275e-06,
+      "loss": 1.5957,
+      "step": 4724
+    },
+    {
+      "epoch": 0.6305044035228182,
+      "grad_norm": 1.0243012844746644,
+      "learning_rate": 6.346015027900951e-06,
+      "loss": 1.6132,
+      "step": 4725
+    },
+    {
+      "epoch": 0.63063784360822,
+      "grad_norm": 0.923480627824626,
+      "learning_rate": 6.3419923189909674e-06,
+      "loss": 1.6316,
+      "step": 4726
+    },
+    {
+      "epoch": 0.6307712836936216,
+      "grad_norm": 0.9845476502020702,
+      "learning_rate": 6.337970293355164e-06,
+      "loss": 1.5867,
+      "step": 4727
+    },
+    {
+      "epoch": 0.6309047237790232,
+      "grad_norm": 1.0867072821365296,
+      "learning_rate": 6.333948951744823e-06,
+      "loss": 1.5594,
+      "step": 4728
+    },
+    {
+      "epoch": 0.6310381638644249,
+      "grad_norm": 0.9725956783048197,
+      "learning_rate": 6.329928294911076e-06,
+      "loss": 1.5799,
+      "step": 4729
+    },
+    {
+      "epoch": 0.6311716039498265,
+      "grad_norm": 0.92191760617083,
+      "learning_rate": 6.325908323604939e-06,
+      "loss": 1.5785,
+      "step": 4730
+    },
+    {
+      "epoch": 0.6313050440352281,
+      "grad_norm": 0.9502477068461954,
+      "learning_rate": 6.321889038577291e-06,
+      "loss": 1.5772,
+      "step": 4731
+    },
+    {
+      "epoch": 0.6314384841206299,
+      "grad_norm": 0.9378807993847438,
+      "learning_rate": 6.317870440578899e-06,
+      "loss": 1.5395,
+      "step": 4732
+    },
+    {
+      "epoch": 0.6315719242060315,
+      "grad_norm": 1.0944522310079952,
+      "learning_rate": 6.313852530360387e-06,
+      "loss": 1.597,
+      "step": 4733
+    },
+    {
+      "epoch": 0.6317053642914331,
+      "grad_norm": 0.9643616082201257,
+      "learning_rate": 6.309835308672248e-06,
+      "loss": 1.5474,
+      "step": 4734
+    },
+    {
+      "epoch": 0.6318388043768348,
+      "grad_norm": 1.1104130886728687,
+      "learning_rate": 6.305818776264864e-06,
+      "loss": 1.5716,
+      "step": 4735
+    },
+    {
+      "epoch": 0.6319722444622364,
+      "grad_norm": 1.0769075057648472,
+      "learning_rate": 6.301802933888472e-06,
+      "loss": 1.5691,
+      "step": 4736
+    },
+    {
+      "epoch": 0.6321056845476382,
+      "grad_norm": 0.9721124377884932,
+      "learning_rate": 6.297787782293188e-06,
+      "loss": 1.5656,
+      "step": 4737
+    },
+    {
+      "epoch": 0.6322391246330398,
+      "grad_norm": 0.945916645197926,
+      "learning_rate": 6.293773322228989e-06,
+      "loss": 1.5976,
+      "step": 4738
+    },
+    {
+      "epoch": 0.6323725647184414,
+      "grad_norm": 1.0327667218788614,
+      "learning_rate": 6.2897595544457426e-06,
+      "loss": 1.5513,
+      "step": 4739
+    },
+    {
+      "epoch": 0.6325060048038431,
+      "grad_norm": 0.9500434363342504,
+      "learning_rate": 6.285746479693166e-06,
+      "loss": 1.5667,
+      "step": 4740
+    },
+    {
+      "epoch": 0.6326394448892447,
+      "grad_norm": 1.1553660667606065,
+      "learning_rate": 6.281734098720862e-06,
+      "loss": 1.5331,
+      "step": 4741
+    },
+    {
+      "epoch": 0.6327728849746463,
+      "grad_norm": 1.0053138525266323,
+      "learning_rate": 6.27772241227829e-06,
+      "loss": 1.6188,
+      "step": 4742
+    },
+    {
+      "epoch": 0.6329063250600481,
+      "grad_norm": 1.117959784943353,
+      "learning_rate": 6.2737114211147946e-06,
+      "loss": 1.6097,
+      "step": 4743
+    },
+    {
+      "epoch": 0.6330397651454497,
+      "grad_norm": 0.91872923216644,
+      "learning_rate": 6.269701125979577e-06,
+      "loss": 1.5345,
+      "step": 4744
+    },
+    {
+      "epoch": 0.6331732052308513,
+      "grad_norm": 1.0580192994245377,
+      "learning_rate": 6.26569152762172e-06,
+      "loss": 1.5928,
+      "step": 4745
+    },
+    {
+      "epoch": 0.633306645316253,
+      "grad_norm": 0.9371703645600504,
+      "learning_rate": 6.261682626790169e-06,
+      "loss": 1.5538,
+      "step": 4746
+    },
+    {
+      "epoch": 0.6334400854016546,
+      "grad_norm": 0.9689827210946913,
+      "learning_rate": 6.257674424233742e-06,
+      "loss": 1.6359,
+      "step": 4747
+    },
+    {
+      "epoch": 0.6335735254870564,
+      "grad_norm": 0.9472008982891928,
+      "learning_rate": 6.253666920701125e-06,
+      "loss": 1.6057,
+      "step": 4748
+    },
+    {
+      "epoch": 0.633706965572458,
+      "grad_norm": 0.9192823024984135,
+      "learning_rate": 6.249660116940869e-06,
+      "loss": 1.5798,
+      "step": 4749
+    },
+    {
+      "epoch": 0.6338404056578596,
+      "grad_norm": 1.031749263396716,
+      "learning_rate": 6.245654013701408e-06,
+      "loss": 1.5517,
+      "step": 4750
+    },
+    {
+      "epoch": 0.6339738457432613,
+      "grad_norm": 0.9624574391068491,
+      "learning_rate": 6.2416486117310326e-06,
+      "loss": 1.6054,
+      "step": 4751
+    },
+    {
+      "epoch": 0.6341072858286629,
+      "grad_norm": 0.9912718125263276,
+      "learning_rate": 6.237643911777902e-06,
+      "loss": 1.5548,
+      "step": 4752
+    },
+    {
+      "epoch": 0.6342407259140646,
+      "grad_norm": 0.9308587514815552,
+      "learning_rate": 6.233639914590056e-06,
+      "loss": 1.5791,
+      "step": 4753
+    },
+    {
+      "epoch": 0.6343741659994663,
+      "grad_norm": 0.9574581547965888,
+      "learning_rate": 6.2296366209153935e-06,
+      "loss": 1.582,
+      "step": 4754
+    },
+    {
+      "epoch": 0.6345076060848679,
+      "grad_norm": 0.9471871838216913,
+      "learning_rate": 6.225634031501682e-06,
+      "loss": 1.5467,
+      "step": 4755
+    },
+    {
+      "epoch": 0.6346410461702695,
+      "grad_norm": 0.9623094686395599,
+      "learning_rate": 6.221632147096557e-06,
+      "loss": 1.5942,
+      "step": 4756
+    },
+    {
+      "epoch": 0.6347744862556712,
+      "grad_norm": 1.0311697985777823,
+      "learning_rate": 6.2176309684475325e-06,
+      "loss": 1.5744,
+      "step": 4757
+    },
+    {
+      "epoch": 0.6349079263410728,
+      "grad_norm": 1.000138982436376,
+      "learning_rate": 6.213630496301979e-06,
+      "loss": 1.6009,
+      "step": 4758
+    },
+    {
+      "epoch": 0.6350413664264745,
+      "grad_norm": 1.0996770869148338,
+      "learning_rate": 6.209630731407138e-06,
+      "loss": 1.5789,
+      "step": 4759
+    },
+    {
+      "epoch": 0.6351748065118762,
+      "grad_norm": 0.9621838976319239,
+      "learning_rate": 6.205631674510122e-06,
+      "loss": 1.5206,
+      "step": 4760
+    },
+    {
+      "epoch": 0.6353082465972778,
+      "grad_norm": 1.0190348806503502,
+      "learning_rate": 6.2016333263579096e-06,
+      "loss": 1.529,
+      "step": 4761
+    },
+    {
+      "epoch": 0.6354416866826795,
+      "grad_norm": 0.9330752296166617,
+      "learning_rate": 6.197635687697345e-06,
+      "loss": 1.5315,
+      "step": 4762
+    },
+    {
+      "epoch": 0.6355751267680811,
+      "grad_norm": 0.9925347326400968,
+      "learning_rate": 6.1936387592751425e-06,
+      "loss": 1.5504,
+      "step": 4763
+    },
+    {
+      "epoch": 0.6357085668534828,
+      "grad_norm": 0.9627856019335367,
+      "learning_rate": 6.189642541837883e-06,
+      "loss": 1.5874,
+      "step": 4764
+    },
+    {
+      "epoch": 0.6358420069388845,
+      "grad_norm": 1.0100325648862956,
+      "learning_rate": 6.185647036132013e-06,
+      "loss": 1.6278,
+      "step": 4765
+    },
+    {
+      "epoch": 0.6359754470242861,
+      "grad_norm": 0.9333049284580013,
+      "learning_rate": 6.181652242903848e-06,
+      "loss": 1.5312,
+      "step": 4766
+    },
+    {
+      "epoch": 0.6361088871096877,
+      "grad_norm": 0.9491077426744001,
+      "learning_rate": 6.177658162899573e-06,
+      "loss": 1.5889,
+      "step": 4767
+    },
+    {
+      "epoch": 0.6362423271950894,
+      "grad_norm": 1.0817082893352956,
+      "learning_rate": 6.1736647968652345e-06,
+      "loss": 1.5957,
+      "step": 4768
+    },
+    {
+      "epoch": 0.6363757672804911,
+      "grad_norm": 1.0229373542272857,
+      "learning_rate": 6.169672145546747e-06,
+      "loss": 1.5911,
+      "step": 4769
+    },
+    {
+      "epoch": 0.6365092073658927,
+      "grad_norm": 1.0467028908683291,
+      "learning_rate": 6.165680209689889e-06,
+      "loss": 1.5631,
+      "step": 4770
+    },
+    {
+      "epoch": 0.6366426474512944,
+      "grad_norm": 0.9981899865143897,
+      "learning_rate": 6.161688990040315e-06,
+      "loss": 1.566,
+      "step": 4771
+    },
+    {
+      "epoch": 0.636776087536696,
+      "grad_norm": 0.9334836157550336,
+      "learning_rate": 6.15769848734354e-06,
+      "loss": 1.5353,
+      "step": 4772
+    },
+    {
+      "epoch": 0.6369095276220976,
+      "grad_norm": 0.9251044403398437,
+      "learning_rate": 6.1537087023449386e-06,
+      "loss": 1.5542,
+      "step": 4773
+    },
+    {
+      "epoch": 0.6370429677074994,
+      "grad_norm": 0.9282614735198017,
+      "learning_rate": 6.149719635789757e-06,
+      "loss": 1.5508,
+      "step": 4774
+    },
+    {
+      "epoch": 0.637176407792901,
+      "grad_norm": 0.9399449871354199,
+      "learning_rate": 6.145731288423114e-06,
+      "loss": 1.5636,
+      "step": 4775
+    },
+    {
+      "epoch": 0.6373098478783027,
+      "grad_norm": 0.9865252792900647,
+      "learning_rate": 6.141743660989984e-06,
+      "loss": 1.5661,
+      "step": 4776
+    },
+    {
+      "epoch": 0.6374432879637043,
+      "grad_norm": 0.9797903050975381,
+      "learning_rate": 6.137756754235205e-06,
+      "loss": 1.5981,
+      "step": 4777
+    },
+    {
+      "epoch": 0.6375767280491059,
+      "grad_norm": 0.9421928709402042,
+      "learning_rate": 6.133770568903497e-06,
+      "loss": 1.5313,
+      "step": 4778
+    },
+    {
+      "epoch": 0.6377101681345076,
+      "grad_norm": 1.0060524251777645,
+      "learning_rate": 6.129785105739427e-06,
+      "loss": 1.6234,
+      "step": 4779
+    },
+    {
+      "epoch": 0.6378436082199093,
+      "grad_norm": 0.9803594189527336,
+      "learning_rate": 6.1258003654874355e-06,
+      "loss": 1.58,
+      "step": 4780
+    },
+    {
+      "epoch": 0.6379770483053109,
+      "grad_norm": 0.9902714344055336,
+      "learning_rate": 6.121816348891822e-06,
+      "loss": 1.5152,
+      "step": 4781
+    },
+    {
+      "epoch": 0.6381104883907126,
+      "grad_norm": 1.057224007473837,
+      "learning_rate": 6.117833056696765e-06,
+      "loss": 1.5827,
+      "step": 4782
+    },
+    {
+      "epoch": 0.6382439284761142,
+      "grad_norm": 1.4743335061623923,
+      "learning_rate": 6.1138504896462915e-06,
+      "loss": 1.593,
+      "step": 4783
+    },
+    {
+      "epoch": 0.6383773685615158,
+      "grad_norm": 0.953837336988305,
+      "learning_rate": 6.1098686484843e-06,
+      "loss": 1.5946,
+      "step": 4784
+    },
+    {
+      "epoch": 0.6385108086469176,
+      "grad_norm": 1.0530003073440923,
+      "learning_rate": 6.105887533954555e-06,
+      "loss": 1.5438,
+      "step": 4785
+    },
+    {
+      "epoch": 0.6386442487323192,
+      "grad_norm": 1.091118899015719,
+      "learning_rate": 6.101907146800683e-06,
+      "loss": 1.5471,
+      "step": 4786
+    },
+    {
+      "epoch": 0.6387776888177208,
+      "grad_norm": 0.9672697431062299,
+      "learning_rate": 6.0979274877661734e-06,
+      "loss": 1.5845,
+      "step": 4787
+    },
+    {
+      "epoch": 0.6389111289031225,
+      "grad_norm": 1.004597787760983,
+      "learning_rate": 6.09394855759438e-06,
+      "loss": 1.5889,
+      "step": 4788
+    },
+    {
+      "epoch": 0.6390445689885241,
+      "grad_norm": 0.9685979515200861,
+      "learning_rate": 6.089970357028528e-06,
+      "loss": 1.5362,
+      "step": 4789
+    },
+    {
+      "epoch": 0.6391780090739259,
+      "grad_norm": 0.9618619817586496,
+      "learning_rate": 6.085992886811696e-06,
+      "loss": 1.5808,
+      "step": 4790
+    },
+    {
+      "epoch": 0.6393114491593275,
+      "grad_norm": 0.928222428021561,
+      "learning_rate": 6.082016147686824e-06,
+      "loss": 1.5076,
+      "step": 4791
+    },
+    {
+      "epoch": 0.6394448892447291,
+      "grad_norm": 0.9491261783995589,
+      "learning_rate": 6.0780401403967346e-06,
+      "loss": 1.5641,
+      "step": 4792
+    },
+    {
+      "epoch": 0.6395783293301308,
+      "grad_norm": 1.0480229583034089,
+      "learning_rate": 6.0740648656840925e-06,
+      "loss": 1.54,
+      "step": 4793
+    },
+    {
+      "epoch": 0.6397117694155324,
+      "grad_norm": 0.9544547807714423,
+      "learning_rate": 6.070090324291436e-06,
+      "loss": 1.5695,
+      "step": 4794
+    },
+    {
+      "epoch": 0.639845209500934,
+      "grad_norm": 0.9172480413703332,
+      "learning_rate": 6.066116516961157e-06,
+      "loss": 1.525,
+      "step": 4795
+    },
+    {
+      "epoch": 0.6399786495863358,
+      "grad_norm": 0.9492147743557356,
+      "learning_rate": 6.062143444435529e-06,
+      "loss": 1.5423,
+      "step": 4796
+    },
+    {
+      "epoch": 0.6401120896717374,
+      "grad_norm": 0.97339963817088,
+      "learning_rate": 6.058171107456672e-06,
+      "loss": 1.5391,
+      "step": 4797
+    },
+    {
+      "epoch": 0.640245529757139,
+      "grad_norm": 1.2402769128852358,
+      "learning_rate": 6.0541995067665675e-06,
+      "loss": 1.5826,
+      "step": 4798
+    },
+    {
+      "epoch": 0.6403789698425407,
+      "grad_norm": 0.9756211394635729,
+      "learning_rate": 6.050228643107074e-06,
+      "loss": 1.5811,
+      "step": 4799
+    },
+    {
+      "epoch": 0.6405124099279423,
+      "grad_norm": 0.923769213113528,
+      "learning_rate": 6.046258517219902e-06,
+      "loss": 1.54,
+      "step": 4800
+    },
+    {
+      "epoch": 0.640645850013344,
+      "grad_norm": 1.1597847530387866,
+      "learning_rate": 6.042289129846623e-06,
+      "loss": 1.5603,
+      "step": 4801
+    },
+    {
+      "epoch": 0.6407792900987457,
+      "grad_norm": 1.1413316657225265,
+      "learning_rate": 6.03832048172867e-06,
+      "loss": 1.5174,
+      "step": 4802
+    },
+    {
+      "epoch": 0.6409127301841473,
+      "grad_norm": 0.9271474721968134,
+      "learning_rate": 6.0343525736073506e-06,
+      "loss": 1.5434,
+      "step": 4803
+    },
+    {
+      "epoch": 0.641046170269549,
+      "grad_norm": 0.9223447060910756,
+      "learning_rate": 6.03038540622382e-06,
+      "loss": 1.5687,
+      "step": 4804
+    },
+    {
+      "epoch": 0.6411796103549506,
+      "grad_norm": 0.9276917949899848,
+      "learning_rate": 6.026418980319098e-06,
+      "loss": 1.5586,
+      "step": 4805
+    },
+    {
+      "epoch": 0.6413130504403523,
+      "grad_norm": 0.9541595431849743,
+      "learning_rate": 6.022453296634069e-06,
+      "loss": 1.5539,
+      "step": 4806
+    },
+    {
+      "epoch": 0.641446490525754,
+      "grad_norm": 7.6768327775349,
+      "learning_rate": 6.01848835590948e-06,
+      "loss": 1.5641,
+      "step": 4807
+    },
+    {
+      "epoch": 0.6415799306111556,
+      "grad_norm": 1.2862124397586092,
+      "learning_rate": 6.0145241588859324e-06,
+      "loss": 1.5405,
+      "step": 4808
+    },
+    {
+      "epoch": 0.6417133706965572,
+      "grad_norm": 0.9605935993165724,
+      "learning_rate": 6.010560706303896e-06,
+      "loss": 1.5141,
+      "step": 4809
+    },
+    {
+      "epoch": 0.6418468107819589,
+      "grad_norm": 0.9542783776622776,
+      "learning_rate": 6.006597998903699e-06,
+      "loss": 1.522,
+      "step": 4810
+    },
+    {
+      "epoch": 0.6419802508673605,
+      "grad_norm": 0.9681776765662814,
+      "learning_rate": 6.002636037425531e-06,
+      "loss": 1.5561,
+      "step": 4811
+    },
+    {
+      "epoch": 0.6421136909527622,
+      "grad_norm": 0.9642447674867372,
+      "learning_rate": 5.998674822609438e-06,
+      "loss": 1.5985,
+      "step": 4812
+    },
+    {
+      "epoch": 0.6422471310381639,
+      "grad_norm": 0.991926581119737,
+      "learning_rate": 5.9947143551953275e-06,
+      "loss": 1.536,
+      "step": 4813
+    },
+    {
+      "epoch": 0.6423805711235655,
+      "grad_norm": 0.9566617525797997,
+      "learning_rate": 5.99075463592298e-06,
+      "loss": 1.5697,
+      "step": 4814
+    },
+    {
+      "epoch": 0.6425140112089672,
+      "grad_norm": 0.9676450282612642,
+      "learning_rate": 5.986795665532017e-06,
+      "loss": 1.5289,
+      "step": 4815
+    },
+    {
+      "epoch": 0.6426474512943688,
+      "grad_norm": 0.9268671621535911,
+      "learning_rate": 5.982837444761929e-06,
+      "loss": 1.5541,
+      "step": 4816
+    },
+    {
+      "epoch": 0.6427808913797705,
+      "grad_norm": 0.9595643521431843,
+      "learning_rate": 5.978879974352072e-06,
+      "loss": 1.5536,
+      "step": 4817
+    },
+    {
+      "epoch": 0.6429143314651722,
+      "grad_norm": 0.9796809856469711,
+      "learning_rate": 5.974923255041653e-06,
+      "loss": 1.5676,
+      "step": 4818
+    },
+    {
+      "epoch": 0.6430477715505738,
+      "grad_norm": 0.9646690903498765,
+      "learning_rate": 5.9709672875697445e-06,
+      "loss": 1.6259,
+      "step": 4819
+    },
+    {
+      "epoch": 0.6431812116359754,
+      "grad_norm": 0.9534358718940747,
+      "learning_rate": 5.967012072675269e-06,
+      "loss": 1.5495,
+      "step": 4820
+    },
+    {
+      "epoch": 0.6433146517213771,
+      "grad_norm": 0.9361561726578995,
+      "learning_rate": 5.963057611097026e-06,
+      "loss": 1.5803,
+      "step": 4821
+    },
+    {
+      "epoch": 0.6434480918067788,
+      "grad_norm": 0.9139126102709447,
+      "learning_rate": 5.9591039035736595e-06,
+      "loss": 1.5603,
+      "step": 4822
+    },
+    {
+      "epoch": 0.6435815318921804,
+      "grad_norm": 1.0664502849150646,
+      "learning_rate": 5.955150950843673e-06,
+      "loss": 1.5331,
+      "step": 4823
+    },
+    {
+      "epoch": 0.6437149719775821,
+      "grad_norm": 1.021561984009056,
+      "learning_rate": 5.951198753645437e-06,
+      "loss": 1.5335,
+      "step": 4824
+    },
+    {
+      "epoch": 0.6438484120629837,
+      "grad_norm": 0.9503166506512712,
+      "learning_rate": 5.9472473127171795e-06,
+      "loss": 1.5493,
+      "step": 4825
+    },
+    {
+      "epoch": 0.6439818521483853,
+      "grad_norm": 0.9149759543673214,
+      "learning_rate": 5.9432966287969805e-06,
+      "loss": 1.5783,
+      "step": 4826
+    },
+    {
+      "epoch": 0.644115292233787,
+      "grad_norm": 0.9900352522766276,
+      "learning_rate": 5.939346702622782e-06,
+      "loss": 1.632,
+      "step": 4827
+    },
+    {
+      "epoch": 0.6442487323191887,
+      "grad_norm": 0.9616514543356586,
+      "learning_rate": 5.93539753493239e-06,
+      "loss": 1.5656,
+      "step": 4828
+    },
+    {
+      "epoch": 0.6443821724045904,
+      "grad_norm": 0.9438961654754505,
+      "learning_rate": 5.93144912646346e-06,
+      "loss": 1.5744,
+      "step": 4829
+    },
+    {
+      "epoch": 0.644515612489992,
+      "grad_norm": 0.914116676055478,
+      "learning_rate": 5.927501477953508e-06,
+      "loss": 1.5665,
+      "step": 4830
+    },
+    {
+      "epoch": 0.6446490525753936,
+      "grad_norm": 1.0053787424529315,
+      "learning_rate": 5.923554590139917e-06,
+      "loss": 1.5984,
+      "step": 4831
+    },
+    {
+      "epoch": 0.6447824926607953,
+      "grad_norm": 1.0553270414804121,
+      "learning_rate": 5.919608463759916e-06,
+      "loss": 1.6098,
+      "step": 4832
+    },
+    {
+      "epoch": 0.644915932746197,
+      "grad_norm": 1.0814676717925145,
+      "learning_rate": 5.915663099550597e-06,
+      "loss": 1.5929,
+      "step": 4833
+    },
+    {
+      "epoch": 0.6450493728315986,
+      "grad_norm": 1.2260986664990163,
+      "learning_rate": 5.911718498248904e-06,
+      "loss": 1.5417,
+      "step": 4834
+    },
+    {
+      "epoch": 0.6451828129170003,
+      "grad_norm": 0.9760168090523013,
+      "learning_rate": 5.907774660591654e-06,
+      "loss": 1.5707,
+      "step": 4835
+    },
+    {
+      "epoch": 0.6453162530024019,
+      "grad_norm": 1.038465167882452,
+      "learning_rate": 5.903831587315505e-06,
+      "loss": 1.5461,
+      "step": 4836
+    },
+    {
+      "epoch": 0.6454496930878035,
+      "grad_norm": 0.9965870568346387,
+      "learning_rate": 5.899889279156976e-06,
+      "loss": 1.5951,
+      "step": 4837
+    },
+    {
+      "epoch": 0.6455831331732053,
+      "grad_norm": 0.921641688093775,
+      "learning_rate": 5.895947736852452e-06,
+      "loss": 1.5367,
+      "step": 4838
+    },
+    {
+      "epoch": 0.6457165732586069,
+      "grad_norm": 0.9488713580551892,
+      "learning_rate": 5.892006961138164e-06,
+      "loss": 1.5896,
+      "step": 4839
+    },
+    {
+      "epoch": 0.6458500133440085,
+      "grad_norm": 0.94133164082473,
+      "learning_rate": 5.8880669527502035e-06,
+      "loss": 1.5983,
+      "step": 4840
+    },
+    {
+      "epoch": 0.6459834534294102,
+      "grad_norm": 0.960447508576979,
+      "learning_rate": 5.884127712424517e-06,
+      "loss": 1.5464,
+      "step": 4841
+    },
+    {
+      "epoch": 0.6461168935148118,
+      "grad_norm": 1.0620853615582413,
+      "learning_rate": 5.880189240896916e-06,
+      "loss": 1.5659,
+      "step": 4842
+    },
+    {
+      "epoch": 0.6462503336002136,
+      "grad_norm": 0.9604672083966226,
+      "learning_rate": 5.876251538903059e-06,
+      "loss": 1.5793,
+      "step": 4843
+    },
+    {
+      "epoch": 0.6463837736856152,
+      "grad_norm": 0.96613702059847,
+      "learning_rate": 5.8723146071784654e-06,
+      "loss": 1.58,
+      "step": 4844
+    },
+    {
+      "epoch": 0.6465172137710168,
+      "grad_norm": 0.9248419108271216,
+      "learning_rate": 5.868378446458503e-06,
+      "loss": 1.5689,
+      "step": 4845
+    },
+    {
+      "epoch": 0.6466506538564185,
+      "grad_norm": 0.9190338150548885,
+      "learning_rate": 5.864443057478411e-06,
+      "loss": 1.5892,
+      "step": 4846
+    },
+    {
+      "epoch": 0.6467840939418201,
+      "grad_norm": 1.0016350457379213,
+      "learning_rate": 5.860508440973269e-06,
+      "loss": 1.5805,
+      "step": 4847
+    },
+    {
+      "epoch": 0.6469175340272217,
+      "grad_norm": 1.0549731633470492,
+      "learning_rate": 5.856574597678019e-06,
+      "loss": 1.5599,
+      "step": 4848
+    },
+    {
+      "epoch": 0.6470509741126235,
+      "grad_norm": 1.0001222985459557,
+      "learning_rate": 5.852641528327463e-06,
+      "loss": 1.5254,
+      "step": 4849
+    },
+    {
+      "epoch": 0.6471844141980251,
+      "grad_norm": 0.9739278595702441,
+      "learning_rate": 5.848709233656249e-06,
+      "loss": 1.6014,
+      "step": 4850
+    },
+    {
+      "epoch": 0.6473178542834267,
+      "grad_norm": 0.9808998489102456,
+      "learning_rate": 5.844777714398884e-06,
+      "loss": 1.5885,
+      "step": 4851
+    },
+    {
+      "epoch": 0.6474512943688284,
+      "grad_norm": 0.9562991254634166,
+      "learning_rate": 5.840846971289733e-06,
+      "loss": 1.5986,
+      "step": 4852
+    },
+    {
+      "epoch": 0.64758473445423,
+      "grad_norm": 0.9170189603539738,
+      "learning_rate": 5.836917005063016e-06,
+      "loss": 1.5917,
+      "step": 4853
+    },
+    {
+      "epoch": 0.6477181745396317,
+      "grad_norm": 1.0146263459940281,
+      "learning_rate": 5.832987816452804e-06,
+      "loss": 1.5877,
+      "step": 4854
+    },
+    {
+      "epoch": 0.6478516146250334,
+      "grad_norm": 1.0944373179261144,
+      "learning_rate": 5.82905940619302e-06,
+      "loss": 1.5499,
+      "step": 4855
+    },
+    {
+      "epoch": 0.647985054710435,
+      "grad_norm": 1.0119646752828362,
+      "learning_rate": 5.825131775017457e-06,
+      "loss": 1.5549,
+      "step": 4856
+    },
+    {
+      "epoch": 0.6481184947958367,
+      "grad_norm": 0.9692668549051643,
+      "learning_rate": 5.8212049236597426e-06,
+      "loss": 1.5815,
+      "step": 4857
+    },
+    {
+      "epoch": 0.6482519348812383,
+      "grad_norm": 0.9836312870853201,
+      "learning_rate": 5.817278852853373e-06,
+      "loss": 1.5055,
+      "step": 4858
+    },
+    {
+      "epoch": 0.64838537496664,
+      "grad_norm": 0.9881357720367061,
+      "learning_rate": 5.813353563331687e-06,
+      "loss": 1.6121,
+      "step": 4859
+    },
+    {
+      "epoch": 0.6485188150520417,
+      "grad_norm": 0.9589370913572998,
+      "learning_rate": 5.809429055827893e-06,
+      "loss": 1.5853,
+      "step": 4860
+    },
+    {
+      "epoch": 0.6486522551374433,
+      "grad_norm": 1.0049866924367945,
+      "learning_rate": 5.805505331075041e-06,
+      "loss": 1.5434,
+      "step": 4861
+    },
+    {
+      "epoch": 0.6487856952228449,
+      "grad_norm": 0.9715555700327767,
+      "learning_rate": 5.801582389806031e-06,
+      "loss": 1.5507,
+      "step": 4862
+    },
+    {
+      "epoch": 0.6489191353082466,
+      "grad_norm": 1.048295893300858,
+      "learning_rate": 5.797660232753635e-06,
+      "loss": 1.6251,
+      "step": 4863
+    },
+    {
+      "epoch": 0.6490525753936482,
+      "grad_norm": 1.1124589689964615,
+      "learning_rate": 5.793738860650462e-06,
+      "loss": 1.5035,
+      "step": 4864
+    },
+    {
+      "epoch": 0.6491860154790499,
+      "grad_norm": 1.0962898619845716,
+      "learning_rate": 5.78981827422898e-06,
+      "loss": 1.5325,
+      "step": 4865
+    },
+    {
+      "epoch": 0.6493194555644516,
+      "grad_norm": 0.9341503031181468,
+      "learning_rate": 5.785898474221505e-06,
+      "loss": 1.4986,
+      "step": 4866
+    },
+    {
+      "epoch": 0.6494528956498532,
+      "grad_norm": 1.042466248631925,
+      "learning_rate": 5.781979461360221e-06,
+      "loss": 1.5031,
+      "step": 4867
+    },
+    {
+      "epoch": 0.6495863357352548,
+      "grad_norm": 0.9970733349938308,
+      "learning_rate": 5.778061236377148e-06,
+      "loss": 1.5859,
+      "step": 4868
+    },
+    {
+      "epoch": 0.6497197758206565,
+      "grad_norm": 0.9565435536603323,
+      "learning_rate": 5.774143800004164e-06,
+      "loss": 1.5419,
+      "step": 4869
+    },
+    {
+      "epoch": 0.6498532159060582,
+      "grad_norm": 0.9836979900956078,
+      "learning_rate": 5.770227152973009e-06,
+      "loss": 1.5494,
+      "step": 4870
+    },
+    {
+      "epoch": 0.6499866559914599,
+      "grad_norm": 1.0135189985173951,
+      "learning_rate": 5.766311296015263e-06,
+      "loss": 1.6009,
+      "step": 4871
+    },
+    {
+      "epoch": 0.6501200960768615,
+      "grad_norm": 0.9506176687117898,
+      "learning_rate": 5.762396229862367e-06,
+      "loss": 1.5465,
+      "step": 4872
+    },
+    {
+      "epoch": 0.6502535361622631,
+      "grad_norm": 0.971050453122327,
+      "learning_rate": 5.758481955245603e-06,
+      "loss": 1.5391,
+      "step": 4873
+    },
+    {
+      "epoch": 0.6503869762476648,
+      "grad_norm": 1.0432819879051074,
+      "learning_rate": 5.754568472896123e-06,
+      "loss": 1.5995,
+      "step": 4874
+    },
+    {
+      "epoch": 0.6505204163330665,
+      "grad_norm": 1.079857789585133,
+      "learning_rate": 5.750655783544914e-06,
+      "loss": 1.5867,
+      "step": 4875
+    },
+    {
+      "epoch": 0.6506538564184681,
+      "grad_norm": 0.9601148800734111,
+      "learning_rate": 5.7467438879228255e-06,
+      "loss": 1.6103,
+      "step": 4876
+    },
+    {
+      "epoch": 0.6507872965038698,
+      "grad_norm": 1.0295074482916855,
+      "learning_rate": 5.7428327867605505e-06,
+      "loss": 1.5321,
+      "step": 4877
+    },
+    {
+      "epoch": 0.6509207365892714,
+      "grad_norm": 0.9531639765490284,
+      "learning_rate": 5.738922480788645e-06,
+      "loss": 1.5331,
+      "step": 4878
+    },
+    {
+      "epoch": 0.651054176674673,
+      "grad_norm": 0.9435011459958462,
+      "learning_rate": 5.7350129707375035e-06,
+      "loss": 1.5342,
+      "step": 4879
+    },
+    {
+      "epoch": 0.6511876167600748,
+      "grad_norm": 0.9604110330797555,
+      "learning_rate": 5.731104257337383e-06,
+      "loss": 1.671,
+      "step": 4880
+    },
+    {
+      "epoch": 0.6513210568454764,
+      "grad_norm": 1.1655933070828943,
+      "learning_rate": 5.727196341318383e-06,
+      "loss": 1.5843,
+      "step": 4881
+    },
+    {
+      "epoch": 0.6514544969308781,
+      "grad_norm": 0.915628984133999,
+      "learning_rate": 5.72328922341046e-06,
+      "loss": 1.5624,
+      "step": 4882
+    },
+    {
+      "epoch": 0.6515879370162797,
+      "grad_norm": 0.9121442794247614,
+      "learning_rate": 5.719382904343411e-06,
+      "loss": 1.5364,
+      "step": 4883
+    },
+    {
+      "epoch": 0.6517213771016813,
+      "grad_norm": 1.0971113356076188,
+      "learning_rate": 5.715477384846906e-06,
+      "loss": 1.606,
+      "step": 4884
+    },
+    {
+      "epoch": 0.651854817187083,
+      "grad_norm": 0.9787600524230616,
+      "learning_rate": 5.7115726656504425e-06,
+      "loss": 1.5761,
+      "step": 4885
+    },
+    {
+      "epoch": 0.6519882572724847,
+      "grad_norm": 0.9284424960740136,
+      "learning_rate": 5.7076687474833795e-06,
+      "loss": 1.5292,
+      "step": 4886
+    },
+    {
+      "epoch": 0.6521216973578863,
+      "grad_norm": 0.9590699643912758,
+      "learning_rate": 5.703765631074922e-06,
+      "loss": 1.5564,
+      "step": 4887
+    },
+    {
+      "epoch": 0.652255137443288,
+      "grad_norm": 0.9384956754557705,
+      "learning_rate": 5.699863317154133e-06,
+      "loss": 1.5851,
+      "step": 4888
+    },
+    {
+      "epoch": 0.6523885775286896,
+      "grad_norm": 0.9240054444337821,
+      "learning_rate": 5.695961806449917e-06,
+      "loss": 1.5768,
+      "step": 4889
+    },
+    {
+      "epoch": 0.6525220176140912,
+      "grad_norm": 1.021644427657407,
+      "learning_rate": 5.692061099691033e-06,
+      "loss": 1.5574,
+      "step": 4890
+    },
+    {
+      "epoch": 0.652655457699493,
+      "grad_norm": 1.017574465371194,
+      "learning_rate": 5.688161197606083e-06,
+      "loss": 1.6171,
+      "step": 4891
+    },
+    {
+      "epoch": 0.6527888977848946,
+      "grad_norm": 0.9327196434111883,
+      "learning_rate": 5.6842621009235345e-06,
+      "loss": 1.5752,
+      "step": 4892
+    },
+    {
+      "epoch": 0.6529223378702962,
+      "grad_norm": 0.9728469536906587,
+      "learning_rate": 5.680363810371687e-06,
+      "loss": 1.6167,
+      "step": 4893
+    },
+    {
+      "epoch": 0.6530557779556979,
+      "grad_norm": 1.1670661476372766,
+      "learning_rate": 5.676466326678697e-06,
+      "loss": 1.5304,
+      "step": 4894
+    },
+    {
+      "epoch": 0.6531892180410995,
+      "grad_norm": 1.0260789492682234,
+      "learning_rate": 5.672569650572574e-06,
+      "loss": 1.5498,
+      "step": 4895
+    },
+    {
+      "epoch": 0.6533226581265013,
+      "grad_norm": 0.942513307680248,
+      "learning_rate": 5.668673782781173e-06,
+      "loss": 1.5392,
+      "step": 4896
+    },
+    {
+      "epoch": 0.6534560982119029,
+      "grad_norm": 0.9337827595786473,
+      "learning_rate": 5.664778724032194e-06,
+      "loss": 1.5591,
+      "step": 4897
+    },
+    {
+      "epoch": 0.6535895382973045,
+      "grad_norm": 0.9280660756568939,
+      "learning_rate": 5.660884475053187e-06,
+      "loss": 1.5647,
+      "step": 4898
+    },
+    {
+      "epoch": 0.6537229783827062,
+      "grad_norm": 0.9775864120631028,
+      "learning_rate": 5.656991036571561e-06,
+      "loss": 1.5598,
+      "step": 4899
+    },
+    {
+      "epoch": 0.6538564184681078,
+      "grad_norm": 0.9090145991125382,
+      "learning_rate": 5.653098409314562e-06,
+      "loss": 1.5081,
+      "step": 4900
+    },
+    {
+      "epoch": 0.6539898585535094,
+      "grad_norm": 1.1498598455869353,
+      "learning_rate": 5.649206594009287e-06,
+      "loss": 1.5525,
+      "step": 4901
+    },
+    {
+      "epoch": 0.6541232986389112,
+      "grad_norm": 0.9452608782280502,
+      "learning_rate": 5.645315591382686e-06,
+      "loss": 1.484,
+      "step": 4902
+    },
+    {
+      "epoch": 0.6542567387243128,
+      "grad_norm": 1.0058072441790753,
+      "learning_rate": 5.641425402161553e-06,
+      "loss": 1.5849,
+      "step": 4903
+    },
+    {
+      "epoch": 0.6543901788097144,
+      "grad_norm": 0.9290237461844418,
+      "learning_rate": 5.6375360270725324e-06,
+      "loss": 1.5293,
+      "step": 4904
+    },
+    {
+      "epoch": 0.6545236188951161,
+      "grad_norm": 1.0552247300637112,
+      "learning_rate": 5.633647466842108e-06,
+      "loss": 1.5393,
+      "step": 4905
+    },
+    {
+      "epoch": 0.6546570589805177,
+      "grad_norm": 0.9311882795001768,
+      "learning_rate": 5.629759722196629e-06,
+      "loss": 1.5945,
+      "step": 4906
+    },
+    {
+      "epoch": 0.6547904990659194,
+      "grad_norm": 6.174325495147954,
+      "learning_rate": 5.625872793862276e-06,
+      "loss": 1.5381,
+      "step": 4907
+    },
+    {
+      "epoch": 0.6549239391513211,
+      "grad_norm": 0.9663081495612666,
+      "learning_rate": 5.621986682565084e-06,
+      "loss": 1.5634,
+      "step": 4908
+    },
+    {
+      "epoch": 0.6550573792367227,
+      "grad_norm": 1.0036068699063934,
+      "learning_rate": 5.61810138903093e-06,
+      "loss": 1.6021,
+      "step": 4909
+    },
+    {
+      "epoch": 0.6551908193221244,
+      "grad_norm": 1.0524583355071078,
+      "learning_rate": 5.614216913985551e-06,
+      "loss": 1.5163,
+      "step": 4910
+    },
+    {
+      "epoch": 0.655324259407526,
+      "grad_norm": 0.9691722893349417,
+      "learning_rate": 5.610333258154519e-06,
+      "loss": 1.5607,
+      "step": 4911
+    },
+    {
+      "epoch": 0.6554576994929276,
+      "grad_norm": 0.9220645447522142,
+      "learning_rate": 5.606450422263251e-06,
+      "loss": 1.5614,
+      "step": 4912
+    },
+    {
+      "epoch": 0.6555911395783294,
+      "grad_norm": 0.9449443848226934,
+      "learning_rate": 5.602568407037025e-06,
+      "loss": 1.6021,
+      "step": 4913
+    },
+    {
+      "epoch": 0.655724579663731,
+      "grad_norm": 0.991397619710688,
+      "learning_rate": 5.598687213200956e-06,
+      "loss": 1.5747,
+      "step": 4914
+    },
+    {
+      "epoch": 0.6558580197491326,
+      "grad_norm": 0.9384924843099448,
+      "learning_rate": 5.594806841480005e-06,
+      "loss": 1.5914,
+      "step": 4915
+    },
+    {
+      "epoch": 0.6559914598345343,
+      "grad_norm": 2.1381936866886178,
+      "learning_rate": 5.5909272925989756e-06,
+      "loss": 1.5951,
+      "step": 4916
+    },
+    {
+      "epoch": 0.6561248999199359,
+      "grad_norm": 0.9393256924385065,
+      "learning_rate": 5.587048567282533e-06,
+      "loss": 1.5611,
+      "step": 4917
+    },
+    {
+      "epoch": 0.6562583400053376,
+      "grad_norm": 1.0866095395539688,
+      "learning_rate": 5.583170666255174e-06,
+      "loss": 1.5516,
+      "step": 4918
+    },
+    {
+      "epoch": 0.6563917800907393,
+      "grad_norm": 0.9771179575753627,
+      "learning_rate": 5.5792935902412485e-06,
+      "loss": 1.6421,
+      "step": 4919
+    },
+    {
+      "epoch": 0.6565252201761409,
+      "grad_norm": 0.9901820279238563,
+      "learning_rate": 5.575417339964944e-06,
+      "loss": 1.573,
+      "step": 4920
+    },
+    {
+      "epoch": 0.6566586602615425,
+      "grad_norm": 0.9453752857587083,
+      "learning_rate": 5.5715419161503085e-06,
+      "loss": 1.5831,
+      "step": 4921
+    },
+    {
+      "epoch": 0.6567921003469442,
+      "grad_norm": 0.9600381275049558,
+      "learning_rate": 5.567667319521222e-06,
+      "loss": 1.5852,
+      "step": 4922
+    },
+    {
+      "epoch": 0.6569255404323459,
+      "grad_norm": 0.9722922529852711,
+      "learning_rate": 5.563793550801418e-06,
+      "loss": 1.5744,
+      "step": 4923
+    },
+    {
+      "epoch": 0.6570589805177476,
+      "grad_norm": 1.088044217413912,
+      "learning_rate": 5.559920610714471e-06,
+      "loss": 1.5541,
+      "step": 4924
+    },
+    {
+      "epoch": 0.6571924206031492,
+      "grad_norm": 1.0635606707693064,
+      "learning_rate": 5.5560484999838005e-06,
+      "loss": 1.5841,
+      "step": 4925
+    },
+    {
+      "epoch": 0.6573258606885508,
+      "grad_norm": 0.9473218174863391,
+      "learning_rate": 5.552177219332671e-06,
+      "loss": 1.5783,
+      "step": 4926
+    },
+    {
+      "epoch": 0.6574593007739525,
+      "grad_norm": 0.9671296980936441,
+      "learning_rate": 5.548306769484199e-06,
+      "loss": 1.5286,
+      "step": 4927
+    },
+    {
+      "epoch": 0.6575927408593542,
+      "grad_norm": 0.9206430753905613,
+      "learning_rate": 5.544437151161339e-06,
+      "loss": 1.5968,
+      "step": 4928
+    },
+    {
+      "epoch": 0.6577261809447558,
+      "grad_norm": 0.954056329951995,
+      "learning_rate": 5.540568365086891e-06,
+      "loss": 1.5636,
+      "step": 4929
+    },
+    {
+      "epoch": 0.6578596210301575,
+      "grad_norm": 0.9276580594697565,
+      "learning_rate": 5.536700411983495e-06,
+      "loss": 1.6237,
+      "step": 4930
+    },
+    {
+      "epoch": 0.6579930611155591,
+      "grad_norm": 0.9240637357436527,
+      "learning_rate": 5.5328332925736495e-06,
+      "loss": 1.5337,
+      "step": 4931
+    },
+    {
+      "epoch": 0.6581265012009607,
+      "grad_norm": 0.9782170394984411,
+      "learning_rate": 5.528967007579684e-06,
+      "loss": 1.5287,
+      "step": 4932
+    },
+    {
+      "epoch": 0.6582599412863624,
+      "grad_norm": 0.9715908393003854,
+      "learning_rate": 5.525101557723773e-06,
+      "loss": 1.6012,
+      "step": 4933
+    },
+    {
+      "epoch": 0.6583933813717641,
+      "grad_norm": 1.0552772715223717,
+      "learning_rate": 5.521236943727948e-06,
+      "loss": 1.6308,
+      "step": 4934
+    },
+    {
+      "epoch": 0.6585268214571657,
+      "grad_norm": 0.9273214286380065,
+      "learning_rate": 5.517373166314068e-06,
+      "loss": 1.5141,
+      "step": 4935
+    },
+    {
+      "epoch": 0.6586602615425674,
+      "grad_norm": 1.167409050528772,
+      "learning_rate": 5.513510226203844e-06,
+      "loss": 1.5365,
+      "step": 4936
+    },
+    {
+      "epoch": 0.658793701627969,
+      "grad_norm": 0.9613274910119957,
+      "learning_rate": 5.5096481241188246e-06,
+      "loss": 1.5552,
+      "step": 4937
+    },
+    {
+      "epoch": 0.6589271417133707,
+      "grad_norm": 1.0120678736060127,
+      "learning_rate": 5.505786860780416e-06,
+      "loss": 1.5682,
+      "step": 4938
+    },
+    {
+      "epoch": 0.6590605817987724,
+      "grad_norm": 0.9499179626379158,
+      "learning_rate": 5.5019264369098535e-06,
+      "loss": 1.5632,
+      "step": 4939
+    },
+    {
+      "epoch": 0.659194021884174,
+      "grad_norm": 0.9642512741416759,
+      "learning_rate": 5.4980668532282185e-06,
+      "loss": 1.5566,
+      "step": 4940
+    },
+    {
+      "epoch": 0.6593274619695757,
+      "grad_norm": 0.934320051094941,
+      "learning_rate": 5.494208110456437e-06,
+      "loss": 1.5399,
+      "step": 4941
+    },
+    {
+      "epoch": 0.6594609020549773,
+      "grad_norm": 1.0972956472557598,
+      "learning_rate": 5.490350209315283e-06,
+      "loss": 1.5875,
+      "step": 4942
+    },
+    {
+      "epoch": 0.6595943421403789,
+      "grad_norm": 0.9405793372237963,
+      "learning_rate": 5.486493150525365e-06,
+      "loss": 1.5474,
+      "step": 4943
+    },
+    {
+      "epoch": 0.6597277822257807,
+      "grad_norm": 0.9802055587437492,
+      "learning_rate": 5.482636934807135e-06,
+      "loss": 1.5946,
+      "step": 4944
+    },
+    {
+      "epoch": 0.6598612223111823,
+      "grad_norm": 0.949526189030937,
+      "learning_rate": 5.478781562880897e-06,
+      "loss": 1.595,
+      "step": 4945
+    },
+    {
+      "epoch": 0.6599946623965839,
+      "grad_norm": 0.9485337756721357,
+      "learning_rate": 5.474927035466789e-06,
+      "loss": 1.6176,
+      "step": 4946
+    },
+    {
+      "epoch": 0.6601281024819856,
+      "grad_norm": 0.917365602362209,
+      "learning_rate": 5.471073353284788e-06,
+      "loss": 1.5559,
+      "step": 4947
+    },
+    {
+      "epoch": 0.6602615425673872,
+      "grad_norm": 0.9752592699248175,
+      "learning_rate": 5.467220517054719e-06,
+      "loss": 1.5717,
+      "step": 4948
+    },
+    {
+      "epoch": 0.660394982652789,
+      "grad_norm": 1.0396268387538175,
+      "learning_rate": 5.463368527496254e-06,
+      "loss": 1.5159,
+      "step": 4949
+    },
+    {
+      "epoch": 0.6605284227381906,
+      "grad_norm": 0.9639723430780847,
+      "learning_rate": 5.459517385328894e-06,
+      "loss": 1.5804,
+      "step": 4950
+    },
+    {
+      "epoch": 0.6606618628235922,
+      "grad_norm": 0.9221523093110253,
+      "learning_rate": 5.4556670912719885e-06,
+      "loss": 1.5503,
+      "step": 4951
+    },
+    {
+      "epoch": 0.6607953029089939,
+      "grad_norm": 0.9723372937636856,
+      "learning_rate": 5.451817646044735e-06,
+      "loss": 1.6308,
+      "step": 4952
+    },
+    {
+      "epoch": 0.6609287429943955,
+      "grad_norm": 0.9472013736532353,
+      "learning_rate": 5.447969050366163e-06,
+      "loss": 1.5603,
+      "step": 4953
+    },
+    {
+      "epoch": 0.6610621830797971,
+      "grad_norm": 1.2387063918292023,
+      "learning_rate": 5.444121304955145e-06,
+      "loss": 1.6083,
+      "step": 4954
+    },
+    {
+      "epoch": 0.6611956231651989,
+      "grad_norm": 1.2148466148047536,
+      "learning_rate": 5.440274410530393e-06,
+      "loss": 1.5535,
+      "step": 4955
+    },
+    {
+      "epoch": 0.6613290632506005,
+      "grad_norm": 0.9349547994328788,
+      "learning_rate": 5.43642836781047e-06,
+      "loss": 1.6053,
+      "step": 4956
+    },
+    {
+      "epoch": 0.6614625033360021,
+      "grad_norm": 0.9546746784572933,
+      "learning_rate": 5.43258317751377e-06,
+      "loss": 1.5645,
+      "step": 4957
+    },
+    {
+      "epoch": 0.6615959434214038,
+      "grad_norm": 0.9571654338821786,
+      "learning_rate": 5.4287388403585275e-06,
+      "loss": 1.5331,
+      "step": 4958
+    },
+    {
+      "epoch": 0.6617293835068054,
+      "grad_norm": 1.0081431486737675,
+      "learning_rate": 5.424895357062826e-06,
+      "loss": 1.5648,
+      "step": 4959
+    },
+    {
+      "epoch": 0.661862823592207,
+      "grad_norm": 1.1640026927531668,
+      "learning_rate": 5.4210527283445824e-06,
+      "loss": 1.5421,
+      "step": 4960
+    },
+    {
+      "epoch": 0.6619962636776088,
+      "grad_norm": 0.9785271528986913,
+      "learning_rate": 5.417210954921557e-06,
+      "loss": 1.5635,
+      "step": 4961
+    },
+    {
+      "epoch": 0.6621297037630104,
+      "grad_norm": 0.9591488914375493,
+      "learning_rate": 5.413370037511347e-06,
+      "loss": 1.5291,
+      "step": 4962
+    },
+    {
+      "epoch": 0.6622631438484121,
+      "grad_norm": 0.9320181389081423,
+      "learning_rate": 5.409529976831392e-06,
+      "loss": 1.5519,
+      "step": 4963
+    },
+    {
+      "epoch": 0.6623965839338137,
+      "grad_norm": 0.9436255066913496,
+      "learning_rate": 5.4056907735989735e-06,
+      "loss": 1.587,
+      "step": 4964
+    },
+    {
+      "epoch": 0.6625300240192153,
+      "grad_norm": 0.9445224040253575,
+      "learning_rate": 5.401852428531212e-06,
+      "loss": 1.616,
+      "step": 4965
+    },
+    {
+      "epoch": 0.6626634641046171,
+      "grad_norm": 1.0814364292819985,
+      "learning_rate": 5.398014942345064e-06,
+      "loss": 1.5734,
+      "step": 4966
+    },
+    {
+      "epoch": 0.6627969041900187,
+      "grad_norm": 0.9584108100106682,
+      "learning_rate": 5.39417831575733e-06,
+      "loss": 1.5685,
+      "step": 4967
+    },
+    {
+      "epoch": 0.6629303442754203,
+      "grad_norm": 0.9665649530241158,
+      "learning_rate": 5.3903425494846485e-06,
+      "loss": 1.6028,
+      "step": 4968
+    },
+    {
+      "epoch": 0.663063784360822,
+      "grad_norm": 0.9365312188882325,
+      "learning_rate": 5.386507644243491e-06,
+      "loss": 1.5945,
+      "step": 4969
+    },
+    {
+      "epoch": 0.6631972244462236,
+      "grad_norm": 0.9729492213835405,
+      "learning_rate": 5.382673600750182e-06,
+      "loss": 1.5775,
+      "step": 4970
+    },
+    {
+      "epoch": 0.6633306645316253,
+      "grad_norm": 1.0805444576468706,
+      "learning_rate": 5.3788404197208744e-06,
+      "loss": 1.5617,
+      "step": 4971
+    },
+    {
+      "epoch": 0.663464104617027,
+      "grad_norm": 0.9617923471221628,
+      "learning_rate": 5.375008101871563e-06,
+      "loss": 1.5363,
+      "step": 4972
+    },
+    {
+      "epoch": 0.6635975447024286,
+      "grad_norm": 0.9481741920611069,
+      "learning_rate": 5.371176647918076e-06,
+      "loss": 1.5512,
+      "step": 4973
+    },
+    {
+      "epoch": 0.6637309847878302,
+      "grad_norm": 0.980896728929079,
+      "learning_rate": 5.367346058576095e-06,
+      "loss": 1.5675,
+      "step": 4974
+    },
+    {
+      "epoch": 0.6638644248732319,
+      "grad_norm": 1.2424778825915792,
+      "learning_rate": 5.363516334561125e-06,
+      "loss": 1.5574,
+      "step": 4975
+    },
+    {
+      "epoch": 0.6639978649586336,
+      "grad_norm": 1.3069446139543355,
+      "learning_rate": 5.359687476588511e-06,
+      "loss": 1.5679,
+      "step": 4976
+    },
+    {
+      "epoch": 0.6641313050440353,
+      "grad_norm": 0.9334235903224387,
+      "learning_rate": 5.355859485373449e-06,
+      "loss": 1.515,
+      "step": 4977
+    },
+    {
+      "epoch": 0.6642647451294369,
+      "grad_norm": 0.9796056460510058,
+      "learning_rate": 5.352032361630959e-06,
+      "loss": 1.5388,
+      "step": 4978
+    },
+    {
+      "epoch": 0.6643981852148385,
+      "grad_norm": 1.0622315589231197,
+      "learning_rate": 5.348206106075906e-06,
+      "loss": 1.6171,
+      "step": 4979
+    },
+    {
+      "epoch": 0.6645316253002402,
+      "grad_norm": 0.9535560005647347,
+      "learning_rate": 5.344380719422985e-06,
+      "loss": 1.5387,
+      "step": 4980
+    },
+    {
+      "epoch": 0.6646650653856419,
+      "grad_norm": 0.9332886144151455,
+      "learning_rate": 5.340556202386743e-06,
+      "loss": 1.5437,
+      "step": 4981
+    },
+    {
+      "epoch": 0.6647985054710435,
+      "grad_norm": 0.926508722116168,
+      "learning_rate": 5.336732555681552e-06,
+      "loss": 1.5249,
+      "step": 4982
+    },
+    {
+      "epoch": 0.6649319455564452,
+      "grad_norm": 0.9597733820765061,
+      "learning_rate": 5.3329097800216244e-06,
+      "loss": 1.6112,
+      "step": 4983
+    },
+    {
+      "epoch": 0.6650653856418468,
+      "grad_norm": 0.9263015315418155,
+      "learning_rate": 5.329087876121016e-06,
+      "loss": 1.5548,
+      "step": 4984
+    },
+    {
+      "epoch": 0.6651988257272484,
+      "grad_norm": 0.9197318999051447,
+      "learning_rate": 5.325266844693611e-06,
+      "loss": 1.5108,
+      "step": 4985
+    },
+    {
+      "epoch": 0.6653322658126501,
+      "grad_norm": 0.9346342407015626,
+      "learning_rate": 5.321446686453137e-06,
+      "loss": 1.5587,
+      "step": 4986
+    },
+    {
+      "epoch": 0.6654657058980518,
+      "grad_norm": 0.9609313535996501,
+      "learning_rate": 5.31762740211315e-06,
+      "loss": 1.5212,
+      "step": 4987
+    },
+    {
+      "epoch": 0.6655991459834534,
+      "grad_norm": 0.9379253903460791,
+      "learning_rate": 5.313808992387057e-06,
+      "loss": 1.5553,
+      "step": 4988
+    },
+    {
+      "epoch": 0.6657325860688551,
+      "grad_norm": 0.9452683100970093,
+      "learning_rate": 5.309991457988091e-06,
+      "loss": 1.5827,
+      "step": 4989
+    },
+    {
+      "epoch": 0.6658660261542567,
+      "grad_norm": 0.9871345858341555,
+      "learning_rate": 5.306174799629317e-06,
+      "loss": 1.5048,
+      "step": 4990
+    },
+    {
+      "epoch": 0.6659994662396584,
+      "grad_norm": 0.9745323174519458,
+      "learning_rate": 5.302359018023656e-06,
+      "loss": 1.5655,
+      "step": 4991
+    },
+    {
+      "epoch": 0.6661329063250601,
+      "grad_norm": 1.0874975316674835,
+      "learning_rate": 5.298544113883845e-06,
+      "loss": 1.5807,
+      "step": 4992
+    },
+    {
+      "epoch": 0.6662663464104617,
+      "grad_norm": 0.9359689028295645,
+      "learning_rate": 5.294730087922464e-06,
+      "loss": 1.555,
+      "step": 4993
+    },
+    {
+      "epoch": 0.6663997864958634,
+      "grad_norm": 1.065007282594557,
+      "learning_rate": 5.290916940851929e-06,
+      "loss": 1.5666,
+      "step": 4994
+    },
+    {
+      "epoch": 0.666533226581265,
+      "grad_norm": 1.1502681142848064,
+      "learning_rate": 5.287104673384498e-06,
+      "loss": 1.5713,
+      "step": 4995
+    },
+    {
+      "epoch": 0.6666666666666666,
+      "grad_norm": 0.9479228632028691,
+      "learning_rate": 5.283293286232254e-06,
+      "loss": 1.5684,
+      "step": 4996
+    },
+    {
+      "epoch": 0.6668001067520684,
+      "grad_norm": 0.9987657361216116,
+      "learning_rate": 5.279482780107119e-06,
+      "loss": 1.5635,
+      "step": 4997
+    },
+    {
+      "epoch": 0.66693354683747,
+      "grad_norm": 1.1782973127657173,
+      "learning_rate": 5.275673155720861e-06,
+      "loss": 1.5947,
+      "step": 4998
+    },
+    {
+      "epoch": 0.6670669869228716,
+      "grad_norm": 0.9525996465669817,
+      "learning_rate": 5.271864413785068e-06,
+      "loss": 1.5657,
+      "step": 4999
+    },
+    {
+      "epoch": 0.6672004270082733,
+      "grad_norm": 0.9683386220692802,
+      "learning_rate": 5.26805655501117e-06,
+      "loss": 1.5778,
+      "step": 5000
+    },
+    {
+      "epoch": 0.6673338670936749,
+      "grad_norm": 1.0500879084213508,
+      "learning_rate": 5.26424958011043e-06,
+      "loss": 1.5636,
+      "step": 5001
+    },
+    {
+      "epoch": 0.6674673071790765,
+      "grad_norm": 0.9531705010447756,
+      "learning_rate": 5.260443489793953e-06,
+      "loss": 1.5954,
+      "step": 5002
+    },
+    {
+      "epoch": 0.6676007472644783,
+      "grad_norm": 0.9421433711002707,
+      "learning_rate": 5.25663828477267e-06,
+      "loss": 1.5517,
+      "step": 5003
+    },
+    {
+      "epoch": 0.6677341873498799,
+      "grad_norm": 0.9278026930537797,
+      "learning_rate": 5.252833965757351e-06,
+      "loss": 1.583,
+      "step": 5004
+    },
+    {
+      "epoch": 0.6678676274352816,
+      "grad_norm": 0.9463961992902421,
+      "learning_rate": 5.249030533458594e-06,
+      "loss": 1.5259,
+      "step": 5005
+    },
+    {
+      "epoch": 0.6680010675206832,
+      "grad_norm": 1.019719000874434,
+      "learning_rate": 5.245227988586845e-06,
+      "loss": 1.5647,
+      "step": 5006
+    },
+    {
+      "epoch": 0.6681345076060848,
+      "grad_norm": 0.9199080705366238,
+      "learning_rate": 5.241426331852373e-06,
+      "loss": 1.5497,
+      "step": 5007
+    },
+    {
+      "epoch": 0.6682679476914866,
+      "grad_norm": 0.9388293679619114,
+      "learning_rate": 5.237625563965285e-06,
+      "loss": 1.512,
+      "step": 5008
+    },
+    {
+      "epoch": 0.6684013877768882,
+      "grad_norm": 0.9772164529249927,
+      "learning_rate": 5.233825685635518e-06,
+      "loss": 1.5984,
+      "step": 5009
+    },
+    {
+      "epoch": 0.6685348278622898,
+      "grad_norm": 1.3066296901302348,
+      "learning_rate": 5.23002669757285e-06,
+      "loss": 1.5724,
+      "step": 5010
+    },
+    {
+      "epoch": 0.6686682679476915,
+      "grad_norm": 1.1126937385329367,
+      "learning_rate": 5.226228600486883e-06,
+      "loss": 1.5821,
+      "step": 5011
+    },
+    {
+      "epoch": 0.6688017080330931,
+      "grad_norm": 0.9120825110494507,
+      "learning_rate": 5.222431395087064e-06,
+      "loss": 1.5863,
+      "step": 5012
+    },
+    {
+      "epoch": 0.6689351481184947,
+      "grad_norm": 0.9636066977274008,
+      "learning_rate": 5.218635082082666e-06,
+      "loss": 1.592,
+      "step": 5013
+    },
+    {
+      "epoch": 0.6690685882038965,
+      "grad_norm": 0.9146764316842598,
+      "learning_rate": 5.214839662182798e-06,
+      "loss": 1.5786,
+      "step": 5014
+    },
+    {
+      "epoch": 0.6692020282892981,
+      "grad_norm": 0.9588946756171576,
+      "learning_rate": 5.211045136096394e-06,
+      "loss": 1.5927,
+      "step": 5015
+    },
+    {
+      "epoch": 0.6693354683746997,
+      "grad_norm": 0.944621271446456,
+      "learning_rate": 5.20725150453224e-06,
+      "loss": 1.51,
+      "step": 5016
+    },
+    {
+      "epoch": 0.6694689084601014,
+      "grad_norm": 0.9688157625406866,
+      "learning_rate": 5.2034587681989364e-06,
+      "loss": 1.5454,
+      "step": 5017
+    },
+    {
+      "epoch": 0.669602348545503,
+      "grad_norm": 0.9487031537844105,
+      "learning_rate": 5.199666927804925e-06,
+      "loss": 1.5289,
+      "step": 5018
+    },
+    {
+      "epoch": 0.6697357886309048,
+      "grad_norm": 0.9892335986551877,
+      "learning_rate": 5.195875984058474e-06,
+      "loss": 1.5918,
+      "step": 5019
+    },
+    {
+      "epoch": 0.6698692287163064,
+      "grad_norm": 0.9044754479784632,
+      "learning_rate": 5.192085937667696e-06,
+      "loss": 1.6101,
+      "step": 5020
+    },
+    {
+      "epoch": 0.670002668801708,
+      "grad_norm": 0.9397019351239638,
+      "learning_rate": 5.188296789340523e-06,
+      "loss": 1.5,
+      "step": 5021
+    },
+    {
+      "epoch": 0.6701361088871097,
+      "grad_norm": 0.9823103967616779,
+      "learning_rate": 5.184508539784725e-06,
+      "loss": 1.6176,
+      "step": 5022
+    },
+    {
+      "epoch": 0.6702695489725113,
+      "grad_norm": 0.9220689248773222,
+      "learning_rate": 5.180721189707908e-06,
+      "loss": 1.5349,
+      "step": 5023
+    },
+    {
+      "epoch": 0.670402989057913,
+      "grad_norm": 0.9481270673997315,
+      "learning_rate": 5.176934739817503e-06,
+      "loss": 1.5678,
+      "step": 5024
+    },
+    {
+      "epoch": 0.6705364291433147,
+      "grad_norm": 1.0396242680439707,
+      "learning_rate": 5.173149190820776e-06,
+      "loss": 1.6279,
+      "step": 5025
+    },
+    {
+      "epoch": 0.6706698692287163,
+      "grad_norm": 0.9646160121743006,
+      "learning_rate": 5.1693645434248216e-06,
+      "loss": 1.5928,
+      "step": 5026
+    },
+    {
+      "epoch": 0.6708033093141179,
+      "grad_norm": 0.9553746604638392,
+      "learning_rate": 5.165580798336575e-06,
+      "loss": 1.5881,
+      "step": 5027
+    },
+    {
+      "epoch": 0.6709367493995196,
+      "grad_norm": 0.9360288710325949,
+      "learning_rate": 5.161797956262793e-06,
+      "loss": 1.6201,
+      "step": 5028
+    },
+    {
+      "epoch": 0.6710701894849213,
+      "grad_norm": 1.0854046179543964,
+      "learning_rate": 5.158016017910064e-06,
+      "loss": 1.5416,
+      "step": 5029
+    },
+    {
+      "epoch": 0.671203629570323,
+      "grad_norm": 0.9591439987959153,
+      "learning_rate": 5.154234983984818e-06,
+      "loss": 1.5665,
+      "step": 5030
+    },
+    {
+      "epoch": 0.6713370696557246,
+      "grad_norm": 1.0881353005554923,
+      "learning_rate": 5.150454855193308e-06,
+      "loss": 1.5806,
+      "step": 5031
+    },
+    {
+      "epoch": 0.6714705097411262,
+      "grad_norm": 1.0095150308218221,
+      "learning_rate": 5.146675632241614e-06,
+      "loss": 1.5884,
+      "step": 5032
+    },
+    {
+      "epoch": 0.6716039498265279,
+      "grad_norm": 0.9853452337837284,
+      "learning_rate": 5.142897315835653e-06,
+      "loss": 1.5928,
+      "step": 5033
+    },
+    {
+      "epoch": 0.6717373899119295,
+      "grad_norm": 0.9114971278210366,
+      "learning_rate": 5.139119906681176e-06,
+      "loss": 1.5162,
+      "step": 5034
+    },
+    {
+      "epoch": 0.6718708299973312,
+      "grad_norm": 0.970853734583226,
+      "learning_rate": 5.135343405483757e-06,
+      "loss": 1.555,
+      "step": 5035
+    },
+    {
+      "epoch": 0.6720042700827329,
+      "grad_norm": 0.9830827365486369,
+      "learning_rate": 5.131567812948805e-06,
+      "loss": 1.5957,
+      "step": 5036
+    },
+    {
+      "epoch": 0.6721377101681345,
+      "grad_norm": 1.0109831996031944,
+      "learning_rate": 5.127793129781551e-06,
+      "loss": 1.6049,
+      "step": 5037
+    },
+    {
+      "epoch": 0.6722711502535361,
+      "grad_norm": 0.9273191953375031,
+      "learning_rate": 5.124019356687073e-06,
+      "loss": 1.5482,
+      "step": 5038
+    },
+    {
+      "epoch": 0.6724045903389378,
+      "grad_norm": 0.9530762057231332,
+      "learning_rate": 5.120246494370264e-06,
+      "loss": 1.557,
+      "step": 5039
+    },
+    {
+      "epoch": 0.6725380304243395,
+      "grad_norm": 1.009716414085073,
+      "learning_rate": 5.116474543535848e-06,
+      "loss": 1.5817,
+      "step": 5040
+    },
+    {
+      "epoch": 0.6726714705097411,
+      "grad_norm": 0.936535223545121,
+      "learning_rate": 5.11270350488839e-06,
+      "loss": 1.5498,
+      "step": 5041
+    },
+    {
+      "epoch": 0.6728049105951428,
+      "grad_norm": 0.9587135356454837,
+      "learning_rate": 5.108933379132272e-06,
+      "loss": 1.6064,
+      "step": 5042
+    },
+    {
+      "epoch": 0.6729383506805444,
+      "grad_norm": 0.9707574935271643,
+      "learning_rate": 5.105164166971714e-06,
+      "loss": 1.4911,
+      "step": 5043
+    },
+    {
+      "epoch": 0.6730717907659461,
+      "grad_norm": 0.9255255572906976,
+      "learning_rate": 5.101395869110755e-06,
+      "loss": 1.5974,
+      "step": 5044
+    },
+    {
+      "epoch": 0.6732052308513478,
+      "grad_norm": 0.9526351690279754,
+      "learning_rate": 5.097628486253278e-06,
+      "loss": 1.583,
+      "step": 5045
+    },
+    {
+      "epoch": 0.6733386709367494,
+      "grad_norm": 0.9231045207080708,
+      "learning_rate": 5.0938620191029865e-06,
+      "loss": 1.571,
+      "step": 5046
+    },
+    {
+      "epoch": 0.6734721110221511,
+      "grad_norm": 0.9515238827519511,
+      "learning_rate": 5.090096468363409e-06,
+      "loss": 1.6318,
+      "step": 5047
+    },
+    {
+      "epoch": 0.6736055511075527,
+      "grad_norm": 0.9356804828343049,
+      "learning_rate": 5.086331834737908e-06,
+      "loss": 1.5706,
+      "step": 5048
+    },
+    {
+      "epoch": 0.6737389911929543,
+      "grad_norm": 1.1128834664193885,
+      "learning_rate": 5.082568118929678e-06,
+      "loss": 1.5895,
+      "step": 5049
+    },
+    {
+      "epoch": 0.673872431278356,
+      "grad_norm": 1.1955255342152211,
+      "learning_rate": 5.078805321641736e-06,
+      "loss": 1.5635,
+      "step": 5050
+    },
+    {
+      "epoch": 0.6740058713637577,
+      "grad_norm": 0.9185509120855007,
+      "learning_rate": 5.075043443576933e-06,
+      "loss": 1.5417,
+      "step": 5051
+    },
+    {
+      "epoch": 0.6741393114491593,
+      "grad_norm": 0.9576951089769443,
+      "learning_rate": 5.071282485437938e-06,
+      "loss": 1.5684,
+      "step": 5052
+    },
+    {
+      "epoch": 0.674272751534561,
+      "grad_norm": 0.9875819130260579,
+      "learning_rate": 5.067522447927262e-06,
+      "loss": 1.5387,
+      "step": 5053
+    },
+    {
+      "epoch": 0.6744061916199626,
+      "grad_norm": 1.0759454029563662,
+      "learning_rate": 5.06376333174723e-06,
+      "loss": 1.5579,
+      "step": 5054
+    },
+    {
+      "epoch": 0.6745396317053642,
+      "grad_norm": 0.9953780319051102,
+      "learning_rate": 5.06000513760001e-06,
+      "loss": 1.5767,
+      "step": 5055
+    },
+    {
+      "epoch": 0.674673071790766,
+      "grad_norm": 0.9538482284664277,
+      "learning_rate": 5.056247866187587e-06,
+      "loss": 1.6003,
+      "step": 5056
+    },
+    {
+      "epoch": 0.6748065118761676,
+      "grad_norm": 0.9578854545173783,
+      "learning_rate": 5.0524915182117754e-06,
+      "loss": 1.5625,
+      "step": 5057
+    },
+    {
+      "epoch": 0.6749399519615693,
+      "grad_norm": 0.9333496781315979,
+      "learning_rate": 5.0487360943742135e-06,
+      "loss": 1.5545,
+      "step": 5058
+    },
+    {
+      "epoch": 0.6750733920469709,
+      "grad_norm": 0.936433357559595,
+      "learning_rate": 5.044981595376383e-06,
+      "loss": 1.571,
+      "step": 5059
+    },
+    {
+      "epoch": 0.6752068321323725,
+      "grad_norm": 0.9366758168083295,
+      "learning_rate": 5.0412280219195746e-06,
+      "loss": 1.6242,
+      "step": 5060
+    },
+    {
+      "epoch": 0.6753402722177743,
+      "grad_norm": 1.0171664676564722,
+      "learning_rate": 5.03747537470491e-06,
+      "loss": 1.578,
+      "step": 5061
+    },
+    {
+      "epoch": 0.6754737123031759,
+      "grad_norm": 1.2233118469857216,
+      "learning_rate": 5.033723654433349e-06,
+      "loss": 1.577,
+      "step": 5062
+    },
+    {
+      "epoch": 0.6756071523885775,
+      "grad_norm": 0.9036663728010922,
+      "learning_rate": 5.029972861805665e-06,
+      "loss": 1.522,
+      "step": 5063
+    },
+    {
+      "epoch": 0.6757405924739792,
+      "grad_norm": 0.9771449490125018,
+      "learning_rate": 5.026222997522465e-06,
+      "loss": 1.5776,
+      "step": 5064
+    },
+    {
+      "epoch": 0.6758740325593808,
+      "grad_norm": 0.9516286727943922,
+      "learning_rate": 5.022474062284177e-06,
+      "loss": 1.5614,
+      "step": 5065
+    },
+    {
+      "epoch": 0.6760074726447824,
+      "grad_norm": 0.9177876450126852,
+      "learning_rate": 5.018726056791068e-06,
+      "loss": 1.5596,
+      "step": 5066
+    },
+    {
+      "epoch": 0.6761409127301842,
+      "grad_norm": 0.9415870237949052,
+      "learning_rate": 5.014978981743216e-06,
+      "loss": 1.5728,
+      "step": 5067
+    },
+    {
+      "epoch": 0.6762743528155858,
+      "grad_norm": 0.9378316877697337,
+      "learning_rate": 5.011232837840534e-06,
+      "loss": 1.5742,
+      "step": 5068
+    },
+    {
+      "epoch": 0.6764077929009874,
+      "grad_norm": 0.9740662444000914,
+      "learning_rate": 5.007487625782755e-06,
+      "loss": 1.5946,
+      "step": 5069
+    },
+    {
+      "epoch": 0.6765412329863891,
+      "grad_norm": 0.955131305489453,
+      "learning_rate": 5.003743346269449e-06,
+      "loss": 1.5925,
+      "step": 5070
+    },
+    {
+      "epoch": 0.6766746730717907,
+      "grad_norm": 0.9317999097948068,
+      "learning_rate": 5.000000000000003e-06,
+      "loss": 1.5604,
+      "step": 5071
+    },
