diff --git "a/trainer_state.json" "b/trainer_state.json"
new file mode 100644--- /dev/null
+++ "b/trainer_state.json"
@@ -0,0 +1,17542 @@
+{
+  "best_metric": null,
+  "best_model_checkpoint": null,
+  "epoch": 1.0,
+  "eval_steps": 5000.0,
+  "global_step": 2500,
+  "is_hyper_param_search": false,
+  "is_local_process_zero": true,
+  "is_world_process_zero": true,
+  "log_history": [
+    {
+      "epoch": 0.0004,
+      "grad_norm": 0.7262070178985596,
+      "learning_rate": 9.333333333333334e-08,
+      "loss": 0.8408,
+      "step": 1
+    },
+    {
+      "epoch": 0.0008,
+      "grad_norm": 0.8701621890068054,
+      "learning_rate": 1.8666666666666667e-07,
+      "loss": 0.974,
+      "step": 2
+    },
+    {
+      "epoch": 0.0012,
+      "grad_norm": 0.7751222848892212,
+      "learning_rate": 2.8e-07,
+      "loss": 0.9699,
+      "step": 3
+    },
+    {
+      "epoch": 0.0016,
+      "grad_norm": 0.9030049443244934,
+      "learning_rate": 3.7333333333333334e-07,
+      "loss": 0.9404,
+      "step": 4
+    },
+    {
+      "epoch": 0.002,
+      "grad_norm": 0.9666240215301514,
+      "learning_rate": 4.6666666666666666e-07,
+      "loss": 0.8808,
+      "step": 5
+    },
+    {
+      "epoch": 0.0024,
+      "grad_norm": 1.1372973918914795,
+      "learning_rate": 5.6e-07,
+      "loss": 1.0387,
+      "step": 6
+    },
+    {
+      "epoch": 0.0028,
+      "grad_norm": 0.8488882780075073,
+      "learning_rate": 6.533333333333334e-07,
+      "loss": 1.0507,
+      "step": 7
+    },
+    {
+      "epoch": 0.0032,
+      "grad_norm": 0.6407755017280579,
+      "learning_rate": 7.466666666666667e-07,
+      "loss": 0.9219,
+      "step": 8
+    },
+    {
+      "epoch": 0.0036,
+      "grad_norm": 0.824164628982544,
+      "learning_rate": 8.4e-07,
+      "loss": 0.9653,
+      "step": 9
+    },
+    {
+      "epoch": 0.004,
+      "grad_norm": 0.9674360156059265,
+      "learning_rate": 9.333333333333333e-07,
+      "loss": 1.0234,
+      "step": 10
+    },
+    {
+      "epoch": 0.0044,
+      "grad_norm": 0.9015031456947327,
+      "learning_rate": 1.0266666666666666e-06,
+      "loss": 0.9666,
+      "step": 11
+    },
+    {
+      "epoch": 0.0048,
+      "grad_norm": 0.8106409311294556,
+      "learning_rate": 1.12e-06,
+      "loss": 0.9502,
+      "step": 12
+    },
+    {
+      "epoch": 0.0052,
+      "grad_norm": 0.9036867022514343,
+      "learning_rate": 1.2133333333333333e-06,
+      "loss": 0.8926,
+      "step": 13
+    },
+    {
+      "epoch": 0.0056,
+      "grad_norm": 0.9781032204627991,
+      "learning_rate": 1.3066666666666667e-06,
+      "loss": 0.9054,
+      "step": 14
+    },
+    {
+      "epoch": 0.006,
+      "grad_norm": 0.6719639897346497,
+      "learning_rate": 1.4000000000000001e-06,
+      "loss": 0.8787,
+      "step": 15
+    },
+    {
+      "epoch": 0.0064,
+      "grad_norm": 0.8836888670921326,
+      "learning_rate": 1.4933333333333334e-06,
+      "loss": 0.9841,
+      "step": 16
+    },
+    {
+      "epoch": 0.0068,
+      "grad_norm": 0.8169286251068115,
+      "learning_rate": 1.5866666666666666e-06,
+      "loss": 0.9571,
+      "step": 17
+    },
+    {
+      "epoch": 0.0072,
+      "grad_norm": 0.8721626400947571,
+      "learning_rate": 1.68e-06,
+      "loss": 0.959,
+      "step": 18
+    },
+    {
+      "epoch": 0.0076,
+      "grad_norm": 0.8413893580436707,
+      "learning_rate": 1.7733333333333334e-06,
+      "loss": 0.8991,
+      "step": 19
+    },
+    {
+      "epoch": 0.008,
+      "grad_norm": 0.7627416253089905,
+      "learning_rate": 1.8666666666666667e-06,
+      "loss": 0.9145,
+      "step": 20
+    },
+    {
+      "epoch": 0.0084,
+      "grad_norm": 0.796474814414978,
+      "learning_rate": 1.9600000000000003e-06,
+      "loss": 1.0169,
+      "step": 21
+    },
+    {
+      "epoch": 0.0088,
+      "grad_norm": 0.7570239901542664,
+      "learning_rate": 2.0533333333333333e-06,
+      "loss": 0.8138,
+      "step": 22
+    },
+    {
+      "epoch": 0.0092,
+      "grad_norm": 0.6165841817855835,
+      "learning_rate": 2.1466666666666663e-06,
+      "loss": 0.8433,
+      "step": 23
+    },
+    {
+      "epoch": 0.0096,
+      "grad_norm": 0.7676166296005249,
+      "learning_rate": 2.24e-06,
+      "loss": 0.9402,
+      "step": 24
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.9886164665222168,
+      "learning_rate": 2.333333333333333e-06,
+      "loss": 0.9609,
+      "step": 25
+    },
+    {
+      "epoch": 0.0104,
+      "grad_norm": 0.6577991843223572,
+      "learning_rate": 2.4266666666666666e-06,
+      "loss": 0.911,
+      "step": 26
+    },
+    {
+      "epoch": 0.0108,
+      "grad_norm": 0.8582744598388672,
+      "learning_rate": 2.52e-06,
+      "loss": 0.868,
+      "step": 27
+    },
+    {
+      "epoch": 0.0112,
+      "grad_norm": 0.7015748023986816,
+      "learning_rate": 2.6133333333333334e-06,
+      "loss": 0.9196,
+      "step": 28
+    },
+    {
+      "epoch": 0.0116,
+      "grad_norm": 0.8589622974395752,
+      "learning_rate": 2.7066666666666664e-06,
+      "loss": 0.8029,
+      "step": 29
+    },
+    {
+      "epoch": 0.012,
+      "grad_norm": 0.708187997341156,
+      "learning_rate": 2.8000000000000003e-06,
+      "loss": 0.8276,
+      "step": 30
+    },
+    {
+      "epoch": 0.0124,
+      "grad_norm": 0.5567463040351868,
+      "learning_rate": 2.8933333333333333e-06,
+      "loss": 0.7957,
+      "step": 31
+    },
+    {
+      "epoch": 0.0128,
+      "grad_norm": 0.6444608569145203,
+      "learning_rate": 2.9866666666666667e-06,
+      "loss": 0.8813,
+      "step": 32
+    },
+    {
+      "epoch": 0.0132,
+      "grad_norm": 0.8051215410232544,
+      "learning_rate": 3.08e-06,
+      "loss": 0.8985,
+      "step": 33
+    },
+    {
+      "epoch": 0.0136,
+      "grad_norm": 0.5687784552574158,
+      "learning_rate": 3.173333333333333e-06,
+      "loss": 0.7884,
+      "step": 34
+    },
+    {
+      "epoch": 0.014,
+      "grad_norm": 0.5885452628135681,
+      "learning_rate": 3.2666666666666666e-06,
+      "loss": 0.8811,
+      "step": 35
+    },
+    {
+      "epoch": 0.0144,
+      "grad_norm": 0.7129512429237366,
+      "learning_rate": 3.36e-06,
+      "loss": 0.8201,
+      "step": 36
+    },
+    {
+      "epoch": 0.0148,
+      "grad_norm": 0.5670292377471924,
+      "learning_rate": 3.4533333333333334e-06,
+      "loss": 0.7139,
+      "step": 37
+    },
+    {
+      "epoch": 0.0152,
+      "grad_norm": 0.5718544125556946,
+      "learning_rate": 3.546666666666667e-06,
+      "loss": 0.8111,
+      "step": 38
+    },
+    {
+      "epoch": 0.0156,
+      "grad_norm": 0.5730088949203491,
+      "learning_rate": 3.64e-06,
+      "loss": 0.8961,
+      "step": 39
+    },
+    {
+      "epoch": 0.016,
+      "grad_norm": 0.6300646662712097,
+      "learning_rate": 3.7333333333333333e-06,
+      "loss": 0.7828,
+      "step": 40
+    },
+    {
+      "epoch": 0.0164,
+      "grad_norm": 0.44455617666244507,
+      "learning_rate": 3.826666666666667e-06,
+      "loss": 0.7466,
+      "step": 41
+    },
+    {
+      "epoch": 0.0168,
+      "grad_norm": 0.5530583262443542,
+      "learning_rate": 3.920000000000001e-06,
+      "loss": 0.7742,
+      "step": 42
+    },
+    {
+      "epoch": 0.0172,
+      "grad_norm": 0.49698424339294434,
+      "learning_rate": 4.013333333333334e-06,
+      "loss": 0.8639,
+      "step": 43
+    },
+    {
+      "epoch": 0.0176,
+      "grad_norm": 0.4044400453567505,
+      "learning_rate": 4.106666666666667e-06,
+      "loss": 0.7938,
+      "step": 44
+    },
+    {
+      "epoch": 0.018,
+      "grad_norm": 0.4309520721435547,
+      "learning_rate": 4.2e-06,
+      "loss": 0.7424,
+      "step": 45
+    },
+    {
+      "epoch": 0.0184,
+      "grad_norm": 0.42502909898757935,
+      "learning_rate": 4.293333333333333e-06,
+      "loss": 0.7268,
+      "step": 46
+    },
+    {
+      "epoch": 0.0188,
+      "grad_norm": 0.4377117455005646,
+      "learning_rate": 4.3866666666666665e-06,
+      "loss": 0.7784,
+      "step": 47
+    },
+    {
+      "epoch": 0.0192,
+      "grad_norm": 0.4521588683128357,
+      "learning_rate": 4.48e-06,
+      "loss": 0.7764,
+      "step": 48
+    },
+    {
+      "epoch": 0.0196,
+      "grad_norm": 0.3998861610889435,
+      "learning_rate": 4.573333333333333e-06,
+      "loss": 0.7351,
+      "step": 49
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.3292255103588104,
+      "learning_rate": 4.666666666666666e-06,
+      "loss": 0.7507,
+      "step": 50
+    },
+    {
+      "epoch": 0.0204,
+      "grad_norm": 0.3537384867668152,
+      "learning_rate": 4.76e-06,
+      "loss": 0.692,
+      "step": 51
+    },
+    {
+      "epoch": 0.0208,
+      "grad_norm": 0.3487699329853058,
+      "learning_rate": 4.853333333333333e-06,
+      "loss": 0.6861,
+      "step": 52
+    },
+    {
+      "epoch": 0.0212,
+      "grad_norm": 0.4202251732349396,
+      "learning_rate": 4.946666666666666e-06,
+      "loss": 0.8594,
+      "step": 53
+    },
+    {
+      "epoch": 0.0216,
+      "grad_norm": 0.318401962518692,
+      "learning_rate": 5.04e-06,
+      "loss": 0.677,
+      "step": 54
+    },
+    {
+      "epoch": 0.022,
+      "grad_norm": 0.3611930310726166,
+      "learning_rate": 5.133333333333333e-06,
+      "loss": 0.8567,
+      "step": 55
+    },
+    {
+      "epoch": 0.0224,
+      "grad_norm": 0.4041845500469208,
+      "learning_rate": 5.226666666666667e-06,
+      "loss": 0.8468,
+      "step": 56
+    },
+    {
+      "epoch": 0.0228,
+      "grad_norm": 0.3687663674354553,
+      "learning_rate": 5.32e-06,
+      "loss": 0.7676,
+      "step": 57
+    },
+    {
+      "epoch": 0.0232,
+      "grad_norm": 0.3392384946346283,
+      "learning_rate": 5.413333333333333e-06,
+      "loss": 0.7888,
+      "step": 58
+    },
+    {
+      "epoch": 0.0236,
+      "grad_norm": 0.3390963077545166,
+      "learning_rate": 5.506666666666666e-06,
+      "loss": 0.7495,
+      "step": 59
+    },
+    {
+      "epoch": 0.024,
+      "grad_norm": 0.34994274377822876,
+      "learning_rate": 5.600000000000001e-06,
+      "loss": 0.7773,
+      "step": 60
+    },
+    {
+      "epoch": 0.0244,
+      "grad_norm": 0.3523801863193512,
+      "learning_rate": 5.693333333333334e-06,
+      "loss": 0.729,
+      "step": 61
+    },
+    {
+      "epoch": 0.0248,
+      "grad_norm": 0.3572114109992981,
+      "learning_rate": 5.786666666666667e-06,
+      "loss": 0.7668,
+      "step": 62
+    },
+    {
+      "epoch": 0.0252,
+      "grad_norm": 0.3471089005470276,
+      "learning_rate": 5.88e-06,
+      "loss": 0.7862,
+      "step": 63
+    },
+    {
+      "epoch": 0.0256,
+      "grad_norm": 0.35437729954719543,
+      "learning_rate": 5.9733333333333335e-06,
+      "loss": 0.7558,
+      "step": 64
+    },
+    {
+      "epoch": 0.026,
+      "grad_norm": 0.33817169070243835,
+      "learning_rate": 6.0666666666666665e-06,
+      "loss": 0.6867,
+      "step": 65
+    },
+    {
+      "epoch": 0.0264,
+      "grad_norm": 0.34083473682403564,
+      "learning_rate": 6.16e-06,
+      "loss": 0.7223,
+      "step": 66
+    },
+    {
+      "epoch": 0.0268,
+      "grad_norm": 0.35583579540252686,
+      "learning_rate": 6.253333333333333e-06,
+      "loss": 0.7972,
+      "step": 67
+    },
+    {
+      "epoch": 0.0272,
+      "grad_norm": 0.3342086374759674,
+      "learning_rate": 6.346666666666666e-06,
+      "loss": 0.6687,
+      "step": 68
+    },
+    {
+      "epoch": 0.0276,
+      "grad_norm": 0.32780104875564575,
+      "learning_rate": 6.44e-06,
+      "loss": 0.7269,
+      "step": 69
+    },
+    {
+      "epoch": 0.028,
+      "grad_norm": 0.3273721933364868,
+      "learning_rate": 6.533333333333333e-06,
+      "loss": 0.7169,
+      "step": 70
+    },
+    {
+      "epoch": 0.0284,
+      "grad_norm": 0.3392617404460907,
+      "learning_rate": 6.626666666666666e-06,
+      "loss": 0.7153,
+      "step": 71
+    },
+    {
+      "epoch": 0.0288,
+      "grad_norm": 0.3380572199821472,
+      "learning_rate": 6.72e-06,
+      "loss": 0.7002,
+      "step": 72
+    },
+    {
+      "epoch": 0.0292,
+      "grad_norm": 0.30406832695007324,
+      "learning_rate": 6.813333333333334e-06,
+      "loss": 0.6909,
+      "step": 73
+    },
+    {
+      "epoch": 0.0296,
+      "grad_norm": 0.31057706475257874,
+      "learning_rate": 6.906666666666667e-06,
+      "loss": 0.6694,
+      "step": 74
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.370216429233551,
+      "learning_rate": 7e-06,
+      "loss": 0.6705,
+      "step": 75
+    },
+    {
+      "epoch": 0.0304,
+      "grad_norm": 0.34453248977661133,
+      "learning_rate": 6.9999970629303366e-06,
+      "loss": 0.7221,
+      "step": 76
+    },
+    {
+      "epoch": 0.0308,
+      "grad_norm": 0.37071311473846436,
+      "learning_rate": 6.999988251726278e-06,
+      "loss": 0.7792,
+      "step": 77
+    },
+    {
+      "epoch": 0.0312,
+      "grad_norm": 0.32805272936820984,
+      "learning_rate": 6.99997356640261e-06,
+      "loss": 0.6478,
+      "step": 78
+    },
+    {
+      "epoch": 0.0316,
+      "grad_norm": 0.34910848736763,
+      "learning_rate": 6.999953006983981e-06,
+      "loss": 0.8052,
+      "step": 79
+    },
+    {
+      "epoch": 0.032,
+      "grad_norm": 0.35916391015052795,
+      "learning_rate": 6.999926573504895e-06,
+      "loss": 0.7894,
+      "step": 80
+    },
+    {
+      "epoch": 0.0324,
+      "grad_norm": 0.335122287273407,
+      "learning_rate": 6.999894266009718e-06,
+      "loss": 0.7175,
+      "step": 81
+    },
+    {
+      "epoch": 0.0328,
+      "grad_norm": 0.3043181598186493,
+      "learning_rate": 6.999856084552671e-06,
+      "loss": 0.683,
+      "step": 82
+    },
+    {
+      "epoch": 0.0332,
+      "grad_norm": 0.3139221966266632,
+      "learning_rate": 6.9998120291978345e-06,
+      "loss": 0.5613,
+      "step": 83
+    },
+    {
+      "epoch": 0.0336,
+      "grad_norm": 0.3567010164260864,
+      "learning_rate": 6.999762100019149e-06,
+      "loss": 0.7053,
+      "step": 84
+    },
+    {
+      "epoch": 0.034,
+      "grad_norm": 0.32007086277008057,
+      "learning_rate": 6.999706297100412e-06,
+      "loss": 0.7367,
+      "step": 85
+    },
+    {
+      "epoch": 0.0344,
+      "grad_norm": 0.34357309341430664,
+      "learning_rate": 6.999644620535277e-06,
+      "loss": 0.713,
+      "step": 86
+    },
+    {
+      "epoch": 0.0348,
+      "grad_norm": 0.3997757136821747,
+      "learning_rate": 6.999577070427259e-06,
+      "loss": 0.7182,
+      "step": 87
+    },
+    {
+      "epoch": 0.0352,
+      "grad_norm": 0.36973533034324646,
+      "learning_rate": 6.999503646889729e-06,
+      "loss": 0.725,
+      "step": 88
+    },
+    {
+      "epoch": 0.0356,
+      "grad_norm": 0.3873033821582794,
+      "learning_rate": 6.999424350045915e-06,
+      "loss": 0.6845,
+      "step": 89
+    },
+    {
+      "epoch": 0.036,
+      "grad_norm": 0.3782697319984436,
+      "learning_rate": 6.9993391800289045e-06,
+      "loss": 0.6828,
+      "step": 90
+    },
+    {
+      "epoch": 0.0364,
+      "grad_norm": 0.3623950481414795,
+      "learning_rate": 6.9992481369816375e-06,
+      "loss": 0.6517,
+      "step": 91
+    },
+    {
+      "epoch": 0.0368,
+      "grad_norm": 0.3495238423347473,
+      "learning_rate": 6.999151221056916e-06,
+      "loss": 0.6296,
+      "step": 92
+    },
+    {
+      "epoch": 0.0372,
+      "grad_norm": 0.35579007863998413,
+      "learning_rate": 6.999048432417397e-06,
+      "loss": 0.7149,
+      "step": 93
+    },
+    {
+      "epoch": 0.0376,
+      "grad_norm": 0.35396984219551086,
+      "learning_rate": 6.9989397712355925e-06,
+      "loss": 0.6782,
+      "step": 94
+    },
+    {
+      "epoch": 0.038,
+      "grad_norm": 0.31575071811676025,
+      "learning_rate": 6.9988252376938705e-06,
+      "loss": 0.5716,
+      "step": 95
+    },
+    {
+      "epoch": 0.0384,
+      "grad_norm": 0.327668696641922,
+      "learning_rate": 6.998704831984456e-06,
+      "loss": 0.6009,
+      "step": 96
+    },
+    {
+      "epoch": 0.0388,
+      "grad_norm": 0.32920071482658386,
+      "learning_rate": 6.998578554309431e-06,
+      "loss": 0.6265,
+      "step": 97
+    },
+    {
+      "epoch": 0.0392,
+      "grad_norm": 0.34935787320137024,
+      "learning_rate": 6.998446404880727e-06,
+      "loss": 0.5991,
+      "step": 98
+    },
+    {
+      "epoch": 0.0396,
+      "grad_norm": 0.3906346261501312,
+      "learning_rate": 6.998308383920136e-06,
+      "loss": 0.6032,
+      "step": 99
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.3648155927658081,
+      "learning_rate": 6.998164491659303e-06,
+      "loss": 0.6669,
+      "step": 100
+    },
+    {
+      "epoch": 0.0404,
+      "grad_norm": 0.35841482877731323,
+      "learning_rate": 6.998014728339723e-06,
+      "loss": 0.6297,
+      "step": 101
+    },
+    {
+      "epoch": 0.0408,
+      "grad_norm": 0.3739035129547119,
+      "learning_rate": 6.997859094212749e-06,
+      "loss": 0.5969,
+      "step": 102
+    },
+    {
+      "epoch": 0.0412,
+      "grad_norm": 0.4193374216556549,
+      "learning_rate": 6.9976975895395875e-06,
+      "loss": 0.6492,
+      "step": 103
+    },
+    {
+      "epoch": 0.0416,
+      "grad_norm": 0.4354825019836426,
+      "learning_rate": 6.997530214591294e-06,
+      "loss": 0.6634,
+      "step": 104
+    },
+    {
+      "epoch": 0.042,
+      "grad_norm": 0.39221763610839844,
+      "learning_rate": 6.997356969648778e-06,
+      "loss": 0.6369,
+      "step": 105
+    },
+    {
+      "epoch": 0.0424,
+      "grad_norm": 0.41635167598724365,
+      "learning_rate": 6.997177855002802e-06,
+      "loss": 0.673,
+      "step": 106
+    },
+    {
+      "epoch": 0.0428,
+      "grad_norm": 0.34026390314102173,
+      "learning_rate": 6.996992870953979e-06,
+      "loss": 0.6071,
+      "step": 107
+    },
+    {
+      "epoch": 0.0432,
+      "grad_norm": 0.30913060903549194,
+      "learning_rate": 6.996802017812771e-06,
+      "loss": 0.539,
+      "step": 108
+    },
+    {
+      "epoch": 0.0436,
+      "grad_norm": 0.40669533610343933,
+      "learning_rate": 6.9966052958994926e-06,
+      "loss": 0.7329,
+      "step": 109
+    },
+    {
+      "epoch": 0.044,
+      "grad_norm": 0.32939252257347107,
+      "learning_rate": 6.996402705544307e-06,
+      "loss": 0.6284,
+      "step": 110
+    },
+    {
+      "epoch": 0.0444,
+      "grad_norm": 0.33887726068496704,
+      "learning_rate": 6.996194247087227e-06,
+      "loss": 0.7118,
+      "step": 111
+    },
+    {
+      "epoch": 0.0448,
+      "grad_norm": 0.40371280908584595,
+      "learning_rate": 6.995979920878112e-06,
+      "loss": 0.7295,
+      "step": 112
+    },
+    {
+      "epoch": 0.0452,
+      "grad_norm": 0.36427947878837585,
+      "learning_rate": 6.995759727276674e-06,
+      "loss": 0.6565,
+      "step": 113
+    },
+    {
+      "epoch": 0.0456,
+      "grad_norm": 0.34461307525634766,
+      "learning_rate": 6.995533666652468e-06,
+      "loss": 0.6354,
+      "step": 114
+    },
+    {
+      "epoch": 0.046,
+      "grad_norm": 0.3352108597755432,
+      "learning_rate": 6.995301739384897e-06,
+      "loss": 0.718,
+      "step": 115
+    },
+    {
+      "epoch": 0.0464,
+      "grad_norm": 0.34925392270088196,
+      "learning_rate": 6.995063945863211e-06,
+      "loss": 0.5494,
+      "step": 116
+    },
+    {
+      "epoch": 0.0468,
+      "grad_norm": 0.3926670253276825,
+      "learning_rate": 6.994820286486506e-06,
+      "loss": 0.7144,
+      "step": 117
+    },
+    {
+      "epoch": 0.0472,
+      "grad_norm": 0.3799579441547394,
+      "learning_rate": 6.99457076166372e-06,
+      "loss": 0.6634,
+      "step": 118
+    },
+    {
+      "epoch": 0.0476,
+      "grad_norm": 0.3147001564502716,
+      "learning_rate": 6.994315371813637e-06,
+      "loss": 0.54,
+      "step": 119
+    },
+    {
+      "epoch": 0.048,
+      "grad_norm": 0.3471606969833374,
+      "learning_rate": 6.994054117364885e-06,
+      "loss": 0.6155,
+      "step": 120
+    },
+    {
+      "epoch": 0.0484,
+      "grad_norm": 0.3033629357814789,
+      "learning_rate": 6.993786998755934e-06,
+      "loss": 0.6386,
+      "step": 121
+    },
+    {
+      "epoch": 0.0488,
+      "grad_norm": 0.3264774680137634,
+      "learning_rate": 6.9935140164350955e-06,
+      "loss": 0.5989,
+      "step": 122
+    },
+    {
+      "epoch": 0.0492,
+      "grad_norm": 0.3133617341518402,
+      "learning_rate": 6.993235170860523e-06,
+      "loss": 0.5887,
+      "step": 123
+    },
+    {
+      "epoch": 0.0496,
+      "grad_norm": 0.30928051471710205,
+      "learning_rate": 6.99295046250021e-06,
+      "loss": 0.6331,
+      "step": 124
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.35159608721733093,
+      "learning_rate": 6.992659891831991e-06,
+      "loss": 0.6124,
+      "step": 125
+    },
+    {
+      "epoch": 0.0504,
+      "grad_norm": 0.31284284591674805,
+      "learning_rate": 6.992363459343536e-06,
+      "loss": 0.5279,
+      "step": 126
+    },
+    {
+      "epoch": 0.0508,
+      "grad_norm": 0.3226817548274994,
+      "learning_rate": 6.992061165532358e-06,
+      "loss": 0.5916,
+      "step": 127
+    },
+    {
+      "epoch": 0.0512,
+      "grad_norm": 0.35078591108322144,
+      "learning_rate": 6.9917530109058e-06,
+      "loss": 0.5826,
+      "step": 128
+    },
+    {
+      "epoch": 0.0516,
+      "grad_norm": 0.33087560534477234,
+      "learning_rate": 6.99143899598105e-06,
+      "loss": 0.508,
+      "step": 129
+    },
+    {
+      "epoch": 0.052,
+      "grad_norm": 0.33635255694389343,
+      "learning_rate": 6.991119121285126e-06,
+      "loss": 0.6149,
+      "step": 130
+    },
+    {
+      "epoch": 0.0524,
+      "grad_norm": 0.3316454589366913,
+      "learning_rate": 6.990793387354881e-06,
+      "loss": 0.5906,
+      "step": 131
+    },
+    {
+      "epoch": 0.0528,
+      "grad_norm": 0.3063321113586426,
+      "learning_rate": 6.990461794737004e-06,
+      "loss": 0.5499,
+      "step": 132
+    },
+    {
+      "epoch": 0.0532,
+      "grad_norm": 0.3659082055091858,
+      "learning_rate": 6.990124343988014e-06,
+      "loss": 0.5934,
+      "step": 133
+    },
+    {
+      "epoch": 0.0536,
+      "grad_norm": 0.35312604904174805,
+      "learning_rate": 6.989781035674264e-06,
+      "loss": 0.6557,
+      "step": 134
+    },
+    {
+      "epoch": 0.054,
+      "grad_norm": 0.30387458205223083,
+      "learning_rate": 6.9894318703719365e-06,
+      "loss": 0.57,
+      "step": 135
+    },
+    {
+      "epoch": 0.0544,
+      "grad_norm": 0.3087642192840576,
+      "learning_rate": 6.989076848667046e-06,
+      "loss": 0.6317,
+      "step": 136
+    },
+    {
+      "epoch": 0.0548,
+      "grad_norm": 0.3203091621398926,
+      "learning_rate": 6.988715971155433e-06,
+      "loss": 0.5949,
+      "step": 137
+    },
+    {
+      "epoch": 0.0552,
+      "grad_norm": 0.37302759289741516,
+      "learning_rate": 6.988349238442766e-06,
+      "loss": 0.7094,
+      "step": 138
+    },
+    {
+      "epoch": 0.0556,
+      "grad_norm": 0.3452549874782562,
+      "learning_rate": 6.987976651144545e-06,
+      "loss": 0.7109,
+      "step": 139
+    },
+    {
+      "epoch": 0.056,
+      "grad_norm": 0.3411109745502472,
+      "learning_rate": 6.987598209886092e-06,
+      "loss": 0.5774,
+      "step": 140
+    },
+    {
+      "epoch": 0.0564,
+      "grad_norm": 0.3314734399318695,
+      "learning_rate": 6.987213915302555e-06,
+      "loss": 0.5756,
+      "step": 141
+    },
+    {
+      "epoch": 0.0568,
+      "grad_norm": 0.34023550152778625,
+      "learning_rate": 6.986823768038903e-06,
+      "loss": 0.7022,
+      "step": 142
+    },
+    {
+      "epoch": 0.0572,
+      "grad_norm": 0.3167857825756073,
+      "learning_rate": 6.986427768749931e-06,
+      "loss": 0.5602,
+      "step": 143
+    },
+    {
+      "epoch": 0.0576,
+      "grad_norm": 0.3189705014228821,
+      "learning_rate": 6.986025918100257e-06,
+      "loss": 0.6159,
+      "step": 144
+    },
+    {
+      "epoch": 0.058,
+      "grad_norm": 0.33322688937187195,
+      "learning_rate": 6.985618216764315e-06,
+      "loss": 0.6296,
+      "step": 145
+    },
+    {
+      "epoch": 0.0584,
+      "grad_norm": 0.3075767457485199,
+      "learning_rate": 6.98520466542636e-06,
+      "loss": 0.5316,
+      "step": 146
+    },
+    {
+      "epoch": 0.0588,
+      "grad_norm": 0.3329537808895111,
+      "learning_rate": 6.984785264780469e-06,
+      "loss": 0.6537,
+      "step": 147
+    },
+    {
+      "epoch": 0.0592,
+      "grad_norm": 0.4022378623485565,
+      "learning_rate": 6.984360015530528e-06,
+      "loss": 0.5802,
+      "step": 148
+    },
+    {
+      "epoch": 0.0596,
+      "grad_norm": 0.37128475308418274,
+      "learning_rate": 6.983928918390249e-06,
+      "loss": 0.6398,
+      "step": 149
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.3873206079006195,
+      "learning_rate": 6.983491974083149e-06,
+      "loss": 0.7127,
+      "step": 150
+    },
+    {
+      "epoch": 0.0604,
+      "grad_norm": 0.39454513788223267,
+      "learning_rate": 6.983049183342565e-06,
+      "loss": 0.7034,
+      "step": 151
+    },
+    {
+      "epoch": 0.0608,
+      "grad_norm": 0.35343465209007263,
+      "learning_rate": 6.982600546911644e-06,
+      "loss": 0.56,
+      "step": 152
+    },
+    {
+      "epoch": 0.0612,
+      "grad_norm": 0.35445618629455566,
+      "learning_rate": 6.982146065543342e-06,
+      "loss": 0.5552,
+      "step": 153
+    },
+    {
+      "epoch": 0.0616,
+      "grad_norm": 0.3582383990287781,
+      "learning_rate": 6.981685740000428e-06,
+      "loss": 0.6675,
+      "step": 154
+    },
+    {
+      "epoch": 0.062,
+      "grad_norm": 0.3691483736038208,
+      "learning_rate": 6.98121957105548e-06,
+      "loss": 0.64,
+      "step": 155
+    },
+    {
+      "epoch": 0.0624,
+      "grad_norm": 0.34397178888320923,
+      "learning_rate": 6.980747559490879e-06,
+      "loss": 0.4971,
+      "step": 156
+    },
+    {
+      "epoch": 0.0628,
+      "grad_norm": 0.3170539438724518,
+      "learning_rate": 6.980269706098813e-06,
+      "loss": 0.5984,
+      "step": 157
+    },
+    {
+      "epoch": 0.0632,
+      "grad_norm": 0.3346475064754486,
+      "learning_rate": 6.979786011681279e-06,
+      "loss": 0.6547,
+      "step": 158
+    },
+    {
+      "epoch": 0.0636,
+      "grad_norm": 0.3395736515522003,
+      "learning_rate": 6.97929647705007e-06,
+      "loss": 0.5189,
+      "step": 159
+    },
+    {
+      "epoch": 0.064,
+      "grad_norm": 0.3688207268714905,
+      "learning_rate": 6.978801103026787e-06,
+      "loss": 0.6356,
+      "step": 160
+    },
+    {
+      "epoch": 0.0644,
+      "grad_norm": 0.3267056345939636,
+      "learning_rate": 6.978299890442828e-06,
+      "loss": 0.6025,
+      "step": 161
+    },
+    {
+      "epoch": 0.0648,
+      "grad_norm": 0.33124905824661255,
+      "learning_rate": 6.977792840139391e-06,
+      "loss": 0.5353,
+      "step": 162
+    },
+    {
+      "epoch": 0.0652,
+      "grad_norm": 0.33149927854537964,
+      "learning_rate": 6.977279952967471e-06,
+      "loss": 0.5713,
+      "step": 163
+    },
+    {
+      "epoch": 0.0656,
+      "grad_norm": 0.3635561466217041,
+      "learning_rate": 6.976761229787861e-06,
+      "loss": 0.6106,
+      "step": 164
+    },
+    {
+      "epoch": 0.066,
+      "grad_norm": 0.3570541739463806,
+      "learning_rate": 6.976236671471146e-06,
+      "loss": 0.6686,
+      "step": 165
+    },
+    {
+      "epoch": 0.0664,
+      "grad_norm": 0.35300779342651367,
+      "learning_rate": 6.975706278897706e-06,
+      "loss": 0.6507,
+      "step": 166
+    },
+    {
+      "epoch": 0.0668,
+      "grad_norm": 0.35414016246795654,
+      "learning_rate": 6.975170052957712e-06,
+      "loss": 0.625,
+      "step": 167
+    },
+    {
+      "epoch": 0.0672,
+      "grad_norm": 0.36714860796928406,
+      "learning_rate": 6.974627994551126e-06,
+      "loss": 0.622,
+      "step": 168
+    },
+    {
+      "epoch": 0.0676,
+      "grad_norm": 0.3507433235645294,
+      "learning_rate": 6.974080104587699e-06,
+      "loss": 0.6359,
+      "step": 169
+    },
+    {
+      "epoch": 0.068,
+      "grad_norm": 0.31815552711486816,
+      "learning_rate": 6.973526383986968e-06,
+      "loss": 0.5666,
+      "step": 170
+    },
+    {
+      "epoch": 0.0684,
+      "grad_norm": 0.3590649962425232,
+      "learning_rate": 6.972966833678257e-06,
+      "loss": 0.6185,
+      "step": 171
+    },
+    {
+      "epoch": 0.0688,
+      "grad_norm": 0.3098425269126892,
+      "learning_rate": 6.9724014546006725e-06,
+      "loss": 0.589,
+      "step": 172
+    },
+    {
+      "epoch": 0.0692,
+      "grad_norm": 0.3326118588447571,
+      "learning_rate": 6.971830247703105e-06,
+      "loss": 0.5494,
+      "step": 173
+    },
+    {
+      "epoch": 0.0696,
+      "grad_norm": 0.3343932330608368,
+      "learning_rate": 6.9712532139442275e-06,
+      "loss": 0.5835,
+      "step": 174
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.3673574924468994,
+      "learning_rate": 6.970670354292488e-06,
+      "loss": 0.6293,
+      "step": 175
+    },
+    {
+      "epoch": 0.0704,
+      "grad_norm": 0.3204609155654907,
+      "learning_rate": 6.9700816697261155e-06,
+      "loss": 0.614,
+      "step": 176
+    },
+    {
+      "epoch": 0.0708,
+      "grad_norm": 0.3640765845775604,
+      "learning_rate": 6.969487161233115e-06,
+      "loss": 0.6177,
+      "step": 177
+    },
+    {
+      "epoch": 0.0712,
+      "grad_norm": 0.3944571912288666,
+      "learning_rate": 6.968886829811264e-06,
+      "loss": 0.6153,
+      "step": 178
+    },
+    {
+      "epoch": 0.0716,
+      "grad_norm": 0.3283381760120392,
+      "learning_rate": 6.968280676468116e-06,
+      "loss": 0.6266,
+      "step": 179
+    },
+    {
+      "epoch": 0.072,
+      "grad_norm": 0.33310529589653015,
+      "learning_rate": 6.967668702220992e-06,
+      "loss": 0.6409,
+      "step": 180
+    },
+    {
+      "epoch": 0.0724,
+      "grad_norm": 0.369968980550766,
+      "learning_rate": 6.967050908096984e-06,
+      "loss": 0.6872,
+      "step": 181
+    },
+    {
+      "epoch": 0.0728,
+      "grad_norm": 0.36735835671424866,
+      "learning_rate": 6.966427295132952e-06,
+      "loss": 0.6841,
+      "step": 182
+    },
+    {
+      "epoch": 0.0732,
+      "grad_norm": 0.3196526765823364,
+      "learning_rate": 6.965797864375522e-06,
+      "loss": 0.5331,
+      "step": 183
+    },
+    {
+      "epoch": 0.0736,
+      "grad_norm": 0.35567739605903625,
+      "learning_rate": 6.965162616881084e-06,
+      "loss": 0.6226,
+      "step": 184
+    },
+    {
+      "epoch": 0.074,
+      "grad_norm": 0.38448742032051086,
+      "learning_rate": 6.964521553715789e-06,
+      "loss": 0.6539,
+      "step": 185
+    },
+    {
+      "epoch": 0.0744,
+      "grad_norm": 0.3042820990085602,
+      "learning_rate": 6.96387467595555e-06,
+      "loss": 0.5079,
+      "step": 186
+    },
+    {
+      "epoch": 0.0748,
+      "grad_norm": 0.37548917531967163,
+      "learning_rate": 6.963221984686039e-06,
+      "loss": 0.5517,
+      "step": 187
+    },
+    {
+      "epoch": 0.0752,
+      "grad_norm": 0.30672183632850647,
+      "learning_rate": 6.962563481002683e-06,
+      "loss": 0.533,
+      "step": 188
+    },
+    {
+      "epoch": 0.0756,
+      "grad_norm": 0.3858972489833832,
+      "learning_rate": 6.9618991660106675e-06,
+      "loss": 0.5961,
+      "step": 189
+    },
+    {
+      "epoch": 0.076,
+      "grad_norm": 0.3610040843486786,
+      "learning_rate": 6.961229040824927e-06,
+      "loss": 0.6161,
+      "step": 190
+    },
+    {
+      "epoch": 0.0764,
+      "grad_norm": 0.33447155356407166,
+      "learning_rate": 6.960553106570152e-06,
+      "loss": 0.5496,
+      "step": 191
+    },
+    {
+      "epoch": 0.0768,
+      "grad_norm": 0.3702964782714844,
+      "learning_rate": 6.959871364380779e-06,
+      "loss": 0.5928,
+      "step": 192
+    },
+    {
+      "epoch": 0.0772,
+      "grad_norm": 0.3681807816028595,
+      "learning_rate": 6.959183815400994e-06,
+      "loss": 0.5623,
+      "step": 193
+    },
+    {
+      "epoch": 0.0776,
+      "grad_norm": 0.3803996741771698,
+      "learning_rate": 6.958490460784728e-06,
+      "loss": 0.7486,
+      "step": 194
+    },
+    {
+      "epoch": 0.078,
+      "grad_norm": 0.37968674302101135,
+      "learning_rate": 6.957791301695654e-06,
+      "loss": 0.6576,
+      "step": 195
+    },
+    {
+      "epoch": 0.0784,
+      "grad_norm": 0.3494243919849396,
+      "learning_rate": 6.95708633930719e-06,
+      "loss": 0.641,
+      "step": 196
+    },
+    {
+      "epoch": 0.0788,
+      "grad_norm": 0.3401181399822235,
+      "learning_rate": 6.956375574802492e-06,
+      "loss": 0.5693,
+      "step": 197
+    },
+    {
+      "epoch": 0.0792,
+      "grad_norm": 0.361893892288208,
+      "learning_rate": 6.955659009374456e-06,
+      "loss": 0.5069,
+      "step": 198
+    },
+    {
+      "epoch": 0.0796,
+      "grad_norm": 0.3742466866970062,
+      "learning_rate": 6.954936644225709e-06,
+      "loss": 0.5492,
+      "step": 199
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.4014841616153717,
+      "learning_rate": 6.954208480568617e-06,
+      "loss": 0.6495,
+      "step": 200
+    },
+    {
+      "epoch": 0.0804,
+      "grad_norm": 0.36528071761131287,
+      "learning_rate": 6.953474519625275e-06,
+      "loss": 0.7123,
+      "step": 201
+    },
+    {
+      "epoch": 0.0808,
+      "grad_norm": 0.36030465364456177,
+      "learning_rate": 6.952734762627509e-06,
+      "loss": 0.5403,
+      "step": 202
+    },
+    {
+      "epoch": 0.0812,
+      "grad_norm": 0.36890923976898193,
+      "learning_rate": 6.95198921081687e-06,
+      "loss": 0.5394,
+      "step": 203
+    },
+    {
+      "epoch": 0.0816,
+      "grad_norm": 0.3089402914047241,
+      "learning_rate": 6.951237865444638e-06,
+      "loss": 0.5311,
+      "step": 204
+    },
+    {
+      "epoch": 0.082,
+      "grad_norm": 0.3627172112464905,
+      "learning_rate": 6.950480727771816e-06,
+      "loss": 0.6114,
+      "step": 205
+    },
+    {
+      "epoch": 0.0824,
+      "grad_norm": 0.35881373286247253,
+      "learning_rate": 6.949717799069127e-06,
+      "loss": 0.5708,
+      "step": 206
+    },
+    {
+      "epoch": 0.0828,
+      "grad_norm": 0.3384152054786682,
+      "learning_rate": 6.9489490806170125e-06,
+      "loss": 0.556,
+      "step": 207
+    },
+    {
+      "epoch": 0.0832,
+      "grad_norm": 0.37565433979034424,
+      "learning_rate": 6.948174573705633e-06,
+      "loss": 0.5681,
+      "step": 208
+    },
+    {
+      "epoch": 0.0836,
+      "grad_norm": 0.3774128556251526,
+      "learning_rate": 6.947394279634864e-06,
+      "loss": 0.5823,
+      "step": 209
+    },
+    {
+      "epoch": 0.084,
+      "grad_norm": 0.35291945934295654,
+      "learning_rate": 6.946608199714292e-06,
+      "loss": 0.6107,
+      "step": 210
+    },
+    {
+      "epoch": 0.0844,
+      "grad_norm": 0.3333498537540436,
+      "learning_rate": 6.945816335263216e-06,
+      "loss": 0.5596,
+      "step": 211
+    },
+    {
+      "epoch": 0.0848,
+      "grad_norm": 0.3444664180278778,
+      "learning_rate": 6.94501868761064e-06,
+      "loss": 0.5874,
+      "step": 212
+    },
+    {
+      "epoch": 0.0852,
+      "grad_norm": 0.3725375235080719,
+      "learning_rate": 6.94421525809528e-06,
+      "loss": 0.6062,
+      "step": 213
+    },
+    {
+      "epoch": 0.0856,
+      "grad_norm": 0.321502149105072,
+      "learning_rate": 6.943406048065549e-06,
+      "loss": 0.5407,
+      "step": 214
+    },
+    {
+      "epoch": 0.086,
+      "grad_norm": 0.37382030487060547,
+      "learning_rate": 6.942591058879567e-06,
+      "loss": 0.6295,
+      "step": 215
+    },
+    {
+      "epoch": 0.0864,
+      "grad_norm": 0.3625917434692383,
+      "learning_rate": 6.941770291905149e-06,
+      "loss": 0.5852,
+      "step": 216
+    },
+    {
+      "epoch": 0.0868,
+      "grad_norm": 0.3712015748023987,
+      "learning_rate": 6.940943748519811e-06,
+      "loss": 0.6046,
+      "step": 217
+    },
+    {
+      "epoch": 0.0872,
+      "grad_norm": 0.3427964448928833,
+      "learning_rate": 6.940111430110762e-06,
+      "loss": 0.5867,
+      "step": 218
+    },
+    {
+      "epoch": 0.0876,
+      "grad_norm": 0.3536797761917114,
+      "learning_rate": 6.939273338074902e-06,
+      "loss": 0.5192,
+      "step": 219
+    },
+    {
+      "epoch": 0.088,
+      "grad_norm": 0.36839908361434937,
+      "learning_rate": 6.938429473818823e-06,
+      "loss": 0.7215,
+      "step": 220
+    },
+    {
+      "epoch": 0.0884,
+      "grad_norm": 0.3873510956764221,
+      "learning_rate": 6.937579838758804e-06,
+      "loss": 0.5555,
+      "step": 221
+    },
+    {
+      "epoch": 0.0888,
+      "grad_norm": 0.371750146150589,
+      "learning_rate": 6.936724434320808e-06,
+      "loss": 0.6247,
+      "step": 222
+    },
+    {
+      "epoch": 0.0892,
+      "grad_norm": 0.35638293623924255,
+      "learning_rate": 6.935863261940484e-06,
+      "loss": 0.6369,
+      "step": 223
+    },
+    {
+      "epoch": 0.0896,
+      "grad_norm": 0.3919801712036133,
+      "learning_rate": 6.934996323063157e-06,
+      "loss": 0.6772,
+      "step": 224
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.4034706652164459,
+      "learning_rate": 6.934123619143837e-06,
+      "loss": 0.655,
+      "step": 225
+    },
+    {
+      "epoch": 0.0904,
+      "grad_norm": 0.36784958839416504,
+      "learning_rate": 6.933245151647201e-06,
+      "loss": 0.6007,
+      "step": 226
+    },
+    {
+      "epoch": 0.0908,
+      "grad_norm": 0.36665797233581543,
+      "learning_rate": 6.932360922047605e-06,
+      "loss": 0.6603,
+      "step": 227
+    },
+    {
+      "epoch": 0.0912,
+      "grad_norm": 0.334515243768692,
+      "learning_rate": 6.931470931829073e-06,
+      "loss": 0.5401,
+      "step": 228
+    },
+    {
+      "epoch": 0.0916,
+      "grad_norm": 0.3468990921974182,
+      "learning_rate": 6.9305751824853e-06,
+      "loss": 0.5969,
+      "step": 229
+    },
+    {
+      "epoch": 0.092,
+      "grad_norm": 0.38072285056114197,
+      "learning_rate": 6.929673675519645e-06,
+      "loss": 0.6163,
+      "step": 230
+    },
+    {
+      "epoch": 0.0924,
+      "grad_norm": 0.33580705523490906,
+      "learning_rate": 6.928766412445129e-06,
+      "loss": 0.5892,
+      "step": 231
+    },
+    {
+      "epoch": 0.0928,
+      "grad_norm": 0.3787344992160797,
+      "learning_rate": 6.927853394784435e-06,
+      "loss": 0.5176,
+      "step": 232
+    },
+    {
+      "epoch": 0.0932,
+      "grad_norm": 0.3536042273044586,
+      "learning_rate": 6.926934624069905e-06,
+      "loss": 0.5962,
+      "step": 233
+    },
+    {
+      "epoch": 0.0936,
+      "grad_norm": 0.3443499207496643,
+      "learning_rate": 6.926010101843533e-06,
+      "loss": 0.5676,
+      "step": 234
+    },
+    {
+      "epoch": 0.094,
+      "grad_norm": 0.3868633210659027,
+      "learning_rate": 6.925079829656971e-06,
+      "loss": 0.6262,
+      "step": 235
+    },
+    {
+      "epoch": 0.0944,
+      "grad_norm": 0.36493581533432007,
+      "learning_rate": 6.924143809071516e-06,
+      "loss": 0.6733,
+      "step": 236
+    },
+    {
+      "epoch": 0.0948,
+      "grad_norm": 0.3501265347003937,
+      "learning_rate": 6.923202041658117e-06,
+      "loss": 0.5836,
+      "step": 237
+    },
+    {
+      "epoch": 0.0952,
+      "grad_norm": 0.319475919008255,
+      "learning_rate": 6.922254528997366e-06,
+      "loss": 0.5541,
+      "step": 238
+    },
+    {
+      "epoch": 0.0956,
+      "grad_norm": 0.3460980951786041,
+      "learning_rate": 6.921301272679497e-06,
+      "loss": 0.5136,
+      "step": 239
+    },
+    {
+      "epoch": 0.096,
+      "grad_norm": 0.3520514965057373,
+      "learning_rate": 6.920342274304385e-06,
+      "loss": 0.525,
+      "step": 240
+    },
+    {
+      "epoch": 0.0964,
+      "grad_norm": 0.36067137122154236,
+      "learning_rate": 6.919377535481542e-06,
+      "loss": 0.5507,
+      "step": 241
+    },
+    {
+      "epoch": 0.0968,
+      "grad_norm": 0.38827893137931824,
+      "learning_rate": 6.918407057830112e-06,
+      "loss": 0.5688,
+      "step": 242
+    },
+    {
+      "epoch": 0.0972,
+      "grad_norm": 0.3810005784034729,
+      "learning_rate": 6.917430842978874e-06,
+      "loss": 0.61,
+      "step": 243
+    },
+    {
+      "epoch": 0.0976,
+      "grad_norm": 0.40279942750930786,
+      "learning_rate": 6.916448892566233e-06,
+      "loss": 0.6187,
+      "step": 244
+    },
+    {
+      "epoch": 0.098,
+      "grad_norm": 0.3923201858997345,
+      "learning_rate": 6.915461208240224e-06,
+      "loss": 0.6091,
+      "step": 245
+    },
+    {
+      "epoch": 0.0984,
+      "grad_norm": 0.36452221870422363,
+      "learning_rate": 6.9144677916585e-06,
+      "loss": 0.5192,
+      "step": 246
+    },
+    {
+      "epoch": 0.0988,
+      "grad_norm": 0.3636743724346161,
+      "learning_rate": 6.9134686444883375e-06,
+      "loss": 0.638,
+      "step": 247
+    },
+    {
+      "epoch": 0.0992,
+      "grad_norm": 0.35182589292526245,
+      "learning_rate": 6.912463768406632e-06,
+      "loss": 0.5919,
+      "step": 248
+    },
+    {
+      "epoch": 0.0996,
+      "grad_norm": 0.4089539647102356,
+      "learning_rate": 6.911453165099894e-06,
+      "loss": 0.6028,
+      "step": 249
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.3915441036224365,
+      "learning_rate": 6.910436836264241e-06,
+      "loss": 0.5141,
+      "step": 250
+    },
+    {
+      "epoch": 0.1004,
+      "grad_norm": 0.36682936549186707,
+      "learning_rate": 6.909414783605407e-06,
+      "loss": 0.5331,
+      "step": 251
+    },
+    {
+      "epoch": 0.1008,
+      "grad_norm": 0.4124221205711365,
+      "learning_rate": 6.908387008838727e-06,
+      "loss": 0.6104,
+      "step": 252
+    },
+    {
+      "epoch": 0.1012,
+      "grad_norm": 0.3299160897731781,
+      "learning_rate": 6.907353513689142e-06,
+      "loss": 0.5446,
+      "step": 253
+    },
+    {
+      "epoch": 0.1016,
+      "grad_norm": 0.3507899045944214,
+      "learning_rate": 6.906314299891195e-06,
+      "loss": 0.5387,
+      "step": 254
+    },
+    {
+      "epoch": 0.102,
+      "grad_norm": 0.3491806983947754,
+      "learning_rate": 6.905269369189024e-06,
+      "loss": 0.5394,
+      "step": 255
+    },
+    {
+      "epoch": 0.1024,
+      "grad_norm": 0.3935128450393677,
+      "learning_rate": 6.904218723336361e-06,
+      "loss": 0.5568,
+      "step": 256
+    },
+    {
+      "epoch": 0.1028,
+      "grad_norm": 0.4095175266265869,
+      "learning_rate": 6.903162364096535e-06,
+      "loss": 0.6018,
+      "step": 257
+    },
+    {
+      "epoch": 0.1032,
+      "grad_norm": 0.42139819264411926,
+      "learning_rate": 6.902100293242458e-06,
+      "loss": 0.7185,
+      "step": 258
+    },
+    {
+      "epoch": 0.1036,
+      "grad_norm": 0.388468474149704,
+      "learning_rate": 6.901032512556632e-06,
+      "loss": 0.5549,
+      "step": 259
+    },
+    {
+      "epoch": 0.104,
+      "grad_norm": 0.35161855816841125,
+      "learning_rate": 6.899959023831139e-06,
+      "loss": 0.5437,
+      "step": 260
+    },
+    {
+      "epoch": 0.1044,
+      "grad_norm": 0.4257499575614929,
+      "learning_rate": 6.898879828867646e-06,
+      "loss": 0.6383,
+      "step": 261
+    },
+    {
+      "epoch": 0.1048,
+      "grad_norm": 0.48008543252944946,
+      "learning_rate": 6.89779492947739e-06,
+      "loss": 0.6421,
+      "step": 262
+    },
+    {
+      "epoch": 0.1052,
+      "grad_norm": 0.3906846046447754,
+      "learning_rate": 6.896704327481186e-06,
+      "loss": 0.6624,
+      "step": 263
+    },
+    {
+      "epoch": 0.1056,
+      "grad_norm": 0.3648388385772705,
+      "learning_rate": 6.895608024709421e-06,
+      "loss": 0.6273,
+      "step": 264
+    },
+    {
+      "epoch": 0.106,
+      "grad_norm": 0.3587093949317932,
+      "learning_rate": 6.894506023002046e-06,
+      "loss": 0.6208,
+      "step": 265
+    },
+    {
+      "epoch": 0.1064,
+      "grad_norm": 0.3674977123737335,
+      "learning_rate": 6.8933983242085795e-06,
+      "loss": 0.5363,
+      "step": 266
+    },
+    {
+      "epoch": 0.1068,
+      "grad_norm": 0.3786182403564453,
+      "learning_rate": 6.8922849301881e-06,
+      "loss": 0.5776,
+      "step": 267
+    },
+    {
+      "epoch": 0.1072,
+      "grad_norm": 0.3598049283027649,
+      "learning_rate": 6.891165842809245e-06,
+      "loss": 0.5218,
+      "step": 268
+    },
+    {
+      "epoch": 0.1076,
+      "grad_norm": 0.352017879486084,
+      "learning_rate": 6.890041063950209e-06,
+      "loss": 0.5722,
+      "step": 269
+    },
+    {
+      "epoch": 0.108,
+      "grad_norm": 0.3780212700366974,
+      "learning_rate": 6.888910595498735e-06,
+      "loss": 0.5742,
+      "step": 270
+    },
+    {
+      "epoch": 0.1084,
+      "grad_norm": 0.37677013874053955,
+      "learning_rate": 6.887774439352119e-06,
+      "loss": 0.6385,
+      "step": 271
+    },
+    {
+      "epoch": 0.1088,
+      "grad_norm": 0.34929612278938293,
+      "learning_rate": 6.886632597417199e-06,
+      "loss": 0.5202,
+      "step": 272
+    },
+    {
+      "epoch": 0.1092,
+      "grad_norm": 0.3872514069080353,
+      "learning_rate": 6.885485071610358e-06,
+      "loss": 0.6136,
+      "step": 273
+    },
+    {
+      "epoch": 0.1096,
+      "grad_norm": 0.3379985988140106,
+      "learning_rate": 6.88433186385752e-06,
+      "loss": 0.4804,
+      "step": 274
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.38230156898498535,
+      "learning_rate": 6.883172976094139e-06,
+      "loss": 0.6076,
+      "step": 275
+    },
+    {
+      "epoch": 0.1104,
+      "grad_norm": 0.3500521183013916,
+      "learning_rate": 6.8820084102652096e-06,
+      "loss": 0.5821,
+      "step": 276
+    },
+    {
+      "epoch": 0.1108,
+      "grad_norm": 0.38483259081840515,
+      "learning_rate": 6.8808381683252494e-06,
+      "loss": 0.5886,
+      "step": 277
+    },
+    {
+      "epoch": 0.1112,
+      "grad_norm": 0.36327463388442993,
+      "learning_rate": 6.879662252238309e-06,
+      "loss": 0.5153,
+      "step": 278
+    },
+    {
+      "epoch": 0.1116,
+      "grad_norm": 0.361081600189209,
+      "learning_rate": 6.878480663977954e-06,
+      "loss": 0.5925,
+      "step": 279
+    },
+    {
+      "epoch": 0.112,
+      "grad_norm": 0.38348904252052307,
+      "learning_rate": 6.877293405527277e-06,
+      "loss": 0.5876,
+      "step": 280
+    },
+    {
+      "epoch": 0.1124,
+      "grad_norm": 0.3530191481113434,
+      "learning_rate": 6.8761004788788834e-06,
+      "loss": 0.5548,
+      "step": 281
+    },
+    {
+      "epoch": 0.1128,
+      "grad_norm": 0.40636175870895386,
+      "learning_rate": 6.874901886034892e-06,
+      "loss": 0.6094,
+      "step": 282
+    },
+    {
+      "epoch": 0.1132,
+      "grad_norm": 0.3557603359222412,
+      "learning_rate": 6.873697629006933e-06,
+      "loss": 0.5073,
+      "step": 283
+    },
+    {
+      "epoch": 0.1136,
+      "grad_norm": 0.3685925304889679,
+      "learning_rate": 6.87248770981614e-06,
+      "loss": 0.5382,
+      "step": 284
+    },
+    {
+      "epoch": 0.114,
+      "grad_norm": 0.3815518319606781,
+      "learning_rate": 6.871272130493154e-06,
+      "loss": 0.5528,
+      "step": 285
+    },
+    {
+      "epoch": 0.1144,
+      "grad_norm": 0.4013282060623169,
+      "learning_rate": 6.870050893078109e-06,
+      "loss": 0.6034,
+      "step": 286
+    },
+    {
+      "epoch": 0.1148,
+      "grad_norm": 0.36867278814315796,
+      "learning_rate": 6.868823999620642e-06,
+      "loss": 0.5778,
+      "step": 287
+    },
+    {
+      "epoch": 0.1152,
+      "grad_norm": 0.37960368394851685,
+      "learning_rate": 6.8675914521798785e-06,
+      "loss": 0.6205,
+      "step": 288
+    },
+    {
+      "epoch": 0.1156,
+      "grad_norm": 0.3762238025665283,
+      "learning_rate": 6.8663532528244335e-06,
+      "loss": 0.6443,
+      "step": 289
+    },
+    {
+      "epoch": 0.116,
+      "grad_norm": 0.3985435366630554,
+      "learning_rate": 6.86510940363241e-06,
+      "loss": 0.5664,
+      "step": 290
+    },
+    {
+      "epoch": 0.1164,
+      "grad_norm": 0.37209147214889526,
+      "learning_rate": 6.86385990669139e-06,
+      "loss": 0.5963,
+      "step": 291
+    },
+    {
+      "epoch": 0.1168,
+      "grad_norm": 0.37561458349227905,
+      "learning_rate": 6.86260476409844e-06,
+      "loss": 0.5929,
+      "step": 292
+    },
+    {
+      "epoch": 0.1172,
+      "grad_norm": 0.3788372576236725,
+      "learning_rate": 6.861343977960093e-06,
+      "loss": 0.5671,
+      "step": 293
+    },
+    {
+      "epoch": 0.1176,
+      "grad_norm": 0.383171021938324,
+      "learning_rate": 6.860077550392362e-06,
+      "loss": 0.5947,
+      "step": 294
+    },
+    {
+      "epoch": 0.118,
+      "grad_norm": 0.33166104555130005,
+      "learning_rate": 6.858805483520723e-06,
+      "loss": 0.5039,
+      "step": 295
+    },
+    {
+      "epoch": 0.1184,
+      "grad_norm": 0.3684594929218292,
+      "learning_rate": 6.8575277794801195e-06,
+      "loss": 0.6029,
+      "step": 296
+    },
+    {
+      "epoch": 0.1188,
+      "grad_norm": 0.40030691027641296,
+      "learning_rate": 6.856244440414953e-06,
+      "loss": 0.6785,
+      "step": 297
+    },
+    {
+      "epoch": 0.1192,
+      "grad_norm": 0.34929394721984863,
+      "learning_rate": 6.8549554684790855e-06,
+      "loss": 0.5659,
+      "step": 298
+    },
+    {
+      "epoch": 0.1196,
+      "grad_norm": 0.3945501446723938,
+      "learning_rate": 6.853660865835831e-06,
+      "loss": 0.6544,
+      "step": 299
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.39090695977211,
+      "learning_rate": 6.852360634657955e-06,
+      "loss": 0.5841,
+      "step": 300
+    },
+    {
+      "epoch": 0.1204,
+      "grad_norm": 0.3813556134700775,
+      "learning_rate": 6.8510547771276675e-06,
+      "loss": 0.6404,
+      "step": 301
+    },
+    {
+      "epoch": 0.1208,
+      "grad_norm": 0.34115681052207947,
+      "learning_rate": 6.849743295436623e-06,
+      "loss": 0.4899,
+      "step": 302
+    },
+    {
+      "epoch": 0.1212,
+      "grad_norm": 0.4142250716686249,
+      "learning_rate": 6.8484261917859155e-06,
+      "loss": 0.6104,
+      "step": 303
+    },
+    {
+      "epoch": 0.1216,
+      "grad_norm": 0.35604000091552734,
+      "learning_rate": 6.847103468386074e-06,
+      "loss": 0.4774,
+      "step": 304
+    },
+    {
+      "epoch": 0.122,
+      "grad_norm": 0.3621678352355957,
+      "learning_rate": 6.845775127457056e-06,
+      "loss": 0.5684,
+      "step": 305
+    },
+    {
+      "epoch": 0.1224,
+      "grad_norm": 0.3987256586551666,
+      "learning_rate": 6.844441171228253e-06,
+      "loss": 0.5834,
+      "step": 306
+    },
+    {
+      "epoch": 0.1228,
+      "grad_norm": 0.4088221490383148,
+      "learning_rate": 6.843101601938477e-06,
+      "loss": 0.672,
+      "step": 307
+    },
+    {
+      "epoch": 0.1232,
+      "grad_norm": 0.3717223107814789,
+      "learning_rate": 6.8417564218359585e-06,
+      "loss": 0.5031,
+      "step": 308
+    },
+    {
+      "epoch": 0.1236,
+      "grad_norm": 0.3734017610549927,
+      "learning_rate": 6.840405633178352e-06,
+      "loss": 0.6068,
+      "step": 309
+    },
+    {
+      "epoch": 0.124,
+      "grad_norm": 0.39380621910095215,
+      "learning_rate": 6.839049238232719e-06,
+      "loss": 0.5903,
+      "step": 310
+    },
+    {
+      "epoch": 0.1244,
+      "grad_norm": 0.38201144337654114,
+      "learning_rate": 6.837687239275532e-06,
+      "loss": 0.5724,
+      "step": 311
+    },
+    {
+      "epoch": 0.1248,
+      "grad_norm": 0.3409825563430786,
+      "learning_rate": 6.836319638592667e-06,
+      "loss": 0.5576,
+      "step": 312
+    },
+    {
+      "epoch": 0.1252,
+      "grad_norm": 0.31887713074684143,
+      "learning_rate": 6.834946438479404e-06,
+      "loss": 0.4632,
+      "step": 313
+    },
+    {
+      "epoch": 0.1256,
+      "grad_norm": 0.40399324893951416,
+      "learning_rate": 6.83356764124042e-06,
+      "loss": 0.5815,
+      "step": 314
+    },
+    {
+      "epoch": 0.126,
+      "grad_norm": 0.4099929630756378,
+      "learning_rate": 6.832183249189787e-06,
+      "loss": 0.5968,
+      "step": 315
+    },
+    {
+      "epoch": 0.1264,
+      "grad_norm": 0.4446236789226532,
+      "learning_rate": 6.830793264650963e-06,
+      "loss": 0.6245,
+      "step": 316
+    },
+    {
+      "epoch": 0.1268,
+      "grad_norm": 0.3838817775249481,
+      "learning_rate": 6.8293976899567955e-06,
+      "loss": 0.5863,
+      "step": 317
+    },
+    {
+      "epoch": 0.1272,
+      "grad_norm": 0.4060188829898834,
+      "learning_rate": 6.827996527449513e-06,
+      "loss": 0.683,
+      "step": 318
+    },
+    {
+      "epoch": 0.1276,
+      "grad_norm": 0.36124998331069946,
+      "learning_rate": 6.826589779480722e-06,
+      "loss": 0.5444,
+      "step": 319
+    },
+    {
+      "epoch": 0.128,
+      "grad_norm": 0.35842403769493103,
+      "learning_rate": 6.825177448411405e-06,
+      "loss": 0.5025,
+      "step": 320
+    },
+    {
+      "epoch": 0.1284,
+      "grad_norm": 0.40654295682907104,
+      "learning_rate": 6.823759536611913e-06,
+      "loss": 0.5622,
+      "step": 321
+    },
+    {
+      "epoch": 0.1288,
+      "grad_norm": 0.42024552822113037,
+      "learning_rate": 6.822336046461962e-06,
+      "loss": 0.5747,
+      "step": 322
+    },
+    {
+      "epoch": 0.1292,
+      "grad_norm": 0.4173870086669922,
+      "learning_rate": 6.820906980350632e-06,
+      "loss": 0.6264,
+      "step": 323
+    },
+    {
+      "epoch": 0.1296,
+      "grad_norm": 0.37095096707344055,
+      "learning_rate": 6.819472340676363e-06,
+      "loss": 0.5924,
+      "step": 324
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.3586147725582123,
+      "learning_rate": 6.818032129846946e-06,
+      "loss": 0.5423,
+      "step": 325
+    },
+    {
+      "epoch": 0.1304,
+      "grad_norm": 0.391284316778183,
+      "learning_rate": 6.816586350279525e-06,
+      "loss": 0.4857,
+      "step": 326
+    },
+    {
+      "epoch": 0.1308,
+      "grad_norm": 0.41726839542388916,
+      "learning_rate": 6.815135004400586e-06,
+      "loss": 0.6486,
+      "step": 327
+    },
+    {
+      "epoch": 0.1312,
+      "grad_norm": 0.3971002697944641,
+      "learning_rate": 6.813678094645962e-06,
+      "loss": 0.5285,
+      "step": 328
+    },
+    {
+      "epoch": 0.1316,
+      "grad_norm": 0.35923510789871216,
+      "learning_rate": 6.812215623460821e-06,
+      "loss": 0.5287,
+      "step": 329
+    },
+    {
+      "epoch": 0.132,
+      "grad_norm": 0.35825857520103455,
+      "learning_rate": 6.810747593299666e-06,
+      "loss": 0.5429,
+      "step": 330
+    },
+    {
+      "epoch": 0.1324,
+      "grad_norm": 0.3599400222301483,
+      "learning_rate": 6.809274006626329e-06,
+      "loss": 0.51,
+      "step": 331
+    },
+    {
+      "epoch": 0.1328,
+      "grad_norm": 0.37383949756622314,
+      "learning_rate": 6.8077948659139695e-06,
+      "loss": 0.5065,
+      "step": 332
+    },
+    {
+      "epoch": 0.1332,
+      "grad_norm": 0.3496291935443878,
+      "learning_rate": 6.8063101736450655e-06,
+      "loss": 0.5497,
+      "step": 333
+    },
+    {
+      "epoch": 0.1336,
+      "grad_norm": 0.33252400159835815,
+      "learning_rate": 6.804819932311415e-06,
+      "loss": 0.5179,
+      "step": 334
+    },
+    {
+      "epoch": 0.134,
+      "grad_norm": 0.3825218975543976,
+      "learning_rate": 6.803324144414127e-06,
+      "loss": 0.5685,
+      "step": 335
+    },
+    {
+      "epoch": 0.1344,
+      "grad_norm": 0.40384697914123535,
+      "learning_rate": 6.801822812463622e-06,
+      "loss": 0.6212,
+      "step": 336
+    },
+    {
+      "epoch": 0.1348,
+      "grad_norm": 0.38337215781211853,
+      "learning_rate": 6.800315938979622e-06,
+      "loss": 0.5862,
+      "step": 337
+    },
+    {
+      "epoch": 0.1352,
+      "grad_norm": 0.37698617577552795,
+      "learning_rate": 6.798803526491154e-06,
+      "loss": 0.6163,
+      "step": 338
+    },
+    {
+      "epoch": 0.1356,
+      "grad_norm": 0.3776755928993225,
+      "learning_rate": 6.797285577536535e-06,
+      "loss": 0.5958,
+      "step": 339
+    },
+    {
+      "epoch": 0.136,
+      "grad_norm": 0.4053572118282318,
+      "learning_rate": 6.795762094663379e-06,
+      "loss": 0.5847,
+      "step": 340
+    },
+    {
+      "epoch": 0.1364,
+      "grad_norm": 0.386665403842926,
+      "learning_rate": 6.794233080428587e-06,
+      "loss": 0.5703,
+      "step": 341
+    },
+    {
+      "epoch": 0.1368,
+      "grad_norm": 0.39310121536254883,
+      "learning_rate": 6.792698537398341e-06,
+      "loss": 0.5127,
+      "step": 342
+    },
+    {
+      "epoch": 0.1372,
+      "grad_norm": 0.3588424623012543,
+      "learning_rate": 6.791158468148105e-06,
+      "loss": 0.5456,
+      "step": 343
+    },
+    {
+      "epoch": 0.1376,
+      "grad_norm": 0.45625194907188416,
+      "learning_rate": 6.789612875262616e-06,
+      "loss": 0.5879,
+      "step": 344
+    },
+    {
+      "epoch": 0.138,
+      "grad_norm": 0.40522149205207825,
+      "learning_rate": 6.788061761335882e-06,
+      "loss": 0.6526,
+      "step": 345
+    },
+    {
+      "epoch": 0.1384,
+      "grad_norm": 0.3919071853160858,
+      "learning_rate": 6.7865051289711785e-06,
+      "loss": 0.4821,
+      "step": 346
+    },
+    {
+      "epoch": 0.1388,
+      "grad_norm": 0.4460979998111725,
+      "learning_rate": 6.784942980781039e-06,
+      "loss": 0.6008,
+      "step": 347
+    },
+    {
+      "epoch": 0.1392,
+      "grad_norm": 0.38495203852653503,
+      "learning_rate": 6.783375319387258e-06,
+      "loss": 0.6104,
+      "step": 348
+    },
+    {
+      "epoch": 0.1396,
+      "grad_norm": 0.3975396752357483,
+      "learning_rate": 6.781802147420881e-06,
+      "loss": 0.5317,
+      "step": 349
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.37524691224098206,
+      "learning_rate": 6.780223467522203e-06,
+      "loss": 0.5957,
+      "step": 350
+    },
+    {
+      "epoch": 0.1404,
+      "grad_norm": 0.3901492953300476,
+      "learning_rate": 6.778639282340763e-06,
+      "loss": 0.5707,
+      "step": 351
+    },
+    {
+      "epoch": 0.1408,
+      "grad_norm": 0.3822447657585144,
+      "learning_rate": 6.777049594535339e-06,
+      "loss": 0.5925,
+      "step": 352
+    },
+    {
+      "epoch": 0.1412,
+      "grad_norm": 0.333995521068573,
+      "learning_rate": 6.775454406773944e-06,
+      "loss": 0.5544,
+      "step": 353
+    },
+    {
+      "epoch": 0.1416,
+      "grad_norm": 0.3573135733604431,
+      "learning_rate": 6.773853721733824e-06,
+      "loss": 0.5393,
+      "step": 354
+    },
+    {
+      "epoch": 0.142,
+      "grad_norm": 0.39041879773139954,
+      "learning_rate": 6.772247542101449e-06,
+      "loss": 0.5191,
+      "step": 355
+    },
+    {
+      "epoch": 0.1424,
+      "grad_norm": 0.3725902736186981,
+      "learning_rate": 6.770635870572511e-06,
+      "loss": 0.5571,
+      "step": 356
+    },
+    {
+      "epoch": 0.1428,
+      "grad_norm": 0.4029092490673065,
+      "learning_rate": 6.7690187098519185e-06,
+      "loss": 0.6127,
+      "step": 357
+    },
+    {
+      "epoch": 0.1432,
+      "grad_norm": 0.37953653931617737,
+      "learning_rate": 6.767396062653795e-06,
+      "loss": 0.5833,
+      "step": 358
+    },
+    {
+      "epoch": 0.1436,
+      "grad_norm": 0.4030202627182007,
+      "learning_rate": 6.76576793170147e-06,
+      "loss": 0.6068,
+      "step": 359
+    },
+    {
+      "epoch": 0.144,
+      "grad_norm": 0.4032868444919586,
+      "learning_rate": 6.7641343197274775e-06,
+      "loss": 0.6152,
+      "step": 360
+    },
+    {
+      "epoch": 0.1444,
+      "grad_norm": 0.4173031151294708,
+      "learning_rate": 6.762495229473551e-06,
+      "loss": 0.6258,
+      "step": 361
+    },
+    {
+      "epoch": 0.1448,
+      "grad_norm": 0.46217605471611023,
+      "learning_rate": 6.760850663690616e-06,
+      "loss": 0.5772,
+      "step": 362
+    },
+    {
+      "epoch": 0.1452,
+      "grad_norm": 0.44322094321250916,
+      "learning_rate": 6.75920062513879e-06,
+      "loss": 0.646,
+      "step": 363
+    },
+    {
+      "epoch": 0.1456,
+      "grad_norm": 0.39636361598968506,
+      "learning_rate": 6.757545116587375e-06,
+      "loss": 0.5529,
+      "step": 364
+    },
+    {
+      "epoch": 0.146,
+      "grad_norm": 0.40710270404815674,
+      "learning_rate": 6.755884140814852e-06,
+      "loss": 0.6251,
+      "step": 365
+    },
+    {
+      "epoch": 0.1464,
+      "grad_norm": 0.4058838188648224,
+      "learning_rate": 6.754217700608879e-06,
+      "loss": 0.6392,
+      "step": 366
+    },
+    {
+      "epoch": 0.1468,
+      "grad_norm": 0.3798361122608185,
+      "learning_rate": 6.752545798766288e-06,
+      "loss": 0.5321,
+      "step": 367
+    },
+    {
+      "epoch": 0.1472,
+      "grad_norm": 0.4159586727619171,
+      "learning_rate": 6.750868438093072e-06,
+      "loss": 0.632,
+      "step": 368
+    },
+    {
+      "epoch": 0.1476,
+      "grad_norm": 0.4010884463787079,
+      "learning_rate": 6.7491856214043896e-06,
+      "loss": 0.5941,
+      "step": 369
+    },
+    {
+      "epoch": 0.148,
+      "grad_norm": 0.3940584063529968,
+      "learning_rate": 6.747497351524553e-06,
+      "loss": 0.544,
+      "step": 370
+    },
+    {
+      "epoch": 0.1484,
+      "grad_norm": 0.3773007094860077,
+      "learning_rate": 6.74580363128703e-06,
+      "loss": 0.5979,
+      "step": 371
+    },
+    {
+      "epoch": 0.1488,
+      "grad_norm": 0.43958812952041626,
+      "learning_rate": 6.744104463534436e-06,
+      "loss": 0.4857,
+      "step": 372
+    },
+    {
+      "epoch": 0.1492,
+      "grad_norm": 0.42353639006614685,
+      "learning_rate": 6.742399851118524e-06,
+      "loss": 0.6292,
+      "step": 373
+    },
+    {
+      "epoch": 0.1496,
+      "grad_norm": 0.3908763825893402,
+      "learning_rate": 6.740689796900191e-06,
+      "loss": 0.5618,
+      "step": 374
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.3558228313922882,
+      "learning_rate": 6.738974303749465e-06,
+      "loss": 0.4899,
+      "step": 375
+    },
+    {
+      "epoch": 0.1504,
+      "grad_norm": 0.42558762431144714,
+      "learning_rate": 6.7372533745455e-06,
+      "loss": 0.6123,
+      "step": 376
+    },
+    {
+      "epoch": 0.1508,
+      "grad_norm": 0.41860321164131165,
+      "learning_rate": 6.735527012176576e-06,
+      "loss": 0.61,
+      "step": 377
+    },
+    {
+      "epoch": 0.1512,
+      "grad_norm": 0.42713692784309387,
+      "learning_rate": 6.733795219540093e-06,
+      "loss": 0.532,
+      "step": 378
+    },
+    {
+      "epoch": 0.1516,
+      "grad_norm": 0.40444377064704895,
+      "learning_rate": 6.73205799954256e-06,
+      "loss": 0.6371,
+      "step": 379
+    },
+    {
+      "epoch": 0.152,
+      "grad_norm": 0.39443764090538025,
+      "learning_rate": 6.730315355099601e-06,
+      "loss": 0.5507,
+      "step": 380
+    },
+    {
+      "epoch": 0.1524,
+      "grad_norm": 0.4197935461997986,
+      "learning_rate": 6.728567289135937e-06,
+      "loss": 0.6081,
+      "step": 381
+    },
+    {
+      "epoch": 0.1528,
+      "grad_norm": 0.388790100812912,
+      "learning_rate": 6.726813804585392e-06,
+      "loss": 0.5826,
+      "step": 382
+    },
+    {
+      "epoch": 0.1532,
+      "grad_norm": 0.4298461675643921,
+      "learning_rate": 6.725054904390888e-06,
+      "loss": 0.5901,
+      "step": 383
+    },
+    {
+      "epoch": 0.1536,
+      "grad_norm": 0.40709060430526733,
+      "learning_rate": 6.7232905915044275e-06,
+      "loss": 0.6201,
+      "step": 384
+    },
+    {
+      "epoch": 0.154,
+      "grad_norm": 0.4251411557197571,
+      "learning_rate": 6.721520868887104e-06,
+      "loss": 0.6158,
+      "step": 385
+    },
+    {
+      "epoch": 0.1544,
+      "grad_norm": 0.3800036311149597,
+      "learning_rate": 6.719745739509088e-06,
+      "loss": 0.4875,
+      "step": 386
+    },
+    {
+      "epoch": 0.1548,
+      "grad_norm": 0.4196547567844391,
+      "learning_rate": 6.717965206349623e-06,
+      "loss": 0.5107,
+      "step": 387
+    },
+    {
+      "epoch": 0.1552,
+      "grad_norm": 0.4319261610507965,
+      "learning_rate": 6.716179272397026e-06,
+      "loss": 0.6308,
+      "step": 388
+    },
+    {
+      "epoch": 0.1556,
+      "grad_norm": 0.38977041840553284,
+      "learning_rate": 6.714387940648673e-06,
+      "loss": 0.514,
+      "step": 389
+    },
+    {
+      "epoch": 0.156,
+      "grad_norm": 0.43319880962371826,
+      "learning_rate": 6.712591214111003e-06,
+      "loss": 0.7118,
+      "step": 390
+    },
+    {
+      "epoch": 0.1564,
+      "grad_norm": 0.4345259964466095,
+      "learning_rate": 6.710789095799508e-06,
+      "loss": 0.6231,
+      "step": 391
+    },
+    {
+      "epoch": 0.1568,
+      "grad_norm": 0.41207200288772583,
+      "learning_rate": 6.708981588738728e-06,
+      "loss": 0.5641,
+      "step": 392
+    },
+    {
+      "epoch": 0.1572,
+      "grad_norm": 0.41626623272895813,
+      "learning_rate": 6.70716869596225e-06,
+      "loss": 0.6129,
+      "step": 393
+    },
+    {
+      "epoch": 0.1576,
+      "grad_norm": 0.3993598222732544,
+      "learning_rate": 6.7053504205126975e-06,
+      "loss": 0.6225,
+      "step": 394
+    },
+    {
+      "epoch": 0.158,
+      "grad_norm": 0.3917825520038605,
+      "learning_rate": 6.703526765441728e-06,
+      "loss": 0.589,
+      "step": 395
+    },
+    {
+      "epoch": 0.1584,
+      "grad_norm": 0.42511889338493347,
+      "learning_rate": 6.7016977338100285e-06,
+      "loss": 0.5113,
+      "step": 396
+    },
+    {
+      "epoch": 0.1588,
+      "grad_norm": 0.4189378321170807,
+      "learning_rate": 6.6998633286873104e-06,
+      "loss": 0.6223,
+      "step": 397
+    },
+    {
+      "epoch": 0.1592,
+      "grad_norm": 0.4240991771221161,
+      "learning_rate": 6.698023553152302e-06,
+      "loss": 0.628,
+      "step": 398
+    },
+    {
+      "epoch": 0.1596,
+      "grad_norm": 0.3946475386619568,
+      "learning_rate": 6.696178410292745e-06,
+      "loss": 0.4924,
+      "step": 399
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.4627389907836914,
+      "learning_rate": 6.694327903205391e-06,
+      "loss": 0.5833,
+      "step": 400
+    },
+    {
+      "epoch": 0.1604,
+      "grad_norm": 0.4218280017375946,
+      "learning_rate": 6.692472034995991e-06,
+      "loss": 0.5969,
+      "step": 401
+    },
+    {
+      "epoch": 0.1608,
+      "grad_norm": 0.4313870072364807,
+      "learning_rate": 6.690610808779299e-06,
+      "loss": 0.6232,
+      "step": 402
+    },
+    {
+      "epoch": 0.1612,
+      "grad_norm": 0.4355848729610443,
+      "learning_rate": 6.688744227679056e-06,
+      "loss": 0.574,
+      "step": 403
+    },
+    {
+      "epoch": 0.1616,
+      "grad_norm": 0.4002700746059418,
+      "learning_rate": 6.686872294827994e-06,
+      "loss": 0.5956,
+      "step": 404
+    },
+    {
+      "epoch": 0.162,
+      "grad_norm": 0.41936513781547546,
+      "learning_rate": 6.6849950133678266e-06,
+      "loss": 0.5749,
+      "step": 405
+    },
+    {
+      "epoch": 0.1624,
+      "grad_norm": 0.4025267958641052,
+      "learning_rate": 6.683112386449241e-06,
+      "loss": 0.6316,
+      "step": 406
+    },
+    {
+      "epoch": 0.1628,
+      "grad_norm": 0.3825077414512634,
+      "learning_rate": 6.681224417231899e-06,
+      "loss": 0.5343,
+      "step": 407
+    },
+    {
+      "epoch": 0.1632,
+      "grad_norm": 0.40748634934425354,
+      "learning_rate": 6.679331108884428e-06,
+      "loss": 0.6376,
+      "step": 408
+    },
+    {
+      "epoch": 0.1636,
+      "grad_norm": 0.43103963136672974,
+      "learning_rate": 6.677432464584414e-06,
+      "loss": 0.6273,
+      "step": 409
+    },
+    {
+      "epoch": 0.164,
+      "grad_norm": 0.37271183729171753,
+      "learning_rate": 6.675528487518402e-06,
+      "loss": 0.5296,
+      "step": 410
+    },
+    {
+      "epoch": 0.1644,
+      "grad_norm": 0.4089418351650238,
+      "learning_rate": 6.673619180881885e-06,
+      "loss": 0.5102,
+      "step": 411
+    },
+    {
+      "epoch": 0.1648,
+      "grad_norm": 0.3724643588066101,
+      "learning_rate": 6.6717045478793e-06,
+      "loss": 0.5508,
+      "step": 412
+    },
+    {
+      "epoch": 0.1652,
+      "grad_norm": 0.4249953627586365,
+      "learning_rate": 6.669784591724026e-06,
+      "loss": 0.6005,
+      "step": 413
+    },
+    {
+      "epoch": 0.1656,
+      "grad_norm": 0.40450453758239746,
+      "learning_rate": 6.667859315638372e-06,
+      "loss": 0.5968,
+      "step": 414
+    },
+    {
+      "epoch": 0.166,
+      "grad_norm": 0.4060114026069641,
+      "learning_rate": 6.665928722853581e-06,
+      "loss": 0.6613,
+      "step": 415
+    },
+    {
+      "epoch": 0.1664,
+      "grad_norm": 0.38569897413253784,
+      "learning_rate": 6.663992816609815e-06,
+      "loss": 0.5891,
+      "step": 416
+    },
+    {
+      "epoch": 0.1668,
+      "grad_norm": 0.41489410400390625,
+      "learning_rate": 6.662051600156154e-06,
+      "loss": 0.5452,
+      "step": 417
+    },
+    {
+      "epoch": 0.1672,
+      "grad_norm": 0.4076942205429077,
+      "learning_rate": 6.660105076750592e-06,
+      "loss": 0.513,
+      "step": 418
+    },
+    {
+      "epoch": 0.1676,
+      "grad_norm": 0.40700528025627136,
+      "learning_rate": 6.658153249660029e-06,
+      "loss": 0.6024,
+      "step": 419
+    },
+    {
+      "epoch": 0.168,
+      "grad_norm": 0.3968103528022766,
+      "learning_rate": 6.656196122160265e-06,
+      "loss": 0.602,
+      "step": 420
+    },
+    {
+      "epoch": 0.1684,
+      "grad_norm": 0.4172203242778778,
+      "learning_rate": 6.654233697535998e-06,
+      "loss": 0.6217,
+      "step": 421
+    },
+    {
+      "epoch": 0.1688,
+      "grad_norm": 0.3906404972076416,
+      "learning_rate": 6.652265979080816e-06,
+      "loss": 0.5455,
+      "step": 422
+    },
+    {
+      "epoch": 0.1692,
+      "grad_norm": 0.4094797372817993,
+      "learning_rate": 6.650292970097189e-06,
+      "loss": 0.5759,
+      "step": 423
+    },
+    {
+      "epoch": 0.1696,
+      "grad_norm": 0.4227938652038574,
+      "learning_rate": 6.64831467389647e-06,
+      "loss": 0.6473,
+      "step": 424
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.3829863965511322,
+      "learning_rate": 6.646331093798884e-06,
+      "loss": 0.5197,
+      "step": 425
+    },
+    {
+      "epoch": 0.1704,
+      "grad_norm": 0.6422174572944641,
+      "learning_rate": 6.644342233133524e-06,
+      "loss": 0.6327,
+      "step": 426
+    },
+    {
+      "epoch": 0.1708,
+      "grad_norm": 0.36776864528656006,
+      "learning_rate": 6.6423480952383445e-06,
+      "loss": 0.597,
+      "step": 427
+    },
+    {
+      "epoch": 0.1712,
+      "grad_norm": 0.4166823625564575,
+      "learning_rate": 6.6403486834601586e-06,
+      "loss": 0.5709,
+      "step": 428
+    },
+    {
+      "epoch": 0.1716,
+      "grad_norm": 0.4036274254322052,
+      "learning_rate": 6.638344001154631e-06,
+      "loss": 0.5106,
+      "step": 429
+    },
+    {
+      "epoch": 0.172,
+      "grad_norm": 0.4060764014720917,
+      "learning_rate": 6.636334051686271e-06,
+      "loss": 0.5733,
+      "step": 430
+    },
+    {
+      "epoch": 0.1724,
+      "grad_norm": 0.4096258282661438,
+      "learning_rate": 6.634318838428427e-06,
+      "loss": 0.5882,
+      "step": 431
+    },
+    {
+      "epoch": 0.1728,
+      "grad_norm": 0.3832622468471527,
+      "learning_rate": 6.632298364763285e-06,
+      "loss": 0.5158,
+      "step": 432
+    },
+    {
+      "epoch": 0.1732,
+      "grad_norm": 0.3626929521560669,
+      "learning_rate": 6.630272634081854e-06,
+      "loss": 0.5444,
+      "step": 433
+    },
+    {
+      "epoch": 0.1736,
+      "grad_norm": 0.41777303814888,
+      "learning_rate": 6.628241649783973e-06,
+      "loss": 0.6262,
+      "step": 434
+    },
+    {
+      "epoch": 0.174,
+      "grad_norm": 0.3933648467063904,
+      "learning_rate": 6.626205415278292e-06,
+      "loss": 0.6132,
+      "step": 435
+    },
+    {
+      "epoch": 0.1744,
+      "grad_norm": 0.35853779315948486,
+      "learning_rate": 6.624163933982277e-06,
+      "loss": 0.535,
+      "step": 436
+    },
+    {
+      "epoch": 0.1748,
+      "grad_norm": 0.428905725479126,
+      "learning_rate": 6.6221172093221975e-06,
+      "loss": 0.6116,
+      "step": 437
+    },
+    {
+      "epoch": 0.1752,
+      "grad_norm": 0.3868679106235504,
+      "learning_rate": 6.620065244733125e-06,
+      "loss": 0.5847,
+      "step": 438
+    },
+    {
+      "epoch": 0.1756,
+      "grad_norm": 0.3978234529495239,
+      "learning_rate": 6.6180080436589235e-06,
+      "loss": 0.6209,
+      "step": 439
+    },
+    {
+      "epoch": 0.176,
+      "grad_norm": 0.37370485067367554,
+      "learning_rate": 6.6159456095522445e-06,
+      "loss": 0.5518,
+      "step": 440
+    },
+    {
+      "epoch": 0.1764,
+      "grad_norm": 0.35376763343811035,
+      "learning_rate": 6.613877945874525e-06,
+      "loss": 0.5174,
+      "step": 441
+    },
+    {
+      "epoch": 0.1768,
+      "grad_norm": 0.47480249404907227,
+      "learning_rate": 6.611805056095978e-06,
+      "loss": 0.6535,
+      "step": 442
+    },
+    {
+      "epoch": 0.1772,
+      "grad_norm": 0.3956994116306305,
+      "learning_rate": 6.6097269436955864e-06,
+      "loss": 0.5688,
+      "step": 443
+    },
+    {
+      "epoch": 0.1776,
+      "grad_norm": 0.40941232442855835,
+      "learning_rate": 6.6076436121611e-06,
+      "loss": 0.5516,
+      "step": 444
+    },
+    {
+      "epoch": 0.178,
+      "grad_norm": 0.3964068591594696,
+      "learning_rate": 6.605555064989027e-06,
+      "loss": 0.5267,
+      "step": 445
+    },
+    {
+      "epoch": 0.1784,
+      "grad_norm": 0.4019598364830017,
+      "learning_rate": 6.6034613056846306e-06,
+      "loss": 0.5498,
+      "step": 446
+    },
+    {
+      "epoch": 0.1788,
+      "grad_norm": 0.3716028928756714,
+      "learning_rate": 6.60136233776192e-06,
+      "loss": 0.5977,
+      "step": 447
+    },
+    {
+      "epoch": 0.1792,
+      "grad_norm": 0.4159116744995117,
+      "learning_rate": 6.599258164743644e-06,
+      "loss": 0.589,
+      "step": 448
+    },
+    {
+      "epoch": 0.1796,
+      "grad_norm": 0.3982093334197998,
+      "learning_rate": 6.597148790161293e-06,
+      "loss": 0.5503,
+      "step": 449
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.44451043009757996,
+      "learning_rate": 6.595034217555082e-06,
+      "loss": 0.5861,
+      "step": 450
+    },
+    {
+      "epoch": 0.1804,
+      "grad_norm": 0.4734339714050293,
+      "learning_rate": 6.5929144504739544e-06,
+      "loss": 0.6438,
+      "step": 451
+    },
+    {
+      "epoch": 0.1808,
+      "grad_norm": 0.45358017086982727,
+      "learning_rate": 6.590789492475569e-06,
+      "loss": 0.6934,
+      "step": 452
+    },
+    {
+      "epoch": 0.1812,
+      "grad_norm": 0.4209020733833313,
+      "learning_rate": 6.588659347126295e-06,
+      "loss": 0.5052,
+      "step": 453
+    },
+    {
+      "epoch": 0.1816,
+      "grad_norm": 0.4025874435901642,
+      "learning_rate": 6.586524018001212e-06,
+      "loss": 0.6219,
+      "step": 454
+    },
+    {
+      "epoch": 0.182,
+      "grad_norm": 0.44042277336120605,
+      "learning_rate": 6.584383508684096e-06,
+      "loss": 0.5807,
+      "step": 455
+    },
+    {
+      "epoch": 0.1824,
+      "grad_norm": 0.3821624517440796,
+      "learning_rate": 6.582237822767418e-06,
+      "loss": 0.6256,
+      "step": 456
+    },
+    {
+      "epoch": 0.1828,
+      "grad_norm": 0.42015254497528076,
+      "learning_rate": 6.58008696385234e-06,
+      "loss": 0.6157,
+      "step": 457
+    },
+    {
+      "epoch": 0.1832,
+      "grad_norm": 0.43211668729782104,
+      "learning_rate": 6.5779309355487e-06,
+      "loss": 0.5517,
+      "step": 458
+    },
+    {
+      "epoch": 0.1836,
+      "grad_norm": 0.40651610493659973,
+      "learning_rate": 6.575769741475019e-06,
+      "loss": 0.5512,
+      "step": 459
+    },
+    {
+      "epoch": 0.184,
+      "grad_norm": 0.42855092883110046,
+      "learning_rate": 6.57360338525848e-06,
+      "loss": 0.5923,
+      "step": 460
+    },
+    {
+      "epoch": 0.1844,
+      "grad_norm": 0.43255260586738586,
+      "learning_rate": 6.571431870534937e-06,
+      "loss": 0.5141,
+      "step": 461
+    },
+    {
+      "epoch": 0.1848,
+      "grad_norm": 0.4311109781265259,
+      "learning_rate": 6.569255200948898e-06,
+      "loss": 0.6245,
+      "step": 462
+    },
+    {
+      "epoch": 0.1852,
+      "grad_norm": 0.3710770606994629,
+      "learning_rate": 6.567073380153522e-06,
+      "loss": 0.5078,
+      "step": 463
+    },
+    {
+      "epoch": 0.1856,
+      "grad_norm": 0.400833398103714,
+      "learning_rate": 6.564886411810617e-06,
+      "loss": 0.4928,
+      "step": 464
+    },
+    {
+      "epoch": 0.186,
+      "grad_norm": 0.4306429624557495,
+      "learning_rate": 6.562694299590625e-06,
+      "loss": 0.5688,
+      "step": 465
+    },
+    {
+      "epoch": 0.1864,
+      "grad_norm": 0.4112854301929474,
+      "learning_rate": 6.5604970471726246e-06,
+      "loss": 0.5235,
+      "step": 466
+    },
+    {
+      "epoch": 0.1868,
+      "grad_norm": 0.49645546078681946,
+      "learning_rate": 6.558294658244321e-06,
+      "loss": 0.6761,
+      "step": 467
+    },
+    {
+      "epoch": 0.1872,
+      "grad_norm": 0.38806602358818054,
+      "learning_rate": 6.55608713650204e-06,
+      "loss": 0.4431,
+      "step": 468
+    },
+    {
+      "epoch": 0.1876,
+      "grad_norm": 0.43465444445610046,
+      "learning_rate": 6.553874485650722e-06,
+      "loss": 0.5803,
+      "step": 469
+    },
+    {
+      "epoch": 0.188,
+      "grad_norm": 0.4388769268989563,
+      "learning_rate": 6.551656709403914e-06,
+      "loss": 0.5055,
+      "step": 470
+    },
+    {
+      "epoch": 0.1884,
+      "grad_norm": 0.47469204664230347,
+      "learning_rate": 6.549433811483768e-06,
+      "loss": 0.6419,
+      "step": 471
+    },
+    {
+      "epoch": 0.1888,
+      "grad_norm": 0.4161481261253357,
+      "learning_rate": 6.54720579562103e-06,
+      "loss": 0.6525,
+      "step": 472
+    },
+    {
+      "epoch": 0.1892,
+      "grad_norm": 0.42730727791786194,
+      "learning_rate": 6.544972665555034e-06,
+      "loss": 0.5436,
+      "step": 473
+    },
+    {
+      "epoch": 0.1896,
+      "grad_norm": 0.41790273785591125,
+      "learning_rate": 6.5427344250337025e-06,
+      "loss": 0.6091,
+      "step": 474
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.4080909192562103,
+      "learning_rate": 6.540491077813529e-06,
+      "loss": 0.5359,
+      "step": 475
+    },
+    {
+      "epoch": 0.1904,
+      "grad_norm": 0.3704690933227539,
+      "learning_rate": 6.53824262765958e-06,
+      "loss": 0.4738,
+      "step": 476
+    },
+    {
+      "epoch": 0.1908,
+      "grad_norm": 0.41908273100852966,
+      "learning_rate": 6.53598907834549e-06,
+      "loss": 0.6515,
+      "step": 477
+    },
+    {
+      "epoch": 0.1912,
+      "grad_norm": 0.40230220556259155,
+      "learning_rate": 6.533730433653446e-06,
+      "loss": 0.6099,
+      "step": 478
+    },
+    {
+      "epoch": 0.1916,
+      "grad_norm": 0.4547681212425232,
+      "learning_rate": 6.5314666973741885e-06,
+      "loss": 0.6308,
+      "step": 479
+    },
+    {
+      "epoch": 0.192,
+      "grad_norm": 0.4127230644226074,
+      "learning_rate": 6.5291978733070066e-06,
+      "loss": 0.605,
+      "step": 480
+    },
+    {
+      "epoch": 0.1924,
+      "grad_norm": 0.4303308427333832,
+      "learning_rate": 6.526923965259722e-06,
+      "loss": 0.5355,
+      "step": 481
+    },
+    {
+      "epoch": 0.1928,
+      "grad_norm": 0.42116066813468933,
+      "learning_rate": 6.524644977048695e-06,
+      "loss": 0.487,
+      "step": 482
+    },
+    {
+      "epoch": 0.1932,
+      "grad_norm": 0.43687164783477783,
+      "learning_rate": 6.5223609124988085e-06,
+      "loss": 0.5955,
+      "step": 483
+    },
+    {
+      "epoch": 0.1936,
+      "grad_norm": 0.39758434891700745,
+      "learning_rate": 6.520071775443468e-06,
+      "loss": 0.5845,
+      "step": 484
+    },
+    {
+      "epoch": 0.194,
+      "grad_norm": 0.4243478775024414,
+      "learning_rate": 6.517777569724588e-06,
+      "loss": 0.5288,
+      "step": 485
+    },
+    {
+      "epoch": 0.1944,
+      "grad_norm": 0.43449336290359497,
+      "learning_rate": 6.5154782991925965e-06,
+      "loss": 0.6107,
+      "step": 486
+    },
+    {
+      "epoch": 0.1948,
+      "grad_norm": 0.47633621096611023,
+      "learning_rate": 6.5131739677064135e-06,
+      "loss": 0.5935,
+      "step": 487
+    },
+    {
+      "epoch": 0.1952,
+      "grad_norm": 0.4590698182582855,
+      "learning_rate": 6.51086457913346e-06,
+      "loss": 0.5879,
+      "step": 488
+    },
+    {
+      "epoch": 0.1956,
+      "grad_norm": 0.43686172366142273,
+      "learning_rate": 6.508550137349641e-06,
+      "loss": 0.6266,
+      "step": 489
+    },
+    {
+      "epoch": 0.196,
+      "grad_norm": 0.4315257966518402,
+      "learning_rate": 6.506230646239344e-06,
+      "loss": 0.5888,
+      "step": 490
+    },
+    {
+      "epoch": 0.1964,
+      "grad_norm": 0.3936934769153595,
+      "learning_rate": 6.50390610969543e-06,
+      "loss": 0.5724,
+      "step": 491
+    },
+    {
+      "epoch": 0.1968,
+      "grad_norm": 0.3999297022819519,
+      "learning_rate": 6.501576531619225e-06,
+      "loss": 0.529,
+      "step": 492
+    },
+    {
+      "epoch": 0.1972,
+      "grad_norm": 0.4457099139690399,
+      "learning_rate": 6.499241915920524e-06,
+      "loss": 0.6437,
+      "step": 493
+    },
+    {
+      "epoch": 0.1976,
+      "grad_norm": 0.38449960947036743,
+      "learning_rate": 6.4969022665175684e-06,
+      "loss": 0.6073,
+      "step": 494
+    },
+    {
+      "epoch": 0.198,
+      "grad_norm": 0.47077676653862,
+      "learning_rate": 6.494557587337054e-06,
+      "loss": 0.6728,
+      "step": 495
+    },
+    {
+      "epoch": 0.1984,
+      "grad_norm": 0.4107373356819153,
+      "learning_rate": 6.492207882314114e-06,
+      "loss": 0.6166,
+      "step": 496
+    },
+    {
+      "epoch": 0.1988,
+      "grad_norm": 0.4460639953613281,
+      "learning_rate": 6.48985315539232e-06,
+      "loss": 0.5883,
+      "step": 497
+    },
+    {
+      "epoch": 0.1992,
+      "grad_norm": 0.42625370621681213,
+      "learning_rate": 6.4874934105236685e-06,
+      "loss": 0.6357,
+      "step": 498
+    },
+    {
+      "epoch": 0.1996,
+      "grad_norm": 0.3935810327529907,
+      "learning_rate": 6.4851286516685816e-06,
+      "loss": 0.531,
+      "step": 499
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.39704349637031555,
+      "learning_rate": 6.482758882795892e-06,
+      "loss": 0.5663,
+      "step": 500
+    },
+    {
+      "epoch": 0.2004,
+      "grad_norm": 0.3972982168197632,
+      "learning_rate": 6.480384107882846e-06,
+      "loss": 0.5619,
+      "step": 501
+    },
+    {
+      "epoch": 0.2008,
+      "grad_norm": 0.4732000529766083,
+      "learning_rate": 6.478004330915086e-06,
+      "loss": 0.6,
+      "step": 502
+    },
+    {
+      "epoch": 0.2012,
+      "grad_norm": 0.450668066740036,
+      "learning_rate": 6.475619555886654e-06,
+      "loss": 0.6062,
+      "step": 503
+    },
+    {
+      "epoch": 0.2016,
+      "grad_norm": 0.4316413998603821,
+      "learning_rate": 6.473229786799979e-06,
+      "loss": 0.5445,
+      "step": 504
+    },
+    {
+      "epoch": 0.202,
+      "grad_norm": 0.415632963180542,
+      "learning_rate": 6.4708350276658704e-06,
+      "loss": 0.6405,
+      "step": 505
+    },
+    {
+      "epoch": 0.2024,
+      "grad_norm": 0.42289984226226807,
+      "learning_rate": 6.468435282503516e-06,
+      "loss": 0.557,
+      "step": 506
+    },
+    {
+      "epoch": 0.2028,
+      "grad_norm": 0.3919405937194824,
+      "learning_rate": 6.466030555340465e-06,
+      "loss": 0.4969,
+      "step": 507
+    },
+    {
+      "epoch": 0.2032,
+      "grad_norm": 0.45820194482803345,
+      "learning_rate": 6.463620850212638e-06,
+      "loss": 0.6139,
+      "step": 508
+    },
+    {
+      "epoch": 0.2036,
+      "grad_norm": 0.4082326591014862,
+      "learning_rate": 6.461206171164299e-06,
+      "loss": 0.6121,
+      "step": 509
+    },
+    {
+      "epoch": 0.204,
+      "grad_norm": 0.42317843437194824,
+      "learning_rate": 6.458786522248069e-06,
+      "loss": 0.5903,
+      "step": 510
+    },
+    {
+      "epoch": 0.2044,
+      "grad_norm": 0.4115223288536072,
+      "learning_rate": 6.456361907524906e-06,
+      "loss": 0.5829,
+      "step": 511
+    },
+    {
+      "epoch": 0.2048,
+      "grad_norm": 0.4232071340084076,
+      "learning_rate": 6.453932331064102e-06,
+      "loss": 0.5426,
+      "step": 512
+    },
+    {
+      "epoch": 0.2052,
+      "grad_norm": 0.3934287428855896,
+      "learning_rate": 6.451497796943277e-06,
+      "loss": 0.5725,
+      "step": 513
+    },
+    {
+      "epoch": 0.2056,
+      "grad_norm": 0.4107876121997833,
+      "learning_rate": 6.449058309248373e-06,
+      "loss": 0.5557,
+      "step": 514
+    },
+    {
+      "epoch": 0.206,
+      "grad_norm": 0.4142535626888275,
+      "learning_rate": 6.446613872073645e-06,
+      "loss": 0.5678,
+      "step": 515
+    },
+    {
+      "epoch": 0.2064,
+      "grad_norm": 0.392920583486557,
+      "learning_rate": 6.4441644895216515e-06,
+      "loss": 0.5124,
+      "step": 516
+    },
+    {
+      "epoch": 0.2068,
+      "grad_norm": 0.4002549946308136,
+      "learning_rate": 6.441710165703256e-06,
+      "loss": 0.4938,
+      "step": 517
+    },
+    {
+      "epoch": 0.2072,
+      "grad_norm": 0.4306708872318268,
+      "learning_rate": 6.439250904737611e-06,
+      "loss": 0.5763,
+      "step": 518
+    },
+    {
+      "epoch": 0.2076,
+      "grad_norm": 0.46144577860832214,
+      "learning_rate": 6.436786710752159e-06,
+      "loss": 0.5427,
+      "step": 519
+    },
+    {
+      "epoch": 0.208,
+      "grad_norm": 0.4587547481060028,
+      "learning_rate": 6.434317587882619e-06,
+      "loss": 0.6381,
+      "step": 520
+    },
+    {
+      "epoch": 0.2084,
+      "grad_norm": 0.42765599489212036,
+      "learning_rate": 6.431843540272982e-06,
+      "loss": 0.6067,
+      "step": 521
+    },
+    {
+      "epoch": 0.2088,
+      "grad_norm": 0.40350085496902466,
+      "learning_rate": 6.429364572075506e-06,
+      "loss": 0.5612,
+      "step": 522
+    },
+    {
+      "epoch": 0.2092,
+      "grad_norm": 0.4018450379371643,
+      "learning_rate": 6.426880687450706e-06,
+      "loss": 0.5882,
+      "step": 523
+    },
+    {
+      "epoch": 0.2096,
+      "grad_norm": 0.4084893465042114,
+      "learning_rate": 6.42439189056735e-06,
+      "loss": 0.5958,
+      "step": 524
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.4008101224899292,
+      "learning_rate": 6.421898185602449e-06,
+      "loss": 0.5665,
+      "step": 525
+    },
+    {
+      "epoch": 0.2104,
+      "grad_norm": 0.48329246044158936,
+      "learning_rate": 6.419399576741251e-06,
+      "loss": 0.6898,
+      "step": 526
+    },
+    {
+      "epoch": 0.2108,
+      "grad_norm": 0.44100847840309143,
+      "learning_rate": 6.416896068177236e-06,
+      "loss": 0.6257,
+      "step": 527
+    },
+    {
+      "epoch": 0.2112,
+      "grad_norm": 0.388071209192276,
+      "learning_rate": 6.414387664112106e-06,
+      "loss": 0.5492,
+      "step": 528
+    },
+    {
+      "epoch": 0.2116,
+      "grad_norm": 0.39122116565704346,
+      "learning_rate": 6.4118743687557784e-06,
+      "loss": 0.5135,
+      "step": 529
+    },
+    {
+      "epoch": 0.212,
+      "grad_norm": 0.3872377276420593,
+      "learning_rate": 6.409356186326382e-06,
+      "loss": 0.5829,
+      "step": 530
+    },
+    {
+      "epoch": 0.2124,
+      "grad_norm": 0.43584463000297546,
+      "learning_rate": 6.406833121050248e-06,
+      "loss": 0.5483,
+      "step": 531
+    },
+    {
+      "epoch": 0.2128,
+      "grad_norm": 0.4248460829257965,
+      "learning_rate": 6.404305177161898e-06,
+      "loss": 0.5424,
+      "step": 532
+    },
+    {
+      "epoch": 0.2132,
+      "grad_norm": 0.4430016875267029,
+      "learning_rate": 6.401772358904047e-06,
+      "loss": 0.6306,
+      "step": 533
+    },
+    {
+      "epoch": 0.2136,
+      "grad_norm": 0.4296245276927948,
+      "learning_rate": 6.399234670527588e-06,
+      "loss": 0.5466,
+      "step": 534
+    },
+    {
+      "epoch": 0.214,
+      "grad_norm": 0.48698973655700684,
+      "learning_rate": 6.396692116291589e-06,
+      "loss": 0.593,
+      "step": 535
+    },
+    {
+      "epoch": 0.2144,
+      "grad_norm": 0.47233495116233826,
+      "learning_rate": 6.394144700463282e-06,
+      "loss": 0.6653,
+      "step": 536
+    },
+    {
+      "epoch": 0.2148,
+      "grad_norm": 0.45071518421173096,
+      "learning_rate": 6.391592427318061e-06,
+      "loss": 0.5932,
+      "step": 537
+    },
+    {
+      "epoch": 0.2152,
+      "grad_norm": 0.4426057040691376,
+      "learning_rate": 6.389035301139472e-06,
+      "loss": 0.5882,
+      "step": 538
+    },
+    {
+      "epoch": 0.2156,
+      "grad_norm": 0.4603918492794037,
+      "learning_rate": 6.386473326219203e-06,
+      "loss": 0.6082,
+      "step": 539
+    },
+    {
+      "epoch": 0.216,
+      "grad_norm": 0.4099114239215851,
+      "learning_rate": 6.3839065068570835e-06,
+      "loss": 0.588,
+      "step": 540
+    },
+    {
+      "epoch": 0.2164,
+      "grad_norm": 0.46051111817359924,
+      "learning_rate": 6.381334847361072e-06,
+      "loss": 0.6374,
+      "step": 541
+    },
+    {
+      "epoch": 0.2168,
+      "grad_norm": 0.4387073516845703,
+      "learning_rate": 6.378758352047249e-06,
+      "loss": 0.5764,
+      "step": 542
+    },
+    {
+      "epoch": 0.2172,
+      "grad_norm": 0.43461674451828003,
+      "learning_rate": 6.376177025239813e-06,
+      "loss": 0.5032,
+      "step": 543
+    },
+    {
+      "epoch": 0.2176,
+      "grad_norm": 0.4845770597457886,
+      "learning_rate": 6.373590871271072e-06,
+      "loss": 0.621,
+      "step": 544
+    },
+    {
+      "epoch": 0.218,
+      "grad_norm": 0.39095914363861084,
+      "learning_rate": 6.370999894481431e-06,
+      "loss": 0.4885,
+      "step": 545
+    },
+    {
+      "epoch": 0.2184,
+      "grad_norm": 0.39494702219963074,
+      "learning_rate": 6.3684040992193955e-06,
+      "loss": 0.5262,
+      "step": 546
+    },
+    {
+      "epoch": 0.2188,
+      "grad_norm": 0.4588661789894104,
+      "learning_rate": 6.365803489841554e-06,
+      "loss": 0.6244,
+      "step": 547
+    },
+    {
+      "epoch": 0.2192,
+      "grad_norm": 0.43723824620246887,
+      "learning_rate": 6.363198070712575e-06,
+      "loss": 0.633,
+      "step": 548
+    },
+    {
+      "epoch": 0.2196,
+      "grad_norm": 0.3847430944442749,
+      "learning_rate": 6.360587846205201e-06,
+      "loss": 0.4945,
+      "step": 549
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.41110801696777344,
+      "learning_rate": 6.357972820700238e-06,
+      "loss": 0.5271,
+      "step": 550
+    },
+    {
+      "epoch": 0.2204,
+      "grad_norm": 0.35954952239990234,
+      "learning_rate": 6.35535299858655e-06,
+      "loss": 0.525,
+      "step": 551
+    },
+    {
+      "epoch": 0.2208,
+      "grad_norm": 0.43042629957199097,
+      "learning_rate": 6.35272838426105e-06,
+      "loss": 0.5427,
+      "step": 552
+    },
+    {
+      "epoch": 0.2212,
+      "grad_norm": 0.3824063837528229,
+      "learning_rate": 6.350098982128698e-06,
+      "loss": 0.5121,
+      "step": 553
+    },
+    {
+      "epoch": 0.2216,
+      "grad_norm": 0.4746267795562744,
+      "learning_rate": 6.347464796602483e-06,
+      "loss": 0.6304,
+      "step": 554
+    },
+    {
+      "epoch": 0.222,
+      "grad_norm": 0.4227291941642761,
+      "learning_rate": 6.34482583210343e-06,
+      "loss": 0.5746,
+      "step": 555
+    },
+    {
+      "epoch": 0.2224,
+      "grad_norm": 0.4328877031803131,
+      "learning_rate": 6.3421820930605774e-06,
+      "loss": 0.5557,
+      "step": 556
+    },
+    {
+      "epoch": 0.2228,
+      "grad_norm": 0.37801414728164673,
+      "learning_rate": 6.33953358391098e-06,
+      "loss": 0.6082,
+      "step": 557
+    },
+    {
+      "epoch": 0.2232,
+      "grad_norm": 0.4196787178516388,
+      "learning_rate": 6.336880309099701e-06,
+      "loss": 0.5608,
+      "step": 558
+    },
+    {
+      "epoch": 0.2236,
+      "grad_norm": 0.4383528232574463,
+      "learning_rate": 6.334222273079796e-06,
+      "loss": 0.583,
+      "step": 559
+    },
+    {
+      "epoch": 0.224,
+      "grad_norm": 0.44536638259887695,
+      "learning_rate": 6.331559480312316e-06,
+      "loss": 0.6434,
+      "step": 560
+    },
+    {
+      "epoch": 0.2244,
+      "grad_norm": 0.40932145714759827,
+      "learning_rate": 6.328891935266295e-06,
+      "loss": 0.5477,
+      "step": 561
+    },
+    {
+      "epoch": 0.2248,
+      "grad_norm": 0.4584387242794037,
+      "learning_rate": 6.3262196424187405e-06,
+      "loss": 0.5553,
+      "step": 562
+    },
+    {
+      "epoch": 0.2252,
+      "grad_norm": 0.45622336864471436,
+      "learning_rate": 6.32354260625463e-06,
+      "loss": 0.6559,
+      "step": 563
+    },
+    {
+      "epoch": 0.2256,
+      "grad_norm": 0.4516793191432953,
+      "learning_rate": 6.3208608312669024e-06,
+      "loss": 0.5715,
+      "step": 564
+    },
+    {
+      "epoch": 0.226,
+      "grad_norm": 0.4983437955379486,
+      "learning_rate": 6.318174321956449e-06,
+      "loss": 0.5955,
+      "step": 565
+    },
+    {
+      "epoch": 0.2264,
+      "grad_norm": 0.42384248971939087,
+      "learning_rate": 6.315483082832107e-06,
+      "loss": 0.5761,
+      "step": 566
+    },
+    {
+      "epoch": 0.2268,
+      "grad_norm": 0.4657883942127228,
+      "learning_rate": 6.312787118410649e-06,
+      "loss": 0.5292,
+      "step": 567
+    },
+    {
+      "epoch": 0.2272,
+      "grad_norm": 0.3928786814212799,
+      "learning_rate": 6.3100864332167856e-06,
+      "loss": 0.5788,
+      "step": 568
+    },
+    {
+      "epoch": 0.2276,
+      "grad_norm": 0.44233494997024536,
+      "learning_rate": 6.307381031783142e-06,
+      "loss": 0.5317,
+      "step": 569
+    },
+    {
+      "epoch": 0.228,
+      "grad_norm": 0.43338441848754883,
+      "learning_rate": 6.304670918650265e-06,
+      "loss": 0.5289,
+      "step": 570
+    },
+    {
+      "epoch": 0.2284,
+      "grad_norm": 0.421744704246521,
+      "learning_rate": 6.301956098366605e-06,
+      "loss": 0.5815,
+      "step": 571
+    },
+    {
+      "epoch": 0.2288,
+      "grad_norm": 0.43918102979660034,
+      "learning_rate": 6.299236575488515e-06,
+      "loss": 0.6108,
+      "step": 572
+    },
+    {
+      "epoch": 0.2292,
+      "grad_norm": 0.4225776791572571,
+      "learning_rate": 6.296512354580237e-06,
+      "loss": 0.5535,
+      "step": 573
+    },
+    {
+      "epoch": 0.2296,
+      "grad_norm": 0.51739102602005,
+      "learning_rate": 6.293783440213905e-06,
+      "loss": 0.5432,
+      "step": 574
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.44613152742385864,
+      "learning_rate": 6.291049836969523e-06,
+      "loss": 0.5376,
+      "step": 575
+    },
+    {
+      "epoch": 0.2304,
+      "grad_norm": 0.45409390330314636,
+      "learning_rate": 6.2883115494349665e-06,
+      "loss": 0.6289,
+      "step": 576
+    },
+    {
+      "epoch": 0.2308,
+      "grad_norm": 0.4217747151851654,
+      "learning_rate": 6.285568582205975e-06,
+      "loss": 0.5261,
+      "step": 577
+    },
+    {
+      "epoch": 0.2312,
+      "grad_norm": 0.41630804538726807,
+      "learning_rate": 6.28282093988614e-06,
+      "loss": 0.5454,
+      "step": 578
+    },
+    {
+      "epoch": 0.2316,
+      "grad_norm": 0.38962697982788086,
+      "learning_rate": 6.280068627086899e-06,
+      "loss": 0.5176,
+      "step": 579
+    },
+    {
+      "epoch": 0.232,
+      "grad_norm": 0.4049234092235565,
+      "learning_rate": 6.277311648427529e-06,
+      "loss": 0.5302,
+      "step": 580
+    },
+    {
+      "epoch": 0.2324,
+      "grad_norm": 0.4065638482570648,
+      "learning_rate": 6.2745500085351384e-06,
+      "loss": 0.5437,
+      "step": 581
+    },
+    {
+      "epoch": 0.2328,
+      "grad_norm": 0.4093839228153229,
+      "learning_rate": 6.2717837120446565e-06,
+      "loss": 0.5592,
+      "step": 582
+    },
+    {
+      "epoch": 0.2332,
+      "grad_norm": 0.44915738701820374,
+      "learning_rate": 6.269012763598831e-06,
+      "loss": 0.6082,
+      "step": 583
+    },
+    {
+      "epoch": 0.2336,
+      "grad_norm": 0.3967624306678772,
+      "learning_rate": 6.266237167848215e-06,
+      "loss": 0.5106,
+      "step": 584
+    },
+    {
+      "epoch": 0.234,
+      "grad_norm": 0.4397246539592743,
+      "learning_rate": 6.26345692945116e-06,
+      "loss": 0.5652,
+      "step": 585
+    },
+    {
+      "epoch": 0.2344,
+      "grad_norm": 0.43744099140167236,
+      "learning_rate": 6.260672053073813e-06,
+      "loss": 0.5719,
+      "step": 586
+    },
+    {
+      "epoch": 0.2348,
+      "grad_norm": 0.4588030278682709,
+      "learning_rate": 6.2578825433901024e-06,
+      "loss": 0.5124,
+      "step": 587
+    },
+    {
+      "epoch": 0.2352,
+      "grad_norm": 0.3984925150871277,
+      "learning_rate": 6.255088405081733e-06,
+      "loss": 0.5531,
+      "step": 588
+    },
+    {
+      "epoch": 0.2356,
+      "grad_norm": 0.42685481905937195,
+      "learning_rate": 6.252289642838181e-06,
+      "loss": 0.5728,
+      "step": 589
+    },
+    {
+      "epoch": 0.236,
+      "grad_norm": 0.4059397280216217,
+      "learning_rate": 6.249486261356676e-06,
+      "loss": 0.513,
+      "step": 590
+    },
+    {
+      "epoch": 0.2364,
+      "grad_norm": 0.45304572582244873,
+      "learning_rate": 6.246678265342208e-06,
+      "loss": 0.5868,
+      "step": 591
+    },
+    {
+      "epoch": 0.2368,
+      "grad_norm": 0.47392749786376953,
+      "learning_rate": 6.243865659507508e-06,
+      "loss": 0.636,
+      "step": 592
+    },
+    {
+      "epoch": 0.2372,
+      "grad_norm": 0.42907243967056274,
+      "learning_rate": 6.241048448573044e-06,
+      "loss": 0.6657,
+      "step": 593
+    },
+    {
+      "epoch": 0.2376,
+      "grad_norm": 0.4455984830856323,
+      "learning_rate": 6.238226637267012e-06,
+      "loss": 0.5975,
+      "step": 594
+    },
+    {
+      "epoch": 0.238,
+      "grad_norm": 0.4236272871494293,
+      "learning_rate": 6.235400230325331e-06,
+      "loss": 0.591,
+      "step": 595
+    },
+    {
+      "epoch": 0.2384,
+      "grad_norm": 0.44643765687942505,
+      "learning_rate": 6.23256923249163e-06,
+      "loss": 0.5875,
+      "step": 596
+    },
+    {
+      "epoch": 0.2388,
+      "grad_norm": 0.3928106725215912,
+      "learning_rate": 6.229733648517248e-06,
+      "loss": 0.6052,
+      "step": 597
+    },
+    {
+      "epoch": 0.2392,
+      "grad_norm": 0.4079158902168274,
+      "learning_rate": 6.2268934831612164e-06,
+      "loss": 0.5558,
+      "step": 598
+    },
+    {
+      "epoch": 0.2396,
+      "grad_norm": 0.422274649143219,
+      "learning_rate": 6.224048741190257e-06,
+      "loss": 0.5032,
+      "step": 599
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.4376719892024994,
+      "learning_rate": 6.221199427378773e-06,
+      "loss": 0.5662,
+      "step": 600
+    },
+    {
+      "epoch": 0.2404,
+      "grad_norm": 0.47152188420295715,
+      "learning_rate": 6.218345546508841e-06,
+      "loss": 0.6223,
+      "step": 601
+    },
+    {
+      "epoch": 0.2408,
+      "grad_norm": 0.4223569631576538,
+      "learning_rate": 6.2154871033702e-06,
+      "loss": 0.5705,
+      "step": 602
+    },
+    {
+      "epoch": 0.2412,
+      "grad_norm": 0.4249934256076813,
+      "learning_rate": 6.21262410276025e-06,
+      "loss": 0.6251,
+      "step": 603
+    },
+    {
+      "epoch": 0.2416,
+      "grad_norm": 0.4933459460735321,
+      "learning_rate": 6.209756549484039e-06,
+      "loss": 0.5981,
+      "step": 604
+    },
+    {
+      "epoch": 0.242,
+      "grad_norm": 0.4324929416179657,
+      "learning_rate": 6.206884448354253e-06,
+      "loss": 0.6305,
+      "step": 605
+    },
+    {
+      "epoch": 0.2424,
+      "grad_norm": 0.40548714995384216,
+      "learning_rate": 6.204007804191214e-06,
+      "loss": 0.5925,
+      "step": 606
+    },
+    {
+      "epoch": 0.2428,
+      "grad_norm": 0.4025042951107025,
+      "learning_rate": 6.201126621822866e-06,
+      "loss": 0.6275,
+      "step": 607
+    },
+    {
+      "epoch": 0.2432,
+      "grad_norm": 0.4443594217300415,
+      "learning_rate": 6.19824090608477e-06,
+      "loss": 0.5989,
+      "step": 608
+    },
+    {
+      "epoch": 0.2436,
+      "grad_norm": 0.4185742437839508,
+      "learning_rate": 6.1953506618201e-06,
+      "loss": 0.5703,
+      "step": 609
+    },
+    {
+      "epoch": 0.244,
+      "grad_norm": 0.4236041009426117,
+      "learning_rate": 6.192455893879624e-06,
+      "loss": 0.509,
+      "step": 610
+    },
+    {
+      "epoch": 0.2444,
+      "grad_norm": 0.4389362037181854,
+      "learning_rate": 6.189556607121704e-06,
+      "loss": 0.5048,
+      "step": 611
+    },
+    {
+      "epoch": 0.2448,
+      "grad_norm": 0.4090619683265686,
+      "learning_rate": 6.18665280641229e-06,
+      "loss": 0.5497,
+      "step": 612
+    },
+    {
+      "epoch": 0.2452,
+      "grad_norm": 0.4614490568637848,
+      "learning_rate": 6.183744496624901e-06,
+      "loss": 0.6688,
+      "step": 613
+    },
+    {
+      "epoch": 0.2456,
+      "grad_norm": 0.4077858030796051,
+      "learning_rate": 6.180831682640632e-06,
+      "loss": 0.4744,
+      "step": 614
+    },
+    {
+      "epoch": 0.246,
+      "grad_norm": 0.42312952876091003,
+      "learning_rate": 6.177914369348129e-06,
+      "loss": 0.5286,
+      "step": 615
+    },
+    {
+      "epoch": 0.2464,
+      "grad_norm": 0.4584427773952484,
+      "learning_rate": 6.174992561643597e-06,
+      "loss": 0.576,
+      "step": 616
+    },
+    {
+      "epoch": 0.2468,
+      "grad_norm": 0.4387916624546051,
+      "learning_rate": 6.172066264430778e-06,
+      "loss": 0.6186,
+      "step": 617
+    },
+    {
+      "epoch": 0.2472,
+      "grad_norm": 0.44402211904525757,
+      "learning_rate": 6.169135482620951e-06,
+      "loss": 0.4917,
+      "step": 618
+    },
+    {
+      "epoch": 0.2476,
+      "grad_norm": 0.4333488941192627,
+      "learning_rate": 6.166200221132923e-06,
+      "loss": 0.5895,
+      "step": 619
+    },
+    {
+      "epoch": 0.248,
+      "grad_norm": 0.4146369695663452,
+      "learning_rate": 6.1632604848930185e-06,
+      "loss": 0.5022,
+      "step": 620
+    },
+    {
+      "epoch": 0.2484,
+      "grad_norm": 0.3817729353904724,
+      "learning_rate": 6.160316278835071e-06,
+      "loss": 0.4959,
+      "step": 621
+    },
+    {
+      "epoch": 0.2488,
+      "grad_norm": 0.3966714143753052,
+      "learning_rate": 6.1573676079004185e-06,
+      "loss": 0.5352,
+      "step": 622
+    },
+    {
+      "epoch": 0.2492,
+      "grad_norm": 0.45970264077186584,
+      "learning_rate": 6.154414477037888e-06,
+      "loss": 0.5645,
+      "step": 623
+    },
+    {
+      "epoch": 0.2496,
+      "grad_norm": 0.48507970571517944,
+      "learning_rate": 6.151456891203796e-06,
+      "loss": 0.6115,
+      "step": 624
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.41141727566719055,
+      "learning_rate": 6.148494855361933e-06,
+      "loss": 0.5251,
+      "step": 625
+    },
+    {
+      "epoch": 0.2504,
+      "grad_norm": 0.41636160016059875,
+      "learning_rate": 6.1455283744835615e-06,
+      "loss": 0.6228,
+      "step": 626
+    },
+    {
+      "epoch": 0.2508,
+      "grad_norm": 0.417447030544281,
+      "learning_rate": 6.1425574535474e-06,
+      "loss": 0.5013,
+      "step": 627
+    },
+    {
+      "epoch": 0.2512,
+      "grad_norm": 0.45456668734550476,
+      "learning_rate": 6.139582097539622e-06,
+      "loss": 0.5358,
+      "step": 628
+    },
+    {
+      "epoch": 0.2516,
+      "grad_norm": 0.44917646050453186,
+      "learning_rate": 6.136602311453844e-06,