+    {
+      "epoch": 0.6768081131571925,
+      "grad_norm": 0.9365536795010426,
+      "learning_rate": 4.9962575876736245e-06,
+      "loss": 1.6174,
+      "step": 5072
+    },
+    {
+      "epoch": 0.6769415532425941,
+      "grad_norm": 1.0580187577949391,
+      "learning_rate": 4.992516109989362e-06,
+      "loss": 1.5542,
+      "step": 5073
+    },
+    {
+      "epoch": 0.6770749933279957,
+      "grad_norm": 0.9475023984254052,
+      "learning_rate": 4.988775567646079e-06,
+      "loss": 1.5899,
+      "step": 5074
+    },
+    {
+      "epoch": 0.6772084334133974,
+      "grad_norm": 1.0674090359857158,
+      "learning_rate": 4.985035961342466e-06,
+      "loss": 1.5677,
+      "step": 5075
+    },
+    {
+      "epoch": 0.677341873498799,
+      "grad_norm": 0.9866266273983164,
+      "learning_rate": 4.981297291777032e-06,
+      "loss": 1.5378,
+      "step": 5076
+    },
+    {
+      "epoch": 0.6774753135842007,
+      "grad_norm": 0.9868827012284229,
+      "learning_rate": 4.977559559648128e-06,
+      "loss": 1.5811,
+      "step": 5077
+    },
+    {
+      "epoch": 0.6776087536696024,
+      "grad_norm": 1.1780046727927997,
+      "learning_rate": 4.973822765653917e-06,
+      "loss": 1.4914,
+      "step": 5078
+    },
+    {
+      "epoch": 0.677742193755004,
+      "grad_norm": 0.95257815072713,
+      "learning_rate": 4.970086910492385e-06,
+      "loss": 1.5842,
+      "step": 5079
+    },
+    {
+      "epoch": 0.6778756338404056,
+      "grad_norm": 1.0663204797517163,
+      "learning_rate": 4.966351994861352e-06,
+      "loss": 1.6202,
+      "step": 5080
+    },
+    {
+      "epoch": 0.6780090739258073,
+      "grad_norm": 0.935486976169621,
+      "learning_rate": 4.962618019458456e-06,
+      "loss": 1.5488,
+      "step": 5081
+    },
+    {
+      "epoch": 0.678142514011209,
+      "grad_norm": 0.9770535662216747,
+      "learning_rate": 4.958884984981163e-06,
+      "loss": 1.5611,
+      "step": 5082
+    },
+    {
+      "epoch": 0.6782759540966106,
+      "grad_norm": 0.918514304568767,
+      "learning_rate": 4.9551528921267545e-06,
+      "loss": 1.5666,
+      "step": 5083
+    },
+    {
+      "epoch": 0.6784093941820123,
+      "grad_norm": 0.9318149736553532,
+      "learning_rate": 4.951421741592353e-06,
+      "loss": 1.5857,
+      "step": 5084
+    },
+    {
+      "epoch": 0.6785428342674139,
+      "grad_norm": 13.688631168582868,
+      "learning_rate": 4.947691534074889e-06,
+      "loss": 1.6078,
+      "step": 5085
+    },
+    {
+      "epoch": 0.6786762743528156,
+      "grad_norm": 1.002752650809866,
+      "learning_rate": 4.94396227027112e-06,
+      "loss": 1.516,
+      "step": 5086
+    },
+    {
+      "epoch": 0.6788097144382172,
+      "grad_norm": 1.023591072044742,
+      "learning_rate": 4.940233950877637e-06,
+      "loss": 1.5348,
+      "step": 5087
+    },
+    {
+      "epoch": 0.6789431545236189,
+      "grad_norm": 1.1764691952827722,
+      "learning_rate": 4.936506576590846e-06,
+      "loss": 1.5717,
+      "step": 5088
+    },
+    {
+      "epoch": 0.6790765946090206,
+      "grad_norm": 1.029025311630692,
+      "learning_rate": 4.932780148106975e-06,
+      "loss": 1.5159,
+      "step": 5089
+    },
+    {
+      "epoch": 0.6792100346944222,
+      "grad_norm": 0.950650193469443,
+      "learning_rate": 4.929054666122081e-06,
+      "loss": 1.5447,
+      "step": 5090
+    },
+    {
+      "epoch": 0.6793434747798238,
+      "grad_norm": 1.1799569092706195,
+      "learning_rate": 4.9253301313320355e-06,
+      "loss": 1.5769,
+      "step": 5091
+    },
+    {
+      "epoch": 0.6794769148652255,
+      "grad_norm": 0.9432993878667583,
+      "learning_rate": 4.921606544432549e-06,
+      "loss": 1.5891,
+      "step": 5092
+    },
+    {
+      "epoch": 0.6796103549506272,
+      "grad_norm": 0.9465085033881826,
+      "learning_rate": 4.91788390611914e-06,
+      "loss": 1.5216,
+      "step": 5093
+    },
+    {
+      "epoch": 0.6797437950360288,
+      "grad_norm": 1.0542649886569853,
+      "learning_rate": 4.914162217087155e-06,
+      "loss": 1.5736,
+      "step": 5094
+    },
+    {
+      "epoch": 0.6798772351214305,
+      "grad_norm": 1.0000819446908524,
+      "learning_rate": 4.910441478031763e-06,
+      "loss": 1.6619,
+      "step": 5095
+    },
+    {
+      "epoch": 0.6800106752068321,
+      "grad_norm": 0.9322943527809388,
+      "learning_rate": 4.906721689647957e-06,
+      "loss": 1.5867,
+      "step": 5096
+    },
+    {
+      "epoch": 0.6801441152922338,
+      "grad_norm": 0.9464543046875021,
+      "learning_rate": 4.903002852630546e-06,
+      "loss": 1.5638,
+      "step": 5097
+    },
+    {
+      "epoch": 0.6802775553776355,
+      "grad_norm": 0.9523084416197721,
+      "learning_rate": 4.899284967674174e-06,
+      "loss": 1.4874,
+      "step": 5098
+    },
+    {
+      "epoch": 0.6804109954630371,
+      "grad_norm": 0.9259089895023304,
+      "learning_rate": 4.895568035473297e-06,
+      "loss": 1.575,
+      "step": 5099
+    },
+    {
+      "epoch": 0.6805444355484388,
+      "grad_norm": 0.913018189144235,
+      "learning_rate": 4.891852056722195e-06,
+      "loss": 1.5759,
+      "step": 5100
+    },
+    {
+      "epoch": 0.6806778756338404,
+      "grad_norm": 0.9351223685450596,
+      "learning_rate": 4.888137032114966e-06,
+      "loss": 1.552,
+      "step": 5101
+    },
+    {
+      "epoch": 0.680811315719242,
+      "grad_norm": 1.1155154706231596,
+      "learning_rate": 4.884422962345542e-06,
+      "loss": 1.5482,
+      "step": 5102
+    },
+    {
+      "epoch": 0.6809447558046438,
+      "grad_norm": 0.9366870347444463,
+      "learning_rate": 4.880709848107666e-06,
+      "loss": 1.5683,
+      "step": 5103
+    },
+    {
+      "epoch": 0.6810781958900454,
+      "grad_norm": 0.972915796830404,
+      "learning_rate": 4.876997690094902e-06,
+      "loss": 1.5796,
+      "step": 5104
+    },
+    {
+      "epoch": 0.681211635975447,
+      "grad_norm": 0.9605649263345989,
+      "learning_rate": 4.873286489000645e-06,
+      "loss": 1.4809,
+      "step": 5105
+    },
+    {
+      "epoch": 0.6813450760608487,
+      "grad_norm": 1.0145828253220486,
+      "learning_rate": 4.869576245518101e-06,
+      "loss": 1.5529,
+      "step": 5106
+    },
+    {
+      "epoch": 0.6814785161462503,
+      "grad_norm": 1.0299976481166175,
+      "learning_rate": 4.865866960340304e-06,
+      "loss": 1.556,
+      "step": 5107
+    },
+    {
+      "epoch": 0.6816119562316519,
+      "grad_norm": 0.9634555693945733,
+      "learning_rate": 4.862158634160101e-06,
+      "loss": 1.5617,
+      "step": 5108
+    },
+    {
+      "epoch": 0.6817453963170537,
+      "grad_norm": 1.0182839375048582,
+      "learning_rate": 4.858451267670171e-06,
+      "loss": 1.5827,
+      "step": 5109
+    },
+    {
+      "epoch": 0.6818788364024553,
+      "grad_norm": 1.1453431192773607,
+      "learning_rate": 4.854744861563007e-06,
+      "loss": 1.5854,
+      "step": 5110
+    },
+    {
+      "epoch": 0.682012276487857,
+      "grad_norm": 0.9803243409985238,
+      "learning_rate": 4.851039416530917e-06,
+      "loss": 1.5329,
+      "step": 5111
+    },
+    {
+      "epoch": 0.6821457165732586,
+      "grad_norm": 1.016213296072417,
+      "learning_rate": 4.847334933266044e-06,
+      "loss": 1.6221,
+      "step": 5112
+    },
+    {
+      "epoch": 0.6822791566586602,
+      "grad_norm": 0.9505609706075121,
+      "learning_rate": 4.843631412460341e-06,
+      "loss": 1.5815,
+      "step": 5113
+    },
+    {
+      "epoch": 0.682412596744062,
+      "grad_norm": 1.0060585154404742,
+      "learning_rate": 4.839928854805583e-06,
+      "loss": 1.5788,
+      "step": 5114
+    },
+    {
+      "epoch": 0.6825460368294636,
+      "grad_norm": 3.7584015873744785,
+      "learning_rate": 4.836227260993361e-06,
+      "loss": 1.5487,
+      "step": 5115
+    },
+    {
+      "epoch": 0.6826794769148652,
+      "grad_norm": 0.9458041427982877,
+      "learning_rate": 4.8325266317150996e-06,
+      "loss": 1.5569,
+      "step": 5116
+    },
+    {
+      "epoch": 0.6828129170002669,
+      "grad_norm": 1.0741204494235426,
+      "learning_rate": 4.828826967662028e-06,
+      "loss": 1.5389,
+      "step": 5117
+    },
+    {
+      "epoch": 0.6829463570856685,
+      "grad_norm": 1.0796007293741565,
+      "learning_rate": 4.8251282695252e-06,
+      "loss": 1.5467,
+      "step": 5118
+    },
+    {
+      "epoch": 0.6830797971710701,
+      "grad_norm": 0.984104078234117,
+      "learning_rate": 4.821430537995497e-06,
+      "loss": 1.5582,
+      "step": 5119
+    },
+    {
+      "epoch": 0.6832132372564719,
+      "grad_norm": 0.9768350794179331,
+      "learning_rate": 4.817733773763609e-06,
+      "loss": 1.5747,
+      "step": 5120
+    },
+    {
+      "epoch": 0.6833466773418735,
+      "grad_norm": 0.97701351831298,
+      "learning_rate": 4.8140379775200494e-06,
+      "loss": 1.562,
+      "step": 5121
+    },
+    {
+      "epoch": 0.6834801174272751,
+      "grad_norm": 1.1798523816680644,
+      "learning_rate": 4.8103431499551474e-06,
+      "loss": 1.5862,
+      "step": 5122
+    },
+    {
+      "epoch": 0.6836135575126768,
+      "grad_norm": 1.0067654037143134,
+      "learning_rate": 4.8066492917590635e-06,
+      "loss": 1.594,
+      "step": 5123
+    },
+    {
+      "epoch": 0.6837469975980784,
+      "grad_norm": 0.9823447933349997,
+      "learning_rate": 4.802956403621763e-06,
+      "loss": 1.5824,
+      "step": 5124
+    },
+    {
+      "epoch": 0.6838804376834802,
+      "grad_norm": 0.9800571652551026,
+      "learning_rate": 4.799264486233031e-06,
+      "loss": 1.5426,
+      "step": 5125
+    },
+    {
+      "epoch": 0.6840138777688818,
+      "grad_norm": 0.9373855164532967,
+      "learning_rate": 4.795573540282484e-06,
+      "loss": 1.5959,
+      "step": 5126
+    },
+    {
+      "epoch": 0.6841473178542834,
+      "grad_norm": 1.1258697616784983,
+      "learning_rate": 4.791883566459546e-06,
+      "loss": 1.5546,
+      "step": 5127
+    },
+    {
+      "epoch": 0.6842807579396851,
+      "grad_norm": 1.0460314104044222,
+      "learning_rate": 4.7881945654534615e-06,
+      "loss": 1.5461,
+      "step": 5128
+    },
+    {
+      "epoch": 0.6844141980250867,
+      "grad_norm": 0.9470018047113221,
+      "learning_rate": 4.78450653795329e-06,
+      "loss": 1.5871,
+      "step": 5129
+    },
+    {
+      "epoch": 0.6845476381104884,
+      "grad_norm": 0.9381571561346944,
+      "learning_rate": 4.780819484647918e-06,
+      "loss": 1.5455,
+      "step": 5130
+    },
+    {
+      "epoch": 0.6846810781958901,
+      "grad_norm": 1.1209989432461438,
+      "learning_rate": 4.777133406226046e-06,
+      "loss": 1.6104,
+      "step": 5131
+    },
+    {
+      "epoch": 0.6848145182812917,
+      "grad_norm": 0.9861145590915344,
+      "learning_rate": 4.773448303376188e-06,
+      "loss": 1.6122,
+      "step": 5132
+    },
+    {
+      "epoch": 0.6849479583666933,
+      "grad_norm": 0.9632711224553404,
+      "learning_rate": 4.769764176786681e-06,
+      "loss": 1.5319,
+      "step": 5133
+    },
+    {
+      "epoch": 0.685081398452095,
+      "grad_norm": 1.0114719533896925,
+      "learning_rate": 4.7660810271456735e-06,
+      "loss": 1.5546,
+      "step": 5134
+    },
+    {
+      "epoch": 0.6852148385374967,
+      "grad_norm": 0.9449667280372048,
+      "learning_rate": 4.762398855141143e-06,
+      "loss": 1.4989,
+      "step": 5135
+    },
+    {
+      "epoch": 0.6853482786228983,
+      "grad_norm": 0.925328878656421,
+      "learning_rate": 4.758717661460874e-06,
+      "loss": 1.5486,
+      "step": 5136
+    },
+    {
+      "epoch": 0.6854817187083,
+      "grad_norm": 0.9589142342169662,
+      "learning_rate": 4.755037446792472e-06,
+      "loss": 1.5787,
+      "step": 5137
+    },
+    {
+      "epoch": 0.6856151587937016,
+      "grad_norm": 0.9539174340381931,
+      "learning_rate": 4.751358211823359e-06,
+      "loss": 1.4947,
+      "step": 5138
+    },
+    {
+      "epoch": 0.6857485988791033,
+      "grad_norm": 0.9935890025844325,
+      "learning_rate": 4.747679957240774e-06,
+      "loss": 1.5835,
+      "step": 5139
+    },
+    {
+      "epoch": 0.685882038964505,
+      "grad_norm": 0.9561138767634939,
+      "learning_rate": 4.744002683731769e-06,
+      "loss": 1.5068,
+      "step": 5140
+    },
+    {
+      "epoch": 0.6860154790499066,
+      "grad_norm": 1.0881008249779185,
+      "learning_rate": 4.740326391983225e-06,
+      "loss": 1.5547,
+      "step": 5141
+    },
+    {
+      "epoch": 0.6861489191353083,
+      "grad_norm": 1.0331157860816333,
+      "learning_rate": 4.736651082681826e-06,
+      "loss": 1.5485,
+      "step": 5142
+    },
+    {
+      "epoch": 0.6862823592207099,
+      "grad_norm": 0.9726155072774376,
+      "learning_rate": 4.732976756514077e-06,
+      "loss": 1.5666,
+      "step": 5143
+    },
+    {
+      "epoch": 0.6864157993061115,
+      "grad_norm": 0.9698903915052955,
+      "learning_rate": 4.729303414166305e-06,
+      "loss": 1.5705,
+      "step": 5144
+    },
+    {
+      "epoch": 0.6865492393915132,
+      "grad_norm": 0.9739962559971576,
+      "learning_rate": 4.7256310563246475e-06,
+      "loss": 1.5984,
+      "step": 5145
+    },
+    {
+      "epoch": 0.6866826794769149,
+      "grad_norm": 1.0147943743702117,
+      "learning_rate": 4.721959683675057e-06,
+      "loss": 1.5452,
+      "step": 5146
+    },
+    {
+      "epoch": 0.6868161195623165,
+      "grad_norm": 0.9785035074553505,
+      "learning_rate": 4.718289296903301e-06,
+      "loss": 1.5578,
+      "step": 5147
+    },
+    {
+      "epoch": 0.6869495596477182,
+      "grad_norm": 0.9329927814076208,
+      "learning_rate": 4.714619896694974e-06,
+      "loss": 1.5801,
+      "step": 5148
+    },
+    {
+      "epoch": 0.6870829997331198,
+      "grad_norm": 0.9379344188241905,
+      "learning_rate": 4.710951483735474e-06,
+      "loss": 1.5543,
+      "step": 5149
+    },
+    {
+      "epoch": 0.6872164398185214,
+      "grad_norm": 1.2323367929741,
+      "learning_rate": 4.7072840587100146e-06,
+      "loss": 1.5424,
+      "step": 5150
+    },
+    {
+      "epoch": 0.6873498799039232,
+      "grad_norm": 0.9866412354494483,
+      "learning_rate": 4.7036176223036375e-06,
+      "loss": 1.5593,
+      "step": 5151
+    },
+    {
+      "epoch": 0.6874833199893248,
+      "grad_norm": 1.0200114322334428,
+      "learning_rate": 4.699952175201187e-06,
+      "loss": 1.5624,
+      "step": 5152
+    },
+    {
+      "epoch": 0.6876167600747265,
+      "grad_norm": 1.0952002464406705,
+      "learning_rate": 4.696287718087326e-06,
+      "loss": 1.5471,
+      "step": 5153
+    },
+    {
+      "epoch": 0.6877502001601281,
+      "grad_norm": 0.9626777049116638,
+      "learning_rate": 4.6926242516465315e-06,
+      "loss": 1.5889,
+      "step": 5154
+    },
+    {
+      "epoch": 0.6878836402455297,
+      "grad_norm": 0.9768971512913973,
+      "learning_rate": 4.688961776563102e-06,
+      "loss": 1.5778,
+      "step": 5155
+    },
+    {
+      "epoch": 0.6880170803309315,
+      "grad_norm": 1.0154475915685601,
+      "learning_rate": 4.685300293521146e-06,
+      "loss": 1.5428,
+      "step": 5156
+    },
+    {
+      "epoch": 0.6881505204163331,
+      "grad_norm": 0.9175407597697326,
+      "learning_rate": 4.68163980320458e-06,
+      "loss": 1.5406,
+      "step": 5157
+    },
+    {
+      "epoch": 0.6882839605017347,
+      "grad_norm": 0.9219686003018891,
+      "learning_rate": 4.677980306297153e-06,
+      "loss": 1.5645,
+      "step": 5158
+    },
+    {
+      "epoch": 0.6884174005871364,
+      "grad_norm": 0.9931694076801224,
+      "learning_rate": 4.674321803482409e-06,
+      "loss": 1.5528,
+      "step": 5159
+    },
+    {
+      "epoch": 0.688550840672538,
+      "grad_norm": 0.9871448513395474,
+      "learning_rate": 4.670664295443718e-06,
+      "loss": 1.5606,
+      "step": 5160
+    },
+    {
+      "epoch": 0.6886842807579396,
+      "grad_norm": 0.9678884161793881,
+      "learning_rate": 4.667007782864256e-06,
+      "loss": 1.6019,
+      "step": 5161
+    },
+    {
+      "epoch": 0.6888177208433414,
+      "grad_norm": 0.9310915342747638,
+      "learning_rate": 4.663352266427026e-06,
+      "loss": 1.6019,
+      "step": 5162
+    },
+    {
+      "epoch": 0.688951160928743,
+      "grad_norm": 0.9779613924239206,
+      "learning_rate": 4.659697746814832e-06,
+      "loss": 1.5946,
+      "step": 5163
+    },
+    {
+      "epoch": 0.6890846010141447,
+      "grad_norm": 0.944018936409953,
+      "learning_rate": 4.656044224710297e-06,
+      "loss": 1.6024,
+      "step": 5164
+    },
+    {
+      "epoch": 0.6892180410995463,
+      "grad_norm": 0.9489255197624183,
+      "learning_rate": 4.652391700795854e-06,
+      "loss": 1.6148,
+      "step": 5165
+    },
+    {
+      "epoch": 0.6893514811849479,
+      "grad_norm": 0.9554477607577454,
+      "learning_rate": 4.64874017575376e-06,
+      "loss": 1.6305,
+      "step": 5166
+    },
+    {
+      "epoch": 0.6894849212703497,
+      "grad_norm": 0.9578515697117463,
+      "learning_rate": 4.645089650266074e-06,
+      "loss": 1.5761,
+      "step": 5167
+    },
+    {
+      "epoch": 0.6896183613557513,
+      "grad_norm": 1.089494909741061,
+      "learning_rate": 4.641440125014672e-06,
+      "loss": 1.552,
+      "step": 5168
+    },
+    {
+      "epoch": 0.6897518014411529,
+      "grad_norm": 0.9764757676054326,
+      "learning_rate": 4.637791600681247e-06,
+      "loss": 1.6047,
+      "step": 5169
+    },
+    {
+      "epoch": 0.6898852415265546,
+      "grad_norm": 0.9704402753837931,
+      "learning_rate": 4.634144077947301e-06,
+      "loss": 1.5901,
+      "step": 5170
+    },
+    {
+      "epoch": 0.6900186816119562,
+      "grad_norm": 0.9450629354753408,
+      "learning_rate": 4.630497557494149e-06,
+      "loss": 1.5672,
+      "step": 5171
+    },
+    {
+      "epoch": 0.6901521216973578,
+      "grad_norm": 0.9265680379145714,
+      "learning_rate": 4.626852040002915e-06,
+      "loss": 1.5683,
+      "step": 5172
+    },
+    {
+      "epoch": 0.6902855617827596,
+      "grad_norm": 1.0282668336579435,
+      "learning_rate": 4.6232075261545476e-06,
+      "loss": 1.5511,
+      "step": 5173
+    },
+    {
+      "epoch": 0.6904190018681612,
+      "grad_norm": 0.9406742239628743,
+      "learning_rate": 4.6195640166297975e-06,
+      "loss": 1.5473,
+      "step": 5174
+    },
+    {
+      "epoch": 0.6905524419535628,
+      "grad_norm": 0.9331259548380249,
+      "learning_rate": 4.61592151210923e-06,
+      "loss": 1.5373,
+      "step": 5175
+    },
+    {
+      "epoch": 0.6906858820389645,
+      "grad_norm": 0.9424269744921971,
+      "learning_rate": 4.612280013273221e-06,
+      "loss": 1.5352,
+      "step": 5176
+    },
+    {
+      "epoch": 0.6908193221243661,
+      "grad_norm": 0.9992155998336348,
+      "learning_rate": 4.608639520801967e-06,
+      "loss": 1.5999,
+      "step": 5177
+    },
+    {
+      "epoch": 0.6909527622097679,
+      "grad_norm": 1.0114712743867609,
+      "learning_rate": 4.605000035375469e-06,
+      "loss": 1.5355,
+      "step": 5178
+    },
+    {
+      "epoch": 0.6910862022951695,
+      "grad_norm": 0.9838800177008744,
+      "learning_rate": 4.6013615576735385e-06,
+      "loss": 1.6047,
+      "step": 5179
+    },
+    {
+      "epoch": 0.6912196423805711,
+      "grad_norm": 0.9546619375864034,
+      "learning_rate": 4.597724088375803e-06,
+      "loss": 1.5293,
+      "step": 5180
+    },
+    {
+      "epoch": 0.6913530824659728,
+      "grad_norm": 1.3199845007698923,
+      "learning_rate": 4.594087628161702e-06,
+      "loss": 1.6278,
+      "step": 5181
+    },
+    {
+      "epoch": 0.6914865225513744,
+      "grad_norm": 0.971042571067602,
+      "learning_rate": 4.590452177710479e-06,
+      "loss": 1.5705,
+      "step": 5182
+    },
+    {
+      "epoch": 0.691619962636776,
+      "grad_norm": 0.9499493704833659,
+      "learning_rate": 4.586817737701204e-06,
+      "loss": 1.5513,
+      "step": 5183
+    },
+    {
+      "epoch": 0.6917534027221778,
+      "grad_norm": 1.0111784613671377,
+      "learning_rate": 4.583184308812742e-06,
+      "loss": 1.5944,
+      "step": 5184
+    },
+    {
+      "epoch": 0.6918868428075794,
+      "grad_norm": 0.9428606553786436,
+      "learning_rate": 4.579551891723781e-06,
+      "loss": 1.5875,
+      "step": 5185
+    },
+    {
+      "epoch": 0.692020282892981,
+      "grad_norm": 1.017385930406295,
+      "learning_rate": 4.575920487112806e-06,
+      "loss": 1.4778,
+      "step": 5186
+    },
+    {
+      "epoch": 0.6921537229783827,
+      "grad_norm": 0.9481626619930875,
+      "learning_rate": 4.572290095658134e-06,
+      "loss": 1.5244,
+      "step": 5187
+    },
+    {
+      "epoch": 0.6922871630637843,
+      "grad_norm": 1.0138210258708102,
+      "learning_rate": 4.568660718037875e-06,
+      "loss": 1.5753,
+      "step": 5188
+    },
+    {
+      "epoch": 0.692420603149186,
+      "grad_norm": 1.0491600377755286,
+      "learning_rate": 4.565032354929951e-06,
+      "loss": 1.5974,
+      "step": 5189
+    },
+    {
+      "epoch": 0.6925540432345877,
+      "grad_norm": 1.2073763312887016,
+      "learning_rate": 4.561405007012109e-06,
+      "loss": 1.5599,
+      "step": 5190
+    },
+    {
+      "epoch": 0.6926874833199893,
+      "grad_norm": 1.0293910333928922,
+      "learning_rate": 4.55777867496189e-06,
+      "loss": 1.5523,
+      "step": 5191
+    },
+    {
+      "epoch": 0.692820923405391,
+      "grad_norm": 0.9189903325399789,
+      "learning_rate": 4.5541533594566535e-06,
+      "loss": 1.5528,
+      "step": 5192
+    },
+    {
+      "epoch": 0.6929543634907926,
+      "grad_norm": 0.9080583203789369,
+      "learning_rate": 4.550529061173562e-06,
+      "loss": 1.5131,
+      "step": 5193
+    },
+    {
+      "epoch": 0.6930878035761943,
+      "grad_norm": 0.952626918394523,
+      "learning_rate": 4.5469057807896e-06,
+      "loss": 1.5493,
+      "step": 5194
+    },
+    {
+      "epoch": 0.693221243661596,
+      "grad_norm": 0.9394419115149728,
+      "learning_rate": 4.5432835189815535e-06,
+      "loss": 1.5554,
+      "step": 5195
+    },
+    {
+      "epoch": 0.6933546837469976,
+      "grad_norm": 0.9546357659052805,
+      "learning_rate": 4.539662276426018e-06,
+      "loss": 1.5209,
+      "step": 5196
+    },
+    {
+      "epoch": 0.6934881238323992,
+      "grad_norm": 1.003350765525599,
+      "learning_rate": 4.536042053799398e-06,
+      "loss": 1.5391,
+      "step": 5197
+    },
+    {
+      "epoch": 0.6936215639178009,
+      "grad_norm": 0.9836741751211436,
+      "learning_rate": 4.532422851777916e-06,
+      "loss": 1.5573,
+      "step": 5198
+    },
+    {
+      "epoch": 0.6937550040032026,
+      "grad_norm": 0.9574473385195087,
+      "learning_rate": 4.528804671037594e-06,
+      "loss": 1.5952,
+      "step": 5199
+    },
+    {
+      "epoch": 0.6938884440886042,
+      "grad_norm": 0.9311362883331841,
+      "learning_rate": 4.525187512254266e-06,
+      "loss": 1.5085,
+      "step": 5200
+    },
+    {
+      "epoch": 0.6940218841740059,
+      "grad_norm": 0.9609583738468779,
+      "learning_rate": 4.521571376103578e-06,
+      "loss": 1.5997,
+      "step": 5201
+    },
+    {
+      "epoch": 0.6941553242594075,
+      "grad_norm": 0.9375491622362763,
+      "learning_rate": 4.517956263260985e-06,
+      "loss": 1.5734,
+      "step": 5202
+    },
+    {
+      "epoch": 0.6942887643448091,
+      "grad_norm": 0.9381515614250917,
+      "learning_rate": 4.514342174401747e-06,
+      "loss": 1.5768,
+      "step": 5203
+    },
+    {
+      "epoch": 0.6944222044302109,
+      "grad_norm": 0.9464641129012018,
+      "learning_rate": 4.510729110200929e-06,
+      "loss": 1.5765,
+      "step": 5204
+    },
+    {
+      "epoch": 0.6945556445156125,
+      "grad_norm": 0.9461797676889874,
+      "learning_rate": 4.50711707133342e-06,
+      "loss": 1.5685,
+      "step": 5205
+    },
+    {
+      "epoch": 0.6946890846010142,
+      "grad_norm": 0.9925712528011031,
+      "learning_rate": 4.503506058473903e-06,
+      "loss": 1.53,
+      "step": 5206
+    },
+    {
+      "epoch": 0.6948225246864158,
+      "grad_norm": 0.932476678094929,
+      "learning_rate": 4.499896072296871e-06,
+      "loss": 1.5794,
+      "step": 5207
+    },
+    {
+      "epoch": 0.6949559647718174,
+      "grad_norm": 1.0005881114895052,
+      "learning_rate": 4.496287113476635e-06,
+      "loss": 1.5897,
+      "step": 5208
+    },
+    {
+      "epoch": 0.6950894048572192,
+      "grad_norm": 0.945425425153403,
+      "learning_rate": 4.492679182687304e-06,
+      "loss": 1.5572,
+      "step": 5209
+    },
+    {
+      "epoch": 0.6952228449426208,
+      "grad_norm": 0.9279450617465351,
+      "learning_rate": 4.489072280602799e-06,
+      "loss": 1.5723,
+      "step": 5210
+    },
+    {
+      "epoch": 0.6953562850280224,
+      "grad_norm": 0.9947713480204747,
+      "learning_rate": 4.485466407896844e-06,
+      "loss": 1.5525,
+      "step": 5211
+    },
+    {
+      "epoch": 0.6954897251134241,
+      "grad_norm": 1.0073089670685307,
+      "learning_rate": 4.481861565242982e-06,
+      "loss": 1.538,
+      "step": 5212
+    },
+    {
+      "epoch": 0.6956231651988257,
+      "grad_norm": 0.9775029711915599,
+      "learning_rate": 4.478257753314554e-06,
+      "loss": 1.478,
+      "step": 5213
+    },
+    {
+      "epoch": 0.6957566052842273,
+      "grad_norm": 1.1051713607759426,
+      "learning_rate": 4.474654972784705e-06,
+      "loss": 1.578,
+      "step": 5214
+    },
+    {
+      "epoch": 0.6958900453696291,
+      "grad_norm": 0.9162253285174159,
+      "learning_rate": 4.471053224326404e-06,
+      "loss": 1.5747,
+      "step": 5215
+    },
+    {
+      "epoch": 0.6960234854550307,
+      "grad_norm": 0.9337462353761629,
+      "learning_rate": 4.46745250861241e-06,
+      "loss": 1.5675,
+      "step": 5216
+    },
+    {
+      "epoch": 0.6961569255404323,
+      "grad_norm": 1.0028329737784027,
+      "learning_rate": 4.463852826315298e-06,
+      "loss": 1.5358,
+      "step": 5217
+    },
+    {
+      "epoch": 0.696290365625834,
+      "grad_norm": 0.9098824085690614,
+      "learning_rate": 4.460254178107446e-06,
+      "loss": 1.5468,
+      "step": 5218
+    },
+    {
+      "epoch": 0.6964238057112356,
+      "grad_norm": 0.954411548951353,
+      "learning_rate": 4.456656564661037e-06,
+      "loss": 1.5395,
+      "step": 5219
+    },
+    {
+      "epoch": 0.6965572457966374,
+      "grad_norm": 0.9723768754629288,
+      "learning_rate": 4.453059986648073e-06,
+      "loss": 1.5383,
+      "step": 5220
+    },
+    {
+      "epoch": 0.696690685882039,
+      "grad_norm": 0.9714260510102523,
+      "learning_rate": 4.449464444740349e-06,
+      "loss": 1.5604,
+      "step": 5221
+    },
+    {
+      "epoch": 0.6968241259674406,
+      "grad_norm": 0.9812563625331769,
+      "learning_rate": 4.445869939609472e-06,
+      "loss": 1.5926,
+      "step": 5222
+    },
+    {
+      "epoch": 0.6969575660528423,
+      "grad_norm": 1.1501049604817195,
+      "learning_rate": 4.442276471926853e-06,
+      "loss": 1.5484,
+      "step": 5223
+    },
+    {
+      "epoch": 0.6970910061382439,
+      "grad_norm": 0.9626957186688674,
+      "learning_rate": 4.4386840423637124e-06,
+      "loss": 1.5799,
+      "step": 5224
+    },
+    {
+      "epoch": 0.6972244462236455,
+      "grad_norm": 1.0216025455702171,
+      "learning_rate": 4.435092651591072e-06,
+      "loss": 1.5409,
+      "step": 5225
+    },
+    {
+      "epoch": 0.6973578863090473,
+      "grad_norm": 0.9485116910427717,
+      "learning_rate": 4.4315023002797675e-06,
+      "loss": 1.5487,
+      "step": 5226
+    },
+    {
+      "epoch": 0.6974913263944489,
+      "grad_norm": 0.9715863993865531,
+      "learning_rate": 4.427912989100434e-06,
+      "loss": 1.596,
+      "step": 5227
+    },
+    {
+      "epoch": 0.6976247664798505,
+      "grad_norm": 0.968567527529441,
+      "learning_rate": 4.42432471872351e-06,
+      "loss": 1.5294,
+      "step": 5228
+    },
+    {
+      "epoch": 0.6977582065652522,
+      "grad_norm": 0.9687730959837667,
+      "learning_rate": 4.420737489819253e-06,
+      "loss": 1.5265,
+      "step": 5229
+    },
+    {
+      "epoch": 0.6978916466506538,
+      "grad_norm": 1.0885904607859849,
+      "learning_rate": 4.417151303057707e-06,
+      "loss": 1.5883,
+      "step": 5230
+    },
+    {
+      "epoch": 0.6980250867360555,
+      "grad_norm": 1.025218940735289,
+      "learning_rate": 4.413566159108736e-06,
+      "loss": 1.514,
+      "step": 5231
+    },
+    {
+      "epoch": 0.6981585268214572,
+      "grad_norm": 2.751796441629812,
+      "learning_rate": 4.409982058641999e-06,
+      "loss": 1.5411,
+      "step": 5232
+    },
+    {
+      "epoch": 0.6982919669068588,
+      "grad_norm": 0.9234092508289905,
+      "learning_rate": 4.406399002326973e-06,
+      "loss": 1.5368,
+      "step": 5233
+    },
+    {
+      "epoch": 0.6984254069922605,
+      "grad_norm": 1.0403344284539555,
+      "learning_rate": 4.402816990832928e-06,
+      "loss": 1.5848,
+      "step": 5234
+    },
+    {
+      "epoch": 0.6985588470776621,
+      "grad_norm": 0.9148227947424791,
+      "learning_rate": 4.3992360248289416e-06,
+      "loss": 1.5145,
+      "step": 5235
+    },
+    {
+      "epoch": 0.6986922871630638,
+      "grad_norm": 0.9779207787661259,
+      "learning_rate": 4.395656104983895e-06,
+      "loss": 1.5853,
+      "step": 5236
+    },
+    {
+      "epoch": 0.6988257272484655,
+      "grad_norm": 1.111013896532024,
+      "learning_rate": 4.392077231966484e-06,
+      "loss": 1.5895,
+      "step": 5237
+    },
+    {
+      "epoch": 0.6989591673338671,
+      "grad_norm": 1.0099682804012988,
+      "learning_rate": 4.388499406445198e-06,
+      "loss": 1.5454,
+      "step": 5238
+    },
+    {
+      "epoch": 0.6990926074192687,
+      "grad_norm": 0.9508070443731435,
+      "learning_rate": 4.384922629088329e-06,
+      "loss": 1.5156,
+      "step": 5239
+    },
+    {
+      "epoch": 0.6992260475046704,
+      "grad_norm": 0.9772725704881033,
+      "learning_rate": 4.381346900563988e-06,
+      "loss": 1.5688,
+      "step": 5240
+    },
+    {
+      "epoch": 0.699359487590072,
+      "grad_norm": 1.1327443354133715,
+      "learning_rate": 4.377772221540073e-06,
+      "loss": 1.5734,
+      "step": 5241
+    },
+    {
+      "epoch": 0.6994929276754737,
+      "grad_norm": 1.080615395417145,
+      "learning_rate": 4.374198592684296e-06,
+      "loss": 1.5525,
+      "step": 5242
+    },
+    {
+      "epoch": 0.6996263677608754,
+      "grad_norm": 0.9705398879426412,
+      "learning_rate": 4.370626014664168e-06,
+      "loss": 1.5905,
+      "step": 5243
+    },
+    {
+      "epoch": 0.699759807846277,
+      "grad_norm": 0.9423985690259806,
+      "learning_rate": 4.36705448814701e-06,
+      "loss": 1.5543,
+      "step": 5244
+    },
+    {
+      "epoch": 0.6998932479316787,
+      "grad_norm": 1.0013253966326623,
+      "learning_rate": 4.36348401379994e-06,
+      "loss": 1.6252,
+      "step": 5245
+    },
+    {
+      "epoch": 0.7000266880170803,
+      "grad_norm": 0.9849762775796472,
+      "learning_rate": 4.359914592289879e-06,
+      "loss": 1.5735,
+      "step": 5246
+    },
+    {
+      "epoch": 0.700160128102482,
+      "grad_norm": 0.9466224944504175,
+      "learning_rate": 4.3563462242835605e-06,
+      "loss": 1.5474,
+      "step": 5247
+    },
+    {
+      "epoch": 0.7002935681878837,
+      "grad_norm": 1.0452884135924572,
+      "learning_rate": 4.352778910447513e-06,
+      "loss": 1.55,
+      "step": 5248
+    },
+    {
+      "epoch": 0.7004270082732853,
+      "grad_norm": 0.9636258481231191,
+      "learning_rate": 4.349212651448067e-06,
+      "loss": 1.5592,
+      "step": 5249
+    },
+    {
+      "epoch": 0.7005604483586869,
+      "grad_norm": 0.9266243703586854,
+      "learning_rate": 4.345647447951359e-06,
+      "loss": 1.5426,
+      "step": 5250
+    },
+    {
+      "epoch": 0.7006938884440886,
+      "grad_norm": 0.9465277944454074,
+      "learning_rate": 4.3420833006233335e-06,
+      "loss": 1.5223,
+      "step": 5251
+    },
+    {
+      "epoch": 0.7008273285294903,
+      "grad_norm": 0.9476657743051038,
+      "learning_rate": 4.338520210129729e-06,
+      "loss": 1.5138,
+      "step": 5252
+    },
+    {
+      "epoch": 0.7009607686148919,
+      "grad_norm": 0.92673177526777,
+      "learning_rate": 4.33495817713609e-06,
+      "loss": 1.5719,
+      "step": 5253
+    },
+    {
+      "epoch": 0.7010942087002936,
+      "grad_norm": 1.1410712640987881,
+      "learning_rate": 4.3313972023077656e-06,
+      "loss": 1.5309,
+      "step": 5254
+    },
+    {
+      "epoch": 0.7012276487856952,
+      "grad_norm": 0.8997294098548428,
+      "learning_rate": 4.327837286309905e-06,
+      "loss": 1.5379,
+      "step": 5255
+    },
+    {
+      "epoch": 0.7013610888710968,
+      "grad_norm": 1.117573499608535,
+      "learning_rate": 4.324278429807459e-06,
+      "loss": 1.5973,
+      "step": 5256
+    },
+    {
+      "epoch": 0.7014945289564986,
+      "grad_norm": 0.9591244120794552,
+      "learning_rate": 4.320720633465178e-06,
+      "loss": 1.576,
+      "step": 5257
+    },
+    {
+      "epoch": 0.7016279690419002,
+      "grad_norm": 1.246892551311886,
+      "learning_rate": 4.317163897947626e-06,
+      "loss": 1.5957,
+      "step": 5258
+    },
+    {
+      "epoch": 0.7017614091273019,
+      "grad_norm": 0.9650925501800929,
+      "learning_rate": 4.3136082239191565e-06,
+      "loss": 1.5404,
+      "step": 5259
+    },
+    {
+      "epoch": 0.7018948492127035,
+      "grad_norm": 0.937754472050194,
+      "learning_rate": 4.310053612043928e-06,
+      "loss": 1.5491,
+      "step": 5260
+    },
+    {
+      "epoch": 0.7020282892981051,
+      "grad_norm": 0.918105960005196,
+      "learning_rate": 4.306500062985903e-06,
+      "loss": 1.5244,
+      "step": 5261
+    },
+    {
+      "epoch": 0.7021617293835068,
+      "grad_norm": 0.9586676779477473,
+      "learning_rate": 4.302947577408839e-06,
+      "loss": 1.6228,
+      "step": 5262
+    },
+    {
+      "epoch": 0.7022951694689085,
+      "grad_norm": 0.9517519387872295,
+      "learning_rate": 4.299396155976308e-06,
+      "loss": 1.6056,
+      "step": 5263
+    },
+    {
+      "epoch": 0.7024286095543101,
+      "grad_norm": 1.495533408695814,
+      "learning_rate": 4.295845799351672e-06,
+      "loss": 1.5421,
+      "step": 5264
+    },
+    {
+      "epoch": 0.7025620496397118,
+      "grad_norm": 0.9864875327078859,
+      "learning_rate": 4.2922965081980964e-06,
+      "loss": 1.5468,
+      "step": 5265
+    },
+    {
+      "epoch": 0.7026954897251134,
+      "grad_norm": 0.9301887930563217,
+      "learning_rate": 4.288748283178546e-06,
+      "loss": 1.5442,
+      "step": 5266
+    },
+    {
+      "epoch": 0.702828929810515,
+      "grad_norm": 0.9633255736311295,
+      "learning_rate": 4.285201124955795e-06,
+      "loss": 1.6113,
+      "step": 5267
+    },
+    {
+      "epoch": 0.7029623698959168,
+      "grad_norm": 0.9458457528853886,
+      "learning_rate": 4.281655034192402e-06,
+      "loss": 1.5682,
+      "step": 5268
+    },
+    {
+      "epoch": 0.7030958099813184,
+      "grad_norm": 0.9396435856415133,
+      "learning_rate": 4.278110011550748e-06,
+      "loss": 1.5206,
+      "step": 5269
+    },
+    {
+      "epoch": 0.70322925006672,
+      "grad_norm": 0.932365193699769,
+      "learning_rate": 4.274566057692996e-06,
+      "loss": 1.5637,
+      "step": 5270
+    },
+    {
+      "epoch": 0.7033626901521217,
+      "grad_norm": 0.9356770658793856,
+      "learning_rate": 4.271023173281116e-06,
+      "loss": 1.5513,
+      "step": 5271
+    },
+    {
+      "epoch": 0.7034961302375233,
+      "grad_norm": 0.9470539843139907,
+      "learning_rate": 4.267481358976883e-06,
+      "loss": 1.5313,
+      "step": 5272
+    },
+    {
+      "epoch": 0.7036295703229251,
+      "grad_norm": 0.936074803599193,
+      "learning_rate": 4.263940615441865e-06,
+      "loss": 1.5915,
+      "step": 5273
+    },
+    {
+      "epoch": 0.7037630104083267,
+      "grad_norm": 0.9601362770329818,
+      "learning_rate": 4.260400943337433e-06,
+      "loss": 1.5683,
+      "step": 5274
+    },
+    {
+      "epoch": 0.7038964504937283,
+      "grad_norm": 0.9200436045608203,
+      "learning_rate": 4.2568623433247534e-06,
+      "loss": 1.5353,
+      "step": 5275
+    },
+    {
+      "epoch": 0.70402989057913,
+      "grad_norm": 1.2183542283592865,
+      "learning_rate": 4.253324816064803e-06,
+      "loss": 1.5893,
+      "step": 5276
+    },
+    {
+      "epoch": 0.7041633306645316,
+      "grad_norm": 0.9608670542983757,
+      "learning_rate": 4.24978836221835e-06,
+      "loss": 1.5719,
+      "step": 5277
+    },
+    {
+      "epoch": 0.7042967707499332,
+      "grad_norm": 1.0831628265033406,
+      "learning_rate": 4.246252982445957e-06,
+      "loss": 1.5496,
+      "step": 5278
+    },
+    {
+      "epoch": 0.704430210835335,
+      "grad_norm": 0.9288733637532146,
+      "learning_rate": 4.242718677408002e-06,
+      "loss": 1.5955,
+      "step": 5279
+    },
+    {
+      "epoch": 0.7045636509207366,
+      "grad_norm": 0.9220556310697627,
+      "learning_rate": 4.23918544776465e-06,
+      "loss": 1.5466,
+      "step": 5280
+    },
+    {
+      "epoch": 0.7046970910061382,
+      "grad_norm": 1.0053641177822847,
+      "learning_rate": 4.235653294175866e-06,
+      "loss": 1.5668,
+      "step": 5281
+    },
+    {
+      "epoch": 0.7048305310915399,
+      "grad_norm": 0.9592861362485582,
+      "learning_rate": 4.232122217301414e-06,
+      "loss": 1.5207,
+      "step": 5282
+    },
+    {
+      "epoch": 0.7049639711769415,
+      "grad_norm": 1.003209125290528,
+      "learning_rate": 4.228592217800865e-06,
+      "loss": 1.5636,
+      "step": 5283
+    },
+    {
+      "epoch": 0.7050974112623432,
+      "grad_norm": 0.9377732106678393,
+      "learning_rate": 4.2250632963335805e-06,
+      "loss": 1.5727,
+      "step": 5284
+    },
+    {
+      "epoch": 0.7052308513477449,
+      "grad_norm": 10.633257878300137,
+      "learning_rate": 4.221535453558718e-06,
+      "loss": 1.6243,
+      "step": 5285
+    },
+    {
+      "epoch": 0.7053642914331465,
+      "grad_norm": 0.9780546181932777,
+      "learning_rate": 4.218008690135247e-06,
+      "loss": 1.5263,
+      "step": 5286
+    },
+    {
+      "epoch": 0.7054977315185482,
+      "grad_norm": 1.016085908744265,
+      "learning_rate": 4.214483006721921e-06,
+      "loss": 1.5835,
+      "step": 5287
+    },
+    {
+      "epoch": 0.7056311716039498,
+      "grad_norm": 0.946908828245464,
+      "learning_rate": 4.2109584039773e-06,
+      "loss": 1.6154,
+      "step": 5288
+    },
+    {
+      "epoch": 0.7057646116893515,
+      "grad_norm": 0.9794302729269099,
+      "learning_rate": 4.207434882559734e-06,
+      "loss": 1.5523,
+      "step": 5289
+    },
+    {
+      "epoch": 0.7058980517747532,
+      "grad_norm": 1.1316618652316528,
+      "learning_rate": 4.2039124431273845e-06,
+      "loss": 1.5279,
+      "step": 5290
+    },
+    {
+      "epoch": 0.7060314918601548,
+      "grad_norm": 0.9378659449314672,
+      "learning_rate": 4.2003910863382005e-06,
+      "loss": 1.565,
+      "step": 5291
+    },
+    {
+      "epoch": 0.7061649319455564,
+      "grad_norm": 0.9503258372477722,
+      "learning_rate": 4.1968708128499256e-06,
+      "loss": 1.5009,
+      "step": 5292
+    },
+    {
+      "epoch": 0.7062983720309581,
+      "grad_norm": 0.9519765695098307,
+      "learning_rate": 4.1933516233201165e-06,
+      "loss": 1.6068,
+      "step": 5293
+    },
+    {
+      "epoch": 0.7064318121163597,
+      "grad_norm": 1.1619371544024075,
+      "learning_rate": 4.189833518406113e-06,
+      "loss": 1.5333,
+      "step": 5294
+    },
+    {
+      "epoch": 0.7065652522017614,
+      "grad_norm": 0.9543539047132761,
+      "learning_rate": 4.1863164987650575e-06,
+      "loss": 1.5793,
+      "step": 5295
+    },
+    {
+      "epoch": 0.7066986922871631,
+      "grad_norm": 0.9515900226999982,
+      "learning_rate": 4.182800565053884e-06,
+      "loss": 1.5893,
+      "step": 5296
+    },
+    {
+      "epoch": 0.7068321323725647,
+      "grad_norm": 0.9355762162950707,
+      "learning_rate": 4.179285717929338e-06,
+      "loss": 1.5553,
+      "step": 5297
+    },
+    {
+      "epoch": 0.7069655724579663,
+      "grad_norm": 0.9421727010797218,
+      "learning_rate": 4.175771958047947e-06,
+      "loss": 1.5825,
+      "step": 5298
+    },
+    {
+      "epoch": 0.707099012543368,
+      "grad_norm": 1.0799208667442957,
+      "learning_rate": 4.172259286066045e-06,
+      "loss": 1.5503,
+      "step": 5299
+    },
+    {
+      "epoch": 0.7072324526287697,
+      "grad_norm": 0.9679399491700288,
+      "learning_rate": 4.168747702639753e-06,
+      "loss": 1.5524,
+      "step": 5300
+    },
+    {
+      "epoch": 0.7073658927141714,
+      "grad_norm": 0.9691270167436293,
+      "learning_rate": 4.165237208425001e-06,
+      "loss": 1.5649,
+      "step": 5301
+    },
+    {
+      "epoch": 0.707499332799573,
+      "grad_norm": 0.9311122890360443,
+      "learning_rate": 4.1617278040775086e-06,
+      "loss": 1.514,
+      "step": 5302
+    },
+    {
+      "epoch": 0.7076327728849746,
+      "grad_norm": 0.8967366726125487,
+      "learning_rate": 4.158219490252791e-06,
+      "loss": 1.5358,
+      "step": 5303
+    },
+    {
+      "epoch": 0.7077662129703763,
+      "grad_norm": 0.9509487136930983,
+      "learning_rate": 4.154712267606161e-06,
+      "loss": 1.5899,
+      "step": 5304
+    },
+    {
+      "epoch": 0.707899653055778,
+      "grad_norm": 0.9712447192685107,
+      "learning_rate": 4.151206136792725e-06,
+      "loss": 1.5993,
+      "step": 5305
+    },
+    {
+      "epoch": 0.7080330931411796,
+      "grad_norm": 0.9453208811579048,
+      "learning_rate": 4.147701098467395e-06,
+      "loss": 1.5903,
+      "step": 5306
+    },
+    {
+      "epoch": 0.7081665332265813,
+      "grad_norm": 0.955132045264726,
+      "learning_rate": 4.144197153284869e-06,
+      "loss": 1.513,
+      "step": 5307
+    },
+    {
+      "epoch": 0.7082999733119829,
+      "grad_norm": 0.9400188884444305,
+      "learning_rate": 4.140694301899645e-06,
+      "loss": 1.551,
+      "step": 5308
+    },
+    {
+      "epoch": 0.7084334133973845,
+      "grad_norm": 0.9193593248190934,
+      "learning_rate": 4.137192544966013e-06,
+      "loss": 1.585,
+      "step": 5309
+    },
+    {
+      "epoch": 0.7085668534827863,
+      "grad_norm": 0.9312434980756322,
+      "learning_rate": 4.13369188313806e-06,
+      "loss": 1.5386,
+      "step": 5310
+    },
+    {
+      "epoch": 0.7087002935681879,
+      "grad_norm": 0.9271931561265911,
+      "learning_rate": 4.130192317069677e-06,
+      "loss": 1.5208,
+      "step": 5311
+    },
+    {
+      "epoch": 0.7088337336535896,
+      "grad_norm": 0.9970719656664363,
+      "learning_rate": 4.126693847414538e-06,
+      "loss": 1.5826,
+      "step": 5312
+    },
+    {
+      "epoch": 0.7089671737389912,
+      "grad_norm": 0.9479786344427346,
+      "learning_rate": 4.123196474826119e-06,
+      "loss": 1.5403,
+      "step": 5313
+    },
+    {
+      "epoch": 0.7091006138243928,
+      "grad_norm": 0.8979702441638332,
+      "learning_rate": 4.119700199957684e-06,
+      "loss": 1.5509,
+      "step": 5314
+    },
+    {
+      "epoch": 0.7092340539097945,
+      "grad_norm": 0.9895920970366872,
+      "learning_rate": 4.116205023462306e-06,
+      "loss": 1.5785,
+      "step": 5315
+    },
+    {
+      "epoch": 0.7093674939951962,
+      "grad_norm": 0.9499405479425314,
+      "learning_rate": 4.112710945992842e-06,
+      "loss": 1.5824,
+      "step": 5316
+    },
+    {
+      "epoch": 0.7095009340805978,
+      "grad_norm": 0.9355717622213187,
+      "learning_rate": 4.109217968201937e-06,
+      "loss": 1.5706,
+      "step": 5317
+    },
+    {
+      "epoch": 0.7096343741659995,
+      "grad_norm": 0.9477283832957076,
+      "learning_rate": 4.1057260907420524e-06,
+      "loss": 1.497,
+      "step": 5318
+    },
+    {
+      "epoch": 0.7097678142514011,
+      "grad_norm": 0.9551646576385708,
+      "learning_rate": 4.102235314265425e-06,
+      "loss": 1.5367,
+      "step": 5319
+    },
+    {
+      "epoch": 0.7099012543368027,
+      "grad_norm": 0.9248106472529288,
+      "learning_rate": 4.098745639424091e-06,
+      "loss": 1.5693,
+      "step": 5320
+    },
+    {
+      "epoch": 0.7100346944222045,
+      "grad_norm": 0.963099005180929,
+      "learning_rate": 4.095257066869881e-06,
+      "loss": 1.573,
+      "step": 5321
+    },
+    {
+      "epoch": 0.7101681345076061,
+      "grad_norm": 0.9687592868939011,
+      "learning_rate": 4.091769597254426e-06,
+      "loss": 1.5924,
+      "step": 5322
+    },
+    {
+      "epoch": 0.7103015745930077,
+      "grad_norm": 0.9896034088236875,
+      "learning_rate": 4.088283231229142e-06,
+      "loss": 1.5621,
+      "step": 5323
+    },
+    {
+      "epoch": 0.7104350146784094,
+      "grad_norm": 0.9341590041901182,
+      "learning_rate": 4.08479796944524e-06,
+      "loss": 1.552,
+      "step": 5324
+    },
+    {
+      "epoch": 0.710568454763811,
+      "grad_norm": 0.9304265266280125,
+      "learning_rate": 4.081313812553734e-06,
+      "loss": 1.5652,
+      "step": 5325
+    },
+    {
+      "epoch": 0.7107018948492128,
+      "grad_norm": 0.9605863893931225,
+      "learning_rate": 4.07783076120542e-06,
+      "loss": 1.5676,
+      "step": 5326
+    },
+    {
+      "epoch": 0.7108353349346144,
+      "grad_norm": 1.0997008119225524,
+      "learning_rate": 4.074348816050895e-06,
+      "loss": 1.6232,
+      "step": 5327
+    },
+    {
+      "epoch": 0.710968775020016,
+      "grad_norm": 0.9045291420444287,
+      "learning_rate": 4.07086797774054e-06,
+      "loss": 1.5255,
+      "step": 5328
+    },
+    {
+      "epoch": 0.7111022151054177,
+      "grad_norm": 0.9985345986185172,
+      "learning_rate": 4.067388246924545e-06,
+      "loss": 1.6026,
+      "step": 5329
+    },
+    {
+      "epoch": 0.7112356551908193,
+      "grad_norm": 0.9841974283402879,
+      "learning_rate": 4.063909624252881e-06,
+      "loss": 1.5454,
+      "step": 5330
+    },
+    {
+      "epoch": 0.7113690952762209,
+      "grad_norm": 0.9526315404945778,
+      "learning_rate": 4.060432110375314e-06,
+      "loss": 1.5019,
+      "step": 5331
+    },
+    {
+      "epoch": 0.7115025353616227,
+      "grad_norm": 0.9491732500227824,
+      "learning_rate": 4.0569557059414e-06,
+      "loss": 1.5423,
+      "step": 5332
+    },
+    {
+      "epoch": 0.7116359754470243,
+      "grad_norm": 0.9491096007519367,
+      "learning_rate": 4.0534804116005e-06,
+      "loss": 1.5221,
+      "step": 5333
+    },
+    {
+      "epoch": 0.7117694155324259,
+      "grad_norm": 0.9284527190069253,
+      "learning_rate": 4.0500062280017545e-06,
+      "loss": 1.5625,
+      "step": 5334
+    },
+    {
+      "epoch": 0.7119028556178276,
+      "grad_norm": 0.9759261683989565,
+      "learning_rate": 4.0465331557941e-06,
+      "loss": 1.5516,
+      "step": 5335
+    },
+    {
+      "epoch": 0.7120362957032292,
+      "grad_norm": 1.0927488467118913,
+      "learning_rate": 4.0430611956262735e-06,
+      "loss": 1.5364,
+      "step": 5336
+    },
+    {
+      "epoch": 0.7121697357886309,
+      "grad_norm": 2.0805229151915445,
+      "learning_rate": 4.039590348146791e-06,
+      "loss": 1.5377,
+      "step": 5337
+    },
+    {
+      "epoch": 0.7123031758740326,
+      "grad_norm": 0.9353765748626967,
+      "learning_rate": 4.036120614003972e-06,
+      "loss": 1.571,
+      "step": 5338
+    },
+    {
+      "epoch": 0.7124366159594342,
+      "grad_norm": 0.9903701860089882,
+      "learning_rate": 4.032651993845917e-06,
+      "loss": 1.5247,
+      "step": 5339
+    },
+    {
+      "epoch": 0.7125700560448359,
+      "grad_norm": 0.9235475362427148,
+      "learning_rate": 4.029184488320531e-06,
+      "loss": 1.515,
+      "step": 5340
+    },
+    {
+      "epoch": 0.7127034961302375,
+      "grad_norm": 0.9026258019298207,
+      "learning_rate": 4.0257180980755025e-06,
+      "loss": 1.5408,
+      "step": 5341
+    },
+    {
+      "epoch": 0.7128369362156391,
+      "grad_norm": 1.2538215164072486,
+      "learning_rate": 4.022252823758311e-06,
+      "loss": 1.6143,
+      "step": 5342
+    },
+    {
+      "epoch": 0.7129703763010409,
+      "grad_norm": 0.9545237012787869,
+      "learning_rate": 4.018788666016236e-06,
+      "loss": 1.5585,
+      "step": 5343
+    },
+    {
+      "epoch": 0.7131038163864425,
+      "grad_norm": 0.9280503299040727,
+      "learning_rate": 4.015325625496339e-06,
+      "loss": 1.573,
+      "step": 5344
+    },
+    {
+      "epoch": 0.7132372564718441,
+      "grad_norm": 0.9990855517950481,
+      "learning_rate": 4.011863702845477e-06,
+      "loss": 1.5783,
+      "step": 5345
+    },
+    {
+      "epoch": 0.7133706965572458,
+      "grad_norm": 0.9614841419311843,
+      "learning_rate": 4.008402898710299e-06,
+      "loss": 1.5716,
+      "step": 5346
+    },
+    {
+      "epoch": 0.7135041366426474,
+      "grad_norm": 0.9450318885739145,
+      "learning_rate": 4.004943213737238e-06,
+      "loss": 1.5561,
+      "step": 5347
+    },
+    {
+      "epoch": 0.7136375767280491,
+      "grad_norm": 0.9957058765830085,
+      "learning_rate": 4.001484648572532e-06,
+      "loss": 1.5563,
+      "step": 5348
+    },
+    {
+      "epoch": 0.7137710168134508,
+      "grad_norm": 1.0067457396072523,
+      "learning_rate": 3.998027203862199e-06,
+      "loss": 1.5962,
+      "step": 5349
+    },
+    {
+      "epoch": 0.7139044568988524,
+      "grad_norm": 0.9853384038154768,
+      "learning_rate": 3.994570880252049e-06,
+      "loss": 1.5547,
+      "step": 5350
+    },
+    {
+      "epoch": 0.714037896984254,
+      "grad_norm": 0.9554713328733841,
+      "learning_rate": 3.991115678387684e-06,
+      "loss": 1.5627,
+      "step": 5351
+    },
+    {
+      "epoch": 0.7141713370696557,
+      "grad_norm": 1.0861226656873586,
+      "learning_rate": 3.987661598914496e-06,
+      "loss": 1.6133,
+      "step": 5352
+    },
+    {
+      "epoch": 0.7143047771550574,
+      "grad_norm": 0.9363068359607128,
+      "learning_rate": 3.984208642477665e-06,
+      "loss": 1.5493,
+      "step": 5353
+    },
+    {
+      "epoch": 0.7144382172404591,
+      "grad_norm": 0.9496779214889088,
+      "learning_rate": 3.9807568097221705e-06,
+      "loss": 1.5819,
+      "step": 5354
+    },
+    {
+      "epoch": 0.7145716573258607,
+      "grad_norm": 0.9293168489795315,
+      "learning_rate": 3.977306101292773e-06,
+      "loss": 1.5272,
+      "step": 5355
+    },
+    {
+      "epoch": 0.7147050974112623,
+      "grad_norm": 0.962632697448326,
+      "learning_rate": 3.973856517834021e-06,
+      "loss": 1.5798,
+      "step": 5356
+    },
+    {
+      "epoch": 0.714838537496664,
+      "grad_norm": 0.9584702412606565,
+      "learning_rate": 3.9704080599902635e-06,
+      "loss": 1.5512,
+      "step": 5357
+    },
+    {
+      "epoch": 0.7149719775820657,
+      "grad_norm": 0.9815244757813827,
+      "learning_rate": 3.966960728405633e-06,
+      "loss": 1.6353,
+      "step": 5358
+    },
+    {
+      "epoch": 0.7151054176674673,
+      "grad_norm": 0.9928728802532544,
+      "learning_rate": 3.963514523724049e-06,
+      "loss": 1.5299,
+      "step": 5359
+    },
+    {
+      "epoch": 0.715238857752869,
+      "grad_norm": 0.9623909333997922,
+      "learning_rate": 3.96006944658922e-06,
+      "loss": 1.5684,
+      "step": 5360
+    },
+    {
+      "epoch": 0.7153722978382706,
+      "grad_norm": 0.9112193396740259,
+      "learning_rate": 3.956625497644655e-06,
+      "loss": 1.5723,
+      "step": 5361
+    },
+    {
+      "epoch": 0.7155057379236722,
+      "grad_norm": 0.9459655781359935,
+      "learning_rate": 3.95318267753364e-06,
+      "loss": 1.5027,
+      "step": 5362
+    },
+    {
+      "epoch": 0.715639178009074,
+      "grad_norm": 1.0216963512845265,
+      "learning_rate": 3.949740986899257e-06,
+      "loss": 1.5619,
+      "step": 5363
+    },
+    {
+      "epoch": 0.7157726180944756,
+      "grad_norm": 14.9027623425261,
+      "learning_rate": 3.946300426384368e-06,
+      "loss": 1.5481,
+      "step": 5364
+    },
+    {
+      "epoch": 0.7159060581798772,
+      "grad_norm": 1.0993613787167713,
+      "learning_rate": 3.942860996631639e-06,
+      "loss": 1.587,
+      "step": 5365
+    },
+    {
+      "epoch": 0.7160394982652789,
+      "grad_norm": 1.0313006265813407,
+      "learning_rate": 3.939422698283513e-06,
+      "loss": 1.5491,
+      "step": 5366
+    },
+    {
+      "epoch": 0.7161729383506805,
+      "grad_norm": 0.9390749627476853,
+      "learning_rate": 3.9359855319822205e-06,
+      "loss": 1.5661,
+      "step": 5367
+    },
+    {
+      "epoch": 0.7163063784360822,
+      "grad_norm": 0.9760457226921616,
+      "learning_rate": 3.932549498369793e-06,
+      "loss": 1.6101,
+      "step": 5368
+    },
+    {
+      "epoch": 0.7164398185214839,
+      "grad_norm": 0.958734457032826,
+      "learning_rate": 3.92911459808804e-06,
+      "loss": 1.5229,
+      "step": 5369
+    },
+    {
+      "epoch": 0.7165732586068855,
+      "grad_norm": 0.9526647222869271,
+      "learning_rate": 3.925680831778559e-06,
+      "loss": 1.5667,
+      "step": 5370
+    },
+    {
+      "epoch": 0.7167066986922872,
+      "grad_norm": 1.071079462379526,
+      "learning_rate": 3.922248200082736e-06,
+      "loss": 1.5439,
+      "step": 5371
+    },
+    {
+      "epoch": 0.7168401387776888,
+      "grad_norm": 1.0754749545538824,
+      "learning_rate": 3.918816703641757e-06,
+      "loss": 1.4997,
+      "step": 5372
+    },
+    {
+      "epoch": 0.7169735788630904,
+      "grad_norm": 0.9839191618810875,
+      "learning_rate": 3.915386343096579e-06,
+      "loss": 1.5861,
+      "step": 5373
+    },
+    {
+      "epoch": 0.7171070189484922,
+      "grad_norm": 0.9696932839295033,
+      "learning_rate": 3.911957119087953e-06,
+      "loss": 1.5767,
+      "step": 5374
+    },
+    {
+      "epoch": 0.7172404590338938,
+      "grad_norm": 1.0878185921141816,
+      "learning_rate": 3.908529032256426e-06,
+      "loss": 1.5695,
+      "step": 5375
+    },
+    {
+      "epoch": 0.7173738991192954,
+      "grad_norm": 1.0356241772862398,
+      "learning_rate": 3.905102083242321e-06,
+      "loss": 1.5772,
+      "step": 5376
+    },
+    {
+      "epoch": 0.7175073392046971,
+      "grad_norm": 0.9534440152222651,
+      "learning_rate": 3.901676272685755e-06,
+      "loss": 1.5578,
+      "step": 5377
+    },
+    {
+      "epoch": 0.7176407792900987,
+      "grad_norm": 1.0211708448584833,
+      "learning_rate": 3.898251601226626e-06,
+      "loss": 1.4683,
+      "step": 5378
+    },
+    {
+      "epoch": 0.7177742193755005,
+      "grad_norm": 1.0468960541282286,
+      "learning_rate": 3.894828069504629e-06,
+      "loss": 1.5454,
+      "step": 5379
+    },
+    {
+      "epoch": 0.7179076594609021,
+      "grad_norm": 0.9536008377149614,
+      "learning_rate": 3.89140567815924e-06,
+      "loss": 1.5293,
+      "step": 5380
+    },
+    {
+      "epoch": 0.7180410995463037,
+      "grad_norm": 1.0458155687344466,
+      "learning_rate": 3.8879844278297164e-06,
+      "loss": 1.6136,
+      "step": 5381
+    },
+    {
+      "epoch": 0.7181745396317054,
+      "grad_norm": 0.9300674323746942,
+      "learning_rate": 3.884564319155119e-06,
+      "loss": 1.5283,
+      "step": 5382
+    },
+    {
+      "epoch": 0.718307979717107,
+      "grad_norm": 1.147948065522506,
+      "learning_rate": 3.881145352774278e-06,
+      "loss": 1.4485,
+      "step": 5383
+    },
+    {
+      "epoch": 0.7184414198025086,
+      "grad_norm": 0.9414728729601566,
+      "learning_rate": 3.877727529325821e-06,
+      "loss": 1.5722,
+      "step": 5384
+    },
+    {
+      "epoch": 0.7185748598879104,
+      "grad_norm": 0.9845671466215753,
+      "learning_rate": 3.874310849448152e-06,
+      "loss": 1.5702,
+      "step": 5385
+    },
+    {
+      "epoch": 0.718708299973312,
+      "grad_norm": 1.2446034917882773,
+      "learning_rate": 3.870895313779477e-06,
+      "loss": 1.5208,
+      "step": 5386
+    },
+    {
+      "epoch": 0.7188417400587136,
+      "grad_norm": 0.9504776390771918,
+      "learning_rate": 3.867480922957775e-06,
+      "loss": 1.5833,
+      "step": 5387
+    },
+    {
+      "epoch": 0.7189751801441153,
+      "grad_norm": 0.9383944211772788,
+      "learning_rate": 3.864067677620815e-06,
+      "loss": 1.5096,
+      "step": 5388
+    },
+    {
+      "epoch": 0.7191086202295169,
+      "grad_norm": 0.9197073694471657,
+      "learning_rate": 3.860655578406153e-06,
+      "loss": 1.4891,
+      "step": 5389
+    },
+    {
+      "epoch": 0.7192420603149186,
+      "grad_norm": 1.293587939218792,
+      "learning_rate": 3.857244625951125e-06,
+      "loss": 1.5577,
+      "step": 5390
+    },
+    {
+      "epoch": 0.7193755004003203,
+      "grad_norm": 0.9425333596458038,
+      "learning_rate": 3.8538348208928675e-06,
+      "loss": 1.547,
+      "step": 5391
+    },
+    {
+      "epoch": 0.7195089404857219,
+      "grad_norm": 0.9187340276232536,
+      "learning_rate": 3.850426163868289e-06,
+      "loss": 1.5602,
+      "step": 5392
+    },
+    {
+      "epoch": 0.7196423805711236,
+      "grad_norm": 1.0066203632714577,
+      "learning_rate": 3.847018655514087e-06,
+      "loss": 1.5686,
+      "step": 5393
+    },
+    {
+      "epoch": 0.7197758206565252,
+      "grad_norm": 0.9101557342470021,
+      "learning_rate": 3.843612296466747e-06,
+      "loss": 1.5074,
+      "step": 5394
+    },
+    {
+      "epoch": 0.7199092607419268,
+      "grad_norm": 1.0091781445483228,
+      "learning_rate": 3.840207087362535e-06,
+      "loss": 1.5506,
+      "step": 5395
+    },
+    {
+      "epoch": 0.7200427008273286,
+      "grad_norm": 0.9766556965941918,
+      "learning_rate": 3.836803028837506e-06,
+      "loss": 1.575,
+      "step": 5396
+    },
+    {
+      "epoch": 0.7201761409127302,
+      "grad_norm": 0.9326384838247909,
+      "learning_rate": 3.833400121527502e-06,
+      "loss": 1.5406,
+      "step": 5397
+    },
+    {
+      "epoch": 0.7203095809981318,
+      "grad_norm": 0.9729805047938272,
+      "learning_rate": 3.829998366068147e-06,
+      "loss": 1.583,
+      "step": 5398
+    },
+    {
+      "epoch": 0.7204430210835335,
+      "grad_norm": 1.010555826446732,
+      "learning_rate": 3.826597763094844e-06,
+      "loss": 1.533,
+      "step": 5399
+    },
+    {
+      "epoch": 0.7205764611689351,
+      "grad_norm": 0.9416257579609061,
+      "learning_rate": 3.823198313242799e-06,
+      "loss": 1.5651,
+      "step": 5400
+    },
+    {
+      "epoch": 0.7207099012543368,
+      "grad_norm": 0.9745957493904607,
+      "learning_rate": 3.81980001714698e-06,
+      "loss": 1.5579,
+      "step": 5401
+    },
+    {
+      "epoch": 0.7208433413397385,
+      "grad_norm": 0.9508022115629777,
+      "learning_rate": 3.816402875442155e-06,
+      "loss": 1.5297,
+      "step": 5402
+    },
+    {
+      "epoch": 0.7209767814251401,
+      "grad_norm": 9.996627671307085,
+      "learning_rate": 3.813006888762867e-06,
+      "loss": 1.5849,
+      "step": 5403
+    },
+    {
+      "epoch": 0.7211102215105417,
+      "grad_norm": 0.9362163294366803,
+      "learning_rate": 3.809612057743454e-06,
+      "loss": 1.5874,
+      "step": 5404
+    },
+    {
+      "epoch": 0.7212436615959434,
+      "grad_norm": 0.9650373486942456,
+      "learning_rate": 3.8062183830180278e-06,
+      "loss": 1.5313,
+      "step": 5405
+    },
+    {
+      "epoch": 0.7213771016813451,
+      "grad_norm": 0.9201142210045637,
+      "learning_rate": 3.802825865220485e-06,
+      "loss": 1.5155,
+      "step": 5406
+    },
+    {
+      "epoch": 0.7215105417667468,
+      "grad_norm": 0.9383591060889741,
+      "learning_rate": 3.799434504984517e-06,
+      "loss": 1.5318,
+      "step": 5407
+    },
+    {
+      "epoch": 0.7216439818521484,
+      "grad_norm": 0.9356509582385691,
+      "learning_rate": 3.7960443029435876e-06,
+      "loss": 1.5374,
+      "step": 5408
+    },
+    {
+      "epoch": 0.72177742193755,
+      "grad_norm": 0.9314768342510902,
+      "learning_rate": 3.7926552597309472e-06,
+      "loss": 1.597,
+      "step": 5409
+    },
+    {
+      "epoch": 0.7219108620229517,
+      "grad_norm": 1.230724666512068,
+      "learning_rate": 3.7892673759796285e-06,
+      "loss": 1.5506,
+      "step": 5410
+    },
+    {
+      "epoch": 0.7220443021083534,
+      "grad_norm": 0.8916809377299305,
+      "learning_rate": 3.7858806523224546e-06,
+      "loss": 1.4881,
+      "step": 5411
+    },
+    {
+      "epoch": 0.722177742193755,
+      "grad_norm": 1.150653767752609,
+      "learning_rate": 3.7824950893920256e-06,
+      "loss": 1.5351,
+      "step": 5412
+    },
+    {
+      "epoch": 0.7223111822791567,
+      "grad_norm": 0.9860122847150068,
+      "learning_rate": 3.7791106878207206e-06,
+      "loss": 1.5517,
+      "step": 5413
+    },
+    {
+      "epoch": 0.7224446223645583,
+      "grad_norm": 0.9187176873452039,
+      "learning_rate": 3.7757274482407146e-06,
+      "loss": 1.5182,
+      "step": 5414
+    },
+    {
+      "epoch": 0.7225780624499599,
+      "grad_norm": 0.9502222725296599,
+      "learning_rate": 3.7723453712839566e-06,
+      "loss": 1.5383,
+      "step": 5415
+    },
+    {
+      "epoch": 0.7227115025353616,
+      "grad_norm": 0.9501143358026882,
+      "learning_rate": 3.768964457582177e-06,
+      "loss": 1.6326,
+      "step": 5416
+    },
+    {
+      "epoch": 0.7228449426207633,
+      "grad_norm": 0.9523336545668618,
+      "learning_rate": 3.7655847077668915e-06,
+      "loss": 1.5764,
+      "step": 5417
+    },
+    {
+      "epoch": 0.7229783827061649,
+      "grad_norm": 1.0389137372207382,
+      "learning_rate": 3.7622061224694038e-06,
+      "loss": 1.5737,
+      "step": 5418
+    },
+    {
+      "epoch": 0.7231118227915666,