+      "loss": 0.5993,
+      "step": 629
+    },
+    {
+      "epoch": 0.252,
+      "grad_norm": 0.4428352415561676,
+      "learning_rate": 6.133618100291116e-06,
+      "loss": 0.6076,
+      "step": 630
+    },
+    {
+      "epoch": 0.2524,
+      "grad_norm": 0.4192942678928375,
+      "learning_rate": 6.130629469059915e-06,
+      "loss": 0.5457,
+      "step": 631
+    },
+    {
+      "epoch": 0.2528,
+      "grad_norm": 0.43102753162384033,
+      "learning_rate": 6.1276364227761394e-06,
+      "loss": 0.59,
+      "step": 632
+    },
+    {
+      "epoch": 0.2532,
+      "grad_norm": 0.44437599182128906,
+      "learning_rate": 6.124638966463093e-06,
+      "loss": 0.5485,
+      "step": 633
+    },
+    {
+      "epoch": 0.2536,
+      "grad_norm": 0.44910112023353577,
+      "learning_rate": 6.1216371051514844e-06,
+      "loss": 0.5859,
+      "step": 634
+    },
+    {
+      "epoch": 0.254,
+      "grad_norm": 0.4694381058216095,
+      "learning_rate": 6.118630843879414e-06,
+      "loss": 0.5947,
+      "step": 635
+    },
+    {
+      "epoch": 0.2544,
+      "grad_norm": 0.45699435472488403,
+      "learning_rate": 6.1156201876923664e-06,
+      "loss": 0.568,
+      "step": 636
+    },
+    {
+      "epoch": 0.2548,
+      "grad_norm": 0.48551276326179504,
+      "learning_rate": 6.1126051416432026e-06,
+      "loss": 0.6698,
+      "step": 637
+    },
+    {
+      "epoch": 0.2552,
+      "grad_norm": 0.4661720395088196,
+      "learning_rate": 6.109585710792152e-06,
+      "loss": 0.6635,
+      "step": 638
+    },
+    {
+      "epoch": 0.2556,
+      "grad_norm": 0.42143744230270386,
+      "learning_rate": 6.106561900206802e-06,
+      "loss": 0.6606,
+      "step": 639
+    },
+    {
+      "epoch": 0.256,
+      "grad_norm": 0.38233682513237,
+      "learning_rate": 6.103533714962091e-06,
+      "loss": 0.5294,
+      "step": 640
+    },
+    {
+      "epoch": 0.2564,
+      "grad_norm": 0.40119943022727966,
+      "learning_rate": 6.1005011601402986e-06,
+      "loss": 0.5762,
+      "step": 641
+    },
+    {
+      "epoch": 0.2568,
+      "grad_norm": 0.45029154419898987,
+      "learning_rate": 6.097464240831041e-06,
+      "loss": 0.579,
+      "step": 642
+    },
+    {
+      "epoch": 0.2572,
+      "grad_norm": 0.4179871678352356,
+      "learning_rate": 6.094422962131257e-06,
+      "loss": 0.4944,
+      "step": 643
+    },
+    {
+      "epoch": 0.2576,
+      "grad_norm": 0.4345018267631531,
+      "learning_rate": 6.091377329145202e-06,
+      "loss": 0.6728,
+      "step": 644
+    },
+    {
+      "epoch": 0.258,
+      "grad_norm": 0.4406702518463135,
+      "learning_rate": 6.0883273469844375e-06,
+      "loss": 0.6143,
+      "step": 645
+    },
+    {
+      "epoch": 0.2584,
+      "grad_norm": 0.3850463330745697,
+      "learning_rate": 6.085273020767829e-06,
+      "loss": 0.5452,
+      "step": 646
+    },
+    {
+      "epoch": 0.2588,
+      "grad_norm": 0.4438881278038025,
+      "learning_rate": 6.08221435562153e-06,
+      "loss": 0.5663,
+      "step": 647
+    },
+    {
+      "epoch": 0.2592,
+      "grad_norm": 0.4332631230354309,
+      "learning_rate": 6.079151356678974e-06,
+      "loss": 0.595,
+      "step": 648
+    },
+    {
+      "epoch": 0.2596,
+      "grad_norm": 0.4043017327785492,
+      "learning_rate": 6.076084029080874e-06,
+      "loss": 0.5761,
+      "step": 649
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.4187481701374054,
+      "learning_rate": 6.073012377975201e-06,
+      "loss": 0.6399,
+      "step": 650
+    },
+    {
+      "epoch": 0.2604,
+      "grad_norm": 0.44606634974479675,
+      "learning_rate": 6.069936408517187e-06,
+      "loss": 0.5939,
+      "step": 651
+    },
+    {
+      "epoch": 0.2608,
+      "grad_norm": 0.4353955388069153,
+      "learning_rate": 6.066856125869309e-06,
+      "loss": 0.5386,
+      "step": 652
+    },
+    {
+      "epoch": 0.2612,
+      "grad_norm": 0.44587963819503784,
+      "learning_rate": 6.063771535201285e-06,
+      "loss": 0.5598,
+      "step": 653
+    },
+    {
+      "epoch": 0.2616,
+      "grad_norm": 0.3983965814113617,
+      "learning_rate": 6.060682641690062e-06,
+      "loss": 0.5234,
+      "step": 654
+    },
+    {
+      "epoch": 0.262,
+      "grad_norm": 0.3644501268863678,
+      "learning_rate": 6.057589450519807e-06,
+      "loss": 0.4878,
+      "step": 655
+    },
+    {
+      "epoch": 0.2624,
+      "grad_norm": 0.489908903837204,
+      "learning_rate": 6.054491966881905e-06,
+      "loss": 0.5665,
+      "step": 656
+    },
+    {
+      "epoch": 0.2628,
+      "grad_norm": 0.4127205014228821,
+      "learning_rate": 6.05139019597494e-06,
+      "loss": 0.5241,
+      "step": 657
+    },
+    {
+      "epoch": 0.2632,
+      "grad_norm": 0.4351557195186615,
+      "learning_rate": 6.048284143004693e-06,
+      "loss": 0.6357,
+      "step": 658
+    },
+    {
+      "epoch": 0.2636,
+      "grad_norm": 0.4123672544956207,
+      "learning_rate": 6.045173813184132e-06,
+      "loss": 0.5769,
+      "step": 659
+    },
+    {
+      "epoch": 0.264,
+      "grad_norm": 0.4411662220954895,
+      "learning_rate": 6.042059211733405e-06,
+      "loss": 0.5479,
+      "step": 660
+    },
+    {
+      "epoch": 0.2644,
+      "grad_norm": 0.4334865212440491,
+      "learning_rate": 6.038940343879824e-06,
+      "loss": 0.5502,
+      "step": 661
+    },
+    {
+      "epoch": 0.2648,
+      "grad_norm": 0.46878618001937866,
+      "learning_rate": 6.035817214857866e-06,
+      "loss": 0.6801,
+      "step": 662
+    },
+    {
+      "epoch": 0.2652,
+      "grad_norm": 0.4362941086292267,
+      "learning_rate": 6.032689829909158e-06,
+      "loss": 0.5267,
+      "step": 663
+    },
+    {
+      "epoch": 0.2656,
+      "grad_norm": 0.40274283289909363,
+      "learning_rate": 6.0295581942824715e-06,
+      "loss": 0.4743,
+      "step": 664
+    },
+    {
+      "epoch": 0.266,
+      "grad_norm": 0.42851364612579346,
+      "learning_rate": 6.026422313233708e-06,
+      "loss": 0.4994,
+      "step": 665
+    },
+    {
+      "epoch": 0.2664,
+      "grad_norm": 0.4585568904876709,
+      "learning_rate": 6.023282192025897e-06,
+      "loss": 0.5265,
+      "step": 666
+    },
+    {
+      "epoch": 0.2668,
+      "grad_norm": 0.41542091965675354,
+      "learning_rate": 6.020137835929185e-06,
+      "loss": 0.4668,
+      "step": 667
+    },
+    {
+      "epoch": 0.2672,
+      "grad_norm": 0.501136839389801,
+      "learning_rate": 6.016989250220826e-06,
+      "loss": 0.6468,
+      "step": 668
+    },
+    {
+      "epoch": 0.2676,
+      "grad_norm": 0.5080068707466125,
+      "learning_rate": 6.013836440185169e-06,
+      "loss": 0.5774,
+      "step": 669
+    },
+    {
+      "epoch": 0.268,
+      "grad_norm": 0.3732645511627197,
+      "learning_rate": 6.010679411113658e-06,
+      "loss": 0.4073,
+      "step": 670
+    },
+    {
+      "epoch": 0.2684,
+      "grad_norm": 0.4972980320453644,
+      "learning_rate": 6.007518168304815e-06,
+      "loss": 0.6443,
+      "step": 671
+    },
+    {
+      "epoch": 0.2688,
+      "grad_norm": 0.43165460228919983,
+      "learning_rate": 6.004352717064234e-06,
+      "loss": 0.5778,
+      "step": 672
+    },
+    {
+      "epoch": 0.2692,
+      "grad_norm": 0.4288862943649292,
+      "learning_rate": 6.001183062704573e-06,
+      "loss": 0.58,
+      "step": 673
+    },
+    {
+      "epoch": 0.2696,
+      "grad_norm": 0.4247485101222992,
+      "learning_rate": 5.998009210545544e-06,
+      "loss": 0.4869,
+      "step": 674
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.49416670203208923,
+      "learning_rate": 5.994831165913903e-06,
+      "loss": 0.6321,
+      "step": 675
+    },
+    {
+      "epoch": 0.2704,
+      "grad_norm": 0.4453592896461487,
+      "learning_rate": 5.991648934143443e-06,
+      "loss": 0.5078,
+      "step": 676
+    },
+    {
+      "epoch": 0.2708,
+      "grad_norm": 0.424652099609375,
+      "learning_rate": 5.988462520574987e-06,
+      "loss": 0.4775,
+      "step": 677
+    },
+    {
+      "epoch": 0.2712,
+      "grad_norm": 0.4466131925582886,
+      "learning_rate": 5.985271930556373e-06,
+      "loss": 0.5983,
+      "step": 678
+    },
+    {
+      "epoch": 0.2716,
+      "grad_norm": 0.4205477237701416,
+      "learning_rate": 5.98207716944245e-06,
+      "loss": 0.5108,
+      "step": 679
+    },
+    {
+      "epoch": 0.272,
+      "grad_norm": 0.40754249691963196,
+      "learning_rate": 5.978878242595065e-06,
+      "loss": 0.5275,
+      "step": 680
+    },
+    {
+      "epoch": 0.2724,
+      "grad_norm": 0.4342345893383026,
+      "learning_rate": 5.975675155383063e-06,
+      "loss": 0.5025,
+      "step": 681
+    },
+    {
+      "epoch": 0.2728,
+      "grad_norm": 0.4328685402870178,
+      "learning_rate": 5.972467913182263e-06,
+      "loss": 0.5089,
+      "step": 682
+    },
+    {
+      "epoch": 0.2732,
+      "grad_norm": 0.3855776786804199,
+      "learning_rate": 5.969256521375463e-06,
+      "loss": 0.4926,
+      "step": 683
+    },
+    {
+      "epoch": 0.2736,
+      "grad_norm": 0.4109788239002228,
+      "learning_rate": 5.966040985352424e-06,
+      "loss": 0.5406,
+      "step": 684
+    },
+    {
+      "epoch": 0.274,
+      "grad_norm": 0.3982260823249817,
+      "learning_rate": 5.962821310509862e-06,
+      "loss": 0.438,
+      "step": 685
+    },
+    {
+      "epoch": 0.2744,
+      "grad_norm": 0.3830687701702118,
+      "learning_rate": 5.9595975022514385e-06,
+      "loss": 0.4475,
+      "step": 686
+    },
+    {
+      "epoch": 0.2748,
+      "grad_norm": 0.40385526418685913,
+      "learning_rate": 5.9563695659877555e-06,
+      "loss": 0.4952,
+      "step": 687
+    },
+    {
+      "epoch": 0.2752,
+      "grad_norm": 0.4147714674472809,
+      "learning_rate": 5.953137507136338e-06,
+      "loss": 0.513,
+      "step": 688
+    },
+    {
+      "epoch": 0.2756,
+      "grad_norm": 0.4347432553768158,
+      "learning_rate": 5.949901331121633e-06,
+      "loss": 0.5845,
+      "step": 689
+    },
+    {
+      "epoch": 0.276,
+      "grad_norm": 0.45100805163383484,
+      "learning_rate": 5.946661043375001e-06,
+      "loss": 0.5552,
+      "step": 690
+    },
+    {
+      "epoch": 0.2764,
+      "grad_norm": 0.4410780668258667,
+      "learning_rate": 5.943416649334695e-06,
+      "loss": 0.6428,
+      "step": 691
+    },
+    {
+      "epoch": 0.2768,
+      "grad_norm": 0.43005675077438354,
+      "learning_rate": 5.940168154445869e-06,
+      "loss": 0.5303,
+      "step": 692
+    },
+    {
+      "epoch": 0.2772,
+      "grad_norm": 0.48011666536331177,
+      "learning_rate": 5.93691556416055e-06,
+      "loss": 0.6382,
+      "step": 693
+    },
+    {
+      "epoch": 0.2776,
+      "grad_norm": 0.40826767683029175,
+      "learning_rate": 5.9336588839376465e-06,
+      "loss": 0.5135,
+      "step": 694
+    },
+    {
+      "epoch": 0.278,
+      "grad_norm": 0.44731247425079346,
+      "learning_rate": 5.930398119242927e-06,
+      "loss": 0.7138,
+      "step": 695
+    },
+    {
+      "epoch": 0.2784,
+      "grad_norm": 0.4752798080444336,
+      "learning_rate": 5.9271332755490155e-06,
+      "loss": 0.6109,
+      "step": 696
+    },
+    {
+      "epoch": 0.2788,
+      "grad_norm": 0.5010511875152588,
+      "learning_rate": 5.923864358335384e-06,
+      "loss": 0.5772,
+      "step": 697
+    },
+    {
+      "epoch": 0.2792,
+      "grad_norm": 0.5007715225219727,
+      "learning_rate": 5.920591373088338e-06,
+      "loss": 0.6622,
+      "step": 698
+    },
+    {
+      "epoch": 0.2796,
+      "grad_norm": 0.39461269974708557,
+      "learning_rate": 5.917314325301015e-06,
+      "loss": 0.5549,
+      "step": 699
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.4943528473377228,
+      "learning_rate": 5.9140332204733655e-06,
+      "loss": 0.6357,
+      "step": 700
+    },
+    {
+      "epoch": 0.2804,
+      "grad_norm": 0.45546188950538635,
+      "learning_rate": 5.910748064112151e-06,
+      "loss": 0.5509,
+      "step": 701
+    },
+    {
+      "epoch": 0.2808,
+      "grad_norm": 0.4762568771839142,
+      "learning_rate": 5.907458861730934e-06,
+      "loss": 0.5395,
+      "step": 702
+    },
+    {
+      "epoch": 0.2812,
+      "grad_norm": 0.4343515932559967,
+      "learning_rate": 5.904165618850068e-06,
+      "loss": 0.5163,
+      "step": 703
+    },
+    {
+      "epoch": 0.2816,
+      "grad_norm": 0.4203025996685028,
+      "learning_rate": 5.900868340996685e-06,
+      "loss": 0.4834,
+      "step": 704
+    },
+    {
+      "epoch": 0.282,
+      "grad_norm": 0.4241425096988678,
+      "learning_rate": 5.897567033704691e-06,
+      "loss": 0.5301,
+      "step": 705
+    },
+    {
+      "epoch": 0.2824,
+      "grad_norm": 0.38161882758140564,
+      "learning_rate": 5.894261702514755e-06,
+      "loss": 0.5134,
+      "step": 706
+    },
+    {
+      "epoch": 0.2828,
+      "grad_norm": 0.45214715600013733,
+      "learning_rate": 5.890952352974299e-06,
+      "loss": 0.5846,
+      "step": 707
+    },
+    {
+      "epoch": 0.2832,
+      "grad_norm": 0.42457085847854614,
+      "learning_rate": 5.887638990637486e-06,
+      "loss": 0.5388,
+      "step": 708
+    },
+    {
+      "epoch": 0.2836,
+      "grad_norm": 0.41017526388168335,
+      "learning_rate": 5.884321621065221e-06,
+      "loss": 0.4365,
+      "step": 709
+    },
+    {
+      "epoch": 0.284,
+      "grad_norm": 0.4165228009223938,
+      "learning_rate": 5.881000249825125e-06,
+      "loss": 0.5469,
+      "step": 710
+    },
+    {
+      "epoch": 0.2844,
+      "grad_norm": 0.4457581341266632,
+      "learning_rate": 5.877674882491543e-06,
+      "loss": 0.5077,
+      "step": 711
+    },
+    {
+      "epoch": 0.2848,
+      "grad_norm": 0.4252013564109802,
+      "learning_rate": 5.874345524645525e-06,
+      "loss": 0.5792,
+      "step": 712
+    },
+    {
+      "epoch": 0.2852,
+      "grad_norm": 0.46879255771636963,
+      "learning_rate": 5.871012181874815e-06,
+      "loss": 0.6151,
+      "step": 713
+    },
+    {
+      "epoch": 0.2856,
+      "grad_norm": 0.45289963483810425,
+      "learning_rate": 5.867674859773849e-06,
+      "loss": 0.5513,
+      "step": 714
+    },
+    {
+      "epoch": 0.286,
+      "grad_norm": 0.4870239198207855,
+      "learning_rate": 5.864333563943737e-06,
+      "loss": 0.6212,
+      "step": 715
+    },
+    {
+      "epoch": 0.2864,
+      "grad_norm": 0.4319377839565277,
+      "learning_rate": 5.860988299992264e-06,
+      "loss": 0.5148,
+      "step": 716
+    },
+    {
+      "epoch": 0.2868,
+      "grad_norm": 0.46285220980644226,
+      "learning_rate": 5.857639073533871e-06,
+      "loss": 0.6536,
+      "step": 717
+    },
+    {
+      "epoch": 0.2872,
+      "grad_norm": 0.43863755464553833,
+      "learning_rate": 5.85428589018965e-06,
+      "loss": 0.6057,
+      "step": 718
+    },
+    {
+      "epoch": 0.2876,
+      "grad_norm": 0.39260825514793396,
+      "learning_rate": 5.850928755587334e-06,
+      "loss": 0.5483,
+      "step": 719
+    },
+    {
+      "epoch": 0.288,
+      "grad_norm": 0.43337124586105347,
+      "learning_rate": 5.847567675361289e-06,
+      "loss": 0.5708,
+      "step": 720
+    },
+    {
+      "epoch": 0.2884,
+      "grad_norm": 0.4293830096721649,
+      "learning_rate": 5.8442026551525e-06,
+      "loss": 0.5817,
+      "step": 721
+    },
+    {
+      "epoch": 0.2888,
+      "grad_norm": 0.4857645034790039,
+      "learning_rate": 5.840833700608566e-06,
+      "loss": 0.6579,
+      "step": 722
+    },
+    {
+      "epoch": 0.2892,
+      "grad_norm": 0.43444594740867615,
+      "learning_rate": 5.837460817383691e-06,
+      "loss": 0.5676,
+      "step": 723
+    },
+    {
+      "epoch": 0.2896,
+      "grad_norm": 0.4489664137363434,
+      "learning_rate": 5.83408401113867e-06,
+      "loss": 0.5545,
+      "step": 724
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.4444904029369354,
+      "learning_rate": 5.8307032875408836e-06,
+      "loss": 0.571,
+      "step": 725
+    },
+    {
+      "epoch": 0.2904,
+      "grad_norm": 0.393856942653656,
+      "learning_rate": 5.8273186522642866e-06,
+      "loss": 0.4803,
+      "step": 726
+    },
+    {
+      "epoch": 0.2908,
+      "grad_norm": 0.4597555994987488,
+      "learning_rate": 5.8239301109893984e-06,
+      "loss": 0.6037,
+      "step": 727
+    },
+    {
+      "epoch": 0.2912,
+      "grad_norm": 0.4321127235889435,
+      "learning_rate": 5.820537669403295e-06,
+      "loss": 0.5774,
+      "step": 728
+    },
+    {
+      "epoch": 0.2916,
+      "grad_norm": 0.4166051149368286,
+      "learning_rate": 5.817141333199596e-06,
+      "loss": 0.5717,
+      "step": 729
+    },
+    {
+      "epoch": 0.292,
+      "grad_norm": 0.436637818813324,
+      "learning_rate": 5.813741108078462e-06,
+      "loss": 0.5879,
+      "step": 730
+    },
+    {
+      "epoch": 0.2924,
+      "grad_norm": 0.43837761878967285,
+      "learning_rate": 5.810336999746573e-06,
+      "loss": 0.5178,
+      "step": 731
+    },
+    {
+      "epoch": 0.2928,
+      "grad_norm": 0.47148534655570984,
+      "learning_rate": 5.806929013917136e-06,
+      "loss": 0.5488,
+      "step": 732
+    },
+    {
+      "epoch": 0.2932,
+      "grad_norm": 0.39476650953292847,
+      "learning_rate": 5.803517156309857e-06,
+      "loss": 0.431,
+      "step": 733
+    },
+    {
+      "epoch": 0.2936,
+      "grad_norm": 0.4546128511428833,
+      "learning_rate": 5.800101432650946e-06,
+      "loss": 0.5669,
+      "step": 734
+    },
+    {
+      "epoch": 0.294,
+      "grad_norm": 0.42975708842277527,
+      "learning_rate": 5.796681848673098e-06,
+      "loss": 0.5621,
+      "step": 735
+    },
+    {
+      "epoch": 0.2944,
+      "grad_norm": 0.46449723839759827,
+      "learning_rate": 5.7932584101154885e-06,
+      "loss": 0.6004,
+      "step": 736
+    },
+    {
+      "epoch": 0.2948,
+      "grad_norm": 0.42449694871902466,
+      "learning_rate": 5.789831122723761e-06,
+      "loss": 0.5197,
+      "step": 737
+    },
+    {
+      "epoch": 0.2952,
+      "grad_norm": 0.39055266976356506,
+      "learning_rate": 5.78639999225002e-06,
+      "loss": 0.5367,
+      "step": 738
+    },
+    {
+      "epoch": 0.2956,
+      "grad_norm": 0.4630034565925598,
+      "learning_rate": 5.78296502445282e-06,
+      "loss": 0.5253,
+      "step": 739
+    },
+    {
+      "epoch": 0.296,
+      "grad_norm": 0.4190136790275574,
+      "learning_rate": 5.779526225097153e-06,
+      "loss": 0.4948,
+      "step": 740
+    },
+    {
+      "epoch": 0.2964,
+      "grad_norm": 0.4667150676250458,
+      "learning_rate": 5.776083599954447e-06,
+      "loss": 0.5638,
+      "step": 741
+    },
+    {
+      "epoch": 0.2968,
+      "grad_norm": 0.45995426177978516,
+      "learning_rate": 5.7726371548025446e-06,
+      "loss": 0.5279,
+      "step": 742
+    },
+    {
+      "epoch": 0.2972,
+      "grad_norm": 0.401014119386673,
+      "learning_rate": 5.769186895425704e-06,
+      "loss": 0.5173,
+      "step": 743
+    },
+    {
+      "epoch": 0.2976,
+      "grad_norm": 0.45050883293151855,
+      "learning_rate": 5.7657328276145845e-06,
+      "loss": 0.576,
+      "step": 744
+    },
+    {
+      "epoch": 0.298,
+      "grad_norm": 0.48369142413139343,
+      "learning_rate": 5.762274957166234e-06,
+      "loss": 0.5987,
+      "step": 745
+    },
+    {
+      "epoch": 0.2984,
+      "grad_norm": 0.5284405946731567,
+      "learning_rate": 5.758813289884086e-06,
+      "loss": 0.5244,
+      "step": 746
+    },
+    {
+      "epoch": 0.2988,
+      "grad_norm": 0.4794248342514038,
+      "learning_rate": 5.755347831577945e-06,
+      "loss": 0.5985,
+      "step": 747
+    },
+    {
+      "epoch": 0.2992,
+      "grad_norm": 0.47695332765579224,
+      "learning_rate": 5.751878588063979e-06,
+      "loss": 0.5536,
+      "step": 748
+    },
+    {
+      "epoch": 0.2996,
+      "grad_norm": 0.435024619102478,
+      "learning_rate": 5.748405565164705e-06,
+      "loss": 0.5141,
+      "step": 749
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.4625321924686432,
+      "learning_rate": 5.7449287687089895e-06,
+      "loss": 0.4985,
+      "step": 750
+    },
+    {
+      "epoch": 0.3004,
+      "grad_norm": 0.4212191700935364,
+      "learning_rate": 5.741448204532028e-06,
+      "loss": 0.4764,
+      "step": 751
+    },
+    {
+      "epoch": 0.3008,
+      "grad_norm": 0.48766613006591797,
+      "learning_rate": 5.7379638784753375e-06,
+      "loss": 0.5614,
+      "step": 752
+    },
+    {
+      "epoch": 0.3012,
+      "grad_norm": 0.5481346249580383,
+      "learning_rate": 5.734475796386754e-06,
+      "loss": 0.6491,
+      "step": 753
+    },
+    {
+      "epoch": 0.3016,
+      "grad_norm": 0.40259304642677307,
+      "learning_rate": 5.730983964120414e-06,
+      "loss": 0.5542,
+      "step": 754
+    },
+    {
+      "epoch": 0.302,
+      "grad_norm": 0.45115047693252563,
+      "learning_rate": 5.727488387536749e-06,
+      "loss": 0.5763,
+      "step": 755
+    },
+    {
+      "epoch": 0.3024,
+      "grad_norm": 0.4455792009830475,
+      "learning_rate": 5.723989072502473e-06,
+      "loss": 0.5814,
+      "step": 756
+    },
+    {
+      "epoch": 0.3028,
+      "grad_norm": 0.422356516122818,
+      "learning_rate": 5.720486024890578e-06,
+      "loss": 0.5086,
+      "step": 757
+    },
+    {
+      "epoch": 0.3032,
+      "grad_norm": 0.4979648292064667,
+      "learning_rate": 5.716979250580316e-06,
+      "loss": 0.5321,
+      "step": 758
+    },
+    {
+      "epoch": 0.3036,
+      "grad_norm": 0.46708184480667114,
+      "learning_rate": 5.713468755457198e-06,
+      "loss": 0.5119,
+      "step": 759
+    },
+    {
+      "epoch": 0.304,
+      "grad_norm": 0.41758787631988525,
+      "learning_rate": 5.709954545412975e-06,
+      "loss": 0.6061,
+      "step": 760
+    },
+    {
+      "epoch": 0.3044,
+      "grad_norm": 0.44567832350730896,
+      "learning_rate": 5.706436626345636e-06,
+      "loss": 0.48,
+      "step": 761
+    },
+    {
+      "epoch": 0.3048,
+      "grad_norm": 0.4240472912788391,
+      "learning_rate": 5.702915004159397e-06,
+      "loss": 0.6254,
+      "step": 762
+    },
+    {
+      "epoch": 0.3052,
+      "grad_norm": 0.4301627278327942,
+      "learning_rate": 5.699389684764685e-06,
+      "loss": 0.5424,
+      "step": 763
+    },
+    {
+      "epoch": 0.3056,
+      "grad_norm": 0.47439152002334595,
+      "learning_rate": 5.6958606740781306e-06,
+      "loss": 0.5885,
+      "step": 764
+    },
+    {
+      "epoch": 0.306,
+      "grad_norm": 0.4316118359565735,
+      "learning_rate": 5.692327978022567e-06,
+      "loss": 0.4796,
+      "step": 765
+    },
+    {
+      "epoch": 0.3064,
+      "grad_norm": 0.4615771472454071,
+      "learning_rate": 5.688791602527005e-06,
+      "loss": 0.6382,
+      "step": 766
+    },
+    {
+      "epoch": 0.3068,
+      "grad_norm": 0.44186869263648987,
+      "learning_rate": 5.6852515535266355e-06,
+      "loss": 0.5634,
+      "step": 767
+    },
+    {
+      "epoch": 0.3072,
+      "grad_norm": 0.46103107929229736,
+      "learning_rate": 5.681707836962812e-06,
+      "loss": 0.576,
+      "step": 768
+    },
+    {
+      "epoch": 0.3076,
+      "grad_norm": 0.47277137637138367,
+      "learning_rate": 5.678160458783045e-06,
+      "loss": 0.5493,
+      "step": 769
+    },
+    {
+      "epoch": 0.308,
+      "grad_norm": 0.49479013681411743,
+      "learning_rate": 5.67460942494099e-06,
+      "loss": 0.5541,
+      "step": 770
+    },
+    {
+      "epoch": 0.3084,
+      "grad_norm": 0.46736854314804077,
+      "learning_rate": 5.6710547413964375e-06,
+      "loss": 0.615,
+      "step": 771
+    },
+    {
+      "epoch": 0.3088,
+      "grad_norm": 0.45999255776405334,
+      "learning_rate": 5.667496414115304e-06,
+      "loss": 0.5378,
+      "step": 772
+    },
+    {
+      "epoch": 0.3092,
+      "grad_norm": 0.4547343850135803,
+      "learning_rate": 5.663934449069619e-06,
+      "loss": 0.5264,
+      "step": 773
+    },
+    {
+      "epoch": 0.3096,
+      "grad_norm": 0.45823800563812256,
+      "learning_rate": 5.6603688522375234e-06,
+      "loss": 0.5644,
+      "step": 774
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.4509381949901581,
+      "learning_rate": 5.656799629603246e-06,
+      "loss": 0.5604,
+      "step": 775
+    },
+    {
+      "epoch": 0.3104,
+      "grad_norm": 0.44573745131492615,
+      "learning_rate": 5.653226787157104e-06,
+      "loss": 0.4764,
+      "step": 776
+    },
+    {
+      "epoch": 0.3108,
+      "grad_norm": 0.4186471402645111,
+      "learning_rate": 5.649650330895492e-06,
+      "loss": 0.5625,
+      "step": 777
+    },
+    {
+      "epoch": 0.3112,
+      "grad_norm": 0.39781829714775085,
+      "learning_rate": 5.646070266820868e-06,
+      "loss": 0.4717,
+      "step": 778
+    },
+    {
+      "epoch": 0.3116,
+      "grad_norm": 0.4090104103088379,
+      "learning_rate": 5.6424866009417425e-06,
+      "loss": 0.547,
+      "step": 779
+    },
+    {
+      "epoch": 0.312,
+      "grad_norm": 0.49576470255851746,
+      "learning_rate": 5.638899339272676e-06,
+      "loss": 0.6078,
+      "step": 780
+    },
+    {
+      "epoch": 0.3124,
+      "grad_norm": 0.4280230402946472,
+      "learning_rate": 5.63530848783426e-06,
+      "loss": 0.521,
+      "step": 781
+    },
+    {
+      "epoch": 0.3128,
+      "grad_norm": 0.4742203950881958,
+      "learning_rate": 5.631714052653111e-06,
+      "loss": 0.5345,
+      "step": 782
+    },
+    {
+      "epoch": 0.3132,
+      "grad_norm": 0.49490198493003845,
+      "learning_rate": 5.628116039761864e-06,
+      "loss": 0.6582,
+      "step": 783
+    },
+    {
+      "epoch": 0.3136,
+      "grad_norm": 0.45912107825279236,
+      "learning_rate": 5.624514455199154e-06,
+      "loss": 0.582,
+      "step": 784
+    },
+    {
+      "epoch": 0.314,
+      "grad_norm": 0.45559161901474,
+      "learning_rate": 5.620909305009612e-06,
+      "loss": 0.5585,
+      "step": 785
+    },
+    {
+      "epoch": 0.3144,
+      "grad_norm": 0.46950197219848633,
+      "learning_rate": 5.617300595243855e-06,
+      "loss": 0.5311,
+      "step": 786
+    },
+    {
+      "epoch": 0.3148,
+      "grad_norm": 0.45106202363967896,
+      "learning_rate": 5.613688331958472e-06,
+      "loss": 0.5609,
+      "step": 787
+    },
+    {
+      "epoch": 0.3152,
+      "grad_norm": 0.4136814773082733,
+      "learning_rate": 5.610072521216017e-06,
+      "loss": 0.5501,
+      "step": 788
+    },
+    {
+      "epoch": 0.3156,
+      "grad_norm": 0.4540536403656006,
+      "learning_rate": 5.606453169084997e-06,
+      "loss": 0.5868,
+      "step": 789
+    },
+    {
+      "epoch": 0.316,
+      "grad_norm": 0.42658835649490356,
+      "learning_rate": 5.602830281639862e-06,
+      "loss": 0.5929,
+      "step": 790
+    },
+    {
+      "epoch": 0.3164,
+      "grad_norm": 0.445243775844574,
+      "learning_rate": 5.599203864961e-06,
+      "loss": 0.6617,
+      "step": 791
+    },
+    {
+      "epoch": 0.3168,
+      "grad_norm": 0.5022698044776917,
+      "learning_rate": 5.595573925134714e-06,
+      "loss": 0.5943,
+      "step": 792
+    },
+    {
+      "epoch": 0.3172,
+      "grad_norm": 0.47502660751342773,
+      "learning_rate": 5.591940468253228e-06,
+      "loss": 0.54,
+      "step": 793
+    },
+    {
+      "epoch": 0.3176,
+      "grad_norm": 0.41046321392059326,
+      "learning_rate": 5.588303500414663e-06,
+      "loss": 0.5199,
+      "step": 794
+    },
+    {
+      "epoch": 0.318,
+      "grad_norm": 0.4040413796901703,
+      "learning_rate": 5.584663027723038e-06,
+      "loss": 0.5715,
+      "step": 795
+    },
+    {
+      "epoch": 0.3184,
+      "grad_norm": 0.48011571168899536,
+      "learning_rate": 5.5810190562882505e-06,
+      "loss": 0.5669,
+      "step": 796
+    },
+    {
+      "epoch": 0.3188,
+      "grad_norm": 0.448673278093338,
+      "learning_rate": 5.577371592226067e-06,
+      "loss": 0.5856,
+      "step": 797
+    },
+    {
+      "epoch": 0.3192,
+      "grad_norm": 0.41867223381996155,
+      "learning_rate": 5.573720641658124e-06,
+      "loss": 0.523,
+      "step": 798
+    },
+    {
+      "epoch": 0.3196,
+      "grad_norm": 0.44206374883651733,
+      "learning_rate": 5.5700662107119035e-06,
+      "loss": 0.5252,
+      "step": 799
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.4805357754230499,
+      "learning_rate": 5.566408305520729e-06,
+      "loss": 0.5956,
+      "step": 800
+    },
+    {
+      "epoch": 0.3204,
+      "grad_norm": 0.42322584986686707,
+      "learning_rate": 5.562746932223757e-06,
+      "loss": 0.5521,
+      "step": 801
+    },
+    {
+      "epoch": 0.3208,
+      "grad_norm": 0.4758487641811371,
+      "learning_rate": 5.559082096965966e-06,
+      "loss": 0.5746,
+      "step": 802
+    },
+    {
+      "epoch": 0.3212,
+      "grad_norm": 0.4096662700176239,
+      "learning_rate": 5.555413805898139e-06,
+      "loss": 0.4617,
+      "step": 803
+    },
+    {
+      "epoch": 0.3216,
+      "grad_norm": 0.40047913789749146,
+      "learning_rate": 5.551742065176864e-06,
+      "loss": 0.4953,
+      "step": 804
+    },
+    {
+      "epoch": 0.322,
+      "grad_norm": 0.4644544720649719,
+      "learning_rate": 5.5480668809645185e-06,
+      "loss": 0.6105,
+      "step": 805
+    },
+    {
+      "epoch": 0.3224,
+      "grad_norm": 0.41639772057533264,
+      "learning_rate": 5.544388259429255e-06,
+      "loss": 0.5031,
+      "step": 806
+    },
+    {
+      "epoch": 0.3228,
+      "grad_norm": 0.48089903593063354,
+      "learning_rate": 5.540706206745e-06,
+      "loss": 0.6029,
+      "step": 807
+    },
+    {
+      "epoch": 0.3232,
+      "grad_norm": 0.4112328588962555,
+      "learning_rate": 5.537020729091436e-06,
+      "loss": 0.5244,
+      "step": 808
+    },
+    {
+      "epoch": 0.3236,
+      "grad_norm": 0.4498112201690674,
+      "learning_rate": 5.533331832653995e-06,
+      "loss": 0.4764,
+      "step": 809
+    },
+    {
+      "epoch": 0.324,
+      "grad_norm": 0.44204649329185486,
+      "learning_rate": 5.529639523623845e-06,
+      "loss": 0.654,
+      "step": 810
+    },
+    {
+      "epoch": 0.3244,
+      "grad_norm": 0.45008304715156555,
+      "learning_rate": 5.5259438081978826e-06,
+      "loss": 0.5443,
+      "step": 811
+    },
+    {
+      "epoch": 0.3248,
+      "grad_norm": 0.4112232029438019,
+      "learning_rate": 5.522244692578722e-06,
+      "loss": 0.5507,
+      "step": 812
+    },
+    {
+      "epoch": 0.3252,
+      "grad_norm": 0.45125049352645874,
+      "learning_rate": 5.518542182974681e-06,
+      "loss": 0.4941,
+      "step": 813
+    },
+    {
+      "epoch": 0.3256,
+      "grad_norm": 0.48122072219848633,
+      "learning_rate": 5.514836285599779e-06,
+      "loss": 0.5857,
+      "step": 814
+    },
+    {
+      "epoch": 0.326,
+      "grad_norm": 0.43177640438079834,
+      "learning_rate": 5.511127006673717e-06,
+      "loss": 0.5111,
+      "step": 815
+    },
+    {
+      "epoch": 0.3264,
+      "grad_norm": 0.37912872433662415,
+      "learning_rate": 5.507414352421872e-06,
+      "loss": 0.5635,
+      "step": 816
+    },
+    {
+      "epoch": 0.3268,
+      "grad_norm": 0.37735727429389954,
+      "learning_rate": 5.503698329075288e-06,
+      "loss": 0.4695,
+      "step": 817
+    },
+    {
+      "epoch": 0.3272,
+      "grad_norm": 0.43305477499961853,
+      "learning_rate": 5.499978942870659e-06,
+      "loss": 0.6414,
+      "step": 818
+    },
+    {
+      "epoch": 0.3276,
+      "grad_norm": 0.4628191888332367,
+      "learning_rate": 5.496256200050329e-06,
+      "loss": 0.5046,
+      "step": 819
+    },
+    {
+      "epoch": 0.328,
+      "grad_norm": 0.4635592997074127,
+      "learning_rate": 5.4925301068622695e-06,
+      "loss": 0.5762,
+      "step": 820
+    },
+    {
+      "epoch": 0.3284,
+      "grad_norm": 0.40314820408821106,
+      "learning_rate": 5.488800669560079e-06,
+      "loss": 0.4439,
+      "step": 821
+    },
+    {
+      "epoch": 0.3288,
+      "grad_norm": 0.508816659450531,
+      "learning_rate": 5.485067894402968e-06,
+      "loss": 0.5022,
+      "step": 822
+    },
+    {
+      "epoch": 0.3292,
+      "grad_norm": 0.4493391811847687,
+      "learning_rate": 5.481331787655747e-06,
+      "loss": 0.5457,
+      "step": 823
+    },
+    {
+      "epoch": 0.3296,
+      "grad_norm": 0.45542025566101074,
+      "learning_rate": 5.477592355588822e-06,
+      "loss": 0.5874,
+      "step": 824
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.4431161880493164,
+      "learning_rate": 5.473849604478173e-06,
+      "loss": 0.5893,
+      "step": 825
+    },
+    {
+      "epoch": 0.3304,
+      "grad_norm": 0.4549485146999359,
+      "learning_rate": 5.470103540605358e-06,
+      "loss": 0.5934,
+      "step": 826
+    },
+    {
+      "epoch": 0.3308,
+      "grad_norm": 0.4524597227573395,
+      "learning_rate": 5.466354170257489e-06,
+      "loss": 0.6158,
+      "step": 827
+    },
+    {
+      "epoch": 0.3312,
+      "grad_norm": 0.46560999751091003,
+      "learning_rate": 5.462601499727233e-06,
+      "loss": 0.5772,
+      "step": 828
+    },
+    {
+      "epoch": 0.3316,
+      "grad_norm": 0.4526055157184601,
+      "learning_rate": 5.4588455353127905e-06,
+      "loss": 0.54,
+      "step": 829
+    },
+    {
+      "epoch": 0.332,
+      "grad_norm": 0.4451562762260437,
+      "learning_rate": 5.455086283317893e-06,
+      "loss": 0.5364,
+      "step": 830
+    },
+    {
+      "epoch": 0.3324,
+      "grad_norm": 0.4619428217411041,
+      "learning_rate": 5.45132375005179e-06,
+      "loss": 0.5536,
+      "step": 831
+    },
+    {
+      "epoch": 0.3328,
+      "grad_norm": 0.45558473467826843,
+      "learning_rate": 5.447557941829236e-06,
+      "loss": 0.5492,
+      "step": 832
+    },
+    {
+      "epoch": 0.3332,
+      "grad_norm": 0.4863041341304779,
+      "learning_rate": 5.443788864970483e-06,
+      "loss": 0.5926,
+      "step": 833
+    },
+    {
+      "epoch": 0.3336,
+      "grad_norm": 0.4615713953971863,
+      "learning_rate": 5.440016525801269e-06,
+      "loss": 0.6397,
+      "step": 834
+    },
+    {
+      "epoch": 0.334,
+      "grad_norm": 0.47101691365242004,
+      "learning_rate": 5.436240930652807e-06,
+      "loss": 0.6057,
+      "step": 835
+    },
+    {
+      "epoch": 0.3344,
+      "grad_norm": 0.43971094489097595,
+      "learning_rate": 5.432462085861777e-06,
+      "loss": 0.4855,
+      "step": 836
+    },
+    {
+      "epoch": 0.3348,
+      "grad_norm": 0.45537781715393066,
+      "learning_rate": 5.428679997770307e-06,
+      "loss": 0.4801,
+      "step": 837
+    },
+    {
+      "epoch": 0.3352,
+      "grad_norm": 0.4680633544921875,
+      "learning_rate": 5.424894672725974e-06,
+      "loss": 0.5117,
+      "step": 838
+    },
+    {
+      "epoch": 0.3356,
+      "grad_norm": 0.4822414219379425,
+      "learning_rate": 5.421106117081785e-06,
+      "loss": 0.6574,
+      "step": 839
+    },
+    {
+      "epoch": 0.336,
+      "grad_norm": 0.464046448469162,
+      "learning_rate": 5.417314337196171e-06,
+      "loss": 0.6024,
+      "step": 840
+    },
+    {
+      "epoch": 0.3364,
+      "grad_norm": 0.47728002071380615,
+      "learning_rate": 5.41351933943297e-06,
+      "loss": 0.5942,
+      "step": 841
+    },
+    {
+      "epoch": 0.3368,
+      "grad_norm": 0.5082221627235413,
+      "learning_rate": 5.409721130161428e-06,
+      "loss": 0.5226,
+      "step": 842
+    },
+    {
+      "epoch": 0.3372,
+      "grad_norm": 0.5022159218788147,
+      "learning_rate": 5.4059197157561715e-06,
+      "loss": 0.5086,
+      "step": 843
+    },
+    {
+      "epoch": 0.3376,
+      "grad_norm": 0.4637428820133209,
+      "learning_rate": 5.402115102597215e-06,
+      "loss": 0.4954,
+      "step": 844
+    },
+    {
+      "epoch": 0.338,
+      "grad_norm": 0.4835737943649292,
+      "learning_rate": 5.398307297069937e-06,
+      "loss": 0.5729,
+      "step": 845
+    },
+    {
+      "epoch": 0.3384,
+      "grad_norm": 0.4542867839336395,
+      "learning_rate": 5.394496305565074e-06,
+      "loss": 0.5587,
+      "step": 846
+    },
+    {
+      "epoch": 0.3388,
+      "grad_norm": 0.43878334760665894,
+      "learning_rate": 5.3906821344787105e-06,
+      "loss": 0.5253,
+      "step": 847
+    },
+    {
+      "epoch": 0.3392,
+      "grad_norm": 0.48306146264076233,
+      "learning_rate": 5.386864790212268e-06,
+      "loss": 0.5476,
+      "step": 848
+    },
+    {
+      "epoch": 0.3396,
+      "grad_norm": 0.455120325088501,
+      "learning_rate": 5.383044279172491e-06,
+      "loss": 0.5114,
+      "step": 849
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.448162317276001,
+      "learning_rate": 5.379220607771444e-06,
+      "loss": 0.5854,
+      "step": 850
+    },
+    {
+      "epoch": 0.3404,
+      "grad_norm": 0.4229871928691864,
+      "learning_rate": 5.375393782426488e-06,
+      "loss": 0.4675,
+      "step": 851
+    },
+    {
+      "epoch": 0.3408,
+      "grad_norm": 0.4494837522506714,
+      "learning_rate": 5.371563809560285e-06,
+      "loss": 0.5747,
+      "step": 852
+    },
+    {
+      "epoch": 0.3412,
+      "grad_norm": 0.4818879961967468,
+      "learning_rate": 5.367730695600774e-06,
+      "loss": 0.5542,
+      "step": 853
+    },
+    {
+      "epoch": 0.3416,
+      "grad_norm": 0.4855602979660034,
+      "learning_rate": 5.363894446981171e-06,
+      "loss": 0.6312,
+      "step": 854
+    },
+    {
+      "epoch": 0.342,
+      "grad_norm": 0.3857225477695465,
+      "learning_rate": 5.360055070139946e-06,
+      "loss": 0.4853,
+      "step": 855
+    },
+    {
+      "epoch": 0.3424,
+      "grad_norm": 0.4498901963233948,
+      "learning_rate": 5.356212571520827e-06,
+      "loss": 0.5814,
+      "step": 856
+    },
+    {
+      "epoch": 0.3428,
+      "grad_norm": 0.3908587098121643,
+      "learning_rate": 5.352366957572774e-06,
+      "loss": 0.4917,
+      "step": 857
+    },
+    {
+      "epoch": 0.3432,
+      "grad_norm": 0.433999627828598,
+      "learning_rate": 5.34851823474998e-06,
+      "loss": 0.6064,
+      "step": 858
+    },
+    {
+      "epoch": 0.3436,
+      "grad_norm": 0.44410914182662964,
+      "learning_rate": 5.344666409511857e-06,
+      "loss": 0.5617,
+      "step": 859
+    },
+    {
+      "epoch": 0.344,
+      "grad_norm": 0.49029380083084106,
+      "learning_rate": 5.340811488323019e-06,
+      "loss": 0.5801,
+      "step": 860
+    },
+    {
+      "epoch": 0.3444,
+      "grad_norm": 0.4502889811992645,
+      "learning_rate": 5.336953477653281e-06,
+      "loss": 0.5024,
+      "step": 861
+    },
+    {
+      "epoch": 0.3448,
+      "grad_norm": 0.4806410074234009,
+      "learning_rate": 5.333092383977638e-06,
+      "loss": 0.6077,
+      "step": 862
+    },
+    {
+      "epoch": 0.3452,
+      "grad_norm": 0.4842955470085144,
+      "learning_rate": 5.329228213776264e-06,
+      "loss": 0.537,
+      "step": 863
+    },
+    {
+      "epoch": 0.3456,
+      "grad_norm": 0.47725391387939453,
+      "learning_rate": 5.325360973534495e-06,
+      "loss": 0.6212,
+      "step": 864
+    },
+    {
+      "epoch": 0.346,
+      "grad_norm": 0.40789616107940674,
+      "learning_rate": 5.321490669742815e-06,
+      "loss": 0.5103,
+      "step": 865
+    },
+    {
+      "epoch": 0.3464,
+      "grad_norm": 0.41891586780548096,
+      "learning_rate": 5.317617308896859e-06,
+      "loss": 0.5349,
+      "step": 866
+    },
+    {
+      "epoch": 0.3468,
+      "grad_norm": 0.4041453003883362,
+      "learning_rate": 5.313740897497384e-06,
+      "loss": 0.4992,
+      "step": 867
+    },
+    {
+      "epoch": 0.3472,
+      "grad_norm": 0.4476974308490753,
+      "learning_rate": 5.309861442050272e-06,
+      "loss": 0.5829,
+      "step": 868
+    },
+    {
+      "epoch": 0.3476,
+      "grad_norm": 0.45309120416641235,
+      "learning_rate": 5.305978949066509e-06,
+      "loss": 0.5601,
+      "step": 869
+    },
+    {
+      "epoch": 0.348,
+      "grad_norm": 0.46513673663139343,
+      "learning_rate": 5.302093425062187e-06,
+      "loss": 0.613,
+      "step": 870
+    },
+    {
+      "epoch": 0.3484,
+      "grad_norm": 0.4027946889400482,
+      "learning_rate": 5.298204876558476e-06,
+      "loss": 0.554,
+      "step": 871
+    },
+    {
+      "epoch": 0.3488,
+      "grad_norm": 0.45852765440940857,
+      "learning_rate": 5.294313310081627e-06,
+      "loss": 0.6269,
+      "step": 872
+    },
+    {
+      "epoch": 0.3492,
+      "grad_norm": 0.4091279208660126,
+      "learning_rate": 5.290418732162957e-06,
+      "loss": 0.565,
+      "step": 873
+    },
+    {
+      "epoch": 0.3496,
+      "grad_norm": 0.4555352032184601,
+      "learning_rate": 5.286521149338833e-06,
+      "loss": 0.5722,
+      "step": 874
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.4212196469306946,
+      "learning_rate": 5.28262056815067e-06,
+      "loss": 0.4969,
+      "step": 875
+    },
+    {
+      "epoch": 0.3504,
+      "grad_norm": 0.4575810730457306,
+      "learning_rate": 5.278716995144912e-06,
+      "loss": 0.5113,
+      "step": 876
+    },
+    {
+      "epoch": 0.3508,
+      "grad_norm": 0.4176642894744873,
+      "learning_rate": 5.274810436873025e-06,
+      "loss": 0.5705,
+      "step": 877
+    },
+    {
+      "epoch": 0.3512,
+      "grad_norm": 0.5067360401153564,
+      "learning_rate": 5.270900899891487e-06,
+      "loss": 0.6339,
+      "step": 878
+    },
+    {
+      "epoch": 0.3516,
+      "grad_norm": 0.46837419271469116,
+      "learning_rate": 5.266988390761771e-06,
+      "loss": 0.4302,
+      "step": 879
+    },
+    {
+      "epoch": 0.352,
+      "grad_norm": 0.4802364408969879,
+      "learning_rate": 5.263072916050343e-06,
+      "loss": 0.5163,
+      "step": 880
+    },
+    {
+      "epoch": 0.3524,
+      "grad_norm": 0.45025691390037537,
+      "learning_rate": 5.259154482328642e-06,
+      "loss": 0.5707,
+      "step": 881
+    },
+    {
+      "epoch": 0.3528,
+      "grad_norm": 0.43300339579582214,
+      "learning_rate": 5.255233096173078e-06,
+      "loss": 0.6326,
+      "step": 882
+    },
+    {
+      "epoch": 0.3532,
+      "grad_norm": 0.47875091433525085,
+      "learning_rate": 5.251308764165012e-06,
+      "loss": 0.5576,
+      "step": 883
+    },
+    {
+      "epoch": 0.3536,
+      "grad_norm": 0.42456284165382385,
+      "learning_rate": 5.247381492890752e-06,
+      "loss": 0.4854,
+      "step": 884
+    },
+    {
+      "epoch": 0.354,
+      "grad_norm": 0.39123857021331787,
+      "learning_rate": 5.243451288941536e-06,
+      "loss": 0.542,
+      "step": 885
+    },
+    {
+      "epoch": 0.3544,
+      "grad_norm": 0.4416203200817108,
+      "learning_rate": 5.239518158913525e-06,
+      "loss": 0.5864,
+      "step": 886
+    },
+    {
+      "epoch": 0.3548,
+      "grad_norm": 0.4715004861354828,
+      "learning_rate": 5.2355821094077935e-06,
+      "loss": 0.6147,
+      "step": 887
+    },
+    {
+      "epoch": 0.3552,
+      "grad_norm": 0.4258134663105011,
+      "learning_rate": 5.231643147030312e-06,
+      "loss": 0.6116,
+      "step": 888
+    },
+    {
+      "epoch": 0.3556,
+      "grad_norm": 0.4103778004646301,
+      "learning_rate": 5.227701278391942e-06,
+      "loss": 0.6069,
+      "step": 889
+    },
+    {
+      "epoch": 0.356,
+      "grad_norm": 0.4473305940628052,
+      "learning_rate": 5.223756510108424e-06,
+      "loss": 0.6424,
+      "step": 890
+    },
+    {
+      "epoch": 0.3564,
+      "grad_norm": 0.4191577136516571,
+      "learning_rate": 5.219808848800361e-06,
+      "loss": 0.559,
+      "step": 891
+    },
+    {
+      "epoch": 0.3568,
+      "grad_norm": 0.46516153216362,
+      "learning_rate": 5.215858301093216e-06,
+      "loss": 0.5549,
+      "step": 892
+    },
+    {
+      "epoch": 0.3572,
+      "grad_norm": 0.4698564410209656,
+      "learning_rate": 5.211904873617292e-06,
+      "loss": 0.5695,
+      "step": 893
+    },
+    {
+      "epoch": 0.3576,
+      "grad_norm": 0.4017867147922516,
+      "learning_rate": 5.207948573007728e-06,
+      "loss": 0.5333,
+      "step": 894
+    },
+    {
+      "epoch": 0.358,
+      "grad_norm": 0.4583630859851837,
+      "learning_rate": 5.203989405904483e-06,
+      "loss": 0.6255,
+      "step": 895
+    },
+    {
+      "epoch": 0.3584,
+      "grad_norm": 0.44605836272239685,
+      "learning_rate": 5.200027378952332e-06,
+      "loss": 0.4985,
+      "step": 896
+    },
+    {
+      "epoch": 0.3588,
+      "grad_norm": 0.4361209571361542,
+      "learning_rate": 5.196062498800842e-06,
+      "loss": 0.6041,
+      "step": 897
+    },
+    {
+      "epoch": 0.3592,
+      "grad_norm": 0.4455268681049347,
+      "learning_rate": 5.192094772104373e-06,
+      "loss": 0.5389,
+      "step": 898
+    },
+    {
+      "epoch": 0.3596,
+      "grad_norm": 0.4654677212238312,
+      "learning_rate": 5.188124205522065e-06,
+      "loss": 0.5827,
+      "step": 899
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.4446803629398346,
+      "learning_rate": 5.184150805717818e-06,
+      "loss": 0.5612,
+      "step": 900
+    },
+    {
+      "epoch": 0.3604,
+      "grad_norm": 0.4743603467941284,
+      "learning_rate": 5.180174579360292e-06,
+      "loss": 0.5681,
+      "step": 901
+    },
+    {
+      "epoch": 0.3608,
+      "grad_norm": 0.45481276512145996,
+      "learning_rate": 5.176195533122887e-06,
+      "loss": 0.5277,
+      "step": 902
+    },
+    {
+      "epoch": 0.3612,
+      "grad_norm": 0.4798752963542938,
+      "learning_rate": 5.172213673683741e-06,
+      "loss": 0.5995,
+      "step": 903
+    },
+    {
+      "epoch": 0.3616,
+      "grad_norm": 0.44253095984458923,
+      "learning_rate": 5.168229007725709e-06,
+      "loss": 0.5572,
+      "step": 904
+    },
+    {
+      "epoch": 0.362,
+      "grad_norm": 0.5103812217712402,
+      "learning_rate": 5.164241541936356e-06,
+      "loss": 0.5673,
+      "step": 905
+    },
+    {
+      "epoch": 0.3624,
+      "grad_norm": 0.45087578892707825,
+      "learning_rate": 5.16025128300795e-06,
+      "loss": 0.4734,
+      "step": 906
+    },
+    {
+      "epoch": 0.3628,
+      "grad_norm": 0.534394383430481,
+      "learning_rate": 5.156258237637444e-06,
+      "loss": 0.5476,
+      "step": 907
+    },
+    {
+      "epoch": 0.3632,
+      "grad_norm": 0.5126708745956421,
+      "learning_rate": 5.152262412526467e-06,
+      "loss": 0.6094,
+      "step": 908
+    },
+    {
+      "epoch": 0.3636,
+      "grad_norm": 0.41151338815689087,
+      "learning_rate": 5.148263814381314e-06,
+      "loss": 0.5243,
+      "step": 909
+    },
+    {
+      "epoch": 0.364,
+      "grad_norm": 0.5132277011871338,
+      "learning_rate": 5.144262449912935e-06,
+      "loss": 0.6284,
+      "step": 910
+    },
+    {
+      "epoch": 0.3644,
+      "grad_norm": 0.4442318379878998,
+      "learning_rate": 5.140258325836923e-06,
+      "loss": 0.4923,
+      "step": 911
+    },
+    {
+      "epoch": 0.3648,
+      "grad_norm": 0.4278882145881653,
+      "learning_rate": 5.1362514488735e-06,
+      "loss": 0.5837,
+      "step": 912
+    },
+    {
+      "epoch": 0.3652,
+      "grad_norm": 0.4440222382545471,
+      "learning_rate": 5.13224182574751e-06,
+      "loss": 0.5436,
+      "step": 913
+    },
+    {
+      "epoch": 0.3656,
+      "grad_norm": 0.4995422661304474,
+      "learning_rate": 5.128229463188406e-06,
+      "loss": 0.5801,
+      "step": 914
+    },
+    {
+      "epoch": 0.366,
+      "grad_norm": 0.4683188796043396,
+      "learning_rate": 5.12421436793024e-06,
+      "loss": 0.6342,
+      "step": 915
+    },
+    {
+      "epoch": 0.3664,
+      "grad_norm": 0.4733324348926544,
+      "learning_rate": 5.120196546711647e-06,
+      "loss": 0.5931,
+      "step": 916
+    },
+    {
+      "epoch": 0.3668,
+      "grad_norm": 0.4991692304611206,
+      "learning_rate": 5.1161760062758395e-06,
+      "loss": 0.6111,
+      "step": 917
+    },
+    {
+      "epoch": 0.3672,
+      "grad_norm": 0.4712377190589905,
+      "learning_rate": 5.112152753370594e-06,
+      "loss": 0.5567,
+      "step": 918
+    },
+    {
+      "epoch": 0.3676,
+      "grad_norm": 0.47813770174980164,
+      "learning_rate": 5.108126794748237e-06,
+      "loss": 0.5108,
+      "step": 919
+    },
+    {
+      "epoch": 0.368,
+      "grad_norm": 0.4006287157535553,
+      "learning_rate": 5.1040981371656384e-06,
+      "loss": 0.4912,
+      "step": 920
+    },
+    {
+      "epoch": 0.3684,
+      "grad_norm": 0.4339583218097687,
+      "learning_rate": 5.100066787384198e-06,
+      "loss": 0.46,
+      "step": 921
+    },
+    {
+      "epoch": 0.3688,
+      "grad_norm": 0.4966609477996826,
+      "learning_rate": 5.096032752169831e-06,
+      "loss": 0.5807,
+      "step": 922
+    },
+    {
+      "epoch": 0.3692,
+      "grad_norm": 0.43865084648132324,
+      "learning_rate": 5.091996038292962e-06,
+      "loss": 0.5623,
+      "step": 923
+    },
+    {
+      "epoch": 0.3696,
+      "grad_norm": 0.4705183506011963,
+      "learning_rate": 5.087956652528513e-06,
+      "loss": 0.5841,
+      "step": 924
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.4557614326477051,
+      "learning_rate": 5.083914601655887e-06,
+      "loss": 0.5623,
+      "step": 925
+    },
+    {
+      "epoch": 0.3704,
+      "grad_norm": 0.45096200704574585,
+      "learning_rate": 5.079869892458959e-06,
+      "loss": 0.629,
+      "step": 926
+    },
+    {
+      "epoch": 0.3708,
+      "grad_norm": 0.48638662695884705,
+      "learning_rate": 5.075822531726071e-06,
+      "loss": 0.5955,
+      "step": 927
+    },
+    {
+      "epoch": 0.3712,
+      "grad_norm": 0.5118933320045471,
+      "learning_rate": 5.071772526250009e-06,
+      "loss": 0.593,
+      "step": 928
+    },
+    {
+      "epoch": 0.3716,
+      "grad_norm": 0.498941570520401,
+      "learning_rate": 5.067719882828004e-06,
+      "loss": 0.4636,
+      "step": 929
+    },
+    {
+      "epoch": 0.372,
+      "grad_norm": 0.4572960138320923,
+      "learning_rate": 5.0636646082617085e-06,
+      "loss": 0.5629,
+      "step": 930
+    },
+    {
+      "epoch": 0.3724,
+      "grad_norm": 0.5548087358474731,
+      "learning_rate": 5.059606709357192e-06,
+      "loss": 0.5777,
+      "step": 931
+    },
+    {
+      "epoch": 0.3728,
+      "grad_norm": 0.46791890263557434,
+      "learning_rate": 5.0555461929249345e-06,
+      "loss": 0.56,
+      "step": 932
+    },
+    {
+      "epoch": 0.3732,
+      "grad_norm": 0.5351516008377075,
+      "learning_rate": 5.0514830657797986e-06,
+      "loss": 0.6033,
+      "step": 933
+    },
+    {
+      "epoch": 0.3736,
+      "grad_norm": 0.46923014521598816,
+      "learning_rate": 5.047417334741038e-06,
+      "loss": 0.4895,
+      "step": 934
+    },
+    {
+      "epoch": 0.374,
+      "grad_norm": 0.4047207534313202,
+      "learning_rate": 5.043349006632271e-06,
+      "loss": 0.5233,
+      "step": 935
+    },
+    {
+      "epoch": 0.3744,
+      "grad_norm": 0.428609699010849,
+      "learning_rate": 5.039278088281479e-06,
+      "loss": 0.509,
+      "step": 936
+    },
+    {
+      "epoch": 0.3748,
+      "grad_norm": 0.4471479654312134,
+      "learning_rate": 5.035204586520985e-06,
+      "loss": 0.4526,
+      "step": 937
+    },
+    {
+      "epoch": 0.3752,
+      "grad_norm": 0.4439519941806793,
+      "learning_rate": 5.031128508187452e-06,
+      "loss": 0.5774,
+      "step": 938
+    },
+    {
+      "epoch": 0.3756,
+      "grad_norm": 0.47322627902030945,
+      "learning_rate": 5.027049860121869e-06,
+      "loss": 0.574,
+      "step": 939
+    },
+    {
+      "epoch": 0.376,
+      "grad_norm": 0.4471387267112732,
+      "learning_rate": 5.022968649169531e-06,
+      "loss": 0.5038,
+      "step": 940
+    },
+    {
+      "epoch": 0.3764,
+      "grad_norm": 0.4897695481777191,
+      "learning_rate": 5.018884882180042e-06,
+      "loss": 0.6207,
+      "step": 941
+    },
+    {
+      "epoch": 0.3768,
+      "grad_norm": 0.40873169898986816,
+      "learning_rate": 5.0147985660072895e-06,
+      "loss": 0.4889,
+      "step": 942
+    },
+    {
+      "epoch": 0.3772,
+      "grad_norm": 0.3923221826553345,
+      "learning_rate": 5.0107097075094445e-06,
+      "loss": 0.5532,
+      "step": 943