+      "grad_norm": 0.9455855773831311,
+      "learning_rate": 3.758828702320794e-06,
+      "loss": 1.5635,
+      "step": 5419
+    },
+    {
+      "epoch": 0.7232452628769682,
+      "grad_norm": 0.9693626405402851,
+      "learning_rate": 3.7554524479519196e-06,
+      "loss": 1.5343,
+      "step": 5420
+    },
+    {
+      "epoch": 0.72337870296237,
+      "grad_norm": 0.9574938541001241,
+      "learning_rate": 3.7520773599934347e-06,
+      "loss": 1.5699,
+      "step": 5421
+    },
+    {
+      "epoch": 0.7235121430477716,
+      "grad_norm": 0.940692924931385,
+      "learning_rate": 3.748703439075764e-06,
+      "loss": 1.5674,
+      "step": 5422
+    },
+    {
+      "epoch": 0.7236455831331732,
+      "grad_norm": 0.9270797016669773,
+      "learning_rate": 3.7453306858291163e-06,
+      "loss": 1.517,
+      "step": 5423
+    },
+    {
+      "epoch": 0.7237790232185749,
+      "grad_norm": 0.9375714221926796,
+      "learning_rate": 3.741959100883481e-06,
+      "loss": 1.5181,
+      "step": 5424
+    },
+    {
+      "epoch": 0.7239124633039765,
+      "grad_norm": 0.9283234071509436,
+      "learning_rate": 3.738588684868639e-06,
+      "loss": 1.5591,
+      "step": 5425
+    },
+    {
+      "epoch": 0.7240459033893781,
+      "grad_norm": 0.9733008805447092,
+      "learning_rate": 3.7352194384141426e-06,
+      "loss": 1.5131,
+      "step": 5426
+    },
+    {
+      "epoch": 0.7241793434747799,
+      "grad_norm": 0.9570280233761831,
+      "learning_rate": 3.731851362149327e-06,
+      "loss": 1.5628,
+      "step": 5427
+    },
+    {
+      "epoch": 0.7243127835601815,
+      "grad_norm": 1.2042047907577633,
+      "learning_rate": 3.7284844567033083e-06,
+      "loss": 1.538,
+      "step": 5428
+    },
+    {
+      "epoch": 0.7244462236455831,
+      "grad_norm": 0.9741593159201551,
+      "learning_rate": 3.7251187227049924e-06,
+      "loss": 1.5553,
+      "step": 5429
+    },
+    {
+      "epoch": 0.7245796637309848,
+      "grad_norm": 0.9197479513774383,
+      "learning_rate": 3.7217541607830576e-06,
+      "loss": 1.5162,
+      "step": 5430
+    },
+    {
+      "epoch": 0.7247131038163864,
+      "grad_norm": 0.9403320524684858,
+      "learning_rate": 3.718390771565964e-06,
+      "loss": 1.5288,
+      "step": 5431
+    },
+    {
+      "epoch": 0.724846543901788,
+      "grad_norm": 0.95182884051891,
+      "learning_rate": 3.7150285556819563e-06,
+      "loss": 1.5433,
+      "step": 5432
+    },
+    {
+      "epoch": 0.7249799839871898,
+      "grad_norm": 1.0304654498405337,
+      "learning_rate": 3.711667513759053e-06,
+      "loss": 1.622,
+      "step": 5433
+    },
+    {
+      "epoch": 0.7251134240725914,
+      "grad_norm": 0.98382338711681,
+      "learning_rate": 3.708307646425068e-06,
+      "loss": 1.5138,
+      "step": 5434
+    },
+    {
+      "epoch": 0.7252468641579931,
+      "grad_norm": 1.058008764362752,
+      "learning_rate": 3.704948954307579e-06,
+      "loss": 1.5539,
+      "step": 5435
+    },
+    {
+      "epoch": 0.7253803042433947,
+      "grad_norm": 1.0067806065661722,
+      "learning_rate": 3.7015914380339545e-06,
+      "loss": 1.5266,
+      "step": 5436
+    },
+    {
+      "epoch": 0.7255137443287963,
+      "grad_norm": 0.9964621996103804,
+      "learning_rate": 3.6982350982313387e-06,
+      "loss": 1.5728,
+      "step": 5437
+    },
+    {
+      "epoch": 0.7256471844141981,
+      "grad_norm": 1.2415140773272089,
+      "learning_rate": 3.694879935526655e-06,
+      "loss": 1.4833,
+      "step": 5438
+    },
+    {
+      "epoch": 0.7257806244995997,
+      "grad_norm": 1.0064068768265315,
+      "learning_rate": 3.691525950546617e-06,
+      "loss": 1.5633,
+      "step": 5439
+    },
+    {
+      "epoch": 0.7259140645850013,
+      "grad_norm": 1.335840952388961,
+      "learning_rate": 3.6881731439177058e-06,
+      "loss": 1.5479,
+      "step": 5440
+    },
+    {
+      "epoch": 0.726047504670403,
+      "grad_norm": 0.9765575128964399,
+      "learning_rate": 3.6848215162661894e-06,
+      "loss": 1.5502,
+      "step": 5441
+    },
+    {
+      "epoch": 0.7261809447558046,
+      "grad_norm": 0.9318261067807464,
+      "learning_rate": 3.6814710682181088e-06,
+      "loss": 1.525,
+      "step": 5442
+    },
+    {
+      "epoch": 0.7263143848412063,
+      "grad_norm": 0.9722119023080255,
+      "learning_rate": 3.6781218003992967e-06,
+      "loss": 1.5279,
+      "step": 5443
+    },
+    {
+      "epoch": 0.726447824926608,
+      "grad_norm": 0.9517245541221244,
+      "learning_rate": 3.6747737134353557e-06,
+      "loss": 1.5848,
+      "step": 5444
+    },
+    {
+      "epoch": 0.7265812650120096,
+      "grad_norm": 9.955006013526363,
+      "learning_rate": 3.6714268079516657e-06,
+      "loss": 1.6209,
+      "step": 5445
+    },
+    {
+      "epoch": 0.7267147050974113,
+      "grad_norm": 0.9547952003339406,
+      "learning_rate": 3.668081084573398e-06,
+      "loss": 1.6035,
+      "step": 5446
+    },
+    {
+      "epoch": 0.7268481451828129,
+      "grad_norm": 0.965245818039803,
+      "learning_rate": 3.6647365439254923e-06,
+      "loss": 1.5966,
+      "step": 5447
+    },
+    {
+      "epoch": 0.7269815852682145,
+      "grad_norm": 0.9878697703561382,
+      "learning_rate": 3.6613931866326704e-06,
+      "loss": 1.5274,
+      "step": 5448
+    },
+    {
+      "epoch": 0.7271150253536163,
+      "grad_norm": 0.9432582883437258,
+      "learning_rate": 3.658051013319429e-06,
+      "loss": 1.5539,
+      "step": 5449
+    },
+    {
+      "epoch": 0.7272484654390179,
+      "grad_norm": 0.9559304328129197,
+      "learning_rate": 3.6547100246100576e-06,
+      "loss": 1.5629,
+      "step": 5450
+    },
+    {
+      "epoch": 0.7273819055244195,
+      "grad_norm": 1.0818678239934452,
+      "learning_rate": 3.6513702211286086e-06,
+      "loss": 1.5481,
+      "step": 5451
+    },
+    {
+      "epoch": 0.7275153456098212,
+      "grad_norm": 0.9568005743025887,
+      "learning_rate": 3.6480316034989173e-06,
+      "loss": 1.5597,
+      "step": 5452
+    },
+    {
+      "epoch": 0.7276487856952228,
+      "grad_norm": 0.9041772250496767,
+      "learning_rate": 3.644694172344606e-06,
+      "loss": 1.534,
+      "step": 5453
+    },
+    {
+      "epoch": 0.7277822257806245,
+      "grad_norm": 0.9169233659753793,
+      "learning_rate": 3.6413579282890655e-06,
+      "loss": 1.5407,
+      "step": 5454
+    },
+    {
+      "epoch": 0.7279156658660262,
+      "grad_norm": 0.9466397250058866,
+      "learning_rate": 3.638022871955469e-06,
+      "loss": 1.5682,
+      "step": 5455
+    },
+    {
+      "epoch": 0.7280491059514278,
+      "grad_norm": 1.1961458887332446,
+      "learning_rate": 3.6346890039667616e-06,
+      "loss": 1.5361,
+      "step": 5456
+    },
+    {
+      "epoch": 0.7281825460368294,
+      "grad_norm": 0.9721270068626529,
+      "learning_rate": 3.6313563249456806e-06,
+      "loss": 1.5553,
+      "step": 5457
+    },
+    {
+      "epoch": 0.7283159861222311,
+      "grad_norm": 0.9791647245219625,
+      "learning_rate": 3.6280248355147273e-06,
+      "loss": 1.5297,
+      "step": 5458
+    },
+    {
+      "epoch": 0.7284494262076328,
+      "grad_norm": 0.9573169731204293,
+      "learning_rate": 3.6246945362961882e-06,
+      "loss": 1.5585,
+      "step": 5459
+    },
+    {
+      "epoch": 0.7285828662930345,
+      "grad_norm": 1.0013874444620199,
+      "learning_rate": 3.62136542791212e-06,
+      "loss": 1.5126,
+      "step": 5460
+    },
+    {
+      "epoch": 0.7287163063784361,
+      "grad_norm": 0.9187983530032887,
+      "learning_rate": 3.6180375109843703e-06,
+      "loss": 1.5512,
+      "step": 5461
+    },
+    {
+      "epoch": 0.7288497464638377,
+      "grad_norm": 1.0234386688020545,
+      "learning_rate": 3.614710786134552e-06,
+      "loss": 1.5312,
+      "step": 5462
+    },
+    {
+      "epoch": 0.7289831865492394,
+      "grad_norm": 0.9315553183193896,
+      "learning_rate": 3.6113852539840567e-06,
+      "loss": 1.5634,
+      "step": 5463
+    },
+    {
+      "epoch": 0.729116626634641,
+      "grad_norm": 0.925019388625177,
+      "learning_rate": 3.6080609151540613e-06,
+      "loss": 1.5594,
+      "step": 5464
+    },
+    {
+      "epoch": 0.7292500667200427,
+      "grad_norm": 0.935710933008227,
+      "learning_rate": 3.604737770265513e-06,
+      "loss": 1.4976,
+      "step": 5465
+    },
+    {
+      "epoch": 0.7293835068054444,
+      "grad_norm": 1.4420915957678724,
+      "learning_rate": 3.6014158199391358e-06,
+      "loss": 1.5603,
+      "step": 5466
+    },
+    {
+      "epoch": 0.729516946890846,
+      "grad_norm": 1.0497135426493642,
+      "learning_rate": 3.5980950647954294e-06,
+      "loss": 1.5248,
+      "step": 5467
+    },
+    {
+      "epoch": 0.7296503869762476,
+      "grad_norm": 0.9269061560231096,
+      "learning_rate": 3.5947755054546795e-06,
+      "loss": 1.5422,
+      "step": 5468
+    },
+    {
+      "epoch": 0.7297838270616493,
+      "grad_norm": 0.9527020311342057,
+      "learning_rate": 3.5914571425369393e-06,
+      "loss": 1.5373,
+      "step": 5469
+    },
+    {
+      "epoch": 0.729917267147051,
+      "grad_norm": 0.9372367174715784,
+      "learning_rate": 3.5881399766620373e-06,
+      "loss": 1.5657,
+      "step": 5470
+    },
+    {
+      "epoch": 0.7300507072324526,
+      "grad_norm": 0.9581054862273867,
+      "learning_rate": 3.58482400844959e-06,
+      "loss": 1.5739,
+      "step": 5471
+    },
+    {
+      "epoch": 0.7301841473178543,
+      "grad_norm": 0.947441074687153,
+      "learning_rate": 3.5815092385189777e-06,
+      "loss": 1.5728,
+      "step": 5472
+    },
+    {
+      "epoch": 0.7303175874032559,
+      "grad_norm": 1.0228746400085942,
+      "learning_rate": 3.5781956674893626e-06,
+      "loss": 1.5334,
+      "step": 5473
+    },
+    {
+      "epoch": 0.7304510274886576,
+      "grad_norm": 0.9536334081802301,
+      "learning_rate": 3.574883295979682e-06,
+      "loss": 1.562,
+      "step": 5474
+    },
+    {
+      "epoch": 0.7305844675740593,
+      "grad_norm": 0.9606672650917066,
+      "learning_rate": 3.5715721246086486e-06,
+      "loss": 1.5526,
+      "step": 5475
+    },
+    {
+      "epoch": 0.7307179076594609,
+      "grad_norm": 0.9577088048584481,
+      "learning_rate": 3.568262153994749e-06,
+      "loss": 1.5837,
+      "step": 5476
+    },
+    {
+      "epoch": 0.7308513477448626,
+      "grad_norm": 0.9719144515399968,
+      "learning_rate": 3.5649533847562544e-06,
+      "loss": 1.5316,
+      "step": 5477
+    },
+    {
+      "epoch": 0.7309847878302642,
+      "grad_norm": 1.004032695088233,
+      "learning_rate": 3.5616458175112013e-06,
+      "loss": 1.593,
+      "step": 5478
+    },
+    {
+      "epoch": 0.7311182279156658,
+      "grad_norm": 0.9494207677824381,
+      "learning_rate": 3.558339452877406e-06,
+      "loss": 1.5683,
+      "step": 5479
+    },
+    {
+      "epoch": 0.7312516680010676,
+      "grad_norm": 0.9698864247819369,
+      "learning_rate": 3.5550342914724613e-06,
+      "loss": 1.585,
+      "step": 5480
+    },
+    {
+      "epoch": 0.7313851080864692,
+      "grad_norm": 0.9508423495123989,
+      "learning_rate": 3.5517303339137267e-06,
+      "loss": 1.5207,
+      "step": 5481
+    },
+    {
+      "epoch": 0.7315185481718708,
+      "grad_norm": 1.2166047965486122,
+      "learning_rate": 3.5484275808183544e-06,
+      "loss": 1.5507,
+      "step": 5482
+    },
+    {
+      "epoch": 0.7316519882572725,
+      "grad_norm": 0.9436409445639349,
+      "learning_rate": 3.5451260328032555e-06,
+      "loss": 1.579,
+      "step": 5483
+    },
+    {
+      "epoch": 0.7317854283426741,
+      "grad_norm": 1.2067580184607811,
+      "learning_rate": 3.541825690485119e-06,
+      "loss": 1.5399,
+      "step": 5484
+    },
+    {
+      "epoch": 0.7319188684280757,
+      "grad_norm": 0.9128886571978432,
+      "learning_rate": 3.5385265544804172e-06,
+      "loss": 1.5099,
+      "step": 5485
+    },
+    {
+      "epoch": 0.7320523085134775,
+      "grad_norm": 0.9232766943936563,
+      "learning_rate": 3.535228625405389e-06,
+      "loss": 1.5688,
+      "step": 5486
+    },
+    {
+      "epoch": 0.7321857485988791,
+      "grad_norm": 0.9443761629982292,
+      "learning_rate": 3.531931903876049e-06,
+      "loss": 1.5515,
+      "step": 5487
+    },
+    {
+      "epoch": 0.7323191886842808,
+      "grad_norm": 0.9281491684957837,
+      "learning_rate": 3.5286363905081843e-06,
+      "loss": 1.5491,
+      "step": 5488
+    },
+    {
+      "epoch": 0.7324526287696824,
+      "grad_norm": 1.0926741045880084,
+      "learning_rate": 3.525342085917366e-06,
+      "loss": 1.5447,
+      "step": 5489
+    },
+    {
+      "epoch": 0.732586068855084,
+      "grad_norm": 0.9386932615205383,
+      "learning_rate": 3.5220489907189293e-06,
+      "loss": 1.5314,
+      "step": 5490
+    },
+    {
+      "epoch": 0.7327195089404858,
+      "grad_norm": 0.9166276573659027,
+      "learning_rate": 3.518757105527988e-06,
+      "loss": 1.5458,
+      "step": 5491
+    },
+    {
+      "epoch": 0.7328529490258874,
+      "grad_norm": 0.9123097547816297,
+      "learning_rate": 3.5154664309594224e-06,
+      "loss": 1.5524,
+      "step": 5492
+    },
+    {
+      "epoch": 0.732986389111289,
+      "grad_norm": 1.1139458388619161,
+      "learning_rate": 3.5121769676279026e-06,
+      "loss": 1.5188,
+      "step": 5493
+    },
+    {
+      "epoch": 0.7331198291966907,
+      "grad_norm": 0.958270690060531,
+      "learning_rate": 3.508888716147859e-06,
+      "loss": 1.5596,
+      "step": 5494
+    },
+    {
+      "epoch": 0.7332532692820923,
+      "grad_norm": 0.9709805252564494,
+      "learning_rate": 3.5056016771334954e-06,
+      "loss": 1.5756,
+      "step": 5495
+    },
+    {
+      "epoch": 0.733386709367494,
+      "grad_norm": 1.2247215933789883,
+      "learning_rate": 3.5023158511988e-06,
+      "loss": 1.597,
+      "step": 5496
+    },
+    {
+      "epoch": 0.7335201494528957,
+      "grad_norm": 0.9334813075768208,
+      "learning_rate": 3.4990312389575253e-06,
+      "loss": 1.5761,
+      "step": 5497
+    },
+    {
+      "epoch": 0.7336535895382973,
+      "grad_norm": 0.9468394042300734,
+      "learning_rate": 3.495747841023198e-06,
+      "loss": 1.547,
+      "step": 5498
+    },
+    {
+      "epoch": 0.7337870296236989,
+      "grad_norm": 0.9478146984584698,
+      "learning_rate": 3.4924656580091176e-06,
+      "loss": 1.6384,
+      "step": 5499
+    },
+    {
+      "epoch": 0.7339204697091006,
+      "grad_norm": 0.967091384617016,
+      "learning_rate": 3.4891846905283645e-06,
+      "loss": 1.5693,
+      "step": 5500
+    },
+    {
+      "epoch": 0.7340539097945022,
+      "grad_norm": 0.9249763240908678,
+      "learning_rate": 3.4859049391937827e-06,
+      "loss": 1.5253,
+      "step": 5501
+    },
+    {
+      "epoch": 0.734187349879904,
+      "grad_norm": 0.96990837542742,
+      "learning_rate": 3.48262640461799e-06,
+      "loss": 1.5783,
+      "step": 5502
+    },
+    {
+      "epoch": 0.7343207899653056,
+      "grad_norm": 0.9655104010410549,
+      "learning_rate": 3.479349087413384e-06,
+      "loss": 1.5925,
+      "step": 5503
+    },
+    {
+      "epoch": 0.7344542300507072,
+      "grad_norm": 0.9165403211065593,
+      "learning_rate": 3.4760729881921286e-06,
+      "loss": 1.4963,
+      "step": 5504
+    },
+    {
+      "epoch": 0.7345876701361089,
+      "grad_norm": 0.9683466520946085,
+      "learning_rate": 3.472798107566161e-06,
+      "loss": 1.6019,
+      "step": 5505
+    },
+    {
+      "epoch": 0.7347211102215105,
+      "grad_norm": 0.9552291626544579,
+      "learning_rate": 3.469524446147189e-06,
+      "loss": 1.5571,
+      "step": 5506
+    },
+    {
+      "epoch": 0.7348545503069122,
+      "grad_norm": 0.9311445516993911,
+      "learning_rate": 3.466252004546702e-06,
+      "loss": 1.5499,
+      "step": 5507
+    },
+    {
+      "epoch": 0.7349879903923139,
+      "grad_norm": 1.1146712718754999,
+      "learning_rate": 3.46298078337595e-06,
+      "loss": 1.5284,
+      "step": 5508
+    },
+    {
+      "epoch": 0.7351214304777155,
+      "grad_norm": 0.9571872321016519,
+      "learning_rate": 3.4597107832459575e-06,
+      "loss": 1.6109,
+      "step": 5509
+    },
+    {
+      "epoch": 0.7352548705631171,
+      "grad_norm": 0.9544626733275546,
+      "learning_rate": 3.456442004767531e-06,
+      "loss": 1.584,
+      "step": 5510
+    },
+    {
+      "epoch": 0.7353883106485188,
+      "grad_norm": 0.9436856603892109,
+      "learning_rate": 3.4531744485512362e-06,
+      "loss": 1.6014,
+      "step": 5511
+    },
+    {
+      "epoch": 0.7355217507339205,
+      "grad_norm": 0.9393458177163648,
+      "learning_rate": 3.4499081152074156e-06,
+      "loss": 1.5539,
+      "step": 5512
+    },
+    {
+      "epoch": 0.7356551908193221,
+      "grad_norm": 0.9422174761137512,
+      "learning_rate": 3.446643005346181e-06,
+      "loss": 1.569,
+      "step": 5513
+    },
+    {
+      "epoch": 0.7357886309047238,
+      "grad_norm": 0.9392923260054774,
+      "learning_rate": 3.4433791195774237e-06,
+      "loss": 1.545,
+      "step": 5514
+    },
+    {
+      "epoch": 0.7359220709901254,
+      "grad_norm": 0.9545864141748973,
+      "learning_rate": 3.440116458510796e-06,
+      "loss": 1.5286,
+      "step": 5515
+    },
+    {
+      "epoch": 0.7360555110755271,
+      "grad_norm": 0.9852651403635309,
+      "learning_rate": 3.4368550227557272e-06,
+      "loss": 1.5654,
+      "step": 5516
+    },
+    {
+      "epoch": 0.7361889511609288,
+      "grad_norm": 0.9339962134756217,
+      "learning_rate": 3.433594812921416e-06,
+      "loss": 1.507,
+      "step": 5517
+    },
+    {
+      "epoch": 0.7363223912463304,
+      "grad_norm": 0.9437597981521707,
+      "learning_rate": 3.4303358296168287e-06,
+      "loss": 1.5623,
+      "step": 5518
+    },
+    {
+      "epoch": 0.7364558313317321,
+      "grad_norm": 0.9428139658289024,
+      "learning_rate": 3.4270780734507136e-06,
+      "loss": 1.514,
+      "step": 5519
+    },
+    {
+      "epoch": 0.7365892714171337,
+      "grad_norm": 0.9476341999117343,
+      "learning_rate": 3.4238215450315794e-06,
+      "loss": 1.5945,
+      "step": 5520
+    },
+    {
+      "epoch": 0.7367227115025353,
+      "grad_norm": 0.9394390246936388,
+      "learning_rate": 3.420566244967707e-06,
+      "loss": 1.5188,
+      "step": 5521
+    },
+    {
+      "epoch": 0.736856151587937,
+      "grad_norm": 0.9506593690055114,
+      "learning_rate": 3.417312173867151e-06,
+      "loss": 1.5694,
+      "step": 5522
+    },
+    {
+      "epoch": 0.7369895916733387,
+      "grad_norm": 1.0129257670360825,
+      "learning_rate": 3.4140593323377336e-06,
+      "loss": 1.6033,
+      "step": 5523
+    },
+    {
+      "epoch": 0.7371230317587403,
+      "grad_norm": 0.9480996004589485,
+      "learning_rate": 3.410807720987046e-06,
+      "loss": 1.5339,
+      "step": 5524
+    },
+    {
+      "epoch": 0.737256471844142,
+      "grad_norm": 0.927903797277357,
+      "learning_rate": 3.4075573404224594e-06,
+      "loss": 1.5274,
+      "step": 5525
+    },
+    {
+      "epoch": 0.7373899119295436,
+      "grad_norm": 0.9379989494588817,
+      "learning_rate": 3.4043081912511033e-06,
+      "loss": 1.5647,
+      "step": 5526
+    },
+    {
+      "epoch": 0.7375233520149453,
+      "grad_norm": 0.9218438261963459,
+      "learning_rate": 3.4010602740798795e-06,
+      "loss": 1.5587,
+      "step": 5527
+    },
+    {
+      "epoch": 0.737656792100347,
+      "grad_norm": 0.9400943430420226,
+      "learning_rate": 3.3978135895154674e-06,
+      "loss": 1.5383,
+      "step": 5528
+    },
+    {
+      "epoch": 0.7377902321857486,
+      "grad_norm": 0.9268874491253772,
+      "learning_rate": 3.394568138164308e-06,
+      "loss": 1.5851,
+      "step": 5529
+    },
+    {
+      "epoch": 0.7379236722711503,
+      "grad_norm": 0.9115511290678305,
+      "learning_rate": 3.3913239206326154e-06,
+      "loss": 1.5734,
+      "step": 5530
+    },
+    {
+      "epoch": 0.7380571123565519,
+      "grad_norm": 0.9222620027598017,
+      "learning_rate": 3.388080937526368e-06,
+      "loss": 1.5531,
+      "step": 5531
+    },
+    {
+      "epoch": 0.7381905524419535,
+      "grad_norm": 0.9826304238549518,
+      "learning_rate": 3.3848391894513253e-06,
+      "loss": 1.5357,
+      "step": 5532
+    },
+    {
+      "epoch": 0.7383239925273553,
+      "grad_norm": 0.9812896742852667,
+      "learning_rate": 3.381598677013006e-06,
+      "loss": 1.6455,
+      "step": 5533
+    },
+    {
+      "epoch": 0.7384574326127569,
+      "grad_norm": 1.1940903464741186,
+      "learning_rate": 3.3783594008166963e-06,
+      "loss": 1.5608,
+      "step": 5534
+    },
+    {
+      "epoch": 0.7385908726981585,
+      "grad_norm": 1.0051527729116323,
+      "learning_rate": 3.3751213614674647e-06,
+      "loss": 1.4906,
+      "step": 5535
+    },
+    {
+      "epoch": 0.7387243127835602,
+      "grad_norm": 0.9448310950987612,
+      "learning_rate": 3.3718845595701356e-06,
+      "loss": 1.5768,
+      "step": 5536
+    },
+    {
+      "epoch": 0.7388577528689618,
+      "grad_norm": 0.9253739341701587,
+      "learning_rate": 3.3686489957293057e-06,
+      "loss": 1.5335,
+      "step": 5537
+    },
+    {
+      "epoch": 0.7389911929543634,
+      "grad_norm": 0.9565771682432299,
+      "learning_rate": 3.3654146705493407e-06,
+      "loss": 1.5332,
+      "step": 5538
+    },
+    {
+      "epoch": 0.7391246330397652,
+      "grad_norm": 0.9417608192865259,
+      "learning_rate": 3.3621815846343797e-06,
+      "loss": 1.5608,
+      "step": 5539
+    },
+    {
+      "epoch": 0.7392580731251668,
+      "grad_norm": 0.9655822978265454,
+      "learning_rate": 3.3589497385883253e-06,
+      "loss": 1.5471,
+      "step": 5540
+    },
+    {
+      "epoch": 0.7393915132105685,
+      "grad_norm": 0.9690529415186309,
+      "learning_rate": 3.355719133014844e-06,
+      "loss": 1.5502,
+      "step": 5541
+    },
+    {
+      "epoch": 0.7395249532959701,
+      "grad_norm": 1.2161957457480974,
+      "learning_rate": 3.3524897685173853e-06,
+      "loss": 1.5819,
+      "step": 5542
+    },
+    {
+      "epoch": 0.7396583933813717,
+      "grad_norm": 0.9441515276869841,
+      "learning_rate": 3.3492616456991524e-06,
+      "loss": 1.5041,
+      "step": 5543
+    },
+    {
+      "epoch": 0.7397918334667735,
+      "grad_norm": 0.960027360606955,
+      "learning_rate": 3.346034765163123e-06,
+      "loss": 1.5315,
+      "step": 5544
+    },
+    {
+      "epoch": 0.7399252735521751,
+      "grad_norm": 0.9330330601144551,
+      "learning_rate": 3.342809127512038e-06,
+      "loss": 1.5264,
+      "step": 5545
+    },
+    {
+      "epoch": 0.7400587136375767,
+      "grad_norm": 1.1030181822089198,
+      "learning_rate": 3.3395847333484153e-06,
+      "loss": 1.5608,
+      "step": 5546
+    },
+    {
+      "epoch": 0.7401921537229784,
+      "grad_norm": 0.9682087080598382,
+      "learning_rate": 3.3363615832745344e-06,
+      "loss": 1.5755,
+      "step": 5547
+    },
+    {
+      "epoch": 0.74032559380838,
+      "grad_norm": 0.9159493952844268,
+      "learning_rate": 3.333139677892436e-06,
+      "loss": 1.555,
+      "step": 5548
+    },
+    {
+      "epoch": 0.7404590338937816,
+      "grad_norm": 0.9659206359188842,
+      "learning_rate": 3.3299190178039464e-06,
+      "loss": 1.6404,
+      "step": 5549
+    },
+    {
+      "epoch": 0.7405924739791834,
+      "grad_norm": 1.1568043350228006,
+      "learning_rate": 3.3266996036106415e-06,
+      "loss": 1.5864,
+      "step": 5550
+    },
+    {
+      "epoch": 0.740725914064585,
+      "grad_norm": 0.923838794736863,
+      "learning_rate": 3.323481435913871e-06,
+      "loss": 1.5286,
+      "step": 5551
+    },
+    {
+      "epoch": 0.7408593541499866,
+      "grad_norm": 0.9701142939931843,
+      "learning_rate": 3.320264515314752e-06,
+      "loss": 1.515,
+      "step": 5552
+    },
+    {
+      "epoch": 0.7409927942353883,
+      "grad_norm": 0.9266738295815183,
+      "learning_rate": 3.3170488424141713e-06,
+      "loss": 1.5394,
+      "step": 5553
+    },
+    {
+      "epoch": 0.74112623432079,
+      "grad_norm": 0.9206286448502721,
+      "learning_rate": 3.31383441781278e-06,
+      "loss": 1.5438,
+      "step": 5554
+    },
+    {
+      "epoch": 0.7412596744061917,
+      "grad_norm": 0.931353809690202,
+      "learning_rate": 3.310621242110994e-06,
+      "loss": 1.5102,
+      "step": 5555
+    },
+    {
+      "epoch": 0.7413931144915933,
+      "grad_norm": 0.9508289819339473,
+      "learning_rate": 3.307409315908995e-06,
+      "loss": 1.5811,
+      "step": 5556
+    },
+    {
+      "epoch": 0.7415265545769949,
+      "grad_norm": 0.959176272650246,
+      "learning_rate": 3.3041986398067416e-06,
+      "loss": 1.599,
+      "step": 5557
+    },
+    {
+      "epoch": 0.7416599946623966,
+      "grad_norm": 0.9375548404744494,
+      "learning_rate": 3.3009892144039478e-06,
+      "loss": 1.5679,
+      "step": 5558
+    },
+    {
+      "epoch": 0.7417934347477982,
+      "grad_norm": 0.9217050733117328,
+      "learning_rate": 3.2977810403000977e-06,
+      "loss": 1.5366,
+      "step": 5559
+    },
+    {
+      "epoch": 0.7419268748331999,
+      "grad_norm": 0.9481402337066958,
+      "learning_rate": 3.29457411809444e-06,
+      "loss": 1.5539,
+      "step": 5560
+    },
+    {
+      "epoch": 0.7420603149186016,
+      "grad_norm": 0.9269200198823844,
+      "learning_rate": 3.291368448385992e-06,
+      "loss": 1.511,
+      "step": 5561
+    },
+    {
+      "epoch": 0.7421937550040032,
+      "grad_norm": 0.9484784210641377,
+      "learning_rate": 3.2881640317735386e-06,
+      "loss": 1.5821,
+      "step": 5562
+    },
+    {
+      "epoch": 0.7423271950894048,
+      "grad_norm": 0.9728501168391839,
+      "learning_rate": 3.2849608688556276e-06,
+      "loss": 1.5608,
+      "step": 5563
+    },
+    {
+      "epoch": 0.7424606351748065,
+      "grad_norm": 0.9582363784642031,
+      "learning_rate": 3.2817589602305732e-06,
+      "loss": 1.5794,
+      "step": 5564
+    },
+    {
+      "epoch": 0.7425940752602082,
+      "grad_norm": 0.9398306842093537,
+      "learning_rate": 3.2785583064964545e-06,
+      "loss": 1.5571,
+      "step": 5565
+    },
+    {
+      "epoch": 0.7427275153456098,
+      "grad_norm": 0.9467894900781275,
+      "learning_rate": 3.2753589082511152e-06,
+      "loss": 1.5512,
+      "step": 5566
+    },
+    {
+      "epoch": 0.7428609554310115,
+      "grad_norm": 0.9367465516942596,
+      "learning_rate": 3.2721607660921716e-06,
+      "loss": 1.5603,
+      "step": 5567
+    },
+    {
+      "epoch": 0.7429943955164131,
+      "grad_norm": 0.9499101812547797,
+      "learning_rate": 3.2689638806169977e-06,
+      "loss": 1.5349,
+      "step": 5568
+    },
+    {
+      "epoch": 0.7431278356018148,
+      "grad_norm": 0.9191430811722798,
+      "learning_rate": 3.265768252422734e-06,
+      "loss": 1.4816,
+      "step": 5569
+    },
+    {
+      "epoch": 0.7432612756872164,
+      "grad_norm": 1.0165076519034235,
+      "learning_rate": 3.2625738821062868e-06,
+      "loss": 1.5599,
+      "step": 5570
+    },
+    {
+      "epoch": 0.7433947157726181,
+      "grad_norm": 0.9387670444451917,
+      "learning_rate": 3.259380770264332e-06,
+      "loss": 1.5309,
+      "step": 5571
+    },
+    {
+      "epoch": 0.7435281558580198,
+      "grad_norm": 0.970178143993145,
+      "learning_rate": 3.256188917493306e-06,
+      "loss": 1.5265,
+      "step": 5572
+    },
+    {
+      "epoch": 0.7436615959434214,
+      "grad_norm": 0.9464180115309929,
+      "learning_rate": 3.2529983243894046e-06,
+      "loss": 1.5373,
+      "step": 5573
+    },
+    {
+      "epoch": 0.743795036028823,
+      "grad_norm": 0.9513946770758279,
+      "learning_rate": 3.2498089915486032e-06,
+      "loss": 1.5375,
+      "step": 5574
+    },
+    {
+      "epoch": 0.7439284761142247,
+      "grad_norm": 0.9395784172257111,
+      "learning_rate": 3.2466209195666266e-06,
+      "loss": 1.5607,
+      "step": 5575
+    },
+    {
+      "epoch": 0.7440619161996264,
+      "grad_norm": 1.081019248269522,
+      "learning_rate": 3.2434341090389734e-06,
+      "loss": 1.5749,
+      "step": 5576
+    },
+    {
+      "epoch": 0.744195356285028,
+      "grad_norm": 0.9327852617129883,
+      "learning_rate": 3.240248560560899e-06,
+      "loss": 1.5236,
+      "step": 5577
+    },
+    {
+      "epoch": 0.7443287963704297,
+      "grad_norm": 0.9878773231924048,
+      "learning_rate": 3.237064274727433e-06,
+      "loss": 1.5515,
+      "step": 5578
+    },
+    {
+      "epoch": 0.7444622364558313,
+      "grad_norm": 1.010601298560983,
+      "learning_rate": 3.233881252133363e-06,
+      "loss": 1.5761,
+      "step": 5579
+    },
+    {
+      "epoch": 0.7445956765412329,
+      "grad_norm": 0.9550690993331595,
+      "learning_rate": 3.230699493373236e-06,
+      "loss": 1.5799,
+      "step": 5580
+    },
+    {
+      "epoch": 0.7447291166266347,
+      "grad_norm": 0.979629622187405,
+      "learning_rate": 3.2275189990413746e-06,
+      "loss": 1.5323,
+      "step": 5581
+    },
+    {
+      "epoch": 0.7448625567120363,
+      "grad_norm": 1.0433957484563827,
+      "learning_rate": 3.224339769731858e-06,
+      "loss": 1.4937,
+      "step": 5582
+    },
+    {
+      "epoch": 0.744995996797438,
+      "grad_norm": 0.9320393784593036,
+      "learning_rate": 3.2211618060385285e-06,
+      "loss": 1.5419,
+      "step": 5583
+    },
+    {
+      "epoch": 0.7451294368828396,
+      "grad_norm": 0.9316119615350978,
+      "learning_rate": 3.2179851085549897e-06,
+      "loss": 1.5457,
+      "step": 5584
+    },
+    {
+      "epoch": 0.7452628769682412,
+      "grad_norm": 0.9283803212880422,
+      "learning_rate": 3.2148096778746195e-06,
+      "loss": 1.5546,
+      "step": 5585
+    },
+    {
+      "epoch": 0.745396317053643,
+      "grad_norm": 0.9313780199697975,
+      "learning_rate": 3.21163551459055e-06,
+      "loss": 1.5485,
+      "step": 5586
+    },
+    {
+      "epoch": 0.7455297571390446,
+      "grad_norm": 1.001529063138306,
+      "learning_rate": 3.2084626192956745e-06,
+      "loss": 1.5779,
+      "step": 5587
+    },
+    {
+      "epoch": 0.7456631972244462,
+      "grad_norm": 1.0662137147201587,
+      "learning_rate": 3.2052909925826604e-06,
+      "loss": 1.5484,
+      "step": 5588
+    },
+    {
+      "epoch": 0.7457966373098479,
+      "grad_norm": 0.9485702607614251,
+      "learning_rate": 3.202120635043928e-06,
+      "loss": 1.5674,
+      "step": 5589
+    },
+    {
+      "epoch": 0.7459300773952495,
+      "grad_norm": 0.9472743543057982,
+      "learning_rate": 3.198951547271665e-06,
+      "loss": 1.5625,
+      "step": 5590
+    },
+    {
+      "epoch": 0.7460635174806511,
+      "grad_norm": 0.9051518915153008,
+      "learning_rate": 3.1957837298578154e-06,
+      "loss": 1.5522,
+      "step": 5591
+    },
+    {
+      "epoch": 0.7461969575660529,
+      "grad_norm": 0.9130917382566023,
+      "learning_rate": 3.1926171833941e-06,
+      "loss": 1.5459,
+      "step": 5592
+    },
+    {
+      "epoch": 0.7463303976514545,
+      "grad_norm": 1.149337316223058,
+      "learning_rate": 3.18945190847199e-06,
+      "loss": 1.5367,
+      "step": 5593
+    },
+    {
+      "epoch": 0.7464638377368562,
+      "grad_norm": 1.0675304023468861,
+      "learning_rate": 3.1862879056827225e-06,
+      "loss": 1.5151,
+      "step": 5594
+    },
+    {
+      "epoch": 0.7465972778222578,
+      "grad_norm": 0.9568872060715193,
+      "learning_rate": 3.1831251756172943e-06,
+      "loss": 1.5452,
+      "step": 5595
+    },
+    {
+      "epoch": 0.7467307179076594,
+      "grad_norm": 1.0314564637633439,
+      "learning_rate": 3.1799637188664736e-06,
+      "loss": 1.5902,
+      "step": 5596
+    },
+    {
+      "epoch": 0.7468641579930612,
+      "grad_norm": 0.9252768296530237,
+      "learning_rate": 3.1768035360207806e-06,
+      "loss": 1.5499,
+      "step": 5597
+    },
+    {
+      "epoch": 0.7469975980784628,
+      "grad_norm": 1.2637363917190654,
+      "learning_rate": 3.173644627670499e-06,
+      "loss": 1.505,
+      "step": 5598
+    },
+    {
+      "epoch": 0.7471310381638644,
+      "grad_norm": 1.0200228137515415,
+      "learning_rate": 3.1704869944056826e-06,
+      "loss": 1.583,
+      "step": 5599
+    },
+    {
+      "epoch": 0.7472644782492661,
+      "grad_norm": 1.2623537137006278,
+      "learning_rate": 3.1673306368161394e-06,
+      "loss": 1.5582,
+      "step": 5600
+    },
+    {
+      "epoch": 0.7473979183346677,
+      "grad_norm": 0.9942148402882126,
+      "learning_rate": 3.16417555549144e-06,
+      "loss": 1.5138,
+      "step": 5601
+    },
+    {
+      "epoch": 0.7475313584200693,
+      "grad_norm": 0.9769665826208432,
+      "learning_rate": 3.1610217510209184e-06,
+      "loss": 1.5111,
+      "step": 5602
+    },
+    {
+      "epoch": 0.7476647985054711,
+      "grad_norm": 1.0229850987939395,
+      "learning_rate": 3.1578692239936693e-06,
+      "loss": 1.5134,
+      "step": 5603
+    },
+    {
+      "epoch": 0.7477982385908727,
+      "grad_norm": 0.9351520451791383,
+      "learning_rate": 3.1547179749985445e-06,
+      "loss": 1.6311,
+      "step": 5604
+    },
+    {
+      "epoch": 0.7479316786762743,
+      "grad_norm": 0.9204017069129004,
+      "learning_rate": 3.1515680046241693e-06,
+      "loss": 1.5303,
+      "step": 5605
+    },
+    {
+      "epoch": 0.748065118761676,
+      "grad_norm": 0.9599046278620328,
+      "learning_rate": 3.148419313458918e-06,
+      "loss": 1.5719,
+      "step": 5606
+    },
+    {
+      "epoch": 0.7481985588470776,
+      "grad_norm": 0.948039514652451,
+      "learning_rate": 3.1452719020909317e-06,
+      "loss": 1.5452,
+      "step": 5607
+    },
+    {
+      "epoch": 0.7483319989324794,
+      "grad_norm": 0.9481417279723116,
+      "learning_rate": 3.1421257711081097e-06,
+      "loss": 1.5983,
+      "step": 5608
+    },
+    {
+      "epoch": 0.748465439017881,
+      "grad_norm": 0.9485239077024988,
+      "learning_rate": 3.13898092109811e-06,
+      "loss": 1.531,
+      "step": 5609
+    },
+    {
+      "epoch": 0.7485988791032826,
+      "grad_norm": 1.1358614513622873,
+      "learning_rate": 3.135837352648362e-06,
+      "loss": 1.5357,
+      "step": 5610
+    },
+    {
+      "epoch": 0.7487323191886843,
+      "grad_norm": 0.970237163017964,
+      "learning_rate": 3.1326950663460466e-06,
+      "loss": 1.5521,
+      "step": 5611
+    },
+    {
+      "epoch": 0.7488657592740859,
+      "grad_norm": 0.9433944906141754,
+      "learning_rate": 3.1295540627781006e-06,
+      "loss": 1.6221,
+      "step": 5612
+    },
+    {
+      "epoch": 0.7489991993594876,
+      "grad_norm": 1.0020127389963633,
+      "learning_rate": 3.1264143425312366e-06,
+      "loss": 1.5551,
+      "step": 5613
+    },
+    {
+      "epoch": 0.7491326394448893,
+      "grad_norm": 0.9670401814605709,
+      "learning_rate": 3.123275906191915e-06,
+      "loss": 1.5776,
+      "step": 5614
+    },
+    {
+      "epoch": 0.7492660795302909,
+      "grad_norm": 0.9361899965589122,
+      "learning_rate": 3.120138754346359e-06,
+      "loss": 1.5674,
+      "step": 5615
+    },
+    {
+      "epoch": 0.7493995196156925,
+      "grad_norm": 0.976493101276714,
+      "learning_rate": 3.1170028875805504e-06,
+      "loss": 1.5411,
+      "step": 5616
+    },
+    {
+      "epoch": 0.7495329597010942,
+      "grad_norm": 0.9634598582821797,
+      "learning_rate": 3.11386830648024e-06,
+      "loss": 1.5704,
+      "step": 5617
+    },
+    {
+      "epoch": 0.7496663997864959,
+      "grad_norm": 0.9640441671842809,
+      "learning_rate": 3.1107350116309275e-06,
+      "loss": 1.5383,
+      "step": 5618
+    },
+    {
+      "epoch": 0.7497998398718975,
+      "grad_norm": 0.9555181724485184,
+      "learning_rate": 3.1076030036178763e-06,
+      "loss": 1.6026,
+      "step": 5619
+    },
+    {
+      "epoch": 0.7499332799572992,
+      "grad_norm": 1.2807705278680843,
+      "learning_rate": 3.104472283026113e-06,
+      "loss": 1.5881,
+      "step": 5620
+    },
+    {
+      "epoch": 0.7500667200427008,
+      "grad_norm": 0.9356127813233456,
+      "learning_rate": 3.1013428504404187e-06,
+      "loss": 1.561,
+      "step": 5621
+    },
+    {
+      "epoch": 0.7502001601281025,
+      "grad_norm": 0.9571532107573171,
+      "learning_rate": 3.098214706445336e-06,
+      "loss": 1.5119,
+      "step": 5622
+    },
+    {
+      "epoch": 0.7503336002135041,
+      "grad_norm": 0.9491196174548282,
+      "learning_rate": 3.0950878516251636e-06,
+      "loss": 1.5096,
+      "step": 5623
+    },
+    {
+      "epoch": 0.7504670402989058,
+      "grad_norm": 0.9417652753789462,
+      "learning_rate": 3.0919622865639677e-06,
+      "loss": 1.5102,
+      "step": 5624
+    },
+    {
+      "epoch": 0.7506004803843075,
+      "grad_norm": 0.9630569879115026,
+      "learning_rate": 3.088838011845566e-06,
+      "loss": 1.5818,
+      "step": 5625
+    },
+    {
+      "epoch": 0.7507339204697091,
+      "grad_norm": 0.9368704080314552,
+      "learning_rate": 3.0857150280535365e-06,
+      "loss": 1.5337,
+      "step": 5626
+    },
+    {
+      "epoch": 0.7508673605551107,
+      "grad_norm": 0.9571714147311596,
+      "learning_rate": 3.082593335771216e-06,
+      "loss": 1.5982,
+      "step": 5627
+    },
+    {
+      "epoch": 0.7510008006405124,
+      "grad_norm": 0.9254871682343206,
+      "learning_rate": 3.079472935581704e-06,
+      "loss": 1.5988,
+      "step": 5628
+    },
+    {
+      "epoch": 0.7511342407259141,
+      "grad_norm": 0.9612391153764122,
+      "learning_rate": 3.0763538280678563e-06,
+      "loss": 1.5205,
+      "step": 5629
+    },
+    {
+      "epoch": 0.7512676808113157,
+      "grad_norm": 0.952348233253775,
+      "learning_rate": 3.0732360138122806e-06,
+      "loss": 1.5686,
+      "step": 5630
+    },
+    {
+      "epoch": 0.7514011208967174,
+      "grad_norm": 0.9494761120662082,
+      "learning_rate": 3.0701194933973566e-06,
+      "loss": 1.5929,
+      "step": 5631
+    },
+    {
+      "epoch": 0.751534560982119,
+      "grad_norm": 0.917699574591178,
+      "learning_rate": 3.0670042674052116e-06,
+      "loss": 1.524,
+      "step": 5632
+    },
+    {
+      "epoch": 0.7516680010675206,
+      "grad_norm": 1.0333026625817825,
+      "learning_rate": 3.0638903364177343e-06,
+      "loss": 1.5439,
+      "step": 5633
+    },
+    {
+      "epoch": 0.7518014411529224,
+      "grad_norm": 0.9417912177877552,
+      "learning_rate": 3.0607777010165683e-06,
+      "loss": 1.6148,
+      "step": 5634
+    },
+    {
+      "epoch": 0.751934881238324,
+      "grad_norm": 0.9406321738897914,
+      "learning_rate": 3.057666361783126e-06,
+      "loss": 1.583,
+      "step": 5635
+    },
+    {
+      "epoch": 0.7520683213237257,
+      "grad_norm": 1.1578117225090305,
+      "learning_rate": 3.054556319298565e-06,
+      "loss": 1.5209,
+      "step": 5636
+    },
+    {
+      "epoch": 0.7522017614091273,
+      "grad_norm": 0.9600519785804953,
+      "learning_rate": 3.051447574143803e-06,
+      "loss": 1.5374,
+      "step": 5637
+    },
+    {
+      "epoch": 0.7523352014945289,
+      "grad_norm": 0.9270313809638928,
+      "learning_rate": 3.048340126899526e-06,
+      "loss": 1.5212,
+      "step": 5638
+    },
+    {
+      "epoch": 0.7524686415799307,
+      "grad_norm": 0.9114207753606299,
+      "learning_rate": 3.045233978146166e-06,
+      "loss": 1.5157,
+      "step": 5639
+    },
+    {
+      "epoch": 0.7526020816653323,
+      "grad_norm": 0.981499892052798,
+      "learning_rate": 3.0421291284639143e-06,
+      "loss": 1.5459,
+      "step": 5640
+    },
+    {
+      "epoch": 0.7527355217507339,
+      "grad_norm": 0.9458653564470572,
+      "learning_rate": 3.0390255784327215e-06,
+      "loss": 1.5687,
+      "step": 5641
+    },
+    {
+      "epoch": 0.7528689618361356,
+      "grad_norm": 1.2124512830726673,
+      "learning_rate": 3.0359233286322986e-06,
+      "loss": 1.5388,
+      "step": 5642
+    },
+    {
+      "epoch": 0.7530024019215372,
+      "grad_norm": 0.9508117630773106,
+      "learning_rate": 3.0328223796421107e-06,
+      "loss": 1.5721,
+      "step": 5643
+    },
+    {
+      "epoch": 0.7531358420069388,
+      "grad_norm": 0.9592182453362529,
+      "learning_rate": 3.029722732041377e-06,
+      "loss": 1.5769,
+      "step": 5644
+    },
+    {
+      "epoch": 0.7532692820923406,
+      "grad_norm": 1.08965444186456,
+      "learning_rate": 3.0266243864090772e-06,
+      "loss": 1.5829,
+      "step": 5645
+    },
+    {
+      "epoch": 0.7534027221777422,
+      "grad_norm": 2.9774916226895147,
+      "learning_rate": 3.0235273433239475e-06,
+      "loss": 1.5859,
+      "step": 5646
+    },
+    {
+      "epoch": 0.7535361622631438,
+      "grad_norm": 1.1074468709821748,
+      "learning_rate": 3.0204316033644765e-06,
+      "loss": 1.5437,
+      "step": 5647
+    },
+    {
+      "epoch": 0.7536696023485455,
+      "grad_norm": 0.9565752685671445,
+      "learning_rate": 3.01733716710892e-06,
+      "loss": 1.5924,
+      "step": 5648
+    },
+    {
+      "epoch": 0.7538030424339471,
+      "grad_norm": 0.9585562387187216,
+      "learning_rate": 3.0142440351352797e-06,
+      "loss": 1.5454,
+      "step": 5649
+    },
+    {
+      "epoch": 0.7539364825193489,
+      "grad_norm": 1.056117358661078,
+      "learning_rate": 3.0111522080213185e-06,
+      "loss": 1.5113,
+      "step": 5650
+    },
+    {
+      "epoch": 0.7540699226047505,
+      "grad_norm": 1.2766508430903627,
+      "learning_rate": 3.008061686344551e-06,
+      "loss": 1.5445,
+      "step": 5651
+    },
+    {
+      "epoch": 0.7542033626901521,
+      "grad_norm": 0.9812676587742001,
+      "learning_rate": 3.0049724706822557e-06,
+      "loss": 1.5591,
+      "step": 5652
+    },
+    {
+      "epoch": 0.7543368027755538,
+      "grad_norm": 0.9349340551684687,
+      "learning_rate": 3.001884561611463e-06,
+      "loss": 1.5658,
+      "step": 5653
+    },
+    {
+      "epoch": 0.7544702428609554,
+      "grad_norm": 1.2097788395210352,
+      "learning_rate": 2.998797959708958e-06,
+      "loss": 1.4703,
+      "step": 5654
+    },
+    {
+      "epoch": 0.754603682946357,
+      "grad_norm": 0.9150765980891424,
+      "learning_rate": 2.995712665551278e-06,
+      "loss": 1.5338,
+      "step": 5655
+    },
+    {
+      "epoch": 0.7547371230317588,
+      "grad_norm": 1.1493932341594215,
+      "learning_rate": 2.9926286797147284e-06,
+      "loss": 1.5119,
+      "step": 5656
+    },
+    {
+      "epoch": 0.7548705631171604,
+      "grad_norm": 0.9295681174421622,
+      "learning_rate": 2.989546002775361e-06,
+      "loss": 1.5785,
+      "step": 5657
+    },
+    {
+      "epoch": 0.755004003202562,
+      "grad_norm": 0.9844513532497011,
+      "learning_rate": 2.9864646353089822e-06,
+      "loss": 1.513,
+      "step": 5658
+    },
+    {
+      "epoch": 0.7551374432879637,
+      "grad_norm": 0.9691752292557538,
+      "learning_rate": 2.983384577891154e-06,
+      "loss": 1.5196,
+      "step": 5659
+    },
+    {
+      "epoch": 0.7552708833733653,
+      "grad_norm": 1.0348255829580888,
+      "learning_rate": 2.980305831097203e-06,
+      "loss": 1.5305,
+      "step": 5660
+    },
+    {
+      "epoch": 0.7554043234587671,
+      "grad_norm": 0.9795461493134596,
+      "learning_rate": 2.9772283955022006e-06,
+      "loss": 1.5619,
+      "step": 5661
+    },
+    {
+      "epoch": 0.7555377635441687,
+      "grad_norm": 0.9480708314986227,
+      "learning_rate": 2.9741522716809734e-06,
+      "loss": 1.5426,
+      "step": 5662
+    },
+    {
+      "epoch": 0.7556712036295703,
+      "grad_norm": 0.9365703829056337,
+      "learning_rate": 2.971077460208113e-06,
+      "loss": 1.5243,
+      "step": 5663
+    },
+    {
+      "epoch": 0.755804643714972,
+      "grad_norm": 0.9664797943870821,
+      "learning_rate": 2.9680039616579558e-06,
+      "loss": 1.5866,
+      "step": 5664
+    },
+    {
+      "epoch": 0.7559380838003736,
+      "grad_norm": 0.9326504176989024,
+      "learning_rate": 2.964931776604596e-06,
+      "loss": 1.5703,
+      "step": 5665
+    },
+    {
+      "epoch": 0.7560715238857753,
+      "grad_norm": 0.9662933536347439,
+      "learning_rate": 2.9618609056218797e-06,
+      "loss": 1.5854,
+      "step": 5666
+    },
+    {
+      "epoch": 0.756204963971177,
+      "grad_norm": 1.0951318602049294,
+      "learning_rate": 2.9587913492834174e-06,
+      "loss": 1.5744,
+      "step": 5667
+    },
+    {
+      "epoch": 0.7563384040565786,
+      "grad_norm": 0.9387753384543569,
+      "learning_rate": 2.9557231081625637e-06,
+      "loss": 1.591,
+      "step": 5668
+    },
+    {
+      "epoch": 0.7564718441419802,
+      "grad_norm": 0.9885761349123516,
+      "learning_rate": 2.9526561828324286e-06,
+      "loss": 1.5488,
+      "step": 5669
+    },
+    {
+      "epoch": 0.7566052842273819,
+      "grad_norm": 0.9272808115972163,
+      "learning_rate": 2.9495905738658846e-06,
+      "loss": 1.5201,
+      "step": 5670
+    },
+    {
+      "epoch": 0.7567387243127836,
+      "grad_norm": 0.9268527670988117,
+      "learning_rate": 2.946526281835549e-06,
+      "loss": 1.571,
+      "step": 5671
+    },
+    {
+      "epoch": 0.7568721643981852,
+      "grad_norm": 1.0147610911957214,
+      "learning_rate": 2.9434633073137976e-06,
+      "loss": 1.586,
+      "step": 5672
+    },
+    {
+      "epoch": 0.7570056044835869,
+      "grad_norm": 0.9383855121743576,
+      "learning_rate": 2.940401650872755e-06,
+      "loss": 1.5405,
+      "step": 5673
+    },
+    {
+      "epoch": 0.7571390445689885,
+      "grad_norm": 1.0683692922351982,
+      "learning_rate": 2.9373413130843108e-06,
+      "loss": 1.5319,
+      "step": 5674
+    },
+    {
+      "epoch": 0.7572724846543902,
+      "grad_norm": 0.9501481416038345,
+      "learning_rate": 2.934282294520099e-06,
+      "loss": 1.5894,
+      "step": 5675
+    },
+    {
+      "epoch": 0.7574059247397918,
+      "grad_norm": 0.9176546375459858,
+      "learning_rate": 2.931224595751504e-06,
+      "loss": 1.4954,
+      "step": 5676
+    },
+    {
+      "epoch": 0.7575393648251935,
+      "grad_norm": 0.947992780591222,
+      "learning_rate": 2.9281682173496764e-06,
+      "loss": 1.5466,
+      "step": 5677
+    },
+    {
+      "epoch": 0.7576728049105952,
+      "grad_norm": 0.9790622488674117,
+      "learning_rate": 2.925113159885511e-06,
+      "loss": 1.5469,
+      "step": 5678
+    },
+    {
+      "epoch": 0.7578062449959968,
+      "grad_norm": 0.9636768377878497,
+      "learning_rate": 2.922059423929656e-06,
+      "loss": 1.5419,
+      "step": 5679
+    },
+    {
+      "epoch": 0.7579396850813984,
+      "grad_norm": 0.9219842776449305,
+      "learning_rate": 2.9190070100525124e-06,
+      "loss": 1.4979,
+      "step": 5680
+    },
+    {
+      "epoch": 0.7580731251668001,
+      "grad_norm": 0.9863967160238304,
+      "learning_rate": 2.9159559188242428e-06,
+      "loss": 1.5115,
+      "step": 5681
+    },
+    {
+      "epoch": 0.7582065652522018,
+      "grad_norm": 0.9540589917238507,
+      "learning_rate": 2.9129061508147514e-06,
+      "loss": 1.5829,
+      "step": 5682
+    },
+    {
+      "epoch": 0.7583400053376034,
+      "grad_norm": 0.9415496333259665,
+      "learning_rate": 2.9098577065937027e-06,
+      "loss": 1.5155,
+      "step": 5683
+    },
+    {
+      "epoch": 0.7584734454230051,
+      "grad_norm": 1.0264214386562416,
+      "learning_rate": 2.906810586730506e-06,
+      "loss": 1.5762,
+      "step": 5684
+    },
+    {
+      "epoch": 0.7586068855084067,
+      "grad_norm": 1.0085987174197029,
+      "learning_rate": 2.903764791794337e-06,
+      "loss": 1.6053,
+      "step": 5685
+    },
+    {
+      "epoch": 0.7587403255938083,
+      "grad_norm": 0.9746022220580336,
+      "learning_rate": 2.9007203223541104e-06,
+      "loss": 1.5505,
+      "step": 5686
+    },
+    {
+      "epoch": 0.75887376567921,
+      "grad_norm": 1.1436325712704063,
+      "learning_rate": 2.8976771789784986e-06,
+      "loss": 1.5532,
+      "step": 5687
+    },
+    {
+      "epoch": 0.7590072057646117,
+      "grad_norm": 0.9458780762382005,
+      "learning_rate": 2.894635362235928e-06,
+      "loss": 1.5335,
+      "step": 5688
+    },
+    {
+      "epoch": 0.7591406458500134,
+      "grad_norm": 0.9586718389826383,
+      "learning_rate": 2.8915948726945696e-06,
+      "loss": 1.5686,
+      "step": 5689
+    },
+    {
+      "epoch": 0.759274085935415,
+      "grad_norm": 0.9374091977052081,
+      "learning_rate": 2.8885557109223615e-06,
+      "loss": 1.5533,
+      "step": 5690
+    },
+    {
+      "epoch": 0.7594075260208166,
+      "grad_norm": 0.9210874426208345,
+      "learning_rate": 2.8855178774869784e-06,
+      "loss": 1.6033,
+      "step": 5691
+    },
+    {
+      "epoch": 0.7595409661062184,
+      "grad_norm": 0.9425967721286689,
+      "learning_rate": 2.882481372955855e-06,
+      "loss": 1.5192,
+      "step": 5692
+    },
+    {
+      "epoch": 0.75967440619162,
+      "grad_norm": 0.9409853280626861,
+      "learning_rate": 2.879446197896176e-06,
+      "loss": 1.5309,
+      "step": 5693
+    },
+    {
+      "epoch": 0.7598078462770216,
+      "grad_norm": 0.9175409560809977,
+      "learning_rate": 2.8764123528748724e-06,
+      "loss": 1.5253,
+      "step": 5694
+    },
+    {
+      "epoch": 0.7599412863624233,
+      "grad_norm": 1.0633079652522899,
+      "learning_rate": 2.8733798384586398e-06,
+      "loss": 1.5564,
+      "step": 5695
+    },
+    {
+      "epoch": 0.7600747264478249,
+      "grad_norm": 1.0083618844062685,
+      "learning_rate": 2.870348655213914e-06,
+      "loss": 1.5223,
+      "step": 5696
+    },
+    {
+      "epoch": 0.7602081665332265,
+      "grad_norm": 1.1484670895236269,
+      "learning_rate": 2.8673188037068843e-06,
+      "loss": 1.5488,
+      "step": 5697
+    },
+    {
+      "epoch": 0.7603416066186283,
+      "grad_norm": 0.9648612510439487,
+      "learning_rate": 2.8642902845034914e-06,
+      "loss": 1.62,
+      "step": 5698
+    },
+    {
+      "epoch": 0.7604750467040299,
+      "grad_norm": 0.9409488162926377,
+      "learning_rate": 2.8612630981694322e-06,
+      "loss": 1.5146,
+      "step": 5699
+    },
+    {
+      "epoch": 0.7606084867894315,
+      "grad_norm": 1.0185521319833666,
+      "learning_rate": 2.85823724527015e-06,
+      "loss": 1.5524,
+      "step": 5700
+    },
+    {
+      "epoch": 0.7607419268748332,
+      "grad_norm": 0.9261998030680365,
+      "learning_rate": 2.855212726370834e-06,
+      "loss": 1.5362,
+      "step": 5701
+    },
+    {
+      "epoch": 0.7608753669602348,
+      "grad_norm": 0.9131576322837639,
+      "learning_rate": 2.852189542036439e-06,
+      "loss": 1.5431,
+      "step": 5702
+    },
+    {
+      "epoch": 0.7610088070456366,
+      "grad_norm": 0.9672674722537751,
+      "learning_rate": 2.849167692831655e-06,
+      "loss": 1.5413,
+      "step": 5703
+    },
+    {
+      "epoch": 0.7611422471310382,
+      "grad_norm": 0.9672341780718317,
+      "learning_rate": 2.8461471793209318e-06,
+      "loss": 1.5816,
+      "step": 5704
+    },
+    {
+      "epoch": 0.7612756872164398,
+      "grad_norm": 1.024093269678312,
+      "learning_rate": 2.8431280020684615e-06,
+      "loss": 1.5202,
+      "step": 5705
+    },
+    {
+      "epoch": 0.7614091273018415,
+      "grad_norm": 0.9730376429468712,
+      "learning_rate": 2.840110161638201e-06,
+      "loss": 1.5619,
+      "step": 5706
+    },
+    {
+      "epoch": 0.7615425673872431,
+      "grad_norm": 0.9277118027287636,
+      "learning_rate": 2.8370936585938425e-06,
+      "loss": 1.5741,
+      "step": 5707
+    },
+    {
+      "epoch": 0.7616760074726447,
+      "grad_norm": 0.9385522263679275,
+      "learning_rate": 2.834078493498833e-06,
+      "loss": 1.5243,
+      "step": 5708
+    },
+    {
+      "epoch": 0.7618094475580465,
+      "grad_norm": 0.921781169116257,
+      "learning_rate": 2.8310646669163777e-06,
+      "loss": 1.5539,
+      "step": 5709
+    },
+    {
+      "epoch": 0.7619428876434481,
+      "grad_norm": 0.9708668936556057,
+      "learning_rate": 2.8280521794094205e-06,
+      "loss": 1.5721,
+      "step": 5710
+    },
+    {
+      "epoch": 0.7620763277288497,
+      "grad_norm": 0.9753523420806991,
+      "learning_rate": 2.8250410315406606e-06,
+      "loss": 1.5386,
+      "step": 5711
+    },
+    {
+      "epoch": 0.7622097678142514,
+      "grad_norm": 1.1177484636132677,
+      "learning_rate": 2.822031223872543e-06,
+      "loss": 1.5526,
+      "step": 5712
+    },
+    {
+      "epoch": 0.762343207899653,
+      "grad_norm": 0.9770528135249951,
+      "learning_rate": 2.8190227569672722e-06,
+      "loss": 1.5853,
+      "step": 5713
+    },
+    {
+      "epoch": 0.7624766479850547,
+      "grad_norm": 1.3032469298982818,
+      "learning_rate": 2.8160156313867915e-06,
+      "loss": 1.6291,
+      "step": 5714
+    },
+    {
+      "epoch": 0.7626100880704564,
+      "grad_norm": 0.9638843608824444,
+      "learning_rate": 2.813009847692795e-06,
+      "loss": 1.5735,
+      "step": 5715
+    },
+    {
+      "epoch": 0.762743528155858,
+      "grad_norm": 1.1791110749699807,
+      "learning_rate": 2.8100054064467355e-06,
+      "loss": 1.5285,
+      "step": 5716
+    },
+    {
+      "epoch": 0.7628769682412597,
+      "grad_norm": 16.86319433750556,
+      "learning_rate": 2.807002308209804e-06,
+      "loss": 1.5693,
+      "step": 5717
+    },
+    {
+      "epoch": 0.7630104083266613,
+      "grad_norm": 0.9439903919518662,
+      "learning_rate": 2.8040005535429472e-06,
+      "loss": 1.5474,
+      "step": 5718
+    },
+    {
+      "epoch": 0.763143848412063,
+      "grad_norm": 0.961074528153183,
+      "learning_rate": 2.8010001430068535e-06,
+      "loss": 1.5005,
+      "step": 5719
+    },
+    {
+      "epoch": 0.7632772884974647,
+      "grad_norm": 0.9524749124010052,
+      "learning_rate": 2.7980010771619727e-06,
+      "loss": 1.5212,
+      "step": 5720
+    },
+    {
+      "epoch": 0.7634107285828663,
+      "grad_norm": 0.9533980355725618,
+      "learning_rate": 2.795003356568492e-06,
+      "loss": 1.5894,
+      "step": 5721
+    },
+    {
+      "epoch": 0.7635441686682679,
+      "grad_norm": 0.9213712133256243,
+      "learning_rate": 2.792006981786354e-06,
+      "loss": 1.5208,
+      "step": 5722
+    },
+    {
+      "epoch": 0.7636776087536696,
+      "grad_norm": 0.9294323574854917,
+      "learning_rate": 2.7890119533752415e-06,
+      "loss": 1.5441,
+      "step": 5723
+    },
+    {
+      "epoch": 0.7638110488390712,
+      "grad_norm": 0.9784510201172563,
+      "learning_rate": 2.7860182718945993e-06,
+      "loss": 1.581,
+      "step": 5724
+    },
+    {
+      "epoch": 0.7639444889244729,
+      "grad_norm": 0.9588683754336024,
+      "learning_rate": 2.7830259379036095e-06,
+      "loss": 1.5307,
+      "step": 5725
+    },
+    {
+      "epoch": 0.7640779290098746,
+      "grad_norm": 0.9143625087310671,
+      "learning_rate": 2.7800349519612023e-06,
+      "loss": 1.5518,
+      "step": 5726
+    },
+    {
+      "epoch": 0.7642113690952762,
+      "grad_norm": 0.9792840961883873,
+      "learning_rate": 2.777045314626068e-06,
+      "loss": 1.4893,
+      "step": 5727
+    },
+    {
+      "epoch": 0.7643448091806779,
+      "grad_norm": 0.9591101744910538,
+      "learning_rate": 2.7740570264566325e-06,
+      "loss": 1.576,
+      "step": 5728
+    },
+    {
+      "epoch": 0.7644782492660795,
+      "grad_norm": 1.0493610713948585,
+      "learning_rate": 2.771070088011073e-06,
+      "loss": 1.5944,
+      "step": 5729
+    },
+    {
+      "epoch": 0.7646116893514812,
+      "grad_norm": 0.940470963878361,
+      "learning_rate": 2.7680844998473176e-06,
+      "loss": 1.6017,
+      "step": 5730
+    },
+    {
+      "epoch": 0.7647451294368829,
+      "grad_norm": 0.9114372371817887,
+      "learning_rate": 2.7651002625230394e-06,
+      "loss": 1.5043,
+      "step": 5731
+    },
+    {
+      "epoch": 0.7648785695222845,
+      "grad_norm": 0.9514804602724689,
+      "learning_rate": 2.7621173765956553e-06,
+      "loss": 1.556,
+      "step": 5732
+    },
+    {
+      "epoch": 0.7650120096076861,
+      "grad_norm": 0.931967453227579,
+      "learning_rate": 2.7591358426223437e-06,
+      "loss": 1.5529,
+      "step": 5733
+    },
+    {
+      "epoch": 0.7651454496930878,
+      "grad_norm": 0.9641034664380269,
+      "learning_rate": 2.756155661160015e-06,
+      "loss": 1.5584,
+      "step": 5734
+    },
+    {
+      "epoch": 0.7652788897784895,
+      "grad_norm": 1.0406330527700243,
+      "learning_rate": 2.753176832765334e-06,
+      "loss": 1.5685,
+      "step": 5735
+    },
+    {
+      "epoch": 0.7654123298638911,
+      "grad_norm": 1.1171214620153784,
+      "learning_rate": 2.750199357994714e-06,
+      "loss": 1.5532,
+      "step": 5736
+    },
+    {
+      "epoch": 0.7655457699492928,
+      "grad_norm": 0.9866430386535575,
+      "learning_rate": 2.7472232374043084e-06,
+      "loss": 1.5261,
+      "step": 5737
+    },
+    {
+      "epoch": 0.7656792100346944,
+      "grad_norm": 0.9361087966486119,
+      "learning_rate": 2.7442484715500286e-06,
+      "loss": 1.5359,
+      "step": 5738
+    },
+    {
+      "epoch": 0.765812650120096,
+      "grad_norm": 0.9609366943677766,
+      "learning_rate": 2.741275060987525e-06,
+      "loss": 1.544,
+      "step": 5739
+    },
+    {
+      "epoch": 0.7659460902054978,
+      "grad_norm": 1.0456283664316062,
+      "learning_rate": 2.7383030062721926e-06,
+      "loss": 1.5098,
+      "step": 5740
+    },
+    {
+      "epoch": 0.7660795302908994,
+      "grad_norm": 0.9071979960109646,
+      "learning_rate": 2.7353323079591842e-06,
+      "loss": 1.5621,
+      "step": 5741
+    },
+    {
+      "epoch": 0.7662129703763011,
+      "grad_norm": 0.9422303587878669,
+      "learning_rate": 2.7323629666033903e-06,
+      "loss": 1.5661,
+      "step": 5742
+    },
+    {
+      "epoch": 0.7663464104617027,
+      "grad_norm": 1.0122385866589452,
+      "learning_rate": 2.729394982759449e-06,
+      "loss": 1.5319,
+      "step": 5743
+    },
+    {
+      "epoch": 0.7664798505471043,
+      "grad_norm": 1.1141139885458986,
+      "learning_rate": 2.726428356981742e-06,
+      "loss": 1.5295,
+      "step": 5744
+    },
+    {
+      "epoch": 0.766613290632506,
+      "grad_norm": 0.982862912585683,
+      "learning_rate": 2.7234630898244084e-06,
+      "loss": 1.5591,
+      "step": 5745
+    },
+    {
+      "epoch": 0.7667467307179077,
+      "grad_norm": 1.075501803864991,
+      "learning_rate": 2.720499181841324e-06,
+      "loss": 1.5615,
+      "step": 5746
+    },
+    {
+      "epoch": 0.7668801708033093,
+      "grad_norm": 1.028976839916406,
+      "learning_rate": 2.7175366335861087e-06,
+      "loss": 1.5477,
+      "step": 5747
+    },
+    {
+      "epoch": 0.767013610888711,
+      "grad_norm": 1.0645308095468675,
+      "learning_rate": 2.714575445612139e-06,
+      "loss": 1.5542,
+      "step": 5748
+    },
+    {
+      "epoch": 0.7671470509741126,
+      "grad_norm": 0.9392883647391629,
+      "learning_rate": 2.7116156184725285e-06,
+      "loss": 1.5699,
+      "step": 5749
+    },
+    {
+      "epoch": 0.7672804910595142,
+      "grad_norm": 0.965446906684568,
+      "learning_rate": 2.708657152720139e-06,
+      "loss": 1.5796,
+      "step": 5750
+    },
+    {
+      "epoch": 0.767413931144916,
+      "grad_norm": 1.0070562909813836,
+      "learning_rate": 2.705700048907576e-06,
+      "loss": 1.5419,
+      "step": 5751
+    },
+    {
+      "epoch": 0.7675473712303176,
+      "grad_norm": 0.9415702130765551,
+      "learning_rate": 2.7027443075871974e-06,
+      "loss": 1.5344,
+      "step": 5752
+    },
+    {
+      "epoch": 0.7676808113157192,
+      "grad_norm": 1.007240476174913,
+      "learning_rate": 2.6997899293110997e-06,
+      "loss": 1.5223,
+      "step": 5753
+    },
+    {
+      "epoch": 0.7678142514011209,
+      "grad_norm": 0.9583583605966772,
+      "learning_rate": 2.696836914631127e-06,
+      "loss": 1.4856,
+      "step": 5754
+    },
+    {
+      "epoch": 0.7679476914865225,
+      "grad_norm": 0.9286247656808373,
+      "learning_rate": 2.6938852640988666e-06,
+      "loss": 1.5845,
+      "step": 5755
+    },
+    {
+      "epoch": 0.7680811315719243,
+      "grad_norm": 1.0066591400064955,
+      "learning_rate": 2.690934978265659e-06,
+      "loss": 1.5767,
+      "step": 5756
+    },
+    {
+      "epoch": 0.7682145716573259,
+      "grad_norm": 1.052076839954032,
+      "learning_rate": 2.68798605768258e-06,
+      "loss": 1.5927,
+      "step": 5757
+    },
+    {
+      "epoch": 0.7683480117427275,
+      "grad_norm": 1.0098347870196225,
+      "learning_rate": 2.685038502900452e-06,
+      "loss": 1.5831,
+      "step": 5758
+    },
+    {
+      "epoch": 0.7684814518281292,
+      "grad_norm": 1.0546240360318806,
+      "learning_rate": 2.682092314469851e-06,
+      "loss": 1.5924,
+      "step": 5759
+    },
+    {
+      "epoch": 0.7686148919135308,
+      "grad_norm": 0.985560507673923,
+      "learning_rate": 2.6791474929410877e-06,
+      "loss": 1.5819,
+      "step": 5760
+    },
+    {
+      "epoch": 0.7687483319989324,
+      "grad_norm": 0.9490509821938993,
+      "learning_rate": 2.6762040388642217e-06,
+      "loss": 1.5278,
+      "step": 5761
+    },
+    {
+      "epoch": 0.7688817720843342,
+      "grad_norm": 0.9668194942766439,
+      "learning_rate": 2.6732619527890547e-06,
+      "loss": 1.5719,
+      "step": 5762
+    },
+    {
+      "epoch": 0.7690152121697358,
+      "grad_norm": 1.0376817559211673,
+      "learning_rate": 2.670321235265139e-06,
+      "loss": 1.5694,
+      "step": 5763
+    },
+    {
+      "epoch": 0.7691486522551374,
+      "grad_norm": 1.1459513474130882,
+      "learning_rate": 2.6673818868417646e-06,
+      "loss": 1.5727,
+      "step": 5764
+    },
+    {
+      "epoch": 0.7692820923405391,
+      "grad_norm": 0.9389863744334982,
+      "learning_rate": 2.6644439080679662e-06,
+      "loss": 1.5679,
+      "step": 5765
+    },
+    {
+      "epoch": 0.7694155324259407,
+      "grad_norm": 0.9437950047313849,
+      "learning_rate": 2.6615072994925308e-06,
+      "loss": 1.5559,
+      "step": 5766
+    },
+    {
+      "epoch": 0.7695489725113424,
+      "grad_norm": 0.955667607095649,
+      "learning_rate": 2.65857206166398e-06,
+      "loss": 1.526,
+      "step": 5767
+    },
+    {
+      "epoch": 0.7696824125967441,
+      "grad_norm": 0.8988996567956318,
+      "learning_rate": 2.655638195130582e-06,
+      "loss": 1.4886,
+      "step": 5768
+    },
+    {
+      "epoch": 0.7698158526821457,
+      "grad_norm": 1.0552144404288448,
+      "learning_rate": 2.652705700440348e-06,
+      "loss": 1.5902,
+      "step": 5769
+    },
+    {
+      "epoch": 0.7699492927675474,
+      "grad_norm": 0.9570299871105503,
+      "learning_rate": 2.64977457814104e-06,
+      "loss": 1.5706,
+      "step": 5770
+    },
+    {
+      "epoch": 0.770082732852949,
+      "grad_norm": 0.9664283679747206,
+      "learning_rate": 2.6468448287801552e-06,
+      "loss": 1.5595,
+      "step": 5771
+    },
+    {
+      "epoch": 0.7702161729383507,
+      "grad_norm": 0.929178580968516,
+      "learning_rate": 2.6439164529049376e-06,
+      "loss": 1.5301,
+      "step": 5772
+    },
+    {
+      "epoch": 0.7703496130237524,
+      "grad_norm": 0.9305117830654621,
+      "learning_rate": 2.6409894510623744e-06,
+      "loss": 1.5678,
+      "step": 5773
+    },
+    {
+      "epoch": 0.770483053109154,
+      "grad_norm": 1.0322434008439814,
+      "learning_rate": 2.6380638237991963e-06,
+      "loss": 1.5298,
+      "step": 5774
+    },
+    {
+      "epoch": 0.7706164931945556,
+      "grad_norm": 0.9590848759112884,
+      "learning_rate": 2.6351395716618746e-06,
+      "loss": 1.5437,
+      "step": 5775
+    },
+    {
+      "epoch": 0.7707499332799573,
+      "grad_norm": 0.9231804585207396,
+      "learning_rate": 2.632216695196631e-06,
+      "loss": 1.4736,
+      "step": 5776
+    },
+    {
+      "epoch": 0.770883373365359,
+      "grad_norm": 0.9786231108386787,
+      "learning_rate": 2.6292951949494237e-06,
+      "loss": 1.5593,
+      "step": 5777
+    },
+    {
+      "epoch": 0.7710168134507606,
+      "grad_norm": 0.9603881195608965,
+      "learning_rate": 2.626375071465955e-06,
+      "loss": 1.5821,
+      "step": 5778
+    },
+    {
+      "epoch": 0.7711502535361623,
+      "grad_norm": 0.9430362885685091,
+      "learning_rate": 2.623456325291669e-06,
+      "loss": 1.5404,
+      "step": 5779
+    },
+    {
+      "epoch": 0.7712836936215639,
+      "grad_norm": 0.9446912811828378,
+      "learning_rate": 2.6205389569717586e-06,
+      "loss": 1.5394,
+      "step": 5780
+    },
+    {
+      "epoch": 0.7714171337069655,
+      "grad_norm": 0.9412298808949228,
+      "learning_rate": 2.6176229670511533e-06,
+      "loss": 1.5538,
+      "step": 5781
+    },
+    {
+      "epoch": 0.7715505737923672,
+      "grad_norm": 1.3095768335252589,
+      "learning_rate": 2.6147083560745257e-06,
+      "loss": 1.5564,
+      "step": 5782
+    },
+    {
+      "epoch": 0.7716840138777689,
+      "grad_norm": 0.9918252751196945,
+      "learning_rate": 2.6117951245862893e-06,
+      "loss": 1.5451,
+      "step": 5783
+    },
+    {
+      "epoch": 0.7718174539631706,
+      "grad_norm": 0.996958048566631,
+      "learning_rate": 2.608883273130609e-06,
+      "loss": 1.5229,
+      "step": 5784
+    },
+    {
+      "epoch": 0.7719508940485722,
+      "grad_norm": 0.957110530564296,
+      "learning_rate": 2.6059728022513832e-06,
+      "loss": 1.55,
+      "step": 5785
+    },
+    {
+      "epoch": 0.7720843341339738,
+      "grad_norm": 1.0399836848407493,
+      "learning_rate": 2.603063712492252e-06,
+      "loss": 1.579,
+      "step": 5786
+    },
+    {
+      "epoch": 0.7722177742193755,
+      "grad_norm": 0.9408064311310185,
+      "learning_rate": 2.6001560043966e-06,
+      "loss": 1.5572,
+      "step": 5787
+    },
+    {
+      "epoch": 0.7723512143047772,
+      "grad_norm": 0.9412201777682861,
+      "learning_rate": 2.5972496785075594e-06,
+      "loss": 1.544,
+      "step": 5788
+    },
+    {
+      "epoch": 0.7724846543901788,
+      "grad_norm": 1.0115310422207648,
+      "learning_rate": 2.594344735367995e-06,
+      "loss": 1.5008,
+      "step": 5789
+    },
+    {
+      "epoch": 0.7726180944755805,
+      "grad_norm": 0.9227438226603408,
+      "learning_rate": 2.591441175520514e-06,
+      "loss": 1.5712,
+      "step": 5790
+    },
+    {
+      "epoch": 0.7727515345609821,
+      "grad_norm": 0.9594213056108911,
+      "learning_rate": 2.5885389995074752e-06,
+      "loss": 1.5903,
+      "step": 5791
+    },
+    {
+      "epoch": 0.7728849746463837,
+      "grad_norm": 1.086165595831062,
+      "learning_rate": 2.5856382078709685e-06,
+      "loss": 1.5691,
+      "step": 5792
+    },
+    {
+      "epoch": 0.7730184147317855,
+      "grad_norm": 0.9505020761449048,
+      "learning_rate": 2.5827388011528296e-06,
+      "loss": 1.5793,
+      "step": 5793
+    },
+    {
+      "epoch": 0.7731518548171871,
+      "grad_norm": 3.5110780857670107,
+      "learning_rate": 2.57984077989463e-06,
+      "loss": 1.5527,
+      "step": 5794
+    },
+    {
+      "epoch": 0.7732852949025887,
+      "grad_norm": 0.9455488863259419,
+      "learning_rate": 2.576944144637694e-06,
+      "loss": 1.5232,
+      "step": 5795
+    },
+    {
+      "epoch": 0.7734187349879904,
+      "grad_norm": 0.9672606415809805,
+      "learning_rate": 2.574048895923078e-06,
+      "loss": 1.5559,
+      "step": 5796
+    },
+    {
+      "epoch": 0.773552175073392,
+      "grad_norm": 0.9385093821290592,
+      "learning_rate": 2.571155034291577e-06,
+      "loss": 1.5565,
+      "step": 5797
+    },
+    {
+      "epoch": 0.7736856151587937,
+      "grad_norm": 0.9520678330195554,
+      "learning_rate": 2.5682625602837384e-06,
+      "loss": 1.5558,
+      "step": 5798
+    },
+    {
+      "epoch": 0.7738190552441954,
+      "grad_norm": 0.99051318050388,
+      "learning_rate": 2.5653714744398393e-06,
+      "loss": 1.5109,
+      "step": 5799
+    },
+    {
+      "epoch": 0.773952495329597,
+      "grad_norm": 0.984714941570732,
+      "learning_rate": 2.5624817772999033e-06,
+      "loss": 1.5115,
+      "step": 5800
+    },
+    {
+      "epoch": 0.7740859354149987,
+      "grad_norm": 0.9194604559966288,
+      "learning_rate": 2.5595934694036882e-06,
+      "loss": 1.4929,
+      "step": 5801
+    },
+    {
+      "epoch": 0.7742193755004003,
+      "grad_norm": 0.9247157704040789,
+      "learning_rate": 2.556706551290704e-06,
+      "loss": 1.5369,
+      "step": 5802
+    },
+    {
+      "epoch": 0.7743528155858019,
+      "grad_norm": 0.9360390213094082,
+      "learning_rate": 2.5538210235001913e-06,
+      "loss": 1.5576,
+      "step": 5803
+    },
+    {
+      "epoch": 0.7744862556712037,
+      "grad_norm": 0.9064471735808803,
+      "learning_rate": 2.5509368865711304e-06,
+      "loss": 1.5415,
+      "step": 5804
+    },
+    {
+      "epoch": 0.7746196957566053,
+      "grad_norm": 0.9513137240894838,
+      "learning_rate": 2.548054141042251e-06,
+      "loss": 1.5708,
+      "step": 5805
+    },
+    {
+      "epoch": 0.7747531358420069,
+      "grad_norm": 1.2175807518906996,
+      "learning_rate": 2.5451727874520148e-06,
+      "loss": 1.514,
+      "step": 5806
+    },
+    {
+      "epoch": 0.7748865759274086,
+      "grad_norm": 0.9892844938681642,
+      "learning_rate": 2.542292826338626e-06,
+      "loss": 1.5524,
+      "step": 5807
+    },
+    {
+      "epoch": 0.7750200160128102,
+      "grad_norm": 1.0937981716259788,
+      "learning_rate": 2.5394142582400238e-06,
+      "loss": 1.4936,
+      "step": 5808
+    },
+    {
+      "epoch": 0.775153456098212,
+      "grad_norm": 1.169626597958446,
+      "learning_rate": 2.5365370836938997e-06,
+      "loss": 1.5551,
+      "step": 5809
+    },
+    {
+      "epoch": 0.7752868961836136,
+      "grad_norm": 0.9516314581228,
+      "learning_rate": 2.5336613032376744e-06,
+      "loss": 1.5305,
+      "step": 5810
+    },
+    {
+      "epoch": 0.7754203362690152,
+      "grad_norm": 0.9404267765234818,
+      "learning_rate": 2.5307869174085086e-06,
+      "loss": 1.5901,
+      "step": 5811
+    },
+    {
+      "epoch": 0.7755537763544169,
+      "grad_norm": 0.9359769036295512,
+      "learning_rate": 2.527913926743305e-06,
+      "loss": 1.5633,
+      "step": 5812
+    },
+    {
+      "epoch": 0.7756872164398185,
+      "grad_norm": 0.9494830051173815,
+      "learning_rate": 2.5250423317787086e-06,
+      "loss": 1.5296,
+      "step": 5813
+    },
+    {
+      "epoch": 0.7758206565252201,
+      "grad_norm": 0.9336095498356144,
+      "learning_rate": 2.522172133051101e-06,
+      "loss": 1.5527,
+      "step": 5814
+    },
+    {
+      "epoch": 0.7759540966106219,
+      "grad_norm": 0.9206894715103576,
+      "learning_rate": 2.5193033310965987e-06,
+      "loss": 1.5765,
+      "step": 5815
+    },
+    {
+      "epoch": 0.7760875366960235,
+      "grad_norm": 0.9285054392212316,
+      "learning_rate": 2.516435926451064e-06,
+      "loss": 1.5186,
+      "step": 5816
+    },
+    {
+      "epoch": 0.7762209767814251,
+      "grad_norm": 0.9469454273921929,
+      "learning_rate": 2.513569919650095e-06,
+      "loss": 1.5748,
+      "step": 5817
+    },
+    {
+      "epoch": 0.7763544168668268,
+      "grad_norm": 1.0422340407226904,
+      "learning_rate": 2.510705311229025e-06,
+      "loss": 1.5367,
+      "step": 5818
+    },
+    {
+      "epoch": 0.7764878569522284,
+      "grad_norm": 0.9341799059535522,
+      "learning_rate": 2.507842101722938e-06,
+      "loss": 1.5679,
+      "step": 5819
+    },
+    {
+      "epoch": 0.77662129703763,
+      "grad_norm": 0.9183397021685292,
+      "learning_rate": 2.5049802916666445e-06,
+      "loss": 1.5037,
+      "step": 5820
+    },
+    {
+      "epoch": 0.7767547371230318,
+      "grad_norm": 0.932957030523527,
+      "learning_rate": 2.5021198815946978e-06,
+      "loss": 1.5195,
+      "step": 5821
+    },
+    {
+      "epoch": 0.7768881772084334,
+      "grad_norm": 0.9241883977653392,
+      "learning_rate": 2.4992608720413893e-06,
+      "loss": 1.5227,
+      "step": 5822
+    },
+    {
+      "epoch": 0.7770216172938351,
+      "grad_norm": 1.2172435346645842,
+      "learning_rate": 2.496403263540752e-06,
+      "loss": 1.5402,
+      "step": 5823
+    },
+    {
+      "epoch": 0.7771550573792367,
+      "grad_norm": 0.9449596951384638,
+      "learning_rate": 2.493547056626554e-06,
+      "loss": 1.5355,
+      "step": 5824
+    },
+    {
+      "epoch": 0.7772884974646384,
+      "grad_norm": 1.0524048742416596,
+      "learning_rate": 2.490692251832302e-06,
+      "loss": 1.5999,
+      "step": 5825
+    },
+    {
+      "epoch": 0.7774219375500401,
+      "grad_norm": 0.984791837305669,
+      "learning_rate": 2.4878388496912377e-06,
+      "loss": 1.5482,
+      "step": 5826
+    },
+    {
+      "epoch": 0.7775553776354417,
+      "grad_norm": 0.9575185903107368,
+      "learning_rate": 2.4849868507363506e-06,
+      "loss": 1.5152,
+      "step": 5827
+    },
+    {
+      "epoch": 0.7776888177208433,
+      "grad_norm": 0.9431849505627886,
+      "learning_rate": 2.482136255500357e-06,
+      "loss": 1.5557,
+      "step": 5828
+    },
+    {
+      "epoch": 0.777822257806245,
+      "grad_norm": 0.9027138718964596,
+      "learning_rate": 2.479287064515714e-06,
+      "loss": 1.5192,
+      "step": 5829
+    },
+    {
+      "epoch": 0.7779556978916466,
+      "grad_norm": 0.9769963398115894,
+      "learning_rate": 2.476439278314624e-06,
+      "loss": 1.5003,
+      "step": 5830
+    },
+    {
+      "epoch": 0.7780891379770483,
+      "grad_norm": 1.0871003730614877,
+      "learning_rate": 2.473592897429018e-06,
+      "loss": 1.5801,
+      "step": 5831
+    },
+    {
+      "epoch": 0.77822257806245,
+      "grad_norm": 0.9763548093718147,
+      "learning_rate": 2.470747922390567e-06,
+      "loss": 1.5494,
+      "step": 5832
+    },
+    {
+      "epoch": 0.7783560181478516,
+      "grad_norm": 0.984785198847868,
+      "learning_rate": 2.467904353730678e-06,
+      "loss": 1.5478,
+      "step": 5833
+    },
+    {
+      "epoch": 0.7784894582332532,
+      "grad_norm": 0.9601413586909818,
+      "learning_rate": 2.465062191980503e-06,
+      "loss": 1.5631,
+      "step": 5834
+    },
+    {
+      "epoch": 0.7786228983186549,
+      "grad_norm": 1.0051540823117064,
+      "learning_rate": 2.462221437670921e-06,
+      "loss": 1.5519,
+      "step": 5835
+    },
+    {
+      "epoch": 0.7787563384040566,
+      "grad_norm": 1.0448166746922907,
+      "learning_rate": 2.4593820913325507e-06,
+      "loss": 1.5673,
+      "step": 5836
+    },
+    {
+      "epoch": 0.7788897784894583,
+      "grad_norm": 1.0626411844360724,
+      "learning_rate": 2.4565441534957558e-06,
+      "loss": 1.5583,
+      "step": 5837
+    },
+    {
+      "epoch": 0.7790232185748599,
+      "grad_norm": 1.0838658214565313,
+      "learning_rate": 2.453707624690628e-06,
+      "loss": 1.5399,
+      "step": 5838
+    },
+    {
+      "epoch": 0.7791566586602615,
+      "grad_norm": 1.0694003456730499,
+      "learning_rate": 2.4508725054469973e-06,
+      "loss": 1.5874,
+      "step": 5839
+    },
+    {
+      "epoch": 0.7792900987456632,
+      "grad_norm": 1.0215015403096692,
+      "learning_rate": 2.4480387962944297e-06,
+      "loss": 1.545,
+      "step": 5840
+    },
+    {
+      "epoch": 0.7794235388310649,
+      "grad_norm": 0.947422869064629,
+      "learning_rate": 2.4452064977622357e-06,
+      "loss": 1.541,
+      "step": 5841
+    },
+    {
+      "epoch": 0.7795569789164665,
+      "grad_norm": 0.9555266023407392,
+      "learning_rate": 2.4423756103794538e-06,
+      "loss": 1.5314,
+      "step": 5842
+    },
+    {
+      "epoch": 0.7796904190018682,
+      "grad_norm": 1.012624888514248,
+      "learning_rate": 2.439546134674858e-06,
+      "loss": 1.5672,
+      "step": 5843
+    },
+    {
+      "epoch": 0.7798238590872698,
+      "grad_norm": 1.2038001979285369,
+      "learning_rate": 2.436718071176969e-06,
+      "loss": 1.5503,
+      "step": 5844
+    },
+    {
+      "epoch": 0.7799572991726714,
+      "grad_norm": 0.9750829570129861,
+      "learning_rate": 2.4338914204140317e-06,
+      "loss": 1.5958,
+      "step": 5845
+    },
+    {
+      "epoch": 0.7800907392580732,
+      "grad_norm": 1.0711410399768173,
+      "learning_rate": 2.4310661829140335e-06,
+      "loss": 1.5094,
+      "step": 5846
+    },
+    {
+      "epoch": 0.7802241793434748,
+      "grad_norm": 0.960236661793123,
+      "learning_rate": 2.4282423592046956e-06,
+      "loss": 1.5084,
+      "step": 5847
+    },
+    {
+      "epoch": 0.7803576194288764,
+      "grad_norm": 0.9702210353839494,
+      "learning_rate": 2.425419949813479e-06,
+      "loss": 1.5571,
+      "step": 5848
+    },
+    {
+      "epoch": 0.7804910595142781,
+      "grad_norm": 0.9253780960251425,
+      "learning_rate": 2.4225989552675756e-06,
+      "loss": 1.579,
+      "step": 5849
+    },
+    {
+      "epoch": 0.7806244995996797,
+      "grad_norm": 0.9233717044203158,
+      "learning_rate": 2.419779376093916e-06,
+      "loss": 1.5668,
+      "step": 5850
+    },
+    {
+      "epoch": 0.7807579396850814,
+      "grad_norm": 0.9348350250160402,
+      "learning_rate": 2.416961212819162e-06,
+      "loss": 1.536,
+      "step": 5851
+    },
+    {
+      "epoch": 0.7808913797704831,
+      "grad_norm": 0.9548401099327721,
+      "learning_rate": 2.41414446596972e-06,
+      "loss": 1.5628,
+      "step": 5852
+    },
+    {
+      "epoch": 0.7810248198558847,
+      "grad_norm": 0.956552292282773,
+      "learning_rate": 2.411329136071724e-06,
+      "loss": 1.6029,
+      "step": 5853
+    },
+    {
+      "epoch": 0.7811582599412864,
+      "grad_norm": 0.9985446373657281,
+      "learning_rate": 2.4085152236510445e-06,
+      "loss": 1.5279,
+      "step": 5854
+    },
+    {
+      "epoch": 0.781291700026688,
+      "grad_norm": 0.9286530398266631,
+      "learning_rate": 2.4057027292332857e-06,
+      "loss": 1.5534,
+      "step": 5855
+    },
+    {
+      "epoch": 0.7814251401120896,
+      "grad_norm": 0.9838982529058777,
+      "learning_rate": 2.4028916533437963e-06,
+      "loss": 1.5773,
+      "step": 5856
+    },
+    {
+      "epoch": 0.7815585801974914,
+      "grad_norm": 0.9685070922452822,
+      "learning_rate": 2.4000819965076506e-06,
+      "loss": 1.5407,
+      "step": 5857
+    },
+    {
+      "epoch": 0.781692020282893,
+      "grad_norm": 0.9289471086756744,
+      "learning_rate": 2.3972737592496576e-06,
+      "loss": 1.5773,
+      "step": 5858
+    },
+    {
+      "epoch": 0.7818254603682946,
+      "grad_norm": 0.9607754330554751,
+      "learning_rate": 2.3944669420943678e-06,
+      "loss": 1.4774,
+      "step": 5859
+    },
+    {
+      "epoch": 0.7819589004536963,
+      "grad_norm": 1.0361591280939995,
+      "learning_rate": 2.3916615455660574e-06,
+      "loss": 1.5427,
+      "step": 5860
+    },
+    {
+      "epoch": 0.7820923405390979,
+      "grad_norm": 0.9200656629737055,
+      "learning_rate": 2.3888575701887483e-06,
+      "loss": 1.5319,
+      "step": 5861
+    },
+    {
+      "epoch": 0.7822257806244995,
+      "grad_norm": 1.2418620489944565,
+      "learning_rate": 2.3860550164861908e-06,
+      "loss": 1.5005,
+      "step": 5862
+    },
+    {
+      "epoch": 0.7823592207099013,
+      "grad_norm": 0.9438642577089267,
+      "learning_rate": 2.3832538849818663e-06,
+      "loss": 1.5266,
+      "step": 5863
+    },
+    {
+      "epoch": 0.7824926607953029,
+      "grad_norm": 0.9488414068156699,
+      "learning_rate": 2.3804541761989974e-06,
+      "loss": 1.568,
+      "step": 5864
+    },
+    {
+      "epoch": 0.7826261008807046,
+      "grad_norm": 0.9217470739999812,
+      "learning_rate": 2.377655890660533e-06,
+      "loss": 1.4917,
+      "step": 5865
+    },
+    {
+      "epoch": 0.7827595409661062,
+      "grad_norm": 0.9403617005912336,
+      "learning_rate": 2.374859028889168e-06,
+      "loss": 1.5904,
+      "step": 5866
+    },
+    {
+      "epoch": 0.7828929810515078,
+      "grad_norm": 0.9377360220679098,
+      "learning_rate": 2.372063591407321e-06,
+      "loss": 1.529,
+      "step": 5867
+    },
+    {
+      "epoch": 0.7830264211369096,
+      "grad_norm": 0.9289834664828129,
+      "learning_rate": 2.3692695787371443e-06,
+      "loss": 1.5424,
+      "step": 5868
+    },
+    {
+      "epoch": 0.7831598612223112,
+      "grad_norm": 0.9284928064699219,
+      "learning_rate": 2.366476991400535e-06,
+      "loss": 1.5683,
+      "step": 5869
+    },
+    {
+      "epoch": 0.7832933013077128,
+      "grad_norm": 0.9414971322161758,
+      "learning_rate": 2.3636858299191113e-06,
+      "loss": 1.5869,
+      "step": 5870
+    },
+    {
+      "epoch": 0.7834267413931145,
+      "grad_norm": 1.049554007951963,
+      "learning_rate": 2.3608960948142334e-06,
+      "loss": 1.581,
+      "step": 5871
+    },
+    {
+      "epoch": 0.7835601814785161,
+      "grad_norm": 0.9261270821311429,
+      "learning_rate": 2.3581077866069868e-06,
+      "loss": 1.5777,
+      "step": 5872
+    },
+    {
+      "epoch": 0.7836936215639178,
+      "grad_norm": 0.9228475363078417,
+      "learning_rate": 2.3553209058182025e-06,
+      "loss": 1.5916,
+      "step": 5873
+    },
+    {
+      "epoch": 0.7838270616493195,
+      "grad_norm": 0.9372723449214597,
+      "learning_rate": 2.3525354529684354e-06,
+      "loss": 1.4841,
+      "step": 5874
+    },
+    {
+      "epoch": 0.7839605017347211,
+      "grad_norm": 1.0271631020714844,
+      "learning_rate": 2.349751428577972e-06,
+      "loss": 1.5588,
+      "step": 5875
+    },
+    {
+      "epoch": 0.7840939418201228,
+      "grad_norm": 0.9083784084217819,
+      "learning_rate": 2.346968833166844e-06,
+      "loss": 1.5394,
+      "step": 5876
+    },
+    {
+      "epoch": 0.7842273819055244,
+      "grad_norm": 0.950058509885086,
+      "learning_rate": 2.3441876672548046e-06,
+      "loss": 1.5469,
+      "step": 5877
+    },
+    {
+      "epoch": 0.784360821990926,
+      "grad_norm": 0.9649329509240152,
+      "learning_rate": 2.341407931361345e-06,
+      "loss": 1.6095,
+      "step": 5878
+    },
+    {
+      "epoch": 0.7844942620763278,
+      "grad_norm": 0.9865778802161118,
+      "learning_rate": 2.338629626005684e-06,
+      "loss": 1.5117,
+      "step": 5879
+    },
+    {
+      "epoch": 0.7846277021617294,
+      "grad_norm": 0.9479757386340179,
+      "learning_rate": 2.335852751706784e-06,
+      "loss": 1.5221,
+      "step": 5880
+    },
+    {
+      "epoch": 0.784761142247131,
+      "grad_norm": 1.0079567810813972,
+      "learning_rate": 2.3330773089833303e-06,
+      "loss": 1.5785,
+      "step": 5881
+    },
+    {
+      "epoch": 0.7848945823325327,
+      "grad_norm": 0.9604536184535951,
+      "learning_rate": 2.3303032983537445e-06,
+      "loss": 1.5498,
+      "step": 5882
+    },
+    {
+      "epoch": 0.7850280224179343,
+      "grad_norm": 0.9554406788386689,
+      "learning_rate": 2.327530720336176e-06,
+      "loss": 1.4614,
+      "step": 5883
+    },
+    {
+      "epoch": 0.785161462503336,
+      "grad_norm": 0.9339548764005035,
+      "learning_rate": 2.3247595754485185e-06,
+      "loss": 1.5512,
+      "step": 5884
+    },
+    {
+      "epoch": 0.7852949025887377,
+      "grad_norm": 0.9499142303394195,
+      "learning_rate": 2.321989864208386e-06,
+      "loss": 1.5241,
+      "step": 5885
+    },
+    {
+      "epoch": 0.7854283426741393,
+      "grad_norm": 0.976577687622226,
+      "learning_rate": 2.319221587133127e-06,
+      "loss": 1.5734,
+      "step": 5886
+    },
+    {
+      "epoch": 0.7855617827595409,
+      "grad_norm": 1.0810206584465782,
+      "learning_rate": 2.316454744739829e-06,
+      "loss": 1.5224,
+      "step": 5887
+    },
+    {
+      "epoch": 0.7856952228449426,
+      "grad_norm": 0.9179787586115475,
+      "learning_rate": 2.313689337545304e-06,
+      "loss": 1.472,
+      "step": 5888
+    },
+    {
+      "epoch": 0.7858286629303443,
+      "grad_norm": 0.9423241752395348,
+      "learning_rate": 2.310925366066099e-06,
+      "loss": 1.5131,
+      "step": 5889
+    },
+    {
+      "epoch": 0.785962103015746,
+      "grad_norm": 1.0074056359823236,
+      "learning_rate": 2.308162830818489e-06,
+      "loss": 1.5842,
+      "step": 5890
+    },
+    {
+      "epoch": 0.7860955431011476,
+      "grad_norm": 1.140943510869096,
+      "learning_rate": 2.3054017323184907e-06,
+      "loss": 1.5383,
+      "step": 5891
+    },
+    {
+      "epoch": 0.7862289831865492,
+      "grad_norm": 0.9686268722021224,
+      "learning_rate": 2.3026420710818433e-06,
+      "loss": 1.569,
+      "step": 5892
+    },
+    {
+      "epoch": 0.7863624232719509,
+      "grad_norm": 1.0903976628772232,
+      "learning_rate": 2.299883847624015e-06,
+      "loss": 1.5378,
+      "step": 5893
+    },
+    {
+      "epoch": 0.7864958633573526,
+      "grad_norm": 1.0249636474092916,
+      "learning_rate": 2.297127062460218e-06,
+      "loss": 1.4946,
+      "step": 5894
+    },
+    {
+      "epoch": 0.7866293034427542,
+      "grad_norm": 0.9160337674447874,
+      "learning_rate": 2.2943717161053856e-06,
+      "loss": 1.563,
+      "step": 5895
+    },
+    {
+      "epoch": 0.7867627435281559,
+      "grad_norm": 0.992905443952185,
+      "learning_rate": 2.2916178090741848e-06,
+      "loss": 1.5425,
+      "step": 5896
+    },
+    {
+      "epoch": 0.7868961836135575,
+      "grad_norm": 1.0099924911665865,
+      "learning_rate": 2.2888653418810114e-06,
+      "loss": 1.5827,
+      "step": 5897
+    },
+    {
+      "epoch": 0.7870296236989591,
+      "grad_norm": 0.9732471135416901,
+      "learning_rate": 2.2861143150400002e-06,
+      "loss": 1.5571,
+      "step": 5898
+    },
+    {
+      "epoch": 0.7871630637843609,
+      "grad_norm": 0.9352056643017831,
+      "learning_rate": 2.2833647290650084e-06,
+      "loss": 1.526,
+      "step": 5899
+    },
+    {
+      "epoch": 0.7872965038697625,
+      "grad_norm": 0.9444113116845154,
+      "learning_rate": 2.2806165844696284e-06,
+      "loss": 1.5729,
+      "step": 5900
+    },
+    {
+      "epoch": 0.7874299439551641,
+      "grad_norm": 0.9816140583735885,
+      "learning_rate": 2.277869881767182e-06,
+      "loss": 1.5734,
+      "step": 5901
+    },
+    {
+      "epoch": 0.7875633840405658,
+      "grad_norm": 1.2377672190640088,
+      "learning_rate": 2.2751246214707223e-06,
+      "loss": 1.5564,
+      "step": 5902
+    },
+    {
+      "epoch": 0.7876968241259674,
+      "grad_norm": 0.9171027423735342,
+      "learning_rate": 2.2723808040930282e-06,
+      "loss": 1.4997,
+      "step": 5903
+    },
+    {
+      "epoch": 0.7878302642113691,
+      "grad_norm": 1.0105329476647573,
+      "learning_rate": 2.2696384301466224e-06,
+      "loss": 1.5311,
+      "step": 5904
+    },
+    {
+      "epoch": 0.7879637042967708,
+      "grad_norm": 0.9299980163921587,
+      "learning_rate": 2.2668975001437433e-06,
+      "loss": 1.5344,
+      "step": 5905
+    },
+    {
+      "epoch": 0.7880971443821724,
+      "grad_norm": 0.9457043818410217,
+      "learning_rate": 2.264158014596366e-06,
+      "loss": 1.5521,
+      "step": 5906
+    },
+    {
+      "epoch": 0.7882305844675741,
+      "grad_norm": 1.2092222254834237,
+      "learning_rate": 2.2614199740161935e-06,
+      "loss": 1.5422,
+      "step": 5907
+    },
+    {
+      "epoch": 0.7883640245529757,
+      "grad_norm": 1.0681054745909435,
+      "learning_rate": 2.2586833789146657e-06,
+      "loss": 1.5889,
+      "step": 5908
+    },
+    {
+      "epoch": 0.7884974646383773,
+      "grad_norm": 0.9368681322416695,
+      "learning_rate": 2.2559482298029447e-06,
+      "loss": 1.5816,
+      "step": 5909
+    },
+    {
+      "epoch": 0.7886309047237791,
+      "grad_norm": 0.9228201723152973,
+      "learning_rate": 2.2532145271919247e-06,
+      "loss": 1.5449,
+      "step": 5910
+    },
+    {
+      "epoch": 0.7887643448091807,
+      "grad_norm": 0.943660701996512,
+      "learning_rate": 2.250482271592228e-06,
+      "loss": 1.5683,
+      "step": 5911
+    },
+    {
+      "epoch": 0.7888977848945823,
+      "grad_norm": 0.929628779619005,
+      "learning_rate": 2.247751463514214e-06,
+      "loss": 1.5122,
+      "step": 5912
+    },
+    {
+      "epoch": 0.789031224979984,
+      "grad_norm": 0.9387161790518015,
+      "learning_rate": 2.245022103467964e-06,
+      "loss": 1.5627,
+      "step": 5913
+    },
+    {
+      "epoch": 0.7891646650653856,
+      "grad_norm": 1.0924957479088928,
+      "learning_rate": 2.242294191963291e-06,
+      "loss": 1.5348,
+      "step": 5914
+    },
+    {
+      "epoch": 0.7892981051507872,
+      "grad_norm": 0.9504151391255395,
+      "learning_rate": 2.2395677295097362e-06,
+      "loss": 1.523,
+      "step": 5915
+    },
+    {
+      "epoch": 0.789431545236189,
+      "grad_norm": 0.9629607665332615,
+      "learning_rate": 2.236842716616575e-06,
+      "loss": 1.549,
+      "step": 5916
+    },
+    {
+      "epoch": 0.7895649853215906,
+      "grad_norm": 1.0660833347570207,
+      "learning_rate": 2.2341191537928087e-06,
+      "loss": 1.5227,
+      "step": 5917
+    },
+    {
+      "epoch": 0.7896984254069923,
+      "grad_norm": 0.953518244934922,
+      "learning_rate": 2.2313970415471643e-06,
+      "loss": 1.5568,
+      "step": 5918
+    },
+    {
+      "epoch": 0.7898318654923939,
+      "grad_norm": 1.0025411935106783,
+      "learning_rate": 2.2286763803881052e-06,
+      "loss": 1.5277,
+      "step": 5919
+    },
+    {
+      "epoch": 0.7899653055777955,
+      "grad_norm": 0.9253963712749954,
+      "learning_rate": 2.2259571708238194e-06,
+      "loss": 1.506,
+      "step": 5920
+    },
+    {
+      "epoch": 0.7900987456631973,
+      "grad_norm": 0.9896713961958903,
+      "learning_rate": 2.2232394133622225e-06,
+      "loss": 1.5484,
+      "step": 5921
+    },
+    {
+      "epoch": 0.7902321857485989,
+      "grad_norm": 1.1168051427517722,
+      "learning_rate": 2.220523108510959e-06,
+      "loss": 1.5591,
+      "step": 5922
+    },
+    {
+      "epoch": 0.7903656258340005,
+      "grad_norm": 0.9692435749687704,
+      "learning_rate": 2.2178082567774086e-06,
+      "loss": 1.5639,
+      "step": 5923
+    },
+    {
+      "epoch": 0.7904990659194022,
+      "grad_norm": 1.0227885427358632,
+      "learning_rate": 2.2150948586686728e-06,
+      "loss": 1.5236,
+      "step": 5924
+    },
+    {
+      "epoch": 0.7906325060048038,
+      "grad_norm": 0.9422956336171604,
+      "learning_rate": 2.21238291469158e-06,
+      "loss": 1.5855,
+      "step": 5925
+    },
+    {
+      "epoch": 0.7907659460902055,
+      "grad_norm": 1.3638108258879078,
+      "learning_rate": 2.2096724253526956e-06,
+      "loss": 1.5657,
+      "step": 5926
+    },
+    {
+      "epoch": 0.7908993861756072,
+      "grad_norm": 0.8997297461212911,
+      "learning_rate": 2.2069633911583067e-06,
+      "loss": 1.5429,
+      "step": 5927
+    },
+    {
+      "epoch": 0.7910328262610088,
+      "grad_norm": 0.964122119886273,
+      "learning_rate": 2.204255812614429e-06,
+      "loss": 1.5413,
+      "step": 5928
+    },
+    {
+      "epoch": 0.7911662663464104,
+      "grad_norm": 0.9482056082342423,
+      "learning_rate": 2.201549690226804e-06,
+      "loss": 1.5358,
+      "step": 5929
+    },
+    {
+      "epoch": 0.7912997064318121,
+      "grad_norm": 1.0861037743666542,
+      "learning_rate": 2.198845024500912e-06,
+      "loss": 1.5515,
+      "step": 5930
+    },
+    {
+      "epoch": 0.7914331465172137,
+      "grad_norm": 0.9464827990376123,
+      "learning_rate": 2.196141815941949e-06,
+      "loss": 1.5723,
+      "step": 5931
+    },
+    {
+      "epoch": 0.7915665866026155,
+      "grad_norm": 0.934897802356449,
+      "learning_rate": 2.193440065054843e-06,
+      "loss": 1.5334,
+      "step": 5932
+    },
+    {
+      "epoch": 0.7917000266880171,
+      "grad_norm": 0.9891163916956548,
+      "learning_rate": 2.1907397723442536e-06,
+      "loss": 1.5606,
+      "step": 5933
+    },
+    {
+      "epoch": 0.7918334667734187,
+      "grad_norm": 0.9667703054129473,
+      "learning_rate": 2.188040938314564e-06,
+      "loss": 1.5315,
+      "step": 5934
+    },
+    {
+      "epoch": 0.7919669068588204,
+      "grad_norm": 0.9296819166809209,
+      "learning_rate": 2.1853435634698837e-06,
+      "loss": 1.5643,
+      "step": 5935
+    },
+    {
+      "epoch": 0.792100346944222,
+      "grad_norm": 0.9552094294250992,
+      "learning_rate": 2.1826476483140503e-06,
+      "loss": 1.5044,
+      "step": 5936
+    },
+    {
+      "epoch": 0.7922337870296237,
+      "grad_norm": 1.0811395195378388,
+      "learning_rate": 2.1799531933506346e-06,
+      "loss": 1.5888,
+      "step": 5937
+    },
+    {
+      "epoch": 0.7923672271150254,
+      "grad_norm": 1.0395532411915318,
+      "learning_rate": 2.177260199082928e-06,
+      "loss": 1.5514,
+      "step": 5938
+    },
+    {
+      "epoch": 0.792500667200427,
+      "grad_norm": 0.9561248759072963,
+      "learning_rate": 2.174568666013951e-06,
+      "loss": 1.5294,
+      "step": 5939
+    },
+    {
+      "epoch": 0.7926341072858286,
+      "grad_norm": 0.9242513920785822,
+      "learning_rate": 2.1718785946464483e-06,
+      "loss": 1.5342,
+      "step": 5940
+    },
+    {
+      "epoch": 0.7927675473712303,
+      "grad_norm": 0.9534837808168115,
+      "learning_rate": 2.1691899854829014e-06,
+      "loss": 1.5325,
+      "step": 5941
+    },
+    {
+      "epoch": 0.792900987456632,
+      "grad_norm": 0.9322289028522723,
+      "learning_rate": 2.1665028390255073e-06,
+      "loss": 1.5002,
+      "step": 5942
+    },
+    {
+      "epoch": 0.7930344275420337,
+      "grad_norm": 1.1723532471213491,
+      "learning_rate": 2.1638171557761943e-06,
+      "loss": 1.5653,
+      "step": 5943
+    },
+    {
+      "epoch": 0.7931678676274353,
+      "grad_norm": 1.1432508740043539,
+      "learning_rate": 2.1611329362366195e-06,
+      "loss": 1.5889,
+      "step": 5944
+    },
+    {
+      "epoch": 0.7933013077128369,
+      "grad_norm": 0.9622094585479455,
+      "learning_rate": 2.1584501809081628e-06,
+      "loss": 1.604,
+      "step": 5945
+    },
+    {
+      "epoch": 0.7934347477982386,
+      "grad_norm": 0.9865589871665827,
+      "learning_rate": 2.1557688902919305e-06,
+      "loss": 1.5216,
+      "step": 5946
+    },
+    {
+      "epoch": 0.7935681878836403,
+      "grad_norm": 1.0241441606366657,
+      "learning_rate": 2.1530890648887628e-06,
+      "loss": 1.5134,
+      "step": 5947
+    },
+    {
+      "epoch": 0.7937016279690419,
+      "grad_norm": 0.9301015988878937,
+      "learning_rate": 2.150410705199216e-06,
+      "loss": 1.5294,
+      "step": 5948
+    },
+    {
+      "epoch": 0.7938350680544436,
+      "grad_norm": 0.9621334353445851,
+      "learning_rate": 2.147733811723579e-06,
+      "loss": 1.5781,
+      "step": 5949
+    },
+    {
+      "epoch": 0.7939685081398452,
+      "grad_norm": 0.9655006569053254,
+      "learning_rate": 2.145058384961862e-06,
+      "loss": 1.5538,
+      "step": 5950
+    },
+    {
+      "epoch": 0.7941019482252468,
+      "grad_norm": 1.0251786646717194,
+      "learning_rate": 2.1423844254138105e-06,
+      "loss": 1.5959,
+      "step": 5951
+    },
+    {
+      "epoch": 0.7942353883106485,
+      "grad_norm": 0.9424616777571333,
+      "learning_rate": 2.139711933578885e-06,
+      "loss": 1.5048,
+      "step": 5952
+    },
+    {
+      "epoch": 0.7943688283960502,
+      "grad_norm": 0.9412547582328011,
+      "learning_rate": 2.1370409099562774e-06,
+      "loss": 1.5633,
+      "step": 5953
+    },
+    {
+      "epoch": 0.7945022684814518,
+      "grad_norm": 1.016228232058081,
+      "learning_rate": 2.134371355044902e-06,
+      "loss": 1.5567,
+      "step": 5954
+    },
+    {
+      "epoch": 0.7946357085668535,
+      "grad_norm": 0.9500451975852726,
+      "learning_rate": 2.131703269343407e-06,
+      "loss": 1.5571,
+      "step": 5955
+    },
+    {
+      "epoch": 0.7947691486522551,
+      "grad_norm": 0.9985865820172933,
+      "learning_rate": 2.1290366533501572e-06,
+      "loss": 1.5439,
+      "step": 5956
+    },
+    {
+      "epoch": 0.7949025887376568,
+      "grad_norm": 1.0616401273574188,
+      "learning_rate": 2.126371507563244e-06,
+      "loss": 1.5213,
+      "step": 5957
+    },
+    {
+      "epoch": 0.7950360288230585,
+      "grad_norm": 0.9523800662623815,
+      "learning_rate": 2.1237078324804906e-06,
+      "loss": 1.5169,
+      "step": 5958
+    },
+    {
+      "epoch": 0.7951694689084601,
+      "grad_norm": 0.9444862343077112,
+      "learning_rate": 2.1210456285994397e-06,
+      "loss": 1.5903,
+      "step": 5959
+    },
+    {
+      "epoch": 0.7953029089938618,
+      "grad_norm": 1.037122868292527,
+      "learning_rate": 2.1183848964173604e-06,
+      "loss": 1.5931,
+      "step": 5960
+    },
+    {
+      "epoch": 0.7954363490792634,
+      "grad_norm": 0.9387631641570781,
+      "learning_rate": 2.115725636431243e-06,
+      "loss": 1.5488,
+      "step": 5961
+    },
+    {
+      "epoch": 0.795569789164665,
+      "grad_norm": 1.1749210941871775,
+      "learning_rate": 2.1130678491378143e-06,
+      "loss": 1.544,
+      "step": 5962
+    },
+    {
+      "epoch": 0.7957032292500668,
+      "grad_norm": 0.9411833484581066,
+      "learning_rate": 2.110411535033515e-06,
+      "loss": 1.5385,
+      "step": 5963
+    },
+    {
+      "epoch": 0.7958366693354684,
+      "grad_norm": 0.9404179032536638,
+      "learning_rate": 2.1077566946145124e-06,
+      "loss": 1.5881,
+      "step": 5964
+    },
+    {
+      "epoch": 0.79597010942087,
+      "grad_norm": 0.9781497662602053,
+      "learning_rate": 2.1051033283767054e-06,
+      "loss": 1.5534,
+      "step": 5965
+    },
+    {
+      "epoch": 0.7961035495062717,
+      "grad_norm": 0.9400404759624681,
+      "learning_rate": 2.102451436815709e-06,
+      "loss": 1.5625,
+      "step": 5966
+    },
+    {
+      "epoch": 0.7962369895916733,
+      "grad_norm": 0.91952723034791,
+      "learning_rate": 2.0998010204268683e-06,
+      "loss": 1.5542,
+      "step": 5967
+    },
+    {
+      "epoch": 0.7963704296770749,
+      "grad_norm": 0.9219414927078943,
+      "learning_rate": 2.0971520797052468e-06,
+      "loss": 1.5401,
+      "step": 5968
+    },
+    {
+      "epoch": 0.7965038697624767,
+      "grad_norm": 0.9602262059129528,
+      "learning_rate": 2.0945046151456417e-06,
+      "loss": 1.5379,
+      "step": 5969
+    },
+    {
+      "epoch": 0.7966373098478783,
+      "grad_norm": 0.9345892591110488,
+      "learning_rate": 2.091858627242568e-06,
+      "loss": 1.5246,
+      "step": 5970
+    },
+    {
+      "epoch": 0.79677074993328,
+      "grad_norm": 0.9407983789526347,
+      "learning_rate": 2.089214116490261e-06,
+      "loss": 1.5373,
+      "step": 5971
+    },
+    {
+      "epoch": 0.7969041900186816,
+      "grad_norm": 0.9833161312725881,
+      "learning_rate": 2.0865710833826936e-06,
+      "loss": 1.547,
+      "step": 5972
+    },
+    {
+      "epoch": 0.7970376301040832,
+      "grad_norm": 1.0593932491674578,
+      "learning_rate": 2.0839295284135486e-06,
+      "loss": 1.5509,
+      "step": 5973
+    },
+    {
+      "epoch": 0.797171070189485,
+      "grad_norm": 1.046152778746016,
+      "learning_rate": 2.0812894520762416e-06,
+      "loss": 1.6058,
+      "step": 5974
+    },
+    {
+      "epoch": 0.7973045102748866,
+      "grad_norm": 1.0912168422367754,
+      "learning_rate": 2.078650854863903e-06,
+      "loss": 1.5468,
+      "step": 5975
+    },
+    {
+      "epoch": 0.7974379503602882,
+      "grad_norm": 0.9464811881645303,
+      "learning_rate": 2.0760137372694e-06,
+      "loss": 1.551,
+      "step": 5976
+    },
+    {
+      "epoch": 0.7975713904456899,
+      "grad_norm": 0.9185730065970497,
+      "learning_rate": 2.0733780997853124e-06,
+      "loss": 1.4854,
+      "step": 5977
+    },
+    {
+      "epoch": 0.7977048305310915,
+      "grad_norm": 0.9526089149618532,
+      "learning_rate": 2.0707439429039454e-06,
+      "loss": 1.6068,
+      "step": 5978
+    },
+    {
+      "epoch": 0.7978382706164932,
+      "grad_norm": 0.9587953235425581,
+      "learning_rate": 2.0681112671173344e-06,
+      "loss": 1.6037,
+      "step": 5979
+    },
+    {
+      "epoch": 0.7979717107018949,
+      "grad_norm": 0.9361644772074595,
+      "learning_rate": 2.065480072917231e-06,
+      "loss": 1.5868,
+      "step": 5980
+    },
+    {
+      "epoch": 0.7981051507872965,
+      "grad_norm": 0.9534441446175175,
+      "learning_rate": 2.062850360795112e-06,
+      "loss": 1.5701,
+      "step": 5981
+    },
+    {
+      "epoch": 0.7982385908726981,
+      "grad_norm": 0.9946923698582439,
+      "learning_rate": 2.060222131242178e-06,
+      "loss": 1.5967,
+      "step": 5982
+    },
+    {
+      "epoch": 0.7983720309580998,
+      "grad_norm": 0.9626053858910402,
+      "learning_rate": 2.057595384749349e-06,
+      "loss": 1.578,
+      "step": 5983
+    },
+    {
+      "epoch": 0.7985054710435014,
+      "grad_norm": 1.1359335916882283,
+      "learning_rate": 2.054970121807278e-06,
+      "loss": 1.5129,
+      "step": 5984
+    },
+    {
+      "epoch": 0.7986389111289032,
+      "grad_norm": 1.1202578382909025,
+      "learning_rate": 2.0523463429063295e-06,
+      "loss": 1.5608,
+      "step": 5985
+    },
+    {
+      "epoch": 0.7987723512143048,
+      "grad_norm": 1.043896917536381,
+      "learning_rate": 2.0497240485365975e-06,
+      "loss": 1.5364,
+      "step": 5986
+    },
+    {
+      "epoch": 0.7989057912997064,
+      "grad_norm": 1.0753022531810645,
+      "learning_rate": 2.0471032391878963e-06,
+      "loss": 1.5591,
+      "step": 5987
+    },
+    {
+      "epoch": 0.7990392313851081,
+      "grad_norm": 1.2319921435864252,
+      "learning_rate": 2.0444839153497618e-06,
+      "loss": 1.5409,
+      "step": 5988
+    },
+    {
+      "epoch": 0.7991726714705097,
+      "grad_norm": 1.00815621813523,
+      "learning_rate": 2.041866077511453e-06,
+      "loss": 1.5381,
+      "step": 5989
+    },
+    {
+      "epoch": 0.7993061115559114,
+      "grad_norm": 0.9256965403165084,
+      "learning_rate": 2.039249726161957e-06,
+      "loss": 1.5051,
+      "step": 5990
+    },
+    {
+      "epoch": 0.7994395516413131,
+      "grad_norm": 0.9415742466538367,
+      "learning_rate": 2.0366348617899745e-06,
+      "loss": 1.4909,
+      "step": 5991
+    },
+    {
+      "epoch": 0.7995729917267147,
+      "grad_norm": 1.204722600477448,
+      "learning_rate": 2.0340214848839347e-06,
+      "loss": 1.5813,
+      "step": 5992
+    },
+    {
+      "epoch": 0.7997064318121163,
+      "grad_norm": 0.9223199394194576,
+      "learning_rate": 2.0314095959319822e-06,
+      "loss": 1.5269,
+      "step": 5993
+    },
+    {
+      "epoch": 0.799839871897518,
+      "grad_norm": 0.9377131565914819,
+      "learning_rate": 2.0287991954219945e-06,
+      "loss": 1.4804,
+      "step": 5994
+    },
+    {
+      "epoch": 0.7999733119829197,
+      "grad_norm": 0.9766250429689536,
+      "learning_rate": 2.0261902838415605e-06,
+      "loss": 1.5481,
+      "step": 5995
+    },
+    {
+      "epoch": 0.8001067520683213,
+      "grad_norm": 0.9549561111606707,
+      "learning_rate": 2.023582861677995e-06,
+      "loss": 1.5329,
+      "step": 5996
+    },
+    {
+      "epoch": 0.800240192153723,
+      "grad_norm": 0.9789891454720687,
+      "learning_rate": 2.020976929418338e-06,
+      "loss": 1.5963,
+      "step": 5997
+    },
+    {
+      "epoch": 0.8003736322391246,
+      "grad_norm": 0.9398469925192877,
+      "learning_rate": 2.018372487549346e-06,
+      "loss": 1.4924,
+      "step": 5998
+    },
+    {
+      "epoch": 0.8005070723245263,
+      "grad_norm": 0.9346327961340182,
+      "learning_rate": 2.015769536557499e-06,
+      "loss": 1.5465,
+      "step": 5999
+    },
+    {
+      "epoch": 0.800640512409928,
+      "grad_norm": 0.9516873692804836,
+      "learning_rate": 2.0131680769289964e-06,
+      "loss": 1.6011,
+      "step": 6000
+    },
+    {
+      "epoch": 0.8007739524953296,
+      "grad_norm": 1.0067538076216085,
+      "learning_rate": 2.0105681091497653e-06,
+      "loss": 1.5703,
+      "step": 6001
+    },
+    {
+      "epoch": 0.8009073925807313,
+      "grad_norm": 0.9585095677361529,
+      "learning_rate": 2.007969633705449e-06,
+      "loss": 1.5311,
+      "step": 6002
+    },
+    {
+      "epoch": 0.8010408326661329,
+      "grad_norm": 0.9360385818007098,
+      "learning_rate": 2.00537265108141e-06,
+      "loss": 1.5563,
+      "step": 6003
+    },
+    {
+      "epoch": 0.8011742727515345,
+      "grad_norm": 0.9620281711842354,
+      "learning_rate": 2.002777161762739e-06,
+      "loss": 1.5605,
+      "step": 6004
+    },
+    {
+      "epoch": 0.8013077128369362,
+      "grad_norm": 0.9717123448452647,
+      "learning_rate": 2.000183166234244e-06,
+      "loss": 1.5998,
+      "step": 6005
+    },
+    {
+      "epoch": 0.8014411529223379,
+      "grad_norm": 1.2665547727823556,
+      "learning_rate": 1.997590664980451e-06,
+      "loss": 1.5412,
+      "step": 6006
+    },
+    {
+      "epoch": 0.8015745930077395,
+      "grad_norm": 0.9436274926419813,
+      "learning_rate": 1.9949996584856093e-06,
+      "loss": 1.5404,
+      "step": 6007
+    },
+    {
+      "epoch": 0.8017080330931412,
+      "grad_norm": 0.9485403732848344,
+      "learning_rate": 1.992410147233692e-06,
+      "loss": 1.4864,
+      "step": 6008
+    },
+    {
+      "epoch": 0.8018414731785428,
+      "grad_norm": 0.9397520552993349,
+      "learning_rate": 1.989822131708391e-06,
+      "loss": 1.5356,
+      "step": 6009
+    },
+    {
+      "epoch": 0.8019749132639445,
+      "grad_norm": 0.9678083881364095,
+      "learning_rate": 1.9872356123931137e-06,
+      "loss": 1.5437,
+      "step": 6010
+    },
+    {
+      "epoch": 0.8021083533493462,
+      "grad_norm": 8.488638953717834,
+      "learning_rate": 1.984650589770998e-06,
+      "loss": 1.5579,
+      "step": 6011
+    },
+    {
+      "epoch": 0.8022417934347478,
+      "grad_norm": 0.9477151839816731,
+      "learning_rate": 1.982067064324893e-06,
+      "loss": 1.5687,
+      "step": 6012
+    },
+    {
+      "epoch": 0.8023752335201495,
+      "grad_norm": 0.9561015940175384,
+      "learning_rate": 1.979485036537373e-06,
+      "loss": 1.5424,
+      "step": 6013
+    },
+    {
+      "epoch": 0.8025086736055511,
+      "grad_norm": 0.940578131890771,
+      "learning_rate": 1.976904506890729e-06,
+      "loss": 1.5604,
+      "step": 6014
+    },
+    {
+      "epoch": 0.8026421136909527,
+      "grad_norm": 0.971107847461901,
+      "learning_rate": 1.9743254758669794e-06,
+      "loss": 1.5516,
+      "step": 6015
+    },
+    {
+      "epoch": 0.8027755537763545,
+      "grad_norm": 1.151724003936446,
+      "learning_rate": 1.971747943947855e-06,
+      "loss": 1.5418,
+      "step": 6016
+    },
+    {
+      "epoch": 0.8029089938617561,
+      "grad_norm": 0.9436979624078382,
+      "learning_rate": 1.969171911614809e-06,
+      "loss": 1.5554,
+      "step": 6017
+    },
+    {
+      "epoch": 0.8030424339471577,
+      "grad_norm": 1.088649547266759,
+      "learning_rate": 1.9665973793490134e-06,
+      "loss": 1.546,
+      "step": 6018
+    },
+    {
+      "epoch": 0.8031758740325594,
+      "grad_norm": 1.1775693108073233,
+      "learning_rate": 1.964024347631367e-06,
+      "loss": 1.5331,
+      "step": 6019
+    },
+    {
+      "epoch": 0.803309314117961,
+      "grad_norm": 0.9338043323421088,
+      "learning_rate": 1.9614528169424784e-06,
+      "loss": 1.5825,
+      "step": 6020
+    },
+    {
+      "epoch": 0.8034427542033626,
+      "grad_norm": 0.9745864204912411,
+      "learning_rate": 1.958882787762678e-06,
+      "loss": 1.5838,
+      "step": 6021
+    },
+    {
+      "epoch": 0.8035761942887644,
+      "grad_norm": 1.5277909690793727,
+      "learning_rate": 1.9563142605720254e-06,
+      "loss": 1.5049,
+      "step": 6022
+    },
+    {
+      "epoch": 0.803709634374166,
+      "grad_norm": 0.9821884824130115,
+      "learning_rate": 1.953747235850287e-06,
+      "loss": 1.5272,
+      "step": 6023
+    },
+    {
+      "epoch": 0.8038430744595677,
+      "grad_norm": 0.9310206254375682,
+      "learning_rate": 1.9511817140769563e-06,
+      "loss": 1.5323,
+      "step": 6024
+    },
+    {
+      "epoch": 0.8039765145449693,
+      "grad_norm": 1.0213597847479,
+      "learning_rate": 1.948617695731242e-06,
+      "loss": 1.5192,
+      "step": 6025
+    },
+    {
+      "epoch": 0.8041099546303709,
+      "grad_norm": 0.9562151575602164,
+      "learning_rate": 1.9460551812920703e-06,
+      "loss": 1.5758,
+      "step": 6026
+    },
+    {
+      "epoch": 0.8042433947157727,
+      "grad_norm": 0.9387251801749865,
+      "learning_rate": 1.943494171238095e-06,
+      "loss": 1.5662,
+      "step": 6027
+    },
+    {
+      "epoch": 0.8043768348011743,
+      "grad_norm": 0.9762010066502882,
+      "learning_rate": 1.9409346660476834e-06,
+      "loss": 1.573,
+      "step": 6028
+    },
+    {
+      "epoch": 0.8045102748865759,
+      "grad_norm": 1.087852765493359,
+      "learning_rate": 1.938376666198919e-06,
+      "loss": 1.5465,
+      "step": 6029
+    },
+    {
+      "epoch": 0.8046437149719776,
+      "grad_norm": 0.938575775358334,
+      "learning_rate": 1.935820172169609e-06,
+      "loss": 1.5516,
+      "step": 6030
+    },
+    {
+      "epoch": 0.8047771550573792,
+      "grad_norm": 0.965176200493375,
+      "learning_rate": 1.933265184437274e-06,
+      "loss": 1.5584,
+      "step": 6031
+    },
+    {
+      "epoch": 0.8049105951427808,
+      "grad_norm": 1.2915982271448758,
+      "learning_rate": 1.930711703479162e-06,
+      "loss": 1.594,
+      "step": 6032
+    },
+    {
+      "epoch": 0.8050440352281826,
+      "grad_norm": 0.9273187149756227,
+      "learning_rate": 1.928159729772231e-06,
+      "loss": 1.5509,
+      "step": 6033
+    },
+    {
+      "epoch": 0.8051774753135842,
+      "grad_norm": 0.9188671634655157,
+      "learning_rate": 1.925609263793162e-06,
+      "loss": 1.5233,
+      "step": 6034
+    },
+    {
+      "epoch": 0.8053109153989858,
+      "grad_norm": 1.009702927758053,
+      "learning_rate": 1.9230603060183493e-06,
+      "loss": 1.5312,
+      "step": 6035
+    },
+    {
+      "epoch": 0.8054443554843875,
+      "grad_norm": 1.100233587871606,
+      "learning_rate": 1.920512856923914e-06,
+      "loss": 1.5386,
+      "step": 6036
+    },
+    {
+      "epoch": 0.8055777955697891,
+      "grad_norm": 0.9308659676568186,
+      "learning_rate": 1.9179669169856896e-06,
+      "loss": 1.537,
+      "step": 6037
+    },
+    {
+      "epoch": 0.8057112356551909,
+      "grad_norm": 0.9407626383495175,
+      "learning_rate": 1.915422486679227e-06,
+      "loss": 1.5401,
+      "step": 6038
+    },
+    {
+      "epoch": 0.8058446757405925,
+      "grad_norm": 0.9436636824443787,
+      "learning_rate": 1.912879566479795e-06,
+      "loss": 1.5742,
+      "step": 6039
+    },
+    {
+      "epoch": 0.8059781158259941,
+      "grad_norm": 0.9418856484629331,
+      "learning_rate": 1.9103381568623868e-06,
+      "loss": 1.5231,
+      "step": 6040
+    },
+    {
+      "epoch": 0.8061115559113958,
+      "grad_norm": 0.9938636042780336,
+      "learning_rate": 1.907798258301707e-06,
+      "loss": 1.5326,
+      "step": 6041
+    },
+    {
+      "epoch": 0.8062449959967974,
+      "grad_norm": 0.9167069778310091,
+      "learning_rate": 1.9052598712721771e-06,
+      "loss": 1.5517,
+      "step": 6042
+    },
+    {
+      "epoch": 0.8063784360821991,
+      "grad_norm": 0.9396867142428339,
+      "learning_rate": 1.9027229962479433e-06,
+      "loss": 1.5704,
+      "step": 6043
+    },
+    {
+      "epoch": 0.8065118761676008,
+      "grad_norm": 0.9421359789583224,
+      "learning_rate": 1.9001876337028635e-06,
+      "loss": 1.5256,
+      "step": 6044
+    },
+    {
+      "epoch": 0.8066453162530024,
+      "grad_norm": 0.973198022271884,
+      "learning_rate": 1.8976537841105136e-06,
+      "loss": 1.528,
+      "step": 6045
+    },
+    {
+      "epoch": 0.806778756338404,
+      "grad_norm": 0.9243015337796414,
+      "learning_rate": 1.895121447944185e-06,
+      "loss": 1.5705,
+      "step": 6046
+    },
+    {
+      "epoch": 0.8069121964238057,
+      "grad_norm": 0.9373269634820086,
+      "learning_rate": 1.8925906256768957e-06,
+      "loss": 1.5274,
+      "step": 6047
+    },
+    {
+      "epoch": 0.8070456365092074,
+      "grad_norm": 1.108272451485646,
+      "learning_rate": 1.8900613177813708e-06,
+      "loss": 1.5485,
+      "step": 6048
+    },
+    {
+      "epoch": 0.807179076594609,
+      "grad_norm": 0.9455270267638988,
+      "learning_rate": 1.8875335247300564e-06,
+      "loss": 1.5525,
+      "step": 6049
+    },
+    {
+      "epoch": 0.8073125166800107,
+      "grad_norm": 0.948942726258177,
+      "learning_rate": 1.8850072469951142e-06,
+      "loss": 1.5362,
+      "step": 6050
+    },
+    {
+      "epoch": 0.8074459567654123,
+      "grad_norm": 1.0549675256800648,
+      "learning_rate": 1.8824824850484269e-06,
+      "loss": 1.5568,
+      "step": 6051
+    },
+    {
+      "epoch": 0.807579396850814,
+      "grad_norm": 0.9693613341191636,
+      "learning_rate": 1.8799592393615906e-06,
+      "loss": 1.5386,
+      "step": 6052
+    },
+    {
+      "epoch": 0.8077128369362157,
+      "grad_norm": 1.002495730710032,
+      "learning_rate": 1.877437510405915e-06,
+      "loss": 1.5567,
+      "step": 6053
+    },
+    {
+      "epoch": 0.8078462770216173,
+      "grad_norm": 0.9399879061201342,
+      "learning_rate": 1.8749172986524378e-06,
+      "loss": 1.5198,
+      "step": 6054
+    },
+    {
+      "epoch": 0.807979717107019,
+      "grad_norm": 0.916955889393747,
+      "learning_rate": 1.8723986045719e-06,
+      "loss": 1.5238,
+      "step": 6055
+    },
+    {
+      "epoch": 0.8081131571924206,
+      "grad_norm": 0.9305269696315253,
+      "learning_rate": 1.8698814286347678e-06,
+      "loss": 1.5412,
+      "step": 6056
+    },
+    {
+      "epoch": 0.8082465972778222,
+      "grad_norm": 0.9234504269114264,
+      "learning_rate": 1.867365771311216e-06,
+      "loss": 1.4926,
+      "step": 6057
+    },
+    {
+      "epoch": 0.808380037363224,
+      "grad_norm": 0.9598446073788602,
+      "learning_rate": 1.8648516330711486e-06,
+      "loss": 1.5225,
+      "step": 6058
+    },
+    {
+      "epoch": 0.8085134774486256,
+      "grad_norm": 1.006044534644564,
+      "learning_rate": 1.8623390143841735e-06,
+      "loss": 1.5641,
+      "step": 6059
+    },
+    {
+      "epoch": 0.8086469175340272,
+      "grad_norm": 0.9891625291510784,
+      "learning_rate": 1.8598279157196164e-06,
+      "loss": 1.5121,
+      "step": 6060
+    },
+    {
+      "epoch": 0.8087803576194289,
+      "grad_norm": 0.9313511828058488,
+      "learning_rate": 1.8573183375465286e-06,
+      "loss": 1.5156,
+      "step": 6061
+    },
+    {
+      "epoch": 0.8089137977048305,
+      "grad_norm": 0.9339926309721247,
+      "learning_rate": 1.8548102803336677e-06,
+      "loss": 1.5326,
+      "step": 6062
+    },
+    {
+      "epoch": 0.8090472377902321,
+      "grad_norm": 0.9664081912314659,
+      "learning_rate": 1.8523037445495095e-06,
+      "loss": 1.5461,
+      "step": 6063
+    },
+    {
+      "epoch": 0.8091806778756339,
+      "grad_norm": 1.12857579777555,
+      "learning_rate": 1.8497987306622446e-06,
+      "loss": 1.5084,
+      "step": 6064
+    },
+    {
+      "epoch": 0.8093141179610355,
+      "grad_norm": 1.0205222142779857,
+      "learning_rate": 1.8472952391397846e-06,
+      "loss": 1.5501,
+      "step": 6065
+    },
+    {
+      "epoch": 0.8094475580464372,
+      "grad_norm": 0.9269906202615147,
+      "learning_rate": 1.844793270449753e-06,
+      "loss": 1.5647,
+      "step": 6066
+    },
+    {
+      "epoch": 0.8095809981318388,
+      "grad_norm": 0.9111441101773377,
+      "learning_rate": 1.8422928250594884e-06,
+      "loss": 1.532,
+      "step": 6067
+    },
+    {
+      "epoch": 0.8097144382172404,
+      "grad_norm": 1.0423582302266032,
+      "learning_rate": 1.8397939034360424e-06,
+      "loss": 1.5852,
+      "step": 6068
+    },
+    {
+      "epoch": 0.8098478783026422,
+      "grad_norm": 0.9495371122585425,
+      "learning_rate": 1.837296506046189e-06,
+      "loss": 1.5211,
+      "step": 6069
+    },
+    {
+      "epoch": 0.8099813183880438,
+      "grad_norm": 0.9361902779942389,
+      "learning_rate": 1.834800633356414e-06,
+      "loss": 1.5148,
+      "step": 6070
+    },
+    {
+      "epoch": 0.8101147584734454,
+      "grad_norm": 1.234996354576864,
+      "learning_rate": 1.8323062858329155e-06,
+      "loss": 1.6077,
+      "step": 6071
+    },
+    {
+      "epoch": 0.8102481985588471,
+      "grad_norm": 0.9278663038573935,
+      "learning_rate": 1.82981346394161e-06,
+      "loss": 1.5558,
+      "step": 6072
+    },
+    {
+      "epoch": 0.8103816386442487,
+      "grad_norm": 0.9476924229244504,
+      "learning_rate": 1.827322168148129e-06,
+      "loss": 1.5618,
+      "step": 6073
+    },
+    {
+      "epoch": 0.8105150787296503,
+      "grad_norm": 0.9319818911496207,
+      "learning_rate": 1.8248323989178151e-06,
+      "loss": 1.5607,
+      "step": 6074
+    },
+    {
+      "epoch": 0.8106485188150521,
+      "grad_norm": 0.9145783092870352,
+      "learning_rate": 1.8223441567157329e-06,
+      "loss": 1.5446,
+      "step": 6075
+    },
+    {
+      "epoch": 0.8107819589004537,
+      "grad_norm": 0.9381651247628452,
+      "learning_rate": 1.8198574420066572e-06,
+      "loss": 1.5931,
+      "step": 6076
+    },
+    {
+      "epoch": 0.8109153989858553,
+      "grad_norm": 0.9248267670752973,
+      "learning_rate": 1.8173722552550766e-06,
+      "loss": 1.5598,
+      "step": 6077
+    },
+    {
+      "epoch": 0.811048839071257,
+      "grad_norm": 0.933228793651718,
+      "learning_rate": 1.8148885969251928e-06,
+      "loss": 1.5767,
+      "step": 6078
+    },
+    {
+      "epoch": 0.8111822791566586,
+      "grad_norm": 0.907509359626052,
+      "learning_rate": 1.8124064674809316e-06,
+      "loss": 1.5065,
+      "step": 6079
+    },
+    {
+      "epoch": 0.8113157192420604,
+      "grad_norm": 0.9228783070319467,
+      "learning_rate": 1.8099258673859221e-06,
+      "loss": 1.4855,
+      "step": 6080
+    },
+    {
+      "epoch": 0.811449159327462,
+      "grad_norm": 1.0177477019581869,
+      "learning_rate": 1.8074467971035136e-06,
+      "loss": 1.4958,
+      "step": 6081
+    },
+    {
+      "epoch": 0.8115825994128636,
+      "grad_norm": 0.9588231215311086,
+      "learning_rate": 1.8049692570967647e-06,
+      "loss": 1.5646,
+      "step": 6082
+    },
+    {
+      "epoch": 0.8117160394982653,
+      "grad_norm": 0.9155594177985604,
+      "learning_rate": 1.8024932478284584e-06,
+      "loss": 1.5231,
+      "step": 6083
+    },
+    {
+      "epoch": 0.8118494795836669,
+      "grad_norm": 0.9565317839445175,
+      "learning_rate": 1.8000187697610804e-06,
+      "loss": 1.6076,
+      "step": 6084
+    },
+    {
+      "epoch": 0.8119829196690685,
+      "grad_norm": 0.9745756544326605,
+      "learning_rate": 1.7975458233568332e-06,
+      "loss": 1.5622,
+      "step": 6085
+    },
+    {
+      "epoch": 0.8121163597544703,
+      "grad_norm": 1.0317394403946698,
+      "learning_rate": 1.7950744090776407e-06,
+      "loss": 1.5838,
+      "step": 6086
+    },
+    {
+      "epoch": 0.8122497998398719,
+      "grad_norm": 0.9537967660036222,
+      "learning_rate": 1.792604527385131e-06,
+      "loss": 1.524,
+      "step": 6087
+    },
+    {
+      "epoch": 0.8123832399252735,
+      "grad_norm": 1.1140810286258387,
+      "learning_rate": 1.7901361787406524e-06,
+      "loss": 1.5551,
+      "step": 6088
+    },
+    {
+      "epoch": 0.8125166800106752,
+      "grad_norm": 0.9970484456453595,
+      "learning_rate": 1.7876693636052588e-06,
+      "loss": 1.563,
+      "step": 6089
+    },
+    {
+      "epoch": 0.8126501200960768,
+      "grad_norm": 0.975647861270303,
+      "learning_rate": 1.785204082439731e-06,
+      "loss": 1.5874,
+      "step": 6090
+    },
+    {
+      "epoch": 0.8127835601814786,
+      "grad_norm": 0.9369203293774957,
+      "learning_rate": 1.7827403357045514e-06,
+      "loss": 1.5696,
+      "step": 6091
+    },
+    {
+      "epoch": 0.8129170002668802,
+      "grad_norm": 1.0641811139775401,
+      "learning_rate": 1.7802781238599164e-06,
+      "loss": 1.5656,
+      "step": 6092
+    },
+    {
+      "epoch": 0.8130504403522818,
+      "grad_norm": 0.9475062980136262,
+      "learning_rate": 1.777817447365746e-06,
+      "loss": 1.5263,
+      "step": 6093
+    },
+    {
+      "epoch": 0.8131838804376835,
+      "grad_norm": 0.9652813026259098,
+      "learning_rate": 1.7753583066816637e-06,
+      "loss": 1.5176,
+      "step": 6094
+    },
+    {
+      "epoch": 0.8133173205230851,
+      "grad_norm": 1.1089594490573167,
+      "learning_rate": 1.7729007022670085e-06,
+      "loss": 1.5231,
+      "step": 6095
+    },
+    {
+      "epoch": 0.8134507606084868,
+      "grad_norm": 0.9258367136985601,
+      "learning_rate": 1.7704446345808312e-06,
+      "loss": 1.5387,
+      "step": 6096
+    },
+    {
+      "epoch": 0.8135842006938885,
+      "grad_norm": 1.0577847232171296,
+      "learning_rate": 1.7679901040819004e-06,
+      "loss": 1.6006,
+      "step": 6097
+    },
+    {
+      "epoch": 0.8137176407792901,
+      "grad_norm": 1.1127882285883361,
+      "learning_rate": 1.7655371112286946e-06,
+      "loss": 1.5717,
+      "step": 6098
+    },
+    {
+      "epoch": 0.8138510808646917,
+      "grad_norm": 1.0673924071414274,
+      "learning_rate": 1.7630856564793996e-06,
+      "loss": 1.5309,
+      "step": 6099
+    },
+    {
+      "epoch": 0.8139845209500934,
+      "grad_norm": 0.9464251733577566,
+      "learning_rate": 1.760635740291926e-06,
+      "loss": 1.5303,
+      "step": 6100
+    },
+    {
+      "epoch": 0.814117961035495,
+      "grad_norm": 0.9319478908646294,
+      "learning_rate": 1.758187363123889e-06,
+      "loss": 1.5401,
+      "step": 6101
+    },
+    {
+      "epoch": 0.8142514011208967,
+      "grad_norm": 0.9287040927493834,
+      "learning_rate": 1.7557405254326144e-06,
+      "loss": 1.5454,
+      "step": 6102
+    },
+    {
+      "epoch": 0.8143848412062984,
+      "grad_norm": 0.9409713573767424,
+      "learning_rate": 1.7532952276751424e-06,
+      "loss": 1.5564,
+      "step": 6103
+    },
+    {
+      "epoch": 0.8145182812917,
+      "grad_norm": 1.1149430849686321,
+      "learning_rate": 1.7508514703082336e-06,
+      "loss": 1.5285,
+      "step": 6104
+    },
+    {
+      "epoch": 0.8146517213771017,
+      "grad_norm": 0.942572754627203,
+      "learning_rate": 1.748409253788349e-06,
+      "loss": 1.5839,
+      "step": 6105
+    },
+    {
+      "epoch": 0.8147851614625033,
+      "grad_norm": 1.0681724159465906,
+      "learning_rate": 1.7459685785716651e-06,
+      "loss": 1.5143,
+      "step": 6106
+    },
+    {
+      "epoch": 0.814918601547905,
+      "grad_norm": 0.9013869560225362,
+      "learning_rate": 1.7435294451140783e-06,
+      "loss": 1.5133,
+      "step": 6107
+    },
+    {
+      "epoch": 0.8150520416333067,
+      "grad_norm": 0.9616884493441265,
+      "learning_rate": 1.7410918538711875e-06,
+      "loss": 1.6119,
+      "step": 6108
+    },
+    {
+      "epoch": 0.8151854817187083,
+      "grad_norm": 0.9680636039329384,
+      "learning_rate": 1.738655805298307e-06,
+      "loss": 1.5284,
+      "step": 6109
+    },
+    {
+      "epoch": 0.8153189218041099,
+      "grad_norm": 0.9024891296925446,
+      "learning_rate": 1.7362212998504635e-06,
+      "loss": 1.5047,
+      "step": 6110
+    },
+    {
+      "epoch": 0.8154523618895116,
+      "grad_norm": 0.9523905191421255,
+      "learning_rate": 1.7337883379823917e-06,
+      "loss": 1.5871,
+      "step": 6111
+    },
+    {