+    },
+    {
+      "epoch": 0.3776,
+      "grad_norm": 0.4386279881000519,
+      "learning_rate": 5.00661831354894e-06,
+      "loss": 0.5483,
+      "step": 944
+    },
+    {
+      "epoch": 0.378,
+      "grad_norm": 0.4298570156097412,
+      "learning_rate": 5.002524390992471e-06,
+      "loss": 0.4718,
+      "step": 945
+    },
+    {
+      "epoch": 0.3784,
+      "grad_norm": 0.4775792360305786,
+      "learning_rate": 4.998427946710967e-06,
+      "loss": 0.6635,
+      "step": 946
+    },
+    {
+      "epoch": 0.3788,
+      "grad_norm": 0.43783479928970337,
+      "learning_rate": 4.9943289875796e-06,
+      "loss": 0.5541,
+      "step": 947
+    },
+    {
+      "epoch": 0.3792,
+      "grad_norm": 0.4827110171318054,
+      "learning_rate": 4.9902275204777545e-06,
+      "loss": 0.6063,
+      "step": 948
+    },
+    {
+      "epoch": 0.3796,
+      "grad_norm": 0.47438400983810425,
+      "learning_rate": 4.9861235522890274e-06,
+      "loss": 0.514,
+      "step": 949
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.485522598028183,
+      "learning_rate": 4.982017089901216e-06,
+      "loss": 0.586,
+      "step": 950
+    },
+    {
+      "epoch": 0.3804,
+      "grad_norm": 0.49923890829086304,
+      "learning_rate": 4.977908140206297e-06,
+      "loss": 0.6124,
+      "step": 951
+    },
+    {
+      "epoch": 0.3808,
+      "grad_norm": 0.46352657675743103,
+      "learning_rate": 4.97379671010043e-06,
+      "loss": 0.6389,
+      "step": 952
+    },
+    {
+      "epoch": 0.3812,
+      "grad_norm": 0.467952162027359,
+      "learning_rate": 4.969682806483929e-06,
+      "loss": 0.581,
+      "step": 953
+    },
+    {
+      "epoch": 0.3816,
+      "grad_norm": 0.5383191108703613,
+      "learning_rate": 4.965566436261266e-06,
+      "loss": 0.6133,
+      "step": 954
+    },
+    {
+      "epoch": 0.382,
+      "grad_norm": 0.4607367217540741,
+      "learning_rate": 4.961447606341048e-06,
+      "loss": 0.5202,
+      "step": 955
+    },
+    {
+      "epoch": 0.3824,
+      "grad_norm": 0.48336759209632874,
+      "learning_rate": 4.957326323636015e-06,
+      "loss": 0.5309,
+      "step": 956
+    },
+    {
+      "epoch": 0.3828,
+      "grad_norm": 0.44407516717910767,
+      "learning_rate": 4.95320259506302e-06,
+      "loss": 0.5205,
+      "step": 957
+    },
+    {
+      "epoch": 0.3832,
+      "grad_norm": 0.45355352759361267,
+      "learning_rate": 4.949076427543021e-06,
+      "loss": 0.616,
+      "step": 958
+    },
+    {
+      "epoch": 0.3836,
+      "grad_norm": 0.42455554008483887,
+      "learning_rate": 4.944947828001071e-06,
+      "loss": 0.5373,
+      "step": 959
+    },
+    {
+      "epoch": 0.384,
+      "grad_norm": 0.3933660089969635,
+      "learning_rate": 4.940816803366304e-06,
+      "loss": 0.4608,
+      "step": 960
+    },
+    {
+      "epoch": 0.3844,
+      "grad_norm": 0.47803884744644165,
+      "learning_rate": 4.936683360571924e-06,
+      "loss": 0.4946,
+      "step": 961
+    },
+    {
+      "epoch": 0.3848,
+      "grad_norm": 0.4601835310459137,
+      "learning_rate": 4.932547506555193e-06,
+      "loss": 0.5556,
+      "step": 962
+    },
+    {
+      "epoch": 0.3852,
+      "grad_norm": 0.5340824127197266,
+      "learning_rate": 4.92840924825742e-06,
+      "loss": 0.5202,
+      "step": 963
+    },
+    {
+      "epoch": 0.3856,
+      "grad_norm": 0.4656091630458832,
+      "learning_rate": 4.924268592623952e-06,
+      "loss": 0.509,
+      "step": 964
+    },
+    {
+      "epoch": 0.386,
+      "grad_norm": 0.4116588234901428,
+      "learning_rate": 4.9201255466041545e-06,
+      "loss": 0.4783,
+      "step": 965
+    },
+    {
+      "epoch": 0.3864,
+      "grad_norm": 0.5430503487586975,
+      "learning_rate": 4.915980117151407e-06,
+      "loss": 0.5982,
+      "step": 966
+    },
+    {
+      "epoch": 0.3868,
+      "grad_norm": 0.43445679545402527,
+      "learning_rate": 4.9118323112230925e-06,
+      "loss": 0.4827,
+      "step": 967
+    },
+    {
+      "epoch": 0.3872,
+      "grad_norm": 0.4689733386039734,
+      "learning_rate": 4.907682135780576e-06,
+      "loss": 0.6506,
+      "step": 968
+    },
+    {
+      "epoch": 0.3876,
+      "grad_norm": 0.4890734553337097,
+      "learning_rate": 4.903529597789206e-06,
+      "loss": 0.5741,
+      "step": 969
+    },
+    {
+      "epoch": 0.388,
+      "grad_norm": 0.4545091688632965,
+      "learning_rate": 4.899374704218291e-06,
+      "loss": 0.5148,
+      "step": 970
+    },
+    {
+      "epoch": 0.3884,
+      "grad_norm": 0.5078072547912598,
+      "learning_rate": 4.8952174620410965e-06,
+      "loss": 0.6291,
+      "step": 971
+    },
+    {
+      "epoch": 0.3888,
+      "grad_norm": 0.4194016754627228,
+      "learning_rate": 4.891057878234827e-06,
+      "loss": 0.504,
+      "step": 972
+    },
+    {
+      "epoch": 0.3892,
+      "grad_norm": 0.4390861988067627,
+      "learning_rate": 4.886895959780618e-06,
+      "loss": 0.5773,
+      "step": 973
+    },
+    {
+      "epoch": 0.3896,
+      "grad_norm": 0.4521550238132477,
+      "learning_rate": 4.882731713663524e-06,
+      "loss": 0.5093,
+      "step": 974
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.4769055247306824,
+      "learning_rate": 4.878565146872505e-06,
+      "loss": 0.5763,
+      "step": 975
+    },
+    {
+      "epoch": 0.3904,
+      "grad_norm": 0.4254240393638611,
+      "learning_rate": 4.874396266400418e-06,
+      "loss": 0.5504,
+      "step": 976
+    },
+    {
+      "epoch": 0.3908,
+      "grad_norm": 0.4530830383300781,
+      "learning_rate": 4.870225079243998e-06,
+      "loss": 0.503,
+      "step": 977
+    },
+    {
+      "epoch": 0.3912,
+      "grad_norm": 0.4451112747192383,
+      "learning_rate": 4.866051592403859e-06,
+      "loss": 0.4848,
+      "step": 978
+    },
+    {
+      "epoch": 0.3916,
+      "grad_norm": 0.5103582739830017,
+      "learning_rate": 4.8618758128844675e-06,
+      "loss": 0.5432,
+      "step": 979
+    },
+    {
+      "epoch": 0.392,
+      "grad_norm": 0.4586694538593292,
+      "learning_rate": 4.857697747694142e-06,
+      "loss": 0.5402,
+      "step": 980
+    },
+    {
+      "epoch": 0.3924,
+      "grad_norm": 0.3923027813434601,
+      "learning_rate": 4.853517403845037e-06,
+      "loss": 0.4839,
+      "step": 981
+    },
+    {
+      "epoch": 0.3928,
+      "grad_norm": 0.43034645915031433,
+      "learning_rate": 4.849334788353128e-06,
+      "loss": 0.5624,
+      "step": 982
+    },
+    {
+      "epoch": 0.3932,
+      "grad_norm": 0.4340299665927887,
+      "learning_rate": 4.8451499082382075e-06,
+      "loss": 0.5928,
+      "step": 983
+    },
+    {
+      "epoch": 0.3936,
+      "grad_norm": 0.5210539698600769,
+      "learning_rate": 4.840962770523865e-06,
+      "loss": 0.5105,
+      "step": 984
+    },
+    {
+      "epoch": 0.394,
+      "grad_norm": 0.4606340527534485,
+      "learning_rate": 4.836773382237481e-06,
+      "loss": 0.5041,
+      "step": 985
+    },
+    {
+      "epoch": 0.3944,
+      "grad_norm": 0.4190974235534668,
+      "learning_rate": 4.832581750410213e-06,
+      "loss": 0.4632,
+      "step": 986
+    },
+    {
+      "epoch": 0.3948,
+      "grad_norm": 0.4932776987552643,
+      "learning_rate": 4.8283878820769835e-06,
+      "loss": 0.5377,
+      "step": 987
+    },
+    {
+      "epoch": 0.3952,
+      "grad_norm": 0.4378507733345032,
+      "learning_rate": 4.824191784276469e-06,
+      "loss": 0.4748,
+      "step": 988
+    },
+    {
+      "epoch": 0.3956,
+      "grad_norm": 0.4637993276119232,
+      "learning_rate": 4.8199934640510875e-06,
+      "loss": 0.6412,
+      "step": 989
+    },
+    {
+      "epoch": 0.396,
+      "grad_norm": 0.5731667280197144,
+      "learning_rate": 4.815792928446986e-06,
+      "loss": 0.6257,
+      "step": 990
+    },
+    {
+      "epoch": 0.3964,
+      "grad_norm": 0.46051129698753357,
+      "learning_rate": 4.811590184514033e-06,
+      "loss": 0.4641,
+      "step": 991
+    },
+    {
+      "epoch": 0.3968,
+      "grad_norm": 0.44054555892944336,
+      "learning_rate": 4.8073852393057995e-06,
+      "loss": 0.4956,
+      "step": 992
+    },
+    {
+      "epoch": 0.3972,
+      "grad_norm": 0.44733765721321106,
+      "learning_rate": 4.803178099879551e-06,
+      "loss": 0.5533,
+      "step": 993
+    },
+    {
+      "epoch": 0.3976,
+      "grad_norm": 0.4734288454055786,
+      "learning_rate": 4.798968773296238e-06,
+      "loss": 0.5923,
+      "step": 994
+    },
+    {
+      "epoch": 0.398,
+      "grad_norm": 0.44051530957221985,
+      "learning_rate": 4.794757266620482e-06,
+      "loss": 0.4936,
+      "step": 995
+    },
+    {
+      "epoch": 0.3984,
+      "grad_norm": 0.48114386200904846,
+      "learning_rate": 4.790543586920561e-06,
+      "loss": 0.6136,
+      "step": 996
+    },
+    {
+      "epoch": 0.3988,
+      "grad_norm": 0.4775163233280182,
+      "learning_rate": 4.7863277412684e-06,
+      "loss": 0.5113,
+      "step": 997
+    },
+    {
+      "epoch": 0.3992,
+      "grad_norm": 0.4406886100769043,
+      "learning_rate": 4.782109736739562e-06,
+      "loss": 0.5045,
+      "step": 998
+    },
+    {
+      "epoch": 0.3996,
+      "grad_norm": 0.5010861754417419,
+      "learning_rate": 4.777889580413232e-06,
+      "loss": 0.5119,
+      "step": 999
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.47519031167030334,
+      "learning_rate": 4.773667279372207e-06,
+      "loss": 0.6024,
+      "step": 1000
+    },
+    {
+      "epoch": 0.4004,
+      "grad_norm": 0.46936264634132385,
+      "learning_rate": 4.769442840702879e-06,
+      "loss": 0.5832,
+      "step": 1001
+    },
+    {
+      "epoch": 0.4008,
+      "grad_norm": 0.4049946665763855,
+      "learning_rate": 4.765216271495233e-06,
+      "loss": 0.5696,
+      "step": 1002
+    },
+    {
+      "epoch": 0.4012,
+      "grad_norm": 0.4819742739200592,
+      "learning_rate": 4.760987578842828e-06,
+      "loss": 0.5264,
+      "step": 1003
+    },
+    {
+      "epoch": 0.4016,
+      "grad_norm": 0.5121580362319946,
+      "learning_rate": 4.756756769842788e-06,
+      "loss": 0.5501,
+      "step": 1004
+    },
+    {
+      "epoch": 0.402,
+      "grad_norm": 0.4060859978199005,
+      "learning_rate": 4.752523851595785e-06,
+      "loss": 0.456,
+      "step": 1005
+    },
+    {
+      "epoch": 0.4024,
+      "grad_norm": 0.5201736688613892,
+      "learning_rate": 4.748288831206037e-06,
+      "loss": 0.6065,
+      "step": 1006
+    },
+    {
+      "epoch": 0.4028,
+      "grad_norm": 0.4155638515949249,
+      "learning_rate": 4.744051715781286e-06,
+      "loss": 0.4812,
+      "step": 1007
+    },
+    {
+      "epoch": 0.4032,
+      "grad_norm": 0.4884546101093292,
+      "learning_rate": 4.739812512432788e-06,
+      "loss": 0.5775,
+      "step": 1008
+    },
+    {
+      "epoch": 0.4036,
+      "grad_norm": 0.4956738352775574,
+      "learning_rate": 4.73557122827531e-06,
+      "loss": 0.6205,
+      "step": 1009
+    },
+    {
+      "epoch": 0.404,
+      "grad_norm": 0.5215380787849426,
+      "learning_rate": 4.731327870427103e-06,
+      "loss": 0.6252,
+      "step": 1010
+    },
+    {
+      "epoch": 0.4044,
+      "grad_norm": 0.47527143359184265,
+      "learning_rate": 4.727082446009909e-06,
+      "loss": 0.56,
+      "step": 1011
+    },
+    {
+      "epoch": 0.4048,
+      "grad_norm": 0.4762853980064392,
+      "learning_rate": 4.722834962148927e-06,
+      "loss": 0.5282,
+      "step": 1012
+    },
+    {
+      "epoch": 0.4052,
+      "grad_norm": 0.42091140151023865,
+      "learning_rate": 4.718585425972819e-06,
+      "loss": 0.495,
+      "step": 1013
+    },
+    {
+      "epoch": 0.4056,
+      "grad_norm": 0.4845111072063446,
+      "learning_rate": 4.714333844613692e-06,
+      "loss": 0.5523,
+      "step": 1014
+    },
+    {
+      "epoch": 0.406,
+      "grad_norm": 0.4628712832927704,
+      "learning_rate": 4.71008022520708e-06,
+      "loss": 0.5445,
+      "step": 1015
+    },
+    {
+      "epoch": 0.4064,
+      "grad_norm": 0.4790785014629364,
+      "learning_rate": 4.705824574891944e-06,
+      "loss": 0.5288,
+      "step": 1016
+    },
+    {
+      "epoch": 0.4068,
+      "grad_norm": 0.5232818126678467,
+      "learning_rate": 4.701566900810648e-06,
+      "loss": 0.6096,
+      "step": 1017
+    },
+    {
+      "epoch": 0.4072,
+      "grad_norm": 0.46365299820899963,
+      "learning_rate": 4.697307210108957e-06,
+      "loss": 0.5674,
+      "step": 1018
+    },
+    {
+      "epoch": 0.4076,
+      "grad_norm": 0.46038663387298584,
+      "learning_rate": 4.693045509936018e-06,
+      "loss": 0.4781,
+      "step": 1019
+    },
+    {
+      "epoch": 0.408,
+      "grad_norm": 0.4882473051548004,
+      "learning_rate": 4.6887818074443514e-06,
+      "loss": 0.5625,
+      "step": 1020
+    },
+    {
+      "epoch": 0.4084,
+      "grad_norm": 0.4895521402359009,
+      "learning_rate": 4.6845161097898376e-06,
+      "loss": 0.5711,
+      "step": 1021
+    },
+    {
+      "epoch": 0.4088,
+      "grad_norm": 0.4484206438064575,
+      "learning_rate": 4.680248424131706e-06,
+      "loss": 0.5218,
+      "step": 1022
+    },
+    {
+      "epoch": 0.4092,
+      "grad_norm": 0.5106765627861023,
+      "learning_rate": 4.675978757632522e-06,
+      "loss": 0.6769,
+      "step": 1023
+    },
+    {
+      "epoch": 0.4096,
+      "grad_norm": 0.45548179745674133,
+      "learning_rate": 4.671707117458176e-06,
+      "loss": 0.4893,
+      "step": 1024
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.4506770074367523,
+      "learning_rate": 4.667433510777872e-06,
+      "loss": 0.5332,
+      "step": 1025
+    },
+    {
+      "epoch": 0.4104,
+      "grad_norm": 0.4675017297267914,
+      "learning_rate": 4.663157944764111e-06,
+      "loss": 0.5545,
+      "step": 1026
+    },
+    {
+      "epoch": 0.4108,
+      "grad_norm": 0.47717100381851196,
+      "learning_rate": 4.658880426592686e-06,
+      "loss": 0.5616,
+      "step": 1027
+    },
+    {
+      "epoch": 0.4112,
+      "grad_norm": 0.4532095789909363,
+      "learning_rate": 4.654600963442665e-06,
+      "loss": 0.5336,
+      "step": 1028
+    },
+    {
+      "epoch": 0.4116,
+      "grad_norm": 0.509066641330719,
+      "learning_rate": 4.65031956249638e-06,
+      "loss": 0.5933,
+      "step": 1029
+    },
+    {
+      "epoch": 0.412,
+      "grad_norm": 0.43408674001693726,
+      "learning_rate": 4.646036230939413e-06,
+      "loss": 0.4959,
+      "step": 1030
+    },
+    {
+      "epoch": 0.4124,
+      "grad_norm": 0.5410246849060059,
+      "learning_rate": 4.641750975960592e-06,
+      "loss": 0.5755,
+      "step": 1031
+    },
+    {
+      "epoch": 0.4128,
+      "grad_norm": 0.47988831996917725,
+      "learning_rate": 4.637463804751969e-06,
+      "loss": 0.4984,
+      "step": 1032
+    },
+    {
+      "epoch": 0.4132,
+      "grad_norm": 0.5014576315879822,
+      "learning_rate": 4.633174724508814e-06,
+      "loss": 0.5499,
+      "step": 1033
+    },
+    {
+      "epoch": 0.4136,
+      "grad_norm": 0.5752670764923096,
+      "learning_rate": 4.628883742429596e-06,
+      "loss": 0.6064,
+      "step": 1034
+    },
+    {
+      "epoch": 0.414,
+      "grad_norm": 0.507512092590332,
+      "learning_rate": 4.624590865715983e-06,
+      "loss": 0.5523,
+      "step": 1035
+    },
+    {
+      "epoch": 0.4144,
+      "grad_norm": 0.5002173781394958,
+      "learning_rate": 4.620296101572819e-06,
+      "loss": 0.6558,
+      "step": 1036
+    },
+    {
+      "epoch": 0.4148,
+      "grad_norm": 0.45491892099380493,
+      "learning_rate": 4.6159994572081155e-06,
+      "loss": 0.5552,
+      "step": 1037
+    },
+    {
+      "epoch": 0.4152,
+      "grad_norm": 0.45507368445396423,
+      "learning_rate": 4.61170093983304e-06,
+      "loss": 0.4994,
+      "step": 1038
+    },
+    {
+      "epoch": 0.4156,
+      "grad_norm": 0.5375739932060242,
+      "learning_rate": 4.607400556661906e-06,
+      "loss": 0.5885,
+      "step": 1039
+    },
+    {
+      "epoch": 0.416,
+      "grad_norm": 0.45972853899002075,
+      "learning_rate": 4.603098314912156e-06,
+      "loss": 0.626,
+      "step": 1040
+    },
+    {
+      "epoch": 0.4164,
+      "grad_norm": 0.446978896856308,
+      "learning_rate": 4.5987942218043484e-06,
+      "loss": 0.4966,
+      "step": 1041
+    },
+    {
+      "epoch": 0.4168,
+      "grad_norm": 0.44826751947402954,
+      "learning_rate": 4.594488284562158e-06,
+      "loss": 0.5734,
+      "step": 1042
+    },
+    {
+      "epoch": 0.4172,
+      "grad_norm": 0.4395397901535034,
+      "learning_rate": 4.590180510412345e-06,
+      "loss": 0.5354,
+      "step": 1043
+    },
+    {
+      "epoch": 0.4176,
+      "grad_norm": 0.4548799395561218,
+      "learning_rate": 4.585870906584758e-06,
+      "loss": 0.6016,
+      "step": 1044
+    },
+    {
+      "epoch": 0.418,
+      "grad_norm": 0.46065983176231384,
+      "learning_rate": 4.581559480312316e-06,
+      "loss": 0.5366,
+      "step": 1045
+    },
+    {
+      "epoch": 0.4184,
+      "grad_norm": 0.45573100447654724,
+      "learning_rate": 4.577246238830995e-06,
+      "loss": 0.6095,
+      "step": 1046
+    },
+    {
+      "epoch": 0.4188,
+      "grad_norm": 0.5116010904312134,
+      "learning_rate": 4.572931189379818e-06,
+      "loss": 0.5555,
+      "step": 1047
+    },
+    {
+      "epoch": 0.4192,
+      "grad_norm": 0.472604900598526,
+      "learning_rate": 4.568614339200843e-06,
+      "loss": 0.5808,
+      "step": 1048
+    },
+    {
+      "epoch": 0.4196,
+      "grad_norm": 0.5294572114944458,
+      "learning_rate": 4.56429569553915e-06,
+      "loss": 0.574,
+      "step": 1049
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.508413553237915,
+      "learning_rate": 4.559975265642828e-06,
+      "loss": 0.5414,
+      "step": 1050
+    },
+    {
+      "epoch": 0.4204,
+      "grad_norm": 0.4205757677555084,
+      "learning_rate": 4.555653056762965e-06,
+      "loss": 0.4573,
+      "step": 1051
+    },
+    {
+      "epoch": 0.4208,
+      "grad_norm": 0.5281642079353333,
+      "learning_rate": 4.551329076153636e-06,
+      "loss": 0.5868,
+      "step": 1052
+    },
+    {
+      "epoch": 0.4212,
+      "grad_norm": 0.46424421668052673,
+      "learning_rate": 4.547003331071886e-06,
+      "loss": 0.5842,
+      "step": 1053
+    },
+    {
+      "epoch": 0.4216,
+      "grad_norm": 0.4419081509113312,
+      "learning_rate": 4.542675828777725e-06,
+      "loss": 0.5302,
+      "step": 1054
+    },
+    {
+      "epoch": 0.422,
+      "grad_norm": 0.46675243973731995,
+      "learning_rate": 4.538346576534109e-06,
+      "loss": 0.5809,
+      "step": 1055
+    },
+    {
+      "epoch": 0.4224,
+      "grad_norm": 0.5093527436256409,
+      "learning_rate": 4.534015581606934e-06,
+      "loss": 0.5996,
+      "step": 1056
+    },
+    {
+      "epoch": 0.4228,
+      "grad_norm": 0.45872142910957336,
+      "learning_rate": 4.529682851265018e-06,
+      "loss": 0.5206,
+      "step": 1057
+    },
+    {
+      "epoch": 0.4232,
+      "grad_norm": 0.4427538514137268,
+      "learning_rate": 4.525348392780094e-06,
+      "loss": 0.5077,
+      "step": 1058
+    },
+    {
+      "epoch": 0.4236,
+      "grad_norm": 0.506925642490387,
+      "learning_rate": 4.5210122134267925e-06,
+      "loss": 0.6141,
+      "step": 1059
+    },
+    {
+      "epoch": 0.424,
+      "grad_norm": 0.5242504477500916,
+      "learning_rate": 4.516674320482636e-06,
+      "loss": 0.5984,
+      "step": 1060
+    },
+    {
+      "epoch": 0.4244,
+      "grad_norm": 0.4801744818687439,
+      "learning_rate": 4.512334721228021e-06,
+      "loss": 0.5656,
+      "step": 1061
+    },
+    {
+      "epoch": 0.4248,
+      "grad_norm": 0.44936510920524597,
+      "learning_rate": 4.507993422946207e-06,
+      "loss": 0.5356,
+      "step": 1062
+    },
+    {
+      "epoch": 0.4252,
+      "grad_norm": 0.5157073140144348,
+      "learning_rate": 4.503650432923304e-06,
+      "loss": 0.6277,
+      "step": 1063
+    },
+    {
+      "epoch": 0.4256,
+      "grad_norm": 0.413225382566452,
+      "learning_rate": 4.499305758448266e-06,
+      "loss": 0.4468,
+      "step": 1064
+    },
+    {
+      "epoch": 0.426,
+      "grad_norm": 0.46837320923805237,
+      "learning_rate": 4.49495940681287e-06,
+      "loss": 0.5491,
+      "step": 1065
+    },
+    {
+      "epoch": 0.4264,
+      "grad_norm": 0.467693567276001,
+      "learning_rate": 4.490611385311707e-06,
+      "loss": 0.6004,
+      "step": 1066
+    },
+    {
+      "epoch": 0.4268,
+      "grad_norm": 0.44910159707069397,
+      "learning_rate": 4.4862617012421765e-06,
+      "loss": 0.4998,
+      "step": 1067
+    },
+    {
+      "epoch": 0.4272,
+      "grad_norm": 0.414615273475647,
+      "learning_rate": 4.481910361904459e-06,
+      "loss": 0.5411,
+      "step": 1068
+    },
+    {
+      "epoch": 0.4276,
+      "grad_norm": 0.48069313168525696,
+      "learning_rate": 4.477557374601523e-06,
+      "loss": 0.5788,
+      "step": 1069
+    },
+    {
+      "epoch": 0.428,
+      "grad_norm": 0.45842450857162476,
+      "learning_rate": 4.473202746639095e-06,
+      "loss": 0.5494,
+      "step": 1070
+    },
+    {
+      "epoch": 0.4284,
+      "grad_norm": 0.48641544580459595,
+      "learning_rate": 4.468846485325661e-06,
+      "loss": 0.5893,
+      "step": 1071
+    },
+    {
+      "epoch": 0.4288,
+      "grad_norm": 0.458332896232605,
+      "learning_rate": 4.464488597972443e-06,
+      "loss": 0.5662,
+      "step": 1072
+    },
+    {
+      "epoch": 0.4292,
+      "grad_norm": 0.5082148909568787,
+      "learning_rate": 4.460129091893396e-06,
+      "loss": 0.664,
+      "step": 1073
+    },
+    {
+      "epoch": 0.4296,
+      "grad_norm": 0.4697319269180298,
+      "learning_rate": 4.455767974405191e-06,
+      "loss": 0.6069,
+      "step": 1074
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.45500311255455017,
+      "learning_rate": 4.4514052528272e-06,
+      "loss": 0.559,
+      "step": 1075
+    },
+    {
+      "epoch": 0.4304,
+      "grad_norm": 0.4315764904022217,
+      "learning_rate": 4.447040934481493e-06,
+      "loss": 0.4619,
+      "step": 1076
+    },
+    {
+      "epoch": 0.4308,
+      "grad_norm": 0.4703432619571686,
+      "learning_rate": 4.442675026692815e-06,
+      "loss": 0.5703,
+      "step": 1077
+    },
+    {
+      "epoch": 0.4312,
+      "grad_norm": 0.49389466643333435,
+      "learning_rate": 4.438307536788581e-06,
+      "loss": 0.5811,
+      "step": 1078
+    },
+    {
+      "epoch": 0.4316,
+      "grad_norm": 0.4419156610965729,
+      "learning_rate": 4.433938472098861e-06,
+      "loss": 0.5763,
+      "step": 1079
+    },
+    {
+      "epoch": 0.432,
+      "grad_norm": 0.5040791034698486,
+      "learning_rate": 4.429567839956368e-06,
+      "loss": 0.5424,
+      "step": 1080
+    },
+    {
+      "epoch": 0.4324,
+      "grad_norm": 0.5160146951675415,
+      "learning_rate": 4.4251956476964445e-06,
+      "loss": 0.584,
+      "step": 1081
+    },
+    {
+      "epoch": 0.4328,
+      "grad_norm": 0.4821451008319855,
+      "learning_rate": 4.420821902657055e-06,
+      "loss": 0.5395,
+      "step": 1082
+    },
+    {
+      "epoch": 0.4332,
+      "grad_norm": 0.5089398622512817,
+      "learning_rate": 4.416446612178762e-06,
+      "loss": 0.6124,
+      "step": 1083
+    },
+    {
+      "epoch": 0.4336,
+      "grad_norm": 0.4727106988430023,
+      "learning_rate": 4.412069783604733e-06,
+      "loss": 0.512,
+      "step": 1084
+    },
+    {
+      "epoch": 0.434,
+      "grad_norm": 0.5604661703109741,
+      "learning_rate": 4.407691424280708e-06,
+      "loss": 0.6138,
+      "step": 1085
+    },
+    {
+      "epoch": 0.4344,
+      "grad_norm": 0.47186407446861267,
+      "learning_rate": 4.403311541555e-06,
+      "loss": 0.5063,
+      "step": 1086
+    },
+    {
+      "epoch": 0.4348,
+      "grad_norm": 0.4935065805912018,
+      "learning_rate": 4.398930142778477e-06,
+      "loss": 0.5855,
+      "step": 1087
+    },
+    {
+      "epoch": 0.4352,
+      "grad_norm": 0.49573445320129395,
+      "learning_rate": 4.394547235304554e-06,
+      "loss": 0.5524,
+      "step": 1088
+    },
+    {
+      "epoch": 0.4356,
+      "grad_norm": 0.484480082988739,
+      "learning_rate": 4.390162826489176e-06,
+      "loss": 0.6309,
+      "step": 1089
+    },
+    {
+      "epoch": 0.436,
+      "grad_norm": 0.4660501182079315,
+      "learning_rate": 4.385776923690807e-06,
+      "loss": 0.5437,
+      "step": 1090
+    },
+    {
+      "epoch": 0.4364,
+      "grad_norm": 0.4966036081314087,
+      "learning_rate": 4.381389534270421e-06,
+      "loss": 0.5881,
+      "step": 1091
+    },
+    {
+      "epoch": 0.4368,
+      "grad_norm": 0.50444096326828,
+      "learning_rate": 4.377000665591484e-06,
+      "loss": 0.5477,
+      "step": 1092
+    },
+    {
+      "epoch": 0.4372,
+      "grad_norm": 0.4618821442127228,
+      "learning_rate": 4.37261032501995e-06,
+      "loss": 0.5197,
+      "step": 1093
+    },
+    {
+      "epoch": 0.4376,
+      "grad_norm": 0.5184645056724548,
+      "learning_rate": 4.368218519924235e-06,
+      "loss": 0.4874,
+      "step": 1094
+    },
+    {
+      "epoch": 0.438,
+      "grad_norm": 0.4814774990081787,
+      "learning_rate": 4.363825257675219e-06,
+      "loss": 0.4983,
+      "step": 1095
+    },
+    {
+      "epoch": 0.4384,
+      "grad_norm": 0.4355922341346741,
+      "learning_rate": 4.359430545646229e-06,
+      "loss": 0.5032,
+      "step": 1096
+    },
+    {
+      "epoch": 0.4388,
+      "grad_norm": 0.5241148471832275,
+      "learning_rate": 4.355034391213018e-06,
+      "loss": 0.5928,
+      "step": 1097
+    },
+    {
+      "epoch": 0.4392,
+      "grad_norm": 0.4663231372833252,
+      "learning_rate": 4.350636801753768e-06,
+      "loss": 0.5468,
+      "step": 1098
+    },
+    {
+      "epoch": 0.4396,
+      "grad_norm": 0.4439452886581421,
+      "learning_rate": 4.346237784649063e-06,
+      "loss": 0.5004,
+      "step": 1099
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.5504818558692932,
+      "learning_rate": 4.341837347281888e-06,
+      "loss": 0.5775,
+      "step": 1100
+    },
+    {
+      "epoch": 0.4404,
+      "grad_norm": 0.45938974618911743,
+      "learning_rate": 4.337435497037606e-06,
+      "loss": 0.5596,
+      "step": 1101
+    },
+    {
+      "epoch": 0.4408,
+      "grad_norm": 0.46923092007637024,
+      "learning_rate": 4.333032241303958e-06,
+      "loss": 0.5178,
+      "step": 1102
+    },
+    {
+      "epoch": 0.4412,
+      "grad_norm": 0.48642316460609436,
+      "learning_rate": 4.328627587471039e-06,
+      "loss": 0.5191,
+      "step": 1103
+    },
+    {
+      "epoch": 0.4416,
+      "grad_norm": 0.5553264617919922,
+      "learning_rate": 4.324221542931292e-06,
+      "loss": 0.7238,
+      "step": 1104
+    },
+    {
+      "epoch": 0.442,
+      "grad_norm": 0.5312342047691345,
+      "learning_rate": 4.319814115079493e-06,
+      "loss": 0.5319,
+      "step": 1105
+    },
+    {
+      "epoch": 0.4424,
+      "grad_norm": 0.5320695638656616,
+      "learning_rate": 4.315405311312743e-06,
+      "loss": 0.6097,
+      "step": 1106
+    },
+    {
+      "epoch": 0.4428,
+      "grad_norm": 0.4854208827018738,
+      "learning_rate": 4.3109951390304484e-06,
+      "loss": 0.5997,
+      "step": 1107
+    },
+    {
+      "epoch": 0.4432,
+      "grad_norm": 0.45754969120025635,
+      "learning_rate": 4.306583605634313e-06,
+      "loss": 0.6103,
+      "step": 1108
+    },
+    {
+      "epoch": 0.4436,
+      "grad_norm": 0.47233521938323975,
+      "learning_rate": 4.3021707185283274e-06,
+      "loss": 0.5375,
+      "step": 1109
+    },
+    {
+      "epoch": 0.444,
+      "grad_norm": 0.5084165930747986,
+      "learning_rate": 4.2977564851187525e-06,
+      "loss": 0.5823,
+      "step": 1110
+    },
+    {
+      "epoch": 0.4444,
+      "grad_norm": 0.47002092003822327,
+      "learning_rate": 4.293340912814108e-06,
+      "loss": 0.5562,
+      "step": 1111
+    },
+    {
+      "epoch": 0.4448,
+      "grad_norm": 0.47349244356155396,
+      "learning_rate": 4.288924009025162e-06,
+      "loss": 0.5875,
+      "step": 1112
+    },
+    {
+      "epoch": 0.4452,
+      "grad_norm": 0.4687560796737671,
+      "learning_rate": 4.284505781164917e-06,
+      "loss": 0.4749,
+      "step": 1113
+    },
+    {
+      "epoch": 0.4456,
+      "grad_norm": 0.4924432635307312,
+      "learning_rate": 4.280086236648599e-06,
+      "loss": 0.5515,
+      "step": 1114
+    },
+    {
+      "epoch": 0.446,
+      "grad_norm": 0.4624880850315094,
+      "learning_rate": 4.2756653828936405e-06,
+      "loss": 0.563,
+      "step": 1115
+    },
+    {
+      "epoch": 0.4464,
+      "grad_norm": 0.4840562343597412,
+      "learning_rate": 4.271243227319673e-06,
+      "loss": 0.4772,
+      "step": 1116
+    },
+    {
+      "epoch": 0.4468,
+      "grad_norm": 0.46091228723526,
+      "learning_rate": 4.266819777348514e-06,
+      "loss": 0.559,
+      "step": 1117
+    },
+    {
+      "epoch": 0.4472,
+      "grad_norm": 0.4594923257827759,
+      "learning_rate": 4.262395040404152e-06,
+      "loss": 0.5677,
+      "step": 1118
+    },
+    {
+      "epoch": 0.4476,
+      "grad_norm": 0.47065865993499756,
+      "learning_rate": 4.257969023912735e-06,
+      "loss": 0.4923,
+      "step": 1119
+    },
+    {
+      "epoch": 0.448,
+      "grad_norm": 0.475721538066864,
+      "learning_rate": 4.253541735302562e-06,
+      "loss": 0.5119,
+      "step": 1120
+    },
+    {
+      "epoch": 0.4484,
+      "grad_norm": 0.5129873752593994,
+      "learning_rate": 4.249113182004063e-06,
+      "loss": 0.5995,
+      "step": 1121
+    },
+    {
+      "epoch": 0.4488,
+      "grad_norm": 0.48000508546829224,
+      "learning_rate": 4.244683371449792e-06,
+      "loss": 0.5229,
+      "step": 1122
+    },
+    {
+      "epoch": 0.4492,
+      "grad_norm": 0.46500471234321594,
+      "learning_rate": 4.240252311074412e-06,
+      "loss": 0.585,
+      "step": 1123
+    },
+    {
+      "epoch": 0.4496,
+      "grad_norm": 0.46694400906562805,
+      "learning_rate": 4.235820008314688e-06,
+      "loss": 0.5224,
+      "step": 1124
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.4840046167373657,
+      "learning_rate": 4.231386470609463e-06,
+      "loss": 0.586,
+      "step": 1125
+    },
+    {
+      "epoch": 0.4504,
+      "grad_norm": 0.48576870560646057,
+      "learning_rate": 4.226951705399659e-06,
+      "loss": 0.5795,
+      "step": 1126
+    },
+    {
+      "epoch": 0.4508,
+      "grad_norm": 0.481308251619339,
+      "learning_rate": 4.222515720128254e-06,
+      "loss": 0.4913,
+      "step": 1127
+    },
+    {
+      "epoch": 0.4512,
+      "grad_norm": 0.46901413798332214,
+      "learning_rate": 4.218078522240276e-06,
+      "loss": 0.5488,
+      "step": 1128
+    },
+    {
+      "epoch": 0.4516,
+      "grad_norm": 0.51523357629776,
+      "learning_rate": 4.21364011918279e-06,
+      "loss": 0.5679,
+      "step": 1129
+    },
+    {
+      "epoch": 0.452,
+      "grad_norm": 0.5236387252807617,
+      "learning_rate": 4.209200518404876e-06,
+      "loss": 0.5059,
+      "step": 1130
+    },
+    {
+      "epoch": 0.4524,
+      "grad_norm": 0.4860883355140686,
+      "learning_rate": 4.2047597273576335e-06,
+      "loss": 0.6122,
+      "step": 1131
+    },
+    {
+      "epoch": 0.4528,
+      "grad_norm": 0.5277671217918396,
+      "learning_rate": 4.200317753494152e-06,
+      "loss": 0.5522,
+      "step": 1132
+    },
+    {
+      "epoch": 0.4532,
+      "grad_norm": 0.45829376578330994,
+      "learning_rate": 4.195874604269514e-06,
+      "loss": 0.5819,
+      "step": 1133
+    },
+    {
+      "epoch": 0.4536,
+      "grad_norm": 0.45748522877693176,
+      "learning_rate": 4.1914302871407655e-06,
+      "loss": 0.5398,
+      "step": 1134
+    },
+    {
+      "epoch": 0.454,
+      "grad_norm": 0.407242089509964,
+      "learning_rate": 4.186984809566921e-06,
+      "loss": 0.4597,
+      "step": 1135
+    },
+    {
+      "epoch": 0.4544,
+      "grad_norm": 0.46401548385620117,
+      "learning_rate": 4.182538179008938e-06,
+      "loss": 0.5768,
+      "step": 1136
+    },
+    {
+      "epoch": 0.4548,
+      "grad_norm": 0.44726067781448364,
+      "learning_rate": 4.178090402929709e-06,
+      "loss": 0.5159,
+      "step": 1137
+    },
+    {
+      "epoch": 0.4552,
+      "grad_norm": 0.4692462086677551,
+      "learning_rate": 4.173641488794052e-06,
+      "loss": 0.569,
+      "step": 1138
+    },
+    {
+      "epoch": 0.4556,
+      "grad_norm": 0.47518929839134216,
+      "learning_rate": 4.16919144406869e-06,
+      "loss": 0.6003,
+      "step": 1139
+    },
+    {
+      "epoch": 0.456,
+      "grad_norm": 0.4495368003845215,
+      "learning_rate": 4.164740276222249e-06,
+      "loss": 0.5276,
+      "step": 1140
+    },
+    {
+      "epoch": 0.4564,
+      "grad_norm": 0.4562149941921234,
+      "learning_rate": 4.160287992725237e-06,
+      "loss": 0.4661,
+      "step": 1141
+    },
+    {
+      "epoch": 0.4568,
+      "grad_norm": 0.469957560300827,
+      "learning_rate": 4.1558346010500365e-06,
+      "loss": 0.6068,
+      "step": 1142
+    },
+    {
+      "epoch": 0.4572,
+      "grad_norm": 0.47348180413246155,
+      "learning_rate": 4.1513801086708865e-06,
+      "loss": 0.5595,
+      "step": 1143
+    },
+    {
+      "epoch": 0.4576,
+      "grad_norm": 0.5264531970024109,
+      "learning_rate": 4.146924523063875e-06,
+      "loss": 0.5726,
+      "step": 1144
+    },
+    {
+      "epoch": 0.458,
+      "grad_norm": 0.41196539998054504,
+      "learning_rate": 4.142467851706926e-06,
+      "loss": 0.4739,
+      "step": 1145
+    },
+    {
+      "epoch": 0.4584,
+      "grad_norm": 0.5362353920936584,
+      "learning_rate": 4.138010102079785e-06,
+      "loss": 0.5435,
+      "step": 1146
+    },
+    {
+      "epoch": 0.4588,
+      "grad_norm": 0.5249359011650085,
+      "learning_rate": 4.133551281664007e-06,
+      "loss": 0.5133,
+      "step": 1147
+    },
+    {
+      "epoch": 0.4592,
+      "grad_norm": 0.47211042046546936,
+      "learning_rate": 4.129091397942944e-06,
+      "loss": 0.4308,
+      "step": 1148
+    },
+    {
+      "epoch": 0.4596,
+      "grad_norm": 0.470157265663147,
+      "learning_rate": 4.124630458401732e-06,
+      "loss": 0.499,
+      "step": 1149
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.4542332589626312,
+      "learning_rate": 4.12016847052728e-06,
+      "loss": 0.5114,
+      "step": 1150
+    },
+    {
+      "epoch": 0.4604,
+      "grad_norm": 0.5682739615440369,
+      "learning_rate": 4.115705441808256e-06,
+      "loss": 0.5619,
+      "step": 1151
+    },
+    {
+      "epoch": 0.4608,
+      "grad_norm": 0.5314604043960571,
+      "learning_rate": 4.111241379735075e-06,
+      "loss": 0.5565,
+      "step": 1152
+    },
+    {
+      "epoch": 0.4612,
+      "grad_norm": 0.47330886125564575,
+      "learning_rate": 4.106776291799887e-06,
+      "loss": 0.5615,
+      "step": 1153
+    },
+    {
+      "epoch": 0.4616,
+      "grad_norm": 0.4496110677719116,
+      "learning_rate": 4.1023101854965626e-06,
+      "loss": 0.5194,
+      "step": 1154
+    },
+    {
+      "epoch": 0.462,
+      "grad_norm": 0.49995559453964233,
+      "learning_rate": 4.097843068320681e-06,
+      "loss": 0.4888,
+      "step": 1155
+    },
+    {
+      "epoch": 0.4624,
+      "grad_norm": 0.4480656087398529,
+      "learning_rate": 4.0933749477695205e-06,
+      "loss": 0.5536,
+      "step": 1156
+    },
+    {
+      "epoch": 0.4628,
+      "grad_norm": 0.4266318082809448,
+      "learning_rate": 4.088905831342042e-06,
+      "loss": 0.4873,
+      "step": 1157
+    },
+    {
+      "epoch": 0.4632,
+      "grad_norm": 0.48847833275794983,
+      "learning_rate": 4.084435726538876e-06,
+      "loss": 0.5844,
+      "step": 1158
+    },
+    {
+      "epoch": 0.4636,
+      "grad_norm": 0.541998028755188,
+      "learning_rate": 4.079964640862315e-06,
+      "loss": 0.528,
+      "step": 1159
+    },
+    {
+      "epoch": 0.464,
+      "grad_norm": 0.5314731001853943,
+      "learning_rate": 4.075492581816295e-06,
+      "loss": 0.6334,
+      "step": 1160
+    },
+    {
+      "epoch": 0.4644,
+      "grad_norm": 0.47492387890815735,
+      "learning_rate": 4.071019556906387e-06,
+      "loss": 0.5644,
+      "step": 1161
+    },
+    {
+      "epoch": 0.4648,
+      "grad_norm": 0.40192219614982605,
+      "learning_rate": 4.066545573639785e-06,
+      "loss": 0.4229,
+      "step": 1162
+    },
+    {
+      "epoch": 0.4652,
+      "grad_norm": 0.5023147463798523,
+      "learning_rate": 4.062070639525285e-06,
+      "loss": 0.5446,
+      "step": 1163
+    },
+    {
+      "epoch": 0.4656,
+      "grad_norm": 0.44182491302490234,
+      "learning_rate": 4.057594762073288e-06,
+      "loss": 0.5071,
+      "step": 1164
+    },
+    {
+      "epoch": 0.466,
+      "grad_norm": 0.45913881063461304,
+      "learning_rate": 4.053117948795769e-06,
+      "loss": 0.5553,
+      "step": 1165
+    },
+    {
+      "epoch": 0.4664,
+      "grad_norm": 0.4556095004081726,
+      "learning_rate": 4.048640207206283e-06,
+      "loss": 0.5374,
+      "step": 1166
+    },
+    {
+      "epoch": 0.4668,
+      "grad_norm": 0.517315149307251,
+      "learning_rate": 4.044161544819933e-06,
+      "loss": 0.6206,
+      "step": 1167
+    },
+    {
+      "epoch": 0.4672,
+      "grad_norm": 0.461018830537796,
+      "learning_rate": 4.0396819691533746e-06,
+      "loss": 0.5424,
+      "step": 1168
+    },
+    {
+      "epoch": 0.4676,
+      "grad_norm": 0.4887098968029022,
+      "learning_rate": 4.035201487724794e-06,
+      "loss": 0.5999,
+      "step": 1169
+    },
+    {
+      "epoch": 0.468,
+      "grad_norm": 0.4762400984764099,
+      "learning_rate": 4.0307201080538974e-06,
+      "loss": 0.5276,
+      "step": 1170
+    },
+    {
+      "epoch": 0.4684,
+      "grad_norm": 0.5469877123832703,
+      "learning_rate": 4.026237837661899e-06,
+      "loss": 0.6401,
+      "step": 1171
+    },
+    {
+      "epoch": 0.4688,
+      "grad_norm": 0.4401659071445465,
+      "learning_rate": 4.021754684071506e-06,
+      "loss": 0.4994,
+      "step": 1172
+    },
+    {
+      "epoch": 0.4692,
+      "grad_norm": 0.49329283833503723,
+      "learning_rate": 4.0172706548069125e-06,
+      "loss": 0.5442,
+      "step": 1173
+    },
+    {
+      "epoch": 0.4696,
+      "grad_norm": 0.4742457866668701,
+      "learning_rate": 4.012785757393776e-06,
+      "loss": 0.6106,
+      "step": 1174
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.4602060914039612,
+      "learning_rate": 4.008299999359216e-06,
+      "loss": 0.572,
+      "step": 1175
+    },
+    {
+      "epoch": 0.4704,
+      "grad_norm": 0.4688628315925598,
+      "learning_rate": 4.003813388231794e-06,
+      "loss": 0.546,
+      "step": 1176
+    },
+    {
+      "epoch": 0.4708,
+      "grad_norm": 0.5030138492584229,
+      "learning_rate": 3.999325931541505e-06,
+      "loss": 0.5704,
+      "step": 1177
+    },
+    {
+      "epoch": 0.4712,
+      "grad_norm": 0.4299745559692383,
+      "learning_rate": 3.994837636819762e-06,
+      "loss": 0.5594,
+      "step": 1178
+    },
+    {
+      "epoch": 0.4716,
+      "grad_norm": 0.5180333852767944,
+      "learning_rate": 3.9903485115993834e-06,
+      "loss": 0.5891,
+      "step": 1179
+    },
+    {
+      "epoch": 0.472,
+      "grad_norm": 0.48325803875923157,
+      "learning_rate": 3.9858585634145845e-06,
+      "loss": 0.5215,
+      "step": 1180
+    },
+    {
+      "epoch": 0.4724,
+      "grad_norm": 0.45841389894485474,
+      "learning_rate": 3.981367799800956e-06,
+      "loss": 0.6235,
+      "step": 1181
+    },
+    {
+      "epoch": 0.4728,
+      "grad_norm": 0.4524870216846466,
+      "learning_rate": 3.976876228295466e-06,
+      "loss": 0.5418,
+      "step": 1182
+    },
+    {
+      "epoch": 0.4732,
+      "grad_norm": 0.509226381778717,
+      "learning_rate": 3.9723838564364305e-06,
+      "loss": 0.5977,
+      "step": 1183
+    },
+    {
+      "epoch": 0.4736,
+      "grad_norm": 0.4991690516471863,
+      "learning_rate": 3.967890691763513e-06,
+      "loss": 0.5788,
+      "step": 1184
+    },
+    {
+      "epoch": 0.474,
+      "grad_norm": 0.5439135432243347,
+      "learning_rate": 3.963396741817706e-06,
+      "loss": 0.5784,
+      "step": 1185
+    },
+    {
+      "epoch": 0.4744,
+      "grad_norm": 0.4289707839488983,
+      "learning_rate": 3.958902014141321e-06,
+      "loss": 0.5169,
+      "step": 1186
+    },
+    {
+      "epoch": 0.4748,
+      "grad_norm": 0.5074586272239685,
+      "learning_rate": 3.954406516277973e-06,
+      "loss": 0.5715,
+      "step": 1187
+    },
+    {
+      "epoch": 0.4752,
+      "grad_norm": 0.4507099390029907,
+      "learning_rate": 3.949910255772572e-06,
+      "loss": 0.6324,
+      "step": 1188
+    },
+    {
+      "epoch": 0.4756,
+      "grad_norm": 0.42596912384033203,
+      "learning_rate": 3.945413240171307e-06,
+      "loss": 0.5326,
+      "step": 1189
+    },
+    {
+      "epoch": 0.476,
+      "grad_norm": 0.4830462336540222,
+      "learning_rate": 3.940915477021632e-06,
+      "loss": 0.6064,
+      "step": 1190
+    },
+    {
+      "epoch": 0.4764,
+      "grad_norm": 0.5114102363586426,
+      "learning_rate": 3.93641697387226e-06,
+      "loss": 0.5712,
+      "step": 1191
+    },
+    {
+      "epoch": 0.4768,
+      "grad_norm": 0.5248522162437439,
+      "learning_rate": 3.931917738273142e-06,
+      "loss": 0.6312,
+      "step": 1192
+    },
+    {
+      "epoch": 0.4772,
+      "grad_norm": 0.48439136147499084,
+      "learning_rate": 3.927417777775461e-06,
+      "loss": 0.5228,
+      "step": 1193
+    },
+    {
+      "epoch": 0.4776,
+      "grad_norm": 0.48873478174209595,
+      "learning_rate": 3.922917099931615e-06,
+      "loss": 0.5546,
+      "step": 1194
+    },
+    {
+      "epoch": 0.478,
+      "grad_norm": 0.48473060131073,
+      "learning_rate": 3.918415712295206e-06,
+      "loss": 0.5022,
+      "step": 1195
+    },
+    {
+      "epoch": 0.4784,
+      "grad_norm": 0.48407095670700073,
+      "learning_rate": 3.913913622421029e-06,
+      "loss": 0.547,
+      "step": 1196
+    },
+    {
+      "epoch": 0.4788,
+      "grad_norm": 0.4828053116798401,
+      "learning_rate": 3.9094108378650566e-06,
+      "loss": 0.5441,
+      "step": 1197
+    },
+    {
+      "epoch": 0.4792,
+      "grad_norm": 0.4596104919910431,
+      "learning_rate": 3.904907366184425e-06,
+      "loss": 0.5507,
+      "step": 1198
+    },
+    {
+      "epoch": 0.4796,
+      "grad_norm": 0.47510096430778503,
+      "learning_rate": 3.900403214937428e-06,
+      "loss": 0.5972,
+      "step": 1199
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.5075947642326355,
+      "learning_rate": 3.8958983916834955e-06,
+      "loss": 0.5585,
+      "step": 1200
+    },
+    {
+      "epoch": 0.4804,
+      "grad_norm": 0.5023922920227051,
+      "learning_rate": 3.891392903983188e-06,
+      "loss": 0.5718,
+      "step": 1201
+    },
+    {
+      "epoch": 0.4808,
+      "grad_norm": 0.5040879249572754,
+      "learning_rate": 3.886886759398181e-06,
+      "loss": 0.4281,
+      "step": 1202
+    },
+    {
+      "epoch": 0.4812,
+      "grad_norm": 0.5060406923294067,
+      "learning_rate": 3.882379965491252e-06,
+      "loss": 0.5222,
+      "step": 1203
+    },
+    {
+      "epoch": 0.4816,
+      "grad_norm": 0.47041571140289307,
+      "learning_rate": 3.877872529826268e-06,
+      "loss": 0.5498,
+      "step": 1204
+    },
+    {
+      "epoch": 0.482,
+      "grad_norm": 0.45117494463920593,
+      "learning_rate": 3.873364459968172e-06,
+      "loss": 0.5662,
+      "step": 1205
+    },
+    {
+      "epoch": 0.4824,
+      "grad_norm": 0.4873691201210022,
+      "learning_rate": 3.8688557634829766e-06,
+      "loss": 0.5611,
+      "step": 1206
+    },
+    {
+      "epoch": 0.4828,
+      "grad_norm": 0.47428447008132935,
+      "learning_rate": 3.8643464479377375e-06,
+      "loss": 0.528,
+      "step": 1207
+    },
+    {
+      "epoch": 0.4832,
+      "grad_norm": 0.4737657606601715,
+      "learning_rate": 3.859836520900556e-06,
+      "loss": 0.4435,
+      "step": 1208
+    },
+    {
+      "epoch": 0.4836,
+      "grad_norm": 0.5013840794563293,
+      "learning_rate": 3.855325989940559e-06,
+      "loss": 0.5217,
+      "step": 1209
+    },
+    {
+      "epoch": 0.484,
+      "grad_norm": 0.46919554471969604,
+      "learning_rate": 3.850814862627884e-06,
+      "loss": 0.6178,
+      "step": 1210
+    },
+    {
+      "epoch": 0.4844,
+      "grad_norm": 0.49590057134628296,
+      "learning_rate": 3.8463031465336715e-06,
+      "loss": 0.5874,
+      "step": 1211
+    },
+    {
+      "epoch": 0.4848,
+      "grad_norm": 0.47950413823127747,
+      "learning_rate": 3.84179084923005e-06,
+      "loss": 0.5478,
+      "step": 1212
+    },
+    {
+      "epoch": 0.4852,
+      "grad_norm": 0.5121056437492371,
+      "learning_rate": 3.837277978290124e-06,
+      "loss": 0.5161,
+      "step": 1213
+    },
+    {
+      "epoch": 0.4856,
+      "grad_norm": 0.44547930359840393,
+      "learning_rate": 3.832764541287958e-06,
+      "loss": 0.5022,
+      "step": 1214
+    },
+    {
+      "epoch": 0.486,
+      "grad_norm": 0.506198525428772,
+      "learning_rate": 3.828250545798571e-06,
+      "loss": 0.5883,
+      "step": 1215
+    },
+    {
+      "epoch": 0.4864,
+      "grad_norm": 0.4669421911239624,
+      "learning_rate": 3.823735999397913e-06,
+      "loss": 0.492,
+      "step": 1216
+    },
+    {
+      "epoch": 0.4868,
+      "grad_norm": 0.46411246061325073,
+      "learning_rate": 3.819220909662867e-06,
+      "loss": 0.5047,
+      "step": 1217
+    },
+    {
+      "epoch": 0.4872,
+      "grad_norm": 0.5031952261924744,
+      "learning_rate": 3.814705284171221e-06,
+      "loss": 0.5952,
+      "step": 1218
+    },
+    {
+      "epoch": 0.4876,
+      "grad_norm": 0.5160549879074097,
+      "learning_rate": 3.8101891305016643e-06,
+      "loss": 0.5459,
+      "step": 1219
+    },
+    {
+      "epoch": 0.488,
+      "grad_norm": 0.5140947699546814,
+      "learning_rate": 3.8056724562337738e-06,
+      "loss": 0.6668,
+      "step": 1220
+    },
+    {
+      "epoch": 0.4884,
+      "grad_norm": 0.4860072433948517,
+      "learning_rate": 3.8011552689479985e-06,
+      "loss": 0.5133,
+      "step": 1221
+    },
+    {
+      "epoch": 0.4888,
+      "grad_norm": 0.4725138247013092,
+      "learning_rate": 3.79663757622565e-06,
+      "loss": 0.5731,
+      "step": 1222
+    },
+    {
+      "epoch": 0.4892,
+      "grad_norm": 0.5040252804756165,
+      "learning_rate": 3.7921193856488865e-06,
+      "loss": 0.585,
+      "step": 1223
+    },
+    {
+      "epoch": 0.4896,
+      "grad_norm": 0.4506980776786804,
+      "learning_rate": 3.787600704800702e-06,
+      "loss": 0.5543,
+      "step": 1224
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.49135822057724,
+      "learning_rate": 3.783081541264915e-06,
+      "loss": 0.5531,
+      "step": 1225
+    },
+    {
+      "epoch": 0.4904,
+      "grad_norm": 0.48195192217826843,
+      "learning_rate": 3.778561902626152e-06,
+      "loss": 0.5561,
+      "step": 1226
+    },
+    {
+      "epoch": 0.4908,
+      "grad_norm": 0.4301593601703644,
+      "learning_rate": 3.7740417964698388e-06,
+      "loss": 0.4825,
+      "step": 1227
+    },
+    {
+      "epoch": 0.4912,
+      "grad_norm": 0.4790424704551697,
+      "learning_rate": 3.769521230382185e-06,
+      "loss": 0.5155,
+      "step": 1228
+    },
+    {
+      "epoch": 0.4916,
+      "grad_norm": 0.4837892949581146,
+      "learning_rate": 3.7650002119501715e-06,
+      "loss": 0.5496,
+      "step": 1229
+    },
+    {
+      "epoch": 0.492,
+      "grad_norm": 0.4896257817745209,
+      "learning_rate": 3.7604787487615388e-06,
+      "loss": 0.5036,
+      "step": 1230
+    },
+    {
+      "epoch": 0.4924,
+      "grad_norm": 0.4826231896877289,
+      "learning_rate": 3.755956848404774e-06,
+      "loss": 0.6042,
+      "step": 1231
+    },
+    {
+      "epoch": 0.4928,
+      "grad_norm": 0.49948158860206604,
+      "learning_rate": 3.751434518469099e-06,
+      "loss": 0.5179,
+      "step": 1232
+    },
+    {
+      "epoch": 0.4932,
+      "grad_norm": 0.5269796252250671,
+      "learning_rate": 3.746911766544454e-06,
+      "loss": 0.5896,
+      "step": 1233
+    },
+    {
+      "epoch": 0.4936,
+      "grad_norm": 0.5144227743148804,
+      "learning_rate": 3.7423886002214895e-06,
+      "loss": 0.5896,
+      "step": 1234
+    },
+    {
+      "epoch": 0.494,
+      "grad_norm": 0.47878497838974,
+      "learning_rate": 3.737865027091551e-06,
+      "loss": 0.5521,
+      "step": 1235
+    },
+    {
+      "epoch": 0.4944,
+      "grad_norm": 0.4803354740142822,
+      "learning_rate": 3.7333410547466667e-06,
+      "loss": 0.5986,
+      "step": 1236
+    },
+    {
+      "epoch": 0.4948,
+      "grad_norm": 0.48093023896217346,
+      "learning_rate": 3.7288166907795362e-06,
+      "loss": 0.566,
+      "step": 1237
+    },
+    {
+      "epoch": 0.4952,
+      "grad_norm": 0.43498966097831726,
+      "learning_rate": 3.7242919427835117e-06,
+      "loss": 0.5302,
+      "step": 1238
+    },
+    {
+      "epoch": 0.4956,
+      "grad_norm": 0.4801994264125824,
+      "learning_rate": 3.719766818352597e-06,
+      "loss": 0.5635,
+      "step": 1239
+    },
+    {
+      "epoch": 0.496,
+      "grad_norm": 0.4880640208721161,
+      "learning_rate": 3.715241325081422e-06,
+      "loss": 0.6138,
+      "step": 1240
+    },
+    {
+      "epoch": 0.4964,
+      "grad_norm": 0.5129235982894897,
+      "learning_rate": 3.7107154705652373e-06,
+      "loss": 0.5623,
+      "step": 1241
+    },
+    {
+      "epoch": 0.4968,
+      "grad_norm": 0.45849037170410156,
+      "learning_rate": 3.7061892623999008e-06,
+      "loss": 0.5122,
+      "step": 1242
+    },
+    {
+      "epoch": 0.4972,
+      "grad_norm": 0.48277243971824646,
+      "learning_rate": 3.701662708181863e-06,
+      "loss": 0.6143,
+      "step": 1243
+    },
+    {
+      "epoch": 0.4976,
+      "grad_norm": 0.45651403069496155,
+      "learning_rate": 3.6971358155081555e-06,
+      "loss": 0.6106,
+      "step": 1244
+    },
+    {
+      "epoch": 0.498,
+      "grad_norm": 0.46248659491539,
+      "learning_rate": 3.692608591976377e-06,
+      "loss": 0.5849,
+      "step": 1245
+    },
+    {
+      "epoch": 0.4984,
+      "grad_norm": 0.4687730371952057,
+      "learning_rate": 3.688081045184684e-06,
+      "loss": 0.544,
+      "step": 1246
+    },
+    {
+      "epoch": 0.4988,
+      "grad_norm": 0.48340800404548645,
+      "learning_rate": 3.683553182731771e-06,
+      "loss": 0.5622,
+      "step": 1247
+    },
+    {
+      "epoch": 0.4992,
+      "grad_norm": 0.48194611072540283,
+      "learning_rate": 3.6790250122168688e-06,
+      "loss": 0.5405,
+      "step": 1248
+    },
+    {
+      "epoch": 0.4996,
+      "grad_norm": 0.5296016335487366,
+      "learning_rate": 3.6744965412397173e-06,
+      "loss": 0.551,
+      "step": 1249
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.5403719544410706,
+      "learning_rate": 3.6699677774005694e-06,
+      "loss": 0.5956,
+      "step": 1250
+    },
+    {
+      "epoch": 0.5004,
+      "grad_norm": 0.4566706120967865,
+      "learning_rate": 3.6654387283001618e-06,
+      "loss": 0.5344,
+      "step": 1251
+    },
+    {
+      "epoch": 0.5008,
+      "grad_norm": 0.5087108612060547,
+      "learning_rate": 3.660909401539714e-06,
+      "loss": 0.5832,
+      "step": 1252
+    },
+    {
+      "epoch": 0.5012,
+      "grad_norm": 0.49442917108535767,
+      "learning_rate": 3.656379804720912e-06,
+      "loss": 0.5847,
+      "step": 1253
+    },
+    {
+      "epoch": 0.5016,
+      "grad_norm": 0.5028738379478455,
+      "learning_rate": 3.6518499454458906e-06,
+      "loss": 0.5638,
+      "step": 1254
+    },
+    {
+      "epoch": 0.502,
+      "grad_norm": 0.5300218462944031,
+      "learning_rate": 3.6473198313172323e-06,
+      "loss": 0.5987,
+      "step": 1255
+    },
+    {
+      "epoch": 0.5024,
+      "grad_norm": 0.5029107928276062,
+      "learning_rate": 3.64278946993794e-06,
+      "loss": 0.6024,