+      "epoch": 0.8155858019749133,
+      "grad_norm": 0.9263795793949324,
+      "learning_rate": 1.731356920148547e-06,
+      "loss": 1.5285,
+      "step": 6112
+    },
+    {
+      "epoch": 0.8157192420603149,
+      "grad_norm": 0.9647766098487651,
+      "learning_rate": 1.7289270468030871e-06,
+      "loss": 1.5834,
+      "step": 6113
+    },
+    {
+      "epoch": 0.8158526821457166,
+      "grad_norm": 0.9197615859806225,
+      "learning_rate": 1.726498718399886e-06,
+      "loss": 1.5364,
+      "step": 6114
+    },
+    {
+      "epoch": 0.8159861222311182,
+      "grad_norm": 0.9626572376096142,
+      "learning_rate": 1.7240719353925262e-06,
+      "loss": 1.5519,
+      "step": 6115
+    },
+    {
+      "epoch": 0.8161195623165198,
+      "grad_norm": 0.9396787834195818,
+      "learning_rate": 1.7216466982343039e-06,
+      "loss": 1.5418,
+      "step": 6116
+    },
+    {
+      "epoch": 0.8162530024019216,
+      "grad_norm": 0.91235670010565,
+      "learning_rate": 1.7192230073782234e-06,
+      "loss": 1.477,
+      "step": 6117
+    },
+    {
+      "epoch": 0.8163864424873232,
+      "grad_norm": 0.9262321266478081,
+      "learning_rate": 1.7168008632770072e-06,
+      "loss": 1.5578,
+      "step": 6118
+    },
+    {
+      "epoch": 0.8165198825727249,
+      "grad_norm": 0.9400952842702311,
+      "learning_rate": 1.7143802663830821e-06,
+      "loss": 1.4842,
+      "step": 6119
+    },
+    {
+      "epoch": 0.8166533226581265,
+      "grad_norm": 0.9299282662667695,
+      "learning_rate": 1.7119612171485877e-06,
+      "loss": 1.5302,
+      "step": 6120
+    },
+    {
+      "epoch": 0.8167867627435281,
+      "grad_norm": 0.9932044315186542,
+      "learning_rate": 1.7095437160253725e-06,
+      "loss": 1.5168,
+      "step": 6121
+    },
+    {
+      "epoch": 0.8169202028289299,
+      "grad_norm": 0.9580569276859812,
+      "learning_rate": 1.707127763465004e-06,
+      "loss": 1.5769,
+      "step": 6122
+    },
+    {
+      "epoch": 0.8170536429143315,
+      "grad_norm": 1.102986018038682,
+      "learning_rate": 1.7047133599187515e-06,
+      "loss": 1.531,
+      "step": 6123
+    },
+    {
+      "epoch": 0.8171870829997331,
+      "grad_norm": 0.9193541580129797,
+      "learning_rate": 1.702300505837594e-06,
+      "loss": 1.5448,
+      "step": 6124
+    },
+    {
+      "epoch": 0.8173205230851348,
+      "grad_norm": 0.9456674504166714,
+      "learning_rate": 1.6998892016722334e-06,
+      "loss": 1.5154,
+      "step": 6125
+    },
+    {
+      "epoch": 0.8174539631705364,
+      "grad_norm": 1.3332667316333,
+      "learning_rate": 1.6974794478730683e-06,
+      "loss": 1.5605,
+      "step": 6126
+    },
+    {
+      "epoch": 0.817587403255938,
+      "grad_norm": 0.9558729747564527,
+      "learning_rate": 1.695071244890215e-06,
+      "loss": 1.5821,
+      "step": 6127
+    },
+    {
+      "epoch": 0.8177208433413398,
+      "grad_norm": 1.2620823710527826,
+      "learning_rate": 1.6926645931734964e-06,
+      "loss": 1.5201,
+      "step": 6128
+    },
+    {
+      "epoch": 0.8178542834267414,
+      "grad_norm": 0.9179178473951229,
+      "learning_rate": 1.6902594931724503e-06,
+      "loss": 1.5354,
+      "step": 6129
+    },
+    {
+      "epoch": 0.817987723512143,
+      "grad_norm": 0.9450627126995589,
+      "learning_rate": 1.687855945336322e-06,
+      "loss": 1.5345,
+      "step": 6130
+    },
+    {
+      "epoch": 0.8181211635975447,
+      "grad_norm": 0.9178343790581889,
+      "learning_rate": 1.6854539501140632e-06,
+      "loss": 1.55,
+      "step": 6131
+    },
+    {
+      "epoch": 0.8182546036829463,
+      "grad_norm": 0.9964264932180854,
+      "learning_rate": 1.6830535079543432e-06,
+      "loss": 1.5697,
+      "step": 6132
+    },
+    {
+      "epoch": 0.8183880437683481,
+      "grad_norm": 0.9890708742005667,
+      "learning_rate": 1.6806546193055374e-06,
+      "loss": 1.5219,
+      "step": 6133
+    },
+    {
+      "epoch": 0.8185214838537497,
+      "grad_norm": 0.9229456288469,
+      "learning_rate": 1.6782572846157285e-06,
+      "loss": 1.5209,
+      "step": 6134
+    },
+    {
+      "epoch": 0.8186549239391513,
+      "grad_norm": 0.9130464886045846,
+      "learning_rate": 1.6758615043327097e-06,
+      "loss": 1.4973,
+      "step": 6135
+    },
+    {
+      "epoch": 0.818788364024553,
+      "grad_norm": 0.953168022773084,
+      "learning_rate": 1.6734672789039907e-06,
+      "loss": 1.5345,
+      "step": 6136
+    },
+    {
+      "epoch": 0.8189218041099546,
+      "grad_norm": 0.9178014077533045,
+      "learning_rate": 1.6710746087767826e-06,
+      "loss": 1.5069,
+      "step": 6137
+    },
+    {
+      "epoch": 0.8190552441953562,
+      "grad_norm": 0.9433818817785761,
+      "learning_rate": 1.668683494398008e-06,
+      "loss": 1.5628,
+      "step": 6138
+    },
+    {
+      "epoch": 0.819188684280758,
+      "grad_norm": 0.955567446026117,
+      "learning_rate": 1.6662939362143028e-06,
+      "loss": 1.5812,
+      "step": 6139
+    },
+    {
+      "epoch": 0.8193221243661596,
+      "grad_norm": 1.0316675276310674,
+      "learning_rate": 1.6639059346720065e-06,
+      "loss": 1.5408,
+      "step": 6140
+    },
+    {
+      "epoch": 0.8194555644515612,
+      "grad_norm": 1.0994368830180041,
+      "learning_rate": 1.6615194902171728e-06,
+      "loss": 1.5808,
+      "step": 6141
+    },
+    {
+      "epoch": 0.8195890045369629,
+      "grad_norm": 0.9318559913000551,
+      "learning_rate": 1.659134603295558e-06,
+      "loss": 1.5227,
+      "step": 6142
+    },
+    {
+      "epoch": 0.8197224446223645,
+      "grad_norm": 0.9700753022715083,
+      "learning_rate": 1.6567512743526383e-06,
+      "loss": 1.5809,
+      "step": 6143
+    },
+    {
+      "epoch": 0.8198558847077662,
+      "grad_norm": 0.9487897185182752,
+      "learning_rate": 1.6543695038335882e-06,
+      "loss": 1.5474,
+      "step": 6144
+    },
+    {
+      "epoch": 0.8199893247931679,
+      "grad_norm": 0.9629831813661353,
+      "learning_rate": 1.651989292183297e-06,
+      "loss": 1.5777,
+      "step": 6145
+    },
+    {
+      "epoch": 0.8201227648785695,
+      "grad_norm": 1.057479149875002,
+      "learning_rate": 1.6496106398463574e-06,
+      "loss": 1.5347,
+      "step": 6146
+    },
+    {
+      "epoch": 0.8202562049639712,
+      "grad_norm": 0.933470698014775,
+      "learning_rate": 1.6472335472670798e-06,
+      "loss": 1.5632,
+      "step": 6147
+    },
+    {
+      "epoch": 0.8203896450493728,
+      "grad_norm": 1.5917968282853159,
+      "learning_rate": 1.6448580148894755e-06,
+      "loss": 1.5685,
+      "step": 6148
+    },
+    {
+      "epoch": 0.8205230851347745,
+      "grad_norm": 0.9576309077077291,
+      "learning_rate": 1.6424840431572652e-06,
+      "loss": 1.5568,
+      "step": 6149
+    },
+    {
+      "epoch": 0.8206565252201762,
+      "grad_norm": 1.0240140958776593,
+      "learning_rate": 1.6401116325138843e-06,
+      "loss": 1.5005,
+      "step": 6150
+    },
+    {
+      "epoch": 0.8207899653055778,
+      "grad_norm": 1.042576719135628,
+      "learning_rate": 1.6377407834024694e-06,
+      "loss": 1.5537,
+      "step": 6151
+    },
+    {
+      "epoch": 0.8209234053909794,
+      "grad_norm": 0.9365442898192005,
+      "learning_rate": 1.6353714962658684e-06,
+      "loss": 1.5258,
+      "step": 6152
+    },
+    {
+      "epoch": 0.8210568454763811,
+      "grad_norm": 6.035336114616692,
+      "learning_rate": 1.6330037715466373e-06,
+      "loss": 1.5402,
+      "step": 6153
+    },
+    {
+      "epoch": 0.8211902855617828,
+      "grad_norm": 1.2568273540776527,
+      "learning_rate": 1.630637609687037e-06,
+      "loss": 1.5612,
+      "step": 6154
+    },
+    {
+      "epoch": 0.8213237256471844,
+      "grad_norm": 0.9408521308120555,
+      "learning_rate": 1.6282730111290446e-06,
+      "loss": 1.5436,
+      "step": 6155
+    },
+    {
+      "epoch": 0.8214571657325861,
+      "grad_norm": 0.9116594461537203,
+      "learning_rate": 1.6259099763143383e-06,
+      "loss": 1.532,
+      "step": 6156
+    },
+    {
+      "epoch": 0.8215906058179877,
+      "grad_norm": 0.9640703464570622,
+      "learning_rate": 1.623548505684306e-06,
+      "loss": 1.5475,
+      "step": 6157
+    },
+    {
+      "epoch": 0.8217240459033894,
+      "grad_norm": 0.9628540017226783,
+      "learning_rate": 1.6211885996800426e-06,
+      "loss": 1.5435,
+      "step": 6158
+    },
+    {
+      "epoch": 0.821857485988791,
+      "grad_norm": 1.0175932870044642,
+      "learning_rate": 1.6188302587423532e-06,
+      "loss": 1.5133,
+      "step": 6159
+    },
+    {
+      "epoch": 0.8219909260741927,
+      "grad_norm": 0.9328929841123439,
+      "learning_rate": 1.6164734833117458e-06,
+      "loss": 1.4553,
+      "step": 6160
+    },
+    {
+      "epoch": 0.8221243661595944,
+      "grad_norm": 0.9326921991444765,
+      "learning_rate": 1.6141182738284444e-06,
+      "loss": 1.58,
+      "step": 6161
+    },
+    {
+      "epoch": 0.822257806244996,
+      "grad_norm": 0.9198207496240384,
+      "learning_rate": 1.611764630732372e-06,
+      "loss": 1.5411,
+      "step": 6162
+    },
+    {
+      "epoch": 0.8223912463303976,
+      "grad_norm": 0.9240771169421748,
+      "learning_rate": 1.60941255446316e-06,
+      "loss": 1.5323,
+      "step": 6163
+    },
+    {
+      "epoch": 0.8225246864157993,
+      "grad_norm": 0.9435650366109184,
+      "learning_rate": 1.607062045460156e-06,
+      "loss": 1.5653,
+      "step": 6164
+    },
+    {
+      "epoch": 0.822658126501201,
+      "grad_norm": 1.0152052475294857,
+      "learning_rate": 1.6047131041624041e-06,
+      "loss": 1.5638,
+      "step": 6165
+    },
+    {
+      "epoch": 0.8227915665866026,
+      "grad_norm": 1.0524123141586792,
+      "learning_rate": 1.6023657310086605e-06,
+      "loss": 1.5782,
+      "step": 6166
+    },
+    {
+      "epoch": 0.8229250066720043,
+      "grad_norm": 0.9311840684459297,
+      "learning_rate": 1.600019926437385e-06,
+      "loss": 1.5414,
+      "step": 6167
+    },
+    {
+      "epoch": 0.8230584467574059,
+      "grad_norm": 0.9019457165743004,
+      "learning_rate": 1.597675690886753e-06,
+      "loss": 1.5351,
+      "step": 6168
+    },
+    {
+      "epoch": 0.8231918868428075,
+      "grad_norm": 0.923450760055443,
+      "learning_rate": 1.5953330247946375e-06,
+      "loss": 1.5313,
+      "step": 6169
+    },
+    {
+      "epoch": 0.8233253269282093,
+      "grad_norm": 1.2831560841203558,
+      "learning_rate": 1.5929919285986195e-06,
+      "loss": 1.5338,
+      "step": 6170
+    },
+    {
+      "epoch": 0.8234587670136109,
+      "grad_norm": 0.9339060300388902,
+      "learning_rate": 1.5906524027359948e-06,
+      "loss": 1.5511,
+      "step": 6171
+    },
+    {
+      "epoch": 0.8235922070990126,
+      "grad_norm": 0.935595003361667,
+      "learning_rate": 1.5883144476437572e-06,
+      "loss": 1.5072,
+      "step": 6172
+    },
+    {
+      "epoch": 0.8237256471844142,
+      "grad_norm": 0.9158207412122499,
+      "learning_rate": 1.5859780637586098e-06,
+      "loss": 1.5835,
+      "step": 6173
+    },
+    {
+      "epoch": 0.8238590872698158,
+      "grad_norm": 1.050328096524351,
+      "learning_rate": 1.5836432515169608e-06,
+      "loss": 1.5829,
+      "step": 6174
+    },
+    {
+      "epoch": 0.8239925273552176,
+      "grad_norm": 1.0263704705140466,
+      "learning_rate": 1.5813100113549307e-06,
+      "loss": 1.5493,
+      "step": 6175
+    },
+    {
+      "epoch": 0.8241259674406192,
+      "grad_norm": 0.9214587200849176,
+      "learning_rate": 1.5789783437083406e-06,
+      "loss": 1.5124,
+      "step": 6176
+    },
+    {
+      "epoch": 0.8242594075260208,
+      "grad_norm": 0.9624753786839196,
+      "learning_rate": 1.5766482490127176e-06,
+      "loss": 1.5424,
+      "step": 6177
+    },
+    {
+      "epoch": 0.8243928476114225,
+      "grad_norm": 0.9397643380771115,
+      "learning_rate": 1.5743197277032974e-06,
+      "loss": 1.5551,
+      "step": 6178
+    },
+    {
+      "epoch": 0.8245262876968241,
+      "grad_norm": 1.0598447563742128,
+      "learning_rate": 1.5719927802150236e-06,
+      "loss": 1.5502,
+      "step": 6179
+    },
+    {
+      "epoch": 0.8246597277822257,
+      "grad_norm": 1.3413789438510157,
+      "learning_rate": 1.5696674069825425e-06,
+      "loss": 1.5785,
+      "step": 6180
+    },
+    {
+      "epoch": 0.8247931678676275,
+      "grad_norm": 1.0362130690683422,
+      "learning_rate": 1.5673436084402039e-06,
+      "loss": 1.5472,
+      "step": 6181
+    },
+    {
+      "epoch": 0.8249266079530291,
+      "grad_norm": 1.190451521816014,
+      "learning_rate": 1.565021385022073e-06,
+      "loss": 1.4974,
+      "step": 6182
+    },
+    {
+      "epoch": 0.8250600480384307,
+      "grad_norm": 1.0064225238687168,
+      "learning_rate": 1.5627007371619107e-06,
+      "loss": 1.5868,
+      "step": 6183
+    },
+    {
+      "epoch": 0.8251934881238324,
+      "grad_norm": 0.9183532599911997,
+      "learning_rate": 1.560381665293189e-06,
+      "loss": 1.5492,
+      "step": 6184
+    },
+    {
+      "epoch": 0.825326928209234,
+      "grad_norm": 0.9451229117253995,
+      "learning_rate": 1.5580641698490805e-06,
+      "loss": 1.5507,
+      "step": 6185
+    },
+    {
+      "epoch": 0.8254603682946358,
+      "grad_norm": 0.9256676754994859,
+      "learning_rate": 1.5557482512624733e-06,
+      "loss": 1.5438,
+      "step": 6186
+    },
+    {
+      "epoch": 0.8255938083800374,
+      "grad_norm": 1.0241913570296761,
+      "learning_rate": 1.5534339099659512e-06,
+      "loss": 1.5336,
+      "step": 6187
+    },
+    {
+      "epoch": 0.825727248465439,
+      "grad_norm": 0.9648855141551541,
+      "learning_rate": 1.551121146391804e-06,
+      "loss": 1.5565,
+      "step": 6188
+    },
+    {
+      "epoch": 0.8258606885508407,
+      "grad_norm": 0.9417547056164751,
+      "learning_rate": 1.5488099609720353e-06,
+      "loss": 1.5198,
+      "step": 6189
+    },
+    {
+      "epoch": 0.8259941286362423,
+      "grad_norm": 1.019075853789808,
+      "learning_rate": 1.546500354138346e-06,
+      "loss": 1.5174,
+      "step": 6190
+    },
+    {
+      "epoch": 0.826127568721644,
+      "grad_norm": 0.950534403894064,
+      "learning_rate": 1.5441923263221426e-06,
+      "loss": 1.5386,
+      "step": 6191
+    },
+    {
+      "epoch": 0.8262610088070457,
+      "grad_norm": 0.9522247872246755,
+      "learning_rate": 1.5418858779545387e-06,
+      "loss": 1.5558,
+      "step": 6192
+    },
+    {
+      "epoch": 0.8263944488924473,
+      "grad_norm": 1.14220578778169,
+      "learning_rate": 1.5395810094663544e-06,
+      "loss": 1.5794,
+      "step": 6193
+    },
+    {
+      "epoch": 0.8265278889778489,
+      "grad_norm": 0.9516184349282555,
+      "learning_rate": 1.537277721288113e-06,
+      "loss": 1.5279,
+      "step": 6194
+    },
+    {
+      "epoch": 0.8266613290632506,
+      "grad_norm": 0.9428708938228508,
+      "learning_rate": 1.5349760138500414e-06,
+      "loss": 1.5303,
+      "step": 6195
+    },
+    {
+      "epoch": 0.8267947691486522,
+      "grad_norm": 0.906052569291839,
+      "learning_rate": 1.5326758875820724e-06,
+      "loss": 1.535,
+      "step": 6196
+    },
+    {
+      "epoch": 0.8269282092340539,
+      "grad_norm": 0.9243034140621643,
+      "learning_rate": 1.5303773429138414e-06,
+      "loss": 1.5628,
+      "step": 6197
+    },
+    {
+      "epoch": 0.8270616493194556,
+      "grad_norm": 0.9422165125262247,
+      "learning_rate": 1.5280803802746947e-06,
+      "loss": 1.5517,
+      "step": 6198
+    },
+    {
+      "epoch": 0.8271950894048572,
+      "grad_norm": 1.0325744245704787,
+      "learning_rate": 1.5257850000936768e-06,
+      "loss": 1.5963,
+      "step": 6199
+    },
+    {
+      "epoch": 0.8273285294902589,
+      "grad_norm": 1.1261838747600312,
+      "learning_rate": 1.5234912027995374e-06,
+      "loss": 1.5573,
+      "step": 6200
+    },
+    {
+      "epoch": 0.8274619695756605,
+      "grad_norm": 1.0318114988538098,
+      "learning_rate": 1.5211989888207335e-06,
+      "loss": 1.5389,
+      "step": 6201
+    },
+    {
+      "epoch": 0.8275954096610622,
+      "grad_norm": 0.9575726653864882,
+      "learning_rate": 1.5189083585854213e-06,
+      "loss": 1.5731,
+      "step": 6202
+    },
+    {
+      "epoch": 0.8277288497464639,
+      "grad_norm": 0.9495250274991178,
+      "learning_rate": 1.516619312521469e-06,
+      "loss": 1.5819,
+      "step": 6203
+    },
+    {
+      "epoch": 0.8278622898318655,
+      "grad_norm": 0.9001039404937866,
+      "learning_rate": 1.514331851056442e-06,
+      "loss": 1.5126,
+      "step": 6204
+    },
+    {
+      "epoch": 0.8279957299172671,
+      "grad_norm": 0.9596548897156311,
+      "learning_rate": 1.512045974617612e-06,
+      "loss": 1.5236,
+      "step": 6205
+    },
+    {
+      "epoch": 0.8281291700026688,
+      "grad_norm": 0.948592313867705,
+      "learning_rate": 1.5097616836319528e-06,
+      "loss": 1.5963,
+      "step": 6206
+    },
+    {
+      "epoch": 0.8282626100880705,
+      "grad_norm": 0.9678395783216994,
+      "learning_rate": 1.507478978526149e-06,
+      "loss": 1.5601,
+      "step": 6207
+    },
+    {
+      "epoch": 0.8283960501734721,
+      "grad_norm": 0.9282502916058483,
+      "learning_rate": 1.5051978597265814e-06,
+      "loss": 1.5221,
+      "step": 6208
+    },
+    {
+      "epoch": 0.8285294902588738,
+      "grad_norm": 1.1106582385865,
+      "learning_rate": 1.502918327659335e-06,
+      "loss": 1.5289,
+      "step": 6209
+    },
+    {
+      "epoch": 0.8286629303442754,
+      "grad_norm": 1.1265979346871835,
+      "learning_rate": 1.5006403827502014e-06,
+      "loss": 1.5341,
+      "step": 6210
+    },
+    {
+      "epoch": 0.828796370429677,
+      "grad_norm": 1.035765628780522,
+      "learning_rate": 1.4983640254246767e-06,
+      "loss": 1.5652,
+      "step": 6211
+    },
+    {
+      "epoch": 0.8289298105150787,
+      "grad_norm": 0.9611613457634671,
+      "learning_rate": 1.4960892561079577e-06,
+      "loss": 1.5555,
+      "step": 6212
+    },
+    {
+      "epoch": 0.8290632506004804,
+      "grad_norm": 0.9101115150391339,
+      "learning_rate": 1.4938160752249431e-06,
+      "loss": 1.5279,
+      "step": 6213
+    },
+    {
+      "epoch": 0.8291966906858821,
+      "grad_norm": 0.9444030552766006,
+      "learning_rate": 1.4915444832002413e-06,
+      "loss": 1.6253,
+      "step": 6214
+    },
+    {
+      "epoch": 0.8293301307712837,
+      "grad_norm": 0.9679370076321139,
+      "learning_rate": 1.4892744804581572e-06,
+      "loss": 1.5131,
+      "step": 6215
+    },
+    {
+      "epoch": 0.8294635708566853,
+      "grad_norm": 0.9687392610701684,
+      "learning_rate": 1.4870060674227016e-06,
+      "loss": 1.5411,
+      "step": 6216
+    },
+    {
+      "epoch": 0.829597010942087,
+      "grad_norm": 0.9324868882302184,
+      "learning_rate": 1.4847392445175867e-06,
+      "loss": 1.5546,
+      "step": 6217
+    },
+    {
+      "epoch": 0.8297304510274887,
+      "grad_norm": 0.9659785539366846,
+      "learning_rate": 1.482474012166234e-06,
+      "loss": 1.5354,
+      "step": 6218
+    },
+    {
+      "epoch": 0.8298638911128903,
+      "grad_norm": 1.0403468201365482,
+      "learning_rate": 1.4802103707917591e-06,
+      "loss": 1.5384,
+      "step": 6219
+    },
+    {
+      "epoch": 0.829997331198292,
+      "grad_norm": 0.9235060298849219,
+      "learning_rate": 1.4779483208169832e-06,
+      "loss": 1.4958,
+      "step": 6220
+    },
+    {
+      "epoch": 0.8301307712836936,
+      "grad_norm": 0.9948663890358413,
+      "learning_rate": 1.4756878626644367e-06,
+      "loss": 1.5092,
+      "step": 6221
+    },
+    {
+      "epoch": 0.8302642113690952,
+      "grad_norm": 1.2438225582984503,
+      "learning_rate": 1.4734289967563442e-06,
+      "loss": 1.5525,
+      "step": 6222
+    },
+    {
+      "epoch": 0.830397651454497,
+      "grad_norm": 0.9394237729404794,
+      "learning_rate": 1.471171723514636e-06,
+      "loss": 1.5704,
+      "step": 6223
+    },
+    {
+      "epoch": 0.8305310915398986,
+      "grad_norm": 0.9220696013800785,
+      "learning_rate": 1.468916043360944e-06,
+      "loss": 1.5649,
+      "step": 6224
+    },
+    {
+      "epoch": 0.8306645316253003,
+      "grad_norm": 0.9661997734554147,
+      "learning_rate": 1.4666619567166074e-06,
+      "loss": 1.5551,
+      "step": 6225
+    },
+    {
+      "epoch": 0.8307979717107019,
+      "grad_norm": 0.9713988388879101,
+      "learning_rate": 1.4644094640026607e-06,
+      "loss": 1.4893,
+      "step": 6226
+    },
+    {
+      "epoch": 0.8309314117961035,
+      "grad_norm": 0.9362262546367968,
+      "learning_rate": 1.4621585656398429e-06,
+      "loss": 1.5044,
+      "step": 6227
+    },
+    {
+      "epoch": 0.8310648518815053,
+      "grad_norm": 1.070256981646457,
+      "learning_rate": 1.459909262048601e-06,
+      "loss": 1.5317,
+      "step": 6228
+    },
+    {
+      "epoch": 0.8311982919669069,
+      "grad_norm": 0.9309802597099771,
+      "learning_rate": 1.4576615536490756e-06,
+      "loss": 1.5314,
+      "step": 6229
+    },
+    {
+      "epoch": 0.8313317320523085,
+      "grad_norm": 0.9534061214073728,
+      "learning_rate": 1.4554154408611142e-06,
+      "loss": 1.5432,
+      "step": 6230
+    },
+    {
+      "epoch": 0.8314651721377102,
+      "grad_norm": 0.9260408668974777,
+      "learning_rate": 1.4531709241042624e-06,
+      "loss": 1.4923,
+      "step": 6231
+    },
+    {
+      "epoch": 0.8315986122231118,
+      "grad_norm": 1.0984291256861187,
+      "learning_rate": 1.4509280037977746e-06,
+      "loss": 1.5121,
+      "step": 6232
+    },
+    {
+      "epoch": 0.8317320523085134,
+      "grad_norm": 1.080816725382994,
+      "learning_rate": 1.4486866803606003e-06,
+      "loss": 1.5719,
+      "step": 6233
+    },
+    {
+      "epoch": 0.8318654923939152,
+      "grad_norm": 0.9403182913647171,
+      "learning_rate": 1.4464469542113924e-06,
+      "loss": 1.5661,
+      "step": 6234
+    },
+    {
+      "epoch": 0.8319989324793168,
+      "grad_norm": 0.9315844916031217,
+      "learning_rate": 1.4442088257685105e-06,
+      "loss": 1.5355,
+      "step": 6235
+    },
+    {
+      "epoch": 0.8321323725647184,
+      "grad_norm": 0.9523132017213204,
+      "learning_rate": 1.4419722954500071e-06,
+      "loss": 1.6324,
+      "step": 6236
+    },
+    {
+      "epoch": 0.8322658126501201,
+      "grad_norm": 0.9122663743215488,
+      "learning_rate": 1.4397373636736435e-06,
+      "loss": 1.51,
+      "step": 6237
+    },
+    {
+      "epoch": 0.8323992527355217,
+      "grad_norm": 1.2517381014691573,
+      "learning_rate": 1.4375040308568765e-06,
+      "loss": 1.6326,
+      "step": 6238
+    },
+    {
+      "epoch": 0.8325326928209235,
+      "grad_norm": 1.0826573600314662,
+      "learning_rate": 1.4352722974168675e-06,
+      "loss": 1.5585,
+      "step": 6239
+    },
+    {
+      "epoch": 0.8326661329063251,
+      "grad_norm": 0.9348054260521418,
+      "learning_rate": 1.4330421637704828e-06,
+      "loss": 1.4989,
+      "step": 6240
+    },
+    {
+      "epoch": 0.8327995729917267,
+      "grad_norm": 0.9451331002738911,
+      "learning_rate": 1.4308136303342835e-06,
+      "loss": 1.5498,
+      "step": 6241
+    },
+    {
+      "epoch": 0.8329330130771284,
+      "grad_norm": 0.9455134123416408,
+      "learning_rate": 1.4285866975245333e-06,
+      "loss": 1.4996,
+      "step": 6242
+    },
+    {
+      "epoch": 0.83306645316253,
+      "grad_norm": 0.9642660018609069,
+      "learning_rate": 1.4263613657571995e-06,
+      "loss": 1.5439,
+      "step": 6243
+    },
+    {
+      "epoch": 0.8331998932479316,
+      "grad_norm": 1.2808082607240958,
+      "learning_rate": 1.4241376354479475e-06,
+      "loss": 1.5403,
+      "step": 6244
+    },
+    {
+      "epoch": 0.8333333333333334,
+      "grad_norm": 1.8043818463976273,
+      "learning_rate": 1.4219155070121438e-06,
+      "loss": 1.5046,
+      "step": 6245
+    },
+    {
+      "epoch": 0.833466773418735,
+      "grad_norm": 0.949954061324863,
+      "learning_rate": 1.4196949808648597e-06,
+      "loss": 1.5568,
+      "step": 6246
+    },
+    {
+      "epoch": 0.8336002135041366,
+      "grad_norm": 0.9241430802845367,
+      "learning_rate": 1.4174760574208634e-06,
+      "loss": 1.5384,
+      "step": 6247
+    },
+    {
+      "epoch": 0.8337336535895383,
+      "grad_norm": 0.9247044084518099,
+      "learning_rate": 1.4152587370946235e-06,
+      "loss": 1.5395,
+      "step": 6248
+    },
+    {
+      "epoch": 0.8338670936749399,
+      "grad_norm": 0.9470481656442147,
+      "learning_rate": 1.4130430203003088e-06,
+      "loss": 1.5524,
+      "step": 6249
+    },
+    {
+      "epoch": 0.8340005337603416,
+      "grad_norm": 0.95349918298535,
+      "learning_rate": 1.4108289074517934e-06,
+      "loss": 1.592,
+      "step": 6250
+    },
+    {
+      "epoch": 0.8341339738457433,
+      "grad_norm": 0.930859710005307,
+      "learning_rate": 1.4086163989626467e-06,
+      "loss": 1.5467,
+      "step": 6251
+    },
+    {
+      "epoch": 0.8342674139311449,
+      "grad_norm": 0.9827202874058659,
+      "learning_rate": 1.4064054952461382e-06,
+      "loss": 1.5415,
+      "step": 6252
+    },
+    {
+      "epoch": 0.8344008540165466,
+      "grad_norm": 1.0609242939930341,
+      "learning_rate": 1.404196196715244e-06,
+      "loss": 1.5581,
+      "step": 6253
+    },
+    {
+      "epoch": 0.8345342941019482,
+      "grad_norm": 0.9879789904417466,
+      "learning_rate": 1.401988503782633e-06,
+      "loss": 1.541,
+      "step": 6254
+    },
+    {
+      "epoch": 0.8346677341873499,
+      "grad_norm": 0.9589864610692524,
+      "learning_rate": 1.3997824168606777e-06,
+      "loss": 1.5305,
+      "step": 6255
+    },
+    {
+      "epoch": 0.8348011742727516,
+      "grad_norm": 0.929209062077639,
+      "learning_rate": 1.397577936361446e-06,
+      "loss": 1.5345,
+      "step": 6256
+    },
+    {
+      "epoch": 0.8349346143581532,
+      "grad_norm": 0.9515351783274919,
+      "learning_rate": 1.3953750626967178e-06,
+      "loss": 1.5299,
+      "step": 6257
+    },
+    {
+      "epoch": 0.8350680544435548,
+      "grad_norm": 1.000540782983064,
+      "learning_rate": 1.39317379627796e-06,
+      "loss": 1.5766,
+      "step": 6258
+    },
+    {
+      "epoch": 0.8352014945289565,
+      "grad_norm": 1.0567054179998463,
+      "learning_rate": 1.3909741375163422e-06,
+      "loss": 1.5723,
+      "step": 6259
+    },
+    {
+      "epoch": 0.8353349346143581,
+      "grad_norm": 0.9549905329939615,
+      "learning_rate": 1.3887760868227396e-06,
+      "loss": 1.5789,
+      "step": 6260
+    },
+    {
+      "epoch": 0.8354683746997598,
+      "grad_norm": 0.9872576585804954,
+      "learning_rate": 1.386579644607723e-06,
+      "loss": 1.5734,
+      "step": 6261
+    },
+    {
+      "epoch": 0.8356018147851615,
+      "grad_norm": 0.9367144021734644,
+      "learning_rate": 1.3843848112815594e-06,
+      "loss": 1.5708,
+      "step": 6262
+    },
+    {
+      "epoch": 0.8357352548705631,
+      "grad_norm": 0.9513673423313426,
+      "learning_rate": 1.3821915872542202e-06,
+      "loss": 1.5313,
+      "step": 6263
+    },
+    {
+      "epoch": 0.8358686949559647,
+      "grad_norm": 0.9567825688256574,
+      "learning_rate": 1.3799999729353764e-06,
+      "loss": 1.5118,
+      "step": 6264
+    },
+    {
+      "epoch": 0.8360021350413664,
+      "grad_norm": 1.0863657842635757,
+      "learning_rate": 1.3778099687343948e-06,
+      "loss": 1.5669,
+      "step": 6265
+    },
+    {
+      "epoch": 0.8361355751267681,
+      "grad_norm": 0.9751182715087741,
+      "learning_rate": 1.375621575060343e-06,
+      "loss": 1.5848,
+      "step": 6266
+    },
+    {
+      "epoch": 0.8362690152121698,
+      "grad_norm": 0.9552963864700986,
+      "learning_rate": 1.3734347923219893e-06,
+      "loss": 1.5594,
+      "step": 6267
+    },
+    {
+      "epoch": 0.8364024552975714,
+      "grad_norm": 0.9334563882233021,
+      "learning_rate": 1.3712496209278004e-06,
+      "loss": 1.5085,
+      "step": 6268
+    },
+    {
+      "epoch": 0.836535895382973,
+      "grad_norm": 0.9418447706556196,
+      "learning_rate": 1.3690660612859397e-06,
+      "loss": 1.484,
+      "step": 6269
+    },
+    {
+      "epoch": 0.8366693354683747,
+      "grad_norm": 1.2452469016859455,
+      "learning_rate": 1.36688411380427e-06,
+      "loss": 1.5939,
+      "step": 6270
+    },
+    {
+      "epoch": 0.8368027755537764,
+      "grad_norm": 0.9630484251295978,
+      "learning_rate": 1.3647037788903582e-06,
+      "loss": 1.5588,
+      "step": 6271
+    },
+    {
+      "epoch": 0.836936215639178,
+      "grad_norm": 0.9163101746800517,
+      "learning_rate": 1.3625250569514636e-06,
+      "loss": 1.543,
+      "step": 6272
+    },
+    {
+      "epoch": 0.8370696557245797,
+      "grad_norm": 1.0159540192644183,
+      "learning_rate": 1.3603479483945482e-06,
+      "loss": 1.5317,
+      "step": 6273
+    },
+    {
+      "epoch": 0.8372030958099813,
+      "grad_norm": 0.9495771661136421,
+      "learning_rate": 1.3581724536262664e-06,
+      "loss": 1.5721,
+      "step": 6274
+    },
+    {
+      "epoch": 0.8373365358953829,
+      "grad_norm": 0.9259908173056178,
+      "learning_rate": 1.3559985730529824e-06,
+      "loss": 1.5375,
+      "step": 6275
+    },
+    {
+      "epoch": 0.8374699759807847,
+      "grad_norm": 1.0746035612986713,
+      "learning_rate": 1.353826307080749e-06,
+      "loss": 1.545,
+      "step": 6276
+    },
+    {
+      "epoch": 0.8376034160661863,
+      "grad_norm": 1.206161587602088,
+      "learning_rate": 1.3516556561153182e-06,
+      "loss": 1.4896,
+      "step": 6277
+    },
+    {
+      "epoch": 0.8377368561515879,
+      "grad_norm": 0.9394100024042479,
+      "learning_rate": 1.3494866205621492e-06,
+      "loss": 1.569,
+      "step": 6278
+    },
+    {
+      "epoch": 0.8378702962369896,
+      "grad_norm": 0.9298549718077799,
+      "learning_rate": 1.347319200826389e-06,
+      "loss": 1.5184,
+      "step": 6279
+    },
+    {
+      "epoch": 0.8380037363223912,
+      "grad_norm": 0.9364031759470136,
+      "learning_rate": 1.3451533973128873e-06,
+      "loss": 1.5611,
+      "step": 6280
+    },
+    {
+      "epoch": 0.838137176407793,
+      "grad_norm": 0.9438099018882347,
+      "learning_rate": 1.3429892104261922e-06,
+      "loss": 1.5884,
+      "step": 6281
+    },
+    {
+      "epoch": 0.8382706164931946,
+      "grad_norm": 1.0963915983521988,
+      "learning_rate": 1.3408266405705462e-06,
+      "loss": 1.5056,
+      "step": 6282
+    },
+    {
+      "epoch": 0.8384040565785962,
+      "grad_norm": 1.0553961263791714,
+      "learning_rate": 1.3386656881498982e-06,
+      "loss": 1.5072,
+      "step": 6283
+    },
+    {
+      "epoch": 0.8385374966639979,
+      "grad_norm": 0.9313093194946619,
+      "learning_rate": 1.3365063535678868e-06,
+      "loss": 1.5504,
+      "step": 6284
+    },
+    {
+      "epoch": 0.8386709367493995,
+      "grad_norm": 0.9472337824960516,
+      "learning_rate": 1.3343486372278502e-06,
+      "loss": 1.5925,
+      "step": 6285
+    },
+    {
+      "epoch": 0.8388043768348011,
+      "grad_norm": 0.9563239409348798,
+      "learning_rate": 1.3321925395328261e-06,
+      "loss": 1.4905,
+      "step": 6286
+    },
+    {
+      "epoch": 0.8389378169202029,
+      "grad_norm": 0.9193497426197157,
+      "learning_rate": 1.33003806088555e-06,
+      "loss": 1.5251,
+      "step": 6287
+    },
+    {
+      "epoch": 0.8390712570056045,
+      "grad_norm": 0.9463362147811565,
+      "learning_rate": 1.3278852016884491e-06,
+      "loss": 1.5755,
+      "step": 6288
+    },
+    {
+      "epoch": 0.8392046970910061,
+      "grad_norm": 1.0465969350369784,
+      "learning_rate": 1.3257339623436606e-06,
+      "loss": 1.558,
+      "step": 6289
+    },
+    {
+      "epoch": 0.8393381371764078,
+      "grad_norm": 0.9078117750928425,
+      "learning_rate": 1.323584343253007e-06,
+      "loss": 1.5181,
+      "step": 6290
+    },
+    {
+      "epoch": 0.8394715772618094,
+      "grad_norm": 0.9471706272387931,
+      "learning_rate": 1.3214363448180111e-06,
+      "loss": 1.5852,
+      "step": 6291
+    },
+    {
+      "epoch": 0.839605017347211,
+      "grad_norm": 0.9438263601613096,
+      "learning_rate": 1.3192899674398985e-06,
+      "loss": 1.5328,
+      "step": 6292
+    },
+    {
+      "epoch": 0.8397384574326128,
+      "grad_norm": 12.634716581956164,
+      "learning_rate": 1.317145211519587e-06,
+      "loss": 1.551,
+      "step": 6293
+    },
+    {
+      "epoch": 0.8398718975180144,
+      "grad_norm": 0.947237397048608,
+      "learning_rate": 1.315002077457692e-06,
+      "loss": 1.523,
+      "step": 6294
+    },
+    {
+      "epoch": 0.8400053376034161,
+      "grad_norm": 0.9268524675831373,
+      "learning_rate": 1.3128605656545245e-06,
+      "loss": 1.5293,
+      "step": 6295
+    },
+    {
+      "epoch": 0.8401387776888177,
+      "grad_norm": 0.9206635334147892,
+      "learning_rate": 1.3107206765100987e-06,
+      "loss": 1.5504,
+      "step": 6296
+    },
+    {
+      "epoch": 0.8402722177742193,
+      "grad_norm": 0.9437804573296683,
+      "learning_rate": 1.3085824104241185e-06,
+      "loss": 1.5366,
+      "step": 6297
+    },
+    {
+      "epoch": 0.8404056578596211,
+      "grad_norm": 0.9532930752867279,
+      "learning_rate": 1.3064457677959874e-06,
+      "loss": 1.5794,
+      "step": 6298
+    },
+    {
+      "epoch": 0.8405390979450227,
+      "grad_norm": 0.9331423826329348,
+      "learning_rate": 1.3043107490248086e-06,
+      "loss": 1.5141,
+      "step": 6299
+    },
+    {
+      "epoch": 0.8406725380304243,
+      "grad_norm": 0.9831369699762285,
+      "learning_rate": 1.3021773545093775e-06,
+      "loss": 1.5939,
+      "step": 6300
+    },
+    {
+      "epoch": 0.840805978115826,
+      "grad_norm": 0.9671730377179253,
+      "learning_rate": 1.3000455846481886e-06,
+      "loss": 1.5556,
+      "step": 6301
+    },
+    {
+      "epoch": 0.8409394182012276,
+      "grad_norm": 0.9400385680698677,
+      "learning_rate": 1.29791543983943e-06,
+      "loss": 1.5407,
+      "step": 6302
+    },
+    {
+      "epoch": 0.8410728582866293,
+      "grad_norm": 0.9409729894045126,
+      "learning_rate": 1.2957869204809925e-06,
+      "loss": 1.5392,
+      "step": 6303
+    },
+    {
+      "epoch": 0.841206298372031,
+      "grad_norm": 0.94097206563248,
+      "learning_rate": 1.2936600269704559e-06,
+      "loss": 1.5605,
+      "step": 6304
+    },
+    {
+      "epoch": 0.8413397384574326,
+      "grad_norm": 0.9949943289521407,
+      "learning_rate": 1.291534759705102e-06,
+      "loss": 1.6005,
+      "step": 6305
+    },
+    {
+      "epoch": 0.8414731785428343,
+      "grad_norm": 0.9355636029437794,
+      "learning_rate": 1.2894111190819025e-06,
+      "loss": 1.5574,
+      "step": 6306
+    },
+    {
+      "epoch": 0.8416066186282359,
+      "grad_norm": 0.9725584670762065,
+      "learning_rate": 1.2872891054975346e-06,
+      "loss": 1.5632,
+      "step": 6307
+    },
+    {
+      "epoch": 0.8417400587136376,
+      "grad_norm": 0.9010547164650611,
+      "learning_rate": 1.2851687193483642e-06,
+      "loss": 1.4956,
+      "step": 6308
+    },
+    {
+      "epoch": 0.8418734987990393,
+      "grad_norm": 0.975052957442709,
+      "learning_rate": 1.2830499610304526e-06,
+      "loss": 1.6136,
+      "step": 6309
+    },
+    {
+      "epoch": 0.8420069388844409,
+      "grad_norm": 0.9766365640836961,
+      "learning_rate": 1.280932830939564e-06,
+      "loss": 1.5266,
+      "step": 6310
+    },
+    {
+      "epoch": 0.8421403789698425,
+      "grad_norm": 0.9456532975205986,
+      "learning_rate": 1.2788173294711526e-06,
+      "loss": 1.5745,
+      "step": 6311
+    },
+    {
+      "epoch": 0.8422738190552442,
+      "grad_norm": 0.9410865383508794,
+      "learning_rate": 1.2767034570203685e-06,
+      "loss": 1.5397,
+      "step": 6312
+    },
+    {
+      "epoch": 0.8424072591406458,
+      "grad_norm": 0.9832507755872817,
+      "learning_rate": 1.2745912139820594e-06,
+      "loss": 1.589,
+      "step": 6313
+    },
+    {
+      "epoch": 0.8425406992260475,
+      "grad_norm": 0.9510871209768055,
+      "learning_rate": 1.2724806007507706e-06,
+      "loss": 1.5622,
+      "step": 6314
+    },
+    {
+      "epoch": 0.8426741393114492,
+      "grad_norm": 0.9418455606671381,
+      "learning_rate": 1.2703716177207393e-06,
+      "loss": 1.5386,
+      "step": 6315
+    },
+    {
+      "epoch": 0.8428075793968508,
+      "grad_norm": 0.9267882316614677,
+      "learning_rate": 1.2682642652858968e-06,
+      "loss": 1.5246,
+      "step": 6316
+    },
+    {
+      "epoch": 0.8429410194822524,
+      "grad_norm": 0.9241066105531126,
+      "learning_rate": 1.2661585438398771e-06,
+      "loss": 1.5255,
+      "step": 6317
+    },
+    {
+      "epoch": 0.8430744595676541,
+      "grad_norm": 0.9299217006327872,
+      "learning_rate": 1.2640544537760035e-06,
+      "loss": 1.5331,
+      "step": 6318
+    },
+    {
+      "epoch": 0.8432078996530558,
+      "grad_norm": 0.9460477011701305,
+      "learning_rate": 1.261951995487295e-06,
+      "loss": 1.5356,
+      "step": 6319
+    },
+    {
+      "epoch": 0.8433413397384575,
+      "grad_norm": 0.99568183474739,
+      "learning_rate": 1.259851169366465e-06,
+      "loss": 1.5396,
+      "step": 6320
+    },
+    {
+      "epoch": 0.8434747798238591,
+      "grad_norm": 0.9531319408928621,
+      "learning_rate": 1.2577519758059286e-06,
+      "loss": 1.5645,
+      "step": 6321
+    },
+    {
+      "epoch": 0.8436082199092607,
+      "grad_norm": 1.4786773551886767,
+      "learning_rate": 1.255654415197789e-06,
+      "loss": 1.5033,
+      "step": 6322
+    },
+    {
+      "epoch": 0.8437416599946624,
+      "grad_norm": 0.9837349009581106,
+      "learning_rate": 1.2535584879338469e-06,
+      "loss": 1.5878,
+      "step": 6323
+    },
+    {
+      "epoch": 0.8438751000800641,
+      "grad_norm": 0.9380148129468954,
+      "learning_rate": 1.2514641944055961e-06,
+      "loss": 1.5127,
+      "step": 6324
+    },
+    {
+      "epoch": 0.8440085401654657,
+      "grad_norm": 0.9800681388655729,
+      "learning_rate": 1.2493715350042267e-06,
+      "loss": 1.5177,
+      "step": 6325
+    },
+    {
+      "epoch": 0.8441419802508674,
+      "grad_norm": 0.9363139844260511,
+      "learning_rate": 1.2472805101206265e-06,
+      "loss": 1.5309,
+      "step": 6326
+    },
+    {
+      "epoch": 0.844275420336269,
+      "grad_norm": 0.9716477758023239,
+      "learning_rate": 1.2451911201453747e-06,
+      "loss": 1.5331,
+      "step": 6327
+    },
+    {
+      "epoch": 0.8444088604216706,
+      "grad_norm": 1.037355252448747,
+      "learning_rate": 1.243103365468743e-06,
+      "loss": 1.5556,
+      "step": 6328
+    },
+    {
+      "epoch": 0.8445423005070724,
+      "grad_norm": 1.127301529246472,
+      "learning_rate": 1.2410172464807024e-06,
+      "loss": 1.5456,
+      "step": 6329
+    },
+    {
+      "epoch": 0.844675740592474,
+      "grad_norm": 0.9421048807023837,
+      "learning_rate": 1.2389327635709136e-06,
+      "loss": 1.4901,
+      "step": 6330
+    },
+    {
+      "epoch": 0.8448091806778756,
+      "grad_norm": 1.1002964846418235,
+      "learning_rate": 1.2368499171287374e-06,
+      "loss": 1.6021,
+      "step": 6331
+    },
+    {
+      "epoch": 0.8449426207632773,
+      "grad_norm": 0.9387328752601601,
+      "learning_rate": 1.2347687075432246e-06,
+      "loss": 1.5047,
+      "step": 6332
+    },
+    {
+      "epoch": 0.8450760608486789,
+      "grad_norm": 0.9123192171247185,
+      "learning_rate": 1.2326891352031223e-06,
+      "loss": 1.5134,
+      "step": 6333
+    },
+    {
+      "epoch": 0.8452095009340806,
+      "grad_norm": 1.1014224330465954,
+      "learning_rate": 1.2306112004968662e-06,
+      "loss": 1.5769,
+      "step": 6334
+    },
+    {
+      "epoch": 0.8453429410194823,
+      "grad_norm": 0.9516250282510863,
+      "learning_rate": 1.2285349038125981e-06,
+      "loss": 1.585,
+      "step": 6335
+    },
+    {
+      "epoch": 0.8454763811048839,
+      "grad_norm": 0.9572784838289615,
+      "learning_rate": 1.2264602455381424e-06,
+      "loss": 1.5436,
+      "step": 6336
+    },
+    {
+      "epoch": 0.8456098211902856,
+      "grad_norm": 0.927276956609323,
+      "learning_rate": 1.2243872260610223e-06,
+      "loss": 1.4966,
+      "step": 6337
+    },
+    {
+      "epoch": 0.8457432612756872,
+      "grad_norm": 0.9439895084993872,
+      "learning_rate": 1.2223158457684526e-06,
+      "loss": 1.5114,
+      "step": 6338
+    },
+    {
+      "epoch": 0.8458767013610888,
+      "grad_norm": 1.0329741808696042,
+      "learning_rate": 1.2202461050473469e-06,
+      "loss": 1.5154,
+      "step": 6339
+    },
+    {
+      "epoch": 0.8460101414464906,
+      "grad_norm": 0.9806884542561266,
+      "learning_rate": 1.2181780042843071e-06,
+      "loss": 1.4914,
+      "step": 6340
+    },
+    {
+      "epoch": 0.8461435815318922,
+      "grad_norm": 1.0466476181426838,
+      "learning_rate": 1.2161115438656301e-06,
+      "loss": 1.5529,
+      "step": 6341
+    },
+    {
+      "epoch": 0.8462770216172938,
+      "grad_norm": 0.9141443517568483,
+      "learning_rate": 1.2140467241773103e-06,
+      "loss": 1.513,
+      "step": 6342
+    },
+    {
+      "epoch": 0.8464104617026955,
+      "grad_norm": 0.9944934691896574,
+      "learning_rate": 1.2119835456050311e-06,
+      "loss": 1.5567,
+      "step": 6343
+    },
+    {
+      "epoch": 0.8465439017880971,
+      "grad_norm": 1.0618279777174555,
+      "learning_rate": 1.2099220085341689e-06,
+      "loss": 1.5453,
+      "step": 6344
+    },
+    {
+      "epoch": 0.8466773418734987,
+      "grad_norm": 0.9826611891622069,
+      "learning_rate": 1.2078621133497958e-06,
+      "loss": 1.5373,
+      "step": 6345
+    },
+    {
+      "epoch": 0.8468107819589005,
+      "grad_norm": 0.9429543528239474,
+      "learning_rate": 1.2058038604366796e-06,
+      "loss": 1.5587,
+      "step": 6346
+    },
+    {
+      "epoch": 0.8469442220443021,
+      "grad_norm": 0.9780300370698021,
+      "learning_rate": 1.2037472501792757e-06,
+      "loss": 1.5841,
+      "step": 6347
+    },
+    {
+      "epoch": 0.8470776621297038,
+      "grad_norm": 0.9485414768814703,
+      "learning_rate": 1.201692282961735e-06,
+      "loss": 1.532,
+      "step": 6348
+    },
+    {
+      "epoch": 0.8472111022151054,
+      "grad_norm": 0.9528273449657707,
+      "learning_rate": 1.1996389591679047e-06,
+      "loss": 1.5611,
+      "step": 6349
+    },
+    {
+      "epoch": 0.847344542300507,
+      "grad_norm": 0.9369740412100928,
+      "learning_rate": 1.1975872791813225e-06,
+      "loss": 1.5032,
+      "step": 6350
+    },
+    {
+      "epoch": 0.8474779823859088,
+      "grad_norm": 0.9223426425400875,
+      "learning_rate": 1.1955372433852163e-06,
+      "loss": 1.5512,
+      "step": 6351
+    },
+    {
+      "epoch": 0.8476114224713104,
+      "grad_norm": 0.9430681441104803,
+      "learning_rate": 1.1934888521625076e-06,
+      "loss": 1.561,
+      "step": 6352
+    },
+    {
+      "epoch": 0.847744862556712,
+      "grad_norm": 0.969082370915136,
+      "learning_rate": 1.1914421058958192e-06,
+      "loss": 1.5353,
+      "step": 6353
+    },
+    {
+      "epoch": 0.8478783026421137,
+      "grad_norm": 1.0413763556214914,
+      "learning_rate": 1.1893970049674553e-06,
+      "loss": 1.5652,
+      "step": 6354
+    },
+    {
+      "epoch": 0.8480117427275153,
+      "grad_norm": 0.9204143577512675,
+      "learning_rate": 1.1873535497594157e-06,
+      "loss": 1.5406,
+      "step": 6355
+    },
+    {
+      "epoch": 0.848145182812917,
+      "grad_norm": 1.1508462678819407,
+      "learning_rate": 1.185311740653401e-06,
+      "loss": 1.5223,
+      "step": 6356
+    },
+    {
+      "epoch": 0.8482786228983187,
+      "grad_norm": 1.0061378039252045,
+      "learning_rate": 1.1832715780307924e-06,
+      "loss": 1.5438,
+      "step": 6357
+    },
+    {
+      "epoch": 0.8484120629837203,
+      "grad_norm": 1.0156019695357563,
+      "learning_rate": 1.181233062272672e-06,
+      "loss": 1.5484,
+      "step": 6358
+    },
+    {
+      "epoch": 0.8485455030691219,
+      "grad_norm": 0.9583797003702161,
+      "learning_rate": 1.1791961937598073e-06,
+      "loss": 1.5511,
+      "step": 6359
+    },
+    {
+      "epoch": 0.8486789431545236,
+      "grad_norm": 1.77815502053269,
+      "learning_rate": 1.1771609728726674e-06,
+      "loss": 1.5944,
+      "step": 6360
+    },
+    {
+      "epoch": 0.8488123832399252,
+      "grad_norm": 0.9601898228216447,
+      "learning_rate": 1.1751273999914059e-06,
+      "loss": 1.5584,
+      "step": 6361
+    },
+    {
+      "epoch": 0.848945823325327,
+      "grad_norm": 1.0337878007106784,
+      "learning_rate": 1.173095475495869e-06,
+      "loss": 1.5255,
+      "step": 6362
+    },
+    {
+      "epoch": 0.8490792634107286,
+      "grad_norm": 1.069797310519057,
+      "learning_rate": 1.171065199765602e-06,
+      "loss": 1.5929,
+      "step": 6363
+    },
+    {
+      "epoch": 0.8492127034961302,
+      "grad_norm": 1.2157650247272633,
+      "learning_rate": 1.1690365731798337e-06,
+      "loss": 1.5466,
+      "step": 6364
+    },
+    {
+      "epoch": 0.8493461435815319,
+      "grad_norm": 1.0078551254584747,
+      "learning_rate": 1.1670095961174889e-06,
+      "loss": 1.5735,
+      "step": 6365
+    },
+    {
+      "epoch": 0.8494795836669335,
+      "grad_norm": 0.9609917847295425,
+      "learning_rate": 1.1649842689571855e-06,
+      "loss": 1.5789,
+      "step": 6366
+    },
+    {
+      "epoch": 0.8496130237523352,
+      "grad_norm": 0.9255278292579112,
+      "learning_rate": 1.1629605920772292e-06,
+      "loss": 1.5726,
+      "step": 6367
+    },
+    {
+      "epoch": 0.8497464638377369,
+      "grad_norm": 0.9461037345263595,
+      "learning_rate": 1.1609385658556183e-06,
+      "loss": 1.572,
+      "step": 6368
+    },
+    {
+      "epoch": 0.8498799039231385,
+      "grad_norm": 0.9622357864602324,
+      "learning_rate": 1.1589181906700498e-06,
+      "loss": 1.5018,
+      "step": 6369
+    },
+    {
+      "epoch": 0.8500133440085401,
+      "grad_norm": 0.9486160725025564,
+      "learning_rate": 1.156899466897904e-06,
+      "loss": 1.5379,
+      "step": 6370
+    },
+    {
+      "epoch": 0.8501467840939418,
+      "grad_norm": 1.0361906044731701,
+      "learning_rate": 1.1548823949162546e-06,
+      "loss": 1.5514,
+      "step": 6371
+    },
+    {
+      "epoch": 0.8502802241793435,
+      "grad_norm": 0.9657512138409541,
+      "learning_rate": 1.1528669751018684e-06,
+      "loss": 1.508,
+      "step": 6372
+    },
+    {
+      "epoch": 0.8504136642647452,
+      "grad_norm": 0.9981316716898595,
+      "learning_rate": 1.1508532078312007e-06,
+      "loss": 1.5569,
+      "step": 6373
+    },
+    {
+      "epoch": 0.8505471043501468,
+      "grad_norm": 0.9476506985506088,
+      "learning_rate": 1.1488410934804051e-06,
+      "loss": 1.539,
+      "step": 6374
+    },
+    {
+      "epoch": 0.8506805444355484,
+      "grad_norm": 1.026783945248529,
+      "learning_rate": 1.1468306324253187e-06,
+      "loss": 1.55,
+      "step": 6375
+    },
+    {
+      "epoch": 0.8508139845209501,
+      "grad_norm": 1.0018298913831105,
+      "learning_rate": 1.1448218250414734e-06,
+      "loss": 1.5283,
+      "step": 6376
+    },
+    {
+      "epoch": 0.8509474246063518,
+      "grad_norm": 0.9541809577080766,
+      "learning_rate": 1.1428146717040888e-06,
+      "loss": 1.4845,
+      "step": 6377
+    },
+    {
+      "epoch": 0.8510808646917534,
+      "grad_norm": 0.94168687520855,
+      "learning_rate": 1.1408091727880822e-06,
+      "loss": 1.5261,
+      "step": 6378
+    },
+    {
+      "epoch": 0.8512143047771551,
+      "grad_norm": 0.9534267240246544,
+      "learning_rate": 1.1388053286680566e-06,
+      "loss": 1.5505,
+      "step": 6379
+    },
+    {
+      "epoch": 0.8513477448625567,
+      "grad_norm": 0.9608021930487494,
+      "learning_rate": 1.1368031397183055e-06,
+      "loss": 1.5709,
+      "step": 6380
+    },
+    {
+      "epoch": 0.8514811849479583,
+      "grad_norm": 1.0827148794751535,
+      "learning_rate": 1.1348026063128193e-06,
+      "loss": 1.54,
+      "step": 6381
+    },
+    {
+      "epoch": 0.85161462503336,
+      "grad_norm": 0.9496776082560817,
+      "learning_rate": 1.1328037288252714e-06,
+      "loss": 1.5394,
+      "step": 6382
+    },
+    {
+      "epoch": 0.8517480651187617,
+      "grad_norm": 0.925719277344387,
+      "learning_rate": 1.1308065076290298e-06,
+      "loss": 1.5546,
+      "step": 6383
+    },
+    {
+      "epoch": 0.8518815052041633,
+      "grad_norm": 0.99649007906056,
+      "learning_rate": 1.1288109430971516e-06,
+      "loss": 1.5796,
+      "step": 6384
+    },
+    {
+      "epoch": 0.852014945289565,
+      "grad_norm": 0.9217233177569367,
+      "learning_rate": 1.1268170356023889e-06,
+      "loss": 1.4987,
+      "step": 6385
+    },
+    {
+      "epoch": 0.8521483853749666,
+      "grad_norm": 0.9795549846257534,
+      "learning_rate": 1.124824785517179e-06,
+      "loss": 1.498,
+      "step": 6386
+    },
+    {
+      "epoch": 0.8522818254603683,
+      "grad_norm": 0.9466460505441463,
+      "learning_rate": 1.122834193213649e-06,
+      "loss": 1.5069,
+      "step": 6387
+    },
+    {
+      "epoch": 0.85241526554577,
+      "grad_norm": 0.9622325848501593,
+      "learning_rate": 1.1208452590636243e-06,
+      "loss": 1.572,
+      "step": 6388
+    },
+    {
+      "epoch": 0.8525487056311716,
+      "grad_norm": 0.907176146573117,
+      "learning_rate": 1.1188579834386116e-06,
+      "loss": 1.5447,
+      "step": 6389
+    },
+    {
+      "epoch": 0.8526821457165733,
+      "grad_norm": 1.227317725567885,
+      "learning_rate": 1.1168723667098115e-06,
+      "loss": 1.5481,
+      "step": 6390
+    },
+    {
+      "epoch": 0.8528155858019749,
+      "grad_norm": 0.9872294854469825,
+      "learning_rate": 1.1148884092481138e-06,
+      "loss": 1.5726,
+      "step": 6391
+    },
+    {
+      "epoch": 0.8529490258873765,
+      "grad_norm": 0.9774988409822314,
+      "learning_rate": 1.1129061114241024e-06,
+      "loss": 1.5505,
+      "step": 6392
+    },
+    {
+      "epoch": 0.8530824659727783,
+      "grad_norm": 0.9342823752805767,
+      "learning_rate": 1.1109254736080456e-06,
+      "loss": 1.5665,
+      "step": 6393
+    },
+    {
+      "epoch": 0.8532159060581799,
+      "grad_norm": 0.945834873384594,
+      "learning_rate": 1.1089464961699025e-06,
+      "loss": 1.5237,
+      "step": 6394
+    },
+    {
+      "epoch": 0.8533493461435815,
+      "grad_norm": 1.0390817853688066,
+      "learning_rate": 1.106969179479328e-06,
+      "loss": 1.559,
+      "step": 6395
+    },
+    {
+      "epoch": 0.8534827862289832,
+      "grad_norm": 0.9915119704103906,
+      "learning_rate": 1.1049935239056598e-06,
+      "loss": 1.5479,
+      "step": 6396
+    },
+    {
+      "epoch": 0.8536162263143848,
+      "grad_norm": 0.9778650816854455,
+      "learning_rate": 1.1030195298179269e-06,
+      "loss": 1.5655,
+      "step": 6397
+    },
+    {
+      "epoch": 0.8537496663997864,
+      "grad_norm": 0.9412423616517623,
+      "learning_rate": 1.101047197584849e-06,
+      "loss": 1.5287,
+      "step": 6398
+    },
+    {
+      "epoch": 0.8538831064851882,
+      "grad_norm": 0.9125544805452201,
+      "learning_rate": 1.0990765275748383e-06,
+      "loss": 1.5672,
+      "step": 6399
+    },
+    {
+      "epoch": 0.8540165465705898,
+      "grad_norm": 0.9887312218815218,
+      "learning_rate": 1.09710752015599e-06,
+      "loss": 1.5265,
+      "step": 6400
+    },
+    {
+      "epoch": 0.8541499866559915,
+      "grad_norm": 1.0483588182325054,
+      "learning_rate": 1.0951401756960934e-06,
+      "loss": 1.5733,
+      "step": 6401
+    },
+    {
+      "epoch": 0.8542834267413931,
+      "grad_norm": 0.9407674603973352,
+      "learning_rate": 1.0931744945626276e-06,
+      "loss": 1.5581,
+      "step": 6402
+    },
+    {
+      "epoch": 0.8544168668267947,
+      "grad_norm": 0.9313568972065337,
+      "learning_rate": 1.0912104771227584e-06,
+      "loss": 1.5329,
+      "step": 6403
+    },
+    {
+      "epoch": 0.8545503069121965,
+      "grad_norm": 0.9194823821651573,
+      "learning_rate": 1.0892481237433405e-06,
+      "loss": 1.5163,
+      "step": 6404
+    },
+    {
+      "epoch": 0.8546837469975981,
+      "grad_norm": 1.0427301724789466,
+      "learning_rate": 1.08728743479092e-06,
+      "loss": 1.5137,
+      "step": 6405
+    },
+    {
+      "epoch": 0.8548171870829997,
+      "grad_norm": 0.9377269194420035,
+      "learning_rate": 1.085328410631733e-06,
+      "loss": 1.4877,
+      "step": 6406
+    },
+    {
+      "epoch": 0.8549506271684014,
+      "grad_norm": 1.021530858485102,
+      "learning_rate": 1.0833710516317009e-06,
+      "loss": 1.599,
+      "step": 6407
+    },
+    {
+      "epoch": 0.855084067253803,
+      "grad_norm": 0.9614653085487166,
+      "learning_rate": 1.081415358156438e-06,
+      "loss": 1.562,
+      "step": 6408
+    },
+    {
+      "epoch": 0.8552175073392047,
+      "grad_norm": 0.9332515315220702,
+      "learning_rate": 1.0794613305712432e-06,
+      "loss": 1.5861,
+      "step": 6409
+    },
+    {
+      "epoch": 0.8553509474246064,
+      "grad_norm": 0.9465013594113041,
+      "learning_rate": 1.0775089692411057e-06,
+      "loss": 1.544,
+      "step": 6410
+    },
+    {
+      "epoch": 0.855484387510008,
+      "grad_norm": 0.9299855464015608,
+      "learning_rate": 1.075558274530709e-06,
+      "loss": 1.5169,
+      "step": 6411
+    },
+    {
+      "epoch": 0.8556178275954096,
+      "grad_norm": 0.9633862952339556,
+      "learning_rate": 1.073609246804418e-06,
+      "loss": 1.5498,
+      "step": 6412
+    },
+    {
+      "epoch": 0.8557512676808113,
+      "grad_norm": 0.9683201294799403,
+      "learning_rate": 1.0716618864262885e-06,
+      "loss": 1.4671,
+      "step": 6413
+    },
+    {
+      "epoch": 0.855884707766213,
+      "grad_norm": 0.9649479650839727,
+      "learning_rate": 1.0697161937600665e-06,
+      "loss": 1.57,
+      "step": 6414
+    },
+    {
+      "epoch": 0.8560181478516147,
+      "grad_norm": 0.9185273611508583,
+      "learning_rate": 1.0677721691691833e-06,
+      "loss": 1.5408,
+      "step": 6415
+    },
+    {
+      "epoch": 0.8561515879370163,
+      "grad_norm": 0.9695842140595766,
+      "learning_rate": 1.0658298130167599e-06,
+      "loss": 1.507,
+      "step": 6416
+    },
+    {
+      "epoch": 0.8562850280224179,
+      "grad_norm": 0.9303971531538695,
+      "learning_rate": 1.0638891256656103e-06,
+      "loss": 1.4942,
+      "step": 6417
+    },
+    {
+      "epoch": 0.8564184681078196,
+      "grad_norm": 1.0549407029198568,
+      "learning_rate": 1.0619501074782313e-06,
+      "loss": 1.5486,
+      "step": 6418
+    },
+    {
+      "epoch": 0.8565519081932212,
+      "grad_norm": 3.80436306285889,
+      "learning_rate": 1.0600127588168063e-06,
+      "loss": 1.5436,
+      "step": 6419
+    },
+    {
+      "epoch": 0.8566853482786229,
+      "grad_norm": 0.9231151613693299,
+      "learning_rate": 1.0580770800432139e-06,
+      "loss": 1.5291,
+      "step": 6420
+    },
+    {
+      "epoch": 0.8568187883640246,
+      "grad_norm": 1.076962491434642,
+      "learning_rate": 1.0561430715190158e-06,
+      "loss": 1.5237,
+      "step": 6421
+    },
+    {
+      "epoch": 0.8569522284494262,
+      "grad_norm": 0.9284052147555532,
+      "learning_rate": 1.054210733605462e-06,
+      "loss": 1.5671,
+      "step": 6422
+    },
+    {
+      "epoch": 0.8570856685348278,
+      "grad_norm": 1.013864860833294,
+      "learning_rate": 1.0522800666634891e-06,
+      "loss": 1.513,
+      "step": 6423
+    },
+    {
+      "epoch": 0.8572191086202295,
+      "grad_norm": 1.0640857162799244,
+      "learning_rate": 1.050351071053729e-06,
+      "loss": 1.5686,
+      "step": 6424
+    },
+    {
+      "epoch": 0.8573525487056312,
+      "grad_norm": 1.6541230030547414,
+      "learning_rate": 1.0484237471364922e-06,
+      "loss": 1.5551,
+      "step": 6425
+    },
+    {
+      "epoch": 0.8574859887910328,
+      "grad_norm": 1.0049096936614765,
+      "learning_rate": 1.0464980952717807e-06,
+      "loss": 1.507,
+      "step": 6426
+    },
+    {
+      "epoch": 0.8576194288764345,
+      "grad_norm": 0.9371213973946467,
+      "learning_rate": 1.0445741158192879e-06,
+      "loss": 1.5747,
+      "step": 6427
+    },
+    {
+      "epoch": 0.8577528689618361,
+      "grad_norm": 1.2240293324907436,
+      "learning_rate": 1.0426518091383886e-06,
+      "loss": 1.6061,
+      "step": 6428
+    },
+    {
+      "epoch": 0.8578863090472378,
+      "grad_norm": 0.9655483000092794,
+      "learning_rate": 1.0407311755881477e-06,
+      "loss": 1.5201,
+      "step": 6429
+    },
+    {
+      "epoch": 0.8580197491326395,
+      "grad_norm": 1.0963997227966682,
+      "learning_rate": 1.0388122155273162e-06,
+      "loss": 1.54,
+      "step": 6430
+    },
+    {
+      "epoch": 0.8581531892180411,
+      "grad_norm": 1.0333139188267229,
+      "learning_rate": 1.0368949293143383e-06,
+      "loss": 1.5161,
+      "step": 6431
+    },
+    {
+      "epoch": 0.8582866293034428,
+      "grad_norm": 0.9470028994674716,
+      "learning_rate": 1.0349793173073386e-06,
+      "loss": 1.5838,
+      "step": 6432
+    },
+    {
+      "epoch": 0.8584200693888444,
+      "grad_norm": 0.945024534455689,
+      "learning_rate": 1.0330653798641288e-06,
+      "loss": 1.5591,
+      "step": 6433
+    },
+    {
+      "epoch": 0.858553509474246,
+      "grad_norm": 0.9404662766746317,
+      "learning_rate": 1.0311531173422172e-06,
+      "loss": 1.5622,
+      "step": 6434
+    },
+    {
+      "epoch": 0.8586869495596477,
+      "grad_norm": 0.9927286739349651,
+      "learning_rate": 1.0292425300987885e-06,
+      "loss": 1.5112,
+      "step": 6435
+    },
+    {
+      "epoch": 0.8588203896450494,
+      "grad_norm": 0.9324477538639491,
+      "learning_rate": 1.027333618490719e-06,
+      "loss": 1.5527,
+      "step": 6436
+    },
+    {
+      "epoch": 0.858953829730451,
+      "grad_norm": 0.9841322098310834,
+      "learning_rate": 1.0254263828745704e-06,
+      "loss": 1.5666,
+      "step": 6437
+    },
+    {
+      "epoch": 0.8590872698158527,
+      "grad_norm": 0.9705280053887522,
+      "learning_rate": 1.0235208236065964e-06,
+      "loss": 1.4959,
+      "step": 6438
+    },
+    {
+      "epoch": 0.8592207099012543,
+      "grad_norm": 0.961832358126531,
+      "learning_rate": 1.0216169410427312e-06,
+      "loss": 1.5583,
+      "step": 6439
+    },
+    {
+      "epoch": 0.859354149986656,
+      "grad_norm": 0.937404869373422,
+      "learning_rate": 1.0197147355385983e-06,
+      "loss": 1.4833,
+      "step": 6440
+    },
+    {
+      "epoch": 0.8594875900720577,
+      "grad_norm": 0.9488407481430322,
+      "learning_rate": 1.0178142074495068e-06,
+      "loss": 1.5485,
+      "step": 6441
+    },
+    {
+      "epoch": 0.8596210301574593,
+      "grad_norm": 0.9318508881776466,
+      "learning_rate": 1.015915357130457e-06,
+      "loss": 1.5505,
+      "step": 6442
+    },
+    {
+      "epoch": 0.859754470242861,
+      "grad_norm": 0.9180739689951576,
+      "learning_rate": 1.0140181849361307e-06,
+      "loss": 1.4841,
+      "step": 6443
+    },
+    {
+      "epoch": 0.8598879103282626,
+      "grad_norm": 0.9168330491896693,
+      "learning_rate": 1.012122691220897e-06,
+      "loss": 1.4934,
+      "step": 6444
+    },
+    {
+      "epoch": 0.8600213504136642,
+      "grad_norm": 0.936550488916631,
+      "learning_rate": 1.0102288763388147e-06,
+      "loss": 1.5639,
+      "step": 6445
+    },
+    {
+      "epoch": 0.860154790499066,
+      "grad_norm": 1.2407428273088301,
+      "learning_rate": 1.0083367406436263e-06,
+      "loss": 1.5286,
+      "step": 6446
+    },
+    {
+      "epoch": 0.8602882305844676,
+      "grad_norm": 0.9564893007922771,
+      "learning_rate": 1.0064462844887613e-06,
+      "loss": 1.5186,
+      "step": 6447
+    },
+    {
+      "epoch": 0.8604216706698692,
+      "grad_norm": 0.9357452140960504,
+      "learning_rate": 1.004557508227333e-06,
+      "loss": 1.5172,
+      "step": 6448
+    },
+    {
+      "epoch": 0.8605551107552709,
+      "grad_norm": 0.8972018016158957,
+      "learning_rate": 1.0026704122121466e-06,
+      "loss": 1.5275,
+      "step": 6449
+    },
+    {
+      "epoch": 0.8606885508406725,
+      "grad_norm": 0.9998053596687703,
+      "learning_rate": 1.0007849967956884e-06,
+      "loss": 1.5544,
+      "step": 6450
+    },
+    {
+      "epoch": 0.8608219909260741,
+      "grad_norm": 0.9489874805787626,
+      "learning_rate": 9.989012623301343e-07,
+      "loss": 1.5508,
+      "step": 6451
+    },
+    {
+      "epoch": 0.8609554310114759,
+      "grad_norm": 0.9471306443210268,
+      "learning_rate": 9.970192091673414e-07,
+      "loss": 1.534,
+      "step": 6452
+    },
+    {
+      "epoch": 0.8610888710968775,
+      "grad_norm": 0.9137682040739373,
+      "learning_rate": 9.951388376588567e-07,
+      "loss": 1.5459,
+      "step": 6453
+    },
+    {
+      "epoch": 0.8612223111822792,
+      "grad_norm": 0.9641556651125667,
+      "learning_rate": 9.932601481559146e-07,
+      "loss": 1.5387,
+      "step": 6454
+    },
+    {
+      "epoch": 0.8613557512676808,
+      "grad_norm": 0.93922770841347,
+      "learning_rate": 9.91383141009431e-07,
+      "loss": 1.5581,
+      "step": 6455
+    },
+    {
+      "epoch": 0.8614891913530824,
+      "grad_norm": 0.9436691135229323,
+      "learning_rate": 9.8950781657001e-07,
+      "loss": 1.5473,
+      "step": 6456
+    },
+    {
+      "epoch": 0.8616226314384842,
+      "grad_norm": 0.9338614269793593,
+      "learning_rate": 9.876341751879404e-07,
+      "loss": 1.5505,
+      "step": 6457
+    },
+    {
+      "epoch": 0.8617560715238858,
+      "grad_norm": 0.9397414325734582,
+      "learning_rate": 9.857622172131952e-07,
+      "loss": 1.5642,
+      "step": 6458
+    },
+    {
+      "epoch": 0.8618895116092874,
+      "grad_norm": 1.0086274613066035,
+      "learning_rate": 9.838919429954386e-07,
+      "loss": 1.5355,
+      "step": 6459
+    },
+    {
+      "epoch": 0.8620229516946891,
+      "grad_norm": 0.9481076250398918,
+      "learning_rate": 9.820233528840151e-07,
+      "loss": 1.5557,
+      "step": 6460
+    },
+    {
+      "epoch": 0.8621563917800907,
+      "grad_norm": 0.9155281021750805,
+      "learning_rate": 9.801564472279557e-07,
+      "loss": 1.4963,
+      "step": 6461
+    },
+    {
+      "epoch": 0.8622898318654924,
+      "grad_norm": 0.9144924099874533,
+      "learning_rate": 9.782912263759748e-07,
+      "loss": 1.5214,
+      "step": 6462
+    },
+    {
+      "epoch": 0.8624232719508941,
+      "grad_norm": 0.9363225569423211,
+      "learning_rate": 9.764276906764792e-07,
+      "loss": 1.5297,
+      "step": 6463
+    },
+    {
+      "epoch": 0.8625567120362957,
+      "grad_norm": 0.9435473694327043,
+      "learning_rate": 9.745658404775537e-07,
+      "loss": 1.5394,
+      "step": 6464
+    },
+    {
+      "epoch": 0.8626901521216973,
+      "grad_norm": 0.9985762329589031,
+      "learning_rate": 9.727056761269693e-07,
+      "loss": 1.5468,
+      "step": 6465
+    },
+    {
+      "epoch": 0.862823592207099,
+      "grad_norm": 0.9290866980377026,
+      "learning_rate": 9.708471979721868e-07,
+      "loss": 1.5702,
+      "step": 6466
+    },
+    {
+      "epoch": 0.8629570322925006,
+      "grad_norm": 1.1072478162840556,
+      "learning_rate": 9.689904063603461e-07,
+      "loss": 1.582,
+      "step": 6467
+    },
+    {
+      "epoch": 0.8630904723779024,
+      "grad_norm": 0.9126478135993858,
+      "learning_rate": 9.671353016382767e-07,
+      "loss": 1.5269,
+      "step": 6468
+    },
+    {
+      "epoch": 0.863223912463304,
+      "grad_norm": 0.9443342912207419,
+      "learning_rate": 9.65281884152487e-07,
+      "loss": 1.5555,
+      "step": 6469
+    },
+    {
+      "epoch": 0.8633573525487056,
+      "grad_norm": 0.9209254675884877,
+      "learning_rate": 9.634301542491798e-07,
+      "loss": 1.5324,
+      "step": 6470
+    },
+    {
+      "epoch": 0.8634907926341073,
+      "grad_norm": 0.9549939583445304,
+      "learning_rate": 9.61580112274234e-07,
+      "loss": 1.5308,
+      "step": 6471
+    },
+    {
+      "epoch": 0.8636242327195089,
+      "grad_norm": 0.9770641027377173,
+      "learning_rate": 9.59731758573217e-07,
+      "loss": 1.544,
+      "step": 6472
+    },
+    {
+      "epoch": 0.8637576728049106,
+      "grad_norm": 0.8817507518794163,
+      "learning_rate": 9.578850934913786e-07,
+      "loss": 1.5241,
+      "step": 6473
+    },
+    {
+      "epoch": 0.8638911128903123,
+      "grad_norm": 0.9128330688258248,
+      "learning_rate": 9.560401173736588e-07,
+      "loss": 1.5328,
+      "step": 6474
+    },
+    {
+      "epoch": 0.8640245529757139,
+      "grad_norm": 0.9271171282021506,
+      "learning_rate": 9.541968305646754e-07,
+      "loss": 1.5089,
+      "step": 6475
+    },
+    {
+      "epoch": 0.8641579930611155,
+      "grad_norm": 0.898516087846143,
+      "learning_rate": 9.523552334087316e-07,
+      "loss": 1.4933,
+      "step": 6476
+    },
+    {
+      "epoch": 0.8642914331465172,
+      "grad_norm": 0.9455820905578254,
+      "learning_rate": 9.505153262498201e-07,
+      "loss": 1.5412,
+      "step": 6477
+    },
+    {
+      "epoch": 0.8644248732319189,
+      "grad_norm": 0.9769692174530545,
+      "learning_rate": 9.486771094316149e-07,
+      "loss": 1.5058,
+      "step": 6478
+    },
+    {
+      "epoch": 0.8645583133173205,
+      "grad_norm": 0.9225921262405403,
+      "learning_rate": 9.468405832974714e-07,
+      "loss": 1.4898,
+      "step": 6479
+    },
+    {
+      "epoch": 0.8646917534027222,
+      "grad_norm": 0.9336861605888874,
+      "learning_rate": 9.450057481904306e-07,
+      "loss": 1.5869,
+      "step": 6480
+    },
+    {
+      "epoch": 0.8648251934881238,
+      "grad_norm": 1.1283612150349465,
+      "learning_rate": 9.431726044532241e-07,
+      "loss": 1.5553,
+      "step": 6481
+    },
+    {
+      "epoch": 0.8649586335735255,
+      "grad_norm": 0.9124111539042673,
+      "learning_rate": 9.41341152428259e-07,
+      "loss": 1.5299,
+      "step": 6482
+    },
+    {
+      "epoch": 0.8650920736589272,
+      "grad_norm": 0.9550521448834836,
+      "learning_rate": 9.395113924576271e-07,
+      "loss": 1.549,
+      "step": 6483
+    },
+    {
+      "epoch": 0.8652255137443288,
+      "grad_norm": 0.9872576025252855,
+      "learning_rate": 9.376833248831119e-07,
+      "loss": 1.5379,
+      "step": 6484
+    },
+    {
+      "epoch": 0.8653589538297305,
+      "grad_norm": 0.9222828401540125,
+      "learning_rate": 9.358569500461734e-07,
+      "loss": 1.5166,
+      "step": 6485
+    },
+    {
+      "epoch": 0.8654923939151321,
+      "grad_norm": 1.0447537425480853,
+      "learning_rate": 9.340322682879577e-07,
+      "loss": 1.5124,
+      "step": 6486
+    },
+    {
+      "epoch": 0.8656258340005337,
+      "grad_norm": 0.9039392609995586,
+      "learning_rate": 9.322092799492921e-07,
+      "loss": 1.5092,
+      "step": 6487
+    },
+    {
+      "epoch": 0.8657592740859354,
+      "grad_norm": 0.9498842193910904,
+      "learning_rate": 9.303879853706955e-07,
+      "loss": 1.5568,
+      "step": 6488
+    },
+    {
+      "epoch": 0.8658927141713371,
+      "grad_norm": 0.9356401186158558,
+      "learning_rate": 9.285683848923599e-07,
+      "loss": 1.5631,
+      "step": 6489
+    },
+    {
+      "epoch": 0.8660261542567387,
+      "grad_norm": 0.9816096213234118,
+      "learning_rate": 9.26750478854167e-07,
+      "loss": 1.5634,
+      "step": 6490
+    },
+    {
+      "epoch": 0.8661595943421404,
+      "grad_norm": 0.9902314871977751,
+      "learning_rate": 9.249342675956841e-07,
+      "loss": 1.5368,
+      "step": 6491
+    },
+    {
+      "epoch": 0.866293034427542,
+      "grad_norm": 0.913769147225799,
+      "learning_rate": 9.231197514561552e-07,
+      "loss": 1.494,
+      "step": 6492
+    },
+    {
+      "epoch": 0.8664264745129436,
+      "grad_norm": 0.9901239051487305,
+      "learning_rate": 9.213069307745137e-07,
+      "loss": 1.529,
+      "step": 6493
+    },
+    {
+      "epoch": 0.8665599145983454,
+      "grad_norm": 0.9186278024861386,
+      "learning_rate": 9.194958058893722e-07,
+      "loss": 1.5525,
+      "step": 6494
+    },
+    {
+      "epoch": 0.866693354683747,
+      "grad_norm": 0.9601791869003774,
+      "learning_rate": 9.176863771390288e-07,
+      "loss": 1.5595,
+      "step": 6495
+    },
+    {
+      "epoch": 0.8668267947691487,
+      "grad_norm": 0.9344491574122844,
+      "learning_rate": 9.158786448614621e-07,
+      "loss": 1.5765,
+      "step": 6496
+    },
+    {
+      "epoch": 0.8669602348545503,
+      "grad_norm": 1.0059039142212696,
+      "learning_rate": 9.140726093943409e-07,
+      "loss": 1.5543,
+      "step": 6497
+    },
+    {
+      "epoch": 0.8670936749399519,
+      "grad_norm": 1.1316836388517142,
+      "learning_rate": 9.122682710750074e-07,
+      "loss": 1.5017,
+      "step": 6498
+    },
+    {
+      "epoch": 0.8672271150253537,
+      "grad_norm": 0.9402893670787278,
+      "learning_rate": 9.104656302404946e-07,
+      "loss": 1.5371,
+      "step": 6499
+    },
+    {
+      "epoch": 0.8673605551107553,
+      "grad_norm": 0.9832968180647175,
+      "learning_rate": 9.086646872275129e-07,
+      "loss": 1.5387,
+      "step": 6500
+    },
+    {
+      "epoch": 0.8674939951961569,
+      "grad_norm": 0.9514391898183131,
+      "learning_rate": 9.068654423724577e-07,
+      "loss": 1.5823,
+      "step": 6501
+    },
+    {
+      "epoch": 0.8676274352815586,
+      "grad_norm": 0.9488421342873816,
+      "learning_rate": 9.050678960114101e-07,
+      "loss": 1.5656,
+      "step": 6502
+    },
+    {
+      "epoch": 0.8677608753669602,
+      "grad_norm": 0.9115733795737984,
+      "learning_rate": 9.032720484801294e-07,
+      "loss": 1.5759,
+      "step": 6503
+    },
+    {
+      "epoch": 0.8678943154523618,
+      "grad_norm": 0.9981354349705964,
+      "learning_rate": 9.014779001140606e-07,
+      "loss": 1.574,
+      "step": 6504
+    },
+    {
+      "epoch": 0.8680277555377636,
+      "grad_norm": 0.9520562761893402,
+      "learning_rate": 8.996854512483277e-07,
+      "loss": 1.6245,
+      "step": 6505
+    },
+    {
+      "epoch": 0.8681611956231652,
+      "grad_norm": 1.4726974542567874,
+      "learning_rate": 8.978947022177431e-07,
+      "loss": 1.525,
+      "step": 6506
+    },
+    {
+      "epoch": 0.8682946357085669,
+      "grad_norm": 1.184012125670187,
+      "learning_rate": 8.961056533567969e-07,
+      "loss": 1.5583,
+      "step": 6507
+    },
+    {
+      "epoch": 0.8684280757939685,
+      "grad_norm": 0.9351096655055477,
+      "learning_rate": 8.943183049996606e-07,
+      "loss": 1.6061,
+      "step": 6508
+    },
+    {
+      "epoch": 0.8685615158793701,
+      "grad_norm": 0.9305843754964107,
+      "learning_rate": 8.925326574801952e-07,
+      "loss": 1.5381,
+      "step": 6509
+    },
+    {
+      "epoch": 0.8686949559647719,
+      "grad_norm": 1.3119179296076262,
+      "learning_rate": 8.907487111319368e-07,
+      "loss": 1.5312,
+      "step": 6510
+    },
+    {
+      "epoch": 0.8688283960501735,
+      "grad_norm": 0.9409917789541443,
+      "learning_rate": 8.889664662881059e-07,
+      "loss": 1.5381,
+      "step": 6511
+    },
+    {
+      "epoch": 0.8689618361355751,
+      "grad_norm": 0.9287019422318122,
+      "learning_rate": 8.871859232816049e-07,
+      "loss": 1.5654,
+      "step": 6512
+    },
+    {
+      "epoch": 0.8690952762209768,
+      "grad_norm": 0.9231687758110173,
+      "learning_rate": 8.854070824450223e-07,
+      "loss": 1.6082,
+      "step": 6513
+    },
+    {
+      "epoch": 0.8692287163063784,
+      "grad_norm": 0.9338271404408033,
+      "learning_rate": 8.836299441106222e-07,
+      "loss": 1.5037,
+      "step": 6514
+    },
+    {
+      "epoch": 0.86936215639178,
+      "grad_norm": 0.9727417740878118,
+      "learning_rate": 8.818545086103536e-07,
+      "loss": 1.5739,
+      "step": 6515
+    },
+    {
+      "epoch": 0.8694955964771818,
+      "grad_norm": 0.95148914834048,
+      "learning_rate": 8.800807762758501e-07,
+      "loss": 1.5617,
+      "step": 6516
+    },
+    {
+      "epoch": 0.8696290365625834,
+      "grad_norm": 0.9361374474665792,
+      "learning_rate": 8.783087474384245e-07,
+      "loss": 1.5317,
+      "step": 6517
+    },
+    {
+      "epoch": 0.869762476647985,
+      "grad_norm": 0.9353510556127035,
+      "learning_rate": 8.765384224290697e-07,
+      "loss": 1.5276,
+      "step": 6518
+    },
+    {
+      "epoch": 0.8698959167333867,
+      "grad_norm": 0.9385834806918207,
+      "learning_rate": 8.747698015784612e-07,
+      "loss": 1.5294,
+      "step": 6519
+    },
+    {
+      "epoch": 0.8700293568187883,
+      "grad_norm": 0.97192379174236,
+      "learning_rate": 8.730028852169614e-07,
+      "loss": 1.516,
+      "step": 6520
+    },
+    {
+      "epoch": 0.8701627969041901,
+      "grad_norm": 0.9442829507665126,
+      "learning_rate": 8.712376736746075e-07,
+      "loss": 1.4664,
+      "step": 6521
+    },
+    {
+      "epoch": 0.8702962369895917,
+      "grad_norm": 1.0520474911116193,
+      "learning_rate": 8.694741672811191e-07,
+      "loss": 1.5211,
+      "step": 6522
+    },
+    {
+      "epoch": 0.8704296770749933,
+      "grad_norm": 0.9648717939397788,
+      "learning_rate": 8.677123663659038e-07,
+      "loss": 1.5513,
+      "step": 6523
+    },
+    {
+      "epoch": 0.870563117160395,
+      "grad_norm": 0.968648455882469,
+      "learning_rate": 8.659522712580437e-07,
+      "loss": 1.5971,
+      "step": 6524
+    },
+    {
+      "epoch": 0.8706965572457966,
+      "grad_norm": 0.9731547427610653,
+      "learning_rate": 8.641938822863039e-07,
+      "loss": 1.5163,
+      "step": 6525
+    },
+    {
+      "epoch": 0.8708299973311983,
+      "grad_norm": 1.3042383135891946,
+      "learning_rate": 8.624371997791292e-07,
+      "loss": 1.5791,
+      "step": 6526
+    },
+    {
+      "epoch": 0.8709634374166,
+      "grad_norm": 0.9354023468065126,
+      "learning_rate": 8.60682224064654e-07,
+      "loss": 1.5338,
+      "step": 6527
+    },
+    {
+      "epoch": 0.8710968775020016,
+      "grad_norm": 1.0952681463977836,
+      "learning_rate": 8.589289554706826e-07,
+      "loss": 1.4912,
+      "step": 6528
+    },
+    {
+      "epoch": 0.8712303175874032,
+      "grad_norm": 0.9461007651182953,
+      "learning_rate": 8.571773943247063e-07,
+      "loss": 1.5812,
+      "step": 6529
+    },
+    {
+      "epoch": 0.8713637576728049,
+      "grad_norm": 0.9328438835651939,
+      "learning_rate": 8.554275409539004e-07,
+      "loss": 1.5688,
+      "step": 6530
+    },
+    {
+      "epoch": 0.8714971977582066,
+      "grad_norm": 0.9452776089490901,
+      "learning_rate": 8.536793956851141e-07,
+      "loss": 1.519,
+      "step": 6531
+    },
+    {
+      "epoch": 0.8716306378436082,
+      "grad_norm": 0.9314474567474438,
+      "learning_rate": 8.519329588448822e-07,
+      "loss": 1.487,
+      "step": 6532
+    },
+    {
+      "epoch": 0.8717640779290099,
+      "grad_norm": 0.9203268868498949,
+      "learning_rate": 8.501882307594167e-07,
+      "loss": 1.5422,
+      "step": 6533
+    },
+    {
+      "epoch": 0.8718975180144115,
+      "grad_norm": 0.9548253170186739,
+      "learning_rate": 8.484452117546171e-07,
+      "loss": 1.5782,
+      "step": 6534
+    },
+    {
+      "epoch": 0.8720309580998132,
+      "grad_norm": 0.9327588816923488,
+      "learning_rate": 8.467039021560575e-07,
+      "loss": 1.5449,
+      "step": 6535
+    },
+    {
+      "epoch": 0.8721643981852149,
+      "grad_norm": 1.116947865086207,
+      "learning_rate": 8.449643022889953e-07,
+      "loss": 1.5331,
+      "step": 6536
+    },
+    {
+      "epoch": 0.8722978382706165,
+      "grad_norm": 0.9154028744010556,
+      "learning_rate": 8.432264124783662e-07,
+      "loss": 1.5358,
+      "step": 6537
+    },
+    {
+      "epoch": 0.8724312783560182,
+      "grad_norm": 0.9584163843597435,
+      "learning_rate": 8.414902330487906e-07,
+      "loss": 1.5348,
+      "step": 6538
+    },
+    {
+      "epoch": 0.8725647184414198,
+      "grad_norm": 0.96915727974722,
+      "learning_rate": 8.397557643245646e-07,
+      "loss": 1.5153,
+      "step": 6539
+    },
+    {
+      "epoch": 0.8726981585268214,
+      "grad_norm": 0.9454548862508139,
+      "learning_rate": 8.38023006629669e-07,
+      "loss": 1.5727,
+      "step": 6540
+    },
+    {
+      "epoch": 0.8728315986122231,
+      "grad_norm": 0.9381925491031865,
+      "learning_rate": 8.36291960287764e-07,
+      "loss": 1.5877,
+      "step": 6541
+    },
+    {
+      "epoch": 0.8729650386976248,
+      "grad_norm": 0.9475264001662446,
+      "learning_rate": 8.345626256221873e-07,
+      "loss": 1.5583,
+      "step": 6542
+    },
+    {
+      "epoch": 0.8730984787830264,
+      "grad_norm": 0.9063919928958889,
+      "learning_rate": 8.328350029559595e-07,
+      "loss": 1.5421,
+      "step": 6543
+    },
+    {
+      "epoch": 0.8732319188684281,
+      "grad_norm": 1.162151324167172,
+      "learning_rate": 8.311090926117793e-07,
+      "loss": 1.5868,
+      "step": 6544
+    },
+    {
+      "epoch": 0.8733653589538297,
+      "grad_norm": 0.9586262421769765,
+      "learning_rate": 8.293848949120309e-07,
+      "loss": 1.5314,
+      "step": 6545
+    },
+    {
+      "epoch": 0.8734987990392313,
+      "grad_norm": 0.9299431509946539,
+      "learning_rate": 8.276624101787733e-07,
+      "loss": 1.4904,
+      "step": 6546
+    },
+    {
+      "epoch": 0.8736322391246331,
+      "grad_norm": 1.1008446800425085,
+      "learning_rate": 8.259416387337437e-07,
+      "loss": 1.5502,
+      "step": 6547
+    },
+    {
+      "epoch": 0.8737656792100347,
+      "grad_norm": 0.9407230063487958,
+      "learning_rate": 8.242225808983684e-07,
+      "loss": 1.5496,
+      "step": 6548
+    },
+    {
+      "epoch": 0.8738991192954364,
+      "grad_norm": 0.9950896376938949,
+      "learning_rate": 8.225052369937436e-07,
+      "loss": 1.569,
+      "step": 6549
+    },
+    {
+      "epoch": 0.874032559380838,
+      "grad_norm": 0.9343428376024484,
+      "learning_rate": 8.207896073406518e-07,
+      "loss": 1.5303,
+      "step": 6550
+    },
+    {
+      "epoch": 0.8741659994662396,
+      "grad_norm": 0.9213894936942306,
+      "learning_rate": 8.1907569225955e-07,
+      "loss": 1.5117,
+      "step": 6551
+    },
+    {
+      "epoch": 0.8742994395516414,
+      "grad_norm": 1.0003551884905675,
+      "learning_rate": 8.17363492070582e-07,
+      "loss": 1.5317,
+      "step": 6552
+    },
+    {
+      "epoch": 0.874432879637043,
+      "grad_norm": 0.952551223522479,
+      "learning_rate": 8.156530070935654e-07,
+      "loss": 1.5314,
+      "step": 6553
+    },
+    {
+      "epoch": 0.8745663197224446,
+      "grad_norm": 0.9370104185043665,
+      "learning_rate": 8.139442376479967e-07,
+      "loss": 1.5173,
+      "step": 6554
+    },
+    {
+      "epoch": 0.8746997598078463,
+      "grad_norm": 1.0396921232304424,
+      "learning_rate": 8.122371840530597e-07,
+      "loss": 1.5413,
+      "step": 6555
+    },
+    {
+      "epoch": 0.8748331998932479,
+      "grad_norm": 0.9443695541062251,
+      "learning_rate": 8.105318466276102e-07,
+      "loss": 1.5138,
+      "step": 6556
+    },
+    {
+      "epoch": 0.8749666399786495,
+      "grad_norm": 0.9529659601953869,
+      "learning_rate": 8.088282256901858e-07,
+      "loss": 1.5592,
+      "step": 6557
+    },
+    {
+      "epoch": 0.8751000800640513,
+      "grad_norm": 0.9345505185968372,
+      "learning_rate": 8.07126321559002e-07,
+      "loss": 1.5475,
+      "step": 6558
+    },
+    {
+      "epoch": 0.8752335201494529,
+      "grad_norm": 0.9949322978905206,
+      "learning_rate": 8.054261345519576e-07,
+      "loss": 1.5437,
+      "step": 6559
+    },
+    {
+      "epoch": 0.8753669602348545,
+      "grad_norm": 1.3528014835986673,
+      "learning_rate": 8.037276649866277e-07,
+      "loss": 1.5508,
+      "step": 6560
+    },
+    {
+      "epoch": 0.8755004003202562,
+      "grad_norm": 0.9126381930458377,
+      "learning_rate": 8.02030913180264e-07,
+      "loss": 1.5883,
+      "step": 6561
+    },
+    {
+      "epoch": 0.8756338404056578,
+      "grad_norm": 0.9059415684652883,
+      "learning_rate": 8.003358794498051e-07,
+      "loss": 1.4654,
+      "step": 6562
+    },
+    {
+      "epoch": 0.8757672804910596,
+      "grad_norm": 0.9317951935750919,
+      "learning_rate": 7.986425641118612e-07,
+      "loss": 1.526,
+      "step": 6563
+    },
+    {
+      "epoch": 0.8759007205764612,
+      "grad_norm": 1.2757830287361367,
+      "learning_rate": 7.969509674827258e-07,
+      "loss": 1.5593,
+      "step": 6564
+    },
+    {
+      "epoch": 0.8760341606618628,
+      "grad_norm": 0.9421033561329558,
+      "learning_rate": 7.952610898783675e-07,
+      "loss": 1.5101,
+      "step": 6565
+    },
+    {
+      "epoch": 0.8761676007472645,
+      "grad_norm": 0.9277438199353042,
+      "learning_rate": 7.935729316144381e-07,
+      "loss": 1.5095,
+      "step": 6566
+    },
+    {
+      "epoch": 0.8763010408326661,
+      "grad_norm": 0.9336579625689782,
+      "learning_rate": 7.918864930062675e-07,
+      "loss": 1.5145,
+      "step": 6567
+    },
+    {
+      "epoch": 0.8764344809180677,
+      "grad_norm": 0.9777046753987705,
+      "learning_rate": 7.902017743688606e-07,
+      "loss": 1.5592,
+      "step": 6568
+    },
+    {
+      "epoch": 0.8765679210034695,
+      "grad_norm": 0.9684594181447197,
+      "learning_rate": 7.885187760169033e-07,
+      "loss": 1.572,
+      "step": 6569
+    },
+    {
+      "epoch": 0.8767013610888711,
+      "grad_norm": 1.267214799587763,
+      "learning_rate": 7.868374982647642e-07,
+      "loss": 1.5555,
+      "step": 6570
+    },
+    {
+      "epoch": 0.8768348011742727,
+      "grad_norm": 0.9419267296602276,
+      "learning_rate": 7.851579414264843e-07,
+      "loss": 1.5705,
+      "step": 6571
+    },
+    {
+      "epoch": 0.8769682412596744,
+      "grad_norm": 0.9304158422245021,
+      "learning_rate": 7.834801058157837e-07,
+      "loss": 1.5284,
+      "step": 6572
+    },
+    {
+      "epoch": 0.877101681345076,
+      "grad_norm": 0.9745745483632863,
+      "learning_rate": 7.818039917460674e-07,
+      "loss": 1.512,
+      "step": 6573
+    },
+    {
+      "epoch": 0.8772351214304777,
+      "grad_norm": 0.9525590369369205,
+      "learning_rate": 7.801295995304125e-07,
+      "loss": 1.532,
+      "step": 6574
+    },
+    {
+      "epoch": 0.8773685615158794,
+      "grad_norm": 0.9198789574840573,
+      "learning_rate": 7.78456929481577e-07,
+      "loss": 1.516,
+      "step": 6575
+    },
+    {
+      "epoch": 0.877502001601281,
+      "grad_norm": 0.901804075657504,
+      "learning_rate": 7.767859819119927e-07,
+      "loss": 1.5354,
+      "step": 6576
+    },
+    {
+      "epoch": 0.8776354416866827,
+      "grad_norm": 0.9437717919984645,
+      "learning_rate": 7.751167571337792e-07,
+      "loss": 1.5562,
+      "step": 6577
+    },
+    {
+      "epoch": 0.8777688817720843,
+      "grad_norm": 0.901156691623584,
+      "learning_rate": 7.734492554587269e-07,
+      "loss": 1.566,
+      "step": 6578
+    },
+    {
+      "epoch": 0.877902321857486,
+      "grad_norm": 0.9507122398403107,
+      "learning_rate": 7.717834771983046e-07,
+      "loss": 1.5809,
+      "step": 6579
+    },
+    {
+      "epoch": 0.8780357619428877,
+      "grad_norm": 0.9154819364254669,
+      "learning_rate": 7.70119422663661e-07,
+      "loss": 1.5289,
+      "step": 6580
+    },
+    {
+      "epoch": 0.8781692020282893,
+      "grad_norm": 0.9235692837642071,
+      "learning_rate": 7.684570921656231e-07,
+      "loss": 1.5468,
+      "step": 6581
+    },
+    {
+      "epoch": 0.8783026421136909,
+      "grad_norm": 0.9431590240253189,
+      "learning_rate": 7.667964860146959e-07,
+      "loss": 1.5715,
+      "step": 6582
+    },
+    {
+      "epoch": 0.8784360821990926,
+      "grad_norm": 0.9841453617835374,
+      "learning_rate": 7.651376045210612e-07,
+      "loss": 1.5703,
+      "step": 6583
+    },
+    {
+      "epoch": 0.8785695222844943,
+      "grad_norm": 0.9280133482016604,
+      "learning_rate": 7.63480447994579e-07,
+      "loss": 1.5325,
+      "step": 6584
+    },
+    {
+      "epoch": 0.8787029623698959,
+      "grad_norm": 0.9042299305841511,
+      "learning_rate": 7.618250167447871e-07,
+      "loss": 1.5151,
+      "step": 6585
+    },
+    {
+      "epoch": 0.8788364024552976,
+      "grad_norm": 0.9303954686816721,
+      "learning_rate": 7.601713110809006e-07,
+      "loss": 1.5533,
+      "step": 6586
+    },
+    {
+      "epoch": 0.8789698425406992,
+      "grad_norm": 0.9450693162515615,
+      "learning_rate": 7.585193313118155e-07,
+      "loss": 1.5242,
+      "step": 6587
+    },
+    {
+      "epoch": 0.8791032826261009,
+      "grad_norm": 0.9224963862036915,
+      "learning_rate": 7.568690777461008e-07,
+      "loss": 1.5231,
+      "step": 6588
+    },
+    {
+      "epoch": 0.8792367227115025,
+      "grad_norm": 1.051976485103453,
+      "learning_rate": 7.552205506920052e-07,
+      "loss": 1.5829,
+      "step": 6589
+    },
+    {
+      "epoch": 0.8793701627969042,
+      "grad_norm": 0.9433620420881246,
+      "learning_rate": 7.535737504574536e-07,
+      "loss": 1.4849,
+      "step": 6590
+    },
+    {
+      "epoch": 0.8795036028823059,
+      "grad_norm": 1.0769771830246422,
+      "learning_rate": 7.519286773500522e-07,
+      "loss": 1.5769,
+      "step": 6591
+    },
+    {
+      "epoch": 0.8796370429677075,
+      "grad_norm": 0.9338463460938362,
+      "learning_rate": 7.502853316770808e-07,
+      "loss": 1.5084,
+      "step": 6592
+    },
+    {
+      "epoch": 0.8797704830531091,
+      "grad_norm": 0.9513071114461021,
+      "learning_rate": 7.48643713745496e-07,
+      "loss": 1.4993,
+      "step": 6593
+    },
+    {
+      "epoch": 0.8799039231385108,
+      "grad_norm": 0.9736174380569763,
+      "learning_rate": 7.47003823861936e-07,
+      "loss": 1.5405,
+      "step": 6594
+    },
+    {
+      "epoch": 0.8800373632239125,
+      "grad_norm": 0.9381351170469457,
+      "learning_rate": 7.453656623327132e-07,
+      "loss": 1.5456,
+      "step": 6595
+    },
+    {
+      "epoch": 0.8801708033093141,
+      "grad_norm": 0.9345529815258716,
+      "learning_rate": 7.437292294638155e-07,
+      "loss": 1.509,
+      "step": 6596
+    },
+    {
+      "epoch": 0.8803042433947158,
+      "grad_norm": 0.9678042966662639,
+      "learning_rate": 7.420945255609102e-07,
+      "loss": 1.4996,
+      "step": 6597
+    },
+    {
+      "epoch": 0.8804376834801174,
+      "grad_norm": 0.9243133790343941,
+      "learning_rate": 7.404615509293444e-07,
+      "loss": 1.5459,
+      "step": 6598
+    },
+    {
+      "epoch": 0.880571123565519,
+      "grad_norm": 0.9421650715900654,
+      "learning_rate": 7.388303058741364e-07,
+      "loss": 1.5101,
+      "step": 6599
+    },
+    {
+      "epoch": 0.8807045636509208,
+      "grad_norm": 0.9521800237087055,
+      "learning_rate": 7.372007906999856e-07,
+      "loss": 1.56,
+      "step": 6600
+    },
+    {
+      "epoch": 0.8808380037363224,
+      "grad_norm": 0.9828304259855544,
+      "learning_rate": 7.355730057112643e-07,
+      "loss": 1.5758,
+      "step": 6601
+    },
+    {
+      "epoch": 0.8809714438217241,
+      "grad_norm": 0.9439889966484398,
+      "learning_rate": 7.339469512120268e-07,
+      "loss": 1.5744,
+      "step": 6602
+    },
+    {
+      "epoch": 0.8811048839071257,
+      "grad_norm": 0.9616630127654491,
+      "learning_rate": 7.323226275060014e-07,
+      "loss": 1.5572,
+      "step": 6603
+    },
+    {
+      "epoch": 0.8812383239925273,
+      "grad_norm": 0.9218601721204037,
+      "learning_rate": 7.307000348965909e-07,
+      "loss": 1.4996,
+      "step": 6604
+    },
+    {
+      "epoch": 0.881371764077929,
+      "grad_norm": 1.0701166098993966,
+      "learning_rate": 7.290791736868819e-07,
+      "loss": 1.5297,
+      "step": 6605
+    },
+    {
+      "epoch": 0.8815052041633307,
+      "grad_norm": 0.9708535489021737,
+      "learning_rate": 7.274600441796287e-07,
+      "loss": 1.519,
+      "step": 6606
+    },
+    {
+      "epoch": 0.8816386442487323,
+      "grad_norm": 0.9728154176736694,
+      "learning_rate": 7.258426466772672e-07,
+      "loss": 1.5832,
+      "step": 6607
+    },
+    {
+      "epoch": 0.881772084334134,
+      "grad_norm": 0.9383164976637655,
+      "learning_rate": 7.242269814819081e-07,
+      "loss": 1.5461,
+      "step": 6608
+    },
+    {
+      "epoch": 0.8819055244195356,
+      "grad_norm": 0.9213707626663843,
+      "learning_rate": 7.226130488953409e-07,
+      "loss": 1.5343,
+      "step": 6609
+    },
+    {
+      "epoch": 0.8820389645049372,
+      "grad_norm": 0.9620469927915996,
+      "learning_rate": 7.210008492190301e-07,
+      "loss": 1.5522,
+      "step": 6610
+    },
+    {
+      "epoch": 0.882172404590339,
+      "grad_norm": 0.9341015762591438,
+      "learning_rate": 7.193903827541127e-07,
+      "loss": 1.4715,
+      "step": 6611
+    },
+    {
+      "epoch": 0.8823058446757406,
+      "grad_norm": 0.9956857458374051,
+      "learning_rate": 7.177816498014101e-07,
+      "loss": 1.5903,
+      "step": 6612
+    },
+    {
+      "epoch": 0.8824392847611422,
+      "grad_norm": 0.9857230476852422,
+      "learning_rate": 7.161746506614142e-07,
+      "loss": 1.5852,
+      "step": 6613
+    },
+    {
+      "epoch": 0.8825727248465439,
+      "grad_norm": 0.9343322357829968,
+      "learning_rate": 7.145693856342928e-07,
+      "loss": 1.5051,
+      "step": 6614
+    },
+    {
+      "epoch": 0.8827061649319455,
+      "grad_norm": 0.9231910471570226,
+      "learning_rate": 7.129658550198892e-07,
+      "loss": 1.5404,
+      "step": 6615
+    },
+    {
+      "epoch": 0.8828396050173473,
+      "grad_norm": 0.9422235323401871,
+      "learning_rate": 7.113640591177296e-07,
+      "loss": 1.5096,
+      "step": 6616
+    },
+    {
+      "epoch": 0.8829730451027489,
+      "grad_norm": 0.9268050292928415,
+      "learning_rate": 7.097639982270077e-07,
+      "loss": 1.5613,
+      "step": 6617
+    },
+    {
+      "epoch": 0.8831064851881505,
+      "grad_norm": 1.0117584913067665,
+      "learning_rate": 7.081656726465968e-07,
+      "loss": 1.5219,
+      "step": 6618
+    },
+    {
+      "epoch": 0.8832399252735522,
+      "grad_norm": 1.0082915428918369,
+      "learning_rate": 7.065690826750482e-07,
+      "loss": 1.564,
+      "step": 6619
+    },
+    {
+      "epoch": 0.8833733653589538,
+      "grad_norm": 0.9226919224076718,
+      "learning_rate": 7.049742286105843e-07,
+      "loss": 1.4563,
+      "step": 6620
+    },
+    {
+      "epoch": 0.8835068054443554,
+      "grad_norm": 0.9235669483501304,
+      "learning_rate": 7.033811107511079e-07,
+      "loss": 1.5549,
+      "step": 6621
+    },
+    {
+      "epoch": 0.8836402455297572,
+      "grad_norm": 0.9675042520116867,
+      "learning_rate": 7.017897293941934e-07,
+      "loss": 1.5154,
+      "step": 6622
+    },
+    {
+      "epoch": 0.8837736856151588,
+      "grad_norm": 0.9532787577491273,
+      "learning_rate": 7.002000848370938e-07,
+      "loss": 1.5967,
+      "step": 6623
+    },
+    {
+      "epoch": 0.8839071257005604,
+      "grad_norm": 0.9172226848899977,
+      "learning_rate": 6.98612177376734e-07,
+      "loss": 1.5259,
+      "step": 6624
+    },
+    {
+      "epoch": 0.8840405657859621,
+      "grad_norm": 0.9383987828753451,
+      "learning_rate": 6.970260073097213e-07,
+      "loss": 1.5119,
+      "step": 6625
+    },
+    {
+      "epoch": 0.8841740058713637,
+      "grad_norm": 1.2105279069704957,
+      "learning_rate": 6.954415749323318e-07,
+      "loss": 1.5682,
+      "step": 6626
+    },
+    {
+      "epoch": 0.8843074459567654,
+      "grad_norm": 0.9799759626641332,
+      "learning_rate": 6.9385888054052e-07,
+      "loss": 1.631,
+      "step": 6627
+    },
+    {
+      "epoch": 0.8844408860421671,
+      "grad_norm": 0.9054994550461375,
+      "learning_rate": 6.92277924429915e-07,
+      "loss": 1.4996,
+      "step": 6628
+    },
+    {
+      "epoch": 0.8845743261275687,
+      "grad_norm": 0.9271188131214871,
+      "learning_rate": 6.906987068958193e-07,
+      "loss": 1.6082,
+      "step": 6629
+    },
+    {
+      "epoch": 0.8847077662129704,
+      "grad_norm": 0.9259197742858476,
+      "learning_rate": 6.89121228233216e-07,
+      "loss": 1.5372,
+      "step": 6630
+    },
+    {
+      "epoch": 0.884841206298372,
+      "grad_norm": 0.9170649924609936,
+      "learning_rate": 6.875454887367605e-07,
+      "loss": 1.5452,
+      "step": 6631
+    },
+    {
+      "epoch": 0.8849746463837737,
+      "grad_norm": 1.2210390039887022,
+      "learning_rate": 6.859714887007796e-07,
+      "loss": 1.5014,
+      "step": 6632
+    },
+    {
+      "epoch": 0.8851080864691754,
+      "grad_norm": 0.9534407665379254,
+      "learning_rate": 6.843992284192802e-07,
+      "loss": 1.5863,
+      "step": 6633
+    },
+    {
+      "epoch": 0.885241526554577,
+      "grad_norm": 0.961514819564435,
+      "learning_rate": 6.82828708185943e-07,
+      "loss": 1.5202,
+      "step": 6634
+    },
+    {
+      "epoch": 0.8853749666399786,
+      "grad_norm": 0.9797343158237848,
+      "learning_rate": 6.812599282941246e-07,
+      "loss": 1.5998,
+      "step": 6635
+    },
+    {
+      "epoch": 0.8855084067253803,
+      "grad_norm": 0.9254630062159546,
+      "learning_rate": 6.796928890368515e-07,
+      "loss": 1.5759,
+      "step": 6636
+    },
+    {
+      "epoch": 0.885641846810782,
+      "grad_norm": 0.921610927618888,
+      "learning_rate": 6.781275907068329e-07,
+      "loss": 1.5506,
+      "step": 6637
+    },
+    {
+      "epoch": 0.8857752868961836,
+      "grad_norm": 1.1965720258668404,
+      "learning_rate": 6.765640335964463e-07,
+      "loss": 1.4921,
+      "step": 6638
+    },
+    {
+      "epoch": 0.8859087269815853,
+      "grad_norm": 0.9199991742942989,
+      "learning_rate": 6.750022179977467e-07,
+      "loss": 1.5276,
+      "step": 6639
+    },
+    {
+      "epoch": 0.8860421670669869,
+      "grad_norm": 0.9414233305947609,
+      "learning_rate": 6.73442144202463e-07,
+      "loss": 1.5265,
+      "step": 6640
+    },
+    {
+      "epoch": 0.8861756071523885,
+      "grad_norm": 0.9303782628792104,
+      "learning_rate": 6.718838125020011e-07,
+      "loss": 1.4829,
+      "step": 6641
+    },
+    {
+      "epoch": 0.8863090472377902,
+      "grad_norm": 1.244221234348905,
+      "learning_rate": 6.703272231874391e-07,
+      "loss": 1.5792,
+      "step": 6642
+    },
+    {
+      "epoch": 0.8864424873231919,
+      "grad_norm": 0.9051064947661962,
+      "learning_rate": 6.687723765495268e-07,
+      "loss": 1.558,
+      "step": 6643
+    },
+    {
+      "epoch": 0.8865759274085936,
+      "grad_norm": 1.0602549890874733,
+      "learning_rate": 6.672192728786964e-07,
+      "loss": 1.5136,
+      "step": 6644
+    },
+    {
+      "epoch": 0.8867093674939952,
+      "grad_norm": 1.015376492992175,
+      "learning_rate": 6.656679124650489e-07,
+      "loss": 1.526,
+      "step": 6645
+    },
+    {
+      "epoch": 0.8868428075793968,
+      "grad_norm": 0.9763385623831438,
+      "learning_rate": 6.641182955983594e-07,
+      "loss": 1.5627,
+      "step": 6646
+    },
+    {
+      "epoch": 0.8869762476647985,
+      "grad_norm": 0.9484361059315187,
+      "learning_rate": 6.625704225680773e-07,
+      "loss": 1.5681,
+      "step": 6647
+    },
+    {
+      "epoch": 0.8871096877502002,
+      "grad_norm": 1.1736786559443448,
+      "learning_rate": 6.610242936633315e-07,
+      "loss": 1.52,
+      "step": 6648
+    },
+    {
+      "epoch": 0.8872431278356018,
+      "grad_norm": 1.0652253170073327,
+      "learning_rate": 6.594799091729198e-07,
+      "loss": 1.5487,
+      "step": 6649
+    },
+    {
+      "epoch": 0.8873765679210035,
+      "grad_norm": 0.9130108805332712,
+      "learning_rate": 6.579372693853137e-07,
+      "loss": 1.5838,
+      "step": 6650
+    },
+    {
+      "epoch": 0.8875100080064051,
+      "grad_norm": 0.9461766707449843,
+      "learning_rate": 6.563963745886637e-07,
+      "loss": 1.5557,
+      "step": 6651
+    },
+    {
+      "epoch": 0.8876434480918067,
+      "grad_norm": 0.9345577948179379,
+      "learning_rate": 6.548572250707896e-07,
+      "loss": 1.5397,
+      "step": 6652
+    },
+    {
+      "epoch": 0.8877768881772085,
+      "grad_norm": 0.9306784187277369,
+      "learning_rate": 6.533198211191871e-07,
+      "loss": 1.5641,
+      "step": 6653
+    },
+    {
+      "epoch": 0.8879103282626101,
+      "grad_norm": 0.9904047827430623,
+      "learning_rate": 6.517841630210254e-07,
+      "loss": 1.5339,
+      "step": 6654
+    },
+    {
+      "epoch": 0.8880437683480118,
+      "grad_norm": 0.9847791868831337,
+      "learning_rate": 6.502502510631503e-07,
+      "loss": 1.5608,
+      "step": 6655
+    },
+    {
+      "epoch": 0.8881772084334134,
+      "grad_norm": 1.2239512456758879,
+      "learning_rate": 6.487180855320762e-07,
+      "loss": 1.5556,
+      "step": 6656
+    },
+    {
+      "epoch": 0.888310648518815,
+      "grad_norm": 0.9473098956072885,
+      "learning_rate": 6.471876667139954e-07,
+      "loss": 1.5671,
+      "step": 6657
+    },
+    {
+      "epoch": 0.8884440886042168,
+      "grad_norm": 0.9150781295209052,
+      "learning_rate": 6.456589948947733e-07,
+      "loss": 1.4855,
+      "step": 6658
+    },
+    {
+      "epoch": 0.8885775286896184,
+      "grad_norm": 0.9262691372999762,
+      "learning_rate": 6.441320703599474e-07,
+      "loss": 1.5296,
+      "step": 6659
+    },
+    {
+      "epoch": 0.88871096877502,
+      "grad_norm": 1.2521772098087953,
+      "learning_rate": 6.426068933947305e-07,
+      "loss": 1.5772,
+      "step": 6660
+    },
+    {
+      "epoch": 0.8888444088604217,
+      "grad_norm": 0.9718277083295738,
+      "learning_rate": 6.410834642840069e-07,
+      "loss": 1.5483,
+      "step": 6661
+    },
+    {
+      "epoch": 0.8889778489458233,
+      "grad_norm": 0.9249506579829485,
+      "learning_rate": 6.395617833123379e-07,
+      "loss": 1.5262,
+      "step": 6662
+    },
+    {
+      "epoch": 0.8891112890312249,
+      "grad_norm": 0.991283938492636,
+      "learning_rate": 6.380418507639563e-07,
+      "loss": 1.5752,
+      "step": 6663
+    },
+    {
+      "epoch": 0.8892447291166267,
+      "grad_norm": 1.145717875907442,
+      "learning_rate": 6.365236669227659e-07,
+      "loss": 1.5203,
+      "step": 6664
+    },
+    {
+      "epoch": 0.8893781692020283,
+      "grad_norm": 0.9459042028552311,
+      "learning_rate": 6.350072320723477e-07,
+      "loss": 1.5782,
+      "step": 6665
+    },
+    {
+      "epoch": 0.8895116092874299,
+      "grad_norm": 0.951409036273284,
+      "learning_rate": 6.33492546495954e-07,
+      "loss": 1.5228,
+      "step": 6666
+    },
+    {
+      "epoch": 0.8896450493728316,
+      "grad_norm": 0.9376247682782157,
+      "learning_rate": 6.319796104765097e-07,
+      "loss": 1.5482,
+      "step": 6667
+    },
+    {
+      "epoch": 0.8897784894582332,
+      "grad_norm": 0.9343467840146905,
+      "learning_rate": 6.304684242966164e-07,
+      "loss": 1.5819,
+      "step": 6668
+    },
+    {
+      "epoch": 0.889911929543635,
+      "grad_norm": 0.9677052840154801,
+      "learning_rate": 6.289589882385461e-07,
+      "loss": 1.5253,
+      "step": 6669
+    },
+    {
+      "epoch": 0.8900453696290366,
+      "grad_norm": 0.9467119490943833,
+      "learning_rate": 6.274513025842421e-07,
+      "loss": 1.5582,
+      "step": 6670
+    },
+    {
+      "epoch": 0.8901788097144382,
+      "grad_norm": 0.9311722512072491,
+      "learning_rate": 6.25945367615326e-07,
+      "loss": 1.5084,
+      "step": 6671
+    },
+    {
+      "epoch": 0.8903122497998399,
+      "grad_norm": 1.1535295894188413,
+      "learning_rate": 6.244411836130848e-07,
+      "loss": 1.5288,
+      "step": 6672
+    },
+    {
+      "epoch": 0.8904456898852415,
+      "grad_norm": 0.9398613520690328,
+      "learning_rate": 6.229387508584872e-07,
+      "loss": 1.541,
+      "step": 6673
+    },
+    {
+      "epoch": 0.8905791299706431,
+      "grad_norm": 0.9460664090823528,
+      "learning_rate": 6.214380696321698e-07,
+      "loss": 1.5259,
+      "step": 6674
+    },
+    {
+      "epoch": 0.8907125700560449,
+      "grad_norm": 0.9192830875852988,
+      "learning_rate": 6.199391402144406e-07,
+      "loss": 1.5079,
+      "step": 6675
+    },
+    {
+      "epoch": 0.8908460101414465,
+      "grad_norm": 0.9624182903854707,
+      "learning_rate": 6.184419628852845e-07,
+      "loss": 1.514,
+      "step": 6676
+    },
+    {
+      "epoch": 0.8909794502268481,
+      "grad_norm": 0.9567443782327006,
+      "learning_rate": 6.169465379243578e-07,
+      "loss": 1.5349,
+      "step": 6677
+    },
+    {
+      "epoch": 0.8911128903122498,
+      "grad_norm": 1.173850544587069,
+      "learning_rate": 6.154528656109871e-07,
+      "loss": 1.5518,
+      "step": 6678
+    },
+    {
+      "epoch": 0.8912463303976514,
+      "grad_norm": 0.9397559763327206,
+      "learning_rate": 6.139609462241724e-07,
+      "loss": 1.524,
+      "step": 6679
+    },
+    {
+      "epoch": 0.8913797704830531,
+      "grad_norm": 0.9739117546526423,
+      "learning_rate": 6.124707800425911e-07,
+      "loss": 1.5663,
+      "step": 6680
+    },
+    {
+      "epoch": 0.8915132105684548,
+      "grad_norm": 0.9526569594291362,
+      "learning_rate": 6.109823673445869e-07,
+      "loss": 1.5959,
+      "step": 6681
+    },
+    {
+      "epoch": 0.8916466506538564,
+      "grad_norm": 1.2208803452356254,
+      "learning_rate": 6.094957084081765e-07,
+      "loss": 1.527,
+      "step": 6682
+    },
+    {
+      "epoch": 0.8917800907392581,
+      "grad_norm": 0.9533030946813461,
+      "learning_rate": 6.080108035110543e-07,
+      "loss": 1.5158,
+      "step": 6683
+    },
+    {
+      "epoch": 0.8919135308246597,
+      "grad_norm": 0.9373914483177651,
+      "learning_rate": 6.06527652930583e-07,
+      "loss": 1.5035,
+      "step": 6684
+    },
+    {
+      "epoch": 0.8920469709100614,
+      "grad_norm": 0.9206005814229758,
+      "learning_rate": 6.050462569437965e-07,
+      "loss": 1.5194,
+      "step": 6685
+    },
+    {
+      "epoch": 0.8921804109954631,
+      "grad_norm": 0.9411608802351985,
+      "learning_rate": 6.035666158274034e-07,
+      "loss": 1.5614,
+      "step": 6686
+    },
+    {
+      "epoch": 0.8923138510808647,
+      "grad_norm": 0.9549233585076925,
+      "learning_rate": 6.020887298577838e-07,
+      "loss": 1.524,
+      "step": 6687
+    },
+    {
+      "epoch": 0.8924472911662663,
+      "grad_norm": 1.198668371629655,
+      "learning_rate": 6.006125993109913e-07,
+      "loss": 1.5915,
+      "step": 6688
+    },
+    {
+      "epoch": 0.892580731251668,
+      "grad_norm": 0.9623095343289453,
+      "learning_rate": 5.991382244627475e-07,
+      "loss": 1.51,
+      "step": 6689
+    },
+    {
+      "epoch": 0.8927141713370697,
+      "grad_norm": 0.9678788722229147,
+      "learning_rate": 5.976656055884522e-07,
+      "loss": 1.5247,
+      "step": 6690
+    },
+    {
+      "epoch": 0.8928476114224713,
+      "grad_norm": 1.0279035033137902,
+      "learning_rate": 5.961947429631721e-07,
+      "loss": 1.5636,
+      "step": 6691
+    },
+    {
+      "epoch": 0.892981051507873,
+      "grad_norm": 0.9304167484501092,
+      "learning_rate": 5.947256368616483e-07,
+      "loss": 1.5038,
+      "step": 6692
+    },
+    {
+      "epoch": 0.8931144915932746,
+      "grad_norm": 0.9374804062894364,
+      "learning_rate": 5.932582875582904e-07,
+      "loss": 1.5582,
+      "step": 6693
+    },
+    {
+      "epoch": 0.8932479316786762,
+      "grad_norm": 0.9034683624937042,
+      "learning_rate": 5.917926953271857e-07,
+      "loss": 1.5244,
+      "step": 6694
+    },
+    {
+      "epoch": 0.893381371764078,
+      "grad_norm": 1.0506006939018857,
+      "learning_rate": 5.903288604420887e-07,
+      "loss": 1.5336,
+      "step": 6695
+    },
+    {
+      "epoch": 0.8935148118494796,
+      "grad_norm": 0.9685577008965324,
+      "learning_rate": 5.888667831764283e-07,
+      "loss": 1.5201,
+      "step": 6696
+    },
+    {
+      "epoch": 0.8936482519348813,
+      "grad_norm": 0.9546554805063521,
+      "learning_rate": 5.874064638033017e-07,
+      "loss": 1.5315,
+      "step": 6697
+    },
+    {
+      "epoch": 0.8937816920202829,
+      "grad_norm": 0.9383844169484069,
+      "learning_rate": 5.859479025954817e-07,
+      "loss": 1.5399,
+      "step": 6698
+    },
+    {
+      "epoch": 0.8939151321056845,
+      "grad_norm": 1.0001219462580064,
+      "learning_rate": 5.844910998254117e-07,
+      "loss": 1.534,
+      "step": 6699
+    },
+    {
+      "epoch": 0.8940485721910862,
+      "grad_norm": 1.0514709520939773,
+      "learning_rate": 5.830360557652026e-07,
+      "loss": 1.5436,
+      "step": 6700
+    },
+    {
+      "epoch": 0.8941820122764879,
+      "grad_norm": 0.9603464211449854,
+      "learning_rate": 5.815827706866439e-07,
+      "loss": 1.6296,
+      "step": 6701
+    },
+    {
+      "epoch": 0.8943154523618895,
+      "grad_norm": 0.9699909980766925,
+      "learning_rate": 5.801312448611907e-07,
+      "loss": 1.5712,
+      "step": 6702
+    },
+    {
+      "epoch": 0.8944488924472912,
+      "grad_norm": 0.9081518549220741,
+      "learning_rate": 5.786814785599715e-07,
+      "loss": 1.5363,
+      "step": 6703
+    },
+    {
+      "epoch": 0.8945823325326928,
+      "grad_norm": 0.9477859920909177,
+      "learning_rate": 5.772334720537854e-07,
+      "loss": 1.5438,
+      "step": 6704
+    },
+    {
+      "epoch": 0.8947157726180944,
+      "grad_norm": 0.9467954261018504,
+      "learning_rate": 5.757872256131048e-07,
+      "loss": 1.5359,
+      "step": 6705
+    },
+    {
+      "epoch": 0.8948492127034962,
+      "grad_norm": 1.0635463978582553,
+      "learning_rate": 5.743427395080736e-07,
+      "loss": 1.5161,
+      "step": 6706
+    },
+    {
+      "epoch": 0.8949826527888978,
+      "grad_norm": 0.9139015356788128,
+      "learning_rate": 5.729000140085017e-07,
+      "loss": 1.5263,
+      "step": 6707
+    },
+    {
+      "epoch": 0.8951160928742994,
+      "grad_norm": 1.2623806658299492,
+      "learning_rate": 5.714590493838768e-07,
+      "loss": 1.5617,
+      "step": 6708
+    },
+    {
+      "epoch": 0.8952495329597011,
+      "grad_norm": 1.0270566802124301,
+      "learning_rate": 5.700198459033535e-07,
+      "loss": 1.5226,
+      "step": 6709
+    },
+    {
+      "epoch": 0.8953829730451027,
+      "grad_norm": 0.9557701098327668,
+      "learning_rate": 5.685824038357568e-07,
+      "loss": 1.554,
+      "step": 6710
+    },
+    {
+      "epoch": 0.8955164131305045,
+      "grad_norm": 0.8951591320979515,
+      "learning_rate": 5.671467234495875e-07,
+      "loss": 1.5101,
+      "step": 6711
+    },
+    {
+      "epoch": 0.8956498532159061,
+      "grad_norm": 0.9290933211646522,
+      "learning_rate": 5.657128050130134e-07,
+      "loss": 1.5454,
+      "step": 6712
+    },
+    {
+      "epoch": 0.8957832933013077,
+      "grad_norm": 0.935571799488784,
+      "learning_rate": 5.642806487938746e-07,
+      "loss": 1.6001,
+      "step": 6713
+    },
+    {
+      "epoch": 0.8959167333867094,
+      "grad_norm": 0.967264821795858,
+      "learning_rate": 5.628502550596781e-07,
+      "loss": 1.4861,
+      "step": 6714
+    },
+    {
+      "epoch": 0.896050173472111,
+      "grad_norm": 0.9487698266935534,
+      "learning_rate": 5.614216240776105e-07,
+      "loss": 1.5693,
+      "step": 6715
+    },
+    {
+      "epoch": 0.8961836135575126,
+      "grad_norm": 1.1670691771030441,
+      "learning_rate": 5.599947561145214e-07,
+      "loss": 1.4934,
+      "step": 6716
+    },
+    {
+      "epoch": 0.8963170536429144,
+      "grad_norm": 0.960589734890621,
+      "learning_rate": 5.585696514369321e-07,
+      "loss": 1.5477,
+      "step": 6717
+    },
+    {
+      "epoch": 0.896450493728316,
+      "grad_norm": 0.935167699678419,
+      "learning_rate": 5.571463103110375e-07,
+      "loss": 1.4821,
+      "step": 6718
+    },
+    {
+      "epoch": 0.8965839338137176,
+      "grad_norm": 0.9947979187682457,
+      "learning_rate": 5.557247330027016e-07,
+      "loss": 1.5684,
+      "step": 6719
+    },
+    {
+      "epoch": 0.8967173738991193,
+      "grad_norm": 0.9593305751018268,
+      "learning_rate": 5.543049197774586e-07,
+      "loss": 1.5996,
+      "step": 6720
+    },
+    {
+      "epoch": 0.8968508139845209,
+      "grad_norm": 0.9375856914820678,
+      "learning_rate": 5.52886870900512e-07,
+      "loss": 1.5154,
+      "step": 6721
+    },
+    {
+      "epoch": 0.8969842540699227,
+      "grad_norm": 1.2862051330318942,
+      "learning_rate": 5.514705866367387e-07,
+      "loss": 1.5196,
+      "step": 6722
+    },
+    {
+      "epoch": 0.8971176941553243,
+      "grad_norm": 0.9340659004674106,
+      "learning_rate": 5.500560672506861e-07,
+      "loss": 1.4784,
+      "step": 6723
+    },
+    {
+      "epoch": 0.8972511342407259,
+      "grad_norm": 0.945137802994412,
+      "learning_rate": 5.486433130065672e-07,
+      "loss": 1.5399,
+      "step": 6724
+    },
+    {
+      "epoch": 0.8973845743261276,
+      "grad_norm": 1.0209524963685872,
+      "learning_rate": 5.472323241682687e-07,
+      "loss": 1.5503,
+      "step": 6725
+    },
+    {
+      "epoch": 0.8975180144115292,
+      "grad_norm": 0.927685929338529,
+      "learning_rate": 5.458231009993498e-07,
+      "loss": 1.5844,
+      "step": 6726
+    },
+    {
+      "epoch": 0.8976514544969308,
+      "grad_norm": 0.9428359886770269,
+      "learning_rate": 5.444156437630365e-07,
+      "loss": 1.5389,
+      "step": 6727
+    },
+    {
+      "epoch": 0.8977848945823326,
+      "grad_norm": 0.9334863494390306,
+      "learning_rate": 5.430099527222244e-07,
+      "loss": 1.5162,
+      "step": 6728
+    },
+    {
+      "epoch": 0.8979183346677342,
+      "grad_norm": 0.9436338050010254,
+      "learning_rate": 5.416060281394797e-07,
+      "loss": 1.5382,
+      "step": 6729
+    },
+    {
+      "epoch": 0.8980517747531358,
+      "grad_norm": 0.9365789196387994,
+      "learning_rate": 5.402038702770418e-07,
+      "loss": 1.5249,
+      "step": 6730
+    },
+    {
+      "epoch": 0.8981852148385375,
+      "grad_norm": 0.9221528407309991,
+      "learning_rate": 5.388034793968189e-07,
+      "loss": 1.5386,
+      "step": 6731
+    },
+    {
+      "epoch": 0.8983186549239391,
+      "grad_norm": 0.9515618324069464,
+      "learning_rate": 5.374048557603828e-07,
+      "loss": 1.5362,
+      "step": 6732
+    },
+    {
+      "epoch": 0.8984520950093408,
+      "grad_norm": 0.9944556911301887,
+      "learning_rate": 5.360079996289869e-07,
+      "loss": 1.5167,
+      "step": 6733
+    },
+    {
+      "epoch": 0.8985855350947425,
+      "grad_norm": 0.97850772806653,
+      "learning_rate": 5.346129112635445e-07,
+      "loss": 1.5102,
+      "step": 6734
+    },
+    {
+      "epoch": 0.8987189751801441,
+      "grad_norm": 0.9540868569995883,
+      "learning_rate": 5.332195909246429e-07,
+      "loss": 1.5752,
+      "step": 6735
+    },
+    {
+      "epoch": 0.8988524152655458,
+      "grad_norm": 0.9221941861156301,
+      "learning_rate": 5.318280388725372e-07,
+      "loss": 1.5118,
+      "step": 6736
+    },
+    {
+      "epoch": 0.8989858553509474,
+      "grad_norm": 1.0270466084981307,
+      "learning_rate": 5.304382553671561e-07,
+      "loss": 1.5436,
+      "step": 6737
+    },
+    {
+      "epoch": 0.899119295436349,
+      "grad_norm": 0.9203067458998816,
+      "learning_rate": 5.290502406680931e-07,
+      "loss": 1.5437,
+      "step": 6738
+    },
+    {
+      "epoch": 0.8992527355217508,
+      "grad_norm": 0.9501536472353344,
+      "learning_rate": 5.276639950346129e-07,
+      "loss": 1.5075,
+      "step": 6739
+    },
+    {
+      "epoch": 0.8993861756071524,
+      "grad_norm": 0.9334210192752896,
+      "learning_rate": 5.262795187256542e-07,
+      "loss": 1.5422,
+      "step": 6740
+    },
+    {
+      "epoch": 0.899519615692554,
+      "grad_norm": 0.9567078114398553,
+      "learning_rate": 5.248968119998188e-07,
+      "loss": 1.5471,
+      "step": 6741
+    },
+    {
+      "epoch": 0.8996530557779557,
+      "grad_norm": 0.9282038591345071,
+      "learning_rate": 5.235158751153801e-07,
+      "loss": 1.4623,
+      "step": 6742
+    },
+    {
+      "epoch": 0.8997864958633573,
+      "grad_norm": 0.9384691548104735,
+      "learning_rate": 5.22136708330282e-07,
+      "loss": 1.5249,
+      "step": 6743
+    },
+    {
+      "epoch": 0.899919935948759,
+      "grad_norm": 1.0717530433261342,
+      "learning_rate": 5.20759311902137e-07,
+      "loss": 1.5446,
+      "step": 6744
+    },
+    {
+      "epoch": 0.9000533760341607,
+      "grad_norm": 0.949517161915402,
+      "learning_rate": 5.193836860882284e-07,
+      "loss": 1.5237,
+      "step": 6745
+    },
+    {
+      "epoch": 0.9001868161195623,
+      "grad_norm": 0.9209721246022604,
+      "learning_rate": 5.180098311455051e-07,
+      "loss": 1.4865,
+      "step": 6746
+    },
+    {
+      "epoch": 0.9003202562049639,
+      "grad_norm": 1.0530255691032593,
+      "learning_rate": 5.166377473305894e-07,
+      "loss": 1.5196,
+      "step": 6747
+    },
+    {
+      "epoch": 0.9004536962903656,
+      "grad_norm": 0.9330131313465537,
+      "learning_rate": 5.15267434899771e-07,
+      "loss": 1.5512,
+      "step": 6748
+    },
+    {
+      "epoch": 0.9005871363757673,
+      "grad_norm": 0.9202256141014292,
+      "learning_rate": 5.138988941090084e-07,
+      "loss": 1.5196,
+      "step": 6749
+    },
+    {
+      "epoch": 0.900720576461169,
+      "grad_norm": 0.9650715822764238,
+      "learning_rate": 5.125321252139282e-07,
+      "loss": 1.5177,
+      "step": 6750
+    },
+    {
+      "epoch": 0.9008540165465706,
+      "grad_norm": 0.9433756615509346,
+      "learning_rate": 5.111671284698283e-07,
+      "loss": 1.5432,
+      "step": 6751
+    },
+    {
+      "epoch": 0.9009874566319722,
+      "grad_norm": 0.9533022305140074,
+      "learning_rate": 5.098039041316738e-07,
+      "loss": 1.5537,
+      "step": 6752
+    },
+    {
+      "epoch": 0.9011208967173739,
+      "grad_norm": 0.9346416532826066,
+      "learning_rate": 5.084424524540999e-07,
+      "loss": 1.5495,
+      "step": 6753
+    },
+    {
+      "epoch": 0.9012543368027756,
+      "grad_norm": 1.1020333932337052,
+      "learning_rate": 5.070827736914119e-07,
+      "loss": 1.5476,
+      "step": 6754
+    },
+    {
+      "epoch": 0.9013877768881772,
+      "grad_norm": 0.9410283837732905,
+      "learning_rate": 5.057248680975802e-07,
+      "loss": 1.5037,
+      "step": 6755
+    },
+    {
+      "epoch": 0.9015212169735789,
+      "grad_norm": 1.0772876014109762,
+      "learning_rate": 5.043687359262472e-07,
+      "loss": 1.5047,
+      "step": 6756
+    },
+    {
+      "epoch": 0.9016546570589805,
+      "grad_norm": 0.9439200358382797,
+      "learning_rate": 5.030143774307205e-07,
+      "loss": 1.5584,
+      "step": 6757
+    },
+    {
+      "epoch": 0.9017880971443821,
+      "grad_norm": 1.0161041213410495,
+      "learning_rate": 5.016617928639822e-07,
+      "loss": 1.5249,
+      "step": 6758
+    },
+    {
+      "epoch": 0.9019215372297839,
+      "grad_norm": 0.9412328793676751,
+      "learning_rate": 5.003109824786789e-07,
+      "loss": 1.5376,
+      "step": 6759
+    },
+    {
+      "epoch": 0.9020549773151855,
+      "grad_norm": 0.9639374913924775,
+      "learning_rate": 4.989619465271245e-07,
+      "loss": 1.5075,
+      "step": 6760
+    },
+    {
+      "epoch": 0.9021884174005871,
+      "grad_norm": 0.9124862517814768,
+      "learning_rate": 4.976146852613062e-07,
+      "loss": 1.5236,
+      "step": 6761
+    },
+    {
+      "epoch": 0.9023218574859888,
+      "grad_norm": 0.9328183814360856,
+      "learning_rate": 4.96269198932876e-07,
+      "loss": 1.5257,
+      "step": 6762
+    },
+    {
+      "epoch": 0.9024552975713904,
+      "grad_norm": 0.9480856180023016,
+      "learning_rate": 4.949254877931564e-07,
+      "loss": 1.6097,
+      "step": 6763
+    },
+    {
+      "epoch": 0.9025887376567922,
+      "grad_norm": 0.9452684747563659,
+      "learning_rate": 4.935835520931342e-07,
+      "loss": 1.5583,
+      "step": 6764
+    },
+    {
+      "epoch": 0.9027221777421938,
+      "grad_norm": 0.9435189294655035,
+      "learning_rate": 4.922433920834713e-07,
+      "loss": 1.5576,
+      "step": 6765
+    },
+    {
+      "epoch": 0.9028556178275954,
+      "grad_norm": 0.9100774226094155,
+      "learning_rate": 4.909050080144928e-07,
+      "loss": 1.5061,
+      "step": 6766
+    },
+    {
+      "epoch": 0.9029890579129971,
+      "grad_norm": 0.9188166446494538,
+      "learning_rate": 4.895684001361933e-07,
+      "loss": 1.525,
+      "step": 6767
+    },
+    {
+      "epoch": 0.9031224979983987,
+      "grad_norm": 0.9513533676277205,
+      "learning_rate": 4.882335686982354e-07,
+      "loss": 1.5194,
+      "step": 6768
+    },
+    {
+      "epoch": 0.9032559380838003,
+      "grad_norm": 0.9901332701917303,
+      "learning_rate": 4.869005139499528e-07,
+      "loss": 1.5135,
+      "step": 6769
+    },
+    {
+      "epoch": 0.9033893781692021,
+      "grad_norm": 0.9252469459313303,
+      "learning_rate": 4.85569236140343e-07,
+      "loss": 1.5236,
+      "step": 6770
+    },
+    {
+      "epoch": 0.9035228182546037,
+      "grad_norm": 1.0114758784078768,
+      "learning_rate": 4.842397355180728e-07,
+      "loss": 1.5166,
+      "step": 6771
+    },
+    {
+      "epoch": 0.9036562583400053,
+      "grad_norm": 0.972194230662802,
+      "learning_rate": 4.829120123314801e-07,
+      "loss": 1.5787,
+      "step": 6772
+    },
+    {
+      "epoch": 0.903789698425407,
+      "grad_norm": 0.917284388104258,
+      "learning_rate": 4.815860668285688e-07,
+      "loss": 1.5353,
+      "step": 6773
+    },
+    {
+      "epoch": 0.9039231385108086,
+      "grad_norm": 0.9549812627537546,
+      "learning_rate": 4.802618992570074e-07,
+      "loss": 1.545,
+      "step": 6774
+    },
+    {
+      "epoch": 0.9040565785962102,
+      "grad_norm": 1.052982946218394,
+      "learning_rate": 4.789395098641359e-07,
+      "loss": 1.5932,
+      "step": 6775
+    },
+    {
+      "epoch": 0.904190018681612,
+      "grad_norm": 1.1732404735888684,
+      "learning_rate": 4.776188988969643e-07,
+      "loss": 1.4899,
+      "step": 6776
+    },
+    {
+      "epoch": 0.9043234587670136,
+      "grad_norm": 1.0066806501063883,
+      "learning_rate": 4.7630006660216665e-07,
+      "loss": 1.5585,
+      "step": 6777
+    },
+    {
+      "epoch": 0.9044568988524153,
+      "grad_norm": 0.9579935580737865,
+      "learning_rate": 4.749830132260824e-07,
+      "loss": 1.5712,
+      "step": 6778
+    },
+    {
+      "epoch": 0.9045903389378169,
+      "grad_norm": 0.9169006378445536,
+      "learning_rate": 4.736677390147271e-07,
+      "loss": 1.5371,
+      "step": 6779
+    },
+    {
+      "epoch": 0.9047237790232185,
+      "grad_norm": 0.9413950267463799,
+      "learning_rate": 4.723542442137774e-07,
+      "loss": 1.5586,
+      "step": 6780
+    },
+    {
+      "epoch": 0.9048572191086203,
+      "grad_norm": 0.9375274450226259,
+      "learning_rate": 4.710425290685772e-07,
+      "loss": 1.526,
+      "step": 6781
+    },
+    {
+      "epoch": 0.9049906591940219,
+      "grad_norm": 0.920951060703472,
+      "learning_rate": 4.697325938241404e-07,
+      "loss": 1.5496,
+      "step": 6782
+    },
+    {
+      "epoch": 0.9051240992794235,
+      "grad_norm": 0.9705124398869889,
+      "learning_rate": 4.684244387251513e-07,
+      "loss": 1.552,
+      "step": 6783
+    },
+    {
+      "epoch": 0.9052575393648252,
+      "grad_norm": 0.9307149713548113,
+      "learning_rate": 4.671180640159545e-07,
+      "loss": 1.5031,
+      "step": 6784
+    },
+    {
+      "epoch": 0.9053909794502268,
+      "grad_norm": 1.0044776590484505,
+      "learning_rate": 4.65813469940567e-07,
+      "loss": 1.5668,
+      "step": 6785
+    },
+    {
+      "epoch": 0.9055244195356285,
+      "grad_norm": 0.8995467573403738,
+      "learning_rate": 4.645106567426738e-07,
+      "loss": 1.4998,
+      "step": 6786
+    },
+    {
+      "epoch": 0.9056578596210302,
+      "grad_norm": 0.9441419162843016,
+      "learning_rate": 4.632096246656237e-07,
+      "loss": 1.5544,
+      "step": 6787
+    },
+    {
+      "epoch": 0.9057912997064318,
+      "grad_norm": 1.37214339045131,
+      "learning_rate": 4.619103739524355e-07,
+      "loss": 1.5009,
+      "step": 6788
+    },
+    {
+      "epoch": 0.9059247397918335,
+      "grad_norm": 0.9569798512304817,
+      "learning_rate": 4.6061290484579304e-07,
+      "loss": 1.5217,
+      "step": 6789
+    },
+    {
+      "epoch": 0.9060581798772351,
+      "grad_norm": 0.9238396452899695,
+      "learning_rate": 4.5931721758805117e-07,
+      "loss": 1.5632,
+      "step": 6790
+    },
+    {
+      "epoch": 0.9061916199626368,
+      "grad_norm": 0.9533994593848583,
+      "learning_rate": 4.5802331242122855e-07,
+      "loss": 1.559,
+      "step": 6791
+    },
+    {
+      "epoch": 0.9063250600480385,
+      "grad_norm": 1.0268411283902035,
+      "learning_rate": 4.567311895870108e-07,
+      "loss": 1.517,
+      "step": 6792
+    },
+    {
+      "epoch": 0.9064585001334401,
+      "grad_norm": 1.0100719989531157,
+      "learning_rate": 4.554408493267537e-07,
+      "loss": 1.5145,
+      "step": 6793
+    },
+    {
+      "epoch": 0.9065919402188417,
+      "grad_norm": 1.1648731959473748,
+      "learning_rate": 4.5415229188147667e-07,
+      "loss": 1.5143,
+      "step": 6794
+    },
+    {
+      "epoch": 0.9067253803042434,
+      "grad_norm": 0.92959204449064,
+      "learning_rate": 4.5286551749186726e-07,
+      "loss": 1.5744,
+      "step": 6795
+    },
+    {
+      "epoch": 0.906858820389645,
+      "grad_norm": 0.9435972338142883,
+      "learning_rate": 4.5158052639828085e-07,
+      "loss": 1.5309,
+      "step": 6796
+    },
+    {
+      "epoch": 0.9069922604750467,
+      "grad_norm": 1.0579708377998769,
+      "learning_rate": 4.5029731884074004e-07,
+      "loss": 1.5452,
+      "step": 6797
+    },
+    {
+      "epoch": 0.9071257005604484,
+      "grad_norm": 1.6099878391111002,
+      "learning_rate": 4.4901589505893297e-07,
+      "loss": 1.535,
+      "step": 6798
+    },
+    {
+      "epoch": 0.90725914064585,
+      "grad_norm": 0.9431159488281261,
+      "learning_rate": 4.4773625529221486e-07,
+      "loss": 1.5271,
+      "step": 6799
+    },
+    {
+      "epoch": 0.9073925807312516,
+      "grad_norm": 0.9302835803444568,
+      "learning_rate": 4.464583997796057e-07,
+      "loss": 1.5343,
+      "step": 6800
+    },
+    {
+      "epoch": 0.9075260208166533,
+      "grad_norm": 0.9727837080368901,
+      "learning_rate": 4.451823287597978e-07,
+      "loss": 1.5523,
+      "step": 6801
+    },
+    {
+      "epoch": 0.907659460902055,
+      "grad_norm": 0.9311133151109089,
+      "learning_rate": 4.4390804247114503e-07,
+      "loss": 1.5352,
+      "step": 6802
+    },
+    {
+      "epoch": 0.9077929009874567,
+      "grad_norm": 0.9834944210354547,
+      "learning_rate": 4.4263554115166805e-07,
+      "loss": 1.5093,
+      "step": 6803
+    },
+    {
+      "epoch": 0.9079263410728583,
+      "grad_norm": 0.924803540849428,
+      "learning_rate": 4.413648250390601e-07,
+      "loss": 1.5702,
+      "step": 6804
+    },
+    {
+      "epoch": 0.9080597811582599,
+      "grad_norm": 1.0516475698411472,
+      "learning_rate": 4.400958943706724e-07,