+      "step": 1256
+    },
+    {
+      "epoch": 0.5028,
+      "grad_norm": 0.44316819310188293,
+      "learning_rate": 3.6382588689114374e-06,
+      "loss": 0.5317,
+      "step": 1257
+    },
+    {
+      "epoch": 0.5032,
+      "grad_norm": 0.4869939088821411,
+      "learning_rate": 3.6337280358415456e-06,
+      "loss": 0.5794,
+      "step": 1258
+    },
+    {
+      "epoch": 0.5036,
+      "grad_norm": 0.48294582962989807,
+      "learning_rate": 3.6291969783324784e-06,
+      "loss": 0.557,
+      "step": 1259
+    },
+    {
+      "epoch": 0.504,
+      "grad_norm": 0.4743523597717285,
+      "learning_rate": 3.624665703988825e-06,
+      "loss": 0.6022,
+      "step": 1260
+    },
+    {
+      "epoch": 0.5044,
+      "grad_norm": 0.48669537901878357,
+      "learning_rate": 3.620134220415539e-06,
+      "loss": 0.5158,
+      "step": 1261
+    },
+    {
+      "epoch": 0.5048,
+      "grad_norm": 0.5077384114265442,
+      "learning_rate": 3.615602535217925e-06,
+      "loss": 0.5877,
+      "step": 1262
+    },
+    {
+      "epoch": 0.5052,
+      "grad_norm": 0.45598161220550537,
+      "learning_rate": 3.6110706560016253e-06,
+      "loss": 0.4984,
+      "step": 1263
+    },
+    {
+      "epoch": 0.5056,
+      "grad_norm": 0.46461182832717896,
+      "learning_rate": 3.6065385903726103e-06,
+      "loss": 0.5788,
+      "step": 1264
+    },
+    {
+      "epoch": 0.506,
+      "grad_norm": 0.4900059998035431,
+      "learning_rate": 3.6020063459371594e-06,
+      "loss": 0.5111,
+      "step": 1265
+    },
+    {
+      "epoch": 0.5064,
+      "grad_norm": 0.45117202401161194,
+      "learning_rate": 3.597473930301854e-06,
+      "loss": 0.4874,
+      "step": 1266
+    },
+    {
+      "epoch": 0.5068,
+      "grad_norm": 0.5490992665290833,
+      "learning_rate": 3.592941351073564e-06,
+      "loss": 0.5638,
+      "step": 1267
+    },
+    {
+      "epoch": 0.5072,
+      "grad_norm": 0.48065051436424255,
+      "learning_rate": 3.588408615859433e-06,
+      "loss": 0.6104,
+      "step": 1268
+    },
+    {
+      "epoch": 0.5076,
+      "grad_norm": 0.5219455361366272,
+      "learning_rate": 3.5838757322668647e-06,
+      "loss": 0.524,
+      "step": 1269
+    },
+    {
+      "epoch": 0.508,
+      "grad_norm": 0.5040216445922852,
+      "learning_rate": 3.5793427079035146e-06,
+      "loss": 0.5249,
+      "step": 1270
+    },
+    {
+      "epoch": 0.5084,
+      "grad_norm": 0.4984973669052124,
+      "learning_rate": 3.5748095503772737e-06,
+      "loss": 0.5192,
+      "step": 1271
+    },
+    {
+      "epoch": 0.5088,
+      "grad_norm": 0.5044708847999573,
+      "learning_rate": 3.5702762672962548e-06,
+      "loss": 0.5929,
+      "step": 1272
+    },
+    {
+      "epoch": 0.5092,
+      "grad_norm": 0.566099226474762,
+      "learning_rate": 3.5657428662687833e-06,
+      "loss": 0.5235,
+      "step": 1273
+    },
+    {
+      "epoch": 0.5096,
+      "grad_norm": 0.49442431330680847,
+      "learning_rate": 3.5612093549033803e-06,
+      "loss": 0.602,
+      "step": 1274
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.4640299379825592,
+      "learning_rate": 3.5566757408087554e-06,
+      "loss": 0.5834,
+      "step": 1275
+    },
+    {
+      "epoch": 0.5104,
+      "grad_norm": 0.42226120829582214,
+      "learning_rate": 3.5521420315937883e-06,
+      "loss": 0.4978,
+      "step": 1276
+    },
+    {
+      "epoch": 0.5108,
+      "grad_norm": 0.4852173626422882,
+      "learning_rate": 3.5476082348675186e-06,
+      "loss": 0.5135,
+      "step": 1277
+    },
+    {
+      "epoch": 0.5112,
+      "grad_norm": 0.4628143310546875,
+      "learning_rate": 3.543074358239133e-06,
+      "loss": 0.4778,
+      "step": 1278
+    },
+    {
+      "epoch": 0.5116,
+      "grad_norm": 0.4509110152721405,
+      "learning_rate": 3.538540409317954e-06,
+      "loss": 0.5291,
+      "step": 1279
+    },
+    {
+      "epoch": 0.512,
+      "grad_norm": 0.4843462109565735,
+      "learning_rate": 3.5340063957134207e-06,
+      "loss": 0.6027,
+      "step": 1280
+    },
+    {
+      "epoch": 0.5124,
+      "grad_norm": 0.49010783433914185,
+      "learning_rate": 3.529472325035087e-06,
+      "loss": 0.549,
+      "step": 1281
+    },
+    {
+      "epoch": 0.5128,
+      "grad_norm": 0.5032756328582764,
+      "learning_rate": 3.524938204892598e-06,
+      "loss": 0.7043,
+      "step": 1282
+    },
+    {
+      "epoch": 0.5132,
+      "grad_norm": 0.4765693247318268,
+      "learning_rate": 3.5204040428956817e-06,
+      "loss": 0.5727,
+      "step": 1283
+    },
+    {
+      "epoch": 0.5136,
+      "grad_norm": 0.42829883098602295,
+      "learning_rate": 3.515869846654141e-06,
+      "loss": 0.5099,
+      "step": 1284
+    },
+    {
+      "epoch": 0.514,
+      "grad_norm": 0.4168989360332489,
+      "learning_rate": 3.5113356237778307e-06,
+      "loss": 0.4861,
+      "step": 1285
+    },
+    {
+      "epoch": 0.5144,
+      "grad_norm": 0.4635244309902191,
+      "learning_rate": 3.5068013818766544e-06,
+      "loss": 0.5594,
+      "step": 1286
+    },
+    {
+      "epoch": 0.5148,
+      "grad_norm": 0.4652838706970215,
+      "learning_rate": 3.5022671285605445e-06,
+      "loss": 0.552,
+      "step": 1287
+    },
+    {
+      "epoch": 0.5152,
+      "grad_norm": 0.47168973088264465,
+      "learning_rate": 3.497732871439455e-06,
+      "loss": 0.5156,
+      "step": 1288
+    },
+    {
+      "epoch": 0.5156,
+      "grad_norm": 0.4864683747291565,
+      "learning_rate": 3.4931986181233455e-06,
+      "loss": 0.5414,
+      "step": 1289
+    },
+    {
+      "epoch": 0.516,
+      "grad_norm": 0.4762532114982605,
+      "learning_rate": 3.4886643762221696e-06,
+      "loss": 0.5165,
+      "step": 1290
+    },
+    {
+      "epoch": 0.5164,
+      "grad_norm": 0.4175041913986206,
+      "learning_rate": 3.48413015334586e-06,
+      "loss": 0.5117,
+      "step": 1291
+    },
+    {
+      "epoch": 0.5168,
+      "grad_norm": 0.48602724075317383,
+      "learning_rate": 3.479595957104319e-06,
+      "loss": 0.5763,
+      "step": 1292
+    },
+    {
+      "epoch": 0.5172,
+      "grad_norm": 0.503019392490387,
+      "learning_rate": 3.475061795107403e-06,
+      "loss": 0.5537,
+      "step": 1293
+    },
+    {
+      "epoch": 0.5176,
+      "grad_norm": 0.45459282398223877,
+      "learning_rate": 3.470527674964913e-06,
+      "loss": 0.6082,
+      "step": 1294
+    },
+    {
+      "epoch": 0.518,
+      "grad_norm": 0.4991267919540405,
+      "learning_rate": 3.4659936042865796e-06,
+      "loss": 0.5081,
+      "step": 1295
+    },
+    {
+      "epoch": 0.5184,
+      "grad_norm": 0.5146558880805969,
+      "learning_rate": 3.461459590682047e-06,
+      "loss": 0.5846,
+      "step": 1296
+    },
+    {
+      "epoch": 0.5188,
+      "grad_norm": 0.4307507872581482,
+      "learning_rate": 3.4569256417608664e-06,
+      "loss": 0.4484,
+      "step": 1297
+    },
+    {
+      "epoch": 0.5192,
+      "grad_norm": 0.43799084424972534,
+      "learning_rate": 3.452391765132481e-06,
+      "loss": 0.5526,
+      "step": 1298
+    },
+    {
+      "epoch": 0.5196,
+      "grad_norm": 0.45588675141334534,
+      "learning_rate": 3.447857968406212e-06,
+      "loss": 0.5245,
+      "step": 1299
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.4409703016281128,
+      "learning_rate": 3.443324259191245e-06,
+      "loss": 0.4505,
+      "step": 1300
+    },
+    {
+      "epoch": 0.5204,
+      "grad_norm": 0.48005324602127075,
+      "learning_rate": 3.4387906450966205e-06,
+      "loss": 0.5844,
+      "step": 1301
+    },
+    {
+      "epoch": 0.5208,
+      "grad_norm": 0.4434102773666382,
+      "learning_rate": 3.4342571337312174e-06,
+      "loss": 0.4502,
+      "step": 1302
+    },
+    {
+      "epoch": 0.5212,
+      "grad_norm": 0.554935872554779,
+      "learning_rate": 3.429723732703745e-06,
+      "loss": 0.5439,
+      "step": 1303
+    },
+    {
+      "epoch": 0.5216,
+      "grad_norm": 0.5339680314064026,
+      "learning_rate": 3.4251904496227266e-06,
+      "loss": 0.5693,
+      "step": 1304
+    },
+    {
+      "epoch": 0.522,
+      "grad_norm": 0.4795043170452118,
+      "learning_rate": 3.4206572920964853e-06,
+      "loss": 0.4937,
+      "step": 1305
+    },
+    {
+      "epoch": 0.5224,
+      "grad_norm": 0.4355802536010742,
+      "learning_rate": 3.4161242677331352e-06,
+      "loss": 0.4897,
+      "step": 1306
+    },
+    {
+      "epoch": 0.5228,
+      "grad_norm": 0.5295188426971436,
+      "learning_rate": 3.4115913841405666e-06,
+      "loss": 0.5563,
+      "step": 1307
+    },
+    {
+      "epoch": 0.5232,
+      "grad_norm": 0.5363090634346008,
+      "learning_rate": 3.4070586489264362e-06,
+      "loss": 0.5842,
+      "step": 1308
+    },
+    {
+      "epoch": 0.5236,
+      "grad_norm": 0.4985086917877197,
+      "learning_rate": 3.4025260696981458e-06,
+      "loss": 0.5373,
+      "step": 1309
+    },
+    {
+      "epoch": 0.524,
+      "grad_norm": 0.461211621761322,
+      "learning_rate": 3.3979936540628418e-06,
+      "loss": 0.5216,
+      "step": 1310
+    },
+    {
+      "epoch": 0.5244,
+      "grad_norm": 0.5300297141075134,
+      "learning_rate": 3.3934614096273896e-06,
+      "loss": 0.6442,
+      "step": 1311
+    },
+    {
+      "epoch": 0.5248,
+      "grad_norm": 0.5206688642501831,
+      "learning_rate": 3.3889293439983737e-06,
+      "loss": 0.5457,
+      "step": 1312
+    },
+    {
+      "epoch": 0.5252,
+      "grad_norm": 0.4997401833534241,
+      "learning_rate": 3.384397464782075e-06,
+      "loss": 0.5301,
+      "step": 1313
+    },
+    {
+      "epoch": 0.5256,
+      "grad_norm": 0.527789294719696,
+      "learning_rate": 3.3798657795844613e-06,
+      "loss": 0.5327,
+      "step": 1314
+    },
+    {
+      "epoch": 0.526,
+      "grad_norm": 0.5220302939414978,
+      "learning_rate": 3.3753342960111753e-06,
+      "loss": 0.6045,
+      "step": 1315
+    },
+    {
+      "epoch": 0.5264,
+      "grad_norm": 0.4488482177257538,
+      "learning_rate": 3.370803021667521e-06,
+      "loss": 0.5192,
+      "step": 1316
+    },
+    {
+      "epoch": 0.5268,
+      "grad_norm": 0.486743688583374,
+      "learning_rate": 3.3662719641584547e-06,
+      "loss": 0.6105,
+      "step": 1317
+    },
+    {
+      "epoch": 0.5272,
+      "grad_norm": 0.45400115847587585,
+      "learning_rate": 3.361741131088563e-06,
+      "loss": 0.5667,
+      "step": 1318
+    },
+    {
+      "epoch": 0.5276,
+      "grad_norm": 0.5188816785812378,
+      "learning_rate": 3.35721053006206e-06,
+      "loss": 0.5929,
+      "step": 1319
+    },
+    {
+      "epoch": 0.528,
+      "grad_norm": 0.4486530125141144,
+      "learning_rate": 3.3526801686827684e-06,
+      "loss": 0.4676,
+      "step": 1320
+    },
+    {
+      "epoch": 0.5284,
+      "grad_norm": 0.46255767345428467,
+      "learning_rate": 3.348150054554109e-06,
+      "loss": 0.4653,
+      "step": 1321
+    },
+    {
+      "epoch": 0.5288,
+      "grad_norm": 0.4916232228279114,
+      "learning_rate": 3.3436201952790886e-06,
+      "loss": 0.4699,
+      "step": 1322
+    },
+    {
+      "epoch": 0.5292,
+      "grad_norm": 0.49927806854248047,
+      "learning_rate": 3.3390905984602865e-06,
+      "loss": 0.5563,
+      "step": 1323
+    },
+    {
+      "epoch": 0.5296,
+      "grad_norm": 0.48269158601760864,
+      "learning_rate": 3.334561271699839e-06,
+      "loss": 0.533,
+      "step": 1324
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.4846993088722229,
+      "learning_rate": 3.330032222599432e-06,
+      "loss": 0.5768,
+      "step": 1325
+    },
+    {
+      "epoch": 0.5304,
+      "grad_norm": 0.47500863671302795,
+      "learning_rate": 3.325503458760282e-06,
+      "loss": 0.4884,
+      "step": 1326
+    },
+    {
+      "epoch": 0.5308,
+      "grad_norm": 0.5686325430870056,
+      "learning_rate": 3.320974987783132e-06,
+      "loss": 0.5854,
+      "step": 1327
+    },
+    {
+      "epoch": 0.5312,
+      "grad_norm": 0.5267006158828735,
+      "learning_rate": 3.3164468172682295e-06,
+      "loss": 0.6222,
+      "step": 1328
+    },
+    {
+      "epoch": 0.5316,
+      "grad_norm": 0.5143467783927917,
+      "learning_rate": 3.3119189548153174e-06,
+      "loss": 0.5843,
+      "step": 1329
+    },
+    {
+      "epoch": 0.532,
+      "grad_norm": 0.47253331542015076,
+      "learning_rate": 3.3073914080236225e-06,
+      "loss": 0.5463,
+      "step": 1330
+    },
+    {
+      "epoch": 0.5324,
+      "grad_norm": 0.4366183280944824,
+      "learning_rate": 3.3028641844918448e-06,
+      "loss": 0.516,
+      "step": 1331
+    },
+    {
+      "epoch": 0.5328,
+      "grad_norm": 0.4651104509830475,
+      "learning_rate": 3.2983372918181374e-06,
+      "loss": 0.5399,
+      "step": 1332
+    },
+    {
+      "epoch": 0.5332,
+      "grad_norm": 0.5560972690582275,
+      "learning_rate": 3.2938107376000996e-06,
+      "loss": 0.4981,
+      "step": 1333
+    },
+    {
+      "epoch": 0.5336,
+      "grad_norm": 0.42216429114341736,
+      "learning_rate": 3.2892845294347635e-06,
+      "loss": 0.5063,
+      "step": 1334
+    },
+    {
+      "epoch": 0.534,
+      "grad_norm": 0.49016714096069336,
+      "learning_rate": 3.2847586749185785e-06,
+      "loss": 0.5798,
+      "step": 1335
+    },
+    {
+      "epoch": 0.5344,
+      "grad_norm": 0.5078908801078796,
+      "learning_rate": 3.2802331816474033e-06,
+      "loss": 0.5965,
+      "step": 1336
+    },
+    {
+      "epoch": 0.5348,
+      "grad_norm": 0.49926018714904785,
+      "learning_rate": 3.2757080572164886e-06,
+      "loss": 0.488,
+      "step": 1337
+    },
+    {
+      "epoch": 0.5352,
+      "grad_norm": 0.5147686004638672,
+      "learning_rate": 3.271183309220465e-06,
+      "loss": 0.553,
+      "step": 1338
+    },
+    {
+      "epoch": 0.5356,
+      "grad_norm": 0.4854147136211395,
+      "learning_rate": 3.2666589452533345e-06,
+      "loss": 0.5469,
+      "step": 1339
+    },
+    {
+      "epoch": 0.536,
+      "grad_norm": 0.4860020875930786,
+      "learning_rate": 3.2621349729084486e-06,
+      "loss": 0.579,
+      "step": 1340
+    },
+    {
+      "epoch": 0.5364,
+      "grad_norm": 0.4679085612297058,
+      "learning_rate": 3.2576113997785113e-06,
+      "loss": 0.5247,
+      "step": 1341
+    },
+    {
+      "epoch": 0.5368,
+      "grad_norm": 0.5016301274299622,
+      "learning_rate": 3.2530882334555464e-06,
+      "loss": 0.4817,
+      "step": 1342
+    },
+    {
+      "epoch": 0.5372,
+      "grad_norm": 0.48567822575569153,
+      "learning_rate": 3.2485654815309026e-06,
+      "loss": 0.444,
+      "step": 1343
+    },
+    {
+      "epoch": 0.5376,
+      "grad_norm": 0.4987831115722656,
+      "learning_rate": 3.244043151595226e-06,
+      "loss": 0.647,
+      "step": 1344
+    },
+    {
+      "epoch": 0.538,
+      "grad_norm": 0.5409462451934814,
+      "learning_rate": 3.2395212512384616e-06,
+      "loss": 0.595,
+      "step": 1345
+    },
+    {
+      "epoch": 0.5384,
+      "grad_norm": 0.4637865126132965,
+      "learning_rate": 3.2349997880498293e-06,
+      "loss": 0.5172,
+      "step": 1346
+    },
+    {
+      "epoch": 0.5388,
+      "grad_norm": 0.498721718788147,
+      "learning_rate": 3.2304787696178154e-06,
+      "loss": 0.5287,
+      "step": 1347
+    },
+    {
+      "epoch": 0.5392,
+      "grad_norm": 0.47601523995399475,
+      "learning_rate": 3.2259582035301615e-06,
+      "loss": 0.514,
+      "step": 1348
+    },
+    {
+      "epoch": 0.5396,
+      "grad_norm": 0.47651252150535583,
+      "learning_rate": 3.2214380973738475e-06,
+      "loss": 0.5567,
+      "step": 1349
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.5265172123908997,
+      "learning_rate": 3.2169184587350856e-06,
+      "loss": 0.6422,
+      "step": 1350
+    },
+    {
+      "epoch": 0.5404,
+      "grad_norm": 0.4211336672306061,
+      "learning_rate": 3.2123992951992984e-06,
+      "loss": 0.4929,
+      "step": 1351
+    },
+    {
+      "epoch": 0.5408,
+      "grad_norm": 0.49535515904426575,
+      "learning_rate": 3.2078806143511146e-06,
+      "loss": 0.5527,
+      "step": 1352
+    },
+    {
+      "epoch": 0.5412,
+      "grad_norm": 0.4549042284488678,
+      "learning_rate": 3.2033624237743506e-06,
+      "loss": 0.5098,
+      "step": 1353
+    },
+    {
+      "epoch": 0.5416,
+      "grad_norm": 0.48842307925224304,
+      "learning_rate": 3.1988447310520005e-06,
+      "loss": 0.5428,
+      "step": 1354
+    },
+    {
+      "epoch": 0.542,
+      "grad_norm": 0.4740166962146759,
+      "learning_rate": 3.194327543766226e-06,
+      "loss": 0.4719,
+      "step": 1355
+    },
+    {
+      "epoch": 0.5424,
+      "grad_norm": 0.47193852066993713,
+      "learning_rate": 3.1898108694983356e-06,
+      "loss": 0.5602,
+      "step": 1356
+    },
+    {
+      "epoch": 0.5428,
+      "grad_norm": 0.5566841959953308,
+      "learning_rate": 3.185294715828779e-06,
+      "loss": 0.5657,
+      "step": 1357
+    },
+    {
+      "epoch": 0.5432,
+      "grad_norm": 0.5171043872833252,
+      "learning_rate": 3.180779090337132e-06,
+      "loss": 0.5435,
+      "step": 1358
+    },
+    {
+      "epoch": 0.5436,
+      "grad_norm": 0.552750825881958,
+      "learning_rate": 3.1762640006020865e-06,
+      "loss": 0.6116,
+      "step": 1359
+    },
+    {
+      "epoch": 0.544,
+      "grad_norm": 0.45572030544281006,
+      "learning_rate": 3.17174945420143e-06,
+      "loss": 0.4724,
+      "step": 1360
+    },
+    {
+      "epoch": 0.5444,
+      "grad_norm": 0.5083634853363037,
+      "learning_rate": 3.1672354587120426e-06,
+      "loss": 0.564,
+      "step": 1361
+    },
+    {
+      "epoch": 0.5448,
+      "grad_norm": 0.50925213098526,
+      "learning_rate": 3.1627220217098767e-06,
+      "loss": 0.5546,
+      "step": 1362
+    },
+    {
+      "epoch": 0.5452,
+      "grad_norm": 0.5205938816070557,
+      "learning_rate": 3.1582091507699496e-06,
+      "loss": 0.5864,
+      "step": 1363
+    },
+    {
+      "epoch": 0.5456,
+      "grad_norm": 0.5350721478462219,
+      "learning_rate": 3.153696853466329e-06,
+      "loss": 0.5921,
+      "step": 1364
+    },
+    {
+      "epoch": 0.546,
+      "grad_norm": 0.5171602368354797,
+      "learning_rate": 3.1491851373721164e-06,
+      "loss": 0.4859,
+      "step": 1365
+    },
+    {
+      "epoch": 0.5464,
+      "grad_norm": 0.5181358456611633,
+      "learning_rate": 3.1446740100594415e-06,
+      "loss": 0.5139,
+      "step": 1366
+    },
+    {
+      "epoch": 0.5468,
+      "grad_norm": 0.4513240158557892,
+      "learning_rate": 3.1401634790994444e-06,
+      "loss": 0.4164,
+      "step": 1367
+    },
+    {
+      "epoch": 0.5472,
+      "grad_norm": 0.47397565841674805,
+      "learning_rate": 3.135653552062263e-06,
+      "loss": 0.5654,
+      "step": 1368
+    },
+    {
+      "epoch": 0.5476,
+      "grad_norm": 0.47725215554237366,
+      "learning_rate": 3.1311442365170238e-06,
+      "loss": 0.5322,
+      "step": 1369
+    },
+    {
+      "epoch": 0.548,
+      "grad_norm": 0.4324577748775482,
+      "learning_rate": 3.1266355400318277e-06,
+      "loss": 0.4717,
+      "step": 1370
+    },
+    {
+      "epoch": 0.5484,
+      "grad_norm": 0.5813025236129761,
+      "learning_rate": 3.1221274701737324e-06,
+      "loss": 0.5871,
+      "step": 1371
+    },
+    {
+      "epoch": 0.5488,
+      "grad_norm": 0.4938691556453705,
+      "learning_rate": 3.117620034508748e-06,
+      "loss": 0.5284,
+      "step": 1372
+    },
+    {
+      "epoch": 0.5492,
+      "grad_norm": 0.4700092077255249,
+      "learning_rate": 3.1131132406018182e-06,
+      "loss": 0.5146,
+      "step": 1373
+    },
+    {
+      "epoch": 0.5496,
+      "grad_norm": 0.4936462938785553,
+      "learning_rate": 3.1086070960168124e-06,
+      "loss": 0.5504,
+      "step": 1374
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.5479834675788879,
+      "learning_rate": 3.1041016083165053e-06,
+      "loss": 0.6068,
+      "step": 1375
+    },
+    {
+      "epoch": 0.5504,
+      "grad_norm": 0.4632750451564789,
+      "learning_rate": 3.0995967850625737e-06,
+      "loss": 0.5139,
+      "step": 1376
+    },
+    {
+      "epoch": 0.5508,
+      "grad_norm": 0.5398275256156921,
+      "learning_rate": 3.0950926338155747e-06,
+      "loss": 0.5622,
+      "step": 1377
+    },
+    {
+      "epoch": 0.5512,
+      "grad_norm": 0.5492286086082458,
+      "learning_rate": 3.0905891621349433e-06,
+      "loss": 0.6452,
+      "step": 1378
+    },
+    {
+      "epoch": 0.5516,
+      "grad_norm": 0.5100933909416199,
+      "learning_rate": 3.0860863775789707e-06,
+      "loss": 0.5928,
+      "step": 1379
+    },
+    {
+      "epoch": 0.552,
+      "grad_norm": 0.44502824544906616,
+      "learning_rate": 3.0815842877047942e-06,
+      "loss": 0.5365,
+      "step": 1380
+    },
+    {
+      "epoch": 0.5524,
+      "grad_norm": 0.46557819843292236,
+      "learning_rate": 3.0770829000683864e-06,
+      "loss": 0.5374,
+      "step": 1381
+    },
+    {
+      "epoch": 0.5528,
+      "grad_norm": 0.5025701522827148,
+      "learning_rate": 3.0725822222245384e-06,
+      "loss": 0.582,
+      "step": 1382
+    },
+    {
+      "epoch": 0.5532,
+      "grad_norm": 0.49238553643226624,
+      "learning_rate": 3.0680822617268583e-06,
+      "loss": 0.4892,
+      "step": 1383
+    },
+    {
+      "epoch": 0.5536,
+      "grad_norm": 0.4772971272468567,
+      "learning_rate": 3.0635830261277404e-06,
+      "loss": 0.5475,
+      "step": 1384
+    },
+    {
+      "epoch": 0.554,
+      "grad_norm": 0.5160426497459412,
+      "learning_rate": 3.0590845229783684e-06,
+      "loss": 0.5744,
+      "step": 1385
+    },
+    {
+      "epoch": 0.5544,
+      "grad_norm": 0.4739657938480377,
+      "learning_rate": 3.054586759828694e-06,
+      "loss": 0.4817,
+      "step": 1386
+    },
+    {
+      "epoch": 0.5548,
+      "grad_norm": 0.4471666216850281,
+      "learning_rate": 3.0500897442274274e-06,
+      "loss": 0.5437,
+      "step": 1387
+    },
+    {
+      "epoch": 0.5552,
+      "grad_norm": 0.49380597472190857,
+      "learning_rate": 3.0455934837220273e-06,
+      "loss": 0.5458,
+      "step": 1388
+    },
+    {
+      "epoch": 0.5556,
+      "grad_norm": 0.4983111023902893,
+      "learning_rate": 3.04109798585868e-06,
+      "loss": 0.5455,
+      "step": 1389
+    },
+    {
+      "epoch": 0.556,
+      "grad_norm": 0.5484086871147156,
+      "learning_rate": 3.036603258182295e-06,
+      "loss": 0.5272,
+      "step": 1390
+    },
+    {
+      "epoch": 0.5564,
+      "grad_norm": 0.5066070556640625,
+      "learning_rate": 3.032109308236486e-06,
+      "loss": 0.56,
+      "step": 1391
+    },
+    {
+      "epoch": 0.5568,
+      "grad_norm": 0.5160626769065857,
+      "learning_rate": 3.0276161435635694e-06,
+      "loss": 0.5745,
+      "step": 1392
+    },
+    {
+      "epoch": 0.5572,
+      "grad_norm": 0.5149394869804382,
+      "learning_rate": 3.0231237717045343e-06,
+      "loss": 0.5562,
+      "step": 1393
+    },
+    {
+      "epoch": 0.5576,
+      "grad_norm": 0.5401614308357239,
+      "learning_rate": 3.0186322001990444e-06,
+      "loss": 0.508,
+      "step": 1394
+    },
+    {
+      "epoch": 0.558,
+      "grad_norm": 0.43853285908699036,
+      "learning_rate": 3.014141436585417e-06,
+      "loss": 0.5448,
+      "step": 1395
+    },
+    {
+      "epoch": 0.5584,
+      "grad_norm": 0.58089679479599,
+      "learning_rate": 3.0096514884006165e-06,
+      "loss": 0.6501,
+      "step": 1396
+    },
+    {
+      "epoch": 0.5588,
+      "grad_norm": 0.49932563304901123,
+      "learning_rate": 3.0051623631802384e-06,
+      "loss": 0.5666,
+      "step": 1397
+    },
+    {
+      "epoch": 0.5592,
+      "grad_norm": 0.5262106657028198,
+      "learning_rate": 3.0006740684584946e-06,
+      "loss": 0.5124,
+      "step": 1398
+    },
+    {
+      "epoch": 0.5596,
+      "grad_norm": 0.5242915749549866,
+      "learning_rate": 2.9961866117682056e-06,
+      "loss": 0.6191,
+      "step": 1399
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.44526514410972595,
+      "learning_rate": 2.9917000006407847e-06,
+      "loss": 0.607,
+      "step": 1400
+    },
+    {
+      "epoch": 0.5604,
+      "grad_norm": 0.45736733078956604,
+      "learning_rate": 2.987214242606224e-06,
+      "loss": 0.5174,
+      "step": 1401
+    },
+    {
+      "epoch": 0.5608,
+      "grad_norm": 0.5021191239356995,
+      "learning_rate": 2.9827293451930882e-06,
+      "loss": 0.5242,
+      "step": 1402
+    },
+    {
+      "epoch": 0.5612,
+      "grad_norm": 0.45563265681266785,
+      "learning_rate": 2.978245315928494e-06,
+      "loss": 0.5253,
+      "step": 1403
+    },
+    {
+      "epoch": 0.5616,
+      "grad_norm": 0.5284848809242249,
+      "learning_rate": 2.9737621623381016e-06,
+      "loss": 0.5263,
+      "step": 1404
+    },
+    {
+      "epoch": 0.562,
+      "grad_norm": 0.5633329749107361,
+      "learning_rate": 2.969279891946102e-06,
+      "loss": 0.6081,
+      "step": 1405
+    },
+    {
+      "epoch": 0.5624,
+      "grad_norm": 0.5600154399871826,
+      "learning_rate": 2.9647985122752057e-06,
+      "loss": 0.5416,
+      "step": 1406
+    },
+    {
+      "epoch": 0.5628,
+      "grad_norm": 0.47047778964042664,
+      "learning_rate": 2.960318030846625e-06,
+      "loss": 0.5438,
+      "step": 1407
+    },
+    {
+      "epoch": 0.5632,
+      "grad_norm": 0.4990597069263458,
+      "learning_rate": 2.955838455180067e-06,
+      "loss": 0.511,
+      "step": 1408
+    },
+    {
+      "epoch": 0.5636,
+      "grad_norm": 0.50710529088974,
+      "learning_rate": 2.9513597927937182e-06,
+      "loss": 0.5696,
+      "step": 1409
+    },
+    {
+      "epoch": 0.564,
+      "grad_norm": 0.5165752172470093,
+      "learning_rate": 2.9468820512042297e-06,
+      "loss": 0.548,
+      "step": 1410
+    },
+    {
+      "epoch": 0.5644,
+      "grad_norm": 0.44777095317840576,
+      "learning_rate": 2.9424052379267115e-06,
+      "loss": 0.4402,
+      "step": 1411
+    },
+    {
+      "epoch": 0.5648,
+      "grad_norm": 0.4853544533252716,
+      "learning_rate": 2.9379293604747146e-06,
+      "loss": 0.5268,
+      "step": 1412
+    },
+    {
+      "epoch": 0.5652,
+      "grad_norm": 0.4983876645565033,
+      "learning_rate": 2.933454426360216e-06,
+      "loss": 0.5174,
+      "step": 1413
+    },
+    {
+      "epoch": 0.5656,
+      "grad_norm": 0.4267212152481079,
+      "learning_rate": 2.928980443093614e-06,
+      "loss": 0.5551,
+      "step": 1414
+    },
+    {
+      "epoch": 0.566,
+      "grad_norm": 0.48267069458961487,
+      "learning_rate": 2.9245074181837055e-06,
+      "loss": 0.494,
+      "step": 1415
+    },
+    {
+      "epoch": 0.5664,
+      "grad_norm": 0.4844222962856293,
+      "learning_rate": 2.9200353591376857e-06,
+      "loss": 0.5433,
+      "step": 1416
+    },
+    {
+      "epoch": 0.5668,
+      "grad_norm": 0.6255803108215332,
+      "learning_rate": 2.9155642734611243e-06,
+      "loss": 0.5762,
+      "step": 1417
+    },
+    {
+      "epoch": 0.5672,
+      "grad_norm": 0.46967852115631104,
+      "learning_rate": 2.911094168657959e-06,
+      "loss": 0.558,
+      "step": 1418
+    },
+    {
+      "epoch": 0.5676,
+      "grad_norm": 0.4525752067565918,
+      "learning_rate": 2.906625052230479e-06,
+      "loss": 0.5339,
+      "step": 1419
+    },
+    {
+      "epoch": 0.568,
+      "grad_norm": 0.6286513805389404,
+      "learning_rate": 2.9021569316793184e-06,
+      "loss": 0.5667,
+      "step": 1420
+    },
+    {
+      "epoch": 0.5684,
+      "grad_norm": 0.47905561327934265,
+      "learning_rate": 2.8976898145034386e-06,
+      "loss": 0.5572,
+      "step": 1421
+    },
+    {
+      "epoch": 0.5688,
+      "grad_norm": 0.4969237446784973,
+      "learning_rate": 2.8932237082001134e-06,
+      "loss": 0.5007,
+      "step": 1422
+    },
+    {
+      "epoch": 0.5692,
+      "grad_norm": 0.568736732006073,
+      "learning_rate": 2.8887586202649257e-06,
+      "loss": 0.597,
+      "step": 1423
+    },
+    {
+      "epoch": 0.5696,
+      "grad_norm": 0.5033916234970093,
+      "learning_rate": 2.8842945581917437e-06,
+      "loss": 0.4943,
+      "step": 1424
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.48971912264823914,
+      "learning_rate": 2.8798315294727205e-06,
+      "loss": 0.534,
+      "step": 1425
+    },
+    {
+      "epoch": 0.5704,
+      "grad_norm": 0.5021529793739319,
+      "learning_rate": 2.8753695415982682e-06,
+      "loss": 0.5684,
+      "step": 1426
+    },
+    {
+      "epoch": 0.5708,
+      "grad_norm": 0.4847045838832855,
+      "learning_rate": 2.8709086020570565e-06,
+      "loss": 0.4761,
+      "step": 1427
+    },
+    {
+      "epoch": 0.5712,
+      "grad_norm": 0.492500364780426,
+      "learning_rate": 2.8664487183359935e-06,
+      "loss": 0.514,
+      "step": 1428
+    },
+    {
+      "epoch": 0.5716,
+      "grad_norm": 0.49166664481163025,
+      "learning_rate": 2.861989897920214e-06,
+      "loss": 0.5007,
+      "step": 1429
+    },
+    {
+      "epoch": 0.572,
+      "grad_norm": 0.5053258538246155,
+      "learning_rate": 2.8575321482930737e-06,
+      "loss": 0.6097,
+      "step": 1430
+    },
+    {
+      "epoch": 0.5724,
+      "grad_norm": 0.5131270289421082,
+      "learning_rate": 2.853075476936125e-06,
+      "loss": 0.5356,
+      "step": 1431
+    },
+    {
+      "epoch": 0.5728,
+      "grad_norm": 0.460875004529953,
+      "learning_rate": 2.848619891329115e-06,
+      "loss": 0.4404,
+      "step": 1432
+    },
+    {
+      "epoch": 0.5732,
+      "grad_norm": 0.5094249248504639,
+      "learning_rate": 2.844165398949963e-06,
+      "loss": 0.6103,
+      "step": 1433
+    },
+    {
+      "epoch": 0.5736,
+      "grad_norm": 0.4838579595088959,
+      "learning_rate": 2.8397120072747624e-06,
+      "loss": 0.5164,
+      "step": 1434
+    },
+    {
+      "epoch": 0.574,
+      "grad_norm": 0.5164437294006348,
+      "learning_rate": 2.8352597237777506e-06,
+      "loss": 0.5857,
+      "step": 1435
+    },
+    {
+      "epoch": 0.5744,
+      "grad_norm": 0.5409811735153198,
+      "learning_rate": 2.830808555931311e-06,
+      "loss": 0.5189,
+      "step": 1436
+    },
+    {
+      "epoch": 0.5748,
+      "grad_norm": 0.5643367767333984,
+      "learning_rate": 2.82635851120595e-06,
+      "loss": 0.5882,
+      "step": 1437
+    },
+    {
+      "epoch": 0.5752,
+      "grad_norm": 0.4767090380191803,
+      "learning_rate": 2.821909597070291e-06,
+      "loss": 0.5102,
+      "step": 1438
+    },
+    {
+      "epoch": 0.5756,
+      "grad_norm": 0.4349215626716614,
+      "learning_rate": 2.8174618209910626e-06,
+      "loss": 0.4364,
+      "step": 1439
+    },
+    {
+      "epoch": 0.576,
+      "grad_norm": 0.46870535612106323,
+      "learning_rate": 2.813015190433079e-06,
+      "loss": 0.5712,
+      "step": 1440
+    },
+    {
+      "epoch": 0.5764,
+      "grad_norm": 0.48113444447517395,
+      "learning_rate": 2.808569712859235e-06,
+      "loss": 0.5533,
+      "step": 1441
+    },
+    {
+      "epoch": 0.5768,
+      "grad_norm": 0.4305316209793091,
+      "learning_rate": 2.8041253957304874e-06,
+      "loss": 0.4946,
+      "step": 1442
+    },
+    {
+      "epoch": 0.5772,
+      "grad_norm": 0.5638675093650818,
+      "learning_rate": 2.799682246505848e-06,
+      "loss": 0.5814,
+      "step": 1443
+    },
+    {
+      "epoch": 0.5776,
+      "grad_norm": 0.4995023310184479,
+      "learning_rate": 2.7952402726423673e-06,
+      "loss": 0.5185,
+      "step": 1444
+    },
+    {
+      "epoch": 0.578,
+      "grad_norm": 0.48982349038124084,
+      "learning_rate": 2.790799481595125e-06,
+      "loss": 0.5735,
+      "step": 1445
+    },
+    {
+      "epoch": 0.5784,
+      "grad_norm": 0.43708935379981995,
+      "learning_rate": 2.7863598808172117e-06,
+      "loss": 0.3894,
+      "step": 1446
+    },
+    {
+      "epoch": 0.5788,
+      "grad_norm": 0.4841836094856262,
+      "learning_rate": 2.7819214777597228e-06,
+      "loss": 0.5465,
+      "step": 1447
+    },
+    {
+      "epoch": 0.5792,
+      "grad_norm": 0.5274856090545654,
+      "learning_rate": 2.777484279871746e-06,
+      "loss": 0.5903,
+      "step": 1448
+    },
+    {
+      "epoch": 0.5796,
+      "grad_norm": 0.46429377794265747,
+      "learning_rate": 2.7730482946003414e-06,
+      "loss": 0.461,
+      "step": 1449
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.46727266907691956,
+      "learning_rate": 2.768613529390537e-06,
+      "loss": 0.5232,
+      "step": 1450
+    },
+    {
+      "epoch": 0.5804,
+      "grad_norm": 0.513526439666748,
+      "learning_rate": 2.7641799916853137e-06,
+      "loss": 0.615,
+      "step": 1451
+    },
+    {
+      "epoch": 0.5808,
+      "grad_norm": 0.5125380754470825,
+      "learning_rate": 2.759747688925587e-06,
+      "loss": 0.593,
+      "step": 1452
+    },
+    {
+      "epoch": 0.5812,
+      "grad_norm": 0.5639067888259888,
+      "learning_rate": 2.7553166285502083e-06,
+      "loss": 0.6349,
+      "step": 1453
+    },
+    {
+      "epoch": 0.5816,
+      "grad_norm": 0.48888012766838074,
+      "learning_rate": 2.7508868179959372e-06,
+      "loss": 0.5049,
+      "step": 1454
+    },
+    {
+      "epoch": 0.582,
+      "grad_norm": 0.5085485577583313,
+      "learning_rate": 2.7464582646974377e-06,
+      "loss": 0.4854,
+      "step": 1455
+    },
+    {
+      "epoch": 0.5824,
+      "grad_norm": 0.4603438973426819,
+      "learning_rate": 2.7420309760872655e-06,
+      "loss": 0.6196,
+      "step": 1456
+    },
+    {
+      "epoch": 0.5828,
+      "grad_norm": 0.48601052165031433,
+      "learning_rate": 2.737604959595849e-06,
+      "loss": 0.4975,
+      "step": 1457
+    },
+    {
+      "epoch": 0.5832,
+      "grad_norm": 0.5217102766036987,
+      "learning_rate": 2.7331802226514865e-06,
+      "loss": 0.5848,
+      "step": 1458
+    },
+    {
+      "epoch": 0.5836,
+      "grad_norm": 0.545604944229126,
+      "learning_rate": 2.728756772680327e-06,
+      "loss": 0.6058,
+      "step": 1459
+    },
+    {
+      "epoch": 0.584,
+      "grad_norm": 0.5154001712799072,
+      "learning_rate": 2.7243346171063607e-06,
+      "loss": 0.5404,
+      "step": 1460
+    },
+    {
+      "epoch": 0.5844,
+      "grad_norm": 0.4819687306880951,
+      "learning_rate": 2.719913763351402e-06,
+      "loss": 0.5809,
+      "step": 1461
+    },
+    {
+      "epoch": 0.5848,
+      "grad_norm": 0.4893083870410919,
+      "learning_rate": 2.715494218835082e-06,
+      "loss": 0.6066,
+      "step": 1462
+    },
+    {
+      "epoch": 0.5852,
+      "grad_norm": 0.48114487528800964,
+      "learning_rate": 2.7110759909748376e-06,
+      "loss": 0.5637,
+      "step": 1463
+    },
+    {
+      "epoch": 0.5856,
+      "grad_norm": 0.4523283541202545,
+      "learning_rate": 2.7066590871858926e-06,
+      "loss": 0.5814,
+      "step": 1464
+    },
+    {
+      "epoch": 0.586,
+      "grad_norm": 0.49530142545700073,
+      "learning_rate": 2.702243514881249e-06,
+      "loss": 0.4947,
+      "step": 1465
+    },
+    {
+      "epoch": 0.5864,
+      "grad_norm": 0.5186442732810974,
+      "learning_rate": 2.697829281471673e-06,
+      "loss": 0.5865,
+      "step": 1466
+    },
+    {
+      "epoch": 0.5868,
+      "grad_norm": 0.47039487957954407,
+      "learning_rate": 2.693416394365687e-06,
+      "loss": 0.5396,
+      "step": 1467
+    },
+    {
+      "epoch": 0.5872,
+      "grad_norm": 0.45197176933288574,
+      "learning_rate": 2.6890048609695523e-06,
+      "loss": 0.4497,
+      "step": 1468
+    },
+    {
+      "epoch": 0.5876,
+      "grad_norm": 0.4874049127101898,
+      "learning_rate": 2.684594688687258e-06,
+      "loss": 0.5222,
+      "step": 1469
+    },
+    {
+      "epoch": 0.588,
+      "grad_norm": 0.51507568359375,
+      "learning_rate": 2.6801858849205074e-06,
+      "loss": 0.5094,
+      "step": 1470
+    },
+    {
+      "epoch": 0.5884,
+      "grad_norm": 0.5144968628883362,
+      "learning_rate": 2.6757784570687083e-06,
+      "loss": 0.5865,
+      "step": 1471
+    },
+    {
+      "epoch": 0.5888,
+      "grad_norm": 0.49269288778305054,
+      "learning_rate": 2.6713724125289613e-06,
+      "loss": 0.4984,
+      "step": 1472
+    },
+    {
+      "epoch": 0.5892,
+      "grad_norm": 0.4876870810985565,
+      "learning_rate": 2.666967758696042e-06,
+      "loss": 0.5238,
+      "step": 1473
+    },
+    {
+      "epoch": 0.5896,
+      "grad_norm": 0.4633502662181854,
+      "learning_rate": 2.662564502962394e-06,
+      "loss": 0.5271,
+      "step": 1474
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.5110370516777039,
+      "learning_rate": 2.658162652718113e-06,
+      "loss": 0.537,
+      "step": 1475
+    },
+    {
+      "epoch": 0.5904,
+      "grad_norm": 0.4866006672382355,
+      "learning_rate": 2.653762215350936e-06,
+      "loss": 0.5239,
+      "step": 1476
+    },
+    {
+      "epoch": 0.5908,
+      "grad_norm": 0.544572114944458,
+      "learning_rate": 2.6493631982462313e-06,
+      "loss": 0.5784,
+      "step": 1477
+    },
+    {
+      "epoch": 0.5912,
+      "grad_norm": 0.5041329264640808,
+      "learning_rate": 2.6449656087869815e-06,
+      "loss": 0.5382,
+      "step": 1478
+    },
+    {
+      "epoch": 0.5916,
+      "grad_norm": 0.4942692220211029,
+      "learning_rate": 2.640569454353772e-06,
+      "loss": 0.5125,
+      "step": 1479
+    },
+    {
+      "epoch": 0.592,
+      "grad_norm": 0.5565232038497925,
+      "learning_rate": 2.63617474232478e-06,
+      "loss": 0.5759,
+      "step": 1480
+    },
+    {
+      "epoch": 0.5924,
+      "grad_norm": 0.48877766728401184,
+      "learning_rate": 2.6317814800757655e-06,
+      "loss": 0.5607,
+      "step": 1481
+    },
+    {
+      "epoch": 0.5928,
+      "grad_norm": 0.4964582622051239,
+      "learning_rate": 2.627389674980051e-06,
+      "loss": 0.5714,
+      "step": 1482
+    },
+    {
+      "epoch": 0.5932,
+      "grad_norm": 0.5051253437995911,
+      "learning_rate": 2.622999334408516e-06,
+      "loss": 0.5631,
+      "step": 1483
+    },
+    {
+      "epoch": 0.5936,
+      "grad_norm": 0.460700124502182,
+      "learning_rate": 2.61861046572958e-06,
+      "loss": 0.5801,
+      "step": 1484
+    },
+    {
+      "epoch": 0.594,
+      "grad_norm": 0.499276727437973,
+      "learning_rate": 2.614223076309193e-06,
+      "loss": 0.5632,
+      "step": 1485
+    },
+    {
+      "epoch": 0.5944,
+      "grad_norm": 0.4617083668708801,
+      "learning_rate": 2.609837173510824e-06,
+      "loss": 0.5333,
+      "step": 1486
+    },
+    {
+      "epoch": 0.5948,
+      "grad_norm": 0.5099096298217773,
+      "learning_rate": 2.6054527646954463e-06,
+      "loss": 0.5713,
+      "step": 1487
+    },
+    {
+      "epoch": 0.5952,
+      "grad_norm": 0.43509137630462646,
+      "learning_rate": 2.601069857221523e-06,
+      "loss": 0.4773,
+      "step": 1488
+    },
+    {
+      "epoch": 0.5956,
+      "grad_norm": 0.4796929359436035,
+      "learning_rate": 2.5966884584450014e-06,
+      "loss": 0.58,
+      "step": 1489
+    },
+    {
+      "epoch": 0.596,
+      "grad_norm": 0.5261090397834778,
+      "learning_rate": 2.5923085757192924e-06,
+      "loss": 0.5708,
+      "step": 1490
+    },
+    {
+      "epoch": 0.5964,
+      "grad_norm": 0.5301308631896973,
+      "learning_rate": 2.5879302163952672e-06,
+      "loss": 0.5811,
+      "step": 1491
+    },
+    {
+      "epoch": 0.5968,
+      "grad_norm": 0.5250400900840759,
+      "learning_rate": 2.583553387821238e-06,
+      "loss": 0.5675,
+      "step": 1492
+    },
+    {
+      "epoch": 0.5972,
+      "grad_norm": 0.5334447622299194,
+      "learning_rate": 2.5791780973429466e-06,
+      "loss": 0.5659,
+      "step": 1493
+    },
+    {
+      "epoch": 0.5976,
+      "grad_norm": 0.4800483286380768,
+      "learning_rate": 2.5748043523035546e-06,
+      "loss": 0.5534,
+      "step": 1494
+    },
+    {
+      "epoch": 0.598,
+      "grad_norm": 0.5284641981124878,
+      "learning_rate": 2.5704321600436317e-06,
+      "loss": 0.5998,
+      "step": 1495
+    },
+    {
+      "epoch": 0.5984,
+      "grad_norm": 0.5287473797798157,
+      "learning_rate": 2.5660615279011386e-06,
+      "loss": 0.4678,
+      "step": 1496
+    },
+    {
+      "epoch": 0.5988,
+      "grad_norm": 0.43308907747268677,
+      "learning_rate": 2.561692463211419e-06,
+      "loss": 0.4371,
+      "step": 1497
+    },
+    {
+      "epoch": 0.5992,
+      "grad_norm": 0.5262799859046936,
+      "learning_rate": 2.557324973307186e-06,
+      "loss": 0.4869,
+      "step": 1498
+    },
+    {
+      "epoch": 0.5996,
+      "grad_norm": 0.4628347158432007,
+      "learning_rate": 2.5529590655185074e-06,
+      "loss": 0.626,
+      "step": 1499
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.5220506191253662,
+      "learning_rate": 2.5485947471727995e-06,
+      "loss": 0.5732,
+      "step": 1500
+    },
+    {
+      "epoch": 0.6004,
+      "grad_norm": 0.46114087104797363,
+      "learning_rate": 2.5442320255948096e-06,
+      "loss": 0.5597,
+      "step": 1501
+    },
+    {
+      "epoch": 0.6008,
+      "grad_norm": 0.5224429368972778,
+      "learning_rate": 2.5398709081066046e-06,
+      "loss": 0.5119,
+      "step": 1502
+    },
+    {
+      "epoch": 0.6012,
+      "grad_norm": 0.4979833960533142,
+      "learning_rate": 2.5355114020275576e-06,
+      "loss": 0.5346,
+      "step": 1503
+    },
+    {
+      "epoch": 0.6016,
+      "grad_norm": 0.4844669699668884,
+      "learning_rate": 2.5311535146743388e-06,
+      "loss": 0.4852,
+      "step": 1504
+    },
+    {
+      "epoch": 0.602,
+      "grad_norm": 0.4876634180545807,
+      "learning_rate": 2.526797253360905e-06,
+      "loss": 0.5701,
+      "step": 1505
+    },
+    {
+      "epoch": 0.6024,
+      "grad_norm": 0.47222307324409485,
+      "learning_rate": 2.5224426253984775e-06,
+      "loss": 0.5319,
+      "step": 1506
+    },
+    {
+      "epoch": 0.6028,
+      "grad_norm": 0.5080312490463257,
+      "learning_rate": 2.5180896380955415e-06,
+      "loss": 0.5992,
+      "step": 1507
+    },
+    {
+      "epoch": 0.6032,
+      "grad_norm": 0.49264171719551086,
+      "learning_rate": 2.5137382987578247e-06,
+      "loss": 0.5549,
+      "step": 1508
+    },
+    {
+      "epoch": 0.6036,
+      "grad_norm": 0.46697530150413513,
+      "learning_rate": 2.5093886146882923e-06,
+      "loss": 0.5624,
+      "step": 1509
+    },
+    {
+      "epoch": 0.604,
+      "grad_norm": 0.5428259968757629,
+      "learning_rate": 2.50504059318713e-06,
+      "loss": 0.5634,
+      "step": 1510
+    },
+    {
+      "epoch": 0.6044,
+      "grad_norm": 0.5281015038490295,
+      "learning_rate": 2.5006942415517338e-06,
+      "loss": 0.6014,
+      "step": 1511
+    },
+    {
+      "epoch": 0.6048,
+      "grad_norm": 0.5557918548583984,
+      "learning_rate": 2.496349567076696e-06,
+      "loss": 0.5203,
+      "step": 1512
+    },
+    {
+      "epoch": 0.6052,
+      "grad_norm": 0.4895382821559906,
+      "learning_rate": 2.492006577053793e-06,
+      "loss": 0.5726,
+      "step": 1513
+    },
+    {
+      "epoch": 0.6056,
+      "grad_norm": 0.563296377658844,
+      "learning_rate": 2.487665278771979e-06,
+      "loss": 0.5745,
+      "step": 1514
+    },
+    {
+      "epoch": 0.606,
+      "grad_norm": 0.4936528205871582,
+      "learning_rate": 2.483325679517363e-06,
+      "loss": 0.5615,
+      "step": 1515
+    },
+    {
+      "epoch": 0.6064,
+      "grad_norm": 0.5365152955055237,
+      "learning_rate": 2.4789877865732082e-06,
+      "loss": 0.5461,
+      "step": 1516
+    },
+    {
+      "epoch": 0.6068,
+      "grad_norm": 0.4697369635105133,
+      "learning_rate": 2.4746516072199076e-06,
+      "loss": 0.5281,
+      "step": 1517
+    },
+    {
+      "epoch": 0.6072,
+      "grad_norm": 0.5144125819206238,
+      "learning_rate": 2.4703171487349826e-06,
+      "loss": 0.5162,
+      "step": 1518
+    },
+    {
+      "epoch": 0.6076,
+      "grad_norm": 0.5662809610366821,
+      "learning_rate": 2.4659844183930663e-06,
+      "loss": 0.6308,
+      "step": 1519
+    },
+    {
+      "epoch": 0.608,
+      "grad_norm": 0.5164036750793457,
+      "learning_rate": 2.4616534234658916e-06,
+      "loss": 0.5497,
+      "step": 1520
+    },
+    {
+      "epoch": 0.6084,
+      "grad_norm": 0.5437076091766357,
+      "learning_rate": 2.457324171222276e-06,
+      "loss": 0.5659,
+      "step": 1521
+    },
+    {
+      "epoch": 0.6088,
+      "grad_norm": 0.4282589852809906,
+      "learning_rate": 2.452996668928115e-06,
+      "loss": 0.4424,
+      "step": 1522
+    },
+    {
+      "epoch": 0.6092,
+      "grad_norm": 0.4498371183872223,
+      "learning_rate": 2.4486709238463642e-06,
+      "loss": 0.4756,
+      "step": 1523
+    },
+    {
+      "epoch": 0.6096,
+      "grad_norm": 0.5019795298576355,
+      "learning_rate": 2.4443469432370347e-06,
+      "loss": 0.5285,
+      "step": 1524
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.47362902760505676,
+      "learning_rate": 2.440024734357173e-06,
+      "loss": 0.6729,
+      "step": 1525
+    },
+    {
+      "epoch": 0.6104,
+      "grad_norm": 0.5483762621879578,
+      "learning_rate": 2.4357043044608513e-06,
+      "loss": 0.5122,
+      "step": 1526
+    },
+    {
+      "epoch": 0.6108,
+      "grad_norm": 0.5297125577926636,
+      "learning_rate": 2.431385660799157e-06,
+      "loss": 0.5444,
+      "step": 1527
+    },
+    {
+      "epoch": 0.6112,
+      "grad_norm": 0.5263299345970154,
+      "learning_rate": 2.4270688106201816e-06,
+      "loss": 0.5762,
+      "step": 1528
+    },
+    {
+      "epoch": 0.6116,
+      "grad_norm": 0.5046945214271545,
+      "learning_rate": 2.4227537611690053e-06,
+      "loss": 0.4573,
+      "step": 1529
+    },
+    {
+      "epoch": 0.612,
+      "grad_norm": 0.5073798894882202,
+      "learning_rate": 2.418440519687684e-06,
+      "loss": 0.5786,
+      "step": 1530
+    },
+    {
+      "epoch": 0.6124,
+      "grad_norm": 0.5219867825508118,
+      "learning_rate": 2.414129093415243e-06,
+      "loss": 0.581,
+      "step": 1531
+    },
+    {
+      "epoch": 0.6128,
+      "grad_norm": 0.46836137771606445,
+      "learning_rate": 2.4098194895876554e-06,
+      "loss": 0.4835,
+      "step": 1532
+    },
+    {
+      "epoch": 0.6132,
+      "grad_norm": 0.4778684079647064,
+      "learning_rate": 2.4055117154378427e-06,
+      "loss": 0.5348,
+      "step": 1533
+    },
+    {
+      "epoch": 0.6136,
+      "grad_norm": 0.4793270230293274,
+      "learning_rate": 2.4012057781956515e-06,
+      "loss": 0.5527,
+      "step": 1534
+    },
+    {
+      "epoch": 0.614,
+      "grad_norm": 0.49084943532943726,
+      "learning_rate": 2.3969016850878457e-06,
+      "loss": 0.5705,
+      "step": 1535
+    },
+    {
+      "epoch": 0.6144,
+      "grad_norm": 0.49898597598075867,
+      "learning_rate": 2.392599443338094e-06,
+      "loss": 0.5809,
+      "step": 1536
+    },
+    {
+      "epoch": 0.6148,
+      "grad_norm": 0.4914925992488861,
+      "learning_rate": 2.3882990601669587e-06,
+      "loss": 0.5114,
+      "step": 1537
+    },
+    {
+      "epoch": 0.6152,
+      "grad_norm": 0.50290846824646,
+      "learning_rate": 2.3840005427918848e-06,
+      "loss": 0.4965,
+      "step": 1538
+    },
+    {
+      "epoch": 0.6156,
+      "grad_norm": 0.4977038502693176,
+      "learning_rate": 2.3797038984271814e-06,
+      "loss": 0.5015,
+      "step": 1539
+    },
+    {
+      "epoch": 0.616,
+      "grad_norm": 0.5011242032051086,
+      "learning_rate": 2.3754091342840174e-06,
+      "loss": 0.5551,
+      "step": 1540
+    },
+    {
+      "epoch": 0.6164,
+      "grad_norm": 0.5203733444213867,
+      "learning_rate": 2.3711162575704035e-06,
+      "loss": 0.499,
+      "step": 1541
+    },
+    {
+      "epoch": 0.6168,
+      "grad_norm": 0.4532087445259094,
+      "learning_rate": 2.3668252754911866e-06,
+      "loss": 0.5067,
+      "step": 1542
+    },
+    {
+      "epoch": 0.6172,
+      "grad_norm": 0.539614200592041,
+      "learning_rate": 2.3625361952480307e-06,
+      "loss": 0.5632,
+      "step": 1543
+    },
+    {
+      "epoch": 0.6176,
+      "grad_norm": 0.5588467717170715,
+      "learning_rate": 2.3582490240394075e-06,
+      "loss": 0.5892,
+      "step": 1544
+    },
+    {
+      "epoch": 0.618,
+      "grad_norm": 0.5543122887611389,
+      "learning_rate": 2.353963769060587e-06,
+      "loss": 0.5045,
+      "step": 1545
+    },
+    {
+      "epoch": 0.6184,
+      "grad_norm": 0.51371830701828,
+      "learning_rate": 2.3496804375036206e-06,
+      "loss": 0.5724,
+      "step": 1546
+    },
+    {
+      "epoch": 0.6188,
+      "grad_norm": 0.6098910570144653,
+      "learning_rate": 2.3453990365573353e-06,
+      "loss": 0.5578,
+      "step": 1547
+    },
+    {
+      "epoch": 0.6192,
+      "grad_norm": 0.46126070618629456,
+      "learning_rate": 2.3411195734073137e-06,
+      "loss": 0.4742,
+      "step": 1548
+    },
+    {
+      "epoch": 0.6196,
+      "grad_norm": 0.5160725712776184,
+      "learning_rate": 2.336842055235889e-06,
+      "loss": 0.4787,
+      "step": 1549
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.4890509247779846,
+      "learning_rate": 2.3325664892221287e-06,
+      "loss": 0.515,
+      "step": 1550
+    },
+    {
+      "epoch": 0.6204,
+      "grad_norm": 0.5041446089744568,
+      "learning_rate": 2.328292882541823e-06,
+      "loss": 0.5502,
+      "step": 1551
+    },
+    {
+      "epoch": 0.6208,
+      "grad_norm": 0.5130775570869446,
+      "learning_rate": 2.324021242367478e-06,
+      "loss": 0.5673,
+      "step": 1552
+    },
+    {
+      "epoch": 0.6212,
+      "grad_norm": 0.5083710551261902,
+      "learning_rate": 2.3197515758682943e-06,
+      "loss": 0.5668,
+      "step": 1553
+    },
+    {
+      "epoch": 0.6216,
+      "grad_norm": 0.49546748399734497,
+      "learning_rate": 2.3154838902101623e-06,
+      "loss": 0.4704,
+      "step": 1554
+    },
+    {
+      "epoch": 0.622,
+      "grad_norm": 0.47747087478637695,
+      "learning_rate": 2.311218192555648e-06,
+      "loss": 0.5373,
+      "step": 1555
+    },
+    {
+      "epoch": 0.6224,
+      "grad_norm": 0.5391461253166199,
+      "learning_rate": 2.306954490063982e-06,
+      "loss": 0.5444,
+      "step": 1556
+    },
+    {
+      "epoch": 0.6228,
+      "grad_norm": 0.5440353751182556,
+      "learning_rate": 2.3026927898910426e-06,
+      "loss": 0.5749,
+      "step": 1557
+    },
+    {
+      "epoch": 0.6232,
+      "grad_norm": 0.4923385977745056,
+      "learning_rate": 2.2984330991893527e-06,
+      "loss": 0.4856,
+      "step": 1558
+    },
+    {
+      "epoch": 0.6236,
+      "grad_norm": 0.47224387526512146,
+      "learning_rate": 2.2941754251080574e-06,
+      "loss": 0.5269,
+      "step": 1559
+    },
+    {
+      "epoch": 0.624,
+      "grad_norm": 0.48879680037498474,
+      "learning_rate": 2.28991977479292e-06,
+      "loss": 0.5758,
+      "step": 1560
+    },
+    {
+      "epoch": 0.6244,
+      "grad_norm": 0.4885936677455902,
+      "learning_rate": 2.285666155386308e-06,
+      "loss": 0.541,
+      "step": 1561
+    },
+    {
+      "epoch": 0.6248,
+      "grad_norm": 0.44462254643440247,
+      "learning_rate": 2.281414574027181e-06,
+      "loss": 0.4438,
+      "step": 1562
+    },
+    {
+      "epoch": 0.6252,
+      "grad_norm": 0.5792527198791504,
+      "learning_rate": 2.2771650378510734e-06,
+      "loss": 0.6358,
+      "step": 1563
+    },
+    {
+      "epoch": 0.6256,
+      "grad_norm": 0.47472530603408813,
+      "learning_rate": 2.2729175539900925e-06,
+      "loss": 0.4899,
+      "step": 1564
+    },
+    {
+      "epoch": 0.626,
+      "grad_norm": 0.46082860231399536,
+      "learning_rate": 2.268672129572896e-06,
+      "loss": 0.6056,
+      "step": 1565
+    },
+    {
+      "epoch": 0.6264,
+      "grad_norm": 0.5208197832107544,
+      "learning_rate": 2.2644287717246906e-06,
+      "loss": 0.5852,
+      "step": 1566
+    },
+    {
+      "epoch": 0.6268,
+      "grad_norm": 0.502346396446228,
+      "learning_rate": 2.2601874875672127e-06,
+      "loss": 0.5329,
+      "step": 1567
+    },
+    {
+      "epoch": 0.6272,
+      "grad_norm": 0.5229620933532715,
+      "learning_rate": 2.2559482842187154e-06,