+      "loss": 1.566,
+      "step": 6805
+    },
+    {
+      "epoch": 0.9081932212436616,
+      "grad_norm": 1.0934743180424145,
+      "learning_rate": 4.388287493835286e-07,
+      "loss": 1.5735,
+      "step": 6806
+    },
+    {
+      "epoch": 0.9083266613290633,
+      "grad_norm": 0.9552392103374737,
+      "learning_rate": 4.3756339031431394e-07,
+      "loss": 1.563,
+      "step": 6807
+    },
+    {
+      "epoch": 0.9084601014144649,
+      "grad_norm": 0.9653006612408468,
+      "learning_rate": 4.362998173993882e-07,
+      "loss": 1.5681,
+      "step": 6808
+    },
+    {
+      "epoch": 0.9085935414998666,
+      "grad_norm": 0.9718108077335248,
+      "learning_rate": 4.3503803087476926e-07,
+      "loss": 1.5478,
+      "step": 6809
+    },
+    {
+      "epoch": 0.9087269815852682,
+      "grad_norm": 0.9602753698421362,
+      "learning_rate": 4.337780309761441e-07,
+      "loss": 1.554,
+      "step": 6810
+    },
+    {
+      "epoch": 0.9088604216706698,
+      "grad_norm": 0.9424771815385352,
+      "learning_rate": 4.3251981793886787e-07,
+      "loss": 1.6088,
+      "step": 6811
+    },
+    {
+      "epoch": 0.9089938617560716,
+      "grad_norm": 0.9274015813847016,
+      "learning_rate": 4.312633919979603e-07,
+      "loss": 1.5803,
+      "step": 6812
+    },
+    {
+      "epoch": 0.9091273018414732,
+      "grad_norm": 0.9230912583513253,
+      "learning_rate": 4.300087533881059e-07,
+      "loss": 1.5684,
+      "step": 6813
+    },
+    {
+      "epoch": 0.9092607419268748,
+      "grad_norm": 0.9305252391203362,
+      "learning_rate": 4.2875590234365825e-07,
+      "loss": 1.4863,
+      "step": 6814
+    },
+    {
+      "epoch": 0.9093941820122765,
+      "grad_norm": 0.9029270378762562,
+      "learning_rate": 4.2750483909863584e-07,
+      "loss": 1.5725,
+      "step": 6815
+    },
+    {
+      "epoch": 0.9095276220976781,
+      "grad_norm": 0.9383970397120545,
+      "learning_rate": 4.2625556388672163e-07,
+      "loss": 1.5339,
+      "step": 6816
+    },
+    {
+      "epoch": 0.9096610621830798,
+      "grad_norm": 0.9345725438722642,
+      "learning_rate": 4.2500807694126677e-07,
+      "loss": 1.5505,
+      "step": 6817
+    },
+    {
+      "epoch": 0.9097945022684815,
+      "grad_norm": 0.9259234911532692,
+      "learning_rate": 4.2376237849528936e-07,
+      "loss": 1.5278,
+      "step": 6818
+    },
+    {
+      "epoch": 0.9099279423538831,
+      "grad_norm": 1.0476543726975824,
+      "learning_rate": 4.2251846878146873e-07,
+      "loss": 1.5572,
+      "step": 6819
+    },
+    {
+      "epoch": 0.9100613824392848,
+      "grad_norm": 0.9337601320766555,
+      "learning_rate": 4.2127634803215576e-07,
+      "loss": 1.5409,
+      "step": 6820
+    },
+    {
+      "epoch": 0.9101948225246864,
+      "grad_norm": 0.9335871299642117,
+      "learning_rate": 4.2003601647936156e-07,
+      "loss": 1.5401,
+      "step": 6821
+    },
+    {
+      "epoch": 0.910328262610088,
+      "grad_norm": 0.9622149934684122,
+      "learning_rate": 4.187974743547674e-07,
+      "loss": 1.5137,
+      "step": 6822
+    },
+    {
+      "epoch": 0.9104617026954898,
+      "grad_norm": 0.9148887346642454,
+      "learning_rate": 4.175607218897204e-07,
+      "loss": 1.5505,
+      "step": 6823
+    },
+    {
+      "epoch": 0.9105951427808914,
+      "grad_norm": 1.0530922369152327,
+      "learning_rate": 4.1632575931522923e-07,
+      "loss": 1.5171,
+      "step": 6824
+    },
+    {
+      "epoch": 0.910728582866293,
+      "grad_norm": 0.9530413630746577,
+      "learning_rate": 4.150925868619726e-07,
+      "loss": 1.6038,
+      "step": 6825
+    },
+    {
+      "epoch": 0.9108620229516947,
+      "grad_norm": 0.9248293389790625,
+      "learning_rate": 4.138612047602919e-07,
+      "loss": 1.53,
+      "step": 6826
+    },
+    {
+      "epoch": 0.9109954630370963,
+      "grad_norm": 0.9435050547492767,
+      "learning_rate": 4.126316132401975e-07,
+      "loss": 1.5497,
+      "step": 6827
+    },
+    {
+      "epoch": 0.911128903122498,
+      "grad_norm": 0.9172974815783576,
+      "learning_rate": 4.1140381253135907e-07,
+      "loss": 1.53,
+      "step": 6828
+    },
+    {
+      "epoch": 0.9112623432078997,
+      "grad_norm": 0.9198332491276605,
+      "learning_rate": 4.101778028631198e-07,
+      "loss": 1.4973,
+      "step": 6829
+    },
+    {
+      "epoch": 0.9113957832933013,
+      "grad_norm": 0.9340095445157393,
+      "learning_rate": 4.0895358446448205e-07,
+      "loss": 1.5311,
+      "step": 6830
+    },
+    {
+      "epoch": 0.911529223378703,
+      "grad_norm": 0.9218669809304821,
+      "learning_rate": 4.077311575641174e-07,
+      "loss": 1.4977,
+      "step": 6831
+    },
+    {
+      "epoch": 0.9116626634641046,
+      "grad_norm": 0.9386440639251815,
+      "learning_rate": 4.0651052239035873e-07,
+      "loss": 1.4714,
+      "step": 6832
+    },
+    {
+      "epoch": 0.9117961035495062,
+      "grad_norm": 0.9188422179185332,
+      "learning_rate": 4.0529167917121024e-07,
+      "loss": 1.5086,
+      "step": 6833
+    },
+    {
+      "epoch": 0.911929543634908,
+      "grad_norm": 0.9929237105805191,
+      "learning_rate": 4.040746281343355e-07,
+      "loss": 1.5435,
+      "step": 6834
+    },
+    {
+      "epoch": 0.9120629837203096,
+      "grad_norm": 0.9262323335599812,
+      "learning_rate": 4.0285936950706705e-07,
+      "loss": 1.543,
+      "step": 6835
+    },
+    {
+      "epoch": 0.9121964238057112,
+      "grad_norm": 0.9413275796185295,
+      "learning_rate": 4.016459035164e-07,
+      "loss": 1.5475,
+      "step": 6836
+    },
+    {
+      "epoch": 0.9123298638911129,
+      "grad_norm": 0.9355976771543438,
+      "learning_rate": 4.0043423038899743e-07,
+      "loss": 1.5646,
+      "step": 6837
+    },
+    {
+      "epoch": 0.9124633039765145,
+      "grad_norm": 0.9308452239118499,
+      "learning_rate": 3.99224350351185e-07,
+      "loss": 1.5282,
+      "step": 6838
+    },
+    {
+      "epoch": 0.9125967440619162,
+      "grad_norm": 0.9352299931869122,
+      "learning_rate": 3.9801626362895527e-07,
+      "loss": 1.5478,
+      "step": 6839
+    },
+    {
+      "epoch": 0.9127301841473179,
+      "grad_norm": 0.920552426822107,
+      "learning_rate": 3.9680997044796445e-07,
+      "loss": 1.571,
+      "step": 6840
+    },
+    {
+      "epoch": 0.9128636242327195,
+      "grad_norm": 0.9614541460432983,
+      "learning_rate": 3.9560547103353553e-07,
+      "loss": 1.5716,
+      "step": 6841
+    },
+    {
+      "epoch": 0.9129970643181211,
+      "grad_norm": 0.921161207602936,
+      "learning_rate": 3.9440276561065306e-07,
+      "loss": 1.5062,
+      "step": 6842
+    },
+    {
+      "epoch": 0.9131305044035228,
+      "grad_norm": 0.9298591480513266,
+      "learning_rate": 3.932018544039717e-07,
+      "loss": 1.5823,
+      "step": 6843
+    },
+    {
+      "epoch": 0.9132639444889245,
+      "grad_norm": 0.9522285441788466,
+      "learning_rate": 3.920027376378055e-07,
+      "loss": 1.5636,
+      "step": 6844
+    },
+    {
+      "epoch": 0.9133973845743262,
+      "grad_norm": 0.9048091631898952,
+      "learning_rate": 3.9080541553613737e-07,
+      "loss": 1.5236,
+      "step": 6845
+    },
+    {
+      "epoch": 0.9135308246597278,
+      "grad_norm": 0.8997967100666988,
+      "learning_rate": 3.8960988832261184e-07,
+      "loss": 1.543,
+      "step": 6846
+    },
+    {
+      "epoch": 0.9136642647451294,
+      "grad_norm": 0.9344054356817217,
+      "learning_rate": 3.884161562205413e-07,
+      "loss": 1.5238,
+      "step": 6847
+    },
+    {
+      "epoch": 0.9137977048305311,
+      "grad_norm": 0.9433085668577765,
+      "learning_rate": 3.872242194529019e-07,
+      "loss": 1.5323,
+      "step": 6848
+    },
+    {
+      "epoch": 0.9139311449159327,
+      "grad_norm": 0.9298922807904612,
+      "learning_rate": 3.8603407824233e-07,
+      "loss": 1.5804,
+      "step": 6849
+    },
+    {
+      "epoch": 0.9140645850013344,
+      "grad_norm": 0.916116285836366,
+      "learning_rate": 3.8484573281113546e-07,
+      "loss": 1.5167,
+      "step": 6850
+    },
+    {
+      "epoch": 0.9141980250867361,
+      "grad_norm": 1.0047053088456148,
+      "learning_rate": 3.8365918338128525e-07,
+      "loss": 1.5509,
+      "step": 6851
+    },
+    {
+      "epoch": 0.9143314651721377,
+      "grad_norm": 0.9283300536875279,
+      "learning_rate": 3.8247443017441323e-07,
+      "loss": 1.5111,
+      "step": 6852
+    },
+    {
+      "epoch": 0.9144649052575393,
+      "grad_norm": 1.1042519917914184,
+      "learning_rate": 3.812914734118167e-07,
+      "loss": 1.5714,
+      "step": 6853
+    },
+    {
+      "epoch": 0.914598345342941,
+      "grad_norm": 0.9081858321179135,
+      "learning_rate": 3.8011031331446125e-07,
+      "loss": 1.5021,
+      "step": 6854
+    },
+    {
+      "epoch": 0.9147317854283427,
+      "grad_norm": 0.9771021703620095,
+      "learning_rate": 3.7893095010297255e-07,
+      "loss": 1.5391,
+      "step": 6855
+    },
+    {
+      "epoch": 0.9148652255137443,
+      "grad_norm": 0.9348735621584969,
+      "learning_rate": 3.7775338399764106e-07,
+      "loss": 1.5353,
+      "step": 6856
+    },
+    {
+      "epoch": 0.914998665599146,
+      "grad_norm": 1.217330780323833,
+      "learning_rate": 3.765776152184264e-07,
+      "loss": 1.5584,
+      "step": 6857
+    },
+    {
+      "epoch": 0.9151321056845476,
+      "grad_norm": 0.9415726884299152,
+      "learning_rate": 3.754036439849451e-07,
+      "loss": 1.5495,
+      "step": 6858
+    },
+    {
+      "epoch": 0.9152655457699493,
+      "grad_norm": 0.9466565821101811,
+      "learning_rate": 3.742314705164829e-07,
+      "loss": 1.5281,
+      "step": 6859
+    },
+    {
+      "epoch": 0.915398985855351,
+      "grad_norm": 0.9439757015604497,
+      "learning_rate": 3.7306109503198797e-07,
+      "loss": 1.6172,
+      "step": 6860
+    },
+    {
+      "epoch": 0.9155324259407526,
+      "grad_norm": 0.9358623022886202,
+      "learning_rate": 3.718925177500743e-07,
+      "loss": 1.5008,
+      "step": 6861
+    },
+    {
+      "epoch": 0.9156658660261543,
+      "grad_norm": 0.9532844740769746,
+      "learning_rate": 3.707257388890195e-07,
+      "loss": 1.5318,
+      "step": 6862
+    },
+    {
+      "epoch": 0.9157993061115559,
+      "grad_norm": 0.9749521833223408,
+      "learning_rate": 3.6956075866676155e-07,
+      "loss": 1.528,
+      "step": 6863
+    },
+    {
+      "epoch": 0.9159327461969575,
+      "grad_norm": 1.1423465704015638,
+      "learning_rate": 3.6839757730090744e-07,
+      "loss": 1.5111,
+      "step": 6864
+    },
+    {
+      "epoch": 0.9160661862823593,
+      "grad_norm": 0.9134442615213153,
+      "learning_rate": 3.672361950087266e-07,
+      "loss": 1.5284,
+      "step": 6865
+    },
+    {
+      "epoch": 0.9161996263677609,
+      "grad_norm": 0.9286686714079571,
+      "learning_rate": 3.660766120071513e-07,
+      "loss": 1.4942,
+      "step": 6866
+    },
+    {
+      "epoch": 0.9163330664531625,
+      "grad_norm": 0.9431823883198829,
+      "learning_rate": 3.649188285127769e-07,
+      "loss": 1.5244,
+      "step": 6867
+    },
+    {
+      "epoch": 0.9164665065385642,
+      "grad_norm": 1.0666489900172478,
+      "learning_rate": 3.637628447418673e-07,
+      "loss": 1.5604,
+      "step": 6868
+    },
+    {
+      "epoch": 0.9165999466239658,
+      "grad_norm": 0.9836730783947925,
+      "learning_rate": 3.626086609103463e-07,
+      "loss": 1.5952,
+      "step": 6869
+    },
+    {
+      "epoch": 0.9167333867093675,
+      "grad_norm": 0.9593749865507312,
+      "learning_rate": 3.614562772338015e-07,
+      "loss": 1.4742,
+      "step": 6870
+    },
+    {
+      "epoch": 0.9168668267947692,
+      "grad_norm": 1.0579631089786683,
+      "learning_rate": 3.6030569392748294e-07,
+      "loss": 1.5043,
+      "step": 6871
+    },
+    {
+      "epoch": 0.9170002668801708,
+      "grad_norm": 0.979103974391311,
+      "learning_rate": 3.5915691120631093e-07,
+      "loss": 1.5323,
+      "step": 6872
+    },
+    {
+      "epoch": 0.9171337069655725,
+      "grad_norm": 1.503246770281354,
+      "learning_rate": 3.5800992928486265e-07,
+      "loss": 1.5514,
+      "step": 6873
+    },
+    {
+      "epoch": 0.9172671470509741,
+      "grad_norm": 0.9405027952328543,
+      "learning_rate": 3.568647483773813e-07,
+      "loss": 1.5276,
+      "step": 6874
+    },
+    {
+      "epoch": 0.9174005871363757,
+      "grad_norm": 1.0134385203890537,
+      "learning_rate": 3.557213686977723e-07,
+      "loss": 1.5814,
+      "step": 6875
+    },
+    {
+      "epoch": 0.9175340272217775,
+      "grad_norm": 0.9200398848804521,
+      "learning_rate": 3.545797904596082e-07,
+      "loss": 1.5544,
+      "step": 6876
+    },
+    {
+      "epoch": 0.9176674673071791,
+      "grad_norm": 0.9685297046655957,
+      "learning_rate": 3.5344001387612293e-07,
+      "loss": 1.5557,
+      "step": 6877
+    },
+    {
+      "epoch": 0.9178009073925807,
+      "grad_norm": 0.9225117368528908,
+      "learning_rate": 3.5230203916021277e-07,
+      "loss": 1.5779,
+      "step": 6878
+    },
+    {
+      "epoch": 0.9179343474779824,
+      "grad_norm": 0.9089807335651439,
+      "learning_rate": 3.511658665244377e-07,
+      "loss": 1.4707,
+      "step": 6879
+    },
+    {
+      "epoch": 0.918067787563384,
+      "grad_norm": 1.0909826397284352,
+      "learning_rate": 3.5003149618102253e-07,
+      "loss": 1.5554,
+      "step": 6880
+    },
+    {
+      "epoch": 0.9182012276487856,
+      "grad_norm": 0.9370669644460864,
+      "learning_rate": 3.4889892834185533e-07,
+      "loss": 1.4784,
+      "step": 6881
+    },
+    {
+      "epoch": 0.9183346677341874,
+      "grad_norm": 0.9322223361726935,
+      "learning_rate": 3.4776816321848596e-07,
+      "loss": 1.5147,
+      "step": 6882
+    },
+    {
+      "epoch": 0.918468107819589,
+      "grad_norm": 0.9329739735097461,
+      "learning_rate": 3.466392010221298e-07,
+      "loss": 1.5383,
+      "step": 6883
+    },
+    {
+      "epoch": 0.9186015479049907,
+      "grad_norm": 0.9151417878518067,
+      "learning_rate": 3.4551204196366263e-07,
+      "loss": 1.5576,
+      "step": 6884
+    },
+    {
+      "epoch": 0.9187349879903923,
+      "grad_norm": 0.951435126300393,
+      "learning_rate": 3.4438668625362493e-07,
+      "loss": 1.5314,
+      "step": 6885
+    },
+    {
+      "epoch": 0.9188684280757939,
+      "grad_norm": 0.913279499004807,
+      "learning_rate": 3.432631341022219e-07,
+      "loss": 1.4513,
+      "step": 6886
+    },
+    {
+      "epoch": 0.9190018681611957,
+      "grad_norm": 0.9441844786118736,
+      "learning_rate": 3.42141385719319e-07,
+      "loss": 1.5387,
+      "step": 6887
+    },
+    {
+      "epoch": 0.9191353082465973,
+      "grad_norm": 0.932767307846722,
+      "learning_rate": 3.410214413144464e-07,
+      "loss": 1.5246,
+      "step": 6888
+    },
+    {
+      "epoch": 0.9192687483319989,
+      "grad_norm": 0.9472847800493374,
+      "learning_rate": 3.399033010967967e-07,
+      "loss": 1.4983,
+      "step": 6889
+    },
+    {
+      "epoch": 0.9194021884174006,
+      "grad_norm": 1.0419370287145757,
+      "learning_rate": 3.3878696527522624e-07,
+      "loss": 1.5103,
+      "step": 6890
+    },
+    {
+      "epoch": 0.9195356285028022,
+      "grad_norm": 0.9681983142545559,
+      "learning_rate": 3.3767243405825487e-07,
+      "loss": 1.5462,
+      "step": 6891
+    },
+    {
+      "epoch": 0.9196690685882039,
+      "grad_norm": 0.9052614017344128,
+      "learning_rate": 3.3655970765406056e-07,
+      "loss": 1.4736,
+      "step": 6892
+    },
+    {
+      "epoch": 0.9198025086736056,
+      "grad_norm": 1.0228191814158987,
+      "learning_rate": 3.354487862704925e-07,
+      "loss": 1.5639,
+      "step": 6893
+    },
+    {
+      "epoch": 0.9199359487590072,
+      "grad_norm": 0.9360422534666102,
+      "learning_rate": 3.3433967011505586e-07,
+      "loss": 1.5466,
+      "step": 6894
+    },
+    {
+      "epoch": 0.9200693888444088,
+      "grad_norm": 0.9264242257227067,
+      "learning_rate": 3.332323593949205e-07,
+      "loss": 1.5214,
+      "step": 6895
+    },
+    {
+      "epoch": 0.9202028289298105,
+      "grad_norm": 0.9785840279808523,
+      "learning_rate": 3.3212685431691983e-07,
+      "loss": 1.521,
+      "step": 6896
+    },
+    {
+      "epoch": 0.9203362690152121,
+      "grad_norm": 0.990110652510918,
+      "learning_rate": 3.310231550875509e-07,
+      "loss": 1.539,
+      "step": 6897
+    },
+    {
+      "epoch": 0.9204697091006139,
+      "grad_norm": 0.9933025587703744,
+      "learning_rate": 3.29921261912971e-07,
+      "loss": 1.5062,
+      "step": 6898
+    },
+    {
+      "epoch": 0.9206031491860155,
+      "grad_norm": 0.924550064236254,
+      "learning_rate": 3.28821174999e-07,
+      "loss": 1.5394,
+      "step": 6899
+    },
+    {
+      "epoch": 0.9207365892714171,
+      "grad_norm": 0.9765816922537065,
+      "learning_rate": 3.277228945511246e-07,
+      "loss": 1.4902,
+      "step": 6900
+    },
+    {
+      "epoch": 0.9208700293568188,
+      "grad_norm": 0.9523736840444362,
+      "learning_rate": 3.266264207744885e-07,
+      "loss": 1.5404,
+      "step": 6901
+    },
+    {
+      "epoch": 0.9210034694422204,
+      "grad_norm": 0.9485498472126853,
+      "learning_rate": 3.2553175387390225e-07,
+      "loss": 1.5547,
+      "step": 6902
+    },
+    {
+      "epoch": 0.9211369095276221,
+      "grad_norm": 0.9345918935854542,
+      "learning_rate": 3.2443889405383564e-07,
+      "loss": 1.4989,
+      "step": 6903
+    },
+    {
+      "epoch": 0.9212703496130238,
+      "grad_norm": 0.9620003150256877,
+      "learning_rate": 3.2334784151842434e-07,
+      "loss": 1.5763,
+      "step": 6904
+    },
+    {
+      "epoch": 0.9214037896984254,
+      "grad_norm": 0.9533414019445823,
+      "learning_rate": 3.2225859647146306e-07,
+      "loss": 1.5578,
+      "step": 6905
+    },
+    {
+      "epoch": 0.921537229783827,
+      "grad_norm": 1.030837787394548,
+      "learning_rate": 3.2117115911640904e-07,
+      "loss": 1.5385,
+      "step": 6906
+    },
+    {
+      "epoch": 0.9216706698692287,
+      "grad_norm": 0.9838369808924554,
+      "learning_rate": 3.200855296563865e-07,
+      "loss": 1.5466,
+      "step": 6907
+    },
+    {
+      "epoch": 0.9218041099546304,
+      "grad_norm": 0.9369615843893399,
+      "learning_rate": 3.1900170829417765e-07,
+      "loss": 1.5616,
+      "step": 6908
+    },
+    {
+      "epoch": 0.921937550040032,
+      "grad_norm": 1.0512907143003218,
+      "learning_rate": 3.179196952322272e-07,
+      "loss": 1.5901,
+      "step": 6909
+    },
+    {
+      "epoch": 0.9220709901254337,
+      "grad_norm": 0.9920000789006053,
+      "learning_rate": 3.168394906726413e-07,
+      "loss": 1.5857,
+      "step": 6910
+    },
+    {
+      "epoch": 0.9222044302108353,
+      "grad_norm": 0.9257587202232741,
+      "learning_rate": 3.1576109481719294e-07,
+      "loss": 1.5313,
+      "step": 6911
+    },
+    {
+      "epoch": 0.922337870296237,
+      "grad_norm": 0.9418319592189258,
+      "learning_rate": 3.146845078673133e-07,
+      "loss": 1.539,
+      "step": 6912
+    },
+    {
+      "epoch": 0.9224713103816387,
+      "grad_norm": 0.9070087342084403,
+      "learning_rate": 3.136097300240948e-07,
+      "loss": 1.4764,
+      "step": 6913
+    },
+    {
+      "epoch": 0.9226047504670403,
+      "grad_norm": 0.965385596189811,
+      "learning_rate": 3.125367614882957e-07,
+      "loss": 1.5912,
+      "step": 6914
+    },
+    {
+      "epoch": 0.922738190552442,
+      "grad_norm": 0.9457846928440798,
+      "learning_rate": 3.114656024603346e-07,
+      "loss": 1.6022,
+      "step": 6915
+    },
+    {
+      "epoch": 0.9228716306378436,
+      "grad_norm": 0.9626967040370848,
+      "learning_rate": 3.1039625314028934e-07,
+      "loss": 1.5788,
+      "step": 6916
+    },
+    {
+      "epoch": 0.9230050707232452,
+      "grad_norm": 0.9262926987232699,
+      "learning_rate": 3.093287137279044e-07,
+      "loss": 1.5076,
+      "step": 6917
+    },
+    {
+      "epoch": 0.923138510808647,
+      "grad_norm": 0.9263671794999331,
+      "learning_rate": 3.0826298442258263e-07,
+      "loss": 1.5376,
+      "step": 6918
+    },
+    {
+      "epoch": 0.9232719508940486,
+      "grad_norm": 1.0802339378714219,
+      "learning_rate": 3.071990654233925e-07,
+      "loss": 1.5194,
+      "step": 6919
+    },
+    {
+      "epoch": 0.9234053909794502,
+      "grad_norm": 0.9517175443121075,
+      "learning_rate": 3.0613695692905955e-07,
+      "loss": 1.5472,
+      "step": 6920
+    },
+    {
+      "epoch": 0.9235388310648519,
+      "grad_norm": 0.9997691999553755,
+      "learning_rate": 3.050766591379739e-07,
+      "loss": 1.5521,
+      "step": 6921
+    },
+    {
+      "epoch": 0.9236722711502535,
+      "grad_norm": 0.927201693441311,
+      "learning_rate": 3.040181722481872e-07,
+      "loss": 1.5507,
+      "step": 6922
+    },
+    {
+      "epoch": 0.9238057112356551,
+      "grad_norm": 0.9436121306544267,
+      "learning_rate": 3.0296149645741344e-07,
+      "loss": 1.5319,
+      "step": 6923
+    },
+    {
+      "epoch": 0.9239391513210569,
+      "grad_norm": 0.9358298860342295,
+      "learning_rate": 3.0190663196302706e-07,
+      "loss": 1.512,
+      "step": 6924
+    },
+    {
+      "epoch": 0.9240725914064585,
+      "grad_norm": 0.9671848262898582,
+      "learning_rate": 3.008535789620648e-07,
+      "loss": 1.5107,
+      "step": 6925
+    },
+    {
+      "epoch": 0.9242060314918602,
+      "grad_norm": 0.9462619779524947,
+      "learning_rate": 2.99802337651226e-07,
+      "loss": 1.5133,
+      "step": 6926
+    },
+    {
+      "epoch": 0.9243394715772618,
+      "grad_norm": 0.9375880200694752,
+      "learning_rate": 2.987529082268692e-07,
+      "loss": 1.5224,
+      "step": 6927
+    },
+    {
+      "epoch": 0.9244729116626634,
+      "grad_norm": 0.9185590934633807,
+      "learning_rate": 2.977052908850142e-07,
+      "loss": 1.5366,
+      "step": 6928
+    },
+    {
+      "epoch": 0.9246063517480652,
+      "grad_norm": 0.925660519615923,
+      "learning_rate": 2.9665948582134783e-07,
+      "loss": 1.5373,
+      "step": 6929
+    },
+    {
+      "epoch": 0.9247397918334668,
+      "grad_norm": 0.9558442289500324,
+      "learning_rate": 2.9561549323121385e-07,
+      "loss": 1.5556,
+      "step": 6930
+    },
+    {
+      "epoch": 0.9248732319188684,
+      "grad_norm": 0.9105723816514041,
+      "learning_rate": 2.9457331330961513e-07,
+      "loss": 1.5277,
+      "step": 6931
+    },
+    {
+      "epoch": 0.9250066720042701,
+      "grad_norm": 0.9402020298370549,
+      "learning_rate": 2.935329462512226e-07,
+      "loss": 1.5444,
+      "step": 6932
+    },
+    {
+      "epoch": 0.9251401120896717,
+      "grad_norm": 0.9389120664607241,
+      "learning_rate": 2.9249439225036313e-07,
+      "loss": 1.5906,
+      "step": 6933
+    },
+    {
+      "epoch": 0.9252735521750733,
+      "grad_norm": 0.9566777094490774,
+      "learning_rate": 2.91457651501027e-07,
+      "loss": 1.5388,
+      "step": 6934
+    },
+    {
+      "epoch": 0.9254069922604751,
+      "grad_norm": 0.9475669400048962,
+      "learning_rate": 2.904227241968638e-07,
+      "loss": 1.5122,
+      "step": 6935
+    },
+    {
+      "epoch": 0.9255404323458767,
+      "grad_norm": 0.9606016426402592,
+      "learning_rate": 2.8938961053118997e-07,
+      "loss": 1.5747,
+      "step": 6936
+    },
+    {
+      "epoch": 0.9256738724312784,
+      "grad_norm": 0.9416160631017734,
+      "learning_rate": 2.8835831069697786e-07,
+      "loss": 1.5942,
+      "step": 6937
+    },
+    {
+      "epoch": 0.92580731251668,
+      "grad_norm": 0.9255067631796212,
+      "learning_rate": 2.873288248868611e-07,
+      "loss": 1.5681,
+      "step": 6938
+    },
+    {
+      "epoch": 0.9259407526020816,
+      "grad_norm": 0.9354963560086351,
+      "learning_rate": 2.86301153293137e-07,
+      "loss": 1.5262,
+      "step": 6939
+    },
+    {
+      "epoch": 0.9260741926874834,
+      "grad_norm": 0.9413294820231983,
+      "learning_rate": 2.852752961077632e-07,
+      "loss": 1.535,
+      "step": 6940
+    },
+    {
+      "epoch": 0.926207632772885,
+      "grad_norm": 0.9362918263402716,
+      "learning_rate": 2.842512535223585e-07,
+      "loss": 1.5211,
+      "step": 6941
+    },
+    {
+      "epoch": 0.9263410728582866,
+      "grad_norm": 0.921974590840743,
+      "learning_rate": 2.832290257282e-07,
+      "loss": 1.5103,
+      "step": 6942
+    },
+    {
+      "epoch": 0.9264745129436883,
+      "grad_norm": 0.9369764170400131,
+      "learning_rate": 2.822086129162305e-07,
+      "loss": 1.5596,
+      "step": 6943
+    },
+    {
+      "epoch": 0.9266079530290899,
+      "grad_norm": 0.9147557346642654,
+      "learning_rate": 2.811900152770519e-07,
+      "loss": 1.5058,
+      "step": 6944
+    },
+    {
+      "epoch": 0.9267413931144916,
+      "grad_norm": 0.9679302190588822,
+      "learning_rate": 2.8017323300092435e-07,
+      "loss": 1.576,
+      "step": 6945
+    },
+    {
+      "epoch": 0.9268748331998933,
+      "grad_norm": 0.9466621416815499,
+      "learning_rate": 2.791582662777725e-07,
+      "loss": 1.5155,
+      "step": 6946
+    },
+    {
+      "epoch": 0.9270082732852949,
+      "grad_norm": 1.0259125716265545,
+      "learning_rate": 2.7814511529718146e-07,
+      "loss": 1.588,
+      "step": 6947
+    },
+    {
+      "epoch": 0.9271417133706965,
+      "grad_norm": 0.9713536021169017,
+      "learning_rate": 2.771337802483953e-07,
+      "loss": 1.505,
+      "step": 6948
+    },
+    {
+      "epoch": 0.9272751534560982,
+      "grad_norm": 0.9124231485781754,
+      "learning_rate": 2.7612426132031966e-07,
+      "loss": 1.5179,
+      "step": 6949
+    },
+    {
+      "epoch": 0.9274085935414998,
+      "grad_norm": 0.9450377082479692,
+      "learning_rate": 2.751165587015214e-07,
+      "loss": 1.516,
+      "step": 6950
+    },
+    {
+      "epoch": 0.9275420336269016,
+      "grad_norm": 0.9430651969927591,
+      "learning_rate": 2.7411067258022896e-07,
+      "loss": 1.5616,
+      "step": 6951
+    },
+    {
+      "epoch": 0.9276754737123032,
+      "grad_norm": 1.1461572084836693,
+      "learning_rate": 2.731066031443275e-07,
+      "loss": 1.5797,
+      "step": 6952
+    },
+    {
+      "epoch": 0.9278089137977048,
+      "grad_norm": 0.9850748413044481,
+      "learning_rate": 2.721043505813692e-07,
+      "loss": 1.6142,
+      "step": 6953
+    },
+    {
+      "epoch": 0.9279423538831065,
+      "grad_norm": 0.9410326470678416,
+      "learning_rate": 2.7110391507856215e-07,
+      "loss": 1.5513,
+      "step": 6954
+    },
+    {
+      "epoch": 0.9280757939685081,
+      "grad_norm": 0.9717540522614107,
+      "learning_rate": 2.7010529682277573e-07,
+      "loss": 1.5271,
+      "step": 6955
+    },
+    {
+      "epoch": 0.9282092340539098,
+      "grad_norm": 1.2527445070154288,
+      "learning_rate": 2.691084960005408e-07,
+      "loss": 1.554,
+      "step": 6956
+    },
+    {
+      "epoch": 0.9283426741393115,
+      "grad_norm": 0.9380409123192367,
+      "learning_rate": 2.681135127980483e-07,
+      "loss": 1.4966,
+      "step": 6957
+    },
+    {
+      "epoch": 0.9284761142247131,
+      "grad_norm": 1.067794635668956,
+      "learning_rate": 2.671203474011508e-07,
+      "loss": 1.4979,
+      "step": 6958
+    },
+    {
+      "epoch": 0.9286095543101147,
+      "grad_norm": 0.924552942569084,
+      "learning_rate": 2.6612899999535867e-07,
+      "loss": 1.6066,
+      "step": 6959
+    },
+    {
+      "epoch": 0.9287429943955164,
+      "grad_norm": 0.9844576271371721,
+      "learning_rate": 2.65139470765845e-07,
+      "loss": 1.558,
+      "step": 6960
+    },
+    {
+      "epoch": 0.9288764344809181,
+      "grad_norm": 0.9671888751234345,
+      "learning_rate": 2.64151759897443e-07,
+      "loss": 1.5542,
+      "step": 6961
+    },
+    {
+      "epoch": 0.9290098745663197,
+      "grad_norm": 0.9178373406964115,
+      "learning_rate": 2.6316586757464513e-07,
+      "loss": 1.5549,
+      "step": 6962
+    },
+    {
+      "epoch": 0.9291433146517214,
+      "grad_norm": 0.9799980886987,
+      "learning_rate": 2.621817939816051e-07,
+      "loss": 1.5763,
+      "step": 6963
+    },
+    {
+      "epoch": 0.929276754737123,
+      "grad_norm": 0.9239659156537983,
+      "learning_rate": 2.6119953930213713e-07,
+      "loss": 1.5654,
+      "step": 6964
+    },
+    {
+      "epoch": 0.9294101948225247,
+      "grad_norm": 1.0117689513131394,
+      "learning_rate": 2.602191037197155e-07,
+      "loss": 1.6097,
+      "step": 6965
+    },
+    {
+      "epoch": 0.9295436349079264,
+      "grad_norm": 0.9259063770945314,
+      "learning_rate": 2.592404874174714e-07,
+      "loss": 1.5442,
+      "step": 6966
+    },
+    {
+      "epoch": 0.929677074993328,
+      "grad_norm": 0.94289731705226,
+      "learning_rate": 2.582636905782032e-07,
+      "loss": 1.5511,
+      "step": 6967
+    },
+    {
+      "epoch": 0.9298105150787297,
+      "grad_norm": 1.2363951568921343,
+      "learning_rate": 2.572887133843638e-07,
+      "loss": 1.5284,
+      "step": 6968
+    },
+    {
+      "epoch": 0.9299439551641313,
+      "grad_norm": 0.9237895244839454,
+      "learning_rate": 2.5631555601806746e-07,
+      "loss": 1.5666,
+      "step": 6969
+    },
+    {
+      "epoch": 0.9300773952495329,
+      "grad_norm": 1.2534199107639514,
+      "learning_rate": 2.5534421866108884e-07,
+      "loss": 1.5386,
+      "step": 6970
+    },
+    {
+      "epoch": 0.9302108353349346,
+      "grad_norm": 0.9440071222676694,
+      "learning_rate": 2.543747014948639e-07,
+      "loss": 1.5252,
+      "step": 6971
+    },
+    {
+      "epoch": 0.9303442754203363,
+      "grad_norm": 0.9410423473034675,
+      "learning_rate": 2.5340700470048664e-07,
+      "loss": 1.5539,
+      "step": 6972
+    },
+    {
+      "epoch": 0.9304777155057379,
+      "grad_norm": 0.9370283821223521,
+      "learning_rate": 2.524411284587114e-07,
+      "loss": 1.5497,
+      "step": 6973
+    },
+    {
+      "epoch": 0.9306111555911396,
+      "grad_norm": 0.9353742776563619,
+      "learning_rate": 2.5147707294995274e-07,
+      "loss": 1.4892,
+      "step": 6974
+    },
+    {
+      "epoch": 0.9307445956765412,
+      "grad_norm": 0.9297881657463101,
+      "learning_rate": 2.505148383542866e-07,
+      "loss": 1.5467,
+      "step": 6975
+    },
+    {
+      "epoch": 0.9308780357619428,
+      "grad_norm": 0.952275461349787,
+      "learning_rate": 2.495544248514459e-07,
+      "loss": 1.4877,
+      "step": 6976
+    },
+    {
+      "epoch": 0.9310114758473446,
+      "grad_norm": 0.9152447233126888,
+      "learning_rate": 2.48595832620826e-07,
+      "loss": 1.5295,
+      "step": 6977
+    },
+    {
+      "epoch": 0.9311449159327462,
+      "grad_norm": 0.91217801833338,
+      "learning_rate": 2.476390618414803e-07,
+      "loss": 1.5292,
+      "step": 6978
+    },
+    {
+      "epoch": 0.9312783560181479,
+      "grad_norm": 0.927701396795192,
+      "learning_rate": 2.4668411269212377e-07,
+      "loss": 1.5662,
+      "step": 6979
+    },
+    {
+      "epoch": 0.9314117961035495,
+      "grad_norm": 0.9404375961303916,
+      "learning_rate": 2.4573098535112913e-07,
+      "loss": 1.5882,
+      "step": 6980
+    },
+    {
+      "epoch": 0.9315452361889511,
+      "grad_norm": 0.9405292351716275,
+      "learning_rate": 2.4477967999652854e-07,
+      "loss": 1.512,
+      "step": 6981
+    },
+    {
+      "epoch": 0.9316786762743529,
+      "grad_norm": 0.9390791133758444,
+      "learning_rate": 2.438301968060186e-07,
+      "loss": 1.5456,
+      "step": 6982
+    },
+    {
+      "epoch": 0.9318121163597545,
+      "grad_norm": 0.9294723456622169,
+      "learning_rate": 2.4288253595694865e-07,
+      "loss": 1.5763,
+      "step": 6983
+    },
+    {
+      "epoch": 0.9319455564451561,
+      "grad_norm": 0.9086002326565813,
+      "learning_rate": 2.419366976263315e-07,
+      "loss": 1.5412,
+      "step": 6984
+    },
+    {
+      "epoch": 0.9320789965305578,
+      "grad_norm": 1.0686971025848702,
+      "learning_rate": 2.409926819908404e-07,
+      "loss": 1.5746,
+      "step": 6985
+    },
+    {
+      "epoch": 0.9322124366159594,
+      "grad_norm": 1.0891885771737662,
+      "learning_rate": 2.400504892268052e-07,
+      "loss": 1.5662,
+      "step": 6986
+    },
+    {
+      "epoch": 0.932345876701361,
+      "grad_norm": 0.9460922151250107,
+      "learning_rate": 2.391101195102175e-07,
+      "loss": 1.557,
+      "step": 6987
+    },
+    {
+      "epoch": 0.9324793167867628,
+      "grad_norm": 0.985651528432505,
+      "learning_rate": 2.3817157301672777e-07,
+      "loss": 1.578,
+      "step": 6988
+    },
+    {
+      "epoch": 0.9326127568721644,
+      "grad_norm": 0.9664112206315217,
+      "learning_rate": 2.372348499216459e-07,
+      "loss": 1.539,
+      "step": 6989
+    },
+    {
+      "epoch": 0.932746196957566,
+      "grad_norm": 0.9490309404868328,
+      "learning_rate": 2.3629995039994082e-07,
+      "loss": 1.5389,
+      "step": 6990
+    },
+    {
+      "epoch": 0.9328796370429677,
+      "grad_norm": 0.9577672566554155,
+      "learning_rate": 2.3536687462624053e-07,
+      "loss": 1.5601,
+      "step": 6991
+    },
+    {
+      "epoch": 0.9330130771283693,
+      "grad_norm": 0.9223528519716557,
+      "learning_rate": 2.3443562277483345e-07,
+      "loss": 1.5533,
+      "step": 6992
+    },
+    {
+      "epoch": 0.9331465172137711,
+      "grad_norm": 0.9427663030847381,
+      "learning_rate": 2.335061950196671e-07,
+      "loss": 1.5562,
+      "step": 6993
+    },
+    {
+      "epoch": 0.9332799572991727,
+      "grad_norm": 0.9523796363575632,
+      "learning_rate": 2.3257859153434815e-07,
+      "loss": 1.556,
+      "step": 6994
+    },
+    {
+      "epoch": 0.9334133973845743,
+      "grad_norm": 0.9514855637959606,
+      "learning_rate": 2.3165281249214133e-07,
+      "loss": 1.5486,
+      "step": 6995
+    },
+    {
+      "epoch": 0.933546837469976,
+      "grad_norm": 0.8960652793709516,
+      "learning_rate": 2.307288580659728e-07,
+      "loss": 1.4996,
+      "step": 6996
+    },
+    {
+      "epoch": 0.9336802775553776,
+      "grad_norm": 0.9468839774939591,
+      "learning_rate": 2.2980672842842665e-07,
+      "loss": 1.5557,
+      "step": 6997
+    },
+    {
+      "epoch": 0.9338137176407793,
+      "grad_norm": 0.9323423502741678,
+      "learning_rate": 2.288864237517463e-07,
+      "loss": 1.5423,
+      "step": 6998
+    },
+    {
+      "epoch": 0.933947157726181,
+      "grad_norm": 0.9477440100001612,
+      "learning_rate": 2.2796794420783198e-07,
+      "loss": 1.5839,
+      "step": 6999
+    },
+    {
+      "epoch": 0.9340805978115826,
+      "grad_norm": 0.9719508936768406,
+      "learning_rate": 2.2705128996824755e-07,
+      "loss": 1.5303,
+      "step": 7000
+    },
+    {
+      "epoch": 0.9342140378969842,
+      "grad_norm": 0.9511975135556928,
+      "learning_rate": 2.2613646120421383e-07,
+      "loss": 1.5622,
+      "step": 7001
+    },
+    {
+      "epoch": 0.9343474779823859,
+      "grad_norm": 0.9164128395943615,
+      "learning_rate": 2.2522345808660861e-07,
+      "loss": 1.5558,
+      "step": 7002
+    },
+    {
+      "epoch": 0.9344809180677875,
+      "grad_norm": 0.9532778700819494,
+      "learning_rate": 2.24312280785971e-07,
+      "loss": 1.5872,
+      "step": 7003
+    },
+    {
+      "epoch": 0.9346143581531893,
+      "grad_norm": 0.9256358437386732,
+      "learning_rate": 2.2340292947250043e-07,
+      "loss": 1.5318,
+      "step": 7004
+    },
+    {
+      "epoch": 0.9347477982385909,
+      "grad_norm": 1.0723225175898141,
+      "learning_rate": 2.2249540431605099e-07,
+      "loss": 1.5304,
+      "step": 7005
+    },
+    {
+      "epoch": 0.9348812383239925,
+      "grad_norm": 0.941909432602312,
+      "learning_rate": 2.2158970548613934e-07,
+      "loss": 1.4993,
+      "step": 7006
+    },
+    {
+      "epoch": 0.9350146784093942,
+      "grad_norm": 0.9580854716526129,
+      "learning_rate": 2.2068583315193902e-07,
+      "loss": 1.5505,
+      "step": 7007
+    },
+    {
+      "epoch": 0.9351481184947958,
+      "grad_norm": 0.9149155992270841,
+      "learning_rate": 2.197837874822839e-07,
+      "loss": 1.5243,
+      "step": 7008
+    },
+    {
+      "epoch": 0.9352815585801975,
+      "grad_norm": 0.931850561675305,
+      "learning_rate": 2.1888356864566363e-07,
+      "loss": 1.5763,
+      "step": 7009
+    },
+    {
+      "epoch": 0.9354149986655992,
+      "grad_norm": 0.945264466148816,
+      "learning_rate": 2.1798517681023257e-07,
+      "loss": 1.5681,
+      "step": 7010
+    },
+    {
+      "epoch": 0.9355484387510008,
+      "grad_norm": 1.1408246656369734,
+      "learning_rate": 2.1708861214379762e-07,
+      "loss": 1.5508,
+      "step": 7011
+    },
+    {
+      "epoch": 0.9356818788364024,
+      "grad_norm": 0.9305152136394855,
+      "learning_rate": 2.1619387481382704e-07,
+      "loss": 1.4831,
+      "step": 7012
+    },
+    {
+      "epoch": 0.9358153189218041,
+      "grad_norm": 0.9374514423774111,
+      "learning_rate": 2.153009649874471e-07,
+      "loss": 1.534,
+      "step": 7013
+    },
+    {
+      "epoch": 0.9359487590072058,
+      "grad_norm": 0.9373740971918221,
+      "learning_rate": 2.144098828314445e-07,
+      "loss": 1.4639,
+      "step": 7014
+    },
+    {
+      "epoch": 0.9360821990926074,
+      "grad_norm": 0.9202329086139532,
+      "learning_rate": 2.1352062851226263e-07,
+      "loss": 1.5189,
+      "step": 7015
+    },
+    {
+      "epoch": 0.9362156391780091,
+      "grad_norm": 0.9251964887586024,
+      "learning_rate": 2.1263320219600426e-07,
+      "loss": 1.5786,
+      "step": 7016
+    },
+    {
+      "epoch": 0.9363490792634107,
+      "grad_norm": 0.9284256574113852,
+      "learning_rate": 2.1174760404843008e-07,
+      "loss": 1.526,
+      "step": 7017
+    },
+    {
+      "epoch": 0.9364825193488124,
+      "grad_norm": 1.882242068390261,
+      "learning_rate": 2.1086383423496004e-07,
+      "loss": 1.5768,
+      "step": 7018
+    },
+    {
+      "epoch": 0.936615959434214,
+      "grad_norm": 0.9901007118211904,
+      "learning_rate": 2.0998189292067316e-07,
+      "loss": 1.5077,
+      "step": 7019
+    },
+    {
+      "epoch": 0.9367493995196157,
+      "grad_norm": 1.0159260253271438,
+      "learning_rate": 2.0910178027030326e-07,
+      "loss": 1.5657,
+      "step": 7020
+    },
+    {
+      "epoch": 0.9368828396050174,
+      "grad_norm": 0.9401229086492431,
+      "learning_rate": 2.082234964482488e-07,
+      "loss": 1.5582,
+      "step": 7021
+    },
+    {
+      "epoch": 0.937016279690419,
+      "grad_norm": 1.0751751887062786,
+      "learning_rate": 2.073470416185619e-07,
+      "loss": 1.5412,
+      "step": 7022
+    },
+    {
+      "epoch": 0.9371497197758206,
+      "grad_norm": 0.9506968457876054,
+      "learning_rate": 2.0647241594495381e-07,
+      "loss": 1.525,
+      "step": 7023
+    },
+    {
+      "epoch": 0.9372831598612223,
+      "grad_norm": 0.9301700959194582,
+      "learning_rate": 2.0559961959079278e-07,
+      "loss": 1.531,
+      "step": 7024
+    },
+    {
+      "epoch": 0.937416599946624,
+      "grad_norm": 0.9323275478960292,
+      "learning_rate": 2.047286527191117e-07,
+      "loss": 1.5362,
+      "step": 7025
+    },
+    {
+      "epoch": 0.9375500400320256,
+      "grad_norm": 0.9191649723560835,
+      "learning_rate": 2.0385951549259486e-07,
+      "loss": 1.5815,
+      "step": 7026
+    },
+    {
+      "epoch": 0.9376834801174273,
+      "grad_norm": 1.032563096761044,
+      "learning_rate": 2.0299220807358578e-07,
+      "loss": 1.5741,
+      "step": 7027
+    },
+    {
+      "epoch": 0.9378169202028289,
+      "grad_norm": 0.987366129369346,
+      "learning_rate": 2.0212673062409038e-07,
+      "loss": 1.5807,
+      "step": 7028
+    },
+    {
+      "epoch": 0.9379503602882305,
+      "grad_norm": 0.9364694767430964,
+      "learning_rate": 2.0126308330576937e-07,
+      "loss": 1.5519,
+      "step": 7029
+    },
+    {
+      "epoch": 0.9380838003736323,
+      "grad_norm": 1.4286340232745998,
+      "learning_rate": 2.004012662799404e-07,
+      "loss": 1.5545,
+      "step": 7030
+    },
+    {
+      "epoch": 0.9382172404590339,
+      "grad_norm": 0.935633863001671,
+      "learning_rate": 1.9954127970758131e-07,
+      "loss": 1.5272,
+      "step": 7031
+    },
+    {
+      "epoch": 0.9383506805444356,
+      "grad_norm": 0.9511612705616507,
+      "learning_rate": 1.986831237493303e-07,
+      "loss": 1.5378,
+      "step": 7032
+    },
+    {
+      "epoch": 0.9384841206298372,
+      "grad_norm": 0.9426846301005706,
+      "learning_rate": 1.9782679856548025e-07,
+      "loss": 1.5306,
+      "step": 7033
+    },
+    {
+      "epoch": 0.9386175607152388,
+      "grad_norm": 1.055465469484963,
+      "learning_rate": 1.969723043159799e-07,
+      "loss": 1.5642,
+      "step": 7034
+    },
+    {
+      "epoch": 0.9387510008006406,
+      "grad_norm": 0.9183296763521461,
+      "learning_rate": 1.961196411604438e-07,
+      "loss": 1.5428,
+      "step": 7035
+    },
+    {
+      "epoch": 0.9388844408860422,
+      "grad_norm": 0.9134482621394847,
+      "learning_rate": 1.9526880925813673e-07,
+      "loss": 1.501,
+      "step": 7036
+    },
+    {
+      "epoch": 0.9390178809714438,
+      "grad_norm": 0.9419067060611698,
+      "learning_rate": 1.9441980876798493e-07,
+      "loss": 1.5601,
+      "step": 7037
+    },
+    {
+      "epoch": 0.9391513210568455,
+      "grad_norm": 0.9369479568180347,
+      "learning_rate": 1.9357263984857044e-07,
+      "loss": 1.5471,
+      "step": 7038
+    },
+    {
+      "epoch": 0.9392847611422471,
+      "grad_norm": 0.9557500312485121,
+      "learning_rate": 1.9272730265813887e-07,
+      "loss": 1.6029,
+      "step": 7039
+    },
+    {
+      "epoch": 0.9394182012276487,
+      "grad_norm": 1.0554599506743223,
+      "learning_rate": 1.9188379735458618e-07,
+      "loss": 1.5511,
+      "step": 7040
+    },
+    {
+      "epoch": 0.9395516413130505,
+      "grad_norm": 1.0458439629457223,
+      "learning_rate": 1.9104212409546964e-07,
+      "loss": 1.5568,
+      "step": 7041
+    },
+    {
+      "epoch": 0.9396850813984521,
+      "grad_norm": 1.0948545675024355,
+      "learning_rate": 1.902022830380068e-07,
+      "loss": 1.537,
+      "step": 7042
+    },
+    {
+      "epoch": 0.9398185214838537,
+      "grad_norm": 1.004814199766514,
+      "learning_rate": 1.8936427433906778e-07,
+      "loss": 1.5177,
+      "step": 7043
+    },
+    {
+      "epoch": 0.9399519615692554,
+      "grad_norm": 0.9607332814184879,
+      "learning_rate": 1.8852809815518514e-07,
+      "loss": 1.5538,
+      "step": 7044
+    },
+    {
+      "epoch": 0.940085401654657,
+      "grad_norm": 0.9487552535897916,
+      "learning_rate": 1.8769375464254503e-07,
+      "loss": 1.5562,
+      "step": 7045
+    },
+    {
+      "epoch": 0.9402188417400588,
+      "grad_norm": 0.9083170430971447,
+      "learning_rate": 1.868612439569939e-07,
+      "loss": 1.5094,
+      "step": 7046
+    },
+    {
+      "epoch": 0.9403522818254604,
+      "grad_norm": 0.9634251033944141,
+      "learning_rate": 1.8603056625403627e-07,
+      "loss": 1.5631,
+      "step": 7047
+    },
+    {
+      "epoch": 0.940485721910862,
+      "grad_norm": 0.9656510203512615,
+      "learning_rate": 1.8520172168883243e-07,
+      "loss": 1.5834,
+      "step": 7048
+    },
+    {
+      "epoch": 0.9406191619962637,
+      "grad_norm": 1.0000660756193365,
+      "learning_rate": 1.843747104162008e-07,
+      "loss": 1.582,
+      "step": 7049
+    },
+    {
+      "epoch": 0.9407526020816653,
+      "grad_norm": 1.1327819091986768,
+      "learning_rate": 1.8354953259061892e-07,
+      "loss": 1.5539,
+      "step": 7050
+    },
+    {
+      "epoch": 0.940886042167067,
+      "grad_norm": 0.9810323631990697,
+      "learning_rate": 1.8272618836621902e-07,
+      "loss": 1.5371,
+      "step": 7051
+    },
+    {
+      "epoch": 0.9410194822524687,
+      "grad_norm": 0.9130660342403393,
+      "learning_rate": 1.819046778967948e-07,
+      "loss": 1.522,
+      "step": 7052
+    },
+    {
+      "epoch": 0.9411529223378703,
+      "grad_norm": 0.9194929376758173,
+      "learning_rate": 1.8108500133579233e-07,
+      "loss": 1.4903,
+      "step": 7053
+    },
+    {
+      "epoch": 0.9412863624232719,
+      "grad_norm": 1.2769265280912176,
+      "learning_rate": 1.8026715883631917e-07,
+      "loss": 1.5293,
+      "step": 7054
+    },
+    {
+      "epoch": 0.9414198025086736,
+      "grad_norm": 0.983577585375118,
+      "learning_rate": 1.7945115055113983e-07,
+      "loss": 1.5471,
+      "step": 7055
+    },
+    {
+      "epoch": 0.9415532425940752,
+      "grad_norm": 1.2831644360579264,
+      "learning_rate": 1.7863697663267231e-07,
+      "loss": 1.5696,
+      "step": 7056
+    },
+    {
+      "epoch": 0.9416866826794769,
+      "grad_norm": 0.9321407428486242,
+      "learning_rate": 1.7782463723299947e-07,
+      "loss": 1.5499,
+      "step": 7057
+    },
+    {
+      "epoch": 0.9418201227648786,
+      "grad_norm": 0.9239769506776327,
+      "learning_rate": 1.7701413250385324e-07,
+      "loss": 1.552,
+      "step": 7058
+    },
+    {
+      "epoch": 0.9419535628502802,
+      "grad_norm": 0.9822924040248578,
+      "learning_rate": 1.7620546259662808e-07,
+      "loss": 1.5701,
+      "step": 7059
+    },
+    {
+      "epoch": 0.9420870029356819,
+      "grad_norm": 0.9264197624873486,
+      "learning_rate": 1.7539862766237536e-07,
+      "loss": 1.5301,
+      "step": 7060
+    },
+    {
+      "epoch": 0.9422204430210835,
+      "grad_norm": 0.9098879290150069,
+      "learning_rate": 1.7459362785180122e-07,
+      "loss": 1.497,
+      "step": 7061
+    },
+    {
+      "epoch": 0.9423538831064852,
+      "grad_norm": 0.9543713844981121,
+      "learning_rate": 1.7379046331527094e-07,
+      "loss": 1.4844,
+      "step": 7062
+    },
+    {
+      "epoch": 0.9424873231918869,
+      "grad_norm": 1.091370085062858,
+      "learning_rate": 1.729891342028067e-07,
+      "loss": 1.5259,
+      "step": 7063
+    },
+    {
+      "epoch": 0.9426207632772885,
+      "grad_norm": 1.0782943171835628,
+      "learning_rate": 1.7218964066408773e-07,
+      "loss": 1.5141,
+      "step": 7064
+    },
+    {
+      "epoch": 0.9427542033626901,
+      "grad_norm": 0.9167701552489729,
+      "learning_rate": 1.7139198284845005e-07,
+      "loss": 1.4914,
+      "step": 7065
+    },
+    {
+      "epoch": 0.9428876434480918,
+      "grad_norm": 0.9315361672522141,
+      "learning_rate": 1.7059616090488561e-07,
+      "loss": 1.553,
+      "step": 7066
+    },
+    {
+      "epoch": 0.9430210835334935,
+      "grad_norm": 0.941731632670938,
+      "learning_rate": 1.6980217498204777e-07,
+      "loss": 1.5184,
+      "step": 7067
+    },
+    {
+      "epoch": 0.9431545236188951,
+      "grad_norm": 1.0951605019526698,
+      "learning_rate": 1.6901002522824228e-07,
+      "loss": 1.518,
+      "step": 7068
+    },
+    {
+      "epoch": 0.9432879637042968,
+      "grad_norm": 0.9413032824460642,
+      "learning_rate": 1.6821971179143415e-07,
+      "loss": 1.4946,
+      "step": 7069
+    },
+    {
+      "epoch": 0.9434214037896984,
+      "grad_norm": 0.9158286385674823,
+      "learning_rate": 1.6743123481924417e-07,
+      "loss": 1.5623,
+      "step": 7070
+    },
+    {
+      "epoch": 0.9435548438751001,
+      "grad_norm": 0.9396586297163264,
+      "learning_rate": 1.666445944589523e-07,
+      "loss": 1.5608,
+      "step": 7071
+    },
+    {
+      "epoch": 0.9436882839605018,
+      "grad_norm": 0.9392711665239271,
+      "learning_rate": 1.6585979085749326e-07,
+      "loss": 1.5315,
+      "step": 7072
+    },
+    {
+      "epoch": 0.9438217240459034,
+      "grad_norm": 0.9382386473834549,
+      "learning_rate": 1.6507682416145865e-07,
+      "loss": 1.578,
+      "step": 7073
+    },
+    {
+      "epoch": 0.9439551641313051,
+      "grad_norm": 0.9007505568219435,
+      "learning_rate": 1.6429569451709925e-07,
+      "loss": 1.5409,
+      "step": 7074
+    },
+    {
+      "epoch": 0.9440886042167067,
+      "grad_norm": 0.9402668564767092,
+      "learning_rate": 1.635164020703206e-07,
+      "loss": 1.5789,
+      "step": 7075
+    },
+    {
+      "epoch": 0.9442220443021083,
+      "grad_norm": 0.9642072770932624,
+      "learning_rate": 1.6273894696668514e-07,
+      "loss": 1.5041,
+      "step": 7076
+    },
+    {
+      "epoch": 0.94435548438751,
+      "grad_norm": 0.9169461721627207,
+      "learning_rate": 1.6196332935141225e-07,
+      "loss": 1.5472,
+      "step": 7077
+    },
+    {
+      "epoch": 0.9444889244729117,
+      "grad_norm": 1.3273948346582414,
+      "learning_rate": 1.6118954936938046e-07,
+      "loss": 1.5359,
+      "step": 7078
+    },
+    {
+      "epoch": 0.9446223645583133,
+      "grad_norm": 1.129474162644648,
+      "learning_rate": 1.604176071651209e-07,
+      "loss": 1.5153,
+      "step": 7079
+    },
+    {
+      "epoch": 0.944755804643715,
+      "grad_norm": 0.9293810385080555,
+      "learning_rate": 1.596475028828237e-07,
+      "loss": 1.5414,
+      "step": 7080
+    },
+    {
+      "epoch": 0.9448892447291166,
+      "grad_norm": 0.9382288057598555,
+      "learning_rate": 1.588792366663383e-07,
+      "loss": 1.5622,
+      "step": 7081
+    },
+    {
+      "epoch": 0.9450226848145182,
+      "grad_norm": 0.989321226514514,
+      "learning_rate": 1.5811280865916435e-07,
+      "loss": 1.5804,
+      "step": 7082
+    },
+    {
+      "epoch": 0.94515612489992,
+      "grad_norm": 0.9390344091231897,
+      "learning_rate": 1.5734821900446507e-07,
+      "loss": 1.5219,
+      "step": 7083
+    },
+    {
+      "epoch": 0.9452895649853216,
+      "grad_norm": 1.0089198056152286,
+      "learning_rate": 1.56585467845054e-07,
+      "loss": 1.5103,
+      "step": 7084
+    },
+    {
+      "epoch": 0.9454230050707233,
+      "grad_norm": 0.9927138491828363,
+      "learning_rate": 1.5582455532340836e-07,
+      "loss": 1.5626,
+      "step": 7085
+    },
+    {
+      "epoch": 0.9455564451561249,
+      "grad_norm": 6.151724858717579,
+      "learning_rate": 1.5506548158165437e-07,
+      "loss": 1.5719,
+      "step": 7086
+    },
+    {
+      "epoch": 0.9456898852415265,
+      "grad_norm": 0.9752024728609847,
+      "learning_rate": 1.5430824676157974e-07,
+      "loss": 1.5091,
+      "step": 7087
+    },
+    {
+      "epoch": 0.9458233253269283,
+      "grad_norm": 0.9228271908165766,
+      "learning_rate": 1.535528510046258e-07,
+      "loss": 1.5471,
+      "step": 7088
+    },
+    {
+      "epoch": 0.9459567654123299,
+      "grad_norm": 0.9142093131081027,
+      "learning_rate": 1.5279929445189523e-07,
+      "loss": 1.56,
+      "step": 7089
+    },
+    {
+      "epoch": 0.9460902054977315,
+      "grad_norm": 1.040271354196889,
+      "learning_rate": 1.5204757724414098e-07,
+      "loss": 1.5649,
+      "step": 7090
+    },
+    {
+      "epoch": 0.9462236455831332,
+      "grad_norm": 0.944030106842017,
+      "learning_rate": 1.512976995217774e-07,
+      "loss": 1.5361,
+      "step": 7091
+    },
+    {
+      "epoch": 0.9463570856685348,
+      "grad_norm": 0.9345127025549201,
+      "learning_rate": 1.5054966142487027e-07,
+      "loss": 1.5856,
+      "step": 7092
+    },
+    {
+      "epoch": 0.9464905257539364,
+      "grad_norm": 0.9282759654458305,
+      "learning_rate": 1.4980346309314776e-07,
+      "loss": 1.5434,
+      "step": 7093
+    },
+    {
+      "epoch": 0.9466239658393382,
+      "grad_norm": 0.9226489628851323,
+      "learning_rate": 1.4905910466598727e-07,
+      "loss": 1.569,
+      "step": 7094
+    },
+    {
+      "epoch": 0.9467574059247398,
+      "grad_norm": 0.9279025309090545,
+      "learning_rate": 1.4831658628243096e-07,
+      "loss": 1.5276,
+      "step": 7095
+    },
+    {
+      "epoch": 0.9468908460101414,
+      "grad_norm": 0.9423761860152485,
+      "learning_rate": 1.4757590808117006e-07,
+      "loss": 1.5249,
+      "step": 7096
+    },
+    {
+      "epoch": 0.9470242860955431,
+      "grad_norm": 1.004907421039197,
+      "learning_rate": 1.4683707020055614e-07,
+      "loss": 1.5448,
+      "step": 7097
+    },
+    {
+      "epoch": 0.9471577261809447,
+      "grad_norm": 0.9258429181621549,
+      "learning_rate": 1.4610007277859328e-07,
+      "loss": 1.5371,
+      "step": 7098
+    },
+    {
+      "epoch": 0.9472911662663465,
+      "grad_norm": 0.9395593033423324,
+      "learning_rate": 1.453649159529469e-07,
+      "loss": 1.5647,
+      "step": 7099
+    },
+    {
+      "epoch": 0.9474246063517481,
+      "grad_norm": 0.9180402841843257,
+      "learning_rate": 1.446315998609349e-07,
+      "loss": 1.5026,
+      "step": 7100
+    },
+    {
+      "epoch": 0.9475580464371497,
+      "grad_norm": 0.9387391943964363,
+      "learning_rate": 1.4390012463953329e-07,
+      "loss": 1.4874,
+      "step": 7101
+    },
+    {
+      "epoch": 0.9476914865225514,
+      "grad_norm": 0.9403053185795006,
+      "learning_rate": 1.4317049042537167e-07,
+      "loss": 1.557,
+      "step": 7102
+    },
+    {
+      "epoch": 0.947824926607953,
+      "grad_norm": 0.9716572056708245,
+      "learning_rate": 1.424426973547377e-07,
+      "loss": 1.5695,
+      "step": 7103
+    },
+    {
+      "epoch": 0.9479583666933546,
+      "grad_norm": 0.9638053337694246,
+      "learning_rate": 1.4171674556357705e-07,
+      "loss": 1.528,
+      "step": 7104
+    },
+    {
+      "epoch": 0.9480918067787564,
+      "grad_norm": 0.9326742291482022,
+      "learning_rate": 1.4099263518748574e-07,
+      "loss": 1.5284,
+      "step": 7105
+    },
+    {
+      "epoch": 0.948225246864158,
+      "grad_norm": 0.9202879189907175,
+      "learning_rate": 1.402703663617222e-07,
+      "loss": 1.5171,
+      "step": 7106
+    },
+    {
+      "epoch": 0.9483586869495596,
+      "grad_norm": 0.9499293926304142,
+      "learning_rate": 1.395499392211963e-07,
+      "loss": 1.5009,
+      "step": 7107
+    },
+    {
+      "epoch": 0.9484921270349613,
+      "grad_norm": 0.9191237857115953,
+      "learning_rate": 1.388313539004771e-07,
+      "loss": 1.5325,
+      "step": 7108
+    },
+    {
+      "epoch": 0.948625567120363,
+      "grad_norm": 0.9281253464726291,
+      "learning_rate": 1.3811461053378606e-07,
+      "loss": 1.5305,
+      "step": 7109
+    },
+    {
+      "epoch": 0.9487590072057646,
+      "grad_norm": 0.9391089740766908,
+      "learning_rate": 1.3739970925500613e-07,
+      "loss": 1.5645,
+      "step": 7110
+    },
+    {
+      "epoch": 0.9488924472911663,
+      "grad_norm": 0.9906753900198737,
+      "learning_rate": 1.3668665019766937e-07,
+      "loss": 1.5169,
+      "step": 7111
+    },
+    {
+      "epoch": 0.9490258873765679,
+      "grad_norm": 0.9307202205615231,
+      "learning_rate": 1.35975433494967e-07,
+      "loss": 1.5244,
+      "step": 7112
+    },
+    {
+      "epoch": 0.9491593274619696,
+      "grad_norm": 0.9307053074686159,
+      "learning_rate": 1.3526605927974946e-07,
+      "loss": 1.5106,
+      "step": 7113
+    },
+    {
+      "epoch": 0.9492927675473712,
+      "grad_norm": 0.9528924062350823,
+      "learning_rate": 1.3455852768451738e-07,
+      "loss": 1.4496,
+      "step": 7114
+    },
+    {
+      "epoch": 0.9494262076327729,
+      "grad_norm": 1.4092694693096828,
+      "learning_rate": 1.3385283884143064e-07,
+      "loss": 1.5641,
+      "step": 7115
+    },
+    {
+      "epoch": 0.9495596477181746,
+      "grad_norm": 0.9508352494225291,
+      "learning_rate": 1.3314899288230154e-07,
+      "loss": 1.5428,
+      "step": 7116
+    },
+    {
+      "epoch": 0.9496930878035762,
+      "grad_norm": 1.0036182199394033,
+      "learning_rate": 1.3244698993860383e-07,
+      "loss": 1.5242,
+      "step": 7117
+    },
+    {
+      "epoch": 0.9498265278889778,
+      "grad_norm": 0.9239362307261929,
+      "learning_rate": 1.317468301414615e-07,
+      "loss": 1.5461,
+      "step": 7118
+    },
+    {
+      "epoch": 0.9499599679743795,
+      "grad_norm": 0.9265412901982545,
+      "learning_rate": 1.310485136216566e-07,
+      "loss": 1.5534,
+      "step": 7119
+    },
+    {
+      "epoch": 0.9500934080597812,
+      "grad_norm": 0.9381184097035782,
+      "learning_rate": 1.3035204050962702e-07,
+      "loss": 1.5232,
+      "step": 7120
+    },
+    {
+      "epoch": 0.9502268481451828,
+      "grad_norm": 1.0080877900305665,
+      "learning_rate": 1.2965741093546757e-07,
+      "loss": 1.5221,
+      "step": 7121
+    },
+    {
+      "epoch": 0.9503602882305845,
+      "grad_norm": 0.9245848574437733,
+      "learning_rate": 1.289646250289245e-07,
+      "loss": 1.5618,
+      "step": 7122
+    },
+    {
+      "epoch": 0.9504937283159861,
+      "grad_norm": 0.9340954109480951,
+      "learning_rate": 1.282736829194031e-07,
+      "loss": 1.539,
+      "step": 7123
+    },
+    {
+      "epoch": 0.9506271684013877,
+      "grad_norm": 0.9413706371278181,
+      "learning_rate": 1.2758458473596468e-07,
+      "loss": 1.5756,
+      "step": 7124
+    },
+    {
+      "epoch": 0.9507606084867894,
+      "grad_norm": 0.9435011931603419,
+      "learning_rate": 1.2689733060732512e-07,
+      "loss": 1.5168,
+      "step": 7125
+    },
+    {
+      "epoch": 0.9508940485721911,
+      "grad_norm": 0.9640763949138873,
+      "learning_rate": 1.2621192066185394e-07,
+      "loss": 1.5767,
+      "step": 7126
+    },
+    {
+      "epoch": 0.9510274886575928,
+      "grad_norm": 1.0666760841057328,
+      "learning_rate": 1.2552835502757765e-07,
+      "loss": 1.5459,
+      "step": 7127
+    },
+    {
+      "epoch": 0.9511609287429944,
+      "grad_norm": 0.9177228398177878,
+      "learning_rate": 1.2484663383217964e-07,
+      "loss": 1.483,
+      "step": 7128
+    },
+    {
+      "epoch": 0.951294368828396,
+      "grad_norm": 0.9400026184127935,
+      "learning_rate": 1.241667572029992e-07,
+      "loss": 1.5753,
+      "step": 7129
+    },
+    {
+      "epoch": 0.9514278089137977,
+      "grad_norm": 1.0492605217878297,
+      "learning_rate": 1.2348872526702693e-07,
+      "loss": 1.5503,
+      "step": 7130
+    },
+    {
+      "epoch": 0.9515612489991994,
+      "grad_norm": 0.9251817708144794,
+      "learning_rate": 1.2281253815091154e-07,
+      "loss": 1.5143,
+      "step": 7131
+    },
+    {
+      "epoch": 0.951694689084601,
+      "grad_norm": 0.940577041105443,
+      "learning_rate": 1.2213819598095979e-07,
+      "loss": 1.5524,
+      "step": 7132
+    },
+    {
+      "epoch": 0.9518281291700027,
+      "grad_norm": 0.9912924069921497,
+      "learning_rate": 1.2146569888312865e-07,
+      "loss": 1.505,
+      "step": 7133
+    },
+    {
+      "epoch": 0.9519615692554043,
+      "grad_norm": 1.0163650859754099,
+      "learning_rate": 1.2079504698303324e-07,
+      "loss": 1.5415,
+      "step": 7134
+    },
+    {
+      "epoch": 0.9520950093408059,
+      "grad_norm": 0.91474346735063,
+      "learning_rate": 1.2012624040594445e-07,
+      "loss": 1.5454,
+      "step": 7135
+    },
+    {
+      "epoch": 0.9522284494262077,
+      "grad_norm": 0.9327035920232858,
+      "learning_rate": 1.194592792767879e-07,
+      "loss": 1.5117,
+      "step": 7136
+    },
+    {
+      "epoch": 0.9523618895116093,
+      "grad_norm": 0.9217306273503845,
+      "learning_rate": 1.1879416372014285e-07,
+      "loss": 1.529,
+      "step": 7137
+    },
+    {
+      "epoch": 0.9524953295970109,
+      "grad_norm": 0.9609918293396442,
+      "learning_rate": 1.1813089386024657e-07,
+      "loss": 1.5807,
+      "step": 7138
+    },
+    {
+      "epoch": 0.9526287696824126,
+      "grad_norm": 1.112063642124847,
+      "learning_rate": 1.1746946982098995e-07,
+      "loss": 1.5532,
+      "step": 7139
+    },
+    {
+      "epoch": 0.9527622097678142,
+      "grad_norm": 0.9466676036098123,
+      "learning_rate": 1.1680989172591972e-07,
+      "loss": 1.52,
+      "step": 7140
+    },
+    {
+      "epoch": 0.952895649853216,
+      "grad_norm": 0.9269239682104546,
+      "learning_rate": 1.1615215969823734e-07,
+      "loss": 1.5417,
+      "step": 7141
+    },
+    {
+      "epoch": 0.9530290899386176,
+      "grad_norm": 1.38625103234689,
+      "learning_rate": 1.1549627386080009e-07,
+      "loss": 1.5533,
+      "step": 7142
+    },
+    {
+      "epoch": 0.9531625300240192,
+      "grad_norm": 0.9665414589291429,
+      "learning_rate": 1.1484223433611885e-07,
+      "loss": 1.5013,
+      "step": 7143
+    },
+    {
+      "epoch": 0.9532959701094209,
+      "grad_norm": 1.1016470069768787,
+      "learning_rate": 1.1419004124636146e-07,
+      "loss": 1.5066,
+      "step": 7144
+    },
+    {
+      "epoch": 0.9534294101948225,
+      "grad_norm": 1.1136680127273073,
+      "learning_rate": 1.1353969471335047e-07,
+      "loss": 1.5206,
+      "step": 7145
+    },
+    {
+      "epoch": 0.9535628502802241,
+      "grad_norm": 0.9664150023916923,
+      "learning_rate": 1.1289119485856315e-07,
+      "loss": 1.5303,
+      "step": 7146
+    },
+    {
+      "epoch": 0.9536962903656259,
+      "grad_norm": 0.9298992429566982,
+      "learning_rate": 1.122445418031315e-07,
+      "loss": 1.5278,
+      "step": 7147
+    },
+    {
+      "epoch": 0.9538297304510275,
+      "grad_norm": 1.1638106219236741,
+      "learning_rate": 1.1159973566784221e-07,
+      "loss": 1.5079,
+      "step": 7148
+    },
+    {
+      "epoch": 0.9539631705364291,
+      "grad_norm": 0.9586600372591042,
+      "learning_rate": 1.1095677657314008e-07,
+      "loss": 1.5023,
+      "step": 7149
+    },
+    {
+      "epoch": 0.9540966106218308,
+      "grad_norm": 0.9231194918030783,
+      "learning_rate": 1.1031566463912014e-07,
+      "loss": 1.5314,
+      "step": 7150
+    },
+    {
+      "epoch": 0.9542300507072324,
+      "grad_norm": 0.9169704635554706,
+      "learning_rate": 1.0967639998553659e-07,
+      "loss": 1.5137,
+      "step": 7151
+    },
+    {
+      "epoch": 0.9543634907926342,