+      "loss": 0.5473,
+      "step": 1568
+    },
+    {
+      "epoch": 0.6276,
+      "grad_norm": 0.49244144558906555,
+      "learning_rate": 2.2517111687939623e-06,
+      "loss": 0.5066,
+      "step": 1569
+    },
+    {
+      "epoch": 0.628,
+      "grad_norm": 0.5355249047279358,
+      "learning_rate": 2.247476148404214e-06,
+      "loss": 0.5816,
+      "step": 1570
+    },
+    {
+      "epoch": 0.6284,
+      "grad_norm": 0.5006580352783203,
+      "learning_rate": 2.243243230157213e-06,
+      "loss": 0.6485,
+      "step": 1571
+    },
+    {
+      "epoch": 0.6288,
+      "grad_norm": 0.521933376789093,
+      "learning_rate": 2.2390124211571718e-06,
+      "loss": 0.6115,
+      "step": 1572
+    },
+    {
+      "epoch": 0.6292,
+      "grad_norm": 0.5163118243217468,
+      "learning_rate": 2.2347837285047677e-06,
+      "loss": 0.4758,
+      "step": 1573
+    },
+    {
+      "epoch": 0.6296,
+      "grad_norm": 0.5321094989776611,
+      "learning_rate": 2.2305571592971214e-06,
+      "loss": 0.584,
+      "step": 1574
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.503724992275238,
+      "learning_rate": 2.226332720627794e-06,
+      "loss": 0.5437,
+      "step": 1575
+    },
+    {
+      "epoch": 0.6304,
+      "grad_norm": 0.5667294263839722,
+      "learning_rate": 2.2221104195867673e-06,
+      "loss": 0.5661,
+      "step": 1576
+    },
+    {
+      "epoch": 0.6308,
+      "grad_norm": 0.48289167881011963,
+      "learning_rate": 2.217890263260437e-06,
+      "loss": 0.5256,
+      "step": 1577
+    },
+    {
+      "epoch": 0.6312,
+      "grad_norm": 0.4751034080982208,
+      "learning_rate": 2.2136722587316e-06,
+      "loss": 0.5426,
+      "step": 1578
+    },
+    {
+      "epoch": 0.6316,
+      "grad_norm": 0.5564718246459961,
+      "learning_rate": 2.209456413079439e-06,
+      "loss": 0.5565,
+      "step": 1579
+    },
+    {
+      "epoch": 0.632,
+      "grad_norm": 0.49462926387786865,
+      "learning_rate": 2.205242733379518e-06,
+      "loss": 0.4651,
+      "step": 1580
+    },
+    {
+      "epoch": 0.6324,
+      "grad_norm": 0.4669487476348877,
+      "learning_rate": 2.2010312267037617e-06,
+      "loss": 0.5236,
+      "step": 1581
+    },
+    {
+      "epoch": 0.6328,
+      "grad_norm": 0.48065274953842163,
+      "learning_rate": 2.1968219001204503e-06,
+      "loss": 0.5738,
+      "step": 1582
+    },
+    {
+      "epoch": 0.6332,
+      "grad_norm": 0.4668215811252594,
+      "learning_rate": 2.192614760694202e-06,
+      "loss": 0.5286,
+      "step": 1583
+    },
+    {
+      "epoch": 0.6336,
+      "grad_norm": 0.5288259983062744,
+      "learning_rate": 2.188409815485967e-06,
+      "loss": 0.5231,
+      "step": 1584
+    },
+    {
+      "epoch": 0.634,
+      "grad_norm": 0.48246291279792786,
+      "learning_rate": 2.1842070715530135e-06,
+      "loss": 0.5433,
+      "step": 1585
+    },
+    {
+      "epoch": 0.6344,
+      "grad_norm": 0.5169889330863953,
+      "learning_rate": 2.180006535948913e-06,
+      "loss": 0.5927,
+      "step": 1586
+    },
+    {
+      "epoch": 0.6348,
+      "grad_norm": 0.45649388432502747,
+      "learning_rate": 2.175808215723531e-06,
+      "loss": 0.4159,
+      "step": 1587
+    },
+    {
+      "epoch": 0.6352,
+      "grad_norm": 0.5261488556861877,
+      "learning_rate": 2.171612117923016e-06,
+      "loss": 0.5426,
+      "step": 1588
+    },
+    {
+      "epoch": 0.6356,
+      "grad_norm": 0.5158681869506836,
+      "learning_rate": 2.167418249589787e-06,
+      "loss": 0.5287,
+      "step": 1589
+    },
+    {
+      "epoch": 0.636,
+      "grad_norm": 0.5511676669120789,
+      "learning_rate": 2.163226617762519e-06,
+      "loss": 0.633,
+      "step": 1590
+    },
+    {
+      "epoch": 0.6364,
+      "grad_norm": 0.45589837431907654,
+      "learning_rate": 2.159037229476136e-06,
+      "loss": 0.5101,
+      "step": 1591
+    },
+    {
+      "epoch": 0.6368,
+      "grad_norm": 0.4977359473705292,
+      "learning_rate": 2.1548500917617933e-06,
+      "loss": 0.4789,
+      "step": 1592
+    },
+    {
+      "epoch": 0.6372,
+      "grad_norm": 0.45518171787261963,
+      "learning_rate": 2.150665211646871e-06,
+      "loss": 0.5169,
+      "step": 1593
+    },
+    {
+      "epoch": 0.6376,
+      "grad_norm": 0.5482791066169739,
+      "learning_rate": 2.1464825961549635e-06,
+      "loss": 0.7014,
+      "step": 1594
+    },
+    {
+      "epoch": 0.638,
+      "grad_norm": 0.4804784059524536,
+      "learning_rate": 2.142302252305857e-06,
+      "loss": 0.5088,
+      "step": 1595
+    },
+    {
+      "epoch": 0.6384,
+      "grad_norm": 0.4980086088180542,
+      "learning_rate": 2.138124187115532e-06,
+      "loss": 0.4547,
+      "step": 1596
+    },
+    {
+      "epoch": 0.6388,
+      "grad_norm": 0.48985469341278076,
+      "learning_rate": 2.133948407596142e-06,
+      "loss": 0.5564,
+      "step": 1597
+    },
+    {
+      "epoch": 0.6392,
+      "grad_norm": 0.5659129023551941,
+      "learning_rate": 2.1297749207560015e-06,
+      "loss": 0.5135,
+      "step": 1598
+    },
+    {
+      "epoch": 0.6396,
+      "grad_norm": 0.5160675048828125,
+      "learning_rate": 2.1256037335995828e-06,
+      "loss": 0.5263,
+      "step": 1599
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.47263601422309875,
+      "learning_rate": 2.1214348531274957e-06,
+      "loss": 0.5682,
+      "step": 1600
+    },
+    {
+      "epoch": 0.6404,
+      "grad_norm": 0.4825802445411682,
+      "learning_rate": 2.1172682863364766e-06,
+      "loss": 0.5353,
+      "step": 1601
+    },
+    {
+      "epoch": 0.6408,
+      "grad_norm": 0.47716841101646423,
+      "learning_rate": 2.113104040219382e-06,
+      "loss": 0.5353,
+      "step": 1602
+    },
+    {
+      "epoch": 0.6412,
+      "grad_norm": 0.47250092029571533,
+      "learning_rate": 2.108942121765173e-06,
+      "loss": 0.5076,
+      "step": 1603
+    },
+    {
+      "epoch": 0.6416,
+      "grad_norm": 0.4839315116405487,
+      "learning_rate": 2.1047825379589038e-06,
+      "loss": 0.56,
+      "step": 1604
+    },
+    {
+      "epoch": 0.642,
+      "grad_norm": 0.5323452949523926,
+      "learning_rate": 2.1006252957817083e-06,
+      "loss": 0.5248,
+      "step": 1605
+    },
+    {
+      "epoch": 0.6424,
+      "grad_norm": 0.5054993033409119,
+      "learning_rate": 2.0964704022107945e-06,
+      "loss": 0.487,
+      "step": 1606
+    },
+    {
+      "epoch": 0.6428,
+      "grad_norm": 0.5133071541786194,
+      "learning_rate": 2.0923178642194233e-06,
+      "loss": 0.6727,
+      "step": 1607
+    },
+    {
+      "epoch": 0.6432,
+      "grad_norm": 0.5333604216575623,
+      "learning_rate": 2.088167688776908e-06,
+      "loss": 0.6296,
+      "step": 1608
+    },
+    {
+      "epoch": 0.6436,
+      "grad_norm": 0.4804375171661377,
+      "learning_rate": 2.0840198828485934e-06,
+      "loss": 0.5005,
+      "step": 1609
+    },
+    {
+      "epoch": 0.644,
+      "grad_norm": 0.4947167634963989,
+      "learning_rate": 2.079874453395847e-06,
+      "loss": 0.5693,
+      "step": 1610
+    },
+    {
+      "epoch": 0.6444,
+      "grad_norm": 0.5010197162628174,
+      "learning_rate": 2.07573140737605e-06,
+      "loss": 0.5887,
+      "step": 1611
+    },
+    {
+      "epoch": 0.6448,
+      "grad_norm": 0.5427845120429993,
+      "learning_rate": 2.07159075174258e-06,
+      "loss": 0.5781,
+      "step": 1612
+    },
+    {
+      "epoch": 0.6452,
+      "grad_norm": 0.4725536108016968,
+      "learning_rate": 2.067452493444808e-06,
+      "loss": 0.533,
+      "step": 1613
+    },
+    {
+      "epoch": 0.6456,
+      "grad_norm": 0.529393196105957,
+      "learning_rate": 2.063316639428077e-06,
+      "loss": 0.4901,
+      "step": 1614
+    },
+    {
+      "epoch": 0.646,
+      "grad_norm": 0.493694007396698,
+      "learning_rate": 2.059183196633697e-06,
+      "loss": 0.5665,
+      "step": 1615
+    },
+    {
+      "epoch": 0.6464,
+      "grad_norm": 0.48902615904808044,
+      "learning_rate": 2.055052171998929e-06,
+      "loss": 0.5396,
+      "step": 1616
+    },
+    {
+      "epoch": 0.6468,
+      "grad_norm": 0.4694744348526001,
+      "learning_rate": 2.0509235724569786e-06,
+      "loss": 0.5601,
+      "step": 1617
+    },
+    {
+      "epoch": 0.6472,
+      "grad_norm": 0.5156325101852417,
+      "learning_rate": 2.04679740493698e-06,
+      "loss": 0.4973,
+      "step": 1618
+    },
+    {
+      "epoch": 0.6476,
+      "grad_norm": 0.44242939352989197,
+      "learning_rate": 2.0426736763639846e-06,
+      "loss": 0.4728,
+      "step": 1619
+    },
+    {
+      "epoch": 0.648,
+      "grad_norm": 0.47373783588409424,
+      "learning_rate": 2.0385523936589525e-06,
+      "loss": 0.5331,
+      "step": 1620
+    },
+    {
+      "epoch": 0.6484,
+      "grad_norm": 0.5373921394348145,
+      "learning_rate": 2.0344335637387345e-06,
+      "loss": 0.5832,
+      "step": 1621
+    },
+    {
+      "epoch": 0.6488,
+      "grad_norm": 0.5912891626358032,
+      "learning_rate": 2.0303171935160704e-06,
+      "loss": 0.5745,
+      "step": 1622
+    },
+    {
+      "epoch": 0.6492,
+      "grad_norm": 0.5122224688529968,
+      "learning_rate": 2.0262032898995707e-06,
+      "loss": 0.5184,
+      "step": 1623
+    },
+    {
+      "epoch": 0.6496,
+      "grad_norm": 0.49968165159225464,
+      "learning_rate": 2.022091859793703e-06,
+      "loss": 0.5457,
+      "step": 1624
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.539306640625,
+      "learning_rate": 2.0179829100987857e-06,
+      "loss": 0.543,
+      "step": 1625
+    },
+    {
+      "epoch": 0.6504,
+      "grad_norm": 0.5262332558631897,
+      "learning_rate": 2.013876447710972e-06,
+      "loss": 0.6301,
+      "step": 1626
+    },
+    {
+      "epoch": 0.6508,
+      "grad_norm": 0.4939981997013092,
+      "learning_rate": 2.009772479522246e-06,
+      "loss": 0.4742,
+      "step": 1627
+    },
+    {
+      "epoch": 0.6512,
+      "grad_norm": 0.5203678011894226,
+      "learning_rate": 2.0056710124204002e-06,
+      "loss": 0.566,
+      "step": 1628
+    },
+    {
+      "epoch": 0.6516,
+      "grad_norm": 0.44473370909690857,
+      "learning_rate": 2.0015720532890332e-06,
+      "loss": 0.5685,
+      "step": 1629
+    },
+    {
+      "epoch": 0.652,
+      "grad_norm": 0.570111095905304,
+      "learning_rate": 1.9974756090075296e-06,
+      "loss": 0.6232,
+      "step": 1630
+    },
+    {
+      "epoch": 0.6524,
+      "grad_norm": 0.47535547614097595,
+      "learning_rate": 1.9933816864510587e-06,
+      "loss": 0.4962,
+      "step": 1631
+    },
+    {
+      "epoch": 0.6528,
+      "grad_norm": 0.5097532868385315,
+      "learning_rate": 1.989290292490556e-06,
+      "loss": 0.4939,
+      "step": 1632
+    },
+    {
+      "epoch": 0.6532,
+      "grad_norm": 0.5101918578147888,
+      "learning_rate": 1.9852014339927113e-06,
+      "loss": 0.4766,
+      "step": 1633
+    },
+    {
+      "epoch": 0.6536,
+      "grad_norm": 0.5294265747070312,
+      "learning_rate": 1.9811151178199597e-06,
+      "loss": 0.5946,
+      "step": 1634
+    },
+    {
+      "epoch": 0.654,
+      "grad_norm": 0.5878704786300659,
+      "learning_rate": 1.9770313508304687e-06,
+      "loss": 0.5594,
+      "step": 1635
+    },
+    {
+      "epoch": 0.6544,
+      "grad_norm": 0.5145576000213623,
+      "learning_rate": 1.9729501398781314e-06,
+      "loss": 0.5903,
+      "step": 1636
+    },
+    {
+      "epoch": 0.6548,
+      "grad_norm": 0.53581702709198,
+      "learning_rate": 1.968871491812547e-06,
+      "loss": 0.4923,
+      "step": 1637
+    },
+    {
+      "epoch": 0.6552,
+      "grad_norm": 0.4903261363506317,
+      "learning_rate": 1.964795413479016e-06,
+      "loss": 0.5131,
+      "step": 1638
+    },
+    {
+      "epoch": 0.6556,
+      "grad_norm": 0.47253820300102234,
+      "learning_rate": 1.960721911718522e-06,
+      "loss": 0.5141,
+      "step": 1639
+    },
+    {
+      "epoch": 0.656,
+      "grad_norm": 0.5105094909667969,
+      "learning_rate": 1.956650993367728e-06,
+      "loss": 0.6333,
+      "step": 1640
+    },
+    {
+      "epoch": 0.6564,
+      "grad_norm": 0.4645703136920929,
+      "learning_rate": 1.9525826652589624e-06,
+      "loss": 0.5166,
+      "step": 1641
+    },
+    {
+      "epoch": 0.6568,
+      "grad_norm": 0.4405498206615448,
+      "learning_rate": 1.9485169342202026e-06,
+      "loss": 0.4998,
+      "step": 1642
+    },
+    {
+      "epoch": 0.6572,
+      "grad_norm": 0.5502030849456787,
+      "learning_rate": 1.944453807075068e-06,
+      "loss": 0.5153,
+      "step": 1643
+    },
+    {
+      "epoch": 0.6576,
+      "grad_norm": 0.5228597521781921,
+      "learning_rate": 1.9403932906428075e-06,
+      "loss": 0.5555,
+      "step": 1644
+    },
+    {
+      "epoch": 0.658,
+      "grad_norm": 0.49661386013031006,
+      "learning_rate": 1.936335391738292e-06,
+      "loss": 0.4916,
+      "step": 1645
+    },
+    {
+      "epoch": 0.6584,
+      "grad_norm": 0.507220447063446,
+      "learning_rate": 1.9322801171719963e-06,
+      "loss": 0.4786,
+      "step": 1646
+    },
+    {
+      "epoch": 0.6588,
+      "grad_norm": 0.5505663752555847,
+      "learning_rate": 1.928227473749991e-06,
+      "loss": 0.5889,
+      "step": 1647
+    },
+    {
+      "epoch": 0.6592,
+      "grad_norm": 0.461161345243454,
+      "learning_rate": 1.924177468273929e-06,
+      "loss": 0.5665,
+      "step": 1648
+    },
+    {
+      "epoch": 0.6596,
+      "grad_norm": 0.47537487745285034,
+      "learning_rate": 1.92013010754104e-06,
+      "loss": 0.5456,
+      "step": 1649
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.5142361521720886,
+      "learning_rate": 1.9160853983441143e-06,
+      "loss": 0.5599,
+      "step": 1650
+    },
+    {
+      "epoch": 0.6604,
+      "grad_norm": 0.498871773481369,
+      "learning_rate": 1.9120433474714876e-06,
+      "loss": 0.5814,
+      "step": 1651
+    },
+    {
+      "epoch": 0.6608,
+      "grad_norm": 0.5202580094337463,
+      "learning_rate": 1.9080039617070384e-06,
+      "loss": 0.6258,
+      "step": 1652
+    },
+    {
+      "epoch": 0.6612,
+      "grad_norm": 0.5052293539047241,
+      "learning_rate": 1.9039672478301698e-06,
+      "loss": 0.5705,
+      "step": 1653
+    },
+    {
+      "epoch": 0.6616,
+      "grad_norm": 0.4765607714653015,
+      "learning_rate": 1.8999332126158018e-06,
+      "loss": 0.5216,
+      "step": 1654
+    },
+    {
+      "epoch": 0.662,
+      "grad_norm": 0.519019365310669,
+      "learning_rate": 1.8959018628343612e-06,
+      "loss": 0.4994,
+      "step": 1655
+    },
+    {
+      "epoch": 0.6624,
+      "grad_norm": 0.5589582920074463,
+      "learning_rate": 1.8918732052517633e-06,
+      "loss": 0.5712,
+      "step": 1656
+    },
+    {
+      "epoch": 0.6628,
+      "grad_norm": 0.5202416777610779,
+      "learning_rate": 1.887847246629406e-06,
+      "loss": 0.5674,
+      "step": 1657
+    },
+    {
+      "epoch": 0.6632,
+      "grad_norm": 0.5023145079612732,
+      "learning_rate": 1.8838239937241604e-06,
+      "loss": 0.5578,
+      "step": 1658
+    },
+    {
+      "epoch": 0.6636,
+      "grad_norm": 0.4727594554424286,
+      "learning_rate": 1.8798034532883528e-06,
+      "loss": 0.5314,
+      "step": 1659
+    },
+    {
+      "epoch": 0.664,
+      "grad_norm": 0.5230659246444702,
+      "learning_rate": 1.8757856320697609e-06,
+      "loss": 0.6046,
+      "step": 1660
+    },
+    {
+      "epoch": 0.6644,
+      "grad_norm": 0.5028753280639648,
+      "learning_rate": 1.8717705368115946e-06,
+      "loss": 0.4558,
+      "step": 1661
+    },
+    {
+      "epoch": 0.6648,
+      "grad_norm": 0.5712804794311523,
+      "learning_rate": 1.8677581742524908e-06,
+      "loss": 0.5368,
+      "step": 1662
+    },
+    {
+      "epoch": 0.6652,
+      "grad_norm": 0.4760363698005676,
+      "learning_rate": 1.8637485511265004e-06,
+      "loss": 0.5177,
+      "step": 1663
+    },
+    {
+      "epoch": 0.6656,
+      "grad_norm": 0.5169165730476379,
+      "learning_rate": 1.8597416741630777e-06,
+      "loss": 0.6252,
+      "step": 1664
+    },
+    {
+      "epoch": 0.666,
+      "grad_norm": 0.48152515292167664,
+      "learning_rate": 1.8557375500870656e-06,
+      "loss": 0.5348,
+      "step": 1665
+    },
+    {
+      "epoch": 0.6664,
+      "grad_norm": 0.4943107068538666,
+      "learning_rate": 1.851736185618686e-06,
+      "loss": 0.5256,
+      "step": 1666
+    },
+    {
+      "epoch": 0.6668,
+      "grad_norm": 0.5224728584289551,
+      "learning_rate": 1.8477375874735342e-06,
+      "loss": 0.5988,
+      "step": 1667
+    },
+    {
+      "epoch": 0.6672,
+      "grad_norm": 0.499190092086792,
+      "learning_rate": 1.8437417623625558e-06,
+      "loss": 0.5087,
+      "step": 1668
+    },
+    {
+      "epoch": 0.6676,
+      "grad_norm": 0.5248624682426453,
+      "learning_rate": 1.8397487169920495e-06,
+      "loss": 0.4826,
+      "step": 1669
+    },
+    {
+      "epoch": 0.668,
+      "grad_norm": 0.5038529634475708,
+      "learning_rate": 1.835758458063644e-06,
+      "loss": 0.614,
+      "step": 1670
+    },
+    {
+      "epoch": 0.6684,
+      "grad_norm": 0.541998028755188,
+      "learning_rate": 1.8317709922742915e-06,
+      "loss": 0.5436,
+      "step": 1671
+    },
+    {
+      "epoch": 0.6688,
+      "grad_norm": 0.5454816222190857,
+      "learning_rate": 1.8277863263162597e-06,
+      "loss": 0.483,
+      "step": 1672
+    },
+    {
+      "epoch": 0.6692,
+      "grad_norm": 0.49650028347969055,
+      "learning_rate": 1.8238044668771123e-06,
+      "loss": 0.4889,
+      "step": 1673
+    },
+    {
+      "epoch": 0.6696,
+      "grad_norm": 0.5571967363357544,
+      "learning_rate": 1.8198254206397094e-06,
+      "loss": 0.6268,
+      "step": 1674
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.5088862776756287,
+      "learning_rate": 1.8158491942821822e-06,
+      "loss": 0.4964,
+      "step": 1675
+    },
+    {
+      "epoch": 0.6704,
+      "grad_norm": 0.4773922264575958,
+      "learning_rate": 1.8118757944779358e-06,
+      "loss": 0.4747,
+      "step": 1676
+    },
+    {
+      "epoch": 0.6708,
+      "grad_norm": 0.5273101925849915,
+      "learning_rate": 1.807905227895626e-06,
+      "loss": 0.5286,
+      "step": 1677
+    },
+    {
+      "epoch": 0.6712,
+      "grad_norm": 0.4688742160797119,
+      "learning_rate": 1.8039375011991588e-06,
+      "loss": 0.4147,
+      "step": 1678
+    },
+    {
+      "epoch": 0.6716,
+      "grad_norm": 0.536142110824585,
+      "learning_rate": 1.7999726210476696e-06,
+      "loss": 0.5764,
+      "step": 1679
+    },
+    {
+      "epoch": 0.672,
+      "grad_norm": 0.5289327502250671,
+      "learning_rate": 1.7960105940955164e-06,
+      "loss": 0.523,
+      "step": 1680
+    },
+    {
+      "epoch": 0.6724,
+      "grad_norm": 0.46021583676338196,
+      "learning_rate": 1.792051426992273e-06,
+      "loss": 0.5167,
+      "step": 1681
+    },
+    {
+      "epoch": 0.6728,
+      "grad_norm": 0.5137984156608582,
+      "learning_rate": 1.7880951263827083e-06,
+      "loss": 0.4866,
+      "step": 1682
+    },
+    {
+      "epoch": 0.6732,
+      "grad_norm": 0.5088829398155212,
+      "learning_rate": 1.7841416989067848e-06,
+      "loss": 0.5143,
+      "step": 1683
+    },
+    {
+      "epoch": 0.6736,
+      "grad_norm": 0.5051218271255493,
+      "learning_rate": 1.7801911511996384e-06,
+      "loss": 0.4959,
+      "step": 1684
+    },
+    {
+      "epoch": 0.674,
+      "grad_norm": 0.5160439610481262,
+      "learning_rate": 1.7762434898915764e-06,
+      "loss": 0.5458,
+      "step": 1685
+    },
+    {
+      "epoch": 0.6744,
+      "grad_norm": 0.47013598680496216,
+      "learning_rate": 1.7722987216080584e-06,
+      "loss": 0.5298,
+      "step": 1686
+    },
+    {
+      "epoch": 0.6748,
+      "grad_norm": 0.507504403591156,
+      "learning_rate": 1.7683568529696885e-06,
+      "loss": 0.4713,
+      "step": 1687
+    },
+    {
+      "epoch": 0.6752,
+      "grad_norm": 0.5315131545066833,
+      "learning_rate": 1.764417890592208e-06,
+      "loss": 0.5132,
+      "step": 1688
+    },
+    {
+      "epoch": 0.6756,
+      "grad_norm": 0.5345341563224792,
+      "learning_rate": 1.7604818410864753e-06,
+      "loss": 0.6151,
+      "step": 1689
+    },
+    {
+      "epoch": 0.676,
+      "grad_norm": 0.5330063700675964,
+      "learning_rate": 1.7565487110584654e-06,
+      "loss": 0.6183,
+      "step": 1690
+    },
+    {
+      "epoch": 0.6764,
+      "grad_norm": 0.5632655620574951,
+      "learning_rate": 1.752618507109248e-06,
+      "loss": 0.5759,
+      "step": 1691
+    },
+    {
+      "epoch": 0.6768,
+      "grad_norm": 0.4822584092617035,
+      "learning_rate": 1.7486912358349875e-06,
+      "loss": 0.4738,
+      "step": 1692
+    },
+    {
+      "epoch": 0.6772,
+      "grad_norm": 0.4986719489097595,
+      "learning_rate": 1.744766903826921e-06,
+      "loss": 0.5118,
+      "step": 1693
+    },
+    {
+      "epoch": 0.6776,
+      "grad_norm": 0.5079983472824097,
+      "learning_rate": 1.7408455176713574e-06,
+      "loss": 0.5201,
+      "step": 1694
+    },
+    {
+      "epoch": 0.678,
+      "grad_norm": 0.5193080902099609,
+      "learning_rate": 1.7369270839496581e-06,
+      "loss": 0.5159,
+      "step": 1695
+    },
+    {
+      "epoch": 0.6784,
+      "grad_norm": 0.4460771679878235,
+      "learning_rate": 1.7330116092382291e-06,
+      "loss": 0.521,
+      "step": 1696
+    },
+    {
+      "epoch": 0.6788,
+      "grad_norm": 0.49275845289230347,
+      "learning_rate": 1.7290991001085142e-06,
+      "loss": 0.4818,
+      "step": 1697
+    },
+    {
+      "epoch": 0.6792,
+      "grad_norm": 0.4485078752040863,
+      "learning_rate": 1.7251895631269744e-06,
+      "loss": 0.5208,
+      "step": 1698
+    },
+    {
+      "epoch": 0.6796,
+      "grad_norm": 0.5298824310302734,
+      "learning_rate": 1.7212830048550881e-06,
+      "loss": 0.5929,
+      "step": 1699
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.5249450206756592,
+      "learning_rate": 1.7173794318493306e-06,
+      "loss": 0.5053,
+      "step": 1700
+    },
+    {
+      "epoch": 0.6804,
+      "grad_norm": 0.5092018842697144,
+      "learning_rate": 1.713478850661167e-06,
+      "loss": 0.5591,
+      "step": 1701
+    },
+    {
+      "epoch": 0.6808,
+      "grad_norm": 0.522066056728363,
+      "learning_rate": 1.7095812678370429e-06,
+      "loss": 0.582,
+      "step": 1702
+    },
+    {
+      "epoch": 0.6812,
+      "grad_norm": 0.48720628023147583,
+      "learning_rate": 1.7056866899183727e-06,
+      "loss": 0.5319,
+      "step": 1703
+    },
+    {
+      "epoch": 0.6816,
+      "grad_norm": 0.5656712055206299,
+      "learning_rate": 1.701795123441525e-06,
+      "loss": 0.5974,
+      "step": 1704
+    },
+    {
+      "epoch": 0.682,
+      "grad_norm": 0.4969402551651001,
+      "learning_rate": 1.697906574937813e-06,
+      "loss": 0.558,
+      "step": 1705
+    },
+    {
+      "epoch": 0.6824,
+      "grad_norm": 0.5818575620651245,
+      "learning_rate": 1.6940210509334889e-06,
+      "loss": 0.628,
+      "step": 1706
+    },
+    {
+      "epoch": 0.6828,
+      "grad_norm": 0.4737216830253601,
+      "learning_rate": 1.6901385579497282e-06,
+      "loss": 0.5157,
+      "step": 1707
+    },
+    {
+      "epoch": 0.6832,
+      "grad_norm": 0.49812376499176025,
+      "learning_rate": 1.6862591025026157e-06,
+      "loss": 0.5326,
+      "step": 1708
+    },
+    {
+      "epoch": 0.6836,
+      "grad_norm": 0.5161290168762207,
+      "learning_rate": 1.6823826911031417e-06,
+      "loss": 0.6267,
+      "step": 1709
+    },
+    {
+      "epoch": 0.684,
+      "grad_norm": 0.5259966850280762,
+      "learning_rate": 1.6785093302571843e-06,
+      "loss": 0.6185,
+      "step": 1710
+    },
+    {
+      "epoch": 0.6844,
+      "grad_norm": 0.4863537847995758,
+      "learning_rate": 1.6746390264655055e-06,
+      "loss": 0.5257,
+      "step": 1711
+    },
+    {
+      "epoch": 0.6848,
+      "grad_norm": 0.5145780444145203,
+      "learning_rate": 1.6707717862237358e-06,
+      "loss": 0.519,
+      "step": 1712
+    },
+    {
+      "epoch": 0.6852,
+      "grad_norm": 0.48994773626327515,
+      "learning_rate": 1.6669076160223625e-06,
+      "loss": 0.441,
+      "step": 1713
+    },
+    {
+      "epoch": 0.6856,
+      "grad_norm": 0.5490663051605225,
+      "learning_rate": 1.6630465223467206e-06,
+      "loss": 0.5786,
+      "step": 1714
+    },
+    {
+      "epoch": 0.686,
+      "grad_norm": 0.4834598898887634,
+      "learning_rate": 1.6591885116769795e-06,
+      "loss": 0.5122,
+      "step": 1715
+    },
+    {
+      "epoch": 0.6864,
+      "grad_norm": 0.5213835835456848,
+      "learning_rate": 1.6553335904881427e-06,
+      "loss": 0.5692,
+      "step": 1716
+    },
+    {
+      "epoch": 0.6868,
+      "grad_norm": 0.5010420680046082,
+      "learning_rate": 1.6514817652500198e-06,
+      "loss": 0.5693,
+      "step": 1717
+    },
+    {
+      "epoch": 0.6872,
+      "grad_norm": 0.5002087354660034,
+      "learning_rate": 1.6476330424272277e-06,
+      "loss": 0.5556,
+      "step": 1718
+    },
+    {
+      "epoch": 0.6876,
+      "grad_norm": 0.4816271960735321,
+      "learning_rate": 1.6437874284791743e-06,
+      "loss": 0.5158,
+      "step": 1719
+    },
+    {
+      "epoch": 0.688,
+      "grad_norm": 0.4585452973842621,
+      "learning_rate": 1.6399449298600533e-06,
+      "loss": 0.523,
+      "step": 1720
+    },
+    {
+      "epoch": 0.6884,
+      "grad_norm": 0.47453686594963074,
+      "learning_rate": 1.6361055530188296e-06,
+      "loss": 0.5716,
+      "step": 1721
+    },
+    {
+      "epoch": 0.6888,
+      "grad_norm": 0.49845659732818604,
+      "learning_rate": 1.632269304399226e-06,
+      "loss": 0.4753,
+      "step": 1722
+    },
+    {
+      "epoch": 0.6892,
+      "grad_norm": 0.5592690706253052,
+      "learning_rate": 1.6284361904397163e-06,
+      "loss": 0.6543,
+      "step": 1723
+    },
+    {
+      "epoch": 0.6896,
+      "grad_norm": 0.5578603744506836,
+      "learning_rate": 1.6246062175735108e-06,
+      "loss": 0.5954,
+      "step": 1724
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.5509665012359619,
+      "learning_rate": 1.6207793922285566e-06,
+      "loss": 0.5676,
+      "step": 1725
+    },
+    {
+      "epoch": 0.6904,
+      "grad_norm": 0.5276341438293457,
+      "learning_rate": 1.6169557208275087e-06,
+      "loss": 0.5759,
+      "step": 1726
+    },
+    {
+      "epoch": 0.6908,
+      "grad_norm": 0.5588827729225159,
+      "learning_rate": 1.6131352097877332e-06,
+      "loss": 0.5559,
+      "step": 1727
+    },
+    {
+      "epoch": 0.6912,
+      "grad_norm": 0.5112521052360535,
+      "learning_rate": 1.6093178655212896e-06,
+      "loss": 0.5852,
+      "step": 1728
+    },
+    {
+      "epoch": 0.6916,
+      "grad_norm": 0.4992835819721222,
+      "learning_rate": 1.6055036944349258e-06,
+      "loss": 0.4917,
+      "step": 1729
+    },
+    {
+      "epoch": 0.692,
+      "grad_norm": 0.4967854917049408,
+      "learning_rate": 1.6016927029300634e-06,
+      "loss": 0.4255,
+      "step": 1730
+    },
+    {
+      "epoch": 0.6924,
+      "grad_norm": 0.5047206878662109,
+      "learning_rate": 1.5978848974027855e-06,
+      "loss": 0.5217,
+      "step": 1731
+    },
+    {
+      "epoch": 0.6928,
+      "grad_norm": 0.49704602360725403,
+      "learning_rate": 1.5940802842438283e-06,
+      "loss": 0.4804,
+      "step": 1732
+    },
+    {
+      "epoch": 0.6932,
+      "grad_norm": 0.509590208530426,
+      "learning_rate": 1.5902788698385736e-06,
+      "loss": 0.5788,
+      "step": 1733
+    },
+    {
+      "epoch": 0.6936,
+      "grad_norm": 0.5450599789619446,
+      "learning_rate": 1.5864806605670296e-06,
+      "loss": 0.5217,
+      "step": 1734
+    },
+    {
+      "epoch": 0.694,
+      "grad_norm": 0.49922382831573486,
+      "learning_rate": 1.58268566280383e-06,
+      "loss": 0.6029,
+      "step": 1735
+    },
+    {
+      "epoch": 0.6944,
+      "grad_norm": 0.5624175667762756,
+      "learning_rate": 1.5788938829182158e-06,
+      "loss": 0.6244,
+      "step": 1736
+    },
+    {
+      "epoch": 0.6948,
+      "grad_norm": 0.46818313002586365,
+      "learning_rate": 1.5751053272740265e-06,
+      "loss": 0.4942,
+      "step": 1737
+    },
+    {
+      "epoch": 0.6952,
+      "grad_norm": 0.6104135513305664,
+      "learning_rate": 1.571320002229693e-06,
+      "loss": 0.5757,
+      "step": 1738
+    },
+    {
+      "epoch": 0.6956,
+      "grad_norm": 0.5003908276557922,
+      "learning_rate": 1.5675379141382236e-06,
+      "loss": 0.5161,
+      "step": 1739
+    },
+    {
+      "epoch": 0.696,
+      "grad_norm": 0.48008689284324646,
+      "learning_rate": 1.5637590693471931e-06,
+      "loss": 0.5857,
+      "step": 1740
+    },
+    {
+      "epoch": 0.6964,
+      "grad_norm": 0.49035823345184326,
+      "learning_rate": 1.559983474198731e-06,
+      "loss": 0.4777,
+      "step": 1741
+    },
+    {
+      "epoch": 0.6968,
+      "grad_norm": 0.5132231116294861,
+      "learning_rate": 1.556211135029518e-06,
+      "loss": 0.5248,
+      "step": 1742
+    },
+    {
+      "epoch": 0.6972,
+      "grad_norm": 0.4993639290332794,
+      "learning_rate": 1.5524420581707644e-06,
+      "loss": 0.5057,
+      "step": 1743
+    },
+    {
+      "epoch": 0.6976,
+      "grad_norm": 0.5058372020721436,
+      "learning_rate": 1.5486762499482106e-06,
+      "loss": 0.5187,
+      "step": 1744
+    },
+    {
+      "epoch": 0.698,
+      "grad_norm": 0.4696761667728424,
+      "learning_rate": 1.5449137166821078e-06,
+      "loss": 0.5289,
+      "step": 1745
+    },
+    {
+      "epoch": 0.6984,
+      "grad_norm": 0.5460999011993408,
+      "learning_rate": 1.5411544646872094e-06,
+      "loss": 0.5688,
+      "step": 1746
+    },
+    {
+      "epoch": 0.6988,
+      "grad_norm": 0.47618138790130615,
+      "learning_rate": 1.5373985002727679e-06,
+      "loss": 0.5779,
+      "step": 1747
+    },
+    {
+      "epoch": 0.6992,
+      "grad_norm": 0.520785391330719,
+      "learning_rate": 1.5336458297425105e-06,
+      "loss": 0.5344,
+      "step": 1748
+    },
+    {
+      "epoch": 0.6996,
+      "grad_norm": 0.49487361311912537,
+      "learning_rate": 1.5298964593946434e-06,
+      "loss": 0.5133,
+      "step": 1749
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.6331418752670288,
+      "learning_rate": 1.526150395521827e-06,
+      "loss": 0.5969,
+      "step": 1750
+    },
+    {
+      "epoch": 0.7004,
+      "grad_norm": 0.513303816318512,
+      "learning_rate": 1.522407644411179e-06,
+      "loss": 0.5477,
+      "step": 1751
+    },
+    {
+      "epoch": 0.7008,
+      "grad_norm": 0.4873301386833191,
+      "learning_rate": 1.5186682123442518e-06,
+      "loss": 0.5146,
+      "step": 1752
+    },
+    {
+      "epoch": 0.7012,
+      "grad_norm": 0.5341429710388184,
+      "learning_rate": 1.5149321055970316e-06,
+      "loss": 0.5552,
+      "step": 1753
+    },
+    {
+      "epoch": 0.7016,
+      "grad_norm": 0.5356207489967346,
+      "learning_rate": 1.5111993304399213e-06,
+      "loss": 0.4732,
+      "step": 1754
+    },
+    {
+      "epoch": 0.702,
+      "grad_norm": 0.5175428986549377,
+      "learning_rate": 1.5074698931377304e-06,
+      "loss": 0.574,
+      "step": 1755
+    },
+    {
+      "epoch": 0.7024,
+      "grad_norm": 0.5094943642616272,
+      "learning_rate": 1.5037437999496718e-06,
+      "loss": 0.5922,
+      "step": 1756
+    },
+    {
+      "epoch": 0.7028,
+      "grad_norm": 0.4792037904262543,
+      "learning_rate": 1.5000210571293403e-06,
+      "loss": 0.4763,
+      "step": 1757
+    },
+    {
+      "epoch": 0.7032,
+      "grad_norm": 0.540791392326355,
+      "learning_rate": 1.4963016709247127e-06,
+      "loss": 0.6105,
+      "step": 1758
+    },
+    {
+      "epoch": 0.7036,
+      "grad_norm": 0.49768590927124023,
+      "learning_rate": 1.4925856475781271e-06,
+      "loss": 0.5767,
+      "step": 1759
+    },
+    {
+      "epoch": 0.704,
+      "grad_norm": 0.5273868441581726,
+      "learning_rate": 1.4888729933262833e-06,
+      "loss": 0.5907,
+      "step": 1760
+    },
+    {
+      "epoch": 0.7044,
+      "grad_norm": 0.5667710900306702,
+      "learning_rate": 1.4851637144002219e-06,
+      "loss": 0.5502,
+      "step": 1761
+    },
+    {
+      "epoch": 0.7048,
+      "grad_norm": 0.5382874608039856,
+      "learning_rate": 1.4814578170253192e-06,
+      "loss": 0.5999,
+      "step": 1762
+    },
+    {
+      "epoch": 0.7052,
+      "grad_norm": 0.4984513819217682,
+      "learning_rate": 1.47775530742128e-06,
+      "loss": 0.4988,
+      "step": 1763
+    },
+    {
+      "epoch": 0.7056,
+      "grad_norm": 0.4884515404701233,
+      "learning_rate": 1.4740561918021178e-06,
+      "loss": 0.474,
+      "step": 1764
+    },
+    {
+      "epoch": 0.706,
+      "grad_norm": 0.46463531255722046,
+      "learning_rate": 1.470360476376156e-06,
+      "loss": 0.4821,
+      "step": 1765
+    },
+    {
+      "epoch": 0.7064,
+      "grad_norm": 0.521081805229187,
+      "learning_rate": 1.466668167346005e-06,
+      "loss": 0.5404,
+      "step": 1766
+    },
+    {
+      "epoch": 0.7068,
+      "grad_norm": 0.5084238648414612,
+      "learning_rate": 1.462979270908564e-06,
+      "loss": 0.5357,
+      "step": 1767
+    },
+    {
+      "epoch": 0.7072,
+      "grad_norm": 0.5334984064102173,
+      "learning_rate": 1.4592937932549993e-06,
+      "loss": 0.5774,
+      "step": 1768
+    },
+    {
+      "epoch": 0.7076,
+      "grad_norm": 0.5306082963943481,
+      "learning_rate": 1.455611740570745e-06,
+      "loss": 0.5521,
+      "step": 1769
+    },
+    {
+      "epoch": 0.708,
+      "grad_norm": 0.5435188412666321,
+      "learning_rate": 1.4519331190354828e-06,
+      "loss": 0.6215,
+      "step": 1770
+    },
+    {
+      "epoch": 0.7084,
+      "grad_norm": 0.46188175678253174,
+      "learning_rate": 1.4482579348231357e-06,
+      "loss": 0.5283,
+      "step": 1771
+    },
+    {
+      "epoch": 0.7088,
+      "grad_norm": 0.47881054878234863,
+      "learning_rate": 1.4445861941018614e-06,
+      "loss": 0.4803,
+      "step": 1772
+    },
+    {
+      "epoch": 0.7092,
+      "grad_norm": 0.5058978796005249,
+      "learning_rate": 1.4409179030340343e-06,
+      "loss": 0.5497,
+      "step": 1773
+    },
+    {
+      "epoch": 0.7096,
+      "grad_norm": 0.4847659468650818,
+      "learning_rate": 1.4372530677762425e-06,
+      "loss": 0.5812,
+      "step": 1774
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.4816116690635681,
+      "learning_rate": 1.4335916944792716e-06,
+      "loss": 0.5475,
+      "step": 1775
+    },
+    {
+      "epoch": 0.7104,
+      "grad_norm": 0.5226293206214905,
+      "learning_rate": 1.429933789288097e-06,
+      "loss": 0.5325,
+      "step": 1776
+    },
+    {
+      "epoch": 0.7108,
+      "grad_norm": 0.5266181230545044,
+      "learning_rate": 1.4262793583418757e-06,
+      "loss": 0.5531,
+      "step": 1777
+    },
+    {
+      "epoch": 0.7112,
+      "grad_norm": 0.49322509765625,
+      "learning_rate": 1.422628407773933e-06,
+      "loss": 0.544,
+      "step": 1778
+    },
+    {
+      "epoch": 0.7116,
+      "grad_norm": 0.5125441551208496,
+      "learning_rate": 1.4189809437117513e-06,
+      "loss": 0.5705,
+      "step": 1779
+    },
+    {
+      "epoch": 0.712,
+      "grad_norm": 0.449192613363266,
+      "learning_rate": 1.4153369722769625e-06,
+      "loss": 0.5128,
+      "step": 1780
+    },
+    {
+      "epoch": 0.7124,
+      "grad_norm": 0.4418649673461914,
+      "learning_rate": 1.4116964995853365e-06,
+      "loss": 0.4764,
+      "step": 1781
+    },
+    {
+      "epoch": 0.7128,
+      "grad_norm": 0.4699050784111023,
+      "learning_rate": 1.4080595317467722e-06,
+      "loss": 0.4901,
+      "step": 1782
+    },
+    {
+      "epoch": 0.7132,
+      "grad_norm": 0.4823940098285675,
+      "learning_rate": 1.4044260748652863e-06,
+      "loss": 0.5816,
+      "step": 1783
+    },
+    {
+      "epoch": 0.7136,
+      "grad_norm": 0.5200097560882568,
+      "learning_rate": 1.4007961350390021e-06,
+      "loss": 0.5738,
+      "step": 1784
+    },
+    {
+      "epoch": 0.714,
+      "grad_norm": 0.4924733638763428,
+      "learning_rate": 1.3971697183601376e-06,
+      "loss": 0.4668,
+      "step": 1785
+    },
+    {
+      "epoch": 0.7144,
+      "grad_norm": 0.47781214118003845,
+      "learning_rate": 1.3935468309150032e-06,
+      "loss": 0.5402,
+      "step": 1786
+    },
+    {
+      "epoch": 0.7148,
+      "grad_norm": 0.537818968296051,
+      "learning_rate": 1.3899274787839834e-06,
+      "loss": 0.5018,
+      "step": 1787
+    },
+    {
+      "epoch": 0.7152,
+      "grad_norm": 0.545964241027832,
+      "learning_rate": 1.3863116680415285e-06,
+      "loss": 0.5537,
+      "step": 1788
+    },
+    {
+      "epoch": 0.7156,
+      "grad_norm": 0.4749377369880676,
+      "learning_rate": 1.3826994047561462e-06,
+      "loss": 0.4985,
+      "step": 1789
+    },
+    {
+      "epoch": 0.716,
+      "grad_norm": 0.4744401276111603,
+      "learning_rate": 1.3790906949903884e-06,
+      "loss": 0.4716,
+      "step": 1790
+    },
+    {
+      "epoch": 0.7164,
+      "grad_norm": 0.5007680654525757,
+      "learning_rate": 1.3754855448008464e-06,
+      "loss": 0.5631,
+      "step": 1791
+    },
+    {
+      "epoch": 0.7168,
+      "grad_norm": 0.5083097815513611,
+      "learning_rate": 1.3718839602381367e-06,
+      "loss": 0.5688,
+      "step": 1792
+    },
+    {
+      "epoch": 0.7172,
+      "grad_norm": 0.5184915065765381,
+      "learning_rate": 1.3682859473468897e-06,
+      "loss": 0.5933,
+      "step": 1793
+    },
+    {
+      "epoch": 0.7176,
+      "grad_norm": 0.4937378466129303,
+      "learning_rate": 1.364691512165741e-06,
+      "loss": 0.5165,
+      "step": 1794
+    },
+    {
+      "epoch": 0.718,
+      "grad_norm": 0.4511565566062927,
+      "learning_rate": 1.3611006607273242e-06,
+      "loss": 0.483,
+      "step": 1795
+    },
+    {
+      "epoch": 0.7184,
+      "grad_norm": 0.5037986040115356,
+      "learning_rate": 1.3575133990582574e-06,
+      "loss": 0.539,
+      "step": 1796
+    },
+    {
+      "epoch": 0.7188,
+      "grad_norm": 0.5806722044944763,
+      "learning_rate": 1.353929733179133e-06,
+      "loss": 0.5049,
+      "step": 1797
+    },
+    {
+      "epoch": 0.7192,
+      "grad_norm": 0.5695548057556152,
+      "learning_rate": 1.3503496691045087e-06,
+      "loss": 0.6149,
+      "step": 1798
+    },
+    {
+      "epoch": 0.7196,
+      "grad_norm": 0.5636582374572754,
+      "learning_rate": 1.346773212842896e-06,
+      "loss": 0.6077,
+      "step": 1799
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.5878316164016724,
+      "learning_rate": 1.3432003703967542e-06,
+      "loss": 0.5881,
+      "step": 1800
+    },
+    {
+      "epoch": 0.7204,
+      "grad_norm": 0.44944626092910767,
+      "learning_rate": 1.339631147762477e-06,
+      "loss": 0.4826,
+      "step": 1801
+    },
+    {
+      "epoch": 0.7208,
+      "grad_norm": 0.551231861114502,
+      "learning_rate": 1.336065550930381e-06,
+      "loss": 0.5654,
+      "step": 1802
+    },
+    {
+      "epoch": 0.7212,
+      "grad_norm": 0.4855599105358124,
+      "learning_rate": 1.3325035858846964e-06,
+      "loss": 0.5741,
+      "step": 1803
+    },
+    {
+      "epoch": 0.7216,
+      "grad_norm": 0.47175052762031555,
+      "learning_rate": 1.328945258603562e-06,
+      "loss": 0.5032,
+      "step": 1804
+    },
+    {
+      "epoch": 0.722,
+      "grad_norm": 0.5383630394935608,
+      "learning_rate": 1.3253905750590099e-06,
+      "loss": 0.5234,
+      "step": 1805
+    },
+    {
+      "epoch": 0.7224,
+      "grad_norm": 0.4824926257133484,
+      "learning_rate": 1.3218395412169554e-06,
+      "loss": 0.5013,
+      "step": 1806
+    },
+    {
+      "epoch": 0.7228,
+      "grad_norm": 0.5814263224601746,
+      "learning_rate": 1.3182921630371892e-06,
+      "loss": 0.6488,
+      "step": 1807
+    },
+    {
+      "epoch": 0.7232,
+      "grad_norm": 0.480724036693573,
+      "learning_rate": 1.314748446473365e-06,
+      "loss": 0.4736,
+      "step": 1808
+    },
+    {
+      "epoch": 0.7236,
+      "grad_norm": 0.5075566172599792,
+      "learning_rate": 1.3112083974729947e-06,
+      "loss": 0.52,
+      "step": 1809
+    },
+    {
+      "epoch": 0.724,
+      "grad_norm": 0.523950457572937,
+      "learning_rate": 1.3076720219774333e-06,
+      "loss": 0.5094,
+      "step": 1810
+    },
+    {
+      "epoch": 0.7244,
+      "grad_norm": 0.45218196511268616,
+      "learning_rate": 1.3041393259218693e-06,
+      "loss": 0.4614,
+      "step": 1811
+    },
+    {
+      "epoch": 0.7248,
+      "grad_norm": 0.5422837138175964,
+      "learning_rate": 1.3006103152353158e-06,
+      "loss": 0.5605,
+      "step": 1812
+    },
+    {
+      "epoch": 0.7252,
+      "grad_norm": 0.46532970666885376,
+      "learning_rate": 1.297084995840602e-06,
+      "loss": 0.5027,
+      "step": 1813
+    },
+    {
+      "epoch": 0.7256,
+      "grad_norm": 0.5618458390235901,
+      "learning_rate": 1.2935633736543628e-06,
+      "loss": 0.514,
+      "step": 1814
+    },
+    {
+      "epoch": 0.726,
+      "grad_norm": 0.5310685634613037,
+      "learning_rate": 1.2900454545870257e-06,
+      "loss": 0.5001,
+      "step": 1815
+    },
+    {
+      "epoch": 0.7264,
+      "grad_norm": 0.5191671252250671,
+      "learning_rate": 1.2865312445428036e-06,
+      "loss": 0.5151,
+      "step": 1816
+    },
+    {
+      "epoch": 0.7268,
+      "grad_norm": 0.5187528133392334,
+      "learning_rate": 1.283020749419684e-06,
+      "loss": 0.4881,
+      "step": 1817
+    },
+    {
+      "epoch": 0.7272,
+      "grad_norm": 0.4712491035461426,
+      "learning_rate": 1.2795139751094222e-06,
+      "loss": 0.4651,
+      "step": 1818
+    },
+    {
+      "epoch": 0.7276,
+      "grad_norm": 0.4946507215499878,
+      "learning_rate": 1.276010927497527e-06,
+      "loss": 0.6037,
+      "step": 1819
+    },
+    {
+      "epoch": 0.728,
+      "grad_norm": 0.4676782786846161,
+      "learning_rate": 1.272511612463252e-06,
+      "loss": 0.5508,
+      "step": 1820
+    },
+    {
+      "epoch": 0.7284,
+      "grad_norm": 0.5122756361961365,
+      "learning_rate": 1.2690160358795858e-06,
+      "loss": 0.5076,
+      "step": 1821
+    },
+    {
+      "epoch": 0.7288,
+      "grad_norm": 0.5033440589904785,
+      "learning_rate": 1.2655242036132466e-06,
+      "loss": 0.5678,
+      "step": 1822
+    },
+    {
+      "epoch": 0.7292,
+      "grad_norm": 0.46391579508781433,
+      "learning_rate": 1.2620361215246624e-06,
+      "loss": 0.5384,
+      "step": 1823
+    },
+    {
+      "epoch": 0.7296,
+      "grad_norm": 0.5027047991752625,
+      "learning_rate": 1.258551795467973e-06,
+      "loss": 0.4817,
+      "step": 1824
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.4803715646266937,
+      "learning_rate": 1.255071231291011e-06,
+      "loss": 0.5046,
+      "step": 1825
+    },
+    {
+      "epoch": 0.7304,
+      "grad_norm": 0.49049100279808044,
+      "learning_rate": 1.2515944348352947e-06,
+      "loss": 0.5937,
+      "step": 1826
+    },
+    {
+      "epoch": 0.7308,
+      "grad_norm": 0.5216216444969177,
+      "learning_rate": 1.2481214119360212e-06,
+      "loss": 0.5592,
+      "step": 1827
+    },
+    {
+      "epoch": 0.7312,
+      "grad_norm": 0.4808286130428314,
+      "learning_rate": 1.244652168422055e-06,
+      "loss": 0.5259,
+      "step": 1828
+    },
+    {
+      "epoch": 0.7316,
+      "grad_norm": 0.49704375863075256,
+      "learning_rate": 1.2411867101159147e-06,
+      "loss": 0.5167,
+      "step": 1829
+    },
+    {
+      "epoch": 0.732,
+      "grad_norm": 0.5660054087638855,
+      "learning_rate": 1.2377250428337665e-06,
+      "loss": 0.5109,
+      "step": 1830
+    },
+    {
+      "epoch": 0.7324,
+      "grad_norm": 0.5108336210250854,
+      "learning_rate": 1.2342671723854164e-06,
+      "loss": 0.5556,
+      "step": 1831
+    },
+    {
+      "epoch": 0.7328,
+      "grad_norm": 0.4856356084346771,
+      "learning_rate": 1.2308131045742956e-06,
+      "loss": 0.528,
+      "step": 1832
+    },
+    {
+      "epoch": 0.7332,
+      "grad_norm": 0.44966772198677063,
+      "learning_rate": 1.2273628451974558e-06,
+      "loss": 0.5224,
+      "step": 1833
+    },
+    {
+      "epoch": 0.7336,
+      "grad_norm": 0.5239558815956116,
+      "learning_rate": 1.2239164000455528e-06,
+      "loss": 0.5129,
+      "step": 1834
+    },
+    {
+      "epoch": 0.734,
+      "grad_norm": 0.5169456005096436,
+      "learning_rate": 1.2204737749028466e-06,
+      "loss": 0.5931,
+      "step": 1835
+    },
+    {
+      "epoch": 0.7344,
+      "grad_norm": 0.441876620054245,
+      "learning_rate": 1.217034975547181e-06,
+      "loss": 0.4571,
+      "step": 1836
+    },
+    {
+      "epoch": 0.7348,
+      "grad_norm": 0.5893264412879944,
+      "learning_rate": 1.21360000774998e-06,
+      "loss": 0.5479,
+      "step": 1837
+    },
+    {
+      "epoch": 0.7352,
+      "grad_norm": 0.5525321960449219,
+      "learning_rate": 1.2101688772762398e-06,
+      "loss": 0.5201,
+      "step": 1838
+    },
+    {
+      "epoch": 0.7356,
+      "grad_norm": 0.5001885294914246,
+      "learning_rate": 1.2067415898845114e-06,
+      "loss": 0.5322,
+      "step": 1839
+    },
+    {
+      "epoch": 0.736,
+      "grad_norm": 0.5684386491775513,
+      "learning_rate": 1.2033181513269025e-06,
+      "loss": 0.5115,
+      "step": 1840
+    },
+    {
+      "epoch": 0.7364,
+      "grad_norm": 0.6078344583511353,
+      "learning_rate": 1.1998985673490533e-06,
+      "loss": 0.6907,
+      "step": 1841
+    },
+    {
+      "epoch": 0.7368,
+      "grad_norm": 0.5336594581604004,
+      "learning_rate": 1.196482843690143e-06,
+      "loss": 0.5895,
+      "step": 1842
+    },
+    {
+      "epoch": 0.7372,
+      "grad_norm": 0.48478782176971436,
+      "learning_rate": 1.1930709860828638e-06,
+      "loss": 0.4413,
+      "step": 1843
+    },
+    {
+      "epoch": 0.7376,
+      "grad_norm": 0.5144538879394531,
+      "learning_rate": 1.1896630002534264e-06,
+      "loss": 0.5178,
+      "step": 1844
+    },
+    {
+      "epoch": 0.738,
+      "grad_norm": 0.4920802712440491,
+      "learning_rate": 1.1862588919215395e-06,
+      "loss": 0.5442,
+      "step": 1845
+    },
+    {
+      "epoch": 0.7384,
+      "grad_norm": 0.46744486689567566,
+      "learning_rate": 1.1828586668004037e-06,
+      "loss": 0.4624,
+      "step": 1846
+    },
+    {
+      "epoch": 0.7388,
+      "grad_norm": 0.4999946355819702,
+      "learning_rate": 1.1794623305967057e-06,
+      "loss": 0.4912,
+      "step": 1847
+    },
+    {
+      "epoch": 0.7392,
+      "grad_norm": 0.5071706175804138,
+      "learning_rate": 1.1760698890106015e-06,
+      "loss": 0.5316,
+      "step": 1848
+    },
+    {
+      "epoch": 0.7396,
+      "grad_norm": 0.5435724854469299,
+      "learning_rate": 1.1726813477357138e-06,
+      "loss": 0.5537,
+      "step": 1849
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.5288483500480652,
+      "learning_rate": 1.169296712459117e-06,
+      "loss": 0.549,
+      "step": 1850
+    },
+    {
+      "epoch": 0.7404,
+      "grad_norm": 0.5118080377578735,
+      "learning_rate": 1.16591598886133e-06,
+      "loss": 0.51,
+      "step": 1851
+    },
+    {
+      "epoch": 0.7408,
+      "grad_norm": 0.49369898438453674,
+      "learning_rate": 1.162539182616309e-06,
+      "loss": 0.5163,
+      "step": 1852
+    },
+    {
+      "epoch": 0.7412,
+      "grad_norm": 0.4700876474380493,
+      "learning_rate": 1.1591662993914344e-06,
+      "loss": 0.4869,
+      "step": 1853
+    },
+    {
+      "epoch": 0.7416,
+      "grad_norm": 0.4725787937641144,
+      "learning_rate": 1.1557973448475015e-06,
+      "loss": 0.4726,
+      "step": 1854
+    },
+    {
+      "epoch": 0.742,
+      "grad_norm": 0.5252476334571838,
+      "learning_rate": 1.1524323246387127e-06,
+      "loss": 0.5113,
+      "step": 1855
+    },
+    {
+      "epoch": 0.7424,
+      "grad_norm": 0.5266048908233643,
+      "learning_rate": 1.1490712444126662e-06,
+      "loss": 0.5853,
+      "step": 1856
+    },
+    {
+      "epoch": 0.7428,
+      "grad_norm": 0.49874410033226013,
+      "learning_rate": 1.1457141098103494e-06,
+      "loss": 0.4883,
+      "step": 1857
+    },
+    {
+      "epoch": 0.7432,
+      "grad_norm": 0.4535691738128662,
+      "learning_rate": 1.142360926466129e-06,
+      "loss": 0.426,
+      "step": 1858
+    },
+    {
+      "epoch": 0.7436,
+      "grad_norm": 0.49276694655418396,
+      "learning_rate": 1.139011700007736e-06,
+      "loss": 0.5052,
+      "step": 1859
+    },
+    {
+      "epoch": 0.744,
+      "grad_norm": 0.4988146126270294,
+      "learning_rate": 1.1356664360562627e-06,
+      "loss": 0.5846,
+      "step": 1860
+    },
+    {
+      "epoch": 0.7444,
+      "grad_norm": 0.574359118938446,
+      "learning_rate": 1.1323251402261509e-06,
+      "loss": 0.5119,
+      "step": 1861
+    },
+    {
+      "epoch": 0.7448,
+      "grad_norm": 0.5316257476806641,
+      "learning_rate": 1.1289878181251847e-06,
+      "loss": 0.5468,
+      "step": 1862
+    },
+    {
+      "epoch": 0.7452,
+      "grad_norm": 0.4588572680950165,
+      "learning_rate": 1.1256544753544755e-06,
+      "loss": 0.4288,
+      "step": 1863
+    },
+    {
+      "epoch": 0.7456,
+      "grad_norm": 0.5082074403762817,
+      "learning_rate": 1.1223251175084573e-06,
+      "loss": 0.5143,
+      "step": 1864
+    },
+    {
+      "epoch": 0.746,
+      "grad_norm": 0.43266239762306213,
+      "learning_rate": 1.1189997501748754e-06,
+      "loss": 0.4633,
+      "step": 1865
+    },
+    {
+      "epoch": 0.7464,
+      "grad_norm": 0.5357276201248169,
+      "learning_rate": 1.11567837893478e-06,
+      "loss": 0.5964,
+      "step": 1866
+    },
+    {
+      "epoch": 0.7468,
+      "grad_norm": 0.5177252292633057,
+      "learning_rate": 1.112361009362514e-06,
+      "loss": 0.5603,
+      "step": 1867
+    },
+    {
+      "epoch": 0.7472,
+      "grad_norm": 0.4995522201061249,
+      "learning_rate": 1.1090476470257024e-06,
+      "loss": 0.5252,
+      "step": 1868
+    },
+    {
+      "epoch": 0.7476,
+      "grad_norm": 0.5316324234008789,
+      "learning_rate": 1.1057382974852448e-06,
+      "loss": 0.4449,
+      "step": 1869
+    },
+    {
+      "epoch": 0.748,
+      "grad_norm": 0.5401836633682251,
+      "learning_rate": 1.1024329662953083e-06,
+      "loss": 0.5806,
+      "step": 1870
+    },
+    {
+      "epoch": 0.7484,
+      "grad_norm": 0.4464450180530548,
+      "learning_rate": 1.0991316590033152e-06,
+      "loss": 0.4032,
+      "step": 1871
+    },
+    {
+      "epoch": 0.7488,
+      "grad_norm": 0.5609473586082458,
+      "learning_rate": 1.095834381149933e-06,
+      "loss": 0.5282,
+      "step": 1872
+    },
+    {
+      "epoch": 0.7492,
+      "grad_norm": 0.46891769766807556,
+      "learning_rate": 1.0925411382690669e-06,
+      "loss": 0.5425,
+      "step": 1873
+    },
+    {
+      "epoch": 0.7496,
+      "grad_norm": 0.5601791739463806,
+      "learning_rate": 1.0892519358878497e-06,
+      "loss": 0.6361,
+      "step": 1874
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.5295112729072571,
+      "learning_rate": 1.0859667795266348e-06,
+      "loss": 0.5997,
+      "step": 1875
+    },
+    {
+      "epoch": 0.7504,
+      "grad_norm": 0.5210896730422974,
+      "learning_rate": 1.0826856746989853e-06,
+      "loss": 0.5908,
+      "step": 1876
+    },
+    {
+      "epoch": 0.7508,
+      "grad_norm": 0.5097439885139465,
+      "learning_rate": 1.0794086269116617e-06,
+      "loss": 0.5604,
+      "step": 1877
+    },
+    {
+      "epoch": 0.7512,
+      "grad_norm": 0.492924302816391,
+      "learning_rate": 1.076135641664616e-06,
+      "loss": 0.5139,
+      "step": 1878
+    },
+    {
+      "epoch": 0.7516,
+      "grad_norm": 0.5186163783073425,
+      "learning_rate": 1.0728667244509831e-06,
+      "loss": 0.5638,
+      "step": 1879
+    },
+    {
+      "epoch": 0.752,
+      "grad_norm": 0.5136158466339111,
+      "learning_rate": 1.069601880757073e-06,
+      "loss": 0.5374,
+      "step": 1880
+    },
+    {
+      "epoch": 0.7524,
+      "grad_norm": 0.5937939882278442,
+      "learning_rate": 1.066341116062354e-06,
+      "loss": 0.6895,
+      "step": 1881
+    },
+    {
+      "epoch": 0.7528,
+      "grad_norm": 0.4891611933708191,
+      "learning_rate": 1.0630844358394507e-06,
+      "loss": 0.5208,
+      "step": 1882
+    },
+    {
+      "epoch": 0.7532,
+      "grad_norm": 0.43400999903678894,
+      "learning_rate": 1.0598318455541317e-06,
+      "loss": 0.4802,
+      "step": 1883
+    },
+    {
+      "epoch": 0.7536,
+      "grad_norm": 0.5274035930633545,
+      "learning_rate": 1.0565833506653034e-06,
+      "loss": 0.5246,
+      "step": 1884
+    },
+    {
+      "epoch": 0.754,
+      "grad_norm": 0.538054883480072,
+      "learning_rate": 1.0533389566249991e-06,
+      "loss": 0.5674,
+      "step": 1885
+    },
+    {
+      "epoch": 0.7544,
+      "grad_norm": 0.44231492280960083,
+      "learning_rate": 1.0500986688783665e-06,
+      "loss": 0.4748,
+      "step": 1886
+    },
+    {
+      "epoch": 0.7548,
+      "grad_norm": 0.5154666900634766,
+      "learning_rate": 1.0468624928636623e-06,
+      "loss": 0.619,
+      "step": 1887
+    },
+    {
+      "epoch": 0.7552,
+      "grad_norm": 0.5303316712379456,
+      "learning_rate": 1.0436304340122446e-06,
+      "loss": 0.5738,
+      "step": 1888
+    },
+    {
+      "epoch": 0.7556,
+      "grad_norm": 0.5067679286003113,
+      "learning_rate": 1.0404024977485612e-06,
+      "loss": 0.498,
+      "step": 1889
+    },
+    {
+      "epoch": 0.756,
+      "grad_norm": 0.5611936450004578,
+      "learning_rate": 1.0371786894901385e-06,
+      "loss": 0.6098,
+      "step": 1890
+    },
+    {
+      "epoch": 0.7564,
+      "grad_norm": 0.5335197448730469,
+      "learning_rate": 1.0339590146475765e-06,
+      "loss": 0.593,
+      "step": 1891
+    },
+    {
+      "epoch": 0.7568,
+      "grad_norm": 0.48436692357063293,
+      "learning_rate": 1.030743478624537e-06,
+      "loss": 0.5192,
+      "step": 1892
+    },
+    {
+      "epoch": 0.7572,
+      "grad_norm": 0.4684750437736511,
+      "learning_rate": 1.0275320868177364e-06,
+      "loss": 0.5287,
+      "step": 1893
+    },
+    {
+      "epoch": 0.7576,
+      "grad_norm": 0.5043955445289612,
+      "learning_rate": 1.0243248446169373e-06,
+      "loss": 0.5264,
+      "step": 1894
+    },
+    {
+      "epoch": 0.758,
+      "grad_norm": 0.5075042247772217,
+      "learning_rate": 1.0211217574049342e-06,
+      "loss": 0.584,
+      "step": 1895
+    },
+    {
+      "epoch": 0.7584,
+      "grad_norm": 0.5217066407203674,
+      "learning_rate": 1.0179228305575502e-06,
+      "loss": 0.5258,
+      "step": 1896
+    },
+    {
+      "epoch": 0.7588,
+      "grad_norm": 0.5070008635520935,
+      "learning_rate": 1.0147280694436272e-06,
+      "loss": 0.5319,
+      "step": 1897
+    },
+    {
+      "epoch": 0.7592,
+      "grad_norm": 0.49269282817840576,
+      "learning_rate": 1.0115374794250126e-06,
+      "loss": 0.4965,
+      "step": 1898
+    },
+    {
+      "epoch": 0.7596,
+      "grad_norm": 0.5396968722343445,
+      "learning_rate": 1.0083510658565568e-06,
+      "loss": 0.4938,
+      "step": 1899
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.5112387537956238,
+      "learning_rate": 1.0051688340860985e-06,
+      "loss": 0.5559,
+      "step": 1900
+    },
+    {
+      "epoch": 0.7604,
+      "grad_norm": 0.5223495960235596,
+      "learning_rate": 1.0019907894544567e-06,
+      "loss": 0.5585,
+      "step": 1901
+    },
+    {
+      "epoch": 0.7608,
+      "grad_norm": 0.47718486189842224,
+      "learning_rate": 9.98816937295426e-07,
+      "loss": 0.5632,
+      "step": 1902
+    },
+    {
+      "epoch": 0.7612,
+      "grad_norm": 0.5641752481460571,
+      "learning_rate": 9.956472829357654e-07,
+      "loss": 0.541,
+      "step": 1903
+    },
+    {
+      "epoch": 0.7616,
+      "grad_norm": 0.5478566884994507,
+      "learning_rate": 9.924818316951848e-07,
+      "loss": 0.5653,
+      "step": 1904
+    },
+    {
+      "epoch": 0.762,
+      "grad_norm": 0.5488006472587585,
+      "learning_rate": 9.893205888863411e-07,
+      "loss": 0.5503,
+      "step": 1905
+    },
+    {
+      "epoch": 0.7624,
+      "grad_norm": 0.5716015696525574,
+      "learning_rate": 9.86163559814831e-07,
+      "loss": 0.5462,
+      "step": 1906
+    },
+    {
+      "epoch": 0.7628,
+      "grad_norm": 0.5241632461547852,
+      "learning_rate": 9.830107497791743e-07,
+      "loss": 0.504,
+      "step": 1907
+    },
+    {
+      "epoch": 0.7632,
+      "grad_norm": 0.5002646446228027,
+      "learning_rate": 9.798621640708148e-07,
+      "loss": 0.5395,
+      "step": 1908
+    },
+    {
+      "epoch": 0.7636,
+      "grad_norm": 0.5340526103973389,
+      "learning_rate": 9.76717807974104e-07,
+      "loss": 0.5614,
+      "step": 1909
+    },
+    {
+      "epoch": 0.764,
+      "grad_norm": 0.45730161666870117,
+      "learning_rate": 9.73577686766293e-07,
+      "loss": 0.4557,
+      "step": 1910
+    },
+    {
+      "epoch": 0.7644,
+      "grad_norm": 0.5114016532897949,
+      "learning_rate": 9.704418057175296e-07,
+      "loss": 0.5585,
+      "step": 1911
+    },
+    {
+      "epoch": 0.7648,
+      "grad_norm": 0.5574830770492554,
+      "learning_rate": 9.673101700908416e-07,
+      "loss": 0.6155,
+      "step": 1912
+    },
+    {
+      "epoch": 0.7652,
+      "grad_norm": 0.5143188238143921,
+      "learning_rate": 9.641827851421342e-07,
+      "loss": 0.559,
+      "step": 1913
+    },
+    {
+      "epoch": 0.7656,
+      "grad_norm": 0.5527118444442749,
+      "learning_rate": 9.610596561201755e-07,
+      "loss": 0.5481,
+      "step": 1914
+    },
+    {
+      "epoch": 0.766,
+      "grad_norm": 0.5203020572662354,
+      "learning_rate": 9.579407882665957e-07,
+      "loss": 0.4671,
+      "step": 1915
+    },
+    {
+      "epoch": 0.7664,
+      "grad_norm": 0.5298575758934021,
+      "learning_rate": 9.548261868158677e-07,
+      "loss": 0.4934,
+      "step": 1916
+    },
+    {
+      "epoch": 0.7668,
+      "grad_norm": 0.48027393221855164,
+      "learning_rate": 9.517158569953065e-07,
+      "loss": 0.5464,
+      "step": 1917
+    },
+    {
+      "epoch": 0.7672,
+      "grad_norm": 0.45088398456573486,
+      "learning_rate": 9.486098040250603e-07,
+      "loss": 0.525,
+      "step": 1918
+    },
+    {
+      "epoch": 0.7676,
+      "grad_norm": 0.488944411277771,
+      "learning_rate": 9.455080331180945e-07,
+      "loss": 0.4518,
+      "step": 1919
+    },
+    {
+      "epoch": 0.768,
+      "grad_norm": 0.4816155433654785,
+      "learning_rate": 9.424105494801924e-07,
+      "loss": 0.4794,
+      "step": 1920
+    },
+    {
+      "epoch": 0.7684,
+      "grad_norm": 0.5139371752738953,
+      "learning_rate": 9.393173583099383e-07,
+      "loss": 0.544,
+      "step": 1921
+    },
+    {
+      "epoch": 0.7688,
+      "grad_norm": 0.5136451125144958,
+      "learning_rate": 9.362284647987152e-07,
+      "loss": 0.5746,
+      "step": 1922
+    },
+    {
+      "epoch": 0.7692,
+      "grad_norm": 0.514636218547821,
+      "learning_rate": 9.331438741306904e-07,
+      "loss": 0.5222,
+      "step": 1923
+    },
+    {
+      "epoch": 0.7696,
+      "grad_norm": 0.5739938020706177,
+      "learning_rate": 9.300635914828133e-07,
+      "loss": 0.5859,
+      "step": 1924
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.47024837136268616,
+      "learning_rate": 9.269876220247995e-07,
+      "loss": 0.4757,
+      "step": 1925
+    },
+    {
+      "epoch": 0.7704,
+      "grad_norm": 0.4559095799922943,
+      "learning_rate": 9.239159709191258e-07,
+      "loss": 0.5346,
+      "step": 1926
+    },
+    {
+      "epoch": 0.7708,
+      "grad_norm": 0.5137647390365601,
+      "learning_rate": 9.208486433210255e-07,
+      "loss": 0.5204,
+      "step": 1927
+    },
+    {
+      "epoch": 0.7712,
+      "grad_norm": 0.47413358092308044,
+      "learning_rate": 9.177856443784699e-07,
+      "loss": 0.5305,
+      "step": 1928
+    },
+    {
+      "epoch": 0.7716,
+      "grad_norm": 0.5467235445976257,
+      "learning_rate": 9.147269792321709e-07,
+      "loss": 0.6016,
+      "step": 1929
+    },
+    {
+      "epoch": 0.772,
+      "grad_norm": 0.4952782094478607,
+      "learning_rate": 9.116726530155633e-07,
+      "loss": 0.5398,
+      "step": 1930
+    },
+    {
+      "epoch": 0.7724,
+      "grad_norm": 0.518825113773346,
+      "learning_rate": 9.086226708547993e-07,
+      "loss": 0.6043,
+      "step": 1931
+    },
+    {
+      "epoch": 0.7728,
+      "grad_norm": 0.5249910950660706,
+      "learning_rate": 9.055770378687427e-07,
+      "loss": 0.5465,
+      "step": 1932
+    },
+    {
+      "epoch": 0.7732,
+      "grad_norm": 0.5710473656654358,
+      "learning_rate": 9.02535759168959e-07,
+      "loss": 0.5921,
+      "step": 1933
+    },
+    {
+      "epoch": 0.7736,
+      "grad_norm": 0.4922305643558502,
+      "learning_rate": 8.994988398597018e-07,
+      "loss": 0.5021,
+      "step": 1934
+    },
+    {
+      "epoch": 0.774,
+      "grad_norm": 0.5559587478637695,
+      "learning_rate": 8.9646628503791e-07,
+      "loss": 0.5457,
+      "step": 1935
+    },
+    {
+      "epoch": 0.7744,
+      "grad_norm": 0.5352874994277954,
+      "learning_rate": 8.934380997931993e-07,
+      "loss": 0.5983,
+      "step": 1936
+    },
+    {
+      "epoch": 0.7748,
+      "grad_norm": 0.5431384444236755,
+      "learning_rate": 8.904142892078484e-07,
+      "loss": 0.5717,
+      "step": 1937
+    },
+    {
+      "epoch": 0.7752,
+      "grad_norm": 0.5083884596824646,
+      "learning_rate": 8.87394858356798e-07,
+      "loss": 0.5722,
+      "step": 1938
+    },
+    {
+      "epoch": 0.7756,
+      "grad_norm": 0.520429790019989,
+      "learning_rate": 8.843798123076346e-07,
+      "loss": 0.4715,
+      "step": 1939
+    },
+    {
+      "epoch": 0.776,
+      "grad_norm": 0.527215838432312,
+      "learning_rate": 8.813691561205864e-07,
+      "loss": 0.5459,
+      "step": 1940
+    },
+    {
+      "epoch": 0.7764,
+      "grad_norm": 0.5186598300933838,
+      "learning_rate": 8.783628948485153e-07,
+      "loss": 0.5213,
+      "step": 1941
+    },
+    {
+      "epoch": 0.7768,
+      "grad_norm": 0.5035976767539978,
+      "learning_rate": 8.75361033536907e-07,
+      "loss": 0.5397,
+      "step": 1942
+    },
+    {
+      "epoch": 0.7772,
+      "grad_norm": 0.5414839386940002,
+      "learning_rate": 8.723635772238613e-07,
+      "loss": 0.4567,
+      "step": 1943
+    },
+    {
+      "epoch": 0.7776,
+      "grad_norm": 0.5333899855613708,
+      "learning_rate": 8.693705309400856e-07,
+      "loss": 0.5362,
+      "step": 1944
+    },
+    {
+      "epoch": 0.778,
+      "grad_norm": 0.5623505711555481,
+      "learning_rate": 8.663818997088838e-07,
+      "loss": 0.6472,
+      "step": 1945
+    },
+    {
+      "epoch": 0.7784,
+      "grad_norm": 0.5305650234222412,
+      "learning_rate": 8.633976885461559e-07,
+      "loss": 0.5869,
+      "step": 1946
+    },
+    {
+      "epoch": 0.7788,
+      "grad_norm": 0.5463561415672302,
+      "learning_rate": 8.604179024603777e-07,
+      "loss": 0.6323,
+      "step": 1947
+    },
+    {
+      "epoch": 0.7792,
+      "grad_norm": 0.5228492617607117,
+      "learning_rate": 8.574425464526005e-07,
+      "loss": 0.5033,
+      "step": 1948
+    },
+    {
+      "epoch": 0.7796,
+      "grad_norm": 0.4840620160102844,
+      "learning_rate": 8.544716255164388e-07,
+      "loss": 0.5229,
+      "step": 1949
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.5401861667633057,
+      "learning_rate": 8.515051446380663e-07,
+      "loss": 0.597,
+      "step": 1950
+    },
+    {
+      "epoch": 0.7804,
+      "grad_norm": 0.5699217319488525,
+      "learning_rate": 8.485431087962045e-07,
+      "loss": 0.5568,
+      "step": 1951
+    },
+    {
+      "epoch": 0.7808,
+      "grad_norm": 0.46224555373191833,
+      "learning_rate": 8.455855229621131e-07,
+      "loss": 0.5103,
+      "step": 1952
+    },
+    {
+      "epoch": 0.7812,
+      "grad_norm": 0.5405699610710144,
+      "learning_rate": 8.426323920995829e-07,
+      "loss": 0.5714,
+      "step": 1953
+    },
+    {
+      "epoch": 0.7816,
+      "grad_norm": 0.5014129877090454,
+      "learning_rate": 8.396837211649276e-07,
+      "loss": 0.5612,
+      "step": 1954
+    },
+    {
+      "epoch": 0.782,
+      "grad_norm": 0.5532823801040649,
+      "learning_rate": 8.36739515106981e-07,
+      "loss": 0.5648,
+      "step": 1955
+    },
+    {
+      "epoch": 0.7824,
+      "grad_norm": 0.49977415800094604,
+      "learning_rate": 8.337997788670768e-07,
+      "loss": 0.486,
+      "step": 1956
+    },
+    {
+      "epoch": 0.7828,
+      "grad_norm": 0.5309237837791443,
+      "learning_rate": 8.308645173790498e-07,
+      "loss": 0.524,
+      "step": 1957
+    },
+    {
+      "epoch": 0.7832,
+      "grad_norm": 0.5683520436286926,
+      "learning_rate": 8.279337355692226e-07,
+      "loss": 0.5687,
+      "step": 1958
+    },
+    {
+      "epoch": 0.7836,
+      "grad_norm": 0.5215836763381958,
+      "learning_rate": 8.250074383564028e-07,
+      "loss": 0.5405,
+      "step": 1959
+    },
+    {
+      "epoch": 0.784,
+      "grad_norm": 0.5125592350959778,
+      "learning_rate": 8.220856306518701e-07,
+      "loss": 0.573,
+      "step": 1960
+    },
+    {
+      "epoch": 0.7844,
+      "grad_norm": 0.4813899099826813,
+      "learning_rate": 8.191683173593684e-07,
+      "loss": 0.4918,
+      "step": 1961
+    },
+    {
+      "epoch": 0.7848,
+      "grad_norm": 0.5178664326667786,
+      "learning_rate": 8.162555033750991e-07,
+      "loss": 0.4384,
+      "step": 1962
+    },
+    {
+      "epoch": 0.7852,
+      "grad_norm": 0.49676766991615295,
+      "learning_rate": 8.133471935877101e-07,
+      "loss": 0.5077,
+      "step": 1963
+    },
+    {
+      "epoch": 0.7856,
+      "grad_norm": 0.4734450876712799,
+      "learning_rate": 8.104433928782957e-07,
+      "loss": 0.5201,
+      "step": 1964
+    },
+    {
+      "epoch": 0.786,
+      "grad_norm": 0.5320947766304016,
+      "learning_rate": 8.07544106120377e-07,
+      "loss": 0.5475,
+      "step": 1965
+    },
+    {
+      "epoch": 0.7864,
+      "grad_norm": 0.5560709238052368,
+      "learning_rate": 8.046493381799007e-07,
+      "loss": 0.5751,
+      "step": 1966
+    },
+    {
+      "epoch": 0.7868,
+      "grad_norm": 0.5085044503211975,
+      "learning_rate": 8.017590939152293e-07,
+      "loss": 0.5635,
+      "step": 1967
+    },
+    {
+      "epoch": 0.7872,
+      "grad_norm": 0.5624022483825684,
+      "learning_rate": 7.988733781771339e-07,
+      "loss": 0.4983,
+      "step": 1968
+    },
+    {
+      "epoch": 0.7876,
+      "grad_norm": 0.5629763603210449,
+      "learning_rate": 7.95992195808786e-07,
+      "loss": 0.6108,
+      "step": 1969
+    },
+    {
+      "epoch": 0.788,
+      "grad_norm": 0.502626895904541,
+      "learning_rate": 7.931155516457466e-07,
+      "loss": 0.5313,
+      "step": 1970
+    },
+    {
+      "epoch": 0.7884,
+      "grad_norm": 0.514829695224762,
+      "learning_rate": 7.902434505159604e-07,
+      "loss": 0.4693,
+      "step": 1971
+    },
+    {
+      "epoch": 0.7888,
+      "grad_norm": 0.5001904964447021,
+      "learning_rate": 7.873758972397493e-07,
+      "loss": 0.4843,
+      "step": 1972
+    },
+    {
+      "epoch": 0.7892,
+      "grad_norm": 0.5443049073219299,
+      "learning_rate": 7.845128966297995e-07,
+      "loss": 0.5105,
+      "step": 1973
+    },
+    {
+      "epoch": 0.7896,
+      "grad_norm": 0.4830927550792694,
+      "learning_rate": 7.816544534911601e-07,
+      "loss": 0.5576,
+      "step": 1974
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.518970787525177,
+      "learning_rate": 7.788005726212282e-07,
+      "loss": 0.6164,
+      "step": 1975
+    },
+    {
+      "epoch": 0.7904,
+      "grad_norm": 0.6245747804641724,
+      "learning_rate": 7.759512588097436e-07,
+      "loss": 0.5214,
+      "step": 1976
+    },
+    {
+      "epoch": 0.7908,
+      "grad_norm": 0.4417863190174103,
+      "learning_rate": 7.731065168387841e-07,
+      "loss": 0.4494,
+      "step": 1977
+    },
+    {
+      "epoch": 0.7912,
+      "grad_norm": 0.5286259651184082,
+      "learning_rate": 7.702663514827517e-07,
+      "loss": 0.5324,
+      "step": 1978
+    },
+    {
+      "epoch": 0.7916,
+      "grad_norm": 0.5066895484924316,
+      "learning_rate": 7.674307675083699e-07,
+      "loss": 0.4509,
+      "step": 1979
+    },
+    {
+      "epoch": 0.792,
+      "grad_norm": 0.4970628321170807,
+      "learning_rate": 7.645997696746695e-07,
+      "loss": 0.6078,
+      "step": 1980
+    },
+    {
+      "epoch": 0.7924,
+      "grad_norm": 0.5808789730072021,
+      "learning_rate": 7.617733627329888e-07,
+      "loss": 0.5558,
+      "step": 1981
+    },
+    {
+      "epoch": 0.7928,
+      "grad_norm": 0.5080152153968811,
+      "learning_rate": 7.589515514269565e-07,
+      "loss": 0.5846,
+      "step": 1982
+    },
+    {
+      "epoch": 0.7932,
+      "grad_norm": 0.4858357608318329,
+      "learning_rate": 7.561343404924919e-07,
+      "loss": 0.5286,
+      "step": 1983
+    },
+    {
+      "epoch": 0.7936,
+      "grad_norm": 0.5149843096733093,
+      "learning_rate": 7.533217346577921e-07,
+      "loss": 0.5441,
+      "step": 1984
+    },
+    {
+      "epoch": 0.794,
+      "grad_norm": 0.5807613134384155,
+      "learning_rate": 7.505137386433237e-07,
+      "loss": 0.5804,
+      "step": 1985
+    },
+    {
+      "epoch": 0.7944,
+      "grad_norm": 0.48946282267570496,
+      "learning_rate": 7.477103571618203e-07,
+      "loss": 0.5502,
+      "step": 1986
+    },
+    {
+      "epoch": 0.7948,
+      "grad_norm": 0.5640960931777954,
+      "learning_rate": 7.449115949182662e-07,
+      "loss": 0.6595,
+      "step": 1987
+    },
+    {
+      "epoch": 0.7952,
+      "grad_norm": 0.5175038576126099,
+      "learning_rate": 7.421174566098976e-07,
+      "loss": 0.4677,
+      "step": 1988
+    },
+    {
+      "epoch": 0.7956,
+      "grad_norm": 0.5714621543884277,
+      "learning_rate": 7.393279469261867e-07,
+      "loss": 0.505,
+      "step": 1989
+    },
+    {
+      "epoch": 0.796,
+      "grad_norm": 0.49548372626304626,
+      "learning_rate": 7.365430705488399e-07,
+      "loss": 0.5464,
+      "step": 1990
+    },
+    {
+      "epoch": 0.7964,
+      "grad_norm": 0.6030895113945007,
+      "learning_rate": 7.337628321517861e-07,
+      "loss": 0.5169,
+      "step": 1991
+    },
+    {
+      "epoch": 0.7968,
+      "grad_norm": 0.5378836989402771,
+      "learning_rate": 7.309872364011688e-07,
+      "loss": 0.5947,
+      "step": 1992
+    },
+    {
+      "epoch": 0.7972,
+      "grad_norm": 0.5616456270217896,
+      "learning_rate": 7.282162879553437e-07,
+      "loss": 0.6075,
+      "step": 1993
+    },
+    {
+      "epoch": 0.7976,
+      "grad_norm": 0.48758378624916077,
+      "learning_rate": 7.254499914648619e-07,
+      "loss": 0.5536,
+      "step": 1994
+    },
+    {
+      "epoch": 0.798,
+      "grad_norm": 0.5290488600730896,
+      "learning_rate": 7.226883515724715e-07,
+      "loss": 0.5948,
+      "step": 1995
+    },
+    {
+      "epoch": 0.7984,
+      "grad_norm": 0.5616166591644287,
+      "learning_rate": 7.19931372913101e-07,
+      "loss": 0.5515,
+      "step": 1996
+    },
+    {
+      "epoch": 0.7988,
+      "grad_norm": 0.4974389672279358,
+      "learning_rate": 7.171790601138605e-07,
+      "loss": 0.5033,
+      "step": 1997
+    },
+    {
+      "epoch": 0.7992,
+      "grad_norm": 0.5275691747665405,
+      "learning_rate": 7.144314177940246e-07,
+      "loss": 0.5464,
+      "step": 1998
+    },
+    {
+      "epoch": 0.7996,
+      "grad_norm": 0.5309053063392639,
+      "learning_rate": 7.116884505650332e-07,
+      "loss": 0.5704,
+      "step": 1999
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.5537293553352356,
+      "learning_rate": 7.089501630304779e-07,
+      "loss": 0.4483,
+      "step": 2000
+    },
+    {
+      "epoch": 0.8004,
+      "grad_norm": 0.467911034822464,
+      "learning_rate": 7.062165597860947e-07,
+      "loss": 0.4794,
+      "step": 2001
+    },
+    {
+      "epoch": 0.8008,
+      "grad_norm": 0.5319057703018188,
+      "learning_rate": 7.034876454197625e-07,
+      "loss": 0.5058,
+      "step": 2002
+    },
+    {
+      "epoch": 0.8012,
+      "grad_norm": 0.5028004050254822,
+      "learning_rate": 7.007634245114858e-07,
+      "loss": 0.5143,
+      "step": 2003
+    },
+    {
+      "epoch": 0.8016,
+      "grad_norm": 0.47743862867355347,
+      "learning_rate": 6.980439016333954e-07,
+      "loss": 0.46,
+      "step": 2004
+    },
+    {
+      "epoch": 0.802,
+      "grad_norm": 0.5376464128494263,
+      "learning_rate": 6.953290813497354e-07,
+      "loss": 0.5481,
+      "step": 2005
+    },
+    {
+      "epoch": 0.8024,
+      "grad_norm": 0.48827072978019714,
+      "learning_rate": 6.926189682168575e-07,
+      "loss": 0.5273,
+      "step": 2006
+    },
+    {
+      "epoch": 0.8028,
+      "grad_norm": 0.5766599178314209,
+      "learning_rate": 6.899135667832138e-07,
+      "loss": 0.553,
+      "step": 2007
+    },
+    {
+      "epoch": 0.8032,
+      "grad_norm": 0.6004101037979126,
+      "learning_rate": 6.872128815893502e-07,
+      "loss": 0.5864,
+      "step": 2008
+    },
+    {
+      "epoch": 0.8036,
+      "grad_norm": 0.5494182705879211,
+      "learning_rate": 6.845169171678945e-07,
+      "loss": 0.6029,
+      "step": 2009
+    },
+    {
+      "epoch": 0.804,
+      "grad_norm": 0.5339188575744629,
+      "learning_rate": 6.81825678043551e-07,
+      "loss": 0.5386,
+      "step": 2010
+    },
+    {
+      "epoch": 0.8044,
+      "grad_norm": 0.5146757960319519,
+      "learning_rate": 6.79139168733098e-07,
+      "loss": 0.5751,
+      "step": 2011
+    },
+    {
+      "epoch": 0.8048,
+      "grad_norm": 0.6246601939201355,
+      "learning_rate": 6.764573937453698e-07,
+      "loss": 0.5433,
+      "step": 2012
+    },
+    {
+      "epoch": 0.8052,
+      "grad_norm": 0.5111002326011658,
+      "learning_rate": 6.737803575812601e-07,
+      "loss": 0.5189,
+      "step": 2013
+    },
+    {
+      "epoch": 0.8056,
+      "grad_norm": 0.45920640230178833,
+      "learning_rate": 6.71108064733706e-07,
+      "loss": 0.4705,
+      "step": 2014
+    },
+    {
+      "epoch": 0.806,
+      "grad_norm": 0.5450352430343628,
+      "learning_rate": 6.684405196876843e-07,
+      "loss": 0.5344,
+      "step": 2015
+    },
+    {
+      "epoch": 0.8064,
+      "grad_norm": 0.5059295892715454,
+      "learning_rate": 6.657777269202041e-07,
+      "loss": 0.466,
+      "step": 2016
+    },
+    {
+      "epoch": 0.8068,
+      "grad_norm": 0.48422038555145264,
+      "learning_rate": 6.631196909002998e-07,
+      "loss": 0.48,
+      "step": 2017
+    },
+    {
+      "epoch": 0.8072,
+      "grad_norm": 0.5188162326812744,
+      "learning_rate": 6.604664160890203e-07,
+      "loss": 0.5654,
+      "step": 2018
+    },
+    {
+      "epoch": 0.8076,
+      "grad_norm": 0.48736605048179626,
+      "learning_rate": 6.57817906939424e-07,
+      "loss": 0.5568,
+      "step": 2019
+    },
+    {
+      "epoch": 0.808,
+      "grad_norm": 0.5381132960319519,
+      "learning_rate": 6.551741678965707e-07,
+      "loss": 0.4822,
+      "step": 2020
+    },
+    {
+      "epoch": 0.8084,
+      "grad_norm": 0.4702557623386383,
+      "learning_rate": 6.525352033975163e-07,
+      "loss": 0.4102,
+      "step": 2021
+    },
+    {
+      "epoch": 0.8088,
+      "grad_norm": 0.5244594216346741,
+      "learning_rate": 6.49901017871303e-07,
+      "loss": 0.5595,
+      "step": 2022
+    },
+    {
+      "epoch": 0.8092,
+      "grad_norm": 0.5497479438781738,
+      "learning_rate": 6.4727161573895e-07,
+      "loss": 0.513,
+      "step": 2023
+    },
+    {
+      "epoch": 0.8096,
+      "grad_norm": 0.5612243413925171,
+      "learning_rate": 6.446470014134504e-07,
+      "loss": 0.5265,
+      "step": 2024
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.576653003692627,
+      "learning_rate": 6.420271792997612e-07,
+      "loss": 0.6046,
+      "step": 2025
+    },
+    {
+      "epoch": 0.8104,
+      "grad_norm": 0.5197085738182068,
+      "learning_rate": 6.394121537947986e-07,
+      "loss": 0.5626,
+      "step": 2026
+    },
+    {
+      "epoch": 0.8108,
+      "grad_norm": 0.5471281409263611,
+      "learning_rate": 6.368019292874246e-07,
+      "loss": 0.5512,
+      "step": 2027
+    },
+    {
+      "epoch": 0.8112,
+      "grad_norm": 0.5379735827445984,
+      "learning_rate": 6.341965101584466e-07,
+      "loss": 0.5143,
+      "step": 2028
+    },
+    {
+      "epoch": 0.8116,
+      "grad_norm": 0.5038474798202515,
+      "learning_rate": 6.315959007806046e-07,
+      "loss": 0.4985,
+      "step": 2029
+    },
+    {
+      "epoch": 0.812,
+      "grad_norm": 0.6646623611450195,
+      "learning_rate": 6.290001055185684e-07,
+      "loss": 0.6237,
+      "step": 2030
+    },
+    {
+      "epoch": 0.8124,
+      "grad_norm": 0.5201054811477661,
+      "learning_rate": 6.264091287289292e-07,
+      "loss": 0.5631,
+      "step": 2031
+    },
+    {
+      "epoch": 0.8128,
+      "grad_norm": 0.4982958137989044,
+      "learning_rate": 6.238229747601876e-07,
+      "loss": 0.5284,
+      "step": 2032
+    },
+    {
+      "epoch": 0.8132,
+      "grad_norm": 0.5728126764297485,
+      "learning_rate": 6.212416479527513e-07,
+      "loss": 0.5772,
+      "step": 2033
+    },
+    {
+      "epoch": 0.8136,
+      "grad_norm": 0.5694175362586975,
+      "learning_rate": 6.186651526389281e-07,
+      "loss": 0.5493,
+      "step": 2034
+    },
+    {
+      "epoch": 0.814,
+      "grad_norm": 0.4679771065711975,
+      "learning_rate": 6.160934931429164e-07,
+      "loss": 0.4774,
+      "step": 2035
+    },
+    {
+      "epoch": 0.8144,
+      "grad_norm": 0.5073397755622864,
+      "learning_rate": 6.135266737807975e-07,
+      "loss": 0.4938,
+      "step": 2036
+    },
+    {
+      "epoch": 0.8148,
+      "grad_norm": 0.5173336863517761,
+      "learning_rate": 6.10964698860529e-07,
+      "loss": 0.5557,
+      "step": 2037
+    },
+    {
+      "epoch": 0.8152,
+      "grad_norm": 0.5217483639717102,
+      "learning_rate": 6.084075726819389e-07,
+      "loss": 0.5521,
+      "step": 2038
+    },
+    {
+      "epoch": 0.8156,
+      "grad_norm": 0.5165453553199768,
+      "learning_rate": 6.058552995367178e-07,
+      "loss": 0.5561,
+      "step": 2039
+    },
+    {
+      "epoch": 0.816,
+      "grad_norm": 0.4597705006599426,
+      "learning_rate": 6.033078837084113e-07,
+      "loss": 0.4374,
+      "step": 2040
+    },
+    {
+      "epoch": 0.8164,
+      "grad_norm": 0.565405547618866,
+      "learning_rate": 6.007653294724123e-07,
+      "loss": 0.6586,
+      "step": 2041
+    },
+    {
+      "epoch": 0.8168,
+      "grad_norm": 0.4685356616973877,
+      "learning_rate": 5.98227641095953e-07,
+      "loss": 0.4905,
+      "step": 2042
+    },
+    {
+      "epoch": 0.8172,
+      "grad_norm": 0.573566734790802,
+      "learning_rate": 5.956948228381015e-07,
+      "loss": 0.6105,
+      "step": 2043
+    },
+    {
+      "epoch": 0.8176,
+      "grad_norm": 0.5715039968490601,
+      "learning_rate": 5.931668789497525e-07,
+      "loss": 0.5855,
+      "step": 2044
+    },
+    {
+      "epoch": 0.818,
+      "grad_norm": 0.48921602964401245,
+      "learning_rate": 5.906438136736182e-07,
+      "loss": 0.5488,
+      "step": 2045
+    },
+    {
+      "epoch": 0.8184,
+      "grad_norm": 0.5068395137786865,
+      "learning_rate": 5.881256312442223e-07,
+      "loss": 0.5313,
+      "step": 2046
+    },
+    {
+      "epoch": 0.8188,
+      "grad_norm": 0.5660580992698669,
+      "learning_rate": 5.856123358878947e-07,
+      "loss": 0.5712,
+      "step": 2047
+    },
+    {
+      "epoch": 0.8192,
+      "grad_norm": 0.48958441615104675,
+      "learning_rate": 5.831039318227639e-07,
+      "loss": 0.5172,
+      "step": 2048
+    },
+    {
+      "epoch": 0.8196,
+      "grad_norm": 0.5002844333648682,
+      "learning_rate": 5.806004232587487e-07,
+      "loss": 0.5371,
+      "step": 2049
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.5133857727050781,
+      "learning_rate": 5.781018143975516e-07,
+      "loss": 0.5908,
+      "step": 2050
+    },
+    {
+      "epoch": 0.8204,
+      "grad_norm": 0.5378307700157166,
+      "learning_rate": 5.756081094326496e-07,
+      "loss": 0.5463,
+      "step": 2051
+    },
+    {
+      "epoch": 0.8208,
+      "grad_norm": 0.574783444404602,
+      "learning_rate": 5.73119312549294e-07,
+      "loss": 0.6224,
+      "step": 2052
+    },
+    {
+      "epoch": 0.8212,
+      "grad_norm": 0.497978538274765,
+      "learning_rate": 5.706354279244942e-07,
+      "loss": 0.5832,
+      "step": 2053
+    },
+    {
+      "epoch": 0.8216,
+      "grad_norm": 0.476604163646698,
+      "learning_rate": 5.681564597270188e-07,
+      "loss": 0.5334,
+      "step": 2054
+    },
+    {
+      "epoch": 0.822,
+      "grad_norm": 0.4663172662258148,
+      "learning_rate": 5.65682412117382e-07,
+      "loss": 0.4863,
+      "step": 2055
+    },
+    {
+      "epoch": 0.8224,
+      "grad_norm": 0.5250645875930786,
+      "learning_rate": 5.632132892478414e-07,
+      "loss": 0.5811,
+      "step": 2056
+    },
+    {
+      "epoch": 0.8228,
+      "grad_norm": 0.48894020915031433,
+      "learning_rate": 5.607490952623886e-07,
+      "loss": 0.5533,
+      "step": 2057
+    },
+    {
+      "epoch": 0.8232,
+      "grad_norm": 0.5382503867149353,
+      "learning_rate": 5.582898342967445e-07,
+      "loss": 0.5632,
+      "step": 2058
+    },
+    {
+      "epoch": 0.8236,
+      "grad_norm": 0.5530444979667664,
+      "learning_rate": 5.558355104783495e-07,
+      "loss": 0.5749,
+      "step": 2059
+    },
+    {
+      "epoch": 0.824,
+      "grad_norm": 0.5484821796417236,
+      "learning_rate": 5.533861279263556e-07,
+      "loss": 0.5641,
+      "step": 2060
+    },
+    {
+      "epoch": 0.8244,
+      "grad_norm": 0.6639271974563599,
+      "learning_rate": 5.509416907516267e-07,
+      "loss": 0.6381,
+      "step": 2061
+    },
+    {
+      "epoch": 0.8248,
+      "grad_norm": 0.49303728342056274,
+      "learning_rate": 5.485022030567224e-07,
+      "loss": 0.5542,
+      "step": 2062
+    },
+    {
+      "epoch": 0.8252,
+      "grad_norm": 0.5660285949707031,
+      "learning_rate": 5.460676689358982e-07,
+      "loss": 0.6456,
+      "step": 2063
+    },
+    {
+      "epoch": 0.8256,
+      "grad_norm": 0.4761320650577545,
+      "learning_rate": 5.436380924750948e-07,
+      "loss": 0.5237,
+      "step": 2064
+    },
+    {
+      "epoch": 0.826,
+      "grad_norm": 0.5708348155021667,
+      "learning_rate": 5.412134777519309e-07,
+      "loss": 0.5548,
+      "step": 2065
+    },
+    {
+      "epoch": 0.8264,
+      "grad_norm": 0.49620676040649414,
+      "learning_rate": 5.387938288357014e-07,
+      "loss": 0.5473,
+      "step": 2066
+    },
+    {
+      "epoch": 0.8268,
+      "grad_norm": 0.5709338784217834,
+      "learning_rate": 5.363791497873632e-07,
+      "loss": 0.6886,
+      "step": 2067
+    },
+    {
+      "epoch": 0.8272,
+      "grad_norm": 0.4843921363353729,
+      "learning_rate": 5.339694446595349e-07,
+      "loss": 0.5347,
+      "step": 2068
+    },
+    {
+      "epoch": 0.8276,
+      "grad_norm": 0.5291857123374939,
+      "learning_rate": 5.315647174964848e-07,
+      "loss": 0.6049,
+      "step": 2069
+    },
+    {
+      "epoch": 0.828,
+      "grad_norm": 0.5444912910461426,
+      "learning_rate": 5.291649723341296e-07,
+      "loss": 0.5096,
+      "step": 2070
+    },
+    {
+      "epoch": 0.8284,
+      "grad_norm": 0.5704883933067322,
+      "learning_rate": 5.267702132000212e-07,
+      "loss": 0.5848,
+      "step": 2071
+    },
+    {
+      "epoch": 0.8288,
+      "grad_norm": 0.45734161138534546,
+      "learning_rate": 5.243804441133465e-07,
+      "loss": 0.6085,
+      "step": 2072
+    },
+    {
+      "epoch": 0.8292,
+      "grad_norm": 0.5407284498214722,
+      "learning_rate": 5.21995669084915e-07,
+      "loss": 0.5605,
+      "step": 2073
+    },
+    {
+      "epoch": 0.8296,
+      "grad_norm": 0.5885137319564819,
+      "learning_rate": 5.196158921171548e-07,
+      "loss": 0.5438,
+      "step": 2074
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.5144458413124084,
+      "learning_rate": 5.172411172041079e-07,
+      "loss": 0.4796,
+      "step": 2075
+    },
+    {
+      "epoch": 0.8304,
+      "grad_norm": 0.4840572774410248,
+      "learning_rate": 5.148713483314182e-07,
+      "loss": 0.4583,
+      "step": 2076
+    },
+    {
+      "epoch": 0.8308,
+      "grad_norm": 0.4647265076637268,
+      "learning_rate": 5.12506589476331e-07,
+      "loss": 0.492,
+      "step": 2077
+    },
+    {
+      "epoch": 0.8312,
+      "grad_norm": 0.5066642761230469,
+      "learning_rate": 5.101468446076796e-07,
+      "loss": 0.473,
+      "step": 2078
+    },
+    {
+      "epoch": 0.8316,
+      "grad_norm": 0.5504907965660095,
+      "learning_rate": 5.077921176858854e-07,
+      "loss": 0.5896,
+      "step": 2079
+    },
+    {
+      "epoch": 0.832,
+      "grad_norm": 0.5531979203224182,
+      "learning_rate": 5.054424126629462e-07,
+      "loss": 0.4968,
+      "step": 2080
+    },
+    {
+      "epoch": 0.8324,
+      "grad_norm": 0.5404284000396729,
+      "learning_rate": 5.030977334824312e-07,
+      "loss": 0.5036,
+      "step": 2081
+    },
+    {
+      "epoch": 0.8328,
+      "grad_norm": 0.4818439185619354,
+      "learning_rate": 5.007580840794761e-07,
+      "loss": 0.4787,
+      "step": 2082
+    },
+    {
+      "epoch": 0.8332,
+      "grad_norm": 0.48733893036842346,
+      "learning_rate": 4.984234683807746e-07,
+      "loss": 0.4894,
+      "step": 2083
+    },
+    {
+      "epoch": 0.8336,
+      "grad_norm": 0.4569380581378937,
+      "learning_rate": 4.960938903045715e-07,
+      "loss": 0.4484,
+      "step": 2084
+    },
+    {
+      "epoch": 0.834,
+      "grad_norm": 0.5140909552574158,
+      "learning_rate": 4.937693537606559e-07,
+      "loss": 0.5013,
+      "step": 2085
+    },
+    {
+      "epoch": 0.8344,
+      "grad_norm": 0.5413423776626587,
+      "learning_rate": 4.914498626503589e-07,
+      "loss": 0.5123,
+      "step": 2086
+    },
+    {
+      "epoch": 0.8348,
+      "grad_norm": 0.4930032789707184,
+      "learning_rate": 4.891354208665396e-07,
+      "loss": 0.557,
+      "step": 2087
+    },
+    {
+      "epoch": 0.8352,
+      "grad_norm": 0.43929794430732727,
+      "learning_rate": 4.868260322935864e-07,
+      "loss": 0.4897,
+      "step": 2088
+    },
+    {
+      "epoch": 0.8356,
+      "grad_norm": 0.5489115118980408,
+      "learning_rate": 4.845217008074045e-07,
+      "loss": 0.5762,
+      "step": 2089
+    },
+    {
+      "epoch": 0.836,
+      "grad_norm": 0.4723738431930542,
+      "learning_rate": 4.82222430275411e-07,
+      "loss": 0.5563,
+      "step": 2090
+    },
+    {
+      "epoch": 0.8364,
+      "grad_norm": 0.5301296710968018,
+      "learning_rate": 4.79928224556532e-07,
+      "loss": 0.51,
+      "step": 2091
+    },
+    {
+      "epoch": 0.8368,
+      "grad_norm": 0.5511886477470398,
+      "learning_rate": 4.776390875011912e-07,
+      "loss": 0.4941,
+      "step": 2092
+    },
+    {
+      "epoch": 0.8372,
+      "grad_norm": 0.5579310059547424,
+      "learning_rate": 4.7535502295130544e-07,
+      "loss": 0.6519,
+      "step": 2093
+    },
+    {
+      "epoch": 0.8376,
+      "grad_norm": 0.48983925580978394,
+      "learning_rate": 4.730760347402788e-07,
+      "loss": 0.5707,
+      "step": 2094
+    },
+    {
+      "epoch": 0.838,
+      "grad_norm": 0.5601329207420349,
+      "learning_rate": 4.708021266929944e-07,
+      "loss": 0.5234,
+      "step": 2095
+    },
+    {
+      "epoch": 0.8384,
+      "grad_norm": 0.5632297396659851,
+      "learning_rate": 4.685333026258109e-07,
+      "loss": 0.6143,
+      "step": 2096
+    },
+    {
+      "epoch": 0.8388,
+      "grad_norm": 0.6144510507583618,
+      "learning_rate": 4.6626956634655434e-07,
+      "loss": 0.5627,
+      "step": 2097
+    },
+    {
+      "epoch": 0.8392,
+      "grad_norm": 0.4851270616054535,
+      "learning_rate": 4.6401092165451024e-07,
+      "loss": 0.5105,
+      "step": 2098
+    },
+    {
+      "epoch": 0.8396,
+      "grad_norm": 0.5638195872306824,
+      "learning_rate": 4.617573723404191e-07,
+      "loss": 0.5906,
+      "step": 2099
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.516826331615448,
+      "learning_rate": 4.59508922186471e-07,
+      "loss": 0.5586,
+      "step": 2100
+    },
+    {
+      "epoch": 0.8404,
+      "grad_norm": 0.5381993055343628,
+      "learning_rate": 4.5726557496629773e-07,
+      "loss": 0.4996,
+      "step": 2101
+    },
+    {
+      "epoch": 0.8408,
+      "grad_norm": 0.5378861427307129,
+      "learning_rate": 4.5502733444496577e-07,
+      "loss": 0.5762,
+      "step": 2102
+    },
+    {
+      "epoch": 0.8412,
+      "grad_norm": 0.5263057351112366,
+      "learning_rate": 4.527942043789708e-07,
+      "loss": 0.6046,
+      "step": 2103
+    },
+    {
+      "epoch": 0.8416,
+      "grad_norm": 0.5146876573562622,
+      "learning_rate": 4.5056618851623173e-07,
+      "loss": 0.6175,
+      "step": 2104
+    },
+    {
+      "epoch": 0.842,
+      "grad_norm": 0.5282055139541626,
+      "learning_rate": 4.4834329059608533e-07,
+      "loss": 0.5349,
+      "step": 2105
+    },
+    {
+      "epoch": 0.8424,
+      "grad_norm": 0.5305773615837097,
+      "learning_rate": 4.461255143492781e-07,
+      "loss": 0.6091,
+      "step": 2106
+    },
+    {
+      "epoch": 0.8428,
+      "grad_norm": 0.5523550510406494,
+      "learning_rate": 4.4391286349796e-07,
+      "loss": 0.6574,
+      "step": 2107
+    },
+    {
+      "epoch": 0.8432,
+      "grad_norm": 0.5279895663261414,
+      "learning_rate": 4.4170534175567863e-07,
+      "loss": 0.4755,
+      "step": 2108
+    },
+    {
+      "epoch": 0.8436,
+      "grad_norm": 0.5235905647277832,
+      "learning_rate": 4.395029528273751e-07,
+      "loss": 0.5776,
+      "step": 2109
+    },
+    {
+      "epoch": 0.844,
+      "grad_norm": 0.5142126083374023,
+      "learning_rate": 4.373057004093756e-07,
+      "loss": 0.4684,
+      "step": 2110
+    },
+    {
+      "epoch": 0.8444,
+      "grad_norm": 0.48986274003982544,
+      "learning_rate": 4.3511358818938384e-07,
+      "loss": 0.5636,
+      "step": 2111
+    },
+    {
+      "epoch": 0.8448,
+      "grad_norm": 0.5454509258270264,
+      "learning_rate": 4.329266198464782e-07,
+      "loss": 0.5735,
+      "step": 2112
+    },
+    {
+      "epoch": 0.8452,
+      "grad_norm": 0.6086745262145996,
+      "learning_rate": 4.3074479905110223e-07,
+      "loss": 0.5865,
+      "step": 2113
+    },
+    {
+      "epoch": 0.8456,
+      "grad_norm": 0.5274353623390198,
+      "learning_rate": 4.285681294650624e-07,
+      "loss": 0.6151,
+      "step": 2114
+    },
+    {
+      "epoch": 0.846,
+      "grad_norm": 0.5232753157615662,
+      "learning_rate": 4.263966147415201e-07,
+      "loss": 0.526,
+      "step": 2115
+    },
+    {
+      "epoch": 0.8464,
+      "grad_norm": 0.5116792917251587,
+      "learning_rate": 4.242302585249821e-07,
+      "loss": 0.5868,
+      "step": 2116
+    },
+    {
+      "epoch": 0.8468,
+      "grad_norm": 0.5277671217918396,
+      "learning_rate": 4.220690644512997e-07,
+      "loss": 0.5121,
+      "step": 2117
+    },
+    {
+      "epoch": 0.8472,
+      "grad_norm": 0.5713874697685242,
+      "learning_rate": 4.199130361476599e-07,
+      "loss": 0.556,
+      "step": 2118
+    },
+    {
+      "epoch": 0.8476,
+      "grad_norm": 0.5311291813850403,
+      "learning_rate": 4.1776217723258134e-07,
+      "loss": 0.4875,
+      "step": 2119
+    },
+    {
+      "epoch": 0.848,
+      "grad_norm": 0.5216573476791382,
+      "learning_rate": 4.1561649131590474e-07,
+      "loss": 0.5525,
+      "step": 2120
+    },
+    {
+      "epoch": 0.8484,
+      "grad_norm": 0.5430652499198914,
+      "learning_rate": 4.1347598199878885e-07,
+      "loss": 0.5158,
+      "step": 2121
+    },
+    {
+      "epoch": 0.8488,
+      "grad_norm": 0.5124344229698181,
+      "learning_rate": 4.113406528737052e-07,
+      "loss": 0.5256,
+      "step": 2122
+    },
+    {
+      "epoch": 0.8492,
+      "grad_norm": 0.5016151070594788,
+      "learning_rate": 4.09210507524431e-07,
+      "loss": 0.5081,
+      "step": 2123
+    },
+    {
+      "epoch": 0.8496,
+      "grad_norm": 0.4942253530025482,
+      "learning_rate": 4.070855495260453e-07,
+      "loss": 0.5332,
+      "step": 2124
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.4644415080547333,
+      "learning_rate": 4.0496578244491793e-07,
+      "loss": 0.5018,
+      "step": 2125
+    },
+    {
+      "epoch": 0.8504,
+      "grad_norm": 0.4998500347137451,
+      "learning_rate": 4.028512098387071e-07,
+      "loss": 0.5553,
+      "step": 2126
+    },
+    {
+      "epoch": 0.8508,
+      "grad_norm": 0.5142858624458313,
+      "learning_rate": 4.007418352563566e-07,
+      "loss": 0.5019,
+      "step": 2127
+    },
+    {
+      "epoch": 0.8512,
+      "grad_norm": 0.5226315855979919,
+      "learning_rate": 3.9863766223808086e-07,
+      "loss": 0.5026,
+      "step": 2128
+    },
+    {
+      "epoch": 0.8516,
+      "grad_norm": 0.518667459487915,
+      "learning_rate": 3.9653869431536953e-07,
+      "loss": 0.5007,
+      "step": 2129
+    },
+    {
+      "epoch": 0.852,
+      "grad_norm": 0.5096085667610168,
+      "learning_rate": 3.944449350109729e-07,
+      "loss": 0.5872,
+      "step": 2130
+    },
+    {
+      "epoch": 0.8524,
+      "grad_norm": 0.5470719933509827,
+      "learning_rate": 3.9235638783889987e-07,
+      "loss": 0.5457,
+      "step": 2131
+    },
+    {
+      "epoch": 0.8528,
+      "grad_norm": 0.5269269943237305,
+      "learning_rate": 3.902730563044133e-07,
+      "loss": 0.4615,
+      "step": 2132
+    },
+    {
+      "epoch": 0.8532,
+      "grad_norm": 0.5528969764709473,
+      "learning_rate": 3.881949439040224e-07,
+      "loss": 0.6155,
+      "step": 2133
+    },
+    {
+      "epoch": 0.8536,
+      "grad_norm": 0.5143035650253296,
+      "learning_rate": 3.8612205412547534e-07,
+      "loss": 0.5154,
+      "step": 2134
+    },
+    {
+      "epoch": 0.854,
+      "grad_norm": 0.5350843667984009,
+      "learning_rate": 3.840543904477557e-07,
+      "loss": 0.4668,
+      "step": 2135
+    },
+    {
+      "epoch": 0.8544,
+      "grad_norm": 0.5625218152999878,
+      "learning_rate": 3.819919563410773e-07,
+      "loss": 0.5487,
+      "step": 2136
+    },
+    {
+      "epoch": 0.8548,
+      "grad_norm": 0.503579318523407,
+      "learning_rate": 3.799347552668745e-07,
+      "loss": 0.5635,
+      "step": 2137
+    },
+    {
+      "epoch": 0.8552,
+      "grad_norm": 0.5495058298110962,
+      "learning_rate": 3.77882790677802e-07,
+      "loss": 0.5593,
+      "step": 2138
+    },
+    {
+      "epoch": 0.8556,
+      "grad_norm": 0.5059952735900879,
+      "learning_rate": 3.7583606601772346e-07,
+      "loss": 0.5202,
+      "step": 2139
+    },
+    {
+      "epoch": 0.856,
+      "grad_norm": 0.5059313774108887,
+      "learning_rate": 3.7379458472170814e-07,
+      "loss": 0.5287,
+      "step": 2140
+    },
+    {
+      "epoch": 0.8564,
+      "grad_norm": 0.5819924473762512,
+      "learning_rate": 3.717583502160283e-07,
+      "loss": 0.5815,
+      "step": 2141
+    },
+    {
+      "epoch": 0.8568,
+      "grad_norm": 0.5166248679161072,
+      "learning_rate": 3.697273659181459e-07,
+      "loss": 0.5107,
+      "step": 2142
+    },
+    {
+      "epoch": 0.8572,
+      "grad_norm": 0.6137151718139648,
+      "learning_rate": 3.67701635236716e-07,
+      "loss": 0.5726,
+      "step": 2143
+    },
+    {
+      "epoch": 0.8576,
+      "grad_norm": 0.468230664730072,
+      "learning_rate": 3.656811615715727e-07,
+      "loss": 0.4374,
+      "step": 2144
+    },
+    {
+      "epoch": 0.858,
+      "grad_norm": 0.4957389831542969,
+      "learning_rate": 3.636659483137291e-07,
+      "loss": 0.5401,
+      "step": 2145
+    },
+    {
+      "epoch": 0.8584,
+      "grad_norm": 0.49795204401016235,
+      "learning_rate": 3.616559988453683e-07,
+      "loss": 0.5141,
+      "step": 2146
+    },
+    {
+      "epoch": 0.8588,
+      "grad_norm": 0.5580368041992188,
+      "learning_rate": 3.596513165398412e-07,
+      "loss": 0.5541,
+      "step": 2147
+    },
+    {
+      "epoch": 0.8592,
+      "grad_norm": 0.5130208134651184,
+      "learning_rate": 3.576519047616564e-07,
+      "loss": 0.4491,
+      "step": 2148
+    },
+    {
+      "epoch": 0.8596,
+      "grad_norm": 0.4993700683116913,
+      "learning_rate": 3.55657766866477e-07,
+      "loss": 0.5044,
+      "step": 2149
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.5155137181282043,
+      "learning_rate": 3.536689062011167e-07,
+      "loss": 0.5642,
+      "step": 2150
+    },
+    {
+      "epoch": 0.8604,
+      "grad_norm": 0.4971082806587219,
+      "learning_rate": 3.516853261035304e-07,
+      "loss": 0.5726,
+      "step": 2151
+    },
+    {
+      "epoch": 0.8608,
+      "grad_norm": 0.515801727771759,
+      "learning_rate": 3.497070299028118e-07,
+      "loss": 0.47,
+      "step": 2152
+    },
+    {
+      "epoch": 0.8612,
+      "grad_norm": 0.5084040760993958,
+      "learning_rate": 3.47734020919185e-07,
+      "loss": 0.5551,
+      "step": 2153
+    },
+    {
+      "epoch": 0.8616,
+      "grad_norm": 0.45051297545433044,
+      "learning_rate": 3.4576630246400246e-07,
+      "loss": 0.4355,
+      "step": 2154
+    },
+    {
+      "epoch": 0.862,
+      "grad_norm": 0.5345770120620728,
+      "learning_rate": 3.43803877839736e-07,
+      "loss": 0.5597,
+      "step": 2155
+    },
+    {
+      "epoch": 0.8624,
+      "grad_norm": 0.5158503651618958,
+      "learning_rate": 3.4184675033997157e-07,
+      "loss": 0.5396,
+      "step": 2156
+    },
+    {
+      "epoch": 0.8628,
+      "grad_norm": 0.5118535757064819,
+      "learning_rate": 3.3989492324940856e-07,
+      "loss": 0.5089,
+      "step": 2157
+    },
+    {
+      "epoch": 0.8632,
+      "grad_norm": 0.521641194820404,
+      "learning_rate": 3.379483998438458e-07,
+      "loss": 0.4735,
+      "step": 2158
+    },
+    {
+      "epoch": 0.8636,
+      "grad_norm": 0.5525158047676086,
+      "learning_rate": 3.360071833901854e-07,
+      "loss": 0.5108,
+      "step": 2159
+    },
+    {
+      "epoch": 0.864,
+      "grad_norm": 0.48484671115875244,
+      "learning_rate": 3.340712771464185e-07,
+      "loss": 0.4106,
+      "step": 2160
+    },
+    {
+      "epoch": 0.8644,
+      "grad_norm": 0.5603184700012207,
+      "learning_rate": 3.3214068436162764e-07,
+      "loss": 0.58,
+      "step": 2161
+    },
+    {
+      "epoch": 0.8648,
+      "grad_norm": 0.5196020603179932,
+      "learning_rate": 3.302154082759741e-07,
+      "loss": 0.5204,
+      "step": 2162
+    },
+    {
+      "epoch": 0.8652,
+      "grad_norm": 0.5511024594306946,
+      "learning_rate": 3.2829545212070014e-07,
+      "loss": 0.5546,
+      "step": 2163
+    },
+    {
+      "epoch": 0.8656,
+      "grad_norm": 0.529928982257843,
+      "learning_rate": 3.263808191181157e-07,
+      "loss": 0.5546,
+      "step": 2164
+    },
+    {
+      "epoch": 0.866,
+      "grad_norm": 0.4980902671813965,
+      "learning_rate": 3.244715124815982e-07,
+      "loss": 0.554,
+      "step": 2165
+    },
+    {
+      "epoch": 0.8664,
+      "grad_norm": 0.5063270926475525,
+      "learning_rate": 3.2256753541558635e-07,
+      "loss": 0.5141,
+      "step": 2166
+    },
+    {
+      "epoch": 0.8668,
+      "grad_norm": 0.5925174355506897,
+      "learning_rate": 3.2066889111557254e-07,
+      "loss": 0.55,
+      "step": 2167
+    },
+    {
+      "epoch": 0.8672,
+      "grad_norm": 0.5630876421928406,
+      "learning_rate": 3.1877558276810124e-07,
+      "loss": 0.5533,
+      "step": 2168
+    },
+    {
+      "epoch": 0.8676,
+      "grad_norm": 0.4962352216243744,
+      "learning_rate": 3.1688761355075977e-07,
+      "loss": 0.5002,
+      "step": 2169
+    },
+    {
+      "epoch": 0.868,
+      "grad_norm": 0.5295508503913879,
+      "learning_rate": 3.1500498663217376e-07,
+      "loss": 0.5161,
+      "step": 2170
+    },
+    {
+      "epoch": 0.8684,
+      "grad_norm": 0.5267398357391357,
+      "learning_rate": 3.13127705172005e-07,
+      "loss": 0.5289,
+      "step": 2171
+    },
+    {
+      "epoch": 0.8688,
+      "grad_norm": 0.4687250852584839,
+      "learning_rate": 3.1125577232094355e-07,
+      "loss": 0.5046,
+      "step": 2172
+    },
+    {
+      "epoch": 0.8692,
+      "grad_norm": 0.5243983864784241,
+      "learning_rate": 3.0938919122070123e-07,
+      "loss": 0.5399,
+      "step": 2173
+    },
+    {
+      "epoch": 0.8696,
+      "grad_norm": 0.5355934500694275,
+      "learning_rate": 3.075279650040092e-07,
+      "loss": 0.5985,
+      "step": 2174
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.5579320788383484,
+      "learning_rate": 3.0567209679460986e-07,
+      "loss": 0.6162,
+      "step": 2175
+    },
+    {
+      "epoch": 0.8704,
+      "grad_norm": 0.5460540056228638,
+      "learning_rate": 3.0382158970725485e-07,
+      "loss": 0.596,
+      "step": 2176
+    },
+    {
+      "epoch": 0.8708,
+      "grad_norm": 0.4951015114784241,
+      "learning_rate": 3.019764468476986e-07,
+      "loss": 0.4641,
+      "step": 2177
+    },
+    {
+      "epoch": 0.8712,
+      "grad_norm": 0.5001224279403687,
+      "learning_rate": 3.001366713126902e-07,
+      "loss": 0.5414,
+      "step": 2178
+    },
+    {
+      "epoch": 0.8716,
+      "grad_norm": 0.5136198997497559,
+      "learning_rate": 2.9830226618997115e-07,
+      "loss": 0.5539,
+      "step": 2179
+    },
+    {
+      "epoch": 0.872,
+      "grad_norm": 0.47107475996017456,
+      "learning_rate": 2.9647323455827183e-07,
+      "loss": 0.494,
+      "step": 2180
+    },
+    {
+      "epoch": 0.8724,
+      "grad_norm": 0.5539179444313049,
+      "learning_rate": 2.946495794873026e-07,
+      "loss": 0.552,
+      "step": 2181
+    },
+    {
+      "epoch": 0.8728,
+      "grad_norm": 0.5295327305793762,
+      "learning_rate": 2.9283130403774985e-07,
+      "loss": 0.6103,
+      "step": 2182
+    },
+    {
+      "epoch": 0.8732,
+      "grad_norm": 0.5619650483131409,
+      "learning_rate": 2.910184112612721e-07,
+      "loss": 0.5496,
+      "step": 2183
+    },
+    {
+      "epoch": 0.8736,
+      "grad_norm": 0.45123714208602905,
+      "learning_rate": 2.8921090420049183e-07,
+      "loss": 0.509,
+      "step": 2184
+    },
+    {
+      "epoch": 0.874,
+      "grad_norm": 0.5038043856620789,
+      "learning_rate": 2.874087858889968e-07,
+      "loss": 0.4946,
+      "step": 2185
+    },
+    {
+      "epoch": 0.8744,
+      "grad_norm": 0.5105406641960144,
+      "learning_rate": 2.856120593513271e-07,
+      "loss": 0.6208,
+      "step": 2186
+    },
+    {
+      "epoch": 0.8748,
+      "grad_norm": 0.5043431520462036,
+      "learning_rate": 2.838207276029743e-07,
+      "loss": 0.5475,
+      "step": 2187
+    },
+    {
+      "epoch": 0.8752,
+      "grad_norm": 0.56167072057724,
+      "learning_rate": 2.8203479365037663e-07,
+      "loss": 0.6108,
+      "step": 2188
+    },
+    {
+      "epoch": 0.8756,
+      "grad_norm": 0.5685185790061951,
+      "learning_rate": 2.802542604909121e-07,
+      "loss": 0.6608,
+      "step": 2189
+    },
+    {
+      "epoch": 0.876,
+      "grad_norm": 0.4665569067001343,
+      "learning_rate": 2.78479131112896e-07,
+      "loss": 0.4223,
+      "step": 2190
+    },
+    {
+      "epoch": 0.8764,
+      "grad_norm": 0.5965784788131714,
+      "learning_rate": 2.7670940849557263e-07,
+      "loss": 0.6361,
+      "step": 2191
+    },
+    {
+      "epoch": 0.8768,
+      "grad_norm": 0.5202442407608032,
+      "learning_rate": 2.7494509560911294e-07,
+      "loss": 0.5137,
+      "step": 2192
+    },
+    {
+      "epoch": 0.8772,