+      "grad_norm": 0.9623890401190696,
+      "learning_rate": 1.0903898273179503e-07,
+      "loss": 1.54,
+      "step": 7152
+    },
+    {
+      "epoch": 0.9544969308780358,
+      "grad_norm": 0.9269400628043825,
+      "learning_rate": 1.0840341299695911e-07,
+      "loss": 1.5609,
+      "step": 7153
+    },
+    {
+      "epoch": 0.9546303709634374,
+      "grad_norm": 0.9465474447152321,
+      "learning_rate": 1.0776969089974609e-07,
+      "loss": 1.5059,
+      "step": 7154
+    },
+    {
+      "epoch": 0.9547638110488391,
+      "grad_norm": 0.9356339177793419,
+      "learning_rate": 1.0713781655852684e-07,
+      "loss": 1.545,
+      "step": 7155
+    },
+    {
+      "epoch": 0.9548972511342407,
+      "grad_norm": 0.927542167441195,
+      "learning_rate": 1.0650779009132917e-07,
+      "loss": 1.5608,
+      "step": 7156
+    },
+    {
+      "epoch": 0.9550306912196423,
+      "grad_norm": 0.9369805873387638,
+      "learning_rate": 1.0587961161583448e-07,
+      "loss": 1.5508,
+      "step": 7157
+    },
+    {
+      "epoch": 0.9551641313050441,
+      "grad_norm": 0.9523970569734105,
+      "learning_rate": 1.0525328124938006e-07,
+      "loss": 1.5117,
+      "step": 7158
+    },
+    {
+      "epoch": 0.9552975713904457,
+      "grad_norm": 0.9097601640170604,
+      "learning_rate": 1.0462879910895674e-07,
+      "loss": 1.5232,
+      "step": 7159
+    },
+    {
+      "epoch": 0.9554310114758473,
+      "grad_norm": 0.9589791998825555,
+      "learning_rate": 1.0400616531121011e-07,
+      "loss": 1.56,
+      "step": 7160
+    },
+    {
+      "epoch": 0.955564451561249,
+      "grad_norm": 0.9522319609125601,
+      "learning_rate": 1.033853799724427e-07,
+      "loss": 1.543,
+      "step": 7161
+    },
+    {
+      "epoch": 0.9556978916466506,
+      "grad_norm": 0.9285567819238145,
+      "learning_rate": 1.0276644320860729e-07,
+      "loss": 1.5558,
+      "step": 7162
+    },
+    {
+      "epoch": 0.9558313317320523,
+      "grad_norm": 0.9595527770779047,
+      "learning_rate": 1.021493551353181e-07,
+      "loss": 1.5226,
+      "step": 7163
+    },
+    {
+      "epoch": 0.955964771817454,
+      "grad_norm": 0.9018522623887555,
+      "learning_rate": 1.0153411586783734e-07,
+      "loss": 1.5229,
+      "step": 7164
+    },
+    {
+      "epoch": 0.9560982119028556,
+      "grad_norm": 1.0712384937201727,
+      "learning_rate": 1.0092072552108534e-07,
+      "loss": 1.5517,
+      "step": 7165
+    },
+    {
+      "epoch": 0.9562316519882573,
+      "grad_norm": 0.9341382301028408,
+      "learning_rate": 1.0030918420963598e-07,
+      "loss": 1.5233,
+      "step": 7166
+    },
+    {
+      "epoch": 0.9563650920736589,
+      "grad_norm": 0.9928816977811972,
+      "learning_rate": 9.969949204772011e-08,
+      "loss": 1.5388,
+      "step": 7167
+    },
+    {
+      "epoch": 0.9564985321590606,
+      "grad_norm": 0.9410197074503386,
+      "learning_rate": 9.909164914921886e-08,
+      "loss": 1.6122,
+      "step": 7168
+    },
+    {
+      "epoch": 0.9566319722444623,
+      "grad_norm": 0.9028807413403246,
+      "learning_rate": 9.848565562767143e-08,
+      "loss": 1.4786,
+      "step": 7169
+    },
+    {
+      "epoch": 0.9567654123298639,
+      "grad_norm": 0.9567117405537122,
+      "learning_rate": 9.788151159627168e-08,
+      "loss": 1.5112,
+      "step": 7170
+    },
+    {
+      "epoch": 0.9568988524152655,
+      "grad_norm": 0.9811327272197673,
+      "learning_rate": 9.727921716786492e-08,
+      "loss": 1.5945,
+      "step": 7171
+    },
+    {
+      "epoch": 0.9570322925006672,
+      "grad_norm": 1.0524690809980066,
+      "learning_rate": 9.667877245495338e-08,
+      "loss": 1.5584,
+      "step": 7172
+    },
+    {
+      "epoch": 0.9571657325860689,
+      "grad_norm": 0.9613936267762925,
+      "learning_rate": 9.608017756969512e-08,
+      "loss": 1.5532,
+      "step": 7173
+    },
+    {
+      "epoch": 0.9572991726714705,
+      "grad_norm": 1.2156726555516593,
+      "learning_rate": 9.548343262389736e-08,
+      "loss": 1.6236,
+      "step": 7174
+    },
+    {
+      "epoch": 0.9574326127568722,
+      "grad_norm": 0.9579182785549096,
+      "learning_rate": 9.488853772902762e-08,
+      "loss": 1.5155,
+      "step": 7175
+    },
+    {
+      "epoch": 0.9575660528422738,
+      "grad_norm": 0.9311509207700649,
+      "learning_rate": 9.429549299620589e-08,
+      "loss": 1.5794,
+      "step": 7176
+    },
+    {
+      "epoch": 0.9576994929276754,
+      "grad_norm": 0.9409890634401489,
+      "learning_rate": 9.370429853620578e-08,
+      "loss": 1.5597,
+      "step": 7177
+    },
+    {
+      "epoch": 0.9578329330130771,
+      "grad_norm": 0.9199706623701136,
+      "learning_rate": 9.311495445945451e-08,
+      "loss": 1.5511,
+      "step": 7178
+    },
+    {
+      "epoch": 0.9579663730984788,
+      "grad_norm": 0.9160829798550142,
+      "learning_rate": 9.252746087603626e-08,
+      "loss": 1.5512,
+      "step": 7179
+    },
+    {
+      "epoch": 0.9580998131838805,
+      "grad_norm": 0.9742308156882673,
+      "learning_rate": 9.194181789568657e-08,
+      "loss": 1.5763,
+      "step": 7180
+    },
+    {
+      "epoch": 0.9582332532692821,
+      "grad_norm": 1.6586965890912353,
+      "learning_rate": 9.135802562779794e-08,
+      "loss": 1.4843,
+      "step": 7181
+    },
+    {
+      "epoch": 0.9583666933546837,
+      "grad_norm": 0.9929783009078508,
+      "learning_rate": 9.077608418141648e-08,
+      "loss": 1.565,
+      "step": 7182
+    },
+    {
+      "epoch": 0.9585001334400854,
+      "grad_norm": 0.9342398271698058,
+      "learning_rate": 9.019599366524079e-08,
+      "loss": 1.524,
+      "step": 7183
+    },
+    {
+      "epoch": 0.9586335735254871,
+      "grad_norm": 0.9169837288541282,
+      "learning_rate": 8.961775418762752e-08,
+      "loss": 1.495,
+      "step": 7184
+    },
+    {
+      "epoch": 0.9587670136108887,
+      "grad_norm": 0.9084724578571916,
+      "learning_rate": 8.904136585658251e-08,
+      "loss": 1.5988,
+      "step": 7185
+    },
+    {
+      "epoch": 0.9589004536962904,
+      "grad_norm": 0.9504512059920357,
+      "learning_rate": 8.846682877977075e-08,
+      "loss": 1.5612,
+      "step": 7186
+    },
+    {
+      "epoch": 0.959033893781692,
+      "grad_norm": 0.9505995914507539,
+      "learning_rate": 8.78941430645075e-08,
+      "loss": 1.5368,
+      "step": 7187
+    },
+    {
+      "epoch": 0.9591673338670936,
+      "grad_norm": 1.1809298932125372,
+      "learning_rate": 8.732330881776608e-08,
+      "loss": 1.5281,
+      "step": 7188
+    },
+    {
+      "epoch": 0.9593007739524954,
+      "grad_norm": 0.9314187782139279,
+      "learning_rate": 8.675432614617008e-08,
+      "loss": 1.5489,
+      "step": 7189
+    },
+    {
+      "epoch": 0.959434214037897,
+      "grad_norm": 0.9232377132180151,
+      "learning_rate": 8.618719515599894e-08,
+      "loss": 1.5449,
+      "step": 7190
+    },
+    {
+      "epoch": 0.9595676541232986,
+      "grad_norm": 1.0270042691413004,
+      "learning_rate": 8.56219159531868e-08,
+      "loss": 1.5986,
+      "step": 7191
+    },
+    {
+      "epoch": 0.9597010942087003,
+      "grad_norm": 1.2061551149492193,
+      "learning_rate": 8.505848864332145e-08,
+      "loss": 1.5768,
+      "step": 7192
+    },
+    {
+      "epoch": 0.9598345342941019,
+      "grad_norm": 0.9340146167897232,
+      "learning_rate": 8.449691333164423e-08,
+      "loss": 1.5385,
+      "step": 7193
+    },
+    {
+      "epoch": 0.9599679743795037,
+      "grad_norm": 0.9296173201446742,
+      "learning_rate": 8.393719012305124e-08,
+      "loss": 1.5808,
+      "step": 7194
+    },
+    {
+      "epoch": 0.9601014144649053,
+      "grad_norm": 0.9802388493036821,
+      "learning_rate": 8.337931912209329e-08,
+      "loss": 1.6037,
+      "step": 7195
+    },
+    {
+      "epoch": 0.9602348545503069,
+      "grad_norm": 0.9331120869630394,
+      "learning_rate": 8.28233004329737e-08,
+      "loss": 1.5445,
+      "step": 7196
+    },
+    {
+      "epoch": 0.9603682946357086,
+      "grad_norm": 1.025665909891443,
+      "learning_rate": 8.22691341595494e-08,
+      "loss": 1.5029,
+      "step": 7197
+    },
+    {
+      "epoch": 0.9605017347211102,
+      "grad_norm": 0.9441651844649627,
+      "learning_rate": 8.171682040533314e-08,
+      "loss": 1.5366,
+      "step": 7198
+    },
+    {
+      "epoch": 0.9606351748065118,
+      "grad_norm": 0.914204298926644,
+      "learning_rate": 8.116635927349126e-08,
+      "loss": 1.4671,
+      "step": 7199
+    },
+    {
+      "epoch": 0.9607686148919136,
+      "grad_norm": 0.9547090251419438,
+      "learning_rate": 8.061775086684376e-08,
+      "loss": 1.5321,
+      "step": 7200
+    },
+    {
+      "epoch": 0.9609020549773152,
+      "grad_norm": 0.9231643897440952,
+      "learning_rate": 8.007099528786311e-08,
+      "loss": 1.4749,
+      "step": 7201
+    },
+    {
+      "epoch": 0.9610354950627168,
+      "grad_norm": 0.9330739495247855,
+      "learning_rate": 7.952609263867872e-08,
+      "loss": 1.549,
+      "step": 7202
+    },
+    {
+      "epoch": 0.9611689351481185,
+      "grad_norm": 0.9436536688017126,
+      "learning_rate": 7.89830430210714e-08,
+      "loss": 1.573,
+      "step": 7203
+    },
+    {
+      "epoch": 0.9613023752335201,
+      "grad_norm": 0.9289138338185312,
+      "learning_rate": 7.844184653647669e-08,
+      "loss": 1.4821,
+      "step": 7204
+    },
+    {
+      "epoch": 0.9614358153189217,
+      "grad_norm": 1.0587966606284311,
+      "learning_rate": 7.79025032859837e-08,
+      "loss": 1.5745,
+      "step": 7205
+    },
+    {
+      "epoch": 0.9615692554043235,
+      "grad_norm": 0.9391981736849577,
+      "learning_rate": 7.736501337033519e-08,
+      "loss": 1.5196,
+      "step": 7206
+    },
+    {
+      "epoch": 0.9617026954897251,
+      "grad_norm": 0.921984477255766,
+      "learning_rate": 7.682937688992975e-08,
+      "loss": 1.5816,
+      "step": 7207
+    },
+    {
+      "epoch": 0.9618361355751268,
+      "grad_norm": 0.9210876386975504,
+      "learning_rate": 7.629559394481622e-08,
+      "loss": 1.5665,
+      "step": 7208
+    },
+    {
+      "epoch": 0.9619695756605284,
+      "grad_norm": 0.9320013031423707,
+      "learning_rate": 7.576366463470042e-08,
+      "loss": 1.4974,
+      "step": 7209
+    },
+    {
+      "epoch": 0.96210301574593,
+      "grad_norm": 1.108953281361308,
+      "learning_rate": 7.523358905894063e-08,
+      "loss": 1.548,
+      "step": 7210
+    },
+    {
+      "epoch": 0.9622364558313318,
+      "grad_norm": 0.9286055208909001,
+      "learning_rate": 7.470536731654876e-08,
+      "loss": 1.4809,
+      "step": 7211
+    },
+    {
+      "epoch": 0.9623698959167334,
+      "grad_norm": 0.9397318541414119,
+      "learning_rate": 7.417899950619035e-08,
+      "loss": 1.5441,
+      "step": 7212
+    },
+    {
+      "epoch": 0.962503336002135,
+      "grad_norm": 0.928266538779687,
+      "learning_rate": 7.36544857261845e-08,
+      "loss": 1.5575,
+      "step": 7213
+    },
+    {
+      "epoch": 0.9626367760875367,
+      "grad_norm": 1.1602227605302893,
+      "learning_rate": 7.31318260745062e-08,
+      "loss": 1.4993,
+      "step": 7214
+    },
+    {
+      "epoch": 0.9627702161729383,
+      "grad_norm": 0.933054423819918,
+      "learning_rate": 7.261102064878067e-08,
+      "loss": 1.5281,
+      "step": 7215
+    },
+    {
+      "epoch": 0.96290365625834,
+      "grad_norm": 0.984048837363307,
+      "learning_rate": 7.2092069546289e-08,
+      "loss": 1.5852,
+      "step": 7216
+    },
+    {
+      "epoch": 0.9630370963437417,
+      "grad_norm": 0.9736412117865043,
+      "learning_rate": 7.157497286396475e-08,
+      "loss": 1.5516,
+      "step": 7217
+    },
+    {
+      "epoch": 0.9631705364291433,
+      "grad_norm": 1.0788935314073798,
+      "learning_rate": 7.105973069839622e-08,
+      "loss": 1.5497,
+      "step": 7218
+    },
+    {
+      "epoch": 0.963303976514545,
+      "grad_norm": 1.0829067853857433,
+      "learning_rate": 7.054634314582531e-08,
+      "loss": 1.5787,
+      "step": 7219
+    },
+    {
+      "epoch": 0.9634374165999466,
+      "grad_norm": 0.9424299932130654,
+      "learning_rate": 7.003481030214642e-08,
+      "loss": 1.5995,
+      "step": 7220
+    },
+    {
+      "epoch": 0.9635708566853483,
+      "grad_norm": 0.9587712887267895,
+      "learning_rate": 6.952513226290758e-08,
+      "loss": 1.5399,
+      "step": 7221
+    },
+    {
+      "epoch": 0.96370429677075,
+      "grad_norm": 0.9129861391767959,
+      "learning_rate": 6.90173091233115e-08,
+      "loss": 1.5154,
+      "step": 7222
+    },
+    {
+      "epoch": 0.9638377368561516,
+      "grad_norm": 1.0939988766367443,
+      "learning_rate": 6.851134097821344e-08,
+      "loss": 1.5632,
+      "step": 7223
+    },
+    {
+      "epoch": 0.9639711769415532,
+      "grad_norm": 0.936664559385459,
+      "learning_rate": 6.800722792212333e-08,
+      "loss": 1.5212,
+      "step": 7224
+    },
+    {
+      "epoch": 0.9641046170269549,
+      "grad_norm": 0.93098974114669,
+      "learning_rate": 6.750497004920253e-08,
+      "loss": 1.5135,
+      "step": 7225
+    },
+    {
+      "epoch": 0.9642380571123566,
+      "grad_norm": 0.9927276595661103,
+      "learning_rate": 6.700456745326822e-08,
+      "loss": 1.5372,
+      "step": 7226
+    },
+    {
+      "epoch": 0.9643714971977582,
+      "grad_norm": 1.0843133033461316,
+      "learning_rate": 6.650602022778785e-08,
+      "loss": 1.5053,
+      "step": 7227
+    },
+    {
+      "epoch": 0.9645049372831599,
+      "grad_norm": 1.1442758247625016,
+      "learning_rate": 6.600932846588692e-08,
+      "loss": 1.5681,
+      "step": 7228
+    },
+    {
+      "epoch": 0.9646383773685615,
+      "grad_norm": 0.9544653994484005,
+      "learning_rate": 6.551449226034124e-08,
+      "loss": 1.4815,
+      "step": 7229
+    },
+    {
+      "epoch": 0.9647718174539631,
+      "grad_norm": 0.9771673417993357,
+      "learning_rate": 6.502151170357906e-08,
+      "loss": 1.5094,
+      "step": 7230
+    },
+    {
+      "epoch": 0.9649052575393648,
+      "grad_norm": 0.9952586049134462,
+      "learning_rate": 6.453038688768454e-08,
+      "loss": 1.5057,
+      "step": 7231
+    },
+    {
+      "epoch": 0.9650386976247665,
+      "grad_norm": 0.9468439711909263,
+      "learning_rate": 6.404111790439427e-08,
+      "loss": 1.5695,
+      "step": 7232
+    },
+    {
+      "epoch": 0.9651721377101682,
+      "grad_norm": 0.9741329945289054,
+      "learning_rate": 6.355370484509848e-08,
+      "loss": 1.5713,
+      "step": 7233
+    },
+    {
+      "epoch": 0.9653055777955698,
+      "grad_norm": 1.18170513807492,
+      "learning_rate": 6.306814780083992e-08,
+      "loss": 1.521,
+      "step": 7234
+    },
+    {
+      "epoch": 0.9654390178809714,
+      "grad_norm": 0.9705328509871132,
+      "learning_rate": 6.258444686231491e-08,
+      "loss": 1.5803,
+      "step": 7235
+    },
+    {
+      "epoch": 0.9655724579663731,
+      "grad_norm": 0.8989950439543595,
+      "learning_rate": 6.21026021198734e-08,
+      "loss": 1.4914,
+      "step": 7236
+    },
+    {
+      "epoch": 0.9657058980517748,
+      "grad_norm": 0.9190782535619162,
+      "learning_rate": 6.162261366351785e-08,
+      "loss": 1.5458,
+      "step": 7237
+    },
+    {
+      "epoch": 0.9658393381371764,
+      "grad_norm": 0.9325432801015195,
+      "learning_rate": 6.114448158290653e-08,
+      "loss": 1.557,
+      "step": 7238
+    },
+    {
+      "epoch": 0.9659727782225781,
+      "grad_norm": 0.9129405790473425,
+      "learning_rate": 6.066820596734801e-08,
+      "loss": 1.5548,
+      "step": 7239
+    },
+    {
+      "epoch": 0.9661062183079797,
+      "grad_norm": 0.9077011984641612,
+      "learning_rate": 6.019378690580447e-08,
+      "loss": 1.515,
+      "step": 7240
+    },
+    {
+      "epoch": 0.9662396583933813,
+      "grad_norm": 0.9878409436375739,
+      "learning_rate": 5.972122448689278e-08,
+      "loss": 1.5372,
+      "step": 7241
+    },
+    {
+      "epoch": 0.966373098478783,
+      "grad_norm": 0.9403367481562539,
+      "learning_rate": 5.925051879888233e-08,
+      "loss": 1.531,
+      "step": 7242
+    },
+    {
+      "epoch": 0.9665065385641847,
+      "grad_norm": 0.9133740011830598,
+      "learning_rate": 5.878166992969503e-08,
+      "loss": 1.5515,
+      "step": 7243
+    },
+    {
+      "epoch": 0.9666399786495863,
+      "grad_norm": 0.9468373367676625,
+      "learning_rate": 5.8314677966906376e-08,
+      "loss": 1.5626,
+      "step": 7244
+    },
+    {
+      "epoch": 0.966773418734988,
+      "grad_norm": 0.9583047387897976,
+      "learning_rate": 5.784954299774548e-08,
+      "loss": 1.5667,
+      "step": 7245
+    },
+    {
+      "epoch": 0.9669068588203896,
+      "grad_norm": 1.1076391186400782,
+      "learning_rate": 5.738626510909506e-08,
+      "loss": 1.5306,
+      "step": 7246
+    },
+    {
+      "epoch": 0.9670402989057914,
+      "grad_norm": 0.9531885401281327,
+      "learning_rate": 5.692484438748924e-08,
+      "loss": 1.5195,
+      "step": 7247
+    },
+    {
+      "epoch": 0.967173738991193,
+      "grad_norm": 0.9286226294276955,
+      "learning_rate": 5.646528091911574e-08,
+      "loss": 1.5468,
+      "step": 7248
+    },
+    {
+      "epoch": 0.9673071790765946,
+      "grad_norm": 1.0932763529633334,
+      "learning_rate": 5.6007574789817e-08,
+      "loss": 1.5583,
+      "step": 7249
+    },
+    {
+      "epoch": 0.9674406191619963,
+      "grad_norm": 0.9368950168319131,
+      "learning_rate": 5.5551726085086854e-08,
+      "loss": 1.5233,
+      "step": 7250
+    },
+    {
+      "epoch": 0.9675740592473979,
+      "grad_norm": 1.0046496115780676,
+      "learning_rate": 5.509773489007164e-08,
+      "loss": 1.5453,
+      "step": 7251
+    },
+    {
+      "epoch": 0.9677074993327995,
+      "grad_norm": 0.9110736411180317,
+      "learning_rate": 5.4645601289572414e-08,
+      "loss": 1.5221,
+      "step": 7252
+    },
+    {
+      "epoch": 0.9678409394182013,
+      "grad_norm": 0.9490066324920002,
+      "learning_rate": 5.419532536804384e-08,
+      "loss": 1.5655,
+      "step": 7253
+    },
+    {
+      "epoch": 0.9679743795036029,
+      "grad_norm": 0.962378240370102,
+      "learning_rate": 5.3746907209590856e-08,
+      "loss": 1.5207,
+      "step": 7254
+    },
+    {
+      "epoch": 0.9681078195890045,
+      "grad_norm": 1.1335586421572892,
+      "learning_rate": 5.3300346897973143e-08,
+      "loss": 1.5354,
+      "step": 7255
+    },
+    {
+      "epoch": 0.9682412596744062,
+      "grad_norm": 0.9713944583777968,
+      "learning_rate": 5.285564451660285e-08,
+      "loss": 1.5339,
+      "step": 7256
+    },
+    {
+      "epoch": 0.9683746997598078,
+      "grad_norm": 0.9407589092657578,
+      "learning_rate": 5.2412800148546884e-08,
+      "loss": 1.5333,
+      "step": 7257
+    },
+    {
+      "epoch": 0.9685081398452094,
+      "grad_norm": 0.9498219588447119,
+      "learning_rate": 5.197181387652128e-08,
+      "loss": 1.5341,
+      "step": 7258
+    },
+    {
+      "epoch": 0.9686415799306112,
+      "grad_norm": 1.184152865294345,
+      "learning_rate": 5.153268578289794e-08,
+      "loss": 1.5766,
+      "step": 7259
+    },
+    {
+      "epoch": 0.9687750200160128,
+      "grad_norm": 0.9171966164362452,
+      "learning_rate": 5.109541594970235e-08,
+      "loss": 1.4931,
+      "step": 7260
+    },
+    {
+      "epoch": 0.9689084601014145,
+      "grad_norm": 0.9294958426402554,
+      "learning_rate": 5.06600044586103e-08,
+      "loss": 1.507,
+      "step": 7261
+    },
+    {
+      "epoch": 0.9690419001868161,
+      "grad_norm": 1.0547850975106445,
+      "learning_rate": 5.022645139095117e-08,
+      "loss": 1.568,
+      "step": 7262
+    },
+    {
+      "epoch": 0.9691753402722177,
+      "grad_norm": 0.9867765801059861,
+      "learning_rate": 4.979475682770907e-08,
+      "loss": 1.516,
+      "step": 7263
+    },
+    {
+      "epoch": 0.9693087803576195,
+      "grad_norm": 0.9697836144717429,
+      "learning_rate": 4.936492084951949e-08,
+      "loss": 1.5803,
+      "step": 7264
+    },
+    {
+      "epoch": 0.9694422204430211,
+      "grad_norm": 0.9720384959061538,
+      "learning_rate": 4.893694353666934e-08,
+      "loss": 1.5052,
+      "step": 7265
+    },
+    {
+      "epoch": 0.9695756605284227,
+      "grad_norm": 0.9188699165623265,
+      "learning_rate": 4.851082496910242e-08,
+      "loss": 1.5465,
+      "step": 7266
+    },
+    {
+      "epoch": 0.9697091006138244,
+      "grad_norm": 0.9508384028938736,
+      "learning_rate": 4.808656522641064e-08,
+      "loss": 1.5879,
+      "step": 7267
+    },
+    {
+      "epoch": 0.969842540699226,
+      "grad_norm": 1.2180128273543447,
+      "learning_rate": 4.766416438784172e-08,
+      "loss": 1.5716,
+      "step": 7268
+    },
+    {
+      "epoch": 0.9699759807846277,
+      "grad_norm": 0.9353158222871983,
+      "learning_rate": 4.7243622532294756e-08,
+      "loss": 1.5542,
+      "step": 7269
+    },
+    {
+      "epoch": 0.9701094208700294,
+      "grad_norm": 0.9324250384056547,
+      "learning_rate": 4.682493973832358e-08,
+      "loss": 1.5415,
+      "step": 7270
+    },
+    {
+      "epoch": 0.970242860955431,
+      "grad_norm": 0.9495787614825245,
+      "learning_rate": 4.6408116084132313e-08,
+      "loss": 1.5307,
+      "step": 7271
+    },
+    {
+      "epoch": 0.9703763010408326,
+      "grad_norm": 0.9474456637216822,
+      "learning_rate": 4.599315164757867e-08,
+      "loss": 1.5638,
+      "step": 7272
+    },
+    {
+      "epoch": 0.9705097411262343,
+      "grad_norm": 0.9360048271462087,
+      "learning_rate": 4.558004650617398e-08,
+      "loss": 1.5106,
+      "step": 7273
+    },
+    {
+      "epoch": 0.970643181211636,
+      "grad_norm": 1.0394175365220302,
+      "learning_rate": 4.5168800737080965e-08,
+      "loss": 1.5319,
+      "step": 7274
+    },
+    {
+      "epoch": 0.9707766212970377,
+      "grad_norm": 0.9512409507257436,
+      "learning_rate": 4.4759414417117064e-08,
+      "loss": 1.542,
+      "step": 7275
+    },
+    {
+      "epoch": 0.9709100613824393,
+      "grad_norm": 0.938780333252122,
+      "learning_rate": 4.435188762274778e-08,
+      "loss": 1.5228,
+      "step": 7276
+    },
+    {
+      "epoch": 0.9710435014678409,
+      "grad_norm": 0.9585863277465944,
+      "learning_rate": 4.3946220430098885e-08,
+      "loss": 1.5299,
+      "step": 7277
+    },
+    {
+      "epoch": 0.9711769415532426,
+      "grad_norm": 1.2292729065904253,
+      "learning_rate": 4.354241291494088e-08,
+      "loss": 1.5671,
+      "step": 7278
+    },
+    {
+      "epoch": 0.9713103816386442,
+      "grad_norm": 0.9701786083988774,
+      "learning_rate": 4.314046515270121e-08,
+      "loss": 1.5487,
+      "step": 7279
+    },
+    {
+      "epoch": 0.9714438217240459,
+      "grad_norm": 0.9290988392315854,
+      "learning_rate": 4.2740377218459805e-08,
+      "loss": 1.5489,
+      "step": 7280
+    },
+    {
+      "epoch": 0.9715772618094476,
+      "grad_norm": 0.9516211408003198,
+      "learning_rate": 4.234214918694912e-08,
+      "loss": 1.5158,
+      "step": 7281
+    },
+    {
+      "epoch": 0.9717107018948492,
+      "grad_norm": 0.9216274492882804,
+      "learning_rate": 4.194578113255188e-08,
+      "loss": 1.4825,
+      "step": 7282
+    },
+    {
+      "epoch": 0.9718441419802508,
+      "grad_norm": 0.956433000469751,
+      "learning_rate": 4.1551273129306624e-08,
+      "loss": 1.5503,
+      "step": 7283
+    },
+    {
+      "epoch": 0.9719775820656525,
+      "grad_norm": 0.959454062575314,
+      "learning_rate": 4.11586252509022e-08,
+      "loss": 1.5901,
+      "step": 7284
+    },
+    {
+      "epoch": 0.9721110221510542,
+      "grad_norm": 0.9623469747948837,
+      "learning_rate": 4.076783757068103e-08,
+      "loss": 1.5469,
+      "step": 7285
+    },
+    {
+      "epoch": 0.9722444622364559,
+      "grad_norm": 0.941165654507255,
+      "learning_rate": 4.037891016163697e-08,
+      "loss": 1.5266,
+      "step": 7286
+    },
+    {
+      "epoch": 0.9723779023218575,
+      "grad_norm": 0.953573546159197,
+      "learning_rate": 3.999184309641857e-08,
+      "loss": 1.5544,
+      "step": 7287
+    },
+    {
+      "epoch": 0.9725113424072591,
+      "grad_norm": 1.0783246600687082,
+      "learning_rate": 3.960663644732465e-08,
+      "loss": 1.515,
+      "step": 7288
+    },
+    {
+      "epoch": 0.9726447824926608,
+      "grad_norm": 0.9455437522062007,
+      "learning_rate": 3.922329028630767e-08,
+      "loss": 1.62,
+      "step": 7289
+    },
+    {
+      "epoch": 0.9727782225780625,
+      "grad_norm": 0.933995861984615,
+      "learning_rate": 3.8841804684972564e-08,
+      "loss": 1.6147,
+      "step": 7290
+    },
+    {
+      "epoch": 0.9729116626634641,
+      "grad_norm": 0.9369601765686556,
+      "learning_rate": 3.846217971457677e-08,
+      "loss": 1.4899,
+      "step": 7291
+    },
+    {
+      "epoch": 0.9730451027488658,
+      "grad_norm": 0.9638712887226217,
+      "learning_rate": 3.808441544602914e-08,
+      "loss": 1.572,
+      "step": 7292
+    },
+    {
+      "epoch": 0.9731785428342674,
+      "grad_norm": 0.9358530817042976,
+      "learning_rate": 3.7708511949891e-08,
+      "loss": 1.5785,
+      "step": 7293
+    },
+    {
+      "epoch": 0.973311982919669,
+      "grad_norm": 0.956879412152412,
+      "learning_rate": 3.7334469296378406e-08,
+      "loss": 1.5398,
+      "step": 7294
+    },
+    {
+      "epoch": 0.9734454230050708,
+      "grad_norm": 0.9452032315669289,
+      "learning_rate": 3.69622875553588e-08,
+      "loss": 1.563,
+      "step": 7295
+    },
+    {
+      "epoch": 0.9735788630904724,
+      "grad_norm": 0.9496476555370976,
+      "learning_rate": 3.659196679634991e-08,
+      "loss": 1.5428,
+      "step": 7296
+    },
+    {
+      "epoch": 0.973712303175874,
+      "grad_norm": 0.9327306321435094,
+      "learning_rate": 3.622350708852307e-08,
+      "loss": 1.5386,
+      "step": 7297
+    },
+    {
+      "epoch": 0.9738457432612757,
+      "grad_norm": 0.9712502348220395,
+      "learning_rate": 3.5856908500704335e-08,
+      "loss": 1.5161,
+      "step": 7298
+    },
+    {
+      "epoch": 0.9739791833466773,
+      "grad_norm": 0.9257919013138982,
+      "learning_rate": 3.5492171101368935e-08,
+      "loss": 1.5558,
+      "step": 7299
+    },
+    {
+      "epoch": 0.974112623432079,
+      "grad_norm": 0.9385635102986536,
+      "learning_rate": 3.512929495864459e-08,
+      "loss": 1.5224,
+      "step": 7300
+    },
+    {
+      "epoch": 0.9742460635174807,
+      "grad_norm": 0.9404378377963817,
+      "learning_rate": 3.476828014031486e-08,
+      "loss": 1.5744,
+      "step": 7301
+    },
+    {
+      "epoch": 0.9743795036028823,
+      "grad_norm": 0.9341637296301126,
+      "learning_rate": 3.440912671381136e-08,
+      "loss": 1.557,
+      "step": 7302
+    },
+    {
+      "epoch": 0.974512943688284,
+      "grad_norm": 0.9208183908614412,
+      "learning_rate": 3.4051834746221534e-08,
+      "loss": 1.5211,
+      "step": 7303
+    },
+    {
+      "epoch": 0.9746463837736856,
+      "grad_norm": 0.9118064472195799,
+      "learning_rate": 3.369640430428089e-08,
+      "loss": 1.5383,
+      "step": 7304
+    },
+    {
+      "epoch": 0.9747798238590872,
+      "grad_norm": 0.9329162779176879,
+      "learning_rate": 3.3342835454382996e-08,
+      "loss": 1.5513,
+      "step": 7305
+    },
+    {
+      "epoch": 0.974913263944489,
+      "grad_norm": 0.951014370362824,
+      "learning_rate": 3.299112826256723e-08,
+      "loss": 1.5102,
+      "step": 7306
+    },
+    {
+      "epoch": 0.9750467040298906,
+      "grad_norm": 1.047900440973857,
+      "learning_rate": 3.264128279453105e-08,
+      "loss": 1.554,
+      "step": 7307
+    },
+    {
+      "epoch": 0.9751801441152922,
+      "grad_norm": 0.9197324429445751,
+      "learning_rate": 3.229329911561996e-08,
+      "loss": 1.5731,
+      "step": 7308
+    },
+    {
+      "epoch": 0.9753135842006939,
+      "grad_norm": 0.9333452848398622,
+      "learning_rate": 3.1947177290834184e-08,
+      "loss": 1.588,
+      "step": 7309
+    },
+    {
+      "epoch": 0.9754470242860955,
+      "grad_norm": 0.9777880188410225,
+      "learning_rate": 3.160291738482535e-08,
+      "loss": 1.5162,
+      "step": 7310
+    },
+    {
+      "epoch": 0.9755804643714971,
+      "grad_norm": 0.9366198386385366,
+      "learning_rate": 3.1260519461896456e-08,
+      "loss": 1.5737,
+      "step": 7311
+    },
+    {
+      "epoch": 0.9757139044568989,
+      "grad_norm": 0.9239853504065914,
+      "learning_rate": 3.091998358600523e-08,
+      "loss": 1.554,
+      "step": 7312
+    },
+    {
+      "epoch": 0.9758473445423005,
+      "grad_norm": 0.9193926283503289,
+      "learning_rate": 3.0581309820757466e-08,
+      "loss": 1.4998,
+      "step": 7313
+    },
+    {
+      "epoch": 0.9759807846277022,
+      "grad_norm": 0.9795340493304713,
+      "learning_rate": 3.0244498229415885e-08,
+      "loss": 1.4955,
+      "step": 7314
+    },
+    {
+      "epoch": 0.9761142247131038,
+      "grad_norm": 0.9333034891767017,
+      "learning_rate": 2.990954887489239e-08,
+      "loss": 1.4969,
+      "step": 7315
+    },
+    {
+      "epoch": 0.9762476647985054,
+      "grad_norm": 0.9282493967471739,
+      "learning_rate": 2.957646181975027e-08,
+      "loss": 1.5149,
+      "step": 7316
+    },
+    {
+      "epoch": 0.9763811048839072,
+      "grad_norm": 0.9838269293925297,
+      "learning_rate": 2.9245237126208638e-08,
+      "loss": 1.5284,
+      "step": 7317
+    },
+    {
+      "epoch": 0.9765145449693088,
+      "grad_norm": 0.9463715264539991,
+      "learning_rate": 2.8915874856134672e-08,
+      "loss": 1.5611,
+      "step": 7318
+    },
+    {
+      "epoch": 0.9766479850547104,
+      "grad_norm": 0.9355403594679113,
+      "learning_rate": 2.858837507105028e-08,
+      "loss": 1.5054,
+      "step": 7319
+    },
+    {
+      "epoch": 0.9767814251401121,
+      "grad_norm": 1.2260606411830046,
+      "learning_rate": 2.826273783212874e-08,
+      "loss": 1.5631,
+      "step": 7320
+    },
+    {
+      "epoch": 0.9769148652255137,
+      "grad_norm": 0.9215276937501206,
+      "learning_rate": 2.7938963200195844e-08,
+      "loss": 1.5625,
+      "step": 7321
+    },
+    {
+      "epoch": 0.9770483053109154,
+      "grad_norm": 1.0251253524014017,
+      "learning_rate": 2.7617051235727666e-08,
+      "loss": 1.5516,
+      "step": 7322
+    },
+    {
+      "epoch": 0.9771817453963171,
+      "grad_norm": 0.9213818791054645,
+      "learning_rate": 2.7297001998854987e-08,
+      "loss": 1.5389,
+      "step": 7323
+    },
+    {
+      "epoch": 0.9773151854817187,
+      "grad_norm": 0.941181131336436,
+      "learning_rate": 2.697881554935888e-08,
+      "loss": 1.5423,
+      "step": 7324
+    },
+    {
+      "epoch": 0.9774486255671203,
+      "grad_norm": 0.8999309030256885,
+      "learning_rate": 2.666249194667292e-08,
+      "loss": 1.5542,
+      "step": 7325
+    },
+    {
+      "epoch": 0.977582065652522,
+      "grad_norm": 0.9331555186069251,
+      "learning_rate": 2.6348031249882057e-08,
+      "loss": 1.5703,
+      "step": 7326
+    },
+    {
+      "epoch": 0.9777155057379237,
+      "grad_norm": 1.0508215046605072,
+      "learning_rate": 2.603543351772486e-08,
+      "loss": 1.5462,
+      "step": 7327
+    },
+    {
+      "epoch": 0.9778489458233254,
+      "grad_norm": 0.9500724069103249,
+      "learning_rate": 2.5724698808591297e-08,
+      "loss": 1.5018,
+      "step": 7328
+    },
+    {
+      "epoch": 0.977982385908727,
+      "grad_norm": 0.950196772138831,
+      "learning_rate": 2.541582718052271e-08,
+      "loss": 1.4982,
+      "step": 7329
+    },
+    {
+      "epoch": 0.9781158259941286,
+      "grad_norm": 1.022399182945929,
+      "learning_rate": 2.5108818691212956e-08,
+      "loss": 1.5155,
+      "step": 7330
+    },
+    {
+      "epoch": 0.9782492660795303,
+      "grad_norm": 1.092331980943047,
+      "learning_rate": 2.4803673398006157e-08,
+      "loss": 1.5725,
+      "step": 7331
+    },
+    {
+      "epoch": 0.978382706164932,
+      "grad_norm": 0.9255819560432912,
+      "learning_rate": 2.4500391357902277e-08,
+      "loss": 1.4885,
+      "step": 7332
+    },
+    {
+      "epoch": 0.9785161462503336,
+      "grad_norm": 0.9513513483680666,
+      "learning_rate": 2.4198972627549335e-08,
+      "loss": 1.5104,
+      "step": 7333
+    },
+    {
+      "epoch": 0.9786495863357353,
+      "grad_norm": 0.8903283569870852,
+      "learning_rate": 2.389941726325007e-08,
+      "loss": 1.5539,
+      "step": 7334
+    },
+    {
+      "epoch": 0.9787830264211369,
+      "grad_norm": 0.9338400501650106,
+      "learning_rate": 2.3601725320957503e-08,
+      "loss": 1.4993,
+      "step": 7335
+    },
+    {
+      "epoch": 0.9789164665065385,
+      "grad_norm": 0.971758384623749,
+      "learning_rate": 2.3305896856277154e-08,
+      "loss": 1.5658,
+      "step": 7336
+    },
+    {
+      "epoch": 0.9790499065919402,
+      "grad_norm": 0.9672566853309181,
+      "learning_rate": 2.3011931924465936e-08,
+      "loss": 1.5645,
+      "step": 7337
+    },
+    {
+      "epoch": 0.9791833466773419,
+      "grad_norm": 0.9218261068570299,
+      "learning_rate": 2.2719830580434366e-08,
+      "loss": 1.5638,
+      "step": 7338
+    },
+    {
+      "epoch": 0.9793167867627435,
+      "grad_norm": 0.9367852380054548,
+      "learning_rate": 2.2429592878742134e-08,
+      "loss": 1.547,
+      "step": 7339
+    },
+    {
+      "epoch": 0.9794502268481452,
+      "grad_norm": 0.9339712727726341,
+      "learning_rate": 2.2141218873602544e-08,
+      "loss": 1.514,
+      "step": 7340
+    },
+    {
+      "epoch": 0.9795836669335468,
+      "grad_norm": 0.9417814446241446,
+      "learning_rate": 2.1854708618882504e-08,
+      "loss": 1.5222,
+      "step": 7341
+    },
+    {
+      "epoch": 0.9797171070189485,
+      "grad_norm": 0.9339447722012021,
+      "learning_rate": 2.1570062168095873e-08,
+      "loss": 1.5335,
+      "step": 7342
+    },
+    {
+      "epoch": 0.9798505471043502,
+      "grad_norm": 1.0402602333871376,
+      "learning_rate": 2.1287279574414566e-08,
+      "loss": 1.534,
+      "step": 7343
+    },
+    {
+      "epoch": 0.9799839871897518,
+      "grad_norm": 0.9326106042217657,
+      "learning_rate": 2.100636089065633e-08,
+      "loss": 1.4977,
+      "step": 7344
+    },
+    {
+      "epoch": 0.9801174272751535,
+      "grad_norm": 0.930957256971068,
+      "learning_rate": 2.0727306169294747e-08,
+      "loss": 1.5236,
+      "step": 7345
+    },
+    {
+      "epoch": 0.9802508673605551,
+      "grad_norm": 0.9473415741877985,
+      "learning_rate": 2.0450115462454788e-08,
+      "loss": 1.5171,
+      "step": 7346
+    },
+    {
+      "epoch": 0.9803843074459567,
+      "grad_norm": 0.9215591618457778,
+      "learning_rate": 2.0174788821911706e-08,
+      "loss": 1.4988,
+      "step": 7347
+    },
+    {
+      "epoch": 0.9805177475313585,
+      "grad_norm": 0.9501836671635087,
+      "learning_rate": 1.990132629909325e-08,
+      "loss": 1.5363,
+      "step": 7348
+    },
+    {
+      "epoch": 0.9806511876167601,
+      "grad_norm": 1.0095366173547495,
+      "learning_rate": 1.9629727945079668e-08,
+      "loss": 1.5768,
+      "step": 7349
+    },
+    {
+      "epoch": 0.9807846277021617,
+      "grad_norm": 0.9307006869728814,
+      "learning_rate": 1.9359993810601495e-08,
+      "loss": 1.5595,
+      "step": 7350
+    },
+    {
+      "epoch": 0.9809180677875634,
+      "grad_norm": 0.9742637399272271,
+      "learning_rate": 1.9092123946042873e-08,
+      "loss": 1.5153,
+      "step": 7351
+    },
+    {
+      "epoch": 0.981051507872965,
+      "grad_norm": 0.9314930322710189,
+      "learning_rate": 1.882611840143822e-08,
+      "loss": 1.5104,
+      "step": 7352
+    },
+    {
+      "epoch": 0.9811849479583667,
+      "grad_norm": 0.9381519532395032,
+      "learning_rate": 1.856197722647557e-08,
+      "loss": 1.5689,
+      "step": 7353
+    },
+    {
+      "epoch": 0.9813183880437684,
+      "grad_norm": 0.9039239250179172,
+      "learning_rate": 1.8299700470492122e-08,
+      "loss": 1.5105,
+      "step": 7354
+    },
+    {
+      "epoch": 0.98145182812917,
+      "grad_norm": 0.9396159105126344,
+      "learning_rate": 1.8039288182478686e-08,
+      "loss": 1.5139,
+      "step": 7355
+    },
+    {
+      "epoch": 0.9815852682145717,
+      "grad_norm": 0.9951647404292933,
+      "learning_rate": 1.778074041107747e-08,
+      "loss": 1.5556,
+      "step": 7356
+    },
+    {
+      "epoch": 0.9817187082999733,
+      "grad_norm": 0.970298389786309,
+      "learning_rate": 1.7524057204582058e-08,
+      "loss": 1.5558,
+      "step": 7357
+    },
+    {
+      "epoch": 0.9818521483853749,
+      "grad_norm": 0.9653437553551486,
+      "learning_rate": 1.7269238610938544e-08,
+      "loss": 1.5561,
+      "step": 7358
+    },
+    {
+      "epoch": 0.9819855884707767,
+      "grad_norm": 1.006023062423882,
+      "learning_rate": 1.701628467774219e-08,
+      "loss": 1.5283,
+      "step": 7359
+    },
+    {
+      "epoch": 0.9821190285561783,
+      "grad_norm": 0.9396565250832867,
+      "learning_rate": 1.6765195452245198e-08,
+      "loss": 1.5434,
+      "step": 7360
+    },
+    {
+      "epoch": 0.9822524686415799,
+      "grad_norm": 0.9608296788404236,
+      "learning_rate": 1.6515970981344498e-08,
+      "loss": 1.4794,
+      "step": 7361
+    },
+    {
+      "epoch": 0.9823859087269816,
+      "grad_norm": 1.0298643297103718,
+      "learning_rate": 1.6268611311595072e-08,
+      "loss": 1.5356,
+      "step": 7362
+    },
+    {
+      "epoch": 0.9825193488123832,
+      "grad_norm": 0.9122880384105183,
+      "learning_rate": 1.6023116489199962e-08,
+      "loss": 1.4973,
+      "step": 7363
+    },
+    {
+      "epoch": 0.9826527888977848,
+      "grad_norm": 0.9280142474276555,
+      "learning_rate": 1.5779486560014713e-08,
+      "loss": 1.5525,
+      "step": 7364
+    },
+    {
+      "epoch": 0.9827862289831866,
+      "grad_norm": 0.9191238933559271,
+      "learning_rate": 1.553772156954736e-08,
+      "loss": 1.4899,
+      "step": 7365
+    },
+    {
+      "epoch": 0.9829196690685882,
+      "grad_norm": 0.9177971381786734,
+      "learning_rate": 1.529782156295512e-08,
+      "loss": 1.5066,
+      "step": 7366
+    },
+    {
+      "epoch": 0.9830531091539899,
+      "grad_norm": 0.9212731057030706,
+      "learning_rate": 1.505978658505103e-08,
+      "loss": 1.502,
+      "step": 7367
+    },
+    {
+      "epoch": 0.9831865492393915,
+      "grad_norm": 0.9287034571182928,
+      "learning_rate": 1.4823616680295083e-08,
+      "loss": 1.5333,
+      "step": 7368
+    },
+    {
+      "epoch": 0.9833199893247931,
+      "grad_norm": 0.9141592195619862,
+      "learning_rate": 1.4589311892801994e-08,
+      "loss": 1.5547,
+      "step": 7369
+    },
+    {
+      "epoch": 0.9834534294101949,
+      "grad_norm": 0.9139700337452249,
+      "learning_rate": 1.4356872266337862e-08,
+      "loss": 1.556,
+      "step": 7370
+    },
+    {
+      "epoch": 0.9835868694955965,
+      "grad_norm": 0.9454395337463968,
+      "learning_rate": 1.4126297844317959e-08,
+      "loss": 1.5366,
+      "step": 7371
+    },
+    {
+      "epoch": 0.9837203095809981,
+      "grad_norm": 0.9532365221783823,
+      "learning_rate": 1.389758866981339e-08,
+      "loss": 1.5516,
+      "step": 7372
+    },
+    {
+      "epoch": 0.9838537496663998,
+      "grad_norm": 0.9276600231944385,
+      "learning_rate": 1.367074478554331e-08,
+      "loss": 1.549,
+      "step": 7373
+    },
+    {
+      "epoch": 0.9839871897518014,
+      "grad_norm": 0.9614937926329741,
+      "learning_rate": 1.3445766233878277e-08,
+      "loss": 1.5544,
+      "step": 7374
+    },
+    {
+      "epoch": 0.984120629837203,
+      "grad_norm": 0.8899635478946575,
+      "learning_rate": 1.322265305684356e-08,
+      "loss": 1.4937,
+      "step": 7375
+    },
+    {
+      "epoch": 0.9842540699226048,
+      "grad_norm": 0.9288487499950765,
+      "learning_rate": 1.3001405296113601e-08,
+      "loss": 1.5814,
+      "step": 7376
+    },
+    {
+      "epoch": 0.9843875100080064,
+      "grad_norm": 1.1647701919689324,
+      "learning_rate": 1.2782022993015342e-08,
+      "loss": 1.5296,
+      "step": 7377
+    },
+    {
+      "epoch": 0.984520950093408,
+      "grad_norm": 0.9358750657202466,
+      "learning_rate": 1.2564506188526004e-08,
+      "loss": 1.5041,
+      "step": 7378
+    },
+    {
+      "epoch": 0.9846543901788097,
+      "grad_norm": 0.9197611351339202,
+      "learning_rate": 1.2348854923275311e-08,
+      "loss": 1.5391,
+      "step": 7379
+    },
+    {
+      "epoch": 0.9847878302642114,
+      "grad_norm": 0.9827592983158957,
+      "learning_rate": 1.2135069237545483e-08,
+      "loss": 1.535,
+      "step": 7380
+    },
+    {
+      "epoch": 0.9849212703496131,
+      "grad_norm": 0.9615691598090853,
+      "learning_rate": 1.1923149171267911e-08,
+      "loss": 1.6116,
+      "step": 7381
+    },
+    {
+      "epoch": 0.9850547104350147,
+      "grad_norm": 0.9273458595588824,
+      "learning_rate": 1.171309476402871e-08,
+      "loss": 1.5576,
+      "step": 7382
+    },
+    {
+      "epoch": 0.9851881505204163,
+      "grad_norm": 0.9460589830522615,
+      "learning_rate": 1.1504906055060938e-08,
+      "loss": 1.5578,
+      "step": 7383
+    },
+    {
+      "epoch": 0.985321590605818,
+      "grad_norm": 0.9954193989279615,
+      "learning_rate": 1.1298583083254599e-08,
+      "loss": 1.5602,
+      "step": 7384
+    },
+    {
+      "epoch": 0.9854550306912196,
+      "grad_norm": 0.9445437464464703,
+      "learning_rate": 1.1094125887146644e-08,
+      "loss": 1.5294,
+      "step": 7385
+    },
+    {
+      "epoch": 0.9855884707766213,
+      "grad_norm": 0.9104360078993541,
+      "learning_rate": 1.0891534504928747e-08,
+      "loss": 1.5428,
+      "step": 7386
+    },
+    {
+      "epoch": 0.985721910862023,
+      "grad_norm": 1.0236669925899005,
+      "learning_rate": 1.0690808974441747e-08,
+      "loss": 1.6108,
+      "step": 7387
+    },
+    {
+      "epoch": 0.9858553509474246,
+      "grad_norm": 0.9071583750469212,
+      "learning_rate": 1.0491949333178986e-08,
+      "loss": 1.5168,
+      "step": 7388
+    },
+    {
+      "epoch": 0.9859887910328262,
+      "grad_norm": 0.9927493294549666,
+      "learning_rate": 1.0294955618285196e-08,
+      "loss": 1.5921,
+      "step": 7389
+    },
+    {
+      "epoch": 0.9861222311182279,
+      "grad_norm": 0.9363046349406777,
+      "learning_rate": 1.00998278665565e-08,
+      "loss": 1.5694,
+      "step": 7390
+    },
+    {
+      "epoch": 0.9862556712036296,
+      "grad_norm": 0.9135114422937861,
+      "learning_rate": 9.906566114440408e-09,
+      "loss": 1.5794,
+      "step": 7391
+    },
+    {
+      "epoch": 0.9863891112890312,
+      "grad_norm": 0.9437284819277648,
+      "learning_rate": 9.715170398036933e-09,
+      "loss": 1.584,
+      "step": 7392
+    },
+    {
+      "epoch": 0.9865225513744329,
+      "grad_norm": 0.9211484661856915,
+      "learning_rate": 9.525640753095256e-09,
+      "loss": 1.5472,
+      "step": 7393
+    },
+    {
+      "epoch": 0.9866559914598345,
+      "grad_norm": 0.9176256061269126,
+      "learning_rate": 9.337977215018169e-09,
+      "loss": 1.4833,
+      "step": 7394
+    },
+    {
+      "epoch": 0.9867894315452362,
+      "grad_norm": 0.9564873181171374,
+      "learning_rate": 9.152179818859853e-09,
+      "loss": 1.5476,
+      "step": 7395
+    },
+    {
+      "epoch": 0.9869228716306379,
+      "grad_norm": 0.89959690847276,
+      "learning_rate": 8.968248599323659e-09,
+      "loss": 1.525,
+      "step": 7396
+    },
+    {
+      "epoch": 0.9870563117160395,
+      "grad_norm": 0.919080132940536,
+      "learning_rate": 8.78618359076544e-09,
+      "loss": 1.5491,
+      "step": 7397
+    },
+    {
+      "epoch": 0.9871897518014412,
+      "grad_norm": 1.4794996154790685,
+      "learning_rate": 8.605984827195767e-09,
+      "loss": 1.5364,
+      "step": 7398
+    },
+    {
+      "epoch": 0.9873231918868428,
+      "grad_norm": 1.0985804202382334,
+      "learning_rate": 8.427652342271053e-09,
+      "loss": 1.529,
+      "step": 7399
+    },
+    {
+      "epoch": 0.9874566319722444,
+      "grad_norm": 1.1736650971200993,
+      "learning_rate": 8.251186169301318e-09,
+      "loss": 1.5596,
+      "step": 7400
+    },
+    {
+      "epoch": 0.9875900720576462,
+      "grad_norm": 1.0303346794805686,
+      "learning_rate": 8.076586341251303e-09,
+      "loss": 1.4777,
+      "step": 7401
+    },
+    {
+      "epoch": 0.9877235121430478,
+      "grad_norm": 0.9872701203867904,
+      "learning_rate": 7.903852890732699e-09,
+      "loss": 1.5144,
+      "step": 7402
+    },
+    {
+      "epoch": 0.9878569522284494,
+      "grad_norm": 0.9850487540902245,
+      "learning_rate": 7.73298585000859e-09,
+      "loss": 1.5604,
+      "step": 7403
+    },
+    {
+      "epoch": 0.9879903923138511,
+      "grad_norm": 0.9471651528476187,
+      "learning_rate": 7.563985250997884e-09,
+      "loss": 1.5555,
+      "step": 7404
+    },
+    {
+      "epoch": 0.9881238323992527,
+      "grad_norm": 0.9420598022378026,
+      "learning_rate": 7.3968511252664466e-09,
+      "loss": 1.517,
+      "step": 7405
+    },
+    {
+      "epoch": 0.9882572724846543,
+      "grad_norm": 0.9419967634572181,
+      "learning_rate": 7.231583504032636e-09,
+      "loss": 1.5475,
+      "step": 7406
+    },
+    {
+      "epoch": 0.9883907125700561,
+      "grad_norm": 1.0474390058108154,
+      "learning_rate": 7.0681824181673134e-09,
+      "loss": 1.5696,
+      "step": 7407
+    },
+    {
+      "epoch": 0.9885241526554577,
+      "grad_norm": 1.036920220314024,
+      "learning_rate": 6.906647898191621e-09,
+      "loss": 1.554,
+      "step": 7408
+    },
+    {
+      "epoch": 0.9886575927408594,
+      "grad_norm": 0.928746553820107,
+      "learning_rate": 6.7469799742780895e-09,
+      "loss": 1.5312,
+      "step": 7409
+    },
+    {
+      "epoch": 0.988791032826261,
+      "grad_norm": 0.9153121582186817,
+      "learning_rate": 6.589178676251751e-09,
+      "loss": 1.528,
+      "step": 7410
+    },
+    {
+      "epoch": 0.9889244729116626,
+      "grad_norm": 1.0836635609150753,
+      "learning_rate": 6.433244033587916e-09,
+      "loss": 1.5573,
+      "step": 7411
+    },
+    {
+      "epoch": 0.9890579129970644,
+      "grad_norm": 1.221340256835362,
+      "learning_rate": 6.279176075412175e-09,
+      "loss": 1.5926,
+      "step": 7412
+    },
+    {
+      "epoch": 0.989191353082466,
+      "grad_norm": 0.9091108579676047,
+      "learning_rate": 6.12697483050484e-09,
+      "loss": 1.5529,
+      "step": 7413
+    },
+    {
+      "epoch": 0.9893247931678676,
+      "grad_norm": 0.92677410973629,
+      "learning_rate": 5.976640327293171e-09,
+      "loss": 1.541,
+      "step": 7414
+    },
+    {
+      "epoch": 0.9894582332532693,
+      "grad_norm": 0.9383472475782342,
+      "learning_rate": 5.828172593858039e-09,
+      "loss": 1.4556,
+      "step": 7415
+    },
+    {
+      "epoch": 0.9895916733386709,
+      "grad_norm": 0.9238065180005837,
+      "learning_rate": 5.681571657933927e-09,
+      "loss": 1.5083,
+      "step": 7416
+    },
+    {
+      "epoch": 0.9897251134240725,
+      "grad_norm": 0.9377466504722105,
+      "learning_rate": 5.536837546902263e-09,
+      "loss": 1.5493,
+      "step": 7417
+    },
+    {
+      "epoch": 0.9898585535094743,
+      "grad_norm": 0.9335726191972902,
+      "learning_rate": 5.393970287796979e-09,
+      "loss": 1.571,
+      "step": 7418
+    },
+    {
+      "epoch": 0.9899919935948759,
+      "grad_norm": 1.1669560380448265,
+      "learning_rate": 5.2529699073067265e-09,
+      "loss": 1.504,
+      "step": 7419
+    },
+    {
+      "epoch": 0.9901254336802775,
+      "grad_norm": 0.9537386546748483,
+      "learning_rate": 5.113836431765995e-09,
+      "loss": 1.5505,
+      "step": 7420
+    },
+    {
+      "epoch": 0.9902588737656792,
+      "grad_norm": 0.9165621627626547,
+      "learning_rate": 4.976569887165106e-09,
+      "loss": 1.5176,
+      "step": 7421
+    },
+    {
+      "epoch": 0.9903923138510808,
+      "grad_norm": 0.9389326153560223,
+      "learning_rate": 4.8411702991435494e-09,
+      "loss": 1.5421,
+      "step": 7422
+    },
+    {
+      "epoch": 0.9905257539364826,
+      "grad_norm": 0.9939334972185406,
+      "learning_rate": 4.707637692992206e-09,
+      "loss": 1.5195,
+      "step": 7423
+    },
+    {
+      "epoch": 0.9906591940218842,
+      "grad_norm": 1.0278495243989596,
+      "learning_rate": 4.575972093653347e-09,
+      "loss": 1.5995,
+      "step": 7424
+    },
+    {
+      "epoch": 0.9907926341072858,
+      "grad_norm": 0.9567686246995181,
+      "learning_rate": 4.4461735257206315e-09,
+      "loss": 1.459,
+      "step": 7425
+    },
+    {
+      "epoch": 0.9909260741926875,
+      "grad_norm": 1.136055181982407,
+      "learning_rate": 4.318242013439111e-09,
+      "loss": 1.4569,
+      "step": 7426
+    },
+    {
+      "epoch": 0.9910595142780891,
+      "grad_norm": 0.9458609375171427,
+      "learning_rate": 4.1921775807041154e-09,
+      "loss": 1.5101,
+      "step": 7427
+    },
+    {
+      "epoch": 0.9911929543634908,
+      "grad_norm": 0.9533736019573524,
+      "learning_rate": 4.067980251064585e-09,
+      "loss": 1.5205,
+      "step": 7428
+    },
+    {
+      "epoch": 0.9913263944488925,
+      "grad_norm": 1.0479939592780765,
+      "learning_rate": 3.94565004771863e-09,
+      "loss": 1.6043,
+      "step": 7429
+    },
+    {
+      "epoch": 0.9914598345342941,
+      "grad_norm": 0.9972412023234489,
+      "learning_rate": 3.825186993515751e-09,
+      "loss": 1.5548,
+      "step": 7430
+    },
+    {
+      "epoch": 0.9915932746196957,
+      "grad_norm": 0.9272943886701682,
+      "learning_rate": 3.7065911109568365e-09,
+      "loss": 1.5379,
+      "step": 7431
+    },
+    {
+      "epoch": 0.9917267147050974,
+      "grad_norm": 0.9473938913490474,
+      "learning_rate": 3.589862422195278e-09,
+      "loss": 1.5721,
+      "step": 7432
+    },
+    {
+      "epoch": 0.991860154790499,
+      "grad_norm": 0.9185317252383915,
+      "learning_rate": 3.4750009490336355e-09,
+      "loss": 1.5664,
+      "step": 7433
+    },
+    {
+      "epoch": 0.9919935948759008,
+      "grad_norm": 0.9791716481267135,
+      "learning_rate": 3.3620067129269683e-09,
+      "loss": 1.5587,
+      "step": 7434
+    },
+    {
+      "epoch": 0.9921270349613024,
+      "grad_norm": 0.9173521049585661,
+      "learning_rate": 3.2508797349817268e-09,
+      "loss": 1.5304,
+      "step": 7435
+    },
+    {
+      "epoch": 0.992260475046704,
+      "grad_norm": 0.9632980233243099,
+      "learning_rate": 3.14162003595353e-09,
+      "loss": 1.5745,
+      "step": 7436
+    },
+    {
+      "epoch": 0.9923939151321057,
+      "grad_norm": 0.9581161698339242,
+      "learning_rate": 3.034227636253828e-09,
+      "loss": 1.5666,
+      "step": 7437
+    },
+    {
+      "epoch": 0.9925273552175073,
+      "grad_norm": 0.9413492862074422,
+      "learning_rate": 2.9287025559399108e-09,
+      "loss": 1.5062,
+      "step": 7438
+    },
+    {
+      "epoch": 0.992660795302909,
+      "grad_norm": 0.9747881941796354,
+      "learning_rate": 2.8250448147237875e-09,
+      "loss": 1.4891,
+      "step": 7439
+    },
+    {
+      "epoch": 0.9927942353883107,
+      "grad_norm": 0.9824468800744949,
+      "learning_rate": 2.723254431967748e-09,
+      "loss": 1.5657,
+      "step": 7440
+    },
+    {
+      "epoch": 0.9929276754737123,
+      "grad_norm": 0.9476470497339562,
+      "learning_rate": 2.623331426683251e-09,
+      "loss": 1.5411,
+      "step": 7441
+    },
+    {
+      "epoch": 0.9930611155591139,
+      "grad_norm": 0.9089786350328128,
+      "learning_rate": 2.525275817536477e-09,
+      "loss": 1.573,
+      "step": 7442
+    },
+    {
+      "epoch": 0.9931945556445156,
+      "grad_norm": 0.9170076888295607,
+      "learning_rate": 2.429087622842774e-09,
+      "loss": 1.5052,
+      "step": 7443
+    },
+    {
+      "epoch": 0.9933279957299173,
+      "grad_norm": 0.9562919014528187,
+      "learning_rate": 2.334766860568882e-09,
+      "loss": 1.5247,
+      "step": 7444
+    },
+    {
+      "epoch": 0.9934614358153189,
+      "grad_norm": 0.9327221876692189,
+      "learning_rate": 2.2423135483329306e-09,
+      "loss": 1.5501,
+      "step": 7445
+    },
+    {
+      "epoch": 0.9935948759007206,
+      "grad_norm": 0.9382241408409218,
+      "learning_rate": 2.151727703404438e-09,
+      "loss": 1.5226,
+      "step": 7446
+    },
+    {
+      "epoch": 0.9937283159861222,
+      "grad_norm": 0.9581003164477122,
+      "learning_rate": 2.0630093427032038e-09,
+      "loss": 1.5585,
+      "step": 7447
+    },
+    {
+      "epoch": 0.9938617560715239,
+      "grad_norm": 1.049522255731219,
+      "learning_rate": 1.9761584828004164e-09,
+      "loss": 1.5329,
+      "step": 7448
+    },
+    {
+      "epoch": 0.9939951961569256,
+      "grad_norm": 0.9641339166472345,
+      "learning_rate": 1.8911751399197655e-09,
+      "loss": 1.5166,
+      "step": 7449
+    },
+    {
+      "epoch": 0.9941286362423272,
+      "grad_norm": 1.1001878062084773,
+      "learning_rate": 1.808059329935219e-09,
+      "loss": 1.5105,
+      "step": 7450
+    },
+    {
+      "epoch": 0.9942620763277289,
+      "grad_norm": 0.9672405162002357,
+      "learning_rate": 1.726811068371026e-09,
+      "loss": 1.5858,
+      "step": 7451
+    },
+    {
+      "epoch": 0.9943955164131305,
+      "grad_norm": 0.9497524896899447,
+      "learning_rate": 1.6474303704039352e-09,
+      "loss": 1.534,
+      "step": 7452
+    },
+    {
+      "epoch": 0.9945289564985321,
+      "grad_norm": 0.9337846143023557,
+      "learning_rate": 1.5699172508620852e-09,
+      "loss": 1.4735,
+      "step": 7453
+    },
+    {
+      "epoch": 0.9946623965839338,
+      "grad_norm": 1.083632648142817,
+      "learning_rate": 1.4942717242227844e-09,
+      "loss": 1.5346,
+      "step": 7454
+    },
+    {
+      "epoch": 0.9947958366693355,
+      "grad_norm": 0.9438573488466988,
+      "learning_rate": 1.4204938046158412e-09,
+      "loss": 1.5054,
+      "step": 7455
+    },
+    {
+      "epoch": 0.9949292767547371,
+      "grad_norm": 0.9556477000553659,
+      "learning_rate": 1.3485835058224538e-09,
+      "loss": 1.5491,
+      "step": 7456
+    },
+    {
+      "epoch": 0.9950627168401388,
+      "grad_norm": 0.9113180378493281,
+      "learning_rate": 1.278540841275211e-09,
+      "loss": 1.4985,
+      "step": 7457
+    },
+    {
+      "epoch": 0.9951961569255404,
+      "grad_norm": 0.9329564036235153,
+      "learning_rate": 1.2103658240569805e-09,
+      "loss": 1.5467,
+      "step": 7458
+    },
+    {
+      "epoch": 0.995329597010942,
+      "grad_norm": 0.931682502757043,
+      "learning_rate": 1.1440584669020205e-09,
+      "loss": 1.5791,
+      "step": 7459
+    },
+    {
+      "epoch": 0.9954630370963438,
+      "grad_norm": 1.1265746496215818,
+      "learning_rate": 1.0796187821959792e-09,
+      "loss": 1.589,
+      "step": 7460
+    },
+    {
+      "epoch": 0.9955964771817454,
+      "grad_norm": 1.139830131670322,
+      "learning_rate": 1.017046781973674e-09,
+      "loss": 1.5046,
+      "step": 7461
+    },
+    {
+      "epoch": 0.9957299172671471,
+      "grad_norm": 0.9401769463698738,
+      "learning_rate": 9.563424779257535e-10,
+      "loss": 1.4786,
+      "step": 7462
+    },
+    {
+      "epoch": 0.9958633573525487,
+      "grad_norm": 1.0561437037692578,
+      "learning_rate": 8.97505881388705e-10,
+      "loss": 1.5136,
+      "step": 7463
+    },
+    {
+      "epoch": 0.9959967974379503,
+      "grad_norm": 1.0139954761378014,
+      "learning_rate": 8.405370033548466e-10,
+      "loss": 1.5199,
+      "step": 7464
+    },
+    {
+      "epoch": 0.9961302375233521,
+      "grad_norm": 0.9447397803000418,
+      "learning_rate": 7.854358544623353e-10,
+      "loss": 1.5971,
+      "step": 7465
+    },
+    {
+      "epoch": 0.9962636776087537,
+      "grad_norm": 0.9299070974900873,
+      "learning_rate": 7.322024450062693e-10,
+      "loss": 1.5235,
+      "step": 7466
+    },
+    {
+      "epoch": 0.9963971176941553,
+      "grad_norm": 0.9261836009002768,
+      "learning_rate": 6.808367849286956e-10,
+      "loss": 1.5404,
+      "step": 7467
+    },
+    {
+      "epoch": 0.996530557779557,
+      "grad_norm": 0.9439837053013257,
+      "learning_rate": 6.313388838230517e-10,
+      "loss": 1.5434,
+      "step": 7468
+    },
+    {
+      "epoch": 0.9966639978649586,
+      "grad_norm": 0.9180840699289173,
+      "learning_rate": 5.83708750937495e-10,
+      "loss": 1.4938,
+      "step": 7469
+    },
+    {
+      "epoch": 0.9967974379503602,
+      "grad_norm": 0.9964338171196607,
+      "learning_rate": 5.379463951671326e-10,
+      "loss": 1.5414,
+      "step": 7470
+    },
+    {
+      "epoch": 0.996930878035762,
+      "grad_norm": 0.967091118915389,
+      "learning_rate": 4.940518250606818e-10,
+      "loss": 1.5766,
+      "step": 7471
+    },
+    {
+      "epoch": 0.9970643181211636,
+      "grad_norm": 0.9025190757694261,
+      "learning_rate": 4.5202504881602936e-10,
+      "loss": 1.5351,
+      "step": 7472
+    },
+    {
+      "epoch": 0.9971977582065652,
+      "grad_norm": 0.92342728170414,
+      "learning_rate": 4.118660742846725e-10,
+      "loss": 1.546,
+      "step": 7473
+    },
+    {
+      "epoch": 0.9973311982919669,
+      "grad_norm": 0.9285850741711503,
+      "learning_rate": 3.735749089661678e-10,
+      "loss": 1.5224,
+      "step": 7474
+    },
+    {
+      "epoch": 0.9974646383773685,
+      "grad_norm": 0.9333801469713987,
+      "learning_rate": 3.371515600147923e-10,
+      "loss": 1.5543,
+      "step": 7475
+    },
+    {
+      "epoch": 0.9975980784627703,
+      "grad_norm": 0.9326257599816116,
+      "learning_rate": 3.0259603423288266e-10,
+      "loss": 1.548,
+      "step": 7476
+    },
+    {
+      "epoch": 0.9977315185481719,
+      "grad_norm": 0.9283677855379775,
+      "learning_rate": 2.699083380741652e-10,
+      "loss": 1.562,
+      "step": 7477
+    },
+    {
+      "epoch": 0.9978649586335735,
+      "grad_norm": 0.9887750864167223,
+      "learning_rate": 2.3908847764597674e-10,
+      "loss": 1.5263,
+      "step": 7478
+    },
+    {
+      "epoch": 0.9979983987189752,
+      "grad_norm": 1.0609376235918477,
+      "learning_rate": 2.101364587048238e-10,
+      "loss": 1.5183,
+      "step": 7479
+    },
+    {
+      "epoch": 0.9981318388043768,
+      "grad_norm": 0.9276456917237319,
+      "learning_rate": 1.8305228665860264e-10,
+      "loss": 1.5592,
+      "step": 7480
+    },
+    {
+      "epoch": 0.9982652788897785,
+      "grad_norm": 0.9511667430507186,
+      "learning_rate": 1.578359665654894e-10,
+      "loss": 1.6094,
+      "step": 7481
+    },
+    {
+      "epoch": 0.9983987189751802,
+      "grad_norm": 0.935192544085142,
+      "learning_rate": 1.3448750313616032e-10,
+      "loss": 1.5063,
+      "step": 7482
+    },
+    {
+      "epoch": 0.9985321590605818,
+      "grad_norm": 0.9411754750534707,
+      "learning_rate": 1.130069007315715e-10,
+      "loss": 1.5261,
+      "step": 7483
+    },
+    {
+      "epoch": 0.9986655991459834,
+      "grad_norm": 1.1219517093294005,
+      "learning_rate": 9.339416336517915e-11,
+      "loss": 1.5611,
+      "step": 7484
+    },
+    {
+      "epoch": 0.9987990392313851,
+      "grad_norm": 1.1069618633698497,
+      "learning_rate": 7.56492946984988e-11,
+      "loss": 1.5211,
+      "step": 7485
+    },
+    {
+      "epoch": 0.9989324793167867,
+      "grad_norm": 0.9326395703982616,
+      "learning_rate": 5.977229804776663e-11,
+      "loss": 1.5316,
+      "step": 7486
+    },
+    {
+      "epoch": 0.9990659194021884,
+      "grad_norm": 1.0224397864563275,
+      "learning_rate": 4.576317637838834e-11,
+      "loss": 1.5785,
+      "step": 7487
+    },
+    {
+      "epoch": 0.9991993594875901,
+      "grad_norm": 1.2901485937141546,
+      "learning_rate": 3.3621932306049375e-11,
+      "loss": 1.5255,
+      "step": 7488
+    },
+    {
+      "epoch": 0.9993327995729917,
+      "grad_norm": 0.9764240130637175,
+      "learning_rate": 2.3348568098935376e-11,
+      "loss": 1.5258,
+      "step": 7489
+    },
+    {
+      "epoch": 0.9994662396583934,
+      "grad_norm": 0.9510632886911196,
+      "learning_rate": 1.494308567662195e-11,
+      "loss": 1.5212,
+      "step": 7490
+    },
+    {
+      "epoch": 0.999599679743795,
+      "grad_norm": 0.9937550478187936,
+      "learning_rate": 8.405486608964452e-12,
+      "loss": 1.5589,
+      "step": 7491
+    },
+    {
+      "epoch": 0.9997331198291967,
+      "grad_norm": 0.9234682778153688,
+      "learning_rate": 3.735772117208214e-12,
+      "loss": 1.501,
+      "step": 7492
+    },
+    {
+      "epoch": 0.9998665599145984,
+      "grad_norm": 0.9297623176763948,
+      "learning_rate": 9.339430728783073e-13,
+      "loss": 1.5671,
+      "step": 7493
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.109193681967266,
+      "learning_rate": 0.0,
+      "loss": 1.5531,
+      "step": 7494
+    },
+    {
+      "epoch": 1.0,
+      "step": 7494,
+      "total_flos": 2.902622270823963e+19,
+      "train_loss": 1.6068721134280217,
+      "train_runtime": 34738.8495,
+      "train_samples_per_second": 36.238,
+      "train_steps_per_second": 0.216
+    }
+  ],
+  "logging_steps": 1.0,
+  "max_steps": 7494,
+  "num_input_tokens_seen": 0,
+  "num_train_epochs": 1,
+  "save_steps": 100000,
+  "stateful_callbacks": {
+    "TrainerControl": {
+      "args": {
+        "should_epoch_stop": false,
+        "should_evaluate": false,
+        "should_log": false,
+        "should_save": false,
+        "should_training_stop": false
+      },
+      "attributes": {}
+    }
+  },
+  "total_flos": 2.902622270823963e+19,
+  "train_batch_size": 24,
+  "trial_name": null,
+  "trial_params": null
+}