+      "grad_norm": 0.5134866833686829,
+      "learning_rate": 2.731861954146065e-07,
+      "loss": 0.5788,
+      "step": 2193
+    },
+    {
+      "epoch": 0.8776,
+      "grad_norm": 0.5923562049865723,
+      "learning_rate": 2.714327108640634e-07,
+      "loss": 0.5261,
+      "step": 2194
+    },
+    {
+      "epoch": 0.878,
+      "grad_norm": 0.49445661902427673,
+      "learning_rate": 2.6968464490039965e-07,
+      "loss": 0.51,
+      "step": 2195
+    },
+    {
+      "epoch": 0.8784,
+      "grad_norm": 0.4957909882068634,
+      "learning_rate": 2.6794200045743957e-07,
+      "loss": 0.4834,
+      "step": 2196
+    },
+    {
+      "epoch": 0.8788,
+      "grad_norm": 0.5249044299125671,
+      "learning_rate": 2.662047804599067e-07,
+      "loss": 0.4612,
+      "step": 2197
+    },
+    {
+      "epoch": 0.8792,
+      "grad_norm": 0.49063214659690857,
+      "learning_rate": 2.6447298782342277e-07,
+      "loss": 0.5146,
+      "step": 2198
+    },
+    {
+      "epoch": 0.8796,
+      "grad_norm": 0.5278512239456177,
+      "learning_rate": 2.6274662545449997e-07,
+      "loss": 0.5328,
+      "step": 2199
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.5886636972427368,
+      "learning_rate": 2.6102569625053557e-07,
+      "loss": 0.5251,
+      "step": 2200
+    },
+    {
+      "epoch": 0.8804,
+      "grad_norm": 0.5029667615890503,
+      "learning_rate": 2.5931020309980934e-07,
+      "loss": 0.5412,
+      "step": 2201
+    },
+    {
+      "epoch": 0.8808,
+      "grad_norm": 0.4710221588611603,
+      "learning_rate": 2.5760014888147604e-07,
+      "loss": 0.4977,
+      "step": 2202
+    },
+    {
+      "epoch": 0.8812,
+      "grad_norm": 0.5477889180183411,
+      "learning_rate": 2.5589553646556463e-07,
+      "loss": 0.5954,
+      "step": 2203
+    },
+    {
+      "epoch": 0.8816,
+      "grad_norm": 0.5483034253120422,
+      "learning_rate": 2.541963687129694e-07,
+      "loss": 0.4608,
+      "step": 2204
+    },
+    {
+      "epoch": 0.882,
+      "grad_norm": 0.46928951144218445,
+      "learning_rate": 2.525026484754473e-07,
+      "loss": 0.5349,
+      "step": 2205
+    },
+    {
+      "epoch": 0.8824,
+      "grad_norm": 0.498003751039505,
+      "learning_rate": 2.508143785956107e-07,
+      "loss": 0.4114,
+      "step": 2206
+    },
+    {
+      "epoch": 0.8828,
+      "grad_norm": 0.4926424026489258,
+      "learning_rate": 2.491315619069274e-07,
+      "loss": 0.5779,
+      "step": 2207
+    },
+    {
+      "epoch": 0.8832,
+      "grad_norm": 0.5372424125671387,
+      "learning_rate": 2.474542012337117e-07,
+      "loss": 0.592,
+      "step": 2208
+    },
+    {
+      "epoch": 0.8836,
+      "grad_norm": 0.5553046464920044,
+      "learning_rate": 2.4578229939112027e-07,
+      "loss": 0.5452,
+      "step": 2209
+    },
+    {
+      "epoch": 0.884,
+      "grad_norm": 0.5457509160041809,
+      "learning_rate": 2.441158591851484e-07,
+      "loss": 0.5042,
+      "step": 2210
+    },
+    {
+      "epoch": 0.8844,
+      "grad_norm": 0.5078551173210144,
+      "learning_rate": 2.424548834126262e-07,
+      "loss": 0.5067,
+      "step": 2211
+    },
+    {
+      "epoch": 0.8848,
+      "grad_norm": 0.5340900421142578,
+      "learning_rate": 2.4079937486121005e-07,
+      "loss": 0.4737,
+      "step": 2212
+    },
+    {
+      "epoch": 0.8852,
+      "grad_norm": 0.532534122467041,
+      "learning_rate": 2.391493363093842e-07,
+      "loss": 0.5541,
+      "step": 2213
+    },
+    {
+      "epoch": 0.8856,
+      "grad_norm": 0.5554724335670471,
+      "learning_rate": 2.3750477052644958e-07,
+      "loss": 0.5777,
+      "step": 2214
+    },
+    {
+      "epoch": 0.886,
+      "grad_norm": 0.5182691216468811,
+      "learning_rate": 2.3586568027252208e-07,
+      "loss": 0.6101,
+      "step": 2215
+    },
+    {
+      "epoch": 0.8864,
+      "grad_norm": 0.5171369314193726,
+      "learning_rate": 2.3423206829852994e-07,
+      "loss": 0.5527,
+      "step": 2216
+    },
+    {
+      "epoch": 0.8868,
+      "grad_norm": 0.5117000937461853,
+      "learning_rate": 2.3260393734620494e-07,
+      "loss": 0.4774,
+      "step": 2217
+    },
+    {
+      "epoch": 0.8872,
+      "grad_norm": 0.4515170753002167,
+      "learning_rate": 2.309812901480816e-07,
+      "loss": 0.4977,
+      "step": 2218
+    },
+    {
+      "epoch": 0.8876,
+      "grad_norm": 0.48999813199043274,
+      "learning_rate": 2.2936412942748946e-07,
+      "loss": 0.4982,
+      "step": 2219
+    },
+    {
+      "epoch": 0.888,
+      "grad_norm": 0.48553434014320374,
+      "learning_rate": 2.2775245789855103e-07,
+      "loss": 0.5583,
+      "step": 2220
+    },
+    {
+      "epoch": 0.8884,
+      "grad_norm": 0.7433586120605469,
+      "learning_rate": 2.2614627826617528e-07,
+      "loss": 0.5753,
+      "step": 2221
+    },
+    {
+      "epoch": 0.8888,
+      "grad_norm": 0.5436974167823792,
+      "learning_rate": 2.2454559322605566e-07,
+      "loss": 0.575,
+      "step": 2222
+    },
+    {
+      "epoch": 0.8892,
+      "grad_norm": 0.4806951880455017,
+      "learning_rate": 2.2295040546466156e-07,
+      "loss": 0.4775,
+      "step": 2223
+    },
+    {
+      "epoch": 0.8896,
+      "grad_norm": 0.4840577244758606,
+      "learning_rate": 2.2136071765923748e-07,
+      "loss": 0.5008,
+      "step": 2224
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.5452485680580139,
+      "learning_rate": 2.197765324777977e-07,
+      "loss": 0.5678,
+      "step": 2225
+    },
+    {
+      "epoch": 0.8904,
+      "grad_norm": 0.5001971125602722,
+      "learning_rate": 2.1819785257911923e-07,
+      "loss": 0.5008,
+      "step": 2226
+    },
+    {
+      "epoch": 0.8908,
+      "grad_norm": 0.5222870707511902,
+      "learning_rate": 2.1662468061274247e-07,
+      "loss": 0.5653,
+      "step": 2227
+    },
+    {
+      "epoch": 0.8912,
+      "grad_norm": 0.5725530982017517,
+      "learning_rate": 2.1505701921896096e-07,
+      "loss": 0.5058,
+      "step": 2228
+    },
+    {
+      "epoch": 0.8916,
+      "grad_norm": 0.5029703378677368,
+      "learning_rate": 2.1349487102882197e-07,
+      "loss": 0.517,
+      "step": 2229
+    },
+    {
+      "epoch": 0.892,
+      "grad_norm": 0.5157403945922852,
+      "learning_rate": 2.119382386641176e-07,
+      "loss": 0.5673,
+      "step": 2230
+    },
+    {
+      "epoch": 0.8924,
+      "grad_norm": 0.5268275141716003,
+      "learning_rate": 2.1038712473738362e-07,
+      "loss": 0.489,
+      "step": 2231
+    },
+    {
+      "epoch": 0.8928,
+      "grad_norm": 0.5436747670173645,
+      "learning_rate": 2.0884153185189492e-07,
+      "loss": 0.5786,
+      "step": 2232
+    },
+    {
+      "epoch": 0.8932,
+      "grad_norm": 0.4886317551136017,
+      "learning_rate": 2.0730146260165904e-07,
+      "loss": 0.5851,
+      "step": 2233
+    },
+    {
+      "epoch": 0.8936,
+      "grad_norm": 0.5080478191375732,
+      "learning_rate": 2.0576691957141374e-07,
+      "loss": 0.5666,
+      "step": 2234
+    },
+    {
+      "epoch": 0.894,
+      "grad_norm": 0.4885810613632202,
+      "learning_rate": 2.0423790533662096e-07,
+      "loss": 0.4711,
+      "step": 2235
+    },
+    {
+      "epoch": 0.8944,
+      "grad_norm": 0.5395861268043518,
+      "learning_rate": 2.0271442246346532e-07,
+      "loss": 0.4948,
+      "step": 2236
+    },
+    {
+      "epoch": 0.8948,
+      "grad_norm": 0.5157225131988525,
+      "learning_rate": 2.0119647350884646e-07,
+      "loss": 0.6207,
+      "step": 2237
+    },
+    {
+      "epoch": 0.8952,
+      "grad_norm": 0.5027074813842773,
+      "learning_rate": 1.9968406102037727e-07,
+      "loss": 0.5474,
+      "step": 2238
+    },
+    {
+      "epoch": 0.8956,
+      "grad_norm": 0.5285701155662537,
+      "learning_rate": 1.9817718753637793e-07,
+      "loss": 0.5634,
+      "step": 2239
+    },
+    {
+      "epoch": 0.896,
+      "grad_norm": 0.5563217997550964,
+      "learning_rate": 1.9667585558587253e-07,
+      "loss": 0.6273,
+      "step": 2240
+    },
+    {
+      "epoch": 0.8964,
+      "grad_norm": 0.5578852891921997,
+      "learning_rate": 1.9518006768858504e-07,
+      "loss": 0.5693,
+      "step": 2241
+    },
+    {
+      "epoch": 0.8968,
+      "grad_norm": 0.567421555519104,
+      "learning_rate": 1.9368982635493408e-07,
+      "loss": 0.609,
+      "step": 2242
+    },
+    {
+      "epoch": 0.8972,
+      "grad_norm": 0.5388922691345215,
+      "learning_rate": 1.9220513408603073e-07,
+      "loss": 0.5307,
+      "step": 2243
+    },
+    {
+      "epoch": 0.8976,
+      "grad_norm": 0.4816891551017761,
+      "learning_rate": 1.9072599337367103e-07,
+      "loss": 0.5399,
+      "step": 2244
+    },
+    {
+      "epoch": 0.898,
+      "grad_norm": 0.5367328524589539,
+      "learning_rate": 1.8925240670033404e-07,
+      "loss": 0.4935,
+      "step": 2245
+    },
+    {
+      "epoch": 0.8984,
+      "grad_norm": 0.4941747188568115,
+      "learning_rate": 1.877843765391789e-07,
+      "loss": 0.5399,
+      "step": 2246
+    },
+    {
+      "epoch": 0.8988,
+      "grad_norm": 0.539661169052124,
+      "learning_rate": 1.8632190535403835e-07,
+      "loss": 0.5073,
+      "step": 2247
+    },
+    {
+      "epoch": 0.8992,
+      "grad_norm": 0.5193576216697693,
+      "learning_rate": 1.8486499559941432e-07,
+      "loss": 0.5519,
+      "step": 2248
+    },
+    {
+      "epoch": 0.8996,
+      "grad_norm": 0.5331642031669617,
+      "learning_rate": 1.834136497204762e-07,
+      "loss": 0.6818,
+      "step": 2249
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.5258761644363403,
+      "learning_rate": 1.819678701530535e-07,
+      "loss": 0.584,
+      "step": 2250
+    },
+    {
+      "epoch": 0.9004,
+      "grad_norm": 0.549641489982605,
+      "learning_rate": 1.805276593236363e-07,
+      "loss": 0.6069,
+      "step": 2251
+    },
+    {
+      "epoch": 0.9008,
+      "grad_norm": 0.5493542551994324,
+      "learning_rate": 1.7909301964936742e-07,
+      "loss": 0.5378,
+      "step": 2252
+    },
+    {
+      "epoch": 0.9012,
+      "grad_norm": 0.5483373403549194,
+      "learning_rate": 1.7766395353803854e-07,
+      "loss": 0.5867,
+      "step": 2253
+    },
+    {
+      "epoch": 0.9016,
+      "grad_norm": 0.5108447670936584,
+      "learning_rate": 1.7624046338808713e-07,
+      "loss": 0.5701,
+      "step": 2254
+    },
+    {
+      "epoch": 0.902,
+      "grad_norm": 0.5171193480491638,
+      "learning_rate": 1.7482255158859445e-07,
+      "loss": 0.5197,
+      "step": 2255
+    },
+    {
+      "epoch": 0.9024,
+      "grad_norm": 0.5908972024917603,
+      "learning_rate": 1.734102205192774e-07,
+      "loss": 0.6117,
+      "step": 2256
+    },
+    {
+      "epoch": 0.9028,
+      "grad_norm": 0.4955964684486389,
+      "learning_rate": 1.7200347255048743e-07,
+      "loss": 0.5254,
+      "step": 2257
+    },
+    {
+      "epoch": 0.9032,
+      "grad_norm": 0.5056068897247314,
+      "learning_rate": 1.7060231004320537e-07,
+      "loss": 0.6004,
+      "step": 2258
+    },
+    {
+      "epoch": 0.9036,
+      "grad_norm": 0.48350733518600464,
+      "learning_rate": 1.69206735349038e-07,
+      "loss": 0.5441,
+      "step": 2259
+    },
+    {
+      "epoch": 0.904,
+      "grad_norm": 0.5320057272911072,
+      "learning_rate": 1.6781675081021348e-07,
+      "loss": 0.513,
+      "step": 2260
+    },
+    {
+      "epoch": 0.9044,
+      "grad_norm": 0.5606966614723206,
+      "learning_rate": 1.6643235875957951e-07,
+      "loss": 0.5428,
+      "step": 2261
+    },
+    {
+      "epoch": 0.9048,
+      "grad_norm": 0.5294163227081299,
+      "learning_rate": 1.650535615205963e-07,
+      "loss": 0.5392,
+      "step": 2262
+    },
+    {
+      "epoch": 0.9052,
+      "grad_norm": 0.5900059938430786,
+      "learning_rate": 1.6368036140733323e-07,
+      "loss": 0.5982,
+      "step": 2263
+    },
+    {
+      "epoch": 0.9056,
+      "grad_norm": 0.5253335237503052,
+      "learning_rate": 1.6231276072446808e-07,
+      "loss": 0.5925,
+      "step": 2264
+    },
+    {
+      "epoch": 0.906,
+      "grad_norm": 0.4768005907535553,
+      "learning_rate": 1.6095076176728062e-07,
+      "loss": 0.5083,
+      "step": 2265
+    },
+    {
+      "epoch": 0.9064,
+      "grad_norm": 0.5210567116737366,
+      "learning_rate": 1.595943668216473e-07,
+      "loss": 0.5637,
+      "step": 2266
+    },
+    {
+      "epoch": 0.9068,
+      "grad_norm": 0.4880336821079254,
+      "learning_rate": 1.58243578164041e-07,
+      "loss": 0.5667,
+      "step": 2267
+    },
+    {
+      "epoch": 0.9072,
+      "grad_norm": 0.48938196897506714,
+      "learning_rate": 1.5689839806152405e-07,
+      "loss": 0.4602,
+      "step": 2268
+    },
+    {
+      "epoch": 0.9076,
+      "grad_norm": 0.5602405667304993,
+      "learning_rate": 1.5555882877174703e-07,
+      "loss": 0.6031,
+      "step": 2269
+    },
+    {
+      "epoch": 0.908,
+      "grad_norm": 0.5172474980354309,
+      "learning_rate": 1.542248725429441e-07,
+      "loss": 0.4896,
+      "step": 2270
+    },
+    {
+      "epoch": 0.9084,
+      "grad_norm": 0.5165852308273315,
+      "learning_rate": 1.5289653161392674e-07,
+      "loss": 0.49,
+      "step": 2271
+    },
+    {
+      "epoch": 0.9088,
+      "grad_norm": 0.5281601548194885,
+      "learning_rate": 1.515738082140839e-07,
+      "loss": 0.5399,
+      "step": 2272
+    },
+    {
+      "epoch": 0.9092,
+      "grad_norm": 0.4762950539588928,
+      "learning_rate": 1.5025670456337596e-07,
+      "loss": 0.5318,
+      "step": 2273
+    },
+    {
+      "epoch": 0.9096,
+      "grad_norm": 0.5265024304389954,
+      "learning_rate": 1.4894522287233226e-07,
+      "loss": 0.5934,
+      "step": 2274
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.49703449010849,
+      "learning_rate": 1.4763936534204537e-07,
+      "loss": 0.5519,
+      "step": 2275
+    },
+    {
+      "epoch": 0.9104,
+      "grad_norm": 0.5251654982566833,
+      "learning_rate": 1.463391341641694e-07,
+      "loss": 0.4945,
+      "step": 2276
+    },
+    {
+      "epoch": 0.9108,
+      "grad_norm": 0.45492956042289734,
+      "learning_rate": 1.4504453152091479e-07,
+      "loss": 0.4593,
+      "step": 2277
+    },
+    {
+      "epoch": 0.9112,
+      "grad_norm": 0.5544923543930054,
+      "learning_rate": 1.4375555958504727e-07,
+      "loss": 0.5885,
+      "step": 2278
+    },
+    {
+      "epoch": 0.9116,
+      "grad_norm": 0.5534620881080627,
+      "learning_rate": 1.4247222051988152e-07,
+      "loss": 0.6237,
+      "step": 2279
+    },
+    {
+      "epoch": 0.912,
+      "grad_norm": 0.49389734864234924,
+      "learning_rate": 1.4119451647927732e-07,
+      "loss": 0.431,
+      "step": 2280
+    },
+    {
+      "epoch": 0.9124,
+      "grad_norm": 0.4699340760707855,
+      "learning_rate": 1.399224496076381e-07,
+      "loss": 0.538,
+      "step": 2281
+    },
+    {
+      "epoch": 0.9128,
+      "grad_norm": 0.5014921426773071,
+      "learning_rate": 1.3865602203990652e-07,
+      "loss": 0.5036,
+      "step": 2282
+    },
+    {
+      "epoch": 0.9132,
+      "grad_norm": 0.46775099635124207,
+      "learning_rate": 1.3739523590156022e-07,
+      "loss": 0.4506,
+      "step": 2283
+    },
+    {
+      "epoch": 0.9136,
+      "grad_norm": 0.5627043843269348,
+      "learning_rate": 1.3614009330860922e-07,
+      "loss": 0.5204,
+      "step": 2284
+    },
+    {
+      "epoch": 0.914,
+      "grad_norm": 0.631554901599884,
+      "learning_rate": 1.348905963675903e-07,
+      "loss": 0.5127,
+      "step": 2285
+    },
+    {
+      "epoch": 0.9144,
+      "grad_norm": 0.5087348818778992,
+      "learning_rate": 1.3364674717556647e-07,
+      "loss": 0.545,
+      "step": 2286
+    },
+    {
+      "epoch": 0.9148,
+      "grad_norm": 0.5642393231391907,
+      "learning_rate": 1.3240854782012195e-07,
+      "loss": 0.5813,
+      "step": 2287
+    },
+    {
+      "epoch": 0.9152,
+      "grad_norm": 0.46905118227005005,
+      "learning_rate": 1.3117600037935827e-07,
+      "loss": 0.4888,
+      "step": 2288
+    },
+    {
+      "epoch": 0.9156,
+      "grad_norm": 0.5647628903388977,
+      "learning_rate": 1.2994910692189132e-07,
+      "loss": 0.5614,
+      "step": 2289
+    },
+    {
+      "epoch": 0.916,
+      "grad_norm": 0.5584691166877747,
+      "learning_rate": 1.2872786950684672e-07,
+      "loss": 0.5206,
+      "step": 2290
+    },
+    {
+      "epoch": 0.9164,
+      "grad_norm": 0.5193395018577576,
+      "learning_rate": 1.2751229018385952e-07,
+      "loss": 0.4593,
+      "step": 2291
+    },
+    {
+      "epoch": 0.9168,
+      "grad_norm": 0.5479791164398193,
+      "learning_rate": 1.2630237099306707e-07,
+      "loss": 0.5706,
+      "step": 2292
+    },
+    {
+      "epoch": 0.9172,
+      "grad_norm": 0.5571714639663696,
+      "learning_rate": 1.25098113965108e-07,
+      "loss": 0.5249,
+      "step": 2293
+    },
+    {
+      "epoch": 0.9176,
+      "grad_norm": 0.5249247550964355,
+      "learning_rate": 1.2389952112111708e-07,
+      "loss": 0.573,
+      "step": 2294
+    },
+    {
+      "epoch": 0.918,
+      "grad_norm": 0.5296669006347656,
+      "learning_rate": 1.2270659447272326e-07,
+      "loss": 0.5324,
+      "step": 2295
+    },
+    {
+      "epoch": 0.9184,
+      "grad_norm": 0.5688275098800659,
+      "learning_rate": 1.2151933602204622e-07,
+      "loss": 0.5893,
+      "step": 2296
+    },
+    {
+      "epoch": 0.9188,
+      "grad_norm": 0.5549254417419434,
+      "learning_rate": 1.2033774776169214e-07,
+      "loss": 0.554,
+      "step": 2297
+    },
+    {
+      "epoch": 0.9192,
+      "grad_norm": 0.5588293671607971,
+      "learning_rate": 1.1916183167475047e-07,
+      "loss": 0.5978,
+      "step": 2298
+    },
+    {
+      "epoch": 0.9196,
+      "grad_norm": 0.515630841255188,
+      "learning_rate": 1.179915897347909e-07,
+      "loss": 0.4945,
+      "step": 2299
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.5234367847442627,
+      "learning_rate": 1.1682702390586102e-07,
+      "loss": 0.5629,
+      "step": 2300
+    },
+    {
+      "epoch": 0.9204,
+      "grad_norm": 0.5144403576850891,
+      "learning_rate": 1.156681361424805e-07,
+      "loss": 0.6108,
+      "step": 2301
+    },
+    {
+      "epoch": 0.9208,
+      "grad_norm": 0.4699039161205292,
+      "learning_rate": 1.1451492838964145e-07,
+      "loss": 0.468,
+      "step": 2302
+    },
+    {
+      "epoch": 0.9212,
+      "grad_norm": 0.5121269822120667,
+      "learning_rate": 1.1336740258280142e-07,
+      "loss": 0.5558,
+      "step": 2303
+    },
+    {
+      "epoch": 0.9216,
+      "grad_norm": 0.5284686088562012,
+      "learning_rate": 1.1222556064788153e-07,
+      "loss": 0.5622,
+      "step": 2304
+    },
+    {
+      "epoch": 0.922,
+      "grad_norm": 0.5822998285293579,
+      "learning_rate": 1.1108940450126481e-07,
+      "loss": 0.5311,
+      "step": 2305
+    },
+    {
+      "epoch": 0.9224,
+      "grad_norm": 0.5090004801750183,
+      "learning_rate": 1.0995893604979124e-07,
+      "loss": 0.5689,
+      "step": 2306
+    },
+    {
+      "epoch": 0.9228,
+      "grad_norm": 0.5638589859008789,
+      "learning_rate": 1.0883415719075457e-07,
+      "loss": 0.6537,
+      "step": 2307
+    },
+    {
+      "epoch": 0.9232,
+      "grad_norm": 0.49607881903648376,
+      "learning_rate": 1.0771506981190004e-07,
+      "loss": 0.5467,
+      "step": 2308
+    },
+    {
+      "epoch": 0.9236,
+      "grad_norm": 0.5245644450187683,
+      "learning_rate": 1.0660167579142088e-07,
+      "loss": 0.4748,
+      "step": 2309
+    },
+    {
+      "epoch": 0.924,
+      "grad_norm": 0.5582762956619263,
+      "learning_rate": 1.0549397699795438e-07,
+      "loss": 0.557,
+      "step": 2310
+    },
+    {
+      "epoch": 0.9244,
+      "grad_norm": 0.5717631578445435,
+      "learning_rate": 1.0439197529057881e-07,
+      "loss": 0.6091,
+      "step": 2311
+    },
+    {
+      "epoch": 0.9248,
+      "grad_norm": 0.4912102222442627,
+      "learning_rate": 1.0329567251881344e-07,
+      "loss": 0.5659,
+      "step": 2312
+    },
+    {
+      "epoch": 0.9252,
+      "grad_norm": 0.52315354347229,
+      "learning_rate": 1.0220507052260997e-07,
+      "loss": 0.6072,
+      "step": 2313
+    },
+    {
+      "epoch": 0.9256,
+      "grad_norm": 0.4884204566478729,
+      "learning_rate": 1.0112017113235444e-07,
+      "loss": 0.4823,
+      "step": 2314
+    },
+    {
+      "epoch": 0.926,
+      "grad_norm": 0.5286775231361389,
+      "learning_rate": 1.0004097616885993e-07,
+      "loss": 0.6211,
+      "step": 2315
+    },
+    {
+      "epoch": 0.9264,
+      "grad_norm": 0.5312598347663879,
+      "learning_rate": 9.896748744336842e-08,
+      "loss": 0.5514,
+      "step": 2316
+    },
+    {
+      "epoch": 0.9268,
+      "grad_norm": 0.4938899278640747,
+      "learning_rate": 9.78997067575419e-08,
+      "loss": 0.5404,
+      "step": 2317
+    },
+    {
+      "epoch": 0.9272,
+      "grad_norm": 0.5511752367019653,
+      "learning_rate": 9.683763590346544e-08,
+      "loss": 0.553,
+      "step": 2318
+    },
+    {
+      "epoch": 0.9276,
+      "grad_norm": 0.5150926113128662,
+      "learning_rate": 9.578127666363872e-08,
+      "loss": 0.4999,
+      "step": 2319
+    },
+    {
+      "epoch": 0.928,
+      "grad_norm": 0.5279099941253662,
+      "learning_rate": 9.473063081097633e-08,
+      "loss": 0.5558,
+      "step": 2320
+    },
+    {
+      "epoch": 0.9284,
+      "grad_norm": 0.5396522283554077,
+      "learning_rate": 9.368570010880467e-08,
+      "loss": 0.5214,
+      "step": 2321
+    },
+    {
+      "epoch": 0.9288,
+      "grad_norm": 0.5016295313835144,
+      "learning_rate": 9.264648631085697e-08,
+      "loss": 0.537,
+      "step": 2322
+    },
+    {
+      "epoch": 0.9292,
+      "grad_norm": 0.5354951024055481,
+      "learning_rate": 9.161299116127324e-08,
+      "loss": 0.5904,
+      "step": 2323
+    },
+    {
+      "epoch": 0.9296,
+      "grad_norm": 0.46822237968444824,
+      "learning_rate": 9.058521639459366e-08,
+      "loss": 0.5,
+      "step": 2324
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.5241938829421997,
+      "learning_rate": 8.956316373575896e-08,
+      "loss": 0.6066,
+      "step": 2325
+    },
+    {
+      "epoch": 0.9304,
+      "grad_norm": 0.5749825835227966,
+      "learning_rate": 8.854683490010656e-08,
+      "loss": 0.5586,
+      "step": 2326
+    },
+    {
+      "epoch": 0.9308,
+      "grad_norm": 0.5310881733894348,
+      "learning_rate": 8.753623159336748e-08,
+      "loss": 0.5686,
+      "step": 2327
+    },
+    {
+      "epoch": 0.9312,
+      "grad_norm": 0.47297438979148865,
+      "learning_rate": 8.653135551166241e-08,
+      "loss": 0.5044,
+      "step": 2328
+    },
+    {
+      "epoch": 0.9316,
+      "grad_norm": 0.49978524446487427,
+      "learning_rate": 8.553220834150055e-08,
+      "loss": 0.4801,
+      "step": 2329
+    },
+    {
+      "epoch": 0.932,
+      "grad_norm": 0.4736655056476593,
+      "learning_rate": 8.453879175977614e-08,
+      "loss": 0.5754,
+      "step": 2330
+    },
+    {
+      "epoch": 0.9324,
+      "grad_norm": 0.5332768559455872,
+      "learning_rate": 8.355110743376648e-08,
+      "loss": 0.5547,
+      "step": 2331
+    },
+    {
+      "epoch": 0.9328,
+      "grad_norm": 0.4920896589756012,
+      "learning_rate": 8.256915702112616e-08,
+      "loss": 0.4997,
+      "step": 2332
+    },
+    {
+      "epoch": 0.9332,
+      "grad_norm": 0.5521984100341797,
+      "learning_rate": 8.159294216988816e-08,
+      "loss": 0.5899,
+      "step": 2333
+    },
+    {
+      "epoch": 0.9336,
+      "grad_norm": 0.5347005128860474,
+      "learning_rate": 8.062246451845845e-08,
+      "loss": 0.5885,
+      "step": 2334
+    },
+    {
+      "epoch": 0.934,
+      "grad_norm": 0.5441831946372986,
+      "learning_rate": 7.96577256956144e-08,
+      "loss": 0.5231,
+      "step": 2335
+    },
+    {
+      "epoch": 0.9344,
+      "grad_norm": 0.5577284097671509,
+      "learning_rate": 7.869872732050253e-08,
+      "loss": 0.5694,
+      "step": 2336
+    },
+    {
+      "epoch": 0.9348,
+      "grad_norm": 0.533072829246521,
+      "learning_rate": 7.774547100263413e-08,
+      "loss": 0.5164,
+      "step": 2337
+    },
+    {
+      "epoch": 0.9352,
+      "grad_norm": 0.5311025381088257,
+      "learning_rate": 7.679795834188296e-08,
+      "loss": 0.562,
+      "step": 2338
+    },
+    {
+      "epoch": 0.9356,
+      "grad_norm": 0.5785207152366638,
+      "learning_rate": 7.585619092848339e-08,
+      "loss": 0.5512,
+      "step": 2339
+    },
+    {
+      "epoch": 0.936,
+      "grad_norm": 0.5167658925056458,
+      "learning_rate": 7.49201703430291e-08,
+      "loss": 0.5515,
+      "step": 2340
+    },
+    {
+      "epoch": 0.9364,
+      "grad_norm": 0.5739880204200745,
+      "learning_rate": 7.398989815646662e-08,
+      "loss": 0.5778,
+      "step": 2341
+    },
+    {
+      "epoch": 0.9368,
+      "grad_norm": 0.5385161638259888,
+      "learning_rate": 7.306537593009555e-08,
+      "loss": 0.5224,
+      "step": 2342
+    },
+    {
+      "epoch": 0.9372,
+      "grad_norm": 0.4797918498516083,
+      "learning_rate": 7.214660521556482e-08,
+      "loss": 0.5574,
+      "step": 2343
+    },
+    {
+      "epoch": 0.9376,
+      "grad_norm": 0.5047261714935303,
+      "learning_rate": 7.123358755487069e-08,
+      "loss": 0.5462,
+      "step": 2344
+    },
+    {
+      "epoch": 0.938,
+      "grad_norm": 0.5187482237815857,
+      "learning_rate": 7.032632448035481e-08,
+      "loss": 0.4779,
+      "step": 2345
+    },
+    {
+      "epoch": 0.9384,
+      "grad_norm": 0.4853595197200775,
+      "learning_rate": 6.942481751469953e-08,
+      "loss": 0.5042,
+      "step": 2346
+    },
+    {
+      "epoch": 0.9388,
+      "grad_norm": 0.5323401093482971,
+      "learning_rate": 6.85290681709268e-08,
+      "loss": 0.5266,
+      "step": 2347
+    },
+    {
+      "epoch": 0.9392,
+      "grad_norm": 0.5271187424659729,
+      "learning_rate": 6.763907795239538e-08,
+      "loss": 0.4735,
+      "step": 2348
+    },
+    {
+      "epoch": 0.9396,
+      "grad_norm": 0.5426302552223206,
+      "learning_rate": 6.675484835279932e-08,
+      "loss": 0.5779,
+      "step": 2349
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.580814778804779,
+      "learning_rate": 6.587638085616332e-08,
+      "loss": 0.5807,
+      "step": 2350
+    },
+    {
+      "epoch": 0.9404,
+      "grad_norm": 0.5777420997619629,
+      "learning_rate": 6.500367693684189e-08,
+      "loss": 0.5932,
+      "step": 2351
+    },
+    {
+      "epoch": 0.9408,
+      "grad_norm": 0.5483967661857605,
+      "learning_rate": 6.413673805951591e-08,
+      "loss": 0.6241,
+      "step": 2352
+    },
+    {
+      "epoch": 0.9412,
+      "grad_norm": 0.49772343039512634,
+      "learning_rate": 6.327556567919146e-08,
+      "loss": 0.4763,
+      "step": 2353
+    },
+    {
+      "epoch": 0.9416,
+      "grad_norm": 0.5433846712112427,
+      "learning_rate": 6.242016124119626e-08,
+      "loss": 0.5387,
+      "step": 2354
+    },
+    {
+      "epoch": 0.942,
+      "grad_norm": 0.5307440161705017,
+      "learning_rate": 6.157052618117703e-08,
+      "loss": 0.5218,
+      "step": 2355
+    },
+    {
+      "epoch": 0.9424,
+      "grad_norm": 0.5752222537994385,
+      "learning_rate": 6.072666192509751e-08,
+      "loss": 0.5562,
+      "step": 2356
+    },
+    {
+      "epoch": 0.9428,
+      "grad_norm": 0.5130374431610107,
+      "learning_rate": 5.988856988923769e-08,
+      "loss": 0.4861,
+      "step": 2357
+    },
+    {
+      "epoch": 0.9432,
+      "grad_norm": 0.5313212871551514,
+      "learning_rate": 5.905625148018834e-08,
+      "loss": 0.5385,
+      "step": 2358
+    },
+    {
+      "epoch": 0.9436,
+      "grad_norm": 0.5523070693016052,
+      "learning_rate": 5.8229708094850676e-08,
+      "loss": 0.5731,
+      "step": 2359
+    },
+    {
+      "epoch": 0.944,
+      "grad_norm": 0.526229202747345,
+      "learning_rate": 5.74089411204336e-08,
+      "loss": 0.5016,
+      "step": 2360
+    },
+    {
+      "epoch": 0.9444,
+      "grad_norm": 0.5088282823562622,
+      "learning_rate": 5.6593951934450994e-08,
+      "loss": 0.5887,
+      "step": 2361
+    },
+    {
+      "epoch": 0.9448,
+      "grad_norm": 0.5603100061416626,
+      "learning_rate": 5.578474190472016e-08,
+      "loss": 0.5791,
+      "step": 2362
+    },
+    {
+      "epoch": 0.9452,
+      "grad_norm": 0.4806862771511078,
+      "learning_rate": 5.4981312389359494e-08,
+      "loss": 0.5215,
+      "step": 2363
+    },
+    {
+      "epoch": 0.9456,
+      "grad_norm": 0.5049216151237488,
+      "learning_rate": 5.41836647367846e-08,
+      "loss": 0.5432,
+      "step": 2364
+    },
+    {
+      "epoch": 0.946,
+      "grad_norm": 0.5518359541893005,
+      "learning_rate": 5.339180028570789e-08,
+      "loss": 0.4736,
+      "step": 2365
+    },
+    {
+      "epoch": 0.9464,
+      "grad_norm": 0.5348437428474426,
+      "learning_rate": 5.260572036513589e-08,
+      "loss": 0.4547,
+      "step": 2366
+    },
+    {
+      "epoch": 0.9468,
+      "grad_norm": 0.5421221852302551,
+      "learning_rate": 5.182542629436648e-08,
+      "loss": 0.5452,
+      "step": 2367
+    },
+    {
+      "epoch": 0.9472,
+      "grad_norm": 0.4720580279827118,
+      "learning_rate": 5.105091938298739e-08,
+      "loss": 0.5686,
+      "step": 2368
+    },
+    {
+      "epoch": 0.9476,
+      "grad_norm": 0.5404186248779297,
+      "learning_rate": 5.028220093087343e-08,
+      "loss": 0.562,
+      "step": 2369
+    },
+    {
+      "epoch": 0.948,
+      "grad_norm": 0.5333725810050964,
+      "learning_rate": 4.951927222818381e-08,
+      "loss": 0.5939,
+      "step": 2370
+    },
+    {
+      "epoch": 0.9484,
+      "grad_norm": 0.5129185914993286,
+      "learning_rate": 4.876213455536171e-08,
+      "loss": 0.5754,
+      "step": 2371
+    },
+    {
+      "epoch": 0.9488,
+      "grad_norm": 0.5559066534042358,
+      "learning_rate": 4.8010789183130066e-08,
+      "loss": 0.5132,
+      "step": 2372
+    },
+    {
+      "epoch": 0.9492,
+      "grad_norm": 0.5458990335464478,
+      "learning_rate": 4.7265237372491883e-08,
+      "loss": 0.5678,
+      "step": 2373
+    },
+    {
+      "epoch": 0.9496,
+      "grad_norm": 0.5720266699790955,
+      "learning_rate": 4.652548037472487e-08,
+      "loss": 0.5237,
+      "step": 2374
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.5585337281227112,
+      "learning_rate": 4.5791519431382544e-08,
+      "loss": 0.5886,
+      "step": 2375
+    },
+    {
+      "epoch": 0.9504,
+      "grad_norm": 0.49161794781684875,
+      "learning_rate": 4.506335577429038e-08,
+      "loss": 0.4594,
+      "step": 2376
+    },
+    {
+      "epoch": 0.9508,
+      "grad_norm": 0.4829513430595398,
+      "learning_rate": 4.434099062554386e-08,
+      "loss": 0.6157,
+      "step": 2377
+    },
+    {
+      "epoch": 0.9512,
+      "grad_norm": 0.4910193085670471,
+      "learning_rate": 4.362442519750692e-08,
+      "loss": 0.5071,
+      "step": 2378
+    },
+    {
+      "epoch": 0.9516,
+      "grad_norm": 0.5328912734985352,
+      "learning_rate": 4.291366069280961e-08,
+      "loss": 0.5709,
+      "step": 2379
+    },
+    {
+      "epoch": 0.952,
+      "grad_norm": 0.6091710329055786,
+      "learning_rate": 4.2208698304346145e-08,
+      "loss": 0.4926,
+      "step": 2380
+    },
+    {
+      "epoch": 0.9524,
+      "grad_norm": 0.5346353054046631,
+      "learning_rate": 4.150953921527262e-08,
+      "loss": 0.5452,
+      "step": 2381
+    },
+    {
+      "epoch": 0.9528,
+      "grad_norm": 0.45982980728149414,
+      "learning_rate": 4.0816184599005764e-08,
+      "loss": 0.5673,
+      "step": 2382
+    },
+    {
+      "epoch": 0.9532,
+      "grad_norm": 0.5378580093383789,
+      "learning_rate": 4.012863561922031e-08,
+      "loss": 0.5315,
+      "step": 2383
+    },
+    {
+      "epoch": 0.9536,
+      "grad_norm": 0.5490142703056335,
+      "learning_rate": 3.944689342984775e-08,
+      "loss": 0.5656,
+      "step": 2384
+    },
+    {
+      "epoch": 0.954,
+      "grad_norm": 0.5665016770362854,
+      "learning_rate": 3.8770959175072506e-08,
+      "loss": 0.5158,
+      "step": 2385
+    },
+    {
+      "epoch": 0.9544,
+      "grad_norm": 0.5378921031951904,
+      "learning_rate": 3.810083398933267e-08,
+      "loss": 0.5287,
+      "step": 2386
+    },
+    {
+      "epoch": 0.9548,
+      "grad_norm": 0.4553784132003784,
+      "learning_rate": 3.7436518997316926e-08,
+      "loss": 0.4685,
+      "step": 2387
+    },
+    {
+      "epoch": 0.9552,
+      "grad_norm": 0.5257003903388977,
+      "learning_rate": 3.6778015313961406e-08,
+      "loss": 0.5858,
+      "step": 2388
+    },
+    {
+      "epoch": 0.9556,
+      "grad_norm": 0.5115745067596436,
+      "learning_rate": 3.6125324044450123e-08,
+      "loss": 0.5712,
+      "step": 2389
+    },
+    {
+      "epoch": 0.956,
+      "grad_norm": 0.5314233899116516,
+      "learning_rate": 3.547844628421104e-08,
+      "loss": 0.5518,
+      "step": 2390
+    },
+    {
+      "epoch": 0.9564,
+      "grad_norm": 0.5423491597175598,
+      "learning_rate": 3.483738311891571e-08,
+      "loss": 0.6051,
+      "step": 2391
+    },
+    {
+      "epoch": 0.9568,
+      "grad_norm": 0.5177218317985535,
+      "learning_rate": 3.420213562447733e-08,
+      "loss": 0.5818,
+      "step": 2392
+    },
+    {
+      "epoch": 0.9572,
+      "grad_norm": 0.5107486844062805,
+      "learning_rate": 3.357270486704722e-08,
+      "loss": 0.5967,
+      "step": 2393
+    },
+    {
+      "epoch": 0.9576,
+      "grad_norm": 0.4815487265586853,
+      "learning_rate": 3.2949091903016033e-08,
+      "loss": 0.5463,
+      "step": 2394
+    },
+    {
+      "epoch": 0.958,
+      "grad_norm": 0.5988988280296326,
+      "learning_rate": 3.2331297779008286e-08,
+      "loss": 0.5769,
+      "step": 2395
+    },
+    {
+      "epoch": 0.9584,
+      "grad_norm": 0.6041003465652466,
+      "learning_rate": 3.171932353188389e-08,
+      "loss": 0.5919,
+      "step": 2396
+    },
+    {
+      "epoch": 0.9588,
+      "grad_norm": 0.4653492569923401,
+      "learning_rate": 3.1113170188735494e-08,
+      "loss": 0.5024,
+      "step": 2397
+    },
+    {
+      "epoch": 0.9592,
+      "grad_norm": 0.4932403266429901,
+      "learning_rate": 3.051283876688493e-08,
+      "loss": 0.5802,
+      "step": 2398
+    },
+    {
+      "epoch": 0.9596,
+      "grad_norm": 0.4975288212299347,
+      "learning_rate": 2.991833027388441e-08,
+      "loss": 0.5104,
+      "step": 2399
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.4812232255935669,
+      "learning_rate": 2.932964570751223e-08,
+      "loss": 0.4926,
+      "step": 2400
+    },
+    {
+      "epoch": 0.9604,
+      "grad_norm": 0.5227335691452026,
+      "learning_rate": 2.87467860557728e-08,
+      "loss": 0.4955,
+      "step": 2401
+    },
+    {
+      "epoch": 0.9608,
+      "grad_norm": 0.5533686876296997,
+      "learning_rate": 2.816975229689428e-08,
+      "loss": 0.4968,
+      "step": 2402
+    },
+    {
+      "epoch": 0.9612,
+      "grad_norm": 0.5150798559188843,
+      "learning_rate": 2.759854539932782e-08,
+      "loss": 0.6334,
+      "step": 2403
+    },
+    {
+      "epoch": 0.9616,
+      "grad_norm": 0.5787931680679321,
+      "learning_rate": 2.7033166321743296e-08,
+      "loss": 0.4845,
+      "step": 2404
+    },
+    {
+      "epoch": 0.962,
+      "grad_norm": 0.5208541750907898,
+      "learning_rate": 2.6473616013032385e-08,
+      "loss": 0.5725,
+      "step": 2405
+    },
+    {
+      "epoch": 0.9624,
+      "grad_norm": 0.49922534823417664,
+      "learning_rate": 2.5919895412301218e-08,
+      "loss": 0.6088,
+      "step": 2406
+    },
+    {
+      "epoch": 0.9628,
+      "grad_norm": 0.48126330971717834,
+      "learning_rate": 2.5372005448873856e-08,
+      "loss": 0.5589,
+      "step": 2407
+    },
+    {
+      "epoch": 0.9632,
+      "grad_norm": 0.4809359610080719,
+      "learning_rate": 2.4829947042288412e-08,
+      "loss": 0.5026,
+      "step": 2408
+    },
+    {
+      "epoch": 0.9636,
+      "grad_norm": 0.5534669756889343,
+      "learning_rate": 2.429372110229433e-08,
+      "loss": 0.5381,
+      "step": 2409
+    },
+    {
+      "epoch": 0.964,
+      "grad_norm": 0.5311241149902344,
+      "learning_rate": 2.3763328528853932e-08,
+      "loss": 0.562,
+      "step": 2410
+    },
+    {
+      "epoch": 0.9644,
+      "grad_norm": 0.5506380200386047,
+      "learning_rate": 2.323877021213894e-08,
+      "loss": 0.5741,
+      "step": 2411
+    },
+    {
+      "epoch": 0.9648,
+      "grad_norm": 0.5539811253547668,
+      "learning_rate": 2.27200470325285e-08,
+      "loss": 0.5573,
+      "step": 2412
+    },
+    {
+      "epoch": 0.9652,
+      "grad_norm": 0.5888808965682983,
+      "learning_rate": 2.2207159860608837e-08,
+      "loss": 0.6698,
+      "step": 2413
+    },
+    {
+      "epoch": 0.9656,
+      "grad_norm": 0.5768120884895325,
+      "learning_rate": 2.1700109557171666e-08,
+      "loss": 0.5487,
+      "step": 2414
+    },
+    {
+      "epoch": 0.966,
+      "grad_norm": 0.5086904168128967,
+      "learning_rate": 2.1198896973212643e-08,
+      "loss": 0.6062,
+      "step": 2415
+    },
+    {
+      "epoch": 0.9664,
+      "grad_norm": 0.5336338877677917,
+      "learning_rate": 2.0703522949929833e-08,
+      "loss": 0.5496,
+      "step": 2416
+    },
+    {
+      "epoch": 0.9668,
+      "grad_norm": 0.5526391863822937,
+      "learning_rate": 2.021398831872173e-08,
+      "loss": 0.5389,
+      "step": 2417
+    },
+    {
+      "epoch": 0.9672,
+      "grad_norm": 0.49860087037086487,
+      "learning_rate": 1.9730293901186522e-08,
+      "loss": 0.5651,
+      "step": 2418
+    },
+    {
+      "epoch": 0.9676,
+      "grad_norm": 0.509070873260498,
+      "learning_rate": 1.925244050912167e-08,
+      "loss": 0.5729,
+      "step": 2419
+    },
+    {
+      "epoch": 0.968,
+      "grad_norm": 0.5186116695404053,
+      "learning_rate": 1.8780428944520035e-08,
+      "loss": 0.5349,
+      "step": 2420
+    },
+    {
+      "epoch": 0.9684,
+      "grad_norm": 0.6276172995567322,
+      "learning_rate": 1.8314259999571426e-08,
+      "loss": 0.6,
+      "step": 2421
+    },
+    {
+      "epoch": 0.9688,
+      "grad_norm": 0.6774277091026306,
+      "learning_rate": 1.7853934456658336e-08,
+      "loss": 0.6064,
+      "step": 2422
+    },
+    {
+      "epoch": 0.9692,
+      "grad_norm": 0.4820622205734253,
+      "learning_rate": 1.739945308835672e-08,
+      "loss": 0.5151,
+      "step": 2423
+    },
+    {
+      "epoch": 0.9696,
+      "grad_norm": 0.5160215497016907,
+      "learning_rate": 1.6950816657435198e-08,
+      "loss": 0.5131,
+      "step": 2424
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.5508131384849548,
+      "learning_rate": 1.650802591685119e-08,
+      "loss": 0.5214,
+      "step": 2425
+    },
+    {
+      "epoch": 0.9704,
+      "grad_norm": 0.5874997973442078,
+      "learning_rate": 1.607108160975168e-08,
+      "loss": 0.615,
+      "step": 2426
+    },
+    {
+      "epoch": 0.9708,
+      "grad_norm": 0.5136599540710449,
+      "learning_rate": 1.563998446947129e-08,
+      "loss": 0.5952,
+      "step": 2427
+    },
+    {
+      "epoch": 0.9712,
+      "grad_norm": 0.5393446087837219,
+      "learning_rate": 1.5214735219531484e-08,
+      "loss": 0.5409,
+      "step": 2428
+    },
+    {
+      "epoch": 0.9716,
+      "grad_norm": 0.5421205759048462,
+      "learning_rate": 1.4795334573639407e-08,
+      "loss": 0.5534,
+      "step": 2429
+    },
+    {
+      "epoch": 0.972,
+      "grad_norm": 0.5334486365318298,
+      "learning_rate": 1.4381783235685175e-08,
+      "loss": 0.5431,
+      "step": 2430
+    },
+    {
+      "epoch": 0.9724,
+      "grad_norm": 0.5072309970855713,
+      "learning_rate": 1.397408189974303e-08,
+      "loss": 0.5234,
+      "step": 2431
+    },
+    {
+      "epoch": 0.9728,
+      "grad_norm": 0.5345374941825867,
+      "learning_rate": 1.3572231250068233e-08,
+      "loss": 0.5247,
+      "step": 2432
+    },
+    {
+      "epoch": 0.9732,
+      "grad_norm": 0.4882015883922577,
+      "learning_rate": 1.3176231961097073e-08,
+      "loss": 0.5452,
+      "step": 2433
+    },
+    {
+      "epoch": 0.9736,
+      "grad_norm": 0.4947294592857361,
+      "learning_rate": 1.2786084697445688e-08,
+      "loss": 0.5102,
+      "step": 2434
+    },
+    {
+      "epoch": 0.974,
+      "grad_norm": 0.49401557445526123,
+      "learning_rate": 1.2401790113907351e-08,
+      "loss": 0.5259,
+      "step": 2435
+    },
+    {
+      "epoch": 0.9744,
+      "grad_norm": 0.4852409064769745,
+      "learning_rate": 1.202334885545403e-08,
+      "loss": 0.5365,
+      "step": 2436
+    },
+    {
+      "epoch": 0.9748,
+      "grad_norm": 0.49361684918403625,
+      "learning_rate": 1.1650761557233268e-08,
+      "loss": 0.5187,
+      "step": 2437
+    },
+    {
+      "epoch": 0.9752,
+      "grad_norm": 0.5144620537757874,
+      "learning_rate": 1.1284028844567417e-08,
+      "loss": 0.5799,
+      "step": 2438
+    },
+    {
+      "epoch": 0.9756,
+      "grad_norm": 0.5108654499053955,
+      "learning_rate": 1.0923151332954406e-08,
+      "loss": 0.5396,
+      "step": 2439
+    },
+    {
+      "epoch": 0.976,
+      "grad_norm": 0.49677518010139465,
+      "learning_rate": 1.0568129628063082e-08,
+      "loss": 0.5351,
+      "step": 2440
+    },
+    {
+      "epoch": 0.9764,
+      "grad_norm": 0.6048544645309448,
+      "learning_rate": 1.0218964325735935e-08,
+      "loss": 0.5381,
+      "step": 2441
+    },
+    {
+      "epoch": 0.9768,
+      "grad_norm": 0.5242604613304138,
+      "learning_rate": 9.875656011985589e-09,
+      "loss": 0.5123,
+      "step": 2442
+    },
+    {
+      "epoch": 0.9772,
+      "grad_norm": 0.538178563117981,
+      "learning_rate": 9.53820526299598e-09,
+      "loss": 0.5112,
+      "step": 2443
+    },
+    {
+      "epoch": 0.9776,
+      "grad_norm": 0.4944842755794525,
+      "learning_rate": 9.206612645118462e-09,
+      "loss": 0.5608,
+      "step": 2444
+    },
+    {
+      "epoch": 0.978,
+      "grad_norm": 0.5592899918556213,
+      "learning_rate": 8.880878714873753e-09,
+      "loss": 0.5691,
+      "step": 2445
+    },
+    {
+      "epoch": 0.9784,
+      "grad_norm": 0.5722050070762634,
+      "learning_rate": 8.561004018949602e-09,
+      "loss": 0.5671,
+      "step": 2446
+    },
+    {
+      "epoch": 0.9788,
+      "grad_norm": 0.4920216500759125,
+      "learning_rate": 8.24698909419963e-09,
+      "loss": 0.5436,
+      "step": 2447
+    },
+    {
+      "epoch": 0.9792,
+      "grad_norm": 0.5159383416175842,
+      "learning_rate": 7.93883446764332e-09,
+      "loss": 0.5927,
+      "step": 2448
+    },
+    {
+      "epoch": 0.9796,
+      "grad_norm": 0.5492022037506104,
+      "learning_rate": 7.636540656463686e-09,
+      "loss": 0.6018,
+      "step": 2449
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.5431396961212158,
+      "learning_rate": 7.3401081680092295e-09,
+      "loss": 0.5015,
+      "step": 2450
+    },
+    {
+      "epoch": 0.9804,
+      "grad_norm": 0.5254392623901367,
+      "learning_rate": 7.04953749978926e-09,
+      "loss": 0.5593,
+      "step": 2451
+    },
+    {
+      "epoch": 0.9808,
+      "grad_norm": 0.46500301361083984,
+      "learning_rate": 6.764829139476625e-09,
+      "loss": 0.4852,
+      "step": 2452
+    },
+    {
+      "epoch": 0.9812,
+      "grad_norm": 0.6032522916793823,
+      "learning_rate": 6.485983564904596e-09,
+      "loss": 0.6172,
+      "step": 2453
+    },
+    {
+      "epoch": 0.9816,
+      "grad_norm": 0.5080193877220154,
+      "learning_rate": 6.213001244066096e-09,
+      "loss": 0.5481,
+      "step": 2454
+    },
+    {
+      "epoch": 0.982,
+      "grad_norm": 0.5079247951507568,
+      "learning_rate": 5.945882635115252e-09,
+      "loss": 0.5238,
+      "step": 2455
+    },
+    {
+      "epoch": 0.9824,
+      "grad_norm": 0.537906289100647,
+      "learning_rate": 5.684628186363116e-09,
+      "loss": 0.6,
+      "step": 2456
+    },
+    {
+      "epoch": 0.9828,
+      "grad_norm": 0.5208892226219177,
+      "learning_rate": 5.429238336280396e-09,
+      "loss": 0.525,
+      "step": 2457
+    },
+    {
+      "epoch": 0.9832,
+      "grad_norm": 0.4783654808998108,
+      "learning_rate": 5.179713513493944e-09,
+      "loss": 0.5296,
+      "step": 2458
+    },
+    {
+      "epoch": 0.9836,
+      "grad_norm": 0.6100749373435974,
+      "learning_rate": 4.936054136788326e-09,
+      "loss": 0.5711,
+      "step": 2459
+    },
+    {
+      "epoch": 0.984,
+      "grad_norm": 0.5242696404457092,
+      "learning_rate": 4.6982606151027005e-09,
+      "loss": 0.5879,
+      "step": 2460
+    },
+    {
+      "epoch": 0.9844,
+      "grad_norm": 0.463133305311203,
+      "learning_rate": 4.466333347531992e-09,
+      "loss": 0.4878,
+      "step": 2461
+    },
+    {
+      "epoch": 0.9848,
+      "grad_norm": 0.5177239775657654,
+      "learning_rate": 4.240272723325722e-09,
+      "loss": 0.4898,
+      "step": 2462
+    },
+    {
+      "epoch": 0.9852,
+      "grad_norm": 0.5330920815467834,
+      "learning_rate": 4.020079121887232e-09,
+      "loss": 0.4809,
+      "step": 2463
+    },
+    {
+      "epoch": 0.9856,
+      "grad_norm": 0.4839600622653961,
+      "learning_rate": 3.805752912773297e-09,
+      "loss": 0.5072,
+      "step": 2464
+    },
+    {
+      "epoch": 0.986,
+      "grad_norm": 0.5248364210128784,
+      "learning_rate": 3.5972944556929563e-09,
+      "loss": 0.5916,
+      "step": 2465
+    },
+    {
+      "epoch": 0.9864,
+      "grad_norm": 0.5435823798179626,
+      "learning_rate": 3.3947041005071287e-09,
+      "loss": 0.5015,
+      "step": 2466
+    },
+    {
+      "epoch": 0.9868,
+      "grad_norm": 0.5040256381034851,
+      "learning_rate": 3.1979821872286097e-09,
+      "loss": 0.5426,
+      "step": 2467
+    },
+    {
+      "epoch": 0.9872,
+      "grad_norm": 0.47600534558296204,
+      "learning_rate": 3.007129046020518e-09,
+      "loss": 0.5207,
+      "step": 2468
+    },
+    {
+      "epoch": 0.9876,
+      "grad_norm": 0.5221179127693176,
+      "learning_rate": 2.8221449971970736e-09,
+      "loss": 0.4919,
+      "step": 2469
+    },
+    {
+      "epoch": 0.988,
+      "grad_norm": 0.4820321500301361,
+      "learning_rate": 2.643030351221265e-09,
+      "loss": 0.5271,
+      "step": 2470
+    },
+    {
+      "epoch": 0.9884,
+      "grad_norm": 0.5176935195922852,
+      "learning_rate": 2.469785408705627e-09,
+      "loss": 0.5299,
+      "step": 2471
+    },
+    {
+      "epoch": 0.9888,
+      "grad_norm": 0.5503344535827637,
+      "learning_rate": 2.302410460412241e-09,
+      "loss": 0.5496,
+      "step": 2472
+    },
+    {
+      "epoch": 0.9892,
+      "grad_norm": 0.5495110154151917,
+      "learning_rate": 2.1409057872504023e-09,
+      "loss": 0.5612,
+      "step": 2473
+    },
+    {
+      "epoch": 0.9896,
+      "grad_norm": 0.5193551182746887,
+      "learning_rate": 1.985271660277399e-09,
+      "loss": 0.5379,
+      "step": 2474
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.5184204578399658,
+      "learning_rate": 1.8355083406977335e-09,
+      "loss": 0.4782,
+      "step": 2475
+    },
+    {
+      "epoch": 0.9904,
+      "grad_norm": 0.5474755167961121,
+      "learning_rate": 1.6916160798635116e-09,
+      "loss": 0.5934,
+      "step": 2476
+    },
+    {
+      "epoch": 0.9908,
+      "grad_norm": 0.49980610609054565,
+      "learning_rate": 1.553595119272888e-09,
+      "loss": 0.4373,
+      "step": 2477
+    },
+    {
+      "epoch": 0.9912,
+      "grad_norm": 0.5719631314277649,
+      "learning_rate": 1.4214456905696782e-09,
+      "loss": 0.6477,
+      "step": 2478
+    },
+    {
+      "epoch": 0.9916,
+      "grad_norm": 0.49593043327331543,
+      "learning_rate": 1.2951680155433575e-09,
+      "loss": 0.4784,
+      "step": 2479
+    },
+    {
+      "epoch": 0.992,
+      "grad_norm": 0.5717119574546814,
+      "learning_rate": 1.174762306129451e-09,
+      "loss": 0.5891,
+      "step": 2480
+    },
+    {
+      "epoch": 0.9924,
+      "grad_norm": 0.4778268039226532,
+      "learning_rate": 1.0602287644075891e-09,
+      "loss": 0.5293,
+      "step": 2481
+    },
+    {
+      "epoch": 0.9928,
+      "grad_norm": 0.5617204308509827,
+      "learning_rate": 9.515675826026748e-10,
+      "loss": 0.6281,
+      "step": 2482
+    },
+    {
+      "epoch": 0.9932,
+      "grad_norm": 0.5070087909698486,
+      "learning_rate": 8.487789430829395e-10,
+      "loss": 0.5313,
+      "step": 2483
+    },
+    {
+      "epoch": 0.9936,
+      "grad_norm": 0.5938708186149597,
+      "learning_rate": 7.518630183618868e-10,
+      "loss": 0.5321,
+      "step": 2484
+    },
+    {
+      "epoch": 0.994,
+      "grad_norm": 0.5089312195777893,
+      "learning_rate": 6.608199710959606e-10,
+      "loss": 0.5176,
+      "step": 2485
+    },
+    {
+      "epoch": 0.9944,
+      "grad_norm": 0.5960400700569153,
+      "learning_rate": 5.75649954084545e-10,
+      "loss": 0.662,
+      "step": 2486
+    },
+    {
+      "epoch": 0.9948,
+      "grad_norm": 0.4653564393520355,
+      "learning_rate": 4.96353110270742e-10,
+      "loss": 0.4459,
+      "step": 2487
+    },
+    {
+      "epoch": 0.9952,
+      "grad_norm": 0.5162554979324341,
+      "learning_rate": 4.229295727405935e-10,
+      "loss": 0.4724,
+      "step": 2488
+    },
+    {
+      "epoch": 0.9956,
+      "grad_norm": 0.5672651529312134,
+      "learning_rate": 3.5537946472308233e-10,
+      "loss": 0.5596,
+      "step": 2489
+    },
+    {
+      "epoch": 0.996,
+      "grad_norm": 0.5480309128761292,
+      "learning_rate": 2.937028995885771e-10,
+      "loss": 0.6553,
+      "step": 2490
+    },
+    {
+      "epoch": 0.9964,
+      "grad_norm": 0.47406649589538574,
+      "learning_rate": 2.3789998085077536e-10,
+      "loss": 0.5226,
+      "step": 2491
+    },
+    {
+      "epoch": 0.9968,
+      "grad_norm": 0.5353025794029236,
+      "learning_rate": 1.8797080216514938e-10,
+      "loss": 0.5022,
+      "step": 2492
+    },
+    {
+      "epoch": 0.9972,
+      "grad_norm": 0.5093353986740112,
+      "learning_rate": 1.4391544732894612e-10,
+      "loss": 0.4815,
+      "step": 2493
+    },
+    {
+      "epoch": 0.9976,
+      "grad_norm": 0.4614766538143158,
+      "learning_rate": 1.0573399028196428e-10,
+      "loss": 0.4883,
+      "step": 2494
+    },
+    {
+      "epoch": 0.998,
+      "grad_norm": 0.48147180676460266,
+      "learning_rate": 7.342649510422295e-11,
+      "loss": 0.565,
+      "step": 2495
+    },
+    {
+      "epoch": 0.9984,
+      "grad_norm": 0.5356802344322205,
+      "learning_rate": 4.699301601907013e-11,
+      "loss": 0.5475,
+      "step": 2496
+    },
+    {
+      "epoch": 0.9988,
+      "grad_norm": 0.5377554297447205,
+      "learning_rate": 2.6433597389685648e-11,
+      "loss": 0.5219,
+      "step": 2497
+    },
+    {
+      "epoch": 0.9992,
+      "grad_norm": 0.4769822061061859,
+      "learning_rate": 1.1748273722189673e-11,
+      "loss": 0.506,
+      "step": 2498
+    },
+    {
+      "epoch": 0.9996,
+      "grad_norm": 0.5197920799255371,
+      "learning_rate": 2.9370696629227312e-12,
+      "loss": 0.603,
+      "step": 2499
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.5427217483520508,
+      "learning_rate": 0.0,
+      "loss": 0.5824,
+      "step": 2500
+    },
+    {
+      "epoch": 1.0,
+      "step": 2500,
+      "total_flos": 1.1471048046280704e+20,
+      "train_loss": 0.06534878476858139,
+      "train_runtime": 3332.6777,
+      "train_samples_per_second": 90.017,
+      "train_steps_per_second": 0.75
+    }
+  ],
+  "logging_steps": 1.0,
+  "max_steps": 2500,
+  "num_input_tokens_seen": 0,
+  "num_train_epochs": 1,
+  "save_steps": 200,
+  "stateful_callbacks": {
+    "TrainerControl": {
+      "args": {
+        "should_epoch_stop": false,
+        "should_evaluate": false,
+        "should_log": false,
+        "should_save": true,
+        "should_training_stop": true
+      },
+      "attributes": {}
+    }
+  },
+  "total_flos": 1.1471048046280704e+20,
+  "train_batch_size": 1,
+  "trial_name": null,
+  "trial_params": null
+